UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trace metals and metallothioneins in rainbow trout (Oncorhynchus mykiss) exposed to various concentrations… Carrier, Raymond Charles 1992

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

Item Metadata


831-ubc_1992_spring_carrier_raymond.pdf [ 6.69MB ]
JSON: 831-1.0086696.json
JSON-LD: 831-1.0086696-ld.json
RDF/XML (Pretty): 831-1.0086696-rdf.xml
RDF/JSON: 831-1.0086696-rdf.json
Turtle: 831-1.0086696-turtle.txt
N-Triples: 831-1.0086696-rdf-ntriples.txt
Original Record: 831-1.0086696-source.json
Full Text

Full Text

to the required standardTRACE METALS AND METALLOTHIONEINSIN RAINBOW TROUT (Oncorhynchus mykiss)EXPOSED TO VARIOUS CONCENTRATIONSOF ACID ROCK DRAINAGEbyRaymond Charles CarrierB.S. The University of Montana, 1986A THESIS SUBMITTED IN PARTIAL FULFILMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIESBio-Resource EngineeringWe accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIADecember 1991© Raymond Charles Carrier, 1991In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission. Department of ei -aix(.t.t.ct.^ .evtie_......."6-The University of British ColumbiaVancouver, CanadaDate a-44-.4a—, 31, /9 19/DE-6 (2/88)ABSTRACTAn in situ experiment was conducted at the Equity Silver mine, near Houston, British Columbia in anattempt to induce hepatic metallothioneins and study metal uptake in rainbow trout (Oncorhynchus mykiss)using acid rock drainage (ARD). A mesocosm apparatus consisting of 16 flow-through troughs were placedover Foxy Creek, and fish were placed tubs at the end of selected troughs. Dilution rates of ARD:stream waterwere 0:1 (control), 1:25000, 1:10000, 1:2500, 1:1000, 1:250, and 1:25. Bioaccumulation of copper, zinc andcadmium was determined in the liver, gill and muscle tissue using atomic absorption spectroscopy. Hepaticmetallothionein concentrations were measured using differential pulse polarography. Metal concentrations inthe troughs were analyzed using inductively coupled argon emission spectroscopy. The water chemistry datawas modelled using the geochemical model MINTEQA2 in an attempt to estimate metal bioavailability.It was concluded that pH and adsorbtion processes in the experimental trou • hs played a very importantrole in controlling the ARD toxicity. Hepatic metallothioneins were not elevated above background levels after23 days of exposure to various concentrations of ARD. Significant metal bioaccumulation occurred with zinc,and the source appeared to be the food which the experimental fish were fed.The 48 hour LC50 was calculated to occur when roughly 1 part ARD was mixed with 1400 parts streamwater. At this mixing rate, the pH of the stream would have been reduced from 7 to roughly 6.8, and copper,zinc, aluminum and cadmium concentrations would have been 0.028, 0.027, 0.277 and 0.0003 mg•L -1respectively. The 23 day LC50 was calculated to occur when roughly 1 part ARD was mixed with 5150 partsstream water. At this dilution, the pH would have not been reduced below neutrality, and the copper, zinc,aluminum and cadmium concentrations would be 0.010, 0.014, 0.113 and 0.0001 mg•L -1 respectively.iiiTABLE OF CONTENTSPageABSTRACT^  iiTABLE OF CONTENTS ^  iiiLIST OF FIGURES  vLIST OF TABLES ^  viACKNOWLEDGEMENTS ^  vii1.0 INTRODUCTION  11.1 Metal Pollution ^  112 Acid Rock Drainage in British Columbia ^  11.3 Acid Rock Drainage Generation ^  21.4 Scope and Rationale of Research  41.5 Objectives ^  51.6 Hypotheses  '72.0 LITERATURE REVIEW ^  102.10 Introduction  102.20 Factors Affecting Metal Toxicity and Metal Speciation ^  102.30 Chemical Principles Affecting Metal/Ligand Binding Patterns  12231 Metal Solubility ^  142.32 Redox Potentials  15233 Ionic Radius  162.34 Reaction Kinetics in Natural Waters ^  172.35 Inorganic Complexation of Metals  202.36 Organic Complexation of Metals  202.37 Complexation Capacities of Natural Waters ^  212.38 Geochemical Modelling as a Means of Predicting Metal Speciation ^ 222.40 Metal Transport Across Biological Membranes  242.41 Metal Transport Across Gills and Body Surface ^  242.42 Metal Uptake Through the Digestive Tract  272.50 Metal Homeostasis Within Organisms ^  292.51 Metallothioneins ^  29252 Non-Metallothionein Metal Binding Proteins ^  312.53 Metals and Metal-Binding Proteins in Experimental Fish ^ 323.0 METHODS AND MATERIALS ^  363.10 Field Methods and Materials  363.11 Mesocosm Description and Operation ^  363.12 Experimental Fish ^  393.20 Laboratory Analyses  403.21 Water Chemistry  403.22 Geochemical Modelling ^  413.23 Metallothionein Analyses  42324 Tissue Analyses ^  43ivTABLE OF CONTENTSPage4.0 RESULTS AND DISCUSSION ^  454.10 Water Chemistry  454.20 Toxicity of ARD to Experimental Fish ^  494.30 Geochemical Modelling ^  494.31 Zinc ^  574.32 Copper  584.33 Aluminum  604.34 Cadmium ^  60435 Iron and Manganese ^  614.40 Tissue Metals and Metallothioneins  624.41 A Priori Hypotheses  634.42 A Posteriori Hypotheses ^  705.0 CONCLUSIONS ^  765.10 Conclusions  765.20 Recommendations ^  786.0 LITERATURE CITED  80APPENDIX 1 ^  93APPENDIX 2  96VLIST OF FIGURESFigure^ Page1. Location of Equity Silver Mine, B.C. ^  32. Hypothetical biological responses to increasing concentrations of metals ^  123. Metal and metalloid classification based on ligand binding preferences due to ionic or covalentbonding ^  134. Time scales of certain physicochemical processes, estimated by their residence time (physicalprocess) or their half-reaction (chemical reaction ^  195. Dissociation of carboxylic and phenolic groups in organic material from Smith Lake as a functionof pH ^  216. A simplified characterization of the environmental interface of an organism, illustratingmechanisms of transport of metals into biological tissues ^  257. Hepatic metallothionein (± one standard deviation) in rainbow trout as a function of heavy metalcontamination ^  348. Schematic design of mesocosm apparatus used at the Equity Silver mine ^  38viLIST OF TABLESTable^ Page1. Effective ionic radii in angstroms of various elements ^  172. Heavy metals found in solution and hepatic metallothionein concentrations found in rainbow troutcollected from the Campbell river watershed  333. Experimental design used at the Equity Silver ARD study  374. Data for samples taken from the experimental troughs in which elements were below or just abovethe I.C.P. limit of detection ^  455. Chemistry results for control troughs and raw ARD at the Equity Silver mine mesocosm study . .^476. Water chemistry results for the Equity Silver mine mesocosm study ^  487. Number of fish exposed to each treatment level and days after initiation of experiment which thefish died ^  498. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in thecontrol troughs  519. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals in the1:25000 treatment troughs ^  5110. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals inthe 1:10,000 treatment troughs  5211. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals inthe 1:2500 treatment troughs ^  5312. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals inthe 1:1,000 treatment trou. hs  5413. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in the1:250 treatment troughs ^  5514. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in the1:25 treatment troughs  5615. National bureau of standards certified oyster tissue metal concentrations and values obtained inthis study ^  6216. Summarized fish tissue chemistry and weight change data from the experimental mesocosm andtwo wild fish stocks near the Equity Silver Mine ^  6517. Regression statistics for total liver metal concentrations versus hepatic metallothioneinconcentrations for the various stocks of fish  6618. Correlation results for analysis between E moles of copper, zinc and cadmium (gmol.g-1 dryliver) versus hepatic metallothioneins ^  6719. Regression statistics for muscle metals versus metal concentrations ^  6820. Regression statistics for gill metals versus metal concentrations in solution  6921. Ranked means and significantly different groups for liver zinc concentrations ^ 7122. Ranked means and significantly different groups for gill zinc concentrations  7223. Ranked means and significantly different groups for muscle zinc concentrations  7324. Ranked means and significantly different groups for gill cadmium concentrations ^ 7525. Raw water chemistry data for the Equity Silver mine mesocosm study ^  9426. Raw data from fish tissue analyses ^  97vuACKNOWLEDGEMENTSI am indebted to many people who have assisted me with this research. Very special thanks go toChris Perrin of Limnotek Research and Development Inc. who made this whole project possible. An IRAPoperating grant for metallothionein analyses and an IRAP student grant were obtained by Chris and constitutedthe majority of this project's funding. My committee members, Dr. Royann Petrell, Dr. Ken Hall, and Dr. AlLewis provided helpful comments on the drafts of this thesis. Benoit Godin of Environment Canada unselfishlyprovided the water chemistry data which was used in this report. Susan Liptak and Paula Parkinson from theEnvironmental Engineering laboratory (Civil Engineering) helped me develop an analytical protocol which wasused by me to analyze fish tissue metal concentrations. Dr. Ken Hall provided funding (NSERC operatinggrant #588935) for gasses and acids used for the digestion and analysis of fish tissue. Dr. L. Lavkulich allowedme to use the soil science microwave and teflon bombs to digest the fish tissue without demanding money inreturn - a courtesy not normally found at U.B.C. Last, and certainly not least, I would like to thank my wifefor her moral and financial support through this thesis. Without her support and patience, this thesis wouldnot have been possible.viii"the most practical lasting benefit science can now offer is to teach man how to avoid destruction of his ownenvironment, and how, by understanding himself with true humility and pride, to find ways to avoid injuriesthat at present he inflicts on himself with such devastating energy."G.E. Hutchinsion (1903-1991)1CHAPTER 11.0 INTRODUCTION1.1 Metal PollutionMetal pollution and the resulting environmental impacts have been documented as far back as 1874in lead mining districts of the British Isles (Leland and Kuwabara 1986). Since then, continuous expansion ofmining and other metal polluting industries have greatly increased the amounts of metals in the biosphere.Some of the major sources of metal pollution include wastes from mines, electroplating factories, emissionsfrom the burning of fossil fuels and emissions from refuse incinerators. Nriagu (1979) estimated that globalanthropogenic emissions exceed natural emissions for copper, nickel and zinc by about 300, 200, and 700percent respectively. Emissions of cadmium and lead attributable to man also exceed natural rates by well overan order of magnitude. Metal pollution is of special concern because metals do not biodegrade, and in certaininstances they may biomagnify (Stumm and Morgan 1981).1.2 Acid Rock Drainage in British ColumbiaIn British Columbia heavy metal pollution derived from mining is a great concern, especially since thisprovince has many massive sulphide-containing ore deposits. Acid rock drainage (ARD) is generated whensulphur-containing ores (e.g. pyrite and pyrrhotite) are allowed to oxidize in the presence of water. Thegeneration of ARD has become a serious problem for five of the 16 metal mines currently operating in BritishColumbia, and ARD may develop in some of the other mines over time (Errington and Ferguson 1987). Fiveabandoned British Columbia mines produce ARD which severely affects approximately 30 km of stream andseveral marine environments. There are over 72,000,000 t of acid generating tailings and over 229,000,000 tof acid generating waste rock in British Columbia (Errington and Ferguson 1987). Collection and treatmentof ARD has been estimated to be a 2 billion dollar liability to the Canadian mining industry (Robertson 1987).These estimates do not account for costs incurred when cleaning-up environmental damage or costs incurredwhen collecting and treating ARD that is derived from sources other than tailings. This is disconcerting2because waste rock dumps and mine-walls can be significant sources of ARD. For instance, Morin (1990)found that at the Equity Silver Mine, mine walls generated roughly 11 mg SO42- per day per m2. Cost effectivemethods for the treatment, prevention and monitoring of ARD are currently being researched by the BritishColumbia Acid Mine Drainage Task Force.Equity Silver mine, near Houston, British Columbia (Figure 1), has faced serious problems associatedwith the containment and treatment of its ARD since 1981. The mine began production in 1980 and will closein 1992. The 800,000 m3 of ARD created annually at Equity Silver has the following approximatecharacteristics: pH 2.5, sulphate 9000 mg•I.: 1, copper 500 mg•I:1, zinc 200 mg•1:1 , cadmium 1.5 mg•1.: 1,and aluminum 1500 mg•L-1 (Remington 1989). Costs for pumping and treating the ARD were $0.85•m 3 in1986 (Patterson 1987). The flow of ARD from Equity Silver could continue for up to 150,000 years after themine has shut down (Wilkes 1987). If pollution of this magnitude were to be released untreated to theenvironment, aquatic communities in both the Bulldey and the Skeena rivers would be severely affected.1.3 Acid Rock Drainage GenerationAcid rock drainage is formed when sulfide-containing ores such as pyrite (FeS 2), chalcopyrite(CuFeS2), or pyrrhotite (Fe i...S) are oxidized in the presence of water. The following reactions denote thegeneration of ARD:(1)FeS2(s) + 7/2 02 = Fe2+ + 2 5042- + 2 H+(2) Fe2+ + 1/4 02 + H + = Fe3+ + 1/2 H2O(3) Fe3+ + 3 H2O = Fe(OH)3(5) + 3 H +(4)^FeS2(s) + 14 Fe3 + + 8 H2O = 15 Fe2+ + 25042- + 16 H+iiTAILINGS PCNOA AMC PCNCSFOXYCe,.^_,•■,GocalyLilco -,-,E SonyC:NJSCU TI-f ERNTAILPI-ANT SITE—1.4;:z/VAT N EZCNE...- I •^MA IN ZCNEWASTE DUMPSUCX CREEK100 rn disBESSEMER CS E.EX 0700061SU= CREEX C.CHTIIOL 0400765efttmlw%3RITISHCZ:L.:AMA31^I^2'^ 3 innr••••••••■••1■MMICEIN=.1 : 100,000Figure 1. Location of Equity Silver Mine, B.C.4At pH 3, the rate limiting step (equation 2) has a half-life of roughly 1000 days (Garrels andThompson 1960). However, the presence of Thiobacillus ferrooxidans can speed the reaction by as much a 1x 106 times (Singer and Stumm 1970). Other bacteria which aid in the formation of ARD include T.thiooxidans, Leptospirillum ferrooxidans and Sulfolo brieleyii (Silver 1989). Browning (1970) believes that upto 80 percent of the ARD generated in the U.S. is due to the activity of these bacteria.Once Fe3+ is present, it can directly oxidize pyrite which releases more acidity. The precipitatedFe(OH)3 typically lines stream beds and is trapped in interstices in rock dumps where it provides a reserve ofFe3+ which can further oxidize pyrite. The oxidation of pyrite to sulphuric acid and ferrous sulfate isexothermic and releases 1440 1C.mol-1 FeS2 (Bennett et a1.1988). Because of these exothermic reactions andthe direct oxidation of pyrite by dissolved ferric iron (see equation 4), the ARD generation process tends tobe self-perpetuating (E.P.S. 1987). Once the low pH is established, metals which were initially immobile comeinto solution and can affect aquatic communities.1.4 Scope and Rationale of ResearchThis research was part of an integrated study conducted by Limnotek Research and Development Ltd.which was designed to document the impact that ARD had on an artificial aquatic ecosystem. Periphytic algaeand benthic invertebrate communities were sampled in experimental troughs lined with gravel which wereexposed to a gradient of ARD. By placing rainbow trout (Oncorhynchus mykiss) in the outfall of theexperimental troughs, an extra trophic level was added to the mesocosm study. Copper, zinc and cadmiumconcentrations in the gill, liver and muscle were analyzed at the end of the study. This was done to determineif the body metal burdens were related to the ARD exposure concentrations.Another parameter which measured at the beginning and end of this study was hepatic metallothioneinconcentrations. Hepatic metallothioneins have been recommended as potential biochemical indicators ofchronic metal stress in fish exposed to ARD in Buttle Lake, British Columbia (Roch and McCarter 1984a).These authors found that hepatic metallothioneins were inducible within a few weeks, and were related to the5level of metal pollution. This experimental work conducted in a different location using ARD derived froma different mine was designed to be a test of the universality of hepatic metallothioneins as biochemicalindicators of metal pollution.Wild fish from Lu Lake and Goosly Lake (Figure 1) were also sampled and analyzed in order to detectpossible differences between fish stocks. Lu Lake is upstream from all potential mining wastes, and GooslyLake has received ARD for at least one season before treatment facilities were constructed at the mine(Patterson 1986). Benoit Godin of Environment Canada was interested in the test results from two stockswhich were found in the same watershed.Brian Wilkes from the Ministry of the Environment (Skeena region) strongly recommended this studyin order to provide insight into mechanisms which affect metal uptake by fish. Many mines exist in the Skeenaregion, some of which have been abandoned; causing severe metal pollution in some instances. Results fromthis thesis may be used to evaluate the usefulness of body metal burdens and hepatic metallothioneins aspotential indicators of metal pollution in fish stocks exposed to ARD.Finally, this study was conducted in the field in order to provide a realistic environment. Organicmatter and various other important complexing and adsorbing agents normally found in natural waters areoften absent in laboratory studies. The availability of ARD and stream water were cruicial in the rationale forsetting up this in situ experiment.1.5 Objectives One of the main objectives of this study was to evaluate the use of hepatic metallothioneins as abiochemical indicator of chronic metal stress. Studies conducted by Roch and McCarter and others in the early1980's have shown that metallothioneins are inducible by placing fish in metal-polluted water (e.g. Roch andMcCarter 1986; McCarter et al. 1982; McCarter and Roch 1984; Roch and McCarter 1984a; Roch andMcCarter 1984b; McCarter and Roch 1983). Deniseger et al. (1990) analyzed community shifts in variousplankton species, metal concentrations in water, metallothionein levels in fish and other parameters during the6metal pollution history of Buttle Lake, British Columbia. They found that fish hepatic metallothioneinconcentrations were positively correlated with the level of metal pollution in the lake.This study was designed to evaluate the concentrations of copper, zinc and cadmium found in muscle,gill and liver tissue from fish which were exposed to a gradient of ARD at the Equity Silver mine. Hepaticmetallothionein concentrations were also measured in order to see if the ARD exposure would cause significantincreases in the concentrations of this cytosolic protein. By monitoring these parameters, their usefulness asindicators of short-term chronic ARD metal exposure in rainbow trout could be evaluated.This research was different from that conducted by Roch and McCarter in that the ARDcharacteristics at the Equity Silver mine were different from the ARD found in Buttle lake. The Equity ARDhad a Zn:Cu:Cd ratio of 60:80:1 whereas the Buttle Lake ARD had a corresponding ratio of 400:20:1. Thechemical characteristics of the water found in the two watershed also differed with Buttle Lake being classifiedas ultra-oligotrophic (Stockner and Shortreed 1985), and the Lu Lake watershed being rather dystrophic innature (pers. obs.). Therefore, the water chemistry and hence, the bioavailability of the metals would havediffered significantly from the studies conducted by Roch et al. This study may provide insight into theuniversality of hepatic metallothioneins as biochemical indicators of chronic metal pollution caused by ARD.The second main objective of this research was to use a geochemical equilibrium model to predict themetal species and hence the bioavailability of the various metals found in the experimental troughs. By utilizingthis type of predictive tool, relationships between estimated metal speciation and metal bioaccumulation willbe discussed. The potential for uptake via the gill was the only potential uptake path that was experimentallymanipulated in this study. Uptake via the digestive tract was normalized by feeding the same amount of foodto the various groups of experimental fish.The fmal objective of this work was to determine in situ LC50 information using rainbow trout.Significant quantities of resident rainbow trout are found in Foxy Creek, and the results from this study wouldprovide data that would be useful to protect those wild stocks.71.6 HypothesesThe first hypothesis was formulated to see if hepatic metallothionein concentrations in the experimentalfish were related to the ARD concentrations which the fish were exposed to. Stated explicitly;Ha :^Fish hepatic metallothionein concentrations in the surviving fish were correlated with the ARDconcentrations which the fish were exposed to.Ho:^No relationship between surviving fish hepatic metallothionein concentrations and ARD concentrationsexisted.Experiments where fish have received intraperitoneal metal injections typically cause metallothioneins to befully induced within a few days (Price-Haughey et al. 1986; Pierson 1985). Fish exposed to metal pollution viaaqueous solutions have shown marked differences in peak hepatic metallothionein induction time. Kito et al.(1982) exposed carp (Cyprinus carpio) to 5 mg•1:1 cadmium and studied the uptake and partitioning of metalsbetween the high-molecular weight protein pool and metallothioneins. After only one day of exposure, anotable metallothionein peak occurred. After 31 days of exposure, the metallothionein concentrations in thehepato-pancreas continued to rise. Roch and McCarter (1984a) conducted studies in which rainbow trout wereexposed to various concentrations of ARD from Buttle lake. They found that a three to four week time periodwas more than sufficient to induce elevated concentrations of metallothioneins.The second hypothesis was formulated to test if zinc, cadmium and/or copper concentrations in thelivers of the experimental fish were related to hepatic metallothionein concentrations. These metals are knownto be strongly sequestered by metallothioneins (Brady 1982). Stated explicitly:Ha :^Hepatic metallothionein concentrations were correlated with zinc, copper and cadmium concentrationsin the liver.Ho:^Hepatic metallothionein concentrations were not correlated with zinc, copper or cadmiumconcentrations in the liver.Metals which bind to high molecular weight proteins may cause the proteins to become dysfunctional, thusinterfering with enzymatic processes. Metallothioneins and other low molecular weight proteins are thoughtto sequester excessive heavy metals, resulting in the protection of the high-molecular protein pool (McCarter8et al. 1982; Brown 1977). The strength of correlations between metal levels and metallothionein concentrationsmay be used to estimate the degree of metal-binding to sensitive high molecular weight proteins in the fishlivers.The third hypothesis was devised to test if zinc, copper or cadmium concentrations in fish muscle wererelated to increasing concentrations of ARD. The general hypothesis for each metal can be stated as:Ha :^The fish muscle metal concentrations were correlated with concentrations of metals in the testsolution.There was no relationship between muscle metal concentration and metal concentration in the testsolutions.The reason for testing this hypothesis was to determine if metal levels would become elevated in the portionof fish which is consumed by humans. Results from these analyses may have significance to humans whoconsume fish from water courses which are polluted with ARD.The fmal hypothesis concerns gill tissue. Gills constitute roughly 65 percent of the surface area ofrainbow trout (Reid 1989), and the diffusion distance between water and plasma is from two to four microns(Bond 1979). Because of these characteristics, the gill is known to be one of the major uptake sites for metalsfound in solution (Pagenkopf 1983; Malian 1985). In this study, no distinction could be made between metalsthat were bound to, or were actually incorporated into the gill tissue. Selective metal binding is known to takeplace by mucous which covers the gill (Part and Lock 1983; Handy and Eddy 1990). Stability constants betweenmetals and organic matter follow the order of copper > > cadmium > zinc (Stokes 1979; Florence 1982; Irvingand Williams 1953; Nriagu 1980). Therefore, it can be hypothesized that the amount of metal found on/in thegill relative to that found in solution should follow this same trend. The fmal hypothesis can be stated for eachmetal in the following form:Ha :^The concentrations of metal found on/in gill tissue will be correlated with the concentration of metalin solution.Ho:^Metal concentrations on/in the gill will have no relationship with the metal concentration in solution.9The strengths of the relationships can be compared by the correlation coefficients and the standard error forthe respective regressions. Several authors (Part et al. 1985, Thomas et al. 1983) have suggested that gills maybe a suitable organ for use as a biological monitoring tool. Gills are easily removed and they are in intimatecontact with receiving waters. Results from the tests conducted in this study may provide further insight intothe use of gill tissues as biological indicators of metal pollution attributable to ARD.10CHAPTER 22.0 LITERATURE REVIEW2.10 IntroductionIn this section metal speciation and how it affects metal bioavailability, metal transport acrossmembranes and the induction of metallothioneins and metallothionein-like proteins once the metals haveentered the organism will be discussed. Due to the plethora of literature on metal pollution and biologicaluptake, emphasis in this review will be placed on freshwater environments and metal uptake by teleosts.2.20 Factors Affecting Metal Toxicity and Metal SpeciationSeveral studies have demonstrated that total metal concentrations in solutions are typically a poorindication of toxicity (Florence 1982; Campbell and Stokes 1985; Hunt 1987). The speciation and concentrationof metals play a crucial role in determining their toxicity. According to Luoma (1983), the toxicity of metalsto aquatic organisms is controlled primarily by:o metal concentrations in solution;o solute speciation of the metal;o metal concentrations in food;o metal partitioning among ligands in food;o the influence of other cations;o temperature (for metals which do not exchange readily); ando pH and redox potential in some situations.Metals in solution can be found in ionic form, in an oxidized or reduced state, complexed by organic substancessuch as chelating agents, adsorbed on inorganic or organic particulate matter, and may be acting singly or incombination with other particulate materials and metals (Waldichuck 1974). O'Donnel (1985) feels that theanalytical determination of metal bioavailability is complicated by experimental errors coupled with biologicaland chemical variations inherent in any study. Uptake of metals by organisms is the only way of accuratelydetermining metal bioavailability, and toxicity tests have been conducted which determine uptake (e.g. Andrewet al. 1977). Analytical techniques have been developed which provide insight into metal bioavailability, but11all are operationally defined. Despite the current analytical difficulties, it is accepted that metal speciation andhow it relates to toxicity are crucial concepts in understanding the chemistry of metal pollution.Certain metals are essential components of proteins and enzymes in biological systems. Metals canbe classified into four basic categories: macro-elements, micro-elements, elements required in trace quantitiesand elements with no known biological function. Macro-elements which are required by animals includenitrogen, sulphur, phosphorus, sodium, calcium, magnesium and iron. Micro-elements include selenium,arsenic, fluorine, chlorine, iodine, manganese, copper and zinc. Elements which are required at very lowconcentrations include silicon, boron, vanadium, chromium, cobalt, nickel, molybdenum and tin (Eichenberger1986). Elements which have no known biological function include cadmium, lead and mercury. The macro-elements are generally of low toxicity, while the micro-elements, trace elements and non-essential elements canbe much more toxic.At very low concentrations, certain elements may be inadequate, and deficiencies may result. However,these same elements may become very toxic at concentrations which are only a couple orders of magnitudegreater than the inadequate concentrations (Eichenberger 1986). Symptoms of excessive or inadequate metalintake include both chronic and acute responses. At high exposure levels, the mechanisms of metal toxicityhave been hypothesized by Eichorn (1974) as:o displacement of an essential metal from an active site by a toxic metal;o binding to undesired parts of macromolecules;o crosslinking, which can produce undesired aggregates;o depolymerization of biological molecules;o misrepair of nucleotide bases and induction of errors in protein synthesis.Chronic responses can be manifested in weeks to generations. Acute responses usually take place quickly, anda lethal end-point is a common measurement in toxicity tests. Two hypothetical responses to these variousmetal concentrations have been put forward by Frausto et al. (1976). For required elements, a parabolic curveimplies that an optimum concentration exists between very low concentrations and very high concentrations(Figure 2). The second response curve represents elements which have no known biological function. Any4—Non-Essential ElementsDeleteriousBeneficialRequired Elements•Concentration12increase in concentration typically causes detrimental effects in the organism. In the following sections,chemical characteristics and some of the toxic manifestations of copper, zinc and cadmium will be discussed.Figure 2. Hypothetical biological responses to increasing concentrations of metals (from Frausto et al. 1976).2.30 Chemical Principles Affecting Metal/Ligand Binding PatternsIn a paper designed to reclassify the nondescript term "heavy metals", Nieboer and Richardson (1980)classified several metals and metalloids by their tendency to bind either ionically or covalently. Ahriand et al.(1958) initially proposed this type of classification. The three classes of elements include class A, class B andborderline elements. Class A elements have an affinity for oxygen > nitrogen > sulphur, and class B elementshave affinities for sulphur > nitrogen > oxygen. The borderline elements were grouped in this fashion becausethey did not always dearly follow the binding order sulphur > nitrogen > oxygen. By examining Figure 3, itcan be noted that class A metals and metalloids form a distinct group and include most of the macro elements,whereas class B and borderline elements have no distinct separation and contain mostly micro elements.0.500^2^4^6^8^10^12CLASS A1^1^1/04"L"'"ii"L'14^16^20^239.01Au*134.5 —• SO+• Be4.• Ag*^Pcr• ••rta+• •TI+^Hil•• Pb"• Ctsa•Cd 44•diCol+W.V.OW*TII+WAIOZna•Ma" ‘,"1.0 Cs•^Be+^• Mg"ilNa+ Sr+•••Ca4+•Gd" Lusf•••• ys+Las+• PbIBD•OF6 4 +^SnI137)Gal+•BORDERLINE•AP+:E 4.0X3.5• 3.0w4:(• 2.50coO 2.0co(9, 1.5• Cu*CLASS 8• Sbffill• AsIMICLASS A OR IONIC INDEX. 22/rFigure 3. Metal and metalloid classification based on ligand binding preferences due to ionic or covalentbonding. From Nieboer and Richardson (1980).In a two part series, Pearson (1968a, 1968b) proposed a classification system for metals which alsoexpanded the principles explained by Ahrland et aL (1958). By using this classification system, many chemicalphenomena could be explained. Cations and anions were classified as hard or soft acids and bases (HSAB14principle). Hard bases hold onto their electrons tightly, whereas soft bases only weakly hold onto their valenceelectrons. Some of the general rules laid out in Pearson's paper include:o hard acids bind tightly to hard bases, and conversely, soft acids bind to soft bases;o cations with high charges and small ionic radii are typically hard acids;o anions which have an increased charge and a decreased radius are strong bases;o elements which have various valence states usually display an increase in hardness as theoxidation state increases;o hard acids bind bases primarily by ionic forces;o soft acids bind bases primarily by covalent bonds, and the two bonded atoms are normally ofsimilar size and electronegativity-,o hard solutes dissolve in hard solvents (e.g. water);o softness of an acceptor increases on going down a column in the periodic table;o hardness of an acceptor increases on going across the periodic table;o reaction kinetics between hard acids and hard bases and soft acids with soft bases are veryquick.Martin (1986) warns that the HSAB classification scheme is an oversimplification of a complicatedsubject, and the rules should be used only as very general guidelines.As can be seen by comparing the two classification methods mentioned, grouping metals by theirtendency to form ionic or covalent bonds is an imprecise science. These classification schemes should be usedto conduct preliminary assessments of aqueous chemical solutions. Much more thorough and analytically basedtechniques must be used before conclusions can be drawn about the toxicity and speciation of metals insolution.2.31 Metal SolubilityAnother factor that controls metal bioavailability is metal solubility. Pagenkopf (1978) states that theconcentrations of many trace metals in natural waters are regulated primarily by the solubility of theirhydroxide and carbonate salts. Other factors which influence the solubilities of metal compounds in the aquaticenvironment include ionic strength, types of ligands present, pE, pH, degree of saturation with respect to agiven substance, pressure and temperature (Stumm and Morgan 1981).The HSAB principle (Pearson 1968a, 1968b) can also be successfully applied to solubility problems.Water acts as both a hard acid and a hard base so consequently, metals classified as hard acids and hard bases15are easily dissolved in water (e.g. Ca2+ , mei) . When soft bases are present in solution (e.g. cr, S), thesewill automatically tend to bind to soft acids (e.g. Cd2+ , Ni2+, Za2+ ). By examining Figure 3, it can be notedthat class A elements are a distinct group which is usually soluble in water. Borderline and class B elementsolubility depends largely on the concentrations of soft bases found in a given solution.2.32 Redox Potentials and pHBoth pH and redox potentials have a great influence on metal speciation, geochemical cycling andbioavailability of metals in aqueous environments. The pH of natural waters is one of the major governingfactors that controls the adsorption and desorption of metals with both organic and inorganic substances. Adecrease in pH as small as 0 5 units can often mean the difference between complete adsorption and completedesorption of metals in aqueous environments near neutrality (Leckie and James 1975). At either end of thepH scale, the hydronium ion or the hydroxide ion can effectively compete with various acids or bases foravailable ligands (Pagenkopf 1978). This explains why cationic metals at a low pH tend to remain in solution,whereas at a high pH, cationic elements are typically precipitated out as carbonates, hydroxides, or oxides.The buffering capacity of natural waters is governed primarily by carbonates, but other types ofcompounds present in solution may also contribute to a lake or stream's buffering capacity. Bufferingcapacities are typically much higher in productive systems which have a higher ionic strength and greateramounts of magnesium and calcium. Oligotrophic systems are often poorly buffered due to low concentrationsof carbonates, and are especially susceptible to the addition of acidic substances (Luoma 1983).Redox potentials in aquatic systems have significant effects on both metal solubility and toxicity.Hydrous ferric and manganese oxides play an important role in trace metal geocycles in aquatic systems(Sholkovitz 1990). Oxidizing environments are characterized by a positive pE which indicates the presence ofoxygen. Under these conditions, Fe3+ and Mn4+ are prevalent and are very insoluble (Pankow and Morgan1981). Amorphous colloids which have a large negatively charged surface area are formed by these two ionswhich provide ideal binding sites for cations in solution. As these colloids grow in size, they tend to settle out16of the water column, and once incorporated into the sediments, they may enter a reducing environment whereiron and manganese oxides are reduced to Fe2+ and Mn2+ . These ions are quite soluble and do not contributeto complexation and/or adsorption (Davison and Woof 1984).In ARD derived from pyrite, the formation of iron and manganese oxides plays a very important rolein metal binding and sequestering. Nordstrom et al. (1979) found that a stream contaminated with untreatedARD had large quantities of iron hydroxide which bound and co-precipitated with other transition elements.If significant discharges of mining wastes enter a productive body of water, an anoxic hypolimnion couldpotentially circulate large quantities of metals into the epilimnion during lake turn-over events. The formationof insoluble metal sulfide complexes in anoxic sediments which are not disturbed can act as a significant sinkfor heavy metals such as zinc, cadmium and copper (Campbell et al. 1988).2.33 Ionic RadiusThe ionic radius and coordination number of an ion (which are also affected by pH and redox)influence the toxicity of a metal. As stated earlier, certain non-essential metals are capable of replacingessential metals on biological macromolecules which causes a disruption of the molecule's function. Byexamining the effective ionic radius for various elements (Table 1), several similarities can be noted (Shannon1976). Cobalt, zinc and copper have remarkably similar ionic radii at all coordination states. Cobalt 2+ isknown to replace Zn2+ in many enzymes without causing damage and is therefore often used to trace zincmetabolism (Martin 1986). The similarity of the ionic radius of these ions would imply that they are effectivelycompeting for binding sites on biological membranes. Another commonly recognized replacement takes placebetween calcium and cadmium (Part and Lock 1983). Cadmium has ionic radii similar to those of calcium atvarious coordination states. The tragic story of cadmium poisoning in Japan is testimony to this phenomena.The Itai-Itai disease ('ouch-ouch' disease) was caused by the release of untreated effluent from a zinc minein the Jintsu River (Mance 1987). Cadmium replaced calcium in human bones and caused them to beweakened to the point where simply touching a victim would cause severe pain.17Table 1. Effective ionic radii in angstroms of various elements (from Shannon 1976). The underlined valuesrepresent the most common coordinate values encountered. Columns numbered 4 through 9 represent thecoordination number of each element.I^Ion 4 5 6 7 8 I^9Ap+ 039 0.48 0.54Fe3+ 0.49 0.58 0,65! 0.78Ni2+ 0.55 0.63 0.69Mg2+ 0.57 0.66 0.72 0.89Cu2+ 0.57 0.65 0.73Co2+ 0.58 0.67 0.74 0.90Zn2+ 0.60 0.68 0.74 0.90Li+ 0.59 0.76 0.92Fe2+ 0.63 0.78b 0.92mn2+ 0.66 0.75 0.83 0.90 0.92Cd2+ 0.78 0.87 0,95 1.03 1.10Ca2+ 1.00 1.06 1.12 1.18Na + 0.99 1.00 102 1.12 1.18 1.24Pb2+ 0.98 1.19 1.23 1.29 135K+ 1.37 1.38 1.46 1.51 1.55Cs + 1.67 1.74 1.78a High Spin, low-spin values is 0.55. b High Spin, low-spin value is Reaction Kinetics in Natural WatersReaction kinetics in natural aquatic systems is a complicated subject that is currently being studied bymany researchers (Stumm 1990). All natural waters are in a state of constant flux; some reactions take placevery quickly while others may take decades or even centuries to complete. Time scales for reactions that takeplace in, or potentially affect aquatic ecosystems range from picoseconds to eons (Figure 4 from Buffle 1988).Because of this variable time scale and the number of chemical reactions that exist, a brief discussion will bedevoted to reaction kinetics that could take place during a short (c.a. < 5 minutes) hydraulic retention timein a mesocosm apparatus.18Stumm and Morgan (1981) state that hydrolysis equilibria usually take place very quickly, as long asthe hydrolysis species are simple. Unidentate reaction kinetics and the simple hydrolysis of ions typically takeonly a few seconds to complete. Polynuclear complexes are often formed rather slowly, and phase changes (i.e.solid/liquid, liquid/gas) take much longer.Factors that influence reaction kinetics include temperature, ionic strength, viscosity and type ofreaction processes (Brezonik 1972; Stumm and Morgan 1981). In natural freshwaters, ionic strength plays alarge role in determining rate constants. Since the concentrations of ions in solution are relatively low, theprobability of an ion coming in contact with another reactive ion is also low (principle of mass action: Gardiner1969). Therefore, in solutions having high ionic strength, the activity coefficients of anions and cationsdecrease (Stone and Morgan 1990) due to the rapid reactions that take place between cations and anions.Hering and Morel (1990) state that metal uptake is dependent on the biological rate of uptake versusthe rate at which chemicals react and become biologically inert or available. For instance, if the desorptionrates from insoluble oxides are much faster than biological uptake rates, a potentially large pool of metal wouldbe biologically available. This pool would not be detectable using analytical techniques or modelled estimates.The converse of this situation applies for adsorption. In a series of studies designed to investigate theadsorption/desorption reaction kinetics with copper and humate, Hering and Morel (1990) found that theconcentrations of the adsorbent and the adsorbate strongly influenced the rates of reactions. For instance, a10 fold increase in the concentration of humic acids resulted in roughly a 100 fold increase in the binding rateswith copper. The second order overall reaction rate constant for copper (2.5 itFL -1) with humic acid (1.1mg•I.:1) was between 9 * 105 to 6 * 104 M-1.see1. These reaction rates are very specific, and should not beextrapolated beyond the experimental conditions.msec^1 sec^1 min^43117)1(0^H7C.03^C4O2R/ 033 CO2^H2CO31 hr^I dayLake efilimnionHorizon,. Verticalmixinglyearseasonal Lakesaid ing Typic. resid. Hite1 rnd lertiumOceanResidence1 psecHS(P414 141125aFt 7) ICH3 COO" CH3COOHat3coott CHICOO19-10^6^-2^0 4 8^10^logft (sec)!Soil water Diffusion at^Glacier. poke ke crap1,74r es . Orme, (3) water-sedrnent Residence timeinterfacetLakes Oceans^TerrestrI------Iialdeep...surface layers water exch. chemical weatheringI^Lakes i Ocean iSettling particles resid. timeMarine Terrestrial 0-1 Ibrologkal cycles^IMicroorganismscroft^ 1•1(H24 Fe(H2Ot^ met al rplaite 1I^CuM2C^ Ni(1204 Fl(H2134.Ft (II) • 02 --e. Fir (OH/3ion solid partici: of solid CACO3Adsorption^Formmn(aut)lo.n^Io ..1.4n02rH C^r^7^I^►^ 1*Mt^993. • SOt-w.^Ni. mat. Nil . oxalate^p pH pH 914,4 r H. F-----4-... IA- oral. Sill) add ation by 02 iMr. CI- •-s• Since^ Nal. glyclnk-wHityc.14nCIA-Mr?"^ ri-glyi-efe.ZnfiTA CaEOTAPbHNTA -..1 h' Ck.^caEllTA04-.14 EDTA(14•Cd,Zn P14rb• HNTACuFA —r Cal. ^i fY4°5 -Z3-PO4acid / base ^-m-- physical mixing processes ---1.-Ftid r a I i on / dehydration^ -.4-- heterogeneous reactions --..-ion-pair formation / dissociation1^Chelate formation /dissociation ishomogenious react ions^1.-Figure 4. Time scales of certain physicochemical processes, estimated by their residence time (physicalprocess) or their half-reaction (chemical reaction) (From Buffle 1988).AIR dissociation^4Mare^Alcorn*decarbaxytation roc' misation202.35 Inorganic Complexation of MetalsInorganic complexing agents in natural waters play a major role in determining the toxicity andbioavailability of metals. Freshwater systems are buffered primarily by carbonate and/or silicate bufferingsystems (Lerman and Childs 1973). Elements which are capable of forming inorganic complexes with cationsinclude carbon, nitrogen, oxygen, fluoride, phosphorus, sulphur, chlorine, arsenic, selenium, bromine, telluriumand iodine (Stumm and Morgan 1981). Of the inorganic compounds, carbonates, bicarbonates and sulphurcompounds typically comprise the greatest abundance of anions in natural unpolluted waters (Faust and Aly1981). Sulfate and other sulphur species are typically leached from the bedrock in watersheds. Dolomite(CaMg(CO3)2 and calcite (CaCO3) are two of the most reactive minerals found in the earth's crust, and theyare also readily leached into freshwaters (Wollast 1990). These two compounds form the majority ofcarbonates and bicarbonates which form the major buffering system in freshwaters (Wetzel 1983).2.36 Organic Complexation of MetalsMany organic ligands exist which can alter the speciation and bioavailablity of metals in naturalfreshwater. Buffle (1988) estimates that over 80 percent of the natural organic matter found in natural watershave yet to be identified. Some of the well-known organic components in natural freshwater include fulvic,tannic and humic acids (Guy and Chakrabarti 1976; Raspor et al. 1984). They originate from both pedogenic(soil-formed) and aquagenic (formed in aqueous environments) processes and typically constitute approximately70 to 80 percent of the organic matter found in natural waters (Buffle et al. 1984).Allochthonous input from the watershed normally makes up the largest portion of organic matter,although substances such as biological exudates are also known to contribute to the organic carbon balance inaqueous environments (Fish and Morel 1983). These organic substances include proteins, carbohydrates, ligninsand humics. Humic substances normally form the greatest portion of organic matter found in aquatic systemsbecause they are very resistant to microbial attack (Perdue 1979). Humic substances can be classified intothree groups: 1) fulvic acids which are soluble in both acidic and basic solutions, 2) humic acid which is21insoluble in acidic solutions and 3) humin which is insoluble in both acidic and basic solutions (Reuter andPerdue 1977). McKnight et al. (1983) state that aquatic fulvic acids account for 30 to 80 percent of thedissolved organic matter in natural waters. The lower values would be typical of groundwater, whereas the highvalues would be typical for organically coloured waters.Fulvic acids typically have two major complexing sites - the carboxylic acid groups and the phenolicgroups (Wilson and Kinney 1977). The carboxylic functional group remains intact from pH 4.0 to 8.5 and thephenolic groups remain intact from pH 8.5 to 12.0 (Figure 5). Because the carboxylic functional group remainsintact at pH values normally found in natural waters, they play the biggest role in metal sequestering undernatural conditions.03.0 4.0 5D 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0pHFigure 5. Dissociation of carboxylic and phenolic groups in organic material from Smith Lake as a functionof pH (taken from Wilson and Kinney 1977). = = portion of functional group that is dissociated.2,37 Complexation Capacities of Natural WatersThe complexation capacity of natural waters is a useful measurement in that it provides an estimateof when naturally occurring ligands become saturated with cations in solution. Once the ligands are fully22complexed, any excessive cations will tend to remain in ionic or hydrated forms which are normally the mosttoxic. Several different techniques have been developed to determine complexation capacities, and each providedifferent estimates (Neubecker and Allen 1983). Methods commonly used include synthetic ion exchangeresins, anodic stripping voltimetry, solubilization techniques, fluorescence, dialysis and inorganic ion-exchange(Jardim and Allen 1984). The complexation capacity of many metals can be determined, but in order toprovide an estimate of the total available number of ligands, copper is usually used (Neubecker and Allen1983). These techniques are all operationally defined, and can be used as estimates of the complexationcapacity of natural waters.The relationship between a water's complexation capacity and the toxicity of a given metal to a certainorganism can be obtained by conducting a 'biological titration'. Under this scenario, test organisms would besubjected to an increasing concentration of test metal, and a response is measured. Theoretically, a break inslope would occur on a plot of the response versus the concentration of added metal. The point of inflectionwould be the complexation capacity of the test solution specific to the test organism. Many authors havesuccessfully used this technique (Borgmann and Ralph 1983; Imber et al. 1984; Florence et al. 1984). Themajor disadvantage of these tests is that they are typically very time-consuming and unmeasurable factors whichare related to the test organisms' health can bias the test results.2.38 Geochemical Modelling as a Means of Predicting Metal SpeciationThe use of geochemical models to predict the fate and bioavailability of metals has been researchedby many people during the last 20 years. Since the use of the first model by Garrels and Thompson (1962),a great degree of sophistication has been incorporated into these models. Nordstrom and Ball (1984) reviewedthe application of 57 models that have been used to model chemical speciation in both fresh and salt waters.Algorithms which can be used to calculate adsorption, mass transfer, redox balances, high geothermaltemperatures and trace element speciation were discussed.23Some of the advantages of using geochemical models to predict metal speciation and bioavailabilityinclude:o the ability to predict potential effects of anthropogenic inputs on natural systems,o the ability to conduct sensitivity analyses to estimate the effect that changing variable constants haveon model results, ando the results are not susceptible to analytical and sample handling errors.A few of the disadvantages of relying on model output to determine speciation and bioavailability havebeen pointed out by Jenne (1979). Some of these problems are that:o researchers cannot fully characterize organic substances in natural systems,o difficulties in accurately determining redox potentials in natural systems often occur,o thermodynamic data is often lacking,o kinetic data is sparse and often lacking for many processes pertinent to natural systems,o little is known about the importance of error estimates for the preceding parameters, ando few model verification studies have been conducted on natural systems.These drawbacks are often outweighed by the ease of model use, the low cost involved with modelling, and thepredictive capabilities that are gained by altering one or more of the chemical parameters in the model. Theuse of models to predict the fate of pollutants will undoubtedly increase over time, especially since a great dealof research is being conducted to eliminate many of the unknowns that currently exist.Chemical equilibrium models attempt to calculate the minimum Gibbs free energy within a definedsystem. The equilibrium constant approach to solving these equations use non-linear equations and massbalance limitations to calculate the minimum Gibbs free energy in a solution (Nordstrom 1984). The modelMINTEQA2 (Allison et al. 1990) was used in this thesis to predict the possible speciation of metals that werepresent in the mesocosm study. This model requires parameters such as total metal and ligand concentrations,redo; pH, temperature, and pressure which is then used to compute mass balance, adsorption, and metalspeciation in solution.242.40 Metal Transport Across Biological MembranesIn order for metals to enter an organism, the compounds or ions must pass through a membranesystem. Epithelial surfaces consist of non-polar phospholipids and are dominated by -COO"- groups which arenegatively charged at neutral pH (Lipman et al. 1966). Although the mechanisms for heavy metal transportacross epithelial membranes is still far from completely understood (Part 1986), Luoma (1983) has proposedthe following general transport mechanisms:o diffusion in non-polar complexes through the lipoid regions of the membrane;o penetration in ionic form by binding to carrier proteins in the membrane (facilitateddiffusion);o endocytosis of particulate fractions - a vacuole is formed from the cell membrane around theparticle and the whole package is then engulfed and digested intracellularly;o metals complexed to nutrients such as amino acids and carboxylic acids are transported intothe organism by specific carriers.A pictorial description of these mechanisms is presented in Figure 6. The rate at which these processes takeplace are determined by factors such as the concentration gradient across the membrane, temperature and thetypes of metals and ligands present at the organism-environment interface.The uptake, biological partitioning and excretion of metals is a very complicated subject. Uptakeversus excretion rates play a large role in understanding the meaning of metal levels found in tissues.Unfortunately, very few of these rates have been determined, and much is left to be studied in the field ofmetal homeostasis. The following sections will discuss two of the recognized uptake pathways for metals infish.2.41 Metal Transport Across Gills and Body SurfaceThe gill surface in rainbow trout accounts for roughly 60 percent of its surface area, and this area canbe exposed to over 48 liters of water per hour (Reid 1989). Several authors have found that the gills are themajor site of metal uptake by freshwater teleosts (Part and Svanberg 1981, Roch et al. 1985, Williams andGiesy 1978), whereas uptake through the skin is virtually non-existent (Bentley 1962). The gill's large surfacearea and short diffusion distance between water and blood (Hughes 1972) may cause the gills to serve as the25main uptake site for dissolved compounds including heavy metals. Metals which are known to be taken up orexchanged primarily at the gill include calcium, cadmium, sodium and potassium (Williams  and Giesey 1978,McWilliams 1982, McWilliams and Potts 1978).ENVIRONMENT^i MEMBRANE TISSUES(Phospholipid-apoiar)Diffusion of anApolar FormM0 ^ M3 1^m°Cornplexation of PolarForm to Carrier M+-00C–R MOOC–R — R-000-EndocytosisI ntracellularDigestionFigure 6. A simplified characterization of the environmental interface of an organism, illustrating mechanismsof transport of metals into biological tissues (from Luoma 1983).Gill membranes consist of phospholipids which provide a surface of net negative charge at near-neutralpH (Pagenkopf 1983). At lower pH values, the hydrogen ion competes with other cations for interaction sites,causing cationic metals to be generally less toxic at lower pH values (Campbell and Stokes 1985, Cusimano etal. 1986). Competition for binding sites at the gill surface is also the reason why metals in hard waters are lesstoxic. Cations such as magnesium and calcium are normally found at concentrations several orders ofmagnitude higher than trace metals, and can successfully out-compete trace metal ions for uptake sites onmembranes.26Both the gills and the body surface of fish are covered with mucus which serve various importantbiological functions. When faced with a low pH and/or elevated heavy metal concentrations, fish secrete mucuson their gills in order to maintain the gill pH (Cusimano et al. 1986, Eisler and Gardner 1973; Skidmore andTovell 1972; Mallat 1985). Mucus on the gill of fish creates an environment which allow ion pumps to maintainthe necessary inflow and outflow of ions (Miller and McKay 1982). The mucus layer also effectively protectsthe gill from the hydronium ion and other cations which are known to cause epithelial damage on the gill(Handy et al. 1989, Mallat 1985, Handy and Eddy 1990). Mucus is known to effectively bind metals when thepH of the surrounding water is above 5. When exposed to water with a pH below 5, the mucus loses anionicsites, and cations are lost from both the gill and the plasma (Handy et al. 1989). This is the reason why at pHless than 5, ion depletion is the most common cause of death in fish (Lauren and McDonald 1985; Lauren andMcDonald 1986).Another function of the gill mucus may be to create a reserve of ions which are available for uptake.Kirschener (1977) suggested that mucus layers may act as a matrix from which essential ions are rapidlymobilized for uptake via the gill epithelium. Part and Lock (1983) found that mucus selectively boundcadmium and mercury while allowing the free passage of calcium. In their study, once the complexationcapacity of the test solution and the mucus was exceeded, a 100 fold increase in cadmium uptake was reported.At lower concentrations, cadmium binding to a specific calcium transport proteins in the gill appeared toenhance cadmium uptake through the mucus layer.Mallatt (1985) reviewed the affect that various toxicants had on gill structural changes. Necrosis andhypertrophy of gill epithelial cells, along with mucus hypersecretion were the most commonly found responsesof gills exposed to heavy metals. Alteration of the osmoregulatory functions of ion pumps in the gills causedby heavy metals distorting carrier molecules was thought to play a major role in causing hypertrophy in thegills.Part and Wikmark (1984) conducted studies using EDTA and citrate to determine some of the factorsthat controlled cadmium uptake across perfused rainbow trout gills. They found that the free Cd 2+ ionpermeated the gill easily, whereas the Cd-EDTA complex was almost completely excluded from the gill plasma.27The Cd-citrate complex was taken up at a rate greater than can be explained by the free Cd2+ concentrationalone, and it was therefore concluded that Cd-citrate complexes were actively taken up across the gill. Theiroverall conclusion was that neither the water's natural complexing capacity or the metal-ligand stabilityconstants could fully explain metal uptake across gill membranes. They suggested that factors such as themolecular radius, complex binding affinities and active uptake of complexes that have metals attached to themmay play important roles in metal transport across the gill.Block and Part (1986) conducted cadmium uptake studies with rainbow trout using several non-polarcompounds. Xanthates are used by mining industries as a flotation agent, and are often released to receivingwaters along with trace metals. They found that several of these organo-cadmium complexes were taken upacross the gill membrane much faster than the free cadmium ion. These compounds, being non-polar, werefound to be highly lipophilic, and the authors concluded that the octanol-water partition coefficient ofcompounds is a crucial parameter which can be used to determine the potential toxicity of a compound. Metalsand metalloids which are capable of being methylated include mercury, tin, palladium, platinum, gold, thallium,selenium, arsenic, tellurium and sulphur (Stumm and Morgan 1981). These elements, once transformed intoalkyl compounds, become non-polar and can easily pass through biological membranes. This can result insignificant biomagnification up the food chain.2.42 Metal Uptake Through the Digestive TractMetal uptake through the digestive tract by fish seems logical considering the high concentrations ofmetals found in food compared to that in the surrounding water. In fish, many different types of digestivetracts exist, and the biochemistry of each differs accordingly. The rainbow trout has a relatively simple, shortdigestive tract which indicates that the fish is mainly piscivorous or insectivorous (Bond 1979). As food passesthrough the pharynx, acid secretions can reduce the pH to values as low as 1.5 in the stomach (Bond 1979).This low pH, along with various enzymes break food particles down into substances which are easily taken upin the intestine. Once the food has passed through the stomach, bile juices elevate the pH to about 8.5 (Conte281%9). At this alkaline pH, nutrients are taken up from the digestive tract and transported to the plasma.While in the digestive tract, low molecular weight complexes which have metals bound to them have been foundto facilitate metal uptake (Evans 1976).In a study designed to document zinc uptake through the digestive tract, Wekell et al. (1983) foundthat body burdens of zinc increased when rainbow trout were fed diets containing 90 itg b•e food or more.Below 90 the fish experienced high mortalities and reduced growth due to zinc deficiencies. Uptakeof zinc via the digestive tract caused concentrations of zinc to increase in the liver, blood and gill. No toxicsymptoms were noticed, even when the zinc concentrations in the food were 1700 Awe.Handy and Eddy (1990) found that dietary zinc accumulates primarily in organs such as gill, liver andkidney (Hardy et al. 1987; Lovegrove and Eddy 1982), and Spry and Wood (1985) found that zinc also appearsin the blood plasma. Hardy and Shearer (1985) found that the uptake of zinc via the digestive tract in rainbowtrout was reduced by the presence of high concentrations of calcium phosphate. Calcium phosphatesprecipitate and absorb zinc when faced with an increased pH as the food passes from the stomach to theintestinal tract.In a study conducted near a sewage treatment plant, Patrick and Loutit (1978) found that fish whichwere fed metal contaminated tubificid worms showed an increase in body metal concentrations within two tofour days. Chromium, lead and to a lesser extent copper, manganese, iron and zinc were all taken up via thecontaminated food source. They felt that the efficiency of the uptake path (water or ingestion) depended onthe concentration gradients across biological membranes and concluded that in some instances, metal uptakeby the food chain may be significant.Singh and Ferns (1978) studied the uptake of metals by rainbow trout which were fed a diet containing30 percent activated sewage sludge. Thirteen elements were measured over a period of 70 days. They foundthat the uptake of elements could be classified into four groups. Elements in the first group (manganese,cobalt, copper and cadmium) showed no increase. The second group consisted of iron which rose initially, andthen reached a constant concentration. The third group (chromium and lead), rose initially but fell to lowlevels near the end of the experiment. The final group (zinc and nickel) increased throughout the duration of29the experiment. The authors felt that the last group of elements may be of concern because there seemed tobe no regulating mechanism in place to control the levels of these metals.2.50 Metal Homeostasis Within OrganismsThe binding and transport of elements upon entering an organism is an area which is still poorlyunderstood. For instance, an incomplete understanding of iron homeostasis still exists, and the biochemistryof this element has been extensively studied for the last 30 years (Petering and Fowler 1986). An even poorerunderstanding exists for trace-element homeostasis. Because of the complexity and poor understanding ofmetal biochemistry, this review will be limited to the discussion of metallothioneins and metallothionein-likeproteins. Special emphasis will be placed on the regulation of copper, zinc and cadmium by metallothioneinsin teleosts.2.51 MetallothioneinsThe discovery of metal-binding proteins was closely related to the search for a biochemical functionof cadmium. Cadmium concentrations in biological tissues were first reported in 1941 by D.P. Maligua. Thisresearcher determined the cadmium to zinc ratio in many biological organisms, and found that the ratio wasa good indication of the relative health of an organism (Maliuga 1941). In 1961, Kagi and Vallee publisheda paper describing some of the physical and chemical characteristics of proteins that were capable of bindingmetals. Based on the high metal and sulphur contents of the proteins, they coined the term metallothionein,and called its apoprotein thionein. Since then, a great deal of metallothionein research has been conductedand after over 30 years of research, an incomplete understanding of this protein's function still exists (Karin1985; Chan et al. 1989).Metallothioneins are ubiquitous in both animal and plant kingdoms, and are found in both eukaryoticand prokaryotic organisms (Brady 1982; Stone and Overnell 1985). In order for a protein to be classified as30a metallothionein, the following characteristics which were described by Kagi and Nordberg (1979) must bemet:o cytosolic location in the cell;o molecular weight of 6,000 to 7,000 daltons and 15,000 daltons from gel-filtrationchromatography;o unique amino acid composition consisting of 25 percent cysteine by composition, 15 cysteineresidues per molecule, and no aromatic amino acids or histidine;o high metal content - six to seven moles of metal atoms per mole of protein;o remain stable at temperatures of 60 °C for 5 minutes.Several authors have noted that typically between 7 and 8 atoms of metal are bound to each mole ofmetallothionein (Olsson and Haux 1985).Metallothioneins are thought to play a major role in the metabolic control of zinc and copper (Kagiand Nordberg 1979; Price-Haughey and Gedamu 1987), and are also responsible for the detoxification ofcadmium and mercury, two elements which have no known biological function (Bremner 1987; Kay et al. 1987).Metallothionein affinities for various metals depends largely on the stability constants between the respectivemetal and sulphur. Displacement of metals can occur in situ (Kito et al. 1982) and the general bindingstrengths are thought to be silver > mercury > copper > cadmium > zinc (Brady 1982; Lauren and McDonald1987).Quantification of metallothioneins and metallothionein-like proteins have been conducted using a widevariety of methods (Klaverkamp et al. 1984; Pierson 1985; Webb 1987; Olafson and Sim 1979; Olafson 1987;Ley et al. 1983; Suzuki 1987; Thompson and Cosson 1984). Gel filtration, radiotracer metal studies, ionexchange chromatography, sulfhydryl analyses, and differential pulse polarography are some of the morecommon methods used to determine metallothionein concentrations. This wide array of analytical techniquesmakes the comparison of different studies difficult, and may contribute to the poor understanding of thefunction of metal-binding proteins (Chan et al. 1989).Experimental induction of metallothioneins in fish have typically been conducted by eitherintraperitoneal injection or exposure to metals in water. The intraperitoneal injections provide a direct doseof metals which are normally partitioned in the organism within five hours (Price-Haughey et al. 1986; Reichert31et al. 1979; Winge et al. 1978). Waterborne exposure to metals also induce metallothionein production, butthe peak concentrations occur within five to 30 days (Bradley et al. 1985; McCarter and Roch 1983).The liver is widely accepted as one of the major sites for metal accumulation in fish (Thomas et al.1985; Lauren and McDonald 1987, Buckley et al. 1982), and metallothioneins are known to be induced there.Metallothioneins are also known to be induced in gill tissues (Olsson et al. 1989; Bradley et al. 1985), in theintestinal tract (Shears and Fletcher 1983) and to a lesser degree in the kidney (Stone and Overnell 1985). Thepresence of metallothioneins in these tissues are thought to function as the first line of defence against metaltoxicity.Recently, several authors have noted that metallothioneins can be induced by factors other than heavymetals. For example, Olsson et al. (1987) tracked the zinc, copper and metallothionein levels in rainbow troutduring a period covering a complete reproductive cycle. They found that metallothioneins along with thecytosolic and total zinc levels paralleled hormone concentrations during the reproductive cycle in the femalefish Seasonal fluctuations of metallothioneins and zinc concentrations were very limited in male fish. Thisresearch clearly documented that factors other than metal concentrations in solution play a significant role incontrolling the levels of hepatic metallothioneins in fish.2.52 Non-Metallothionein Metal Binding ProteinsAlong with metallothioneins, several other metal inducible proteins have been reported (Price-Haugheyet al. 1986; Thomas et al. 1983; Thomas et al. 1985; Roberts et al. 1979; Stone and Overnell 1985).Metallothioneins and metal binding proteins are different due to the presence or absence of functional groupswhich in turn affects the molecular weight of the protein. For instance, Thomas et al. (1983) found thatcadmium was consistently eluted from Sephadex 75 columns at a molecular weight of approximately 10,000daltons whereas metallothioneins were typically eluted at roughly 7000 daltons. In an review of non-metallothionein cadmium binding proteins, Stone and Overnell (1985) conclude that non-thiol proteins are verydiverse and are common in many organisms. In a study of cultured fish cells, Price-Haughey et al. (1986)32found that metallothioneins are by no means the only metal-binding proteins induced during exposure to heavymetals. Some of these non-metallothionein proteins have the potential to be good indicators of metalpollution, but more studies are required to clarify their role in metal homeostasis.2.53 Metals and Metal-Binding in rim nt FishIn this section, studies in which hepatic metallothionein concentrations were obtained from fish exposedto environmental metal pollution will be reviewed. Many metallothionein studies have been conducted in whichexperimental fish have received intraperitoneal injections containing heavy metals. This route of administration,and the subsequent metallothionein induction is not representative of metal exposure via food ingestion orwater exposure (Olsson et al. 1989; Thomas et al. 1983). For this reason, studies which have assessedmetallothionein and metallothionein-like protein induction when fish were exposed to aqueous heavy metalcontamination will be reviewed.Possibly the most frequently cited work on metallothionein induction in wild fish was conducted byRoch and McCarter (and others) in the early 1980's. Buttle lake, British Columbia, has received ARD fromthe Westmin Resources mine at Myra creek. The typical waste from this mine consisted of a 400:20:1 mixtureof zinc, copper and cadmium respectively. Deniseger et al. (1990) conducted a historical review of the researchconducted on Buttle lake since the mine opened in 1966. They found that hepatic metallothioneins and metallevels in salmonid muscle tissues increased in parallel with the water borne metal concentrations.Phytoplankton and zooplankton species diversity decreased during the most polluted time period in the lake.Results from some of the specific studies conducted at Buttle Lake will now be discussed.In a survey conducted on the Campbell river watershed, Roch et al. (1982) collected rainbow trout andanalyzed tissues for zinc, copper, cadmium and metallothionein concentrations. Significant differences werefound between hepatic metallothionein concentrations in the various fish stocks, with the highest levels beingfound in Buttle Lake. No significant differences were found between the hepatic zinc concentrations, but livercadmium and copper concentrations were correlated with the aqueous metal pollution which the fish were33exposed to (Table 2). Roch et al. (1982) felt that the high degree of variation in the liver total metalconcentrations between sampling sites made total hepatic metal concentrations a poor pollution indicator.However, their sample size of 5 for each lake was rather small to draw such a conclusion.Roch and McCarter (1984a) also conducted a four week-long caged fish study in the same lakes asthey sampled in 1982. They found that the correlation between hepatic metallothioneins was significant (F <0.01), but no correlation coefficient was given. The relationships between hepatic metallothioneins and zincconcentrations in solution from several studies in the Campbell river watershed show that a great deal ofvariation was present (Figure 7). Careful examination of Figure 7 reveals that for the caged fish, statisticallysignificant differences between the various groups did not exist.Table 2. Heavy metals found in solution and hepatic metallothionein concentrations found in rainbow troutcollected from the Campbell river watershed. Values are expressed as mean value ± one standard deviation.Lake WaterCopperitg-L4WaterZincug.L-1WaterCadmiumAg-L-1MetallothioneinAmol.g".1 wet massSouth Buttle 539 ± 124 168 ± 50 19.4 ± 6.1 269 ± 23Upper Campbell 496 ± 238 161 ± 20 15.9 ± 7.0 164 ± 37John Hart 1% ± 39 123 ± 33 8.6 ± 2.9 94 ± 18Upper Quinsam 35 ± 14 173 ± 21 4.0 ± 0.6 58 ± 14In a study designed to see if metal toxicity was between different fish stocks, Roch and McCarter(1986) conducted laboratory studies using several stocks of rainbow trout which were exposed to a 400:20:1mixture of zinc, copper and cadmium respectively. Three hatchery stocks were used to conduct 96 hour LC 50studies at the University of Victoria and simultaneously at Corvallis, Oregon. The objective was to see if waterquality and/or experimental equipment may have contributed to differences between the LC50 values foundfor the different stocks. The conclusions were that water quality and/or differences in the experimentalapparatus could not explain the differences found between the three stocks of hatchery fish Age of the fish34and inherent differences between stocks were cited as possible reasons for differing LC50 values. Thesedifferences underline the difficulties associated with comparing toxicity results from various stocks of fish Zinc concentration to q /1)Zn. Cu :^4OO:2O: IFigure 7. Hepatic metallothionein (± one standard deviation) in rainbow trout as a function of heavy metalcontamination. Wild trout July 1981 (o), wild trout August 1981 (0), caged trout May-June 1982 (0), cagedtrout February 1983 (II). (taken from Roch and McCarter 1984).Hamiliton and Mehrle (1986) reviewed the applicability of using metallothioneins as biochemicalpollution indicators in fish which were exposed to varying concentrations of metals. They recommended thatthe spill-over hypothesis which was first proposed by Winge et al. (1974) could be used to further improve thevalue of metallothionein data. The spill-over hypothesis states that metals exert toxic symptoms only whenbound to high molecular weight proteins. When this occurs, the replacement of essential elements inmetalloenzymes by non-essential elements such and mercury or cadmium cause the proteins to becomedisfigured. This disfiguration prevents the metalloenzymes from properly binding to substrate molecules, andcause the metalloenzymes to become dysfunctional. Metallothioneins, which are low molecular weight proteinsare known to bind strongly to mercury, cadmium, copper and zinc. The metallothioneins may preferentiallybind excess quantities of these transition metals, thus allowing the high molecular weight proteins to function35normally. Similar conclusions were obtained by Brown (1977) and Brown and Parsons (1978) when exposingchum salmon (Oncorhynchus keta) to mercury in a controlled ecosystem.In a subsequent paper, Hamilton et al. (1987) conducted studies on uptake and metallothioneininduction caused by water-borne cadmium in brook trout (Salvilinus fontinalis). They concluded thatmetallothioneins can be used as biological indicators of metal stress, but emphasize that the metallothioneindata alone can often be misleading. They highly recommended the fractionation and separate analysis of thehigh- and low-molecular weight protein pools to determine where the metals were being sequestered. Bindingof heavy metals to the high molecular-weight protein pool was found to be extremely toxic to the brook troutin their experiments.36CHAPTER 33.0 METHODS AND MATERIALSField and laboratory methods and materials will be discussed in two separate sections. Standardanalytical methods are referred to when appropriate while discussing laboratory methods and materials.3.10 Field Methods and MaterialsAs mentioned earlier, the experiment was conducted at the Equity Silver mine near Houston, BritishColumbia (Figure 1). The mesocosm apparatus was installed over Foxy Creek and was operational by July 22,1990. The experiment was terminated on September 8, 1990. Fish were placed in the troughs on August 15,1990 and were sacrificed on September 8, 1990.3.11 Mesocosm Description and OperationThe water passed through the mesocosm was diverted from Foxy creek by a 15.25 cm diameterpolyvinyl chloride pipe (PVC) which was run approximately 100 m upstream. A box constructed from plywoodwas placed at the intake end of the water delivery pipe in order to prevent excessive debris from entering themesocosm. The 100 m length of pipe was connected to a fibreglass-lined wooden head tank which had sixteenoutlets in the form of stand-pipes which were used to maintain the flow in each mesocosm at 18 ± 1 L perminute (Figure 8). Water was delivered from the stand-pipe to each trough by 5 cm I.D. nylon pipe. Troughflow rates were measured and calibrated at least 4 times a week using a calibrated V-notch weir.Each trough was constructed from 0.64 cm plexiglass and had dimensions of 1.52 m long by 02 mwide. The troughs had a mixing baffle where the ARD was introduced and a laminar flow area which waslined with 2.5 cm diameter washed gravel to a depth of 7 cm. A weir plate was placed at the end of each37trough to maintain a 6 cm column of water over the gravel substrate. These troll. hs were designed forsampling benthic invertebrates and periphytic algae.Raw ARD was piped down from a seepage pond via a gravity-fed 5 cm PVC pipe. Treatments wererandomly assigned to the troughs and ARD was introduced to the mesocosms starting August 14 using aTechnikon peristaltic pump. The two highest treatments required more ARD than the peristaltic pump couldsupply so precision needle valves were used to deliver the ARD. The ARD delivery flow rates were checkedand adjusted using a stop watch and a graduated cylinder at least 4 times a week.The controls and highest treatment levels were triplicated because it was thought that these treatmentswould experience the greatest degree of variation. All other treatment levels were replicated. TheARD:stream water dilutions were chosen on a logarithmic scale in order to cover a wide range ofconcentrations which would ensure measurable differences. The ARD:stream water dilutions were set at 0:1(control), 1:25000, 1:10000, 1:2500, 1:1000, 1:100 and 1:25. Based on calculations using the ARD chemistrydata, it was decided that exposing fish to the two highest treatments would be wasteful because almost instantdeath would have occurred. The experimental design, the number of trou • hs and the number of fish exposedto each treatment are shown in Table 3.Table 3. Experimental design used at the Equity Silver ARD study.Number of Troughs Number of Fish Exposedto Each TreatmentTreatment ARD:Stream Water3 18 0:1 (control)2 12 1:25,0002 12 1:10,0002 12 1:2,5002 6 1:1,0002 0 1:2503 0 1:25ov38Figure 8. Schematic design of mesocosm apparatus used at the Equity Silver mine. A: head tank assembly showing inflow pipe, water distribution insidethe tank, outflow regulated by standpipes, and mixing chamber of each trough showing water inflow and chemical injection line; B: reservoir of ARD thatwas used as the supply for troughs receiving the ARD delivery by pump; C: trough outflow fitted with drift net, fish tank, and waste line that dischargedto a collection basin which was continuously emptied into the upper seepage pond using a submersible pump (Taken from Perrin, 1991).393.12 Experimental FishThe fish that were used in this experiment were obtained from the Abbotsford Provincial Hatchery.Nine fish were sacrificed and frozen at the Abbotsford hatchery and 75 fish were flown up to Smithers onAugust 13, 1990. Fish were placed in a hatchery facility overnight near Smithers, British Columbia, to recoverfrom the flight. The fish were then transported to the experimental site using an insulated plastic tank whichhad an oxygen aeration system installed.Prior to placing the fish in 36 L Rubbermaid nylon tubs at the experiment site, they were anaesthetizedusing 2-phenoxyethanol in order to accurately measure their mass and length. The average length of the fishwas 11.8 ± 0.9 cm)and the average mass of the fish was 16.8 ± 3.4 g at the beginning of the experiment. Finclippings were also made in order to identify the fish. After masses and lengths were taken, the fish wereplaced in plastic tubs at the end of each trough. A nylon funnel was used to collect all water from the troughsand pass it through the fish tubs. The total hydraulic retention time of the troughs and the fish tubs wasroughly 3.2 minutes. Plastic spigots were installed at the outlet end of the fish tubs in order to collect theexperimental effluent which was pumped back to a seepage pond at the minesite.The fish were fed 1.5 percent of their body weight twice daily. Excess food in the tubs was removedevery second day with a dip net. Daily mortalities were recorded, and starting August 26, dead fish were frozenfor later analyses. On September 8, the fish were killed with a sharp blow to the head. Fish were thenweighed, measured and immediately frozen. The fish were later transported to Vancouver on dry ice.Wild fish were collected from two lakes near the Equity Silver mine. Lu Lake (Figure 1) which isupstream from the mine and Goosly lake which is downstream from the mine were sampled using fly-fishinggear. Lu Lake supplies the potable water to the mine and Goosly Lake has received significant quantities ofARD from the Equity Silver mine in the past (Patterson 1986). Nine rainbow trout were sampled from LuLake on September 5, and 16 rainbow trout were sampled from Goosly Lake on September 10. Fish that werecollected for metallothionein analyses were placed directly on dry ice and shipped to Vancouver.403.20 Laboratory AnalysesSamples collected at the mesocosm experiment were analyzed at a number of laboratories in BritishColumbia. Standard methods references are referred to where appropriate.3,21 Water ChemistryWater samples from the mesocosm apparatus were collected on August 16, August 23, August 30 andSeptember 6, 1990. Water samples for metal analyses were filtered through Sartorius 0.45 um filters using aNalgene hand-held vacuum pump. Care was taken to ensure that the vacuum on the filter was less than 15p.s.i. in order to prevent cell lysis. The samples were collected in 250 mL plastic bottles which had 1 mL ultra-pure nitric acid added to them for preservation. Samples were then shipped via courier to the West VancouverEnvironmental Protection Service (E.P.S.) laboratory where metal scans were conducted using inductivelycoupled argon plasma emission spectroscopy (I.C.P.). Graphite furnace atomic absorption spectroscopy wasused for trace levels on copper and cadmium (E.P.S. 1988). A quality assurance program was conducted bythe E.P.S. laboratory to ensure accuracy of the data.Alkalinity / conductivity samples were collected in 300 mL plastic bottles that were slowly filled tomaximum capacity in order to minimize gas exchange within the bottle. Samples were then shipped to Can-Test Ltd. for analyses. The analyses were carried out in accordance with standard procedures (A.P.HA. 1985).In situ pH readings were taken on the water sampling days. The pH meter was calibrated usingbuffers that were at the same temperature as the stream water. Readin&s were taken after the probe had beenallowed to equilibrate in each trough for exactly ten minutes.The Equity Silver mine conducted analyses on the raw ARD twice weekly during the study.Parameters monitored included pH, sulphate, conductivity, total acidity and metal analyses. Dissolved (10.45gm) and total metal samples were collected for analysis of copper, zinc and iron using atomic absorptionspectroscopy. All procedures were conducted according to standard methods (A.P.HA. 1985). Total metalsamples were digested on a hot plate by heating concentrated nitric acid to near dryness. Nitric acid was added41again to the sample and heated until a gently refluxing action took place. The sample was then made up to50 mL and directly aspirated using flame atomic absorption spectroscopy.Once all the data was generated, it was entered into a spreadsheet for manipulation. All of thechemical data collected in this experiment are presented in Table 25 in Appendix 1. Statistical calculations forthe water chemistry data were conducted using SPSS/PC+ V2.0 software (Norusis 1988). While calculatingmeans and standard deviations for the water chemistry data, values which were below the limit of detection(LOD) were assigned a value 1 µg•L-1 less than the actual detection limit (Gilbert 1987).3.22 Geochemical ModellingThe MINTEQA2 geochemical model was run in order to estimate the bioavailability of metals in theexperimental troughs. Predicted concentrations (mg•L -1) for aluminum, copper, cadmium, iron, manganeseand zinc were calculated by multiplying the raw ARD total concentrations by the dilution factor for eachtreatment level and adding this value to the respective background concentration (Table 5). While running themodel, the pH and alkalinity was fixed at the appropriate field reading and the pE was fixed at 13 milli-volts.This redox potential is representative of oxygen-saturated systems (Stumm and Morgan 1981).The diffuse layer model adsorption algorithm was used to predict the effect that complexation had onmetal speciation in the troughs. In order to run this algorithm, parameters such as the number of adsorptionsites-nm-2, surface area of adsorbent (m2-g-1) and the concentration of the adsorbent present in solution(mgL-1) must be known. Iron oxide was assumed to have a specific surface area of 320 m 2-g-1 with 11 sitesper nm2 (Jenne 1968). Anderson and Benjamin (1990) measured the surface area of iron oxides and obtainedan estimate of 186-201 m2 .g-1 . Cowan et al. (1991) estimated that amorphous iron oxide had a surface areaof 600 m2 .g-1 . A medium estimate for surface area was used to compensate for adsorption reactions that werenot modelled for silica, manganese and aluminum. The highest estimate of 600 m1/4 .1 was not used becauseit is unlikely that the retention time in the troughs would have been sufficient to allow full adsorption on ironoxides in the troughs.42Thermodynamic information existed in version 3.0 of MINTEQA2 for iron oxide adsorption with zinc,cadmium, copper, beryllium, nickel, lead, calcium, barium, sulphur, phosphorus, arsenic and boron (Allison etal. 1990). Thermodynamic information did not exist for adorption reactions with manganes, silica andaluminum. Thermodynamic information was also lacking with respect to complexation reactions that may havetaken place between trace metals and the organic matter found in the stream. This lacking information limitsthe applicability of the model results.Another problem with the model results is that the troughs were not in a state of equilibrium. Thedegree of complexation and adsorption remained unknown because the 3 minute retention time would not havebeen adequate to allow all reactions to carry on to completion. The kinetic component of this experiment mayhave been confounding because complexation and adsorption reactions would have taken place faster at thehighest treatment levels due to Gardiner's (1969) law of mass action. This kinetic problem could not beresolved. Because of these problems, the model output data should be used to obtain order of magnitudeestimates. Progress is currently being made to improve the databases in geochemical models, and the utilityof them will improve in the future.3.23 Metallothionein AnalysesOnce the frozen fish samples arrived in Vancouver, they were kept frozen at -20 °C. Livers weredissected while the fish were still frozen and were shipped on dry ice to C.B. Research International in SidneyBritish Columbia for hepatic metallothionein analyses. These analyses determined the level of thiolic proteinfound in the cytosol of this liver, and in the remainder of this report, will be referred to as metallothioneins.The following analytical protocol was taken from the report prepared by C.B.R. International (1990).Livers were slowly thawed and kept on ice for all manipulations. Liver mass was measured andhomogenization buffer (50 mM Tris, pH 8.6 containing 2- mercaptoethanol as an antioxidant, and 0.02 percentNaN3 as a microbial inhibitor) was added to create a ratio of roughly 5:1 buffer:liver tissue. Samples werehomogenized for thirty seconds using a Brinkman Polytron Tissue Dismembrator at medium speed.43Following homogenization, samples were separated for two treatments: lyophilization andmetallothionein (MT) analysis. The remaining homogenate was re-frozen for total liver metal analyses. Forlyophilization, aliquots of 0.5 g homogenate were frozen and dried in a Labconoco freeze dryer.Samples were prepared for MT analysis by denaturation of 100 to 400 mg with an equal weight of 95percent ethanol. The cytosolic denatured homogenate was allowed to stand for two hours at 5 °C, thencentrifuged at 20,000 g for thirty minutes at 5 °C. The supernatant was retained for MT determination.Differential pulse polarographic analysis was carried out using a Metrohm (Brinkman) PolarographVA 646 Processor and VA 647 Stand. Twenty mL of Brdicka support electrolyte, 0.001 M of(NH3)6CO(III)CL3 .3H20 prepared in 1 M NH4OH/1 M NH4C1 with 100 id. of a 0.5 percent Triton-X-100surfactant (secondary maximum suppressant) was contained in a jacketed cell cooled to 5°C(±.5°C). Abaseline scan was run prior to the addition of sample or standard. Analyses were performed on serial duplicateadditions of appropriate volumes (1.5 to 20 gL) denatured ethanol extract (comprised mostly ofmetallothionein) using fresh electrolyte for each sample. Printed curves of signal intensity (nanoAmps, NA)as a function of applied potential were obtained. Sample values were interpolated on a working curve preparedfrom measurements of a range of standard additions of a solution containing 50 mg•L -1 of rabbitmetallothionein MT-II (Sigma Chemicals) prepared in homogenization buffer.324 Tissue AnalysesTissues were analyzed for metals using a Thermo Jarrel Ash Video 22 flame atomic absorptionspectrophotometer (FAA). Tissues analyzed include liver, gill and muscle. Muscle and gill were dissected fromthe fish using stainless steel instruments. These tissues were then placed in a freeze dryer and allowed to drycompletely. The freeze-dried material was then homogenized in a stainless steel coffee grinder.The microwave oven used for tissue digestions was a CEM model MDS 81D. Approximately 30 mgdried muscle or gill was carefully weighed into 60 mL teflon bombs, and then 4 mL HNO 3 and lmL HCl addedto them. Similarly, 30 mg dried liver tissue was carefully weighed into 6 mL teflon bombs which had 0.4 mL44HNO3 and 0.1 mL HCl added to them. All acids were Analar reagent grade. The bombs were allowed topredigest overnight with the lids only loosely screwed on. This allowed excessive CO 2 to vent which wasderived from easily oxidizable organic matter. The caps were then tighten to a standard torque and the entireteflon bomb was weighed.The bombs were then placed in the microwave oven for digestion. Based on the volume of acids inthe microwave chamber, calculations for power settings were made which allowed an approximate internaltemperature of 170 °C to be reached in the teflon bombs. Power settings were then reduced in order tomaintain this temperature for 5 minutes. The bombs were then removed, allowed to cool for at least 30minutes, and then re-weighed. The microwave program settings for the 60 mL bombs was 2 minutes at 100percent power (660 watts), 6 minutes at 40 percent power, 4 minutes at 20 percent power and a final cool downperiod of 10 minutes (0 percent power). The smaller bombs had one-tenth of the power used for the 60 mLbombs applied to them for the same amounts of time.Once the samples had cooled, the gill and muscle samples were made up to 25 mL, and the liversamples were made up to 2.5 mL with distilled deionized water. Blanks along with National Bureau ofStandards (NBS) certified oyster tissue standards were processed in a fashion identical to the samples foranalysis as part of a quality assurance program.Approximately 20 muscle tissue samples from the experimental fish, along with several samples fromthe wild stock were run for cadmium analysis and all were below the LOD. For this reason the remainingsamples were not analyzed and the muscle cadmium data was omitted.45CHAPTER 44.0 RESULTS AND DISCUSSION4.10 Water ChemistryElements which were below or near the I.C.P. LOD are listed in Table 4. When the LOD wasexceeded or equalled, the exception is listed. Silver, molybdenum, lead, antimony, tin, titanium and vanadiumwere all below their respective limits of detection at all treatment levels.Table 4. Data for samples taken from the experimental troughs in which elements were below or just abovethe I.C.P. limit of detection (n=64).Element Detection LimitAg-L-1CommentsAg 10 all values below LODMo 10 all values below LODPb 50 all values below LODSb 50 all values below LODSn 50 most values below LOD except for- Aug. 23 trough #3 = 70 Ag•L -1- Aug. 30 trough #7 = 50 Ag•L -1Ti 2 all values below LODV 10 all values below LODChemical parameters that were measured in the raw ARD and the control troll  • hs are listed in Table5. The raw ARD had total iron, sulphate, copper and zinc concentrations around 1000, 6500, 45, and 25 mg•L"1 respectively. Sulphate, pH, iron, zinc, copper and conductivity data collected by the Equity Silver mine showthat the composition of the ARD did not change significantly during the experiment (Table 5, n = 9). In theraw ARD, dissolved metal concentrations were not significantly different from the total metal concentrations(F < 0.001), and this was likely due to the pH of 2.5 which would have caused the metals to remain in solution.46The control trough chemistry data (n =12) are also presented in Table 5 in order to depict thedifferences between the ARD and the stream background levels. The humic and fulvic acid concentrations,along with the cyanide concentrations were determined by C.B. Research in Sidney, British Columbia. Thesesamples were collected from the experimental troughs on August 23, 1990, and were placed on ice and shippedvia air express to C.B. Research Ltd. for analyses.The summarized water chemistry data from the mesocosm study are presented in Table 6. Theaddition of ARD had little or no effect on the concentrations of several elements in the experiment. Bariumand potassium remained around 0.019 mg•L -1 and 2.8 mg•L-1 respectively. Elements that remained at theirdetection limit and increased only at the highest treatment levels include arsenic, beryllium, cobalt, chromium,nickel, phosphorus, calcium, magnesium, sodium, silica and strontium. Elements which increased steadily overall treatment levels included aluminum, cadmium, copper, iron, manganese and zinc. Alkalinity (mg CaCO31:1), conductivity (AS-cm -1), total hardness (mg CaCO31.: 1) and calcium and magnesium hardness (mgCaCO31:1) all increased only at the highest treatment levels. Chromium and selenium remained near theirdetection limits in all treatments. Boron was the only element which decreased at the two highest treatmentlevels. As would be expected, the in situ pH readings decreased as greater volumes of ARD were added tothe experimental troughs.47Table 5. Chemistry results for control troughs and raw ARD at the Equity Silver mine mesocosm study.Values listed are as mg•L -1 (± one standard deviation) unless noted otherwise. Values for the raw ARDsamples are reported as total metal concentrations, whereas the control trough data are dissolved metalconcentrations (i.e. 5 0.45 Am). Total hardness, Ca + Mg hardness and alkalinity are all expressed as mgCaCO3•1.:1 .ParameterEquity SilverRaw ARDmg•1,-'ControlTroughsmg•L-'pH 2S2 t 0.05 7.08 ± 0.31ConductivityMS-cm - '6946 ± 168 166 t 9Total Hardness . 73 ± 8Ca + Mg Hardness . 73 t 8Alkalinity . 36SO4- 7187 ± 325 53S ± 20.2Ag < 0.015 < 0.010Al 360 < 0.050As 6.98 < 0.050B <0.100 <0.010Ba < 0.010 0.019Be 0.037 <0.001Ca 419 20 ± 3Cd 0520 < 0.0001Co 1.6 < 0.005Cr <0.015 <0.005Cu 40.1 t 1.7 0.003 ± 0.001Fe 1003 ± 119 0.055 ± 0.025K . 3 ± 1Mg 271 6 ± 0.23Mn 38 0.003 ± 0.002Mo < 0.030 < 0.010Na . 3.43 t 0.198Ni 4.08 < 0.020P 12.9 < 0.100Pb < 0.050 < 0.050Sb < 0.200 < 0.050Se <0.200 <0.050Si 385 8.1 ± 0.56Sr 3.17 0.199 t 0.010V 0.152 < 0.010i21 28.9 t 1.2 0.009 ± 0.011Fulvic Acid 0.050Humic Acid . 0.050CN- . 0.50048Table 6. Water chemistry results for the Equity Silver mine mesocosm study. All values are in mg•I. -1 unless noted otherwise. Ratios in the top rowreflect the ARD:Foxy Creek water dilution ratio. Alkalinity, total hardness and Ca + Mg hardness are all expressed as mg CaCO 3•l. -1 .Analysis Control 1:25000 1:10000 1:2500 1:1000 1:250 1:25pH 7.08 t 0.31 6.95 t 0.35 6.95 t 0.22 6.88 t 0.21 6.34 t 0.48 5.72 t 0.25 3.48 t 0.28Alkalinity 36 t 3 39 t 10 30 t 11 35 t 5 31 t 3 23 t 7 0ConductivityµS• cm"'166 t 9 166 t 8 169 t 8 163 ± 15 172 t 6 190 ± 10 724 t 216Total Hardness 73.4 t 8 72.9 t 10 75.3 t 9 76.3 t 10 76.6 t 10 82.3 t 12 234 t 29Ca + MgHardness73.0 t 8 72.2 t 10 74.0 t 9 74.8 t 10 73.2 t 10 80.4 t 11 156 t 19Al <0.050 0.068 t 0.032 0.086 t 0.052 0.136 t 0.043 0.346 t 0.112 0.103 t 0.65 14.67 t 2.41As 0.054 t 0.013 < 0.050 < 0.050 0.050 t 0.004 < 0.050 0.051 t 0.004 0.198 t 0.078B 0.011 t 0.006 0.014 t 0.014 0.011 ± 0.004 0.010 t 0.004 0.017 t 0.018 < 0.010 0.010 t 0.003Ba 0.019 t 0.001 0.019 t 0.002 0.019 t 0.001 0.020 t 0.001 0.019 t 0.001 0.019 t 0.002 0.019 t 0.001Be < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.002 t 0.001Ca 19.6 t 3.0 19.4 t 3.7 19.9 t 3.0 20.2 t 3.6 19.5 t 3.5 21.1 t 3.6 35.6 t 5.6Cd < 0.0001 < 0.0001 0.0002 t 0.0001 0.0001 t 0.00005 0.0002 10.00008 0.0006 t 0.0003 0.017 t 0.004Co < 0.005 < 0.005 < 0.005 < 0.005 0.005 t 0.001 0.007 ± 0.003 0.119 t 0.068Cu 0.003 t 0.001 0.007 t 0.004 0.013 t 0.011 0.014 t 0.005 0.037 t 0.012 0.051 t 0.027 1.530 t 0.247Cr < 0.005 < 0.005 < 0.005 < 0.005 < 0.005 < 0.005 0.005 t 0.002Fe 0.055 t 0.025 0.134 t 0.060 0.308 t 0.277 0.322 t 0.161 0.788 t 0.422 0.722 t 0.423 26.53 t 13.73K 2.8 t 1.2 3.3 t 2.4 2.8 t 1.5 2.7 t 1.8 3.0 t 1.6 2.9 t 1.8 2.8 t 1.6Mg 5.8 t 0.2 5.8 t 0.3 5.9 t 0.4 5.9 t 0.3 6.0 t 0.3 6.7 t 0.6 16.3 t 1.6Mn 0.003 t 0.002 0.009 t 0.006 0.027 t 0.044 0.018 t 0.006 0.050 t 0.016 0.159 t 0.070 1.657 t 0.595Na 3.4 t 0.2 3.3 t 0.1 3.5 t 0.1 3.4 t 0.1 3.6 t 0.1 4.1 t 0.4 11.7 t 1.7Ni < 0.020 < 0.020 < 0.020 < 0.020 < 0.020 0.018 t 0.004 0.189 t 0.031Se 0.053 ± 0.009 0.052 t 0.007 < 0.050 0.050 t 0.004 0.053 t 0.011 < 0.050 < 0.050P < 0.100 < 0.100 < 0.100 < 0.100 0.112 t 0.036 < 0.100 0.392 t 0.117Si 8.1 t 0.6 7.9 t 0.5 8.1 t 0.3 8.2 t 0.5 8.0 t 0.3 8.1 t 0.4 9.7 t 0.6Sr 0.199 ± 0.010 0.197 t 0.012 0.200 t 0.008 0.202 t 0.010 0.198 t 0.007 0.210 t 0.012 0.303 t 0.021Zn _^_ 0.0089 t 0.011 0.009 t 0.008 0.024 t 0.027 0.012 t 0.0130.031 t 0.032 0.095 t 0.048 1.147 t 0.222494.20 Toxicity of ARD to Experimental FishThe concentration at which 50 percent of the test organisms died (LC 50) was calculated using Probitanalyses with the Statistical Analysis System (SAS). The number of fish exposed to each treatment and thenumber of days which the fish survived are listed in Table 7. Each number in the last column of Table 7depicts the number of days a fish survived the experimental treatment. The 96 hour LC 50 was calculated tooccur when 11.35 mL ARD • minute-1 would be added to a trough. This is equal to 1 part ARD added to1400 parts stream water. At this rate of ARD addition, the concentrations of copper, zinc, aluminum andcadmium would be 0.028, 0.027, 0.277 and 0.0003 mg•L4 respectively. The ARD addition rate at whichroughly half of the fish died during the 23 day experiment was calculated to be 3.15 mL • minute 1 . This rateof 1 part ARD to 5150 parts stream water corresponds to copper, zinc, aluminum and cadmium concentrationsof 0.010, 0.014, 0.113 and 0.0001 mg•L-1 respectively. Thus, if groundwater discharges or spills occur wherethe above dilutions are reached, indegenous fish would be expected to die.Table 7. Number of fish exposed to each treatment level and days after initiation of experiment which the fishdied. Each number in the last column denotes a single mortality.Treatment Number of FishExposedNumber of DaysSurvivedControl 18 211:25,000 12 101:10,000 12 •1:2,500 12 4, 7, 8, 10, 11, 11, 11, 131:1,000 6 1, 2, 4, 7, 11, 124.30 Geochemical ModellingIn the following section, the results from the MINTEQA2 geochemical model will be discussed. Thefocus of the discussion will be placed on the 6 elements which rose in a consistent manner with increasing ARDconcentrations. The problems associated with reaction kinetics, lacking thermodynamic information and lacking50stability constants for organic matter should be kept in mind while reading this section. As mentioned earlier,the model output should be used to obtain order of magnitude estimates.By inspecting Tables 8 through 14, it may be noted that all of the experimental troughs were predictedto be saturated with respect to iron. A gradient of iron oxide precipitates was evident on the periphyton platesat the end of the troughs signifying that iron oxides were precipitating at all treatment levels. All of the testsolutions except the highest treatment were predicted to be saturated with respect to aluminum concentrations.Both iron and aluminum oxides are very insoluble in water (Weast et al. 1986).In the control troughs (Table 8), 99.8 percent of the silica and 100 percent of the iron would haveprecipitated according to the model. Because these troughs received no raw ARD, it is likely that they werein at least a pseudo-equilibrium state. Elements which would have adsorbed onto iron oxide included copper(91.1 percent of the total) and zinc (36.6 percent of the total). Because cadmium was below the LOD, it wasexcluded from the control model. One fish died after 21 days of exposure to this treatment level (Table 7),and it is unlikely that metal toxicity was responsible for this.In the 1:25,000 treatment (Table 9), adsorption of copper (92.2 percent), cadmium (13.2 percent) andzinc (34.6 percent) onto iron oxide may have played a crucial role in reducing the toxicity of these elements.Measured concentrations of aluminum, copper, iron and manganese were all slightly higher than the predictedconcentrations. Slight stream flow variations and analytical errors may have been responsible for the deviationsfrom predicted values. All calculated values in Table 8 were within the error bounds of the measuredconcentrations, therefore it was concluded that the measured concentrations were not significantly differentfrom the calculated concentrations. Both the measured and predicted cadmium concentrations were below thegraphite furnace LOD. The pH of 6.95 ± 0.35 did not differ significantly from the control trough value of 7.08t 0.31 (F = 0.93). At the 1:25,000 treatment, one mortality occurred on day 10 (Table 7). Aluminum,copper, cadmium and zinc concentrations were again below any reported concentrations which are know to betoxic.51Table 8. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in the controltroughs. Alkalinity values are expressed as mg CaCO31: 1 and pH is in relative units.Element MeasuredConcentrationmg1:1PredictedConcentrationmg1:1PrecipitatedConcentrationmg1:1AdsorbedConcentrationmg-L-1pH 7.08 ± 031 . . .Alkalinity 36 ± 3 . . .Al <0.010 . 0.010Cd < 0.00001 . .Cu 0.003 ± 0.001 . 0 0.0027Fe 0.055 ± 0.025 . 0.055 .Mn 0.003 ± 0.002 . • .Zn 0.009 ± 0.011 . 0 0.003Table 9. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals in the1:25000 treatment tron• hs. Alkalinity values are expressed as mg CaCO30 and pH is in relative units.Element MeasuredConcentrationmg1:1PredictedConcentrationmg1:1PrecipitatedConcentrationmg1:1AdsorbedConcentrationmg.L4PH 6.95 ± 0.35 . .Alkalinity 39 t 10 . . .Al 0.068 ± 0.032 0.064 0.064 0Cd < 0.00001 0.00002 0 2.64' 10Cu 0.007 ± 0.004 0.004 ± 0.000 0 0.0036Fe 0.134 ± 0.060 0.095 ± 0.005 0.095 .Mn 0.009 ± 0.006 0.0041 . .Zn 0.009 ± 0.008 0.010 ± 0.000 0 0.0037At the 1:10,000 treatment level, elements that were elevated above background concentrations includealuminum, copper, iron, manganese and zinc (Table 10). The error bounds on the measured values fully52encompassed the predicted values for copper, iron, manganese, cadmium, aluminum and zinc. Predictedadsorption of copper (92.2 percent), zinc (34.6 percent) and cadmium (13.2 percent) again reveal the potentialimportance that adsorbing surfaces may have had on metal speciation. The pH of 6.95 t 0.22 was notsignificantly different from the control trough pH of 7.08 t 0.31 (F = 0.87), indicating that the bufferingcapacity of the stream was not severely affected by the addition of one part ARD to 10,000 parts stream water.There were no mortalities at this treatment level during the 23 day experiment (Table 7), and this supports thehypothesis that the mortalities at the lower treatment levels may have been caused by factors other than metaltoxicity.Table 10. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals in the1:10,000 treatment troughs. Alkalinity values are expressed as mg CaCO3L4 and pH is in relative units.Element MeasuredConcentrationmg4:1PredictedConcentrationmg.1:1PrecipitatedConcentrationmg.r1AdsorbedConcentrationmg.I.:1PH 6.95 t 0.22 . . .Alkalinity 30 ± 11 . . .Al 0.086 t 0.052 0.088 0.088 0Cd 0.0002 t 0.0004 0.00005 0 0.000006Cu 0.013 ± 0.011 0.007 ± 0.000 0 0.006Fe 0.308 t 0.277 0.155 I 0.012 0.155 .Mn 0.0085 t 0.006 0.0058 . .Zn 0.009 t 0.008 0.011 t 0.000 0 0.006Elements which increased above background concentrations at the 1:2,500 dilution level includedaluminum, copper, iron, manganese and zinc (Table 11). The trough pH (6.88 t 0.21) was lower than thepH of 7.08 found in the control troughs (F = 0.12). At this treatment level, the buffering capacity of thestream seems to have been affected. Predicted values for copper, iron and zinc were within one standarddeviation of measured values. Adsorption of copper (95 percent), zinc (39.9 percent) and cadmium (15.7percent) was predicted using the model. In this treatment, 8 out of 12 fish died during the experiment (Table537), signifying that at the 1:2500 dilution ratio, the ARD was acutely toxic. For the mortalities, the averagesurvival time was just over 9 days.Table 11. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals in the1:2500 treatment troughs. Alkalinity values are expressed as mg CaCO3L-1 and pH is in relative units.Element MeasuredConcentrationmg-L-1PredictedConcentrationmg-L-1PrecipitatedConcentrationmg-L-1AdsorbedConcentrationmg-L-1pH 6.88 ± 0.21 . . .Alkalinity 35 ± 5 . . .Al 0.136 ± 0.043 0.188 0.188 0Cd 0.0001 ± 0.0004 0.0002 0 0.00003Cu 0.014 ± 0.005 0.019 ± 0.001 0 .018Fe 0.379 ± 0.130 0.456 ± 0.048 0.456 .Mn 0.018 ± 0.006 0.0217 . .Zn 0.012 ± 0.013 0.020 ± 0.000 0 0.008By examining the predicted versus the measured concentrations, it can be noted that all of themeasured concentrations were below the predicted concentrations (Table 11). This may have been due toprecipitation events in which elements were bound onto particles larger that 0.45 Am. These particles mayhave settled out of solution or may have been excluded from the filtrate by the filter. The reduced pH wasnot low enough to prevent the complexation and/or adsorption of metals onto adsorbing and complexingagents. However, the solution was still acutely toxic.One possible explanation of this is that the metal-containing particles may have been adsorbed ontothe gill filaments of the experimental fish Once on the gill, these substances may have caused gill irritationor toxic effects. Howarth and Sprague (1978) hypothesized that metals and metal complexes may becomeentrapped in the mucous surrounding the gill. Excretion of carbon dioxide and ammonia from the gill reducesthe pH directly adjacent to the gill and may cause some of the relatively benign forms of metals to be54transformed into their most toxic ionic form. This type of metal transformation may have taken place at the1:2500 treatment level.Table 12 displays the results from the 1:1,000 treatment. At this treatment level, the pH (6.34 ± 0.48)was significantly lower than the pH found in the control troughs (F = 0.04). The adsorption of 99.2 percentof the copper, 385 percent of the cadmium and 68.8 percent of the zinc onto hydrous iron oxide was predictedusing the model. Measured copper, cadmium, iron and zinc concentrations were all lower than the predictedtotal concentration, signifying that precipitation events were significant. At this treatment level, all of the testfish were dead within 12 days (Table 7), with the mean survival time being 6 days.Table 12. Comparison of the measured, predicted, precipitated and adsorbed concentrations of metals in the1:1,000 treatment troughs. Alkalinity values are expressed as mg CaCO31: 1 and pH is in relative units.Element MeasuredConcentrationmg-L-1PredictedConcentrationmg-L-1PrecipitatedConcentrationmg-L-1AdsorbedConcentrationmg-L-1PH 6.34 ± 0.48 . . .Alkalinity 31 ± 3 . . .Al 0.346 ± 0.111 0.294 0.294 0Cd 0.00024 t 0.0001 0.0005 0 0.00008Cu 0.037 ± 0.012 0.043 ± 0.002 0 0.0427Fe 0.788 ± 0.422 1.058 ± 0.119 1.058 .Mn 0.050 ± 0.016 0.041 . .Zn 0.031 t 0.032 0.037 ± 0.001 0 0.015Table 13 reveals the results from the 1:250 treatment. The pH of 5.72 ± 0.25 was substantially lowerthan the control pH (F < 0.001). Measured copper and iron concentrations were significantly lower than thepredicted concentrations. The measured aluminum concentration was less than 10 percent of the predictedvalue. Aluminum readily precipitates out at pH 5.5 to 6.0 due to the formation of aluminum hydroxide(Herrmann 1987). The trot] , hs at this treatment typically had a white opaque colour, which may have beendue to aluminum hydroxide precipitation. The model again predicted significant amounts of the copper (98.355percent) and zinc (45.4 percent) would be adsorbed onto hydrous iron oxide. Cadmium was predicted to loseonly 19 percent of its total concentration to adsorption with iron oxide. The very low measured concentrationsof iron, cadmium and copper relative to the predicted concentrations found in this treatment may have be duein part to the co-precipitation of these elements with aluminum hydroxide and iron oxides. No fish wereexposed to this treatment.Table 13. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in the 1:250treatment troughs. Alkalinity values are expressed as mg CaCO30.Element MeasuredConcentrationmg-L-1PredictedConcentrationmg.1:1PrecipitatedConcentrationmg.L4AdsorbedConcentrationmg.L4pH 5.72 ± 0.25 . . .Alkalinity 23 ± 7 . . .Al 0.103 ± 0.065 1.434 1.434 0Cd 0.00063 ± 0.0003 0.0021 0 0.0004Cu 0.051 ± 0.027 0.163 ± 0.007 0 .160Fe 0.722 ± 0.422 4.069 ± 0.475 4.069 .Mn 0.159 ± 0.070 0.144 . .Zn 0.95 ± 0.48 0.124 ± 0.005 0 .0645At the 1:25 treatment (Table 14), the pH was reduced to 3.48 ± 0.28. The alkalinity of this solutionwas 0 (Table 5), signifying the complete depletion of the stream's carbonate buffering capacity. The substantialpH reduction caused many metals to remain in solution. A thick layer (5 cm) of what appeared to be ironoxide was present in the bottom of the trough, indicating that even at the low pH of 3.5, the troughs were stillsaturated with respect to iron. Measured values for iron, copper, manganese, aluminum and zinc were notsignificantly different from the predicted values. Results from the model predict that adsorption would nothave played a major role in binding several potentially toxic elements. The pH of 3.5 would have preventedthe adsorption and precipitation of zinc and cadmium, and only 2.3 percent of the copper would have been56adsorbed onto iron oxides. This solution would have been instantaneously toxic so no test fish were exposedto this treatment.Table 14. Comparison of measured, predicted, precipitated and adsorbed concentrations of metals in the 1:25treatment troughs. Alkalinity values are expressed as mg CaCO31: 1 and pH is in relative units.Element MeasuredConcentrationmg.1:1PredictedConcentrationmg-I:1PrecipitatedConcentrationmg-L4AdsorbedConcentrationmg-I:1pH 3.48 ± 0.28 . . .Alkalinity 0 . . .Al 14.675 ± 2.412 13.890 4.292 0Cd 0.017 ± 0.004 0.0205 0 0Cu 1.530 ± 0.247 1.607 ± 0.068 0 0.037Fe 26530 ± 13.730 40.193 ± 4.751 40.193 .Mn 1.675 ± 0.595 1.843 0 0Zn 1.147 ± 0.222 1.165 ± 0.050 0 0Predicted metal concentrations are in general agreement with the measured concentrations for mostof the elements shown in Tables 8 through 14. Predicted and measured values may have been in closeagreement because colloidal and other small particles (i.e. 10 - 500 nm) are known to pass through 0.45 Amfilters (Florence and Batley 1980). Colloidal materials include both inorganic and organic complexes whichmetals may be bound to. The importance of colloidal materials in aqueous solutions has been of great concernto aquatic toxicologists and aquatic chemists for some time (Morrison 1987). Colloidal particles are generallythough to be non-toxic because of their tendency to flocculate with other particles which then may settle outof solution.574.31. ZincZinc increased from background levels of 0.008 mg•L-1 to 1.147 mg•L-1 in the highest treatment level.Standard deviations for zinc are ± 100 percent for all treatment levels except the top two. All calculatedconcentrations for zinc were within one standard deviation of the measured dissolved concentration.The variation in the zinc data may have been due to the fact that the E.P.S. laboratory is not a cleanlaboratory (Millward 1991). The determination of zinc is very susceptible to contamination in laboratoriesbecause rubber, paints, paper, dust, fumes, human skin, hair and clothing particles are all known to containconsiderable concentrations of zinc (Florence 1980). Martin et al. (1980) warn that zinc is commonly foundin almost all unpurified reagents, water supplies, on the walls of glass and plastic storage containers, and inairborne particles. They advise that extreme precautions should be taken when trying to analyze tracequantities of zinc, and warn that zinc and lead are the two elements most susceptible to laboratorycontamination.Zinc exists almost exclusively as the aquo ion [Zn(H20)6]2+ at pH < 7.0 (moderate alkalinity) (Spear1981). In this form, zinc is commonly thought to be most toxic (Skidmore 1964). At low alkalinities, the aquoion is expected to constitute 50 to 100 percent of the total zinc concentration, with hydroxyl and carbonatecomplexes becoming significant only at pH > 7.5. Studies conducted by Lloyd (1960) have shown that zinccarbonate was toxic to rainbow trout. Bradley and Sprague (1985) concluded that zinc hydroxide [Zn(OH) 2]was not toxic and go on to question the fmdings from Lloyd's 1960 work. In the study conducted by Bradleyand Sprague (1985), pH, water hardness and alkalinity were altered in order to investigate factors that causedacute lethality. Zinc is thought to be toxic when in the Zn 2+ and the Zn0H + forms, and in solutions with lowalkalinity and neutral pH (Pagenkopf 1980). Pagenkopf (1980), felt that competition between aquated zinc,the hydronium ion and the hardness cations resulted in competition for binding sites on the gill in hard wateror in water below neutrality. This had the overall effect of reducing the toxicity of the aquated form of zinc.Zinc forms reasonably strong complexes with chloride and cysteine, and powerful complexes withsulphide. Zinc also adsorbs strongly on ferric hydroxide at pH values above 7; on silica and alumina; onmanganese dioxide; and on clays and organically coated minerals (Florence 1980). Stability constants for zinc58and organic matter are generally quite low, implying that organics do not strongly sequester zinc in naturalwaters (Irving and Williams 1948). Randhawa and Broadbent (1965) found that at the optimum pH of 8.5, only10 to 15 percent of the total zinc was bound to humic substances. Wilson and Kinney (1977) found that humicsubstances did influence the speciation of zinc in pure solutions, but they felt that the overwhelming presenceof cations such as calcium and magnesium in natural waters would effectively out-compete zinc for binding siteson organic matter.The model output for the 1:1000 treatment level predicted that if the complexation reactions wouldhave completed, there would have been only 10 itg•L -1 Zn2+ . If no complexation events would have occurred,the predicted Zn2+ concentration would have reached 33 Ag•L -1 . Based on these estimates, it is notanticipated that zinc would have played a major role in the experimental trough toxicity. Zinc is expected tobe detrimental to aquatic organisms only at concentrations greater that 0.100 mg•L -1 (Spear 1981). Zincexceeded this level only at the two highest treatment levels, and it is suspected that copper, cadmium and thereduced pH would have played a greater role in the toxic effects at those treatment levels.4.32 CopperCopper increased gradually in the experimental trou • hs from background concentrations of roughly0.003 mg•L-1 to nearly 2 mg•L4 in the highest treatment. The variation in the copper data was relatively low,and this may be attributed to the good analytical characteristics of copper. This element displays a highsignal/noise ratio on analytical instruments so precise readings are readily obtained.Copper is known to bind very strongly to organics (Irving and Williams 1953), and as much as 58 to98 percent of copper found in aquatic systems is normally bound to organics (Stiff 1971; Florence 1977;McKnight et al. 1983; Sunda and Lewis 1978). The fulvic acid concentration in the experimental troughs mayhave complexed with some of the copper, and thus would have caused it to become relatively non-toxic.However, complexation kinetics with multidentate organic molecules are typically slow. This, coupled with the59relatively low concentrations of fulvic acids (50 gg•L-1), may have caused the fulvic acids to play a relativelysmall role in rendering the copper inert.Inorganic forms of copper that have been shown to be toxic to fish include Cu2+ , Cu0H+ and possiblyCu2(OH) 22+ (Chakoumakos et al. 1979; Howarth and Sprague 1978). Copper carbonate (CuCO 3) is thoughtto be the most important inorganic form of copper at a pH of 7 in freshwaters (Leckie and Davis 1979), andis suspected to be relatively non-toxic (O'Donnel et al. 1985). At pH below 5, almost all copper is found inthe Cu2+ form (Leckie and James 1975). Model results from the 1:1000 treatment estimate that if theadsorption reactions between copper and iron oxides would have gone to completion, there would have beenonly 0.3 µg•L-1 toxic forms of copper. If no adsorption reactions would have taken place, roughly 30 µg•L -1of toxic copper forms would have been present. Based on these estimates, copper is most likely the elementthat contributed the largest toxic effect in the mesocosm study. In soft water, invertebrates and coho salmonhave shown ill side-effects from copper concentrations ranging from 0.0007 to 0.020 mg•L -1 (Spear and Pierce1979). The relatively short retention times in the experimental troughs probably caused some of the copperto remain in its most toxic ionic form. Miller and MacKay (1980) found that the 15 day copper LC 50 forrainbow trout was 0.048 mg•L -1 . The alkalinity of their test solution was 28 mg•L -1 as CaCO3 and thehardness was 49 mg•L-1 as CaCO3 . The 96 hour (4 day) LC50 copper concentration obtained in this studywas 0.028 mg•L-1 which is lower than Miller and MacKays' 15 day LC50 concentration. However, in thisexperiment, the pH that would have accompanied the addition of 0.028 mg•L -1 Cu would have beenapproximately 6.5 which is lower than the pH of 73 used in the study conducted by Miller and MacKay (1980).Chakoumakos et al. (1979) exposed cutthroat trout (Oncorhynchus clarki) to various copperconcentrations under different combinations of alkalinity, hardness and pH regimes. With an alkalinity of 23mg•L-1 CaCO3, hardness of 74 mg•L-1 CaCO3 and a pH of 7.6, they found that the 96 hour copper LC50 was0.0444 mg•L-1 . Although the alkalinity was lower in their experiments, the high pH could have easilycompensated for the greater resistance to copper.604.33 Aluminum Aluminum concentrations varied from under 0.010 mg•L -1 in the control troughs to over 15 mg•L -1at the highest treatment level. At pH 8 to 9, aluminum precipitates can be activated to aluminate ions whichare toxic to fish (Freeman and Everhart 1971). Concentrations of aluminum below 0.200 mg•L-1 have beenfound to be toxic to aquatic organisms in oligotrophic lakes (Havas and Jaworski 1986). Nevelle (1985) foundrainbow trout 96 hour LC75 concentrations near 0.75 mg Al•L -1 at pH levels from 4.0 to 6.1. Aluminumconcentrations in the 1:1,000 treatment troughs averaged at 0.346 mg•L -1 , and this may have contributed tothe toxicity of the ARD.4.34 CadmiumThe concentrations of cadmium ranged from less than 0.0001 mg•L -1 in the control troughs to 0.017mg•L-1 at the highest treatment level. Martin et al. (1980) state that the analytical characteristics for cadmiumare very good, and contamination is not normally a problem.The chemistry of cadmium is similar to that of zinc. Cadmium is commonly encountered with a formaloxidation state of +2 (Pagenkopf 1978), and is normally found in either ionic or aquated form (Cd0H + ) issoft waters (Weber and Poselt 1974). Cadmium binds weakly to organics, and may remain relatively mobilein the biosphere (Nriagu 1980). Cadmium is not an essential element, and this also contributes to its toxicity.Cadmium is one of the more toxic trace metals, and rainbow trout fry 96 hour LC 50 concentrationshave been reported to be as low as 0.0014 mg•L -1 (Alabaster and Lloyd 1984). Calamari et al. (1980)determined that the cadmium LC50 was 0.010 mg•L-1 for rainbow trout reared in water with a hardness of 20mg CaCO3•L-1 . Cadmium by itself was potentially acutely toxic only at the highest treatment level where theconcentration averaged 0.017 mg•L -1 , but no fish were exposed to that treatment. However, chronic symptomsin rainbow trout have been recorded at concentrations as low as 0.0009 mg•L -1 (Knittel 1980). Thus, it is notlikely that cadmium concentrations alone would have caused 96 hour acute toxicity because fish were exposedto treatment which achieved average cadmium concentrations 0.00024 mg•L-1.614.35 Iron and ManganeseThe fmal two elements that will be included in this discussion include iron and manganese. Theseelements will be discussed together because they are known to play crucial roles in regulating trace metals infresh waters (Florence and Batley 1980). Nordstrom et al. (1979) found that the addition of ARD to streamwater led to the formation of significant quantities of iron and manganese oxides and hydroxides. These metaloxides and hydroxides produced significant quantities of excellent surfaces for other metals to bind to inadsorption and precipitation events (Navas and Jaworski 1986; Spear and Pierce 1979; Spear 1981). Due totheir large negatively charged surface area, their multidentate nature, and their low solubility, hydrous oxidesof manganese and iron are efficient scavengers of heavy metals at normal pH (i.e. 6-8) in aerobic aquaticenvironments (Singh and Subramanian 1983).Iron is common in the earth's crust, but natural waters contain only minor amounts due to limitedsolubility. The Fe3+ oxidation state is predominant in oxygenated waters near neutral pH and readilyhydrolyzes causing iron to precipitate as hydrous ferric oxide (Pagenkopf 1978). Iron is of low toxicity and over7 mg•L4 were required to be lethal to Daphnia in 48 hr EC50 studies (Khangarot and Ray 1989).Manganese is also thought to be relatively non-toxic and over 8 mg•L-1 were required to cause toxiceffects to Daphnia (Khangarot and Ray 1989). However, Nix and Ingols (1981) attributed increased rainbowtrout mortality in a hatchery to a manganese concentration of 0.500 mg•L -1 . In a literature review onmanganese, Hanna and Speyer (1988) found that rainbow trout are one of the most sensitive organisms tomanganese toxicity. The tetravalent state of manganese forms an insoluble oxide in natural waters andprecipitates or is colloidally dispersed due to its very low solubility.Both manganese and iron reached potentially acutely toxic concentrations in the 1:25 treatment level.No fish were exposed to this treatment, so it is unlikely that manganese and/or iron played a significant rolein the toxicity of the test solutions. These two elements have very low stability constants with organic matter(LaZerte and Burling 1990; Koenings 1976), so organic complexation is not generally thought to be important.62The ARD which these experimental fish were exposed to contained significant quantities of adsorbingagents which may have helped to reduce the toxicity of the metals found in solution. Specifically,concentrations of iron alone would have been sufficient to adsorb almost all of the copper if the model resultsare even close to being realistic. Other complexes formed with organic matter, manganese oxides, aluminumoxides, and silica oxides would have also contributed to the adsorptive and complexing capabilities of the testsolutions.4.40 Tissue Metals and MetallothioneinsData from the quality assurance program for tissue analyses are listed in Table 15. All values obtainedin this study encompass the N.B.S. certified standard values. The average recovery for zinc was 87 percent,copper was 106 percent and cadmium was 109 percent. The error bounds of all values overlap so it wasconcluded that the tissue metal concentrations determined in this experiment were not significantly differentfrom the certified tissue concentrations.Table 15. National bureau of standards certified oyster tissue metal concentrations and values obtained in thisstudy. All values are in µg•g -1 dry weight t 1 standard deviation.Element N.B.S. Value Study ValueZn 852± 14 742 ± 99Cu 63 ± 3.5 67 ± 1.5Cd 3.5 ± 0.4 3.8 i 0.7After analyzing a series of plots, a liver sample from the experimental fish group was omitted fromthe analyses. This one sample explained a significant amount of the error in the correlation and regressionmodels. For example, by removing it from the hepatic metallothionein versus liver copper correlation, the Fvalue decreased from 0.184 to 0.005, and the correlation coefficient (r 2) increased from 0.042 to 0.176 Similartrends were noted for the zinc and cadmium regressions, so this liver sample was omitted from all statisticalanalyses.63Summarized results from the tissue digestions, metallothionein analyses and the changes in fish weightsare presented in Table 16. The complete data set is listed in Table 26 in Appendix 2. For the statisticalanalyses, all fish which died during the experiment were categorized in a separate group because the highconcentrations of metals in their tissues would have biased the results.4.41 A Priori Hypotheses In this section, hypotheses which were established at the onset of this experiment will be addressed.Hypotheses which were formulated after examining the data will be discussed in the a posteriori section of thediscussion.The first hypothesis was developed to test if hepatic metallothionein concentrations in the experimentalfish were related to the ARD concentrations in the trou • hs. A regression model was used to test thishypothesis with the ARD concentration as the independent variable and hepatic metallothionein concentrationas the dependent variable. The correlation coefficient (r2) was equal to 0.0006 and the slope was insignificant(F = 0.90). Thus, the first hypothesis that increasing ARD concentrations would cause an increase inmetallothionein concentrations in the surviving fish was rejected.A one way analysis of variance (ANOVA) was also run to see if significant hepatic metallothioneinconcentrations existed between the various groups of fish. Livers from experimental mortalities were notsubmitted because the amount of liver degradation would have been unknown. No significant differences inhepatic metallothionein concentrations were found between the various groups of fish (F = 0.14).The lack of a relationship between these parameters can be possibly explained by the highconcentrations of adsorbing and possibly complexing materials in the test solutions. According to theMINTEQA2 model results, many of the potentially toxic metals would have been either adsorbed orcomplexed, thus rendering them non-toxic. Gill mucous would have also been capable of providing an extraline of defense against metal uptake via the gill. Once the pH in the test solutions was lowered to a point64where metals became toxic, the short exposure time may have prevented significant quantities ofmetallothioneins from being induced.In the studies conducted by Roch et al., the metal of concern was zinc. Zinc does not bind stronglyto organics, and in an ultra-oligotrophic lake like Buttle Lake, a large portion of this metal would have beenionic (i.e. in the Zn2+ ) form. Concentrations of total zinc reached 140 µg•1: 1 at the most contaminated sitein Buttle Lake. This relatively high concentration of zinc, along with the long exposure times that the residentfish would have received, may explain why their studies found very significant trends.The second hypothesis was formulated to test if liver copper, zinc or cadmium concentrations werecorrelated with hepatic metallothionein concentrations in the various groups of fish. Correlation coefficients,number of samples, slope of the correlation equation and the probability of obtaining a slope greater than zeroby chance alone are listed in Table 17.By examining the F values, it was concluded that in three of the four fish stocks hepaticmetallothionein values were correlated with various metal levels in the liver (F < 0.05). All significantcorrelations had positive slopes, suggesting that increasing concentrations of liver metals were accompanied byincreasing concentrations of hepatic metallothionein.In both wild stocks, copper concentrations in the liver were correlated (r 2 > 0.60) with hepaticmetallothionein concentrations. Copper is known to be sequestered strongly by metallothioneins (Brady 1982),and it is therefore not surprising that a strong relationship existed between these two parameters. Theequilibrium state in which the wild fish were with respect to their environment may have also strengthenedthese correlations.65Table 16. Summarized fish tissue chemistry and weight change data from the experimental mesocosm and two wildfish stocks near the Equity Silver Mine. All means are listed as Ag.g-1 dry weight ± one standard deviation,except for change in mass which is expressed in grams.StockorTreatm'tInitialWeightgInitialLengthcmWeightchange9Metallo-thioneinpaol.eGillCuA9'9"MuscleCuA9'9-4LiverCuA9'91GillZnA9'9"MuscleZnA9'9"LiverZnA9'9"LiverCdA9'9"GillCdA9'9"Control . . . 1.3 t 0.4 6.2 t 2.1 3.2 t 0.5 33 t 6 372 t 50 54 t 6 395 t 235 5 t 2 1.8 ± 0.4Control 16.3 t 3.7 11.8 ± 0.8 1.2 ± 3.0 2.0 t 0.6 6.4 t 1.7 3.8 t 0.4 35 ± 12 400 t 110 60 t 9 1045 t 650 14 t 5 0.7 t 0.6Control 16.0 t 3.1 11.5^t^1.1 2.5 t 2.0 1.5 t 0.4 8.8 t 4.8 2.9 t 0.9 28 t 10 516 t 239 84 t 44 435 t 83 8 t 4 1.3 t 0.51:25,000 . . . 1.6 t 0.9 11.3 t 2.8 2.8 t 0.5 46 t 38 483 t 115 75 t 10 1148 t 909 14 t 11 1.5 t 0.71:25,000 18.8 t 1.9 12.4 t^1.1 0.0 t 2.0 1.6 t 0.4 4.8 t 2.2 2.9 t 0.7 40 ± 15 404 t 111 62 t 13 732 t 363 9 t 3 1.4 t^11:10,000 19.9 t 3.3 12.4 t 0.3 1.2 ± 3.0 1.4 t 0.3 7.8 t 2.2 3.7 t^1.1 29 t 16 356 t 95 57 t 5 303 t 120 3 t 2 2.1^t 0.51:10,000 16.3 ± 3.5 11.5 t^1.2 1.2 i^1.0 2.1 t 0.5 6.2 t 1.6 3.2 t 0.8 32 t 13 396 t 88 65 t 12 434 t 118 2 t 1 1.3 ± 0.21:2,500 15.4 * 2.8 11.3 t 0.5 2.2 t 1.0 1.6 t 0.5 7.0 t 3.2 3.0 t 0.5 42 t 2 444 t 171 69 t 12 815 t 158 12 t 1 1.3 t^11:2,500 19.3 t 3.7 12.2 t 0.8 2.6 1.2 8.8 2.6 37.2 1^362 37 403 3 1.9Morts . . 1.0 t^1.0 3.3 33.2 t 35 4.3 t 1.9 49 t 20 403 ± 175 72 t 39 412 ± 408 6 t 3 2.2 t 0.7Goosly . . . 1.6 t 0.7 4.7 t 1.5 2.5 t 0.7 22 t 17 162 t 48 38 t 11 16 t 6 4 t 2 1.7 t 0.8Lu . . . 1.3 t 0.8 6.5 t 1.3 3.0 ± 0.7 18 ± 10 229 t 50 31 t 2 11 t 4 3 t 1 0.9 t 0.6Abbots. . . . 1.7 t 0.8 5.0 t^1.1 3.5 ± 0.6 51 t 29 373 t 58 49 t 11 137 t 249 4 t 2 1.6 t 0.366Table 17. Regression statistics for total liver metal concentrations versus hepatic metallothionein concentrationsfor the various stocks of fish. * denotes that the correlation was significant at the «= 0.05 level.Stock Element n r2 I^Slope P > FExperiment Zinc 43 0.180 348 0.005*Experiment Copper 43 0.176 8.9 0.005*Experiment Cadmium 43 0.130 3.4 0.174Abbotsford Zinc 8 0.235 -157 0.186Abbotsford Copper 8 0.020 -5.2 0.718Abbotsford Cadmium 8 0.061 -0.7 0.523Goosly Lake Zinc 9 0.384 5.2 0.056Goosly Lake Copper 9 0.612 17.5 0.007*Goosly Lake Cadmium 9 0.032 0.3 0.621Lu Lake Zinc 8 0.065 -1.4 0.509Lu Lake Copper 8 0.663 1.5 0.007*Lu Lake Cadmium 8 0.706 1.3 0.005*The Abbotsford stock was the only group of fish in which none of the liver metals were significantlyrelated to metallothionein concentrations. The experimental fish had significant relationships between hepaticmetallothioneins and liver zinc and copper concentrations. However, the low r2 values (i.e. all less than 0.25)show that only a small fraction of the variation could be explained by the suggested relationship This supportsthe theory that the hepatic metallothionein concentrations are at best, only loosely related to total liver metalconcentrations in the experimental stocks.The relationships between hepatic metallothioneins and liver metal concentrations were also analyzedby summing the moles of copper, zinc and cadmium and correlating this sum with the moles ofmetallothioneins found in the fish livers. The slopes of the significant regressions represent the number ofmoles of metal that were bound to each mole of metallothionein. Results from these regression analyses arepresented in Table 18.67Table 18. Correlation results for analysis between E moles of copper, zinc and cadmium (iimol•g -1 dry liver)versus hepatic metallothioneins (iLmol.g -1 dry liver). * denotes a significant relationship at «= 0.05.Stock n r2 Slope P > FAbbotsford 9 0.209 -0.0838 0.2158Goosly Lake 10 0.656 1.865 0.0045*Lu Lake 9 0.489 3.158 0.0360*Experimental 43 0.213 0.0337 0.0039*Controls 17 0.501 0.0527 0.0015*1:25000 10 0.091 0.2106 0.39751:10000 12 0.457 0.1683 0.0158*1:2500 4 0.287 0.6383 0.4640Table 18 shows that the regression slopes decrease in the following manner: Lu Lake stock > GooslyLake stock > experimental stock. This implies that fish which have low concentrations of metals bound tometallothionein would by default, have higher concentrations of metal bound to the high molecular weightprotein pool. Although the spill-over hypothesis put forth by Winge et al. (1973) is still under considerabledebate, Brown (1977) and Brown and Parsons (1978) found a strong relationship between the incidence oftumours in flounders (Parophys vetulus) and metals in the high molecular weight protein pool. McCarter etal. (1982) also found that the concentrations of metals which bound to the high molecular weight protein poolwas typically a good indication of toxicity. If this relationship existed for all stocks analyzed, it can behypothesized that the Lu Lake stock was the healthiest, the Goosly stock were of intermediate health and theexperimental fish were in poor condition. Again, low ? values point to the inconclusive nature of theseanalyses. Inherent differences between stocks of fish or differences in their environments may have also beenresponsible for the different metal to metallothionein ratios.The third hypothesis was designed to test if muscle metal concentrations were correlated withconcentrations of elements in the experimental troughs. The wild fish stocks were not included in theseanalyses due to the lacking water chemistry data. Simple regression models were run with the respectiveelement concentration in solution as the independent variable and the tissue element concentration as the68dependent variable. Separate models were run for both the mortalities and the fish which survived theexperiment. The results are listed in Table 19.Table 19. Regression statistics for muscle metals versus metal concentrations in solution.Stock Element n I^r2 Slope P > FSurvivors Zinc 44 0.062 -0.993 0.102Survivors Copper 44 0.006 0.491 0.623Mortalities Zinc 16 0.109 0.107 0.385Mortalities Copper 16 0.011 -0.853 0.788Zinc and copper concentrations found in both the survivors and the mortalities had no significantrelationships with the metal concentrations in the test solution. Zinc was almost significant at the a = 0.10level.The insignificant correlations between fish muscle metal concentrations and test water metalconcentrations is in agreement with several other studies. Kito et al. (1982) exposed carp (Cyprinus carpio)to water borne cadmium and zinc. They found that insignificant amounts of metals accumulated in the muscletissue of fish after 30 days of exposure. Similar results were found by Lauren and McDonald (1987) whoexposed rainbow trout to solutions containing 0.055 mg Cu•L-1 . They found that the distribution of metalsin several organs changed over time with only a very slight increase in muscle copper concentration after 28days of exposure. Thus, for relatively short-term experiments (i.e. less than 30 days), the muscle tissue seemslike a poor tissue to assay for metal uptake studies.The fmal a priori hypothesis was designed to see if relationships existed between metal concentrationson/in the gill tissue and the metal concentrations in the test solution. Results from the statistical analyses forboth the survivors and the mortalities from the experiment are listed in Table 20. All water cadmiumconcentrations below, the LOD were omitted from the analyses.69In the surviving fish, both the zinc and copper concentrations in the gill were not correlated with thetest solution zinc and copper concentrations. The gill cadmium concentrations were significantly (P = 0.011)correlated with the solution cadmium concentrations (r2 = 0.488), however, only two concentrations (i.e. 0.1and 02 ag•L4) of cadmium were detectable in treatments in which all fish survived. By regressing gillcadmium concentrations against two points, an unrealistically high regression coefficients was bound to occur.For this reason, this relationship is suspect.Part and Lock (1983) found that cadmium uptake across rainbow trout gills increased over 100 foldwhen the concentrations of cadmium in solution increased from 0.0056 ug•L-1 to 0.056 µg•L-1 . They felt thatthe uptake of metals via the gill depends not only the concentration of free ions, but also the strength of themetal-ligand binding capability. Cadmium is known to be very labile (Forstner 1987), and uptake of this metaltherefore can occur at very low concentrations.Table 20. Regression statistics for gill metals versus metal concentrations in solution. * denotes a significantrelationship at oc =0.05.Stock Element n r2 Slope I^P > FSurvivors Zinc 44 0.073 -4.42 0.075Survivors Copper 44 < 0.00 -0.01 0.947Survivors Cadmium 12 0.448 6.93 0.011*Mortalities Zinc 16 0.113 4.66 0.377Mortalities Copper 16 0.945 2.23 < 0.000*Mortalities Cadmium 6 0.202 6.32 0.371In the mortalities, copper concentrations in/on the gill were strongly correlated (r 2 = 0.945) with thesolution copper concentrations, whereas zinc and cadmium levels were not significantly correlated with the testsolution metal concentrations (Table 20). The high correlation between gill copper concentrations and the testsolution copper concentrations supports the hypothesis that the gill mucous effectively bound copper in thesurviving fish, and it may have been sloughed off during coughing events. Copper is known to bind very70strongly with organic matter (Irving and Williams 1983), and the very high correlation between water copperand gill copper concentrations in the mortalities (r 2 = 0.945) may be explained by the inability of themortalities to cough and thus slough copper-laden mucous from the gill. No relationship existed betweensurviving fish gill copper concentrations and the test solution concentrations. It can be hypothesized that thecontinuous sloughing of mucus from the gill to prevent epithelial damage (Mallat 1985) kept the copperconcentrations on the gill relatively low.4.42 A Posteriori HypothesesOnce the data was collected and analyzed, several other notable trends became evident. Some of themore significant trends will now be discussed.The first a posteriori hypothesis investigated related to the mass change in the experimental fish. Masschange and hence food ingestion can be used as an index of fish health. The instantaneous daily growth rate(g•day-1) was calculated for the groups of fish which had their mass obtained prior to and after the experiment(Table 16). There were no significant differences in growth between the various groups (F = 0.75), and thismay have been due to the large degree of variation within groups. The average increase in mass during the23 day experiment was roughly 15 grams, and this converted into an instantaneous daily growth rate of 0.065g-day-1 . Buckley et al. (1982) exposed coho salmon (Oncorhynchus kisutch) to copper concentrations of 70and 140 µg•L-1 . These fish ceased to feed immediately after being exposed to the copper, and regainednormal growth rates after 20 days. The instantaneous growth rate of the control fish in their experiment was0.15 g•d-1 . This is roughly twice the growth rate which the fish experienced in this experiment, but an initialperiod where the fish may have ceased to feed may be responsible for the low growth rates.The second trend that was evident upon examination of the tissue data was the very high zincconcentrations found in the fish which were exposed to the experimental troughs. Fish which were exposedto the experiment had muscle, gill and liver zinc concentrations that were much higher than fish which weresacrificed at the Abbotsford hatchery.71Table 21 reveals the results from a Duncan's multiple range test on the liver zinc concentrations foundin the various fish stocks. The experimental fish had liver zinc concentrations over an order of magnitudehi • her than the wild fish stocks. The fish which were exposed to the experimental troughs had liver zincconcentrations at least twice as high as the fish which were sacrificed at the Abbotsford hatchery. TheAbbotsford stock had roil . hly 130 gg Zn-g-1 in their livers, and the experimental stock had from 300 to 1044gg Zirg-1 liver tissue. This change in liver zinc concentration took place within 23 days.Gill zinc concentrations in the experimental and the Abbotsford fish were roughly 100 gg•g4 higherthan concentrations found in the wild fish gills (Table 22). The differences between the experimental groupsof fish and the Abbotsford fish were not significant, but the experimental fish had at least more zinc in theirgill tissues than the Abbotsford fish still existed.Table 21.^Ranked means and significantly different groups for liver zincconcentrations.^(*) significantly different at the 0.05 level.LGA 1 C 1 M1 C 1 1 1Cuob:o:o:o:::oobln 2 r 1 n2 2 2 nso 0 t 5 t 0 t 5 5 5 tlt 0 r 0^Or 0 0 0 rys 0 o 0^0 o 0 0 0 o0 1^0 1 0^0 1Mean S.D n Group11.0000 4.5 9 Lu15.9231 5.7 13 Goosly136.6667 249.5 9 Abbots302.8333 120.1 6 1:10000394.6000 235.1 5 Control *403.0000 . 1 1:2500412.0000 407.6 9 Mort *^*434.0000 117.7 6 1:10000 *^*435.5000 83.4 6 Control *^*731.6667 362.7 6 1:25000 *^* *815.3333 158.2 3 1:2500 *^*^*945.0000 909.6 4 1:25000 *^*^* *^*^* * *1044.6667 649.7 6 Control * * * * *^* * *72Table 22.^Ranked means and significantly different groups for gill zincconcentrations.^(*) significantly different at the 0.05 level.GL 1 1 CA 1 CM 1 1 1 Cou::ob:oo:::oo^1 2 nblnr 2 2 2 ns^0 5 toOtt 5 5 5 t1^0 OrtOr^0 0 0 ry^0 0 osOo^0 0 0 o0^1^0 1^0^0 1Mean S.D n Group161.8481 47.6 16 Goosly228.6725 49.5 8 Lu355.8317 94.7 6 1:10000 * *362.4000 . 1 1:2500372.8260 50.2 5 Control * *373.8463 57.7 8 Abbots * *395.7933 87.7 6 1:10000 * *400.0450 110.5 6 Control * *402.8056 174.7 9 Mort * *404.1367 111.3 6 1:25000 * *444.2667 170.6 3 1:2500 * *482.6700 115.6 5 1:25000 * *515.7567 239.6 6 Control * * *A similar trend existed for the zinc concentrations found in the experimental fish muscle tissue. Theresults from a Duncan's multiple range test on the muscle zinc concentrations in the various groups of fish ispresented in Table 23. Although the differences in concentrations were not as extreme as those found in thelivers, the same general trend existed. The Abbotsford stock and the two wild fish stocks had muscle zincconcentrations that contained from 5 to 46 li•e less copper per gram dry weight.The first hypothesis was that the zinc found in the liver was obtained from the media in which the fishwere reared. Regressions were run using water zinc concentrations as the independent variable and liver zincconcentrations as the dependent variable. The relationship was found to be insignificant (F = 0.15), and thecorrelation coefficient was found to be low (r2 = 0.05). Based on this lacking relationship, it is likely that zincfound in the experimental media did not play a major role in the bioaccumulation of zinc in the experimentalfish livers. The relatively low concentrations (8.5 Ag•L -1 to 31 ;L0:1 ), along with the short exposure periodof just over 3 weeks probably limited the amount of uptake via the water.73Table 23.^Ranked meansconcentrations.^(*)and significantly different groups for muscle zincsignificantly different at the 0.050 level.LlGAC1C111M1Cu:obo:o:::o:o2 obnln212 r 2 n5 sot 0 t 505 t 5 tOltrOr 000^OrOyso 0 o 000^Oa1 0 1 0 0^0 1Mean S.D n Group30.7733 2.4 9 Lu37.2000 . 1 1:250038.3900 11.2 14 Goosly48.8433 11.2 9 Abbots53.8380 6.1 5 Control56.9483 5.0 6 1:10000 *60.4867 8.6 6 Control *^*62.5350 13.2 6 1:25000 *^*64.6650 12.2 6 1:10000 *^*69.0867 12.4 3 1:2500 *^*72.0433 38.8 9 Mort *^* *75.4500 10.2 5 1:25000 *^* *84.1133 44.5 6 Control *^* * * *The second hypothesis formulated was that uptake from the food which the fish were fed during theexperiment was significant. The food used at the experimental site contained 90 Age zinc and 12 ligecopper dry weight. Wekell et al. (1983) found that zinc concentrations in the gill and liver from rainbow troutincreased significantly when the fish were fed diets containing 90 ug Zn •g1 fish food or more. No toxiceffects were found even when the food contained 1700 Age zinc. This is in agreement with Buell (1991)who also found that increased metal loads in chum salmon (Oncorhynchus keta) which were exposed to minetailings was primarily due to the food which the fish ingested. This study, along with others (Patrick and Loutit1978; Buell 1991) have found that the uptake of metals can be primarily via food ingestion. In aquaticecosystems, zooplankton and phytoplankton may provide binding sites for certain elements. Once these foodparticles are ingested by fish, elements may then be passed directly to the planktivores. It is for this reasonsthat these organisms have been recommended as biological indicator organisms (Luoma 1983). Fish whichhave very long digestive tracts have also been found to bioaccumulate more elements from food due to thelonger retention time and the more efficient digestion of food (Bendell-Young et al. 1986).74The fmal a posteriori hypothesis that was formulated relates to the cadmium concentrations that wereanalyzed in the gill tissues of the various groups of fish. The results from a Duncan's multiple range test onthe mean gill cadmium concentrations are shown in Table 24. No strong trends were evident in theexperimental groups, but significantly different concentrations of gill cadmium were found between the Gooslylake stock and the Lu lake stock. The Lu Lake stock had average gill cadmium concentrations of 0.95 ± 0.6while the Goosly Lake stock had gill cadmium concentrations of 1.7 t 0.8 ;4•' 1 . Although thesamples sizes were small (8 for Lu and 16 for Goosly), this trend may be worthy of further study. The loadingrates of cadmium to Goosly Lake are roughly 3 kg per year (Patterson 1986), and this may explain thedifferences in the two stocks.In a study conducted on a stream ecosystem polluted by ARD in Montana, Moore et al. (1991) foundthat the mobilization of elements downstream from the pollution outfall were cadmium zinc > copper >arsenic nickel. The order of re-mobilization was inversely related to the element's tendency to form covalentbonds. The area they studied contained several beaver dams and marshes indicating a high organic content.The Equity Silver mine discharges its effluent through a marsh area which then enters Bessemer creek. Thiscreek then flows into Goosly lake. Although the sample sizes were relatively small, the significantly differentconcentrations of cadmium in the gills of the two fish stocks may indicate that the Goosly lake fish stock isbeing exposed to higher concentrations of cadmium.75Table 24. Ranked means and significantly different groups for gill cadmium concentrations. (*) Denotessignificantly different groups at the 0.05 level.CLC 1 1 1 1 AGC 1 1Mouo::::boo::on n2 1 2 2 bon 2 lrt t 5 0 5 5 ost 5 Otr r 0 0 0 0 t 1 r 0 0o o 0 0 0 0 syo 0 01 1 0 0 0 1 0Mean S.D n Group.6917 0.6 6 Control.9538 0.6 8 Lu1.3017 0.5 6 Control1.3300 1.5 3 1:25001.3540 0.2 5 1:100001.4400 1.0 6 1:250001.4700 0.8 5 1:250001.6288 0.3 8 Abbots *1.6963 0.8 16 Goosly * *1.8040 0.4 5 Control *1.9100 . 1 1:25002.0750 0.5 6 1:10000 * *2.1671 0.7 7 Mort * *76CHAPTER 55.0 CONCLUSIONS AND RECOMMENDATIONS5.10 ConclusionsA relationship between hepatic metallothionein concentrations and the various metals found in solutionwas non-existent. The surviving fish, although in some instances exposed to rather high concentrations ofcopper, zinc, aluminum and cadmium did not respond to the metals in solution by increasing metallothioneinconcentrations. This would lead to the conclusion that the metals were not highly bioavailable. Because theARD had high concentrations of adsorbing agents such as iron, manganese, aluminum and silica, the toxicityof metals may have been reduced due to the formation of non-toxic compounds. Complexation events withorganic matter may have also reduced the toxicity of the metals to the fish. Finally, the secretion and metal-binding characteristics of mucous from the fish may have also reduced the possibility of the metals enteringthe fish's body and inducing metallothioneins.The amount of metal bound to hepatic metallothioneins versus the summed moles of copper, zinc andcadmium in the liver may be a reasonable indicator of fish health. By determining the number of moles ofcopper, zinc and cadmium in the homogenized liver and regressing the sum of these values with the numberof moles of metallothionein, the amount of metals bound to high molecular weight proteins can be determinedindirectly. Relatively small samples sizes are required for these analyses, and this technique could prove to beuseful when studying fish that are small in size. However, this technique would have to be thoroughly testedin an experimental design before it could be implemented as a method of metal toxicity analysis.Based on the results from this experiment, it can be concluded that fish muscle tissue is not a goodindicator tissue for metal uptake. Zinc was most likely bio-concentrated from the food in this study, andrelatively small increases in muscle zinc concentrations were noted. The liver and gill both showed muchgreater increases in zinc concentrations, but there were no relationships with the test solution metal77concentrations. It appears that the gill and liver both responded to increases in metal uptake, and these tissuesshould therefore be used to determine the level of bioaccumulation.The utility of using gill tissue as a metal pollution monitoring tool seems to depend on the type ofmetal that the fish is being exposed to. Water hardness, pH and complexation capacity of the water may alsoinfluence the amount of metals that bind to mucous and/or gill tissue. Fish which died during the experimenthad gill copper concentrations which were strongly related to water copper concentrations. This supports thehypothesis that copper binds strongly with organic matter on the gill. Relationships between gill zinc and gillcadmium concentrations with the respective solution metal concentrations were very weak or non-existent.Therefore, the use of the gill as a metal pollution indicator tissue should be used with great caution.The differences in gill cadmium concentrations between the two wild stocks of fish warrants furtherstudy. Cadmium is a non-essential element, it is one of the most mobile divalent cations and it is known tobe taken up primarily via the gill. Due to these characteristics, this metal could be used as a potential indicatorof cadmium exposure between various fish stocks. Again, further studies would be necessary to confirm thishypothesis.The 96 hour rainbow trout LC50 ARD concentration was calculated to occur when 1 part ARD wouldbe added to 1586 parts stream water. At this dilution, the concentrations of copper, zinc, aluminum andcadmium would be 0.028, 0.027, 0.277 and 0.0003 mg•L-1 respectively. The pH would be near 6.8 at thatdilution ratio. The ARD dilution rate at which roughly half of the fish died during the 23 day experiment wascalculated to be 1 part ARD to 5714 parts stream water. That rate of ARD addition would correspond tocopper, zinc, aluminum and cadmium concentrations of 0.010, 0.014, 0.113 and 0.0001 mg•1: 1 respectively.At the latter dilution ratio, the stream pH would not have been reduced significantly below neutrality. Thisdata could be used to establish guidelines for the receiving waters near the Equity Silver mine.Interactions between pH and adsorptions processes played a major role in controlling the toxicity ofheavy metals to fish in these experiments. As the pH was reduced, metals such as copper, aluminum, zinc andcadmium would have remained in a simple aquated or ionic form for longer periods of time. A reduction inpH of 0.2 units, along with increases in heavy metal concentrations caused the ARD to be acutely toxic. It is78impossible to positively identify the metals and/or species of metals that caused the ARD solutions to be toxic.Metal toxicity is typically thought to be additive, so concentrations of zinc, cadmium and aluminum may haveall contributed to the toxicity of the ARD.Because the pH in these experiments played such a major role in controlling the toxicity of metals tofish, it might be feasible to place in situ pH probes in streams which receive ARD. By monitoring this singlevariable, it would be possible to determine if the stream is basic or acidic, and thus relate it to adsorption andtoxicity. If nothing else, it might provide a first warning signal for increasing acidification and metal toxicity.5.20 RecommendationsThe results from this study have provided the information necessary to generate a series of testablehypotheses. One of the ways to reduce the chances of uncontrollable problems is to conduct a series ofexperiments in the laboratory. Under controlled conditions, single variables could be isolated and controlledwhich would possibly generate clearer results.The first, and possibly most important study that should be conducted would determine the effects thatcomplexing and/or adsorbing agents have on metal toxicity. Simple bioassays using various concentrations ofiron oxides coupled with the same concentration of toxicant would highlight the importance of the adsorbingagents. If a well-defined media were used, geochemical models could be used to calculate metal speciation.Analytical techniques could then be used in an attempt to verify the model results. Concentrations of metalsadsorbed onto the precipitated oxides could also be conducted in order to obtain a mass balance for the metals.The importance of reaction kinetics in toxicity experiments cannot be over-emphasized. Studies shouldbe conducted to determine the reaction rates that are taking place while the experiments are being conducted.For instance, toxicity studies could be conducted at the mixing zone where metals are in their most toxic form.Studies could also be conducted once the reaction kinetics have been allowed to go to completion. Kineticinformation obtained in this fashion could then be incorporated into a geochemical model which then couldbe used to estimate "downstream effects" of metals in natural systems.79Results from this study also support the hypothesis that food ingestion is an important pathway formetal uptake. Various metals could be used at various concentrations to see how these variables affectbioaccumulation. A potential problem with those types of studies is that it is very difficult to determine exactlyhow much food each fish ingests. Force-feeding with a syringe or feeding only limited quantities of food toensure that the food is being ingested might be necessary. Studies conducted in that manner could beconducted over several months and a time-series of metal uptake and partitioning in various tissues could befollowed. This type of study could also be conducted using isotopes, enabling very low levels to be accuratelymeasured.The final recommendation would be to conduct a series of studies designed to induce metallothioneinsusing solutions that are representative of Buttle Lake and Foxy Creek. Fish food metal concentrations shouldbe normalized, and various concentrations and types of metals could be used. Lbile metals such and zinc andcadmium should induce metallothioneins readily, whereas metals such as copper would probably bind tocomplexing agents prior to entering the fish's body.806.0 LITERATURE CITEDAhrland, S., J. Chatt, and N.R. Davies. 1958. The relative affinities of ligand atoms for acceptor moleculesand ions. Q. Rev. Chem. Soc., 12:265-276.Alabaster, J.S. and R. Lloyd. 1984. Water quality criteria for freshwater fish. Butterworths, Toronto. 361 pp.Allison, J.D., D.S. Brown and K.J. Novo-Gradac. 1990. MinteqA2/ProdefA2, A geochemical assessmentmodel for environmental systems. Environmental Researc Laboratory, U.S. E.PA., Athens, Georgia.U.S.A. 105 pp.Anderson, P.R. and M.M. Benjamin. 1990. Modelling adsorption in aluminum-iron binary oxide suspensions.Environ. Sci. Technol. 24: 1586-1592.Andrew, R.W., K.E. Biesinger and G.E. Glass. 1977. Effects of inorganic complexing on the toxicity of copperto Daphnia magna. Water Res. 11: 309-315.A.P.HA. 1985. Standard Methods for the Examination of Water and Wastewater. (16th Edition).Bendell-Young, L.I., H.H. Harvey and J.E. Young. 1986. Accumulation of cadmium by white suckers(Catostomus commersoni) in relation to fish growth and lake acidification. Can. J. Fish. Aquat. Sci. 43:806-811.Bentley, P.J. 1962. Permeability of the skin of the cyclostome Lampetra fluviatilis to water and electrolytes.Comp. Biochem. Physiol. 6: 95-97.Bennett, J.W., J.R. Harries and A.I.M. Ritchie. 1988. Rehabilitation of waste rock dumps at the Rum Junglemine spoil. Proceedings of the Mine Drainage and Surface Mine Reclamation Conference. Vol. 1: MineWater and Mine Waste. April 19-21, Pittsburgh, PA. pp. 104-108.Block, M. and P. Part. 1986. Increased availability of cadmium to perfused rainbow trout (Salmo gairdneri,Rich.) gills in the presence of the complexing agents diethyl dithiocarbamate, ethyl xanthate andisoprophylxanthate. Aquat. Toxicol. 8: 295-302.Bond, C.E. 1979. Biology of Fishes. Saunders College Publishing, Philadelphia. 514 pp.Borgmann, U. and K.M. Ralph. 1983. Complexation and toxicity of copper and the free metal bioassaytechnique. Water Res. 17(11): 1697-1703.Brady, F.O. 1982. The physiological function of metallothionein. Trends Biochem. Sci. 7: 143-145.Bradley, R.W. and J.B. Sprague. 1985. The influence of pH, water hardness, and alkalinity on the acutelethality of zinc to rainbow trout (Salmo gairdneri). Can. J. Fish. Aquat. Sci. Vol. 42:731-735.Bradley, R.W., C. DuQuesnay, and J.B. Sprague. 1985. Acclimation of rainbow trout, Salmo gairdneriRichardson, to zinc: kinetics and mechanism of enhanced tolerance induction. J. Fish Biol. 27: 367-379.Bremner, I. 1987. Nutritional and physiological significance of metallothionein. In: Metallothionein II.Proceedings of the Second International Meeting on Metallothionein and Other Low Molecular WeightMetal-Binding Proteins. Zurich, August 21-24. pp. 81-107.81Brenzonik, P 1972. Chemical kinetics and dynamics in natural water systems. In: L.L. Ciaccio (Ed.). Waterand Water Pollution Handbook. Vol. 3. Chapter 16. Marcel Dekker Inc. New York.Brown, D.A. 1977. Increases of Cd and the Cd:Zn ratio in the high molecular weight protein pool fromapparently normal liver tumor-bearing flounders (Parophyrys vetulus). Mar. Biol. 44: 203-209.Brown, DA. and T.R. Parsons. 1978. Relationship between cytoplasmic distribution of mercury and toxiceffects to zooplankton and chum salmon (Oncorhynchus keta) exposed to mercury in a controlledecosystem. J. Fish. Res. Board Can. 35: 880-884.Browning, J.E. 1970. Freshening acid mine-waters. Chemical Engineering. 77: 40-42.Brezonik, P.L. 1972. Chemical kinetics and dynamics in natural water systems. In: L.L. Ciaccio (Ed.). Waterand Water Pollution Handbook. Vol. 3. pp. 832-913.Buckley, J.T., M. Roch, JA. McCarter, and CA. Rendell. 1982. Chronic exposure of coho salmon to sublethalconcentrations of copper - I. Effect on growth, on accumulation and distribution of copper, and on coppertolerance. Comp. Biochem. Physiol. 72C(1): 15-19.Buffle, J. 1988. Complexation Reactions in Aquatic Systems: an Analytical Approach. John Wiley and Sons,Toronto. 692 pp.Buffle, J., A. Tessier and W. Haerdi. 1984. Interpretation of trace metal complexation by aquatic organicmatter. In: C.J.M. Kramer and J.C. Duinker (Eds.). Complexation of Trace Metals in Natural Waters.Martinus Nijhoff/Dr. W. Junk Publishers. Boston. pp. 301-306.Buell, J.W. 1991. Consequences of episodic exposure of Chum Salmon eggs/alevins to expected A-J Minetailings pond concentrations of Ag, Cd, Cu, Pb and Zn. Proceedings of the Seventeenth Annual AquaticToxicity Workshop; November 5-7, 1990. Can. Tech. Rep. of Fish. and Aquatic Sci. 1774(1): 208-237.Calamari, D., R. Marchetti and G. Vailati. 1980. Influence of water hardness on cadmium toxicity to Salmogairdneri Richland. Water Res. 14(10): 141-146.Campbell, P.G.C. and P.M. Stokes. 1985. Acidification and toxicity of metals to aquatic biota. Can. J. Fish.Aquat. Sci. 42: 2034-2049.Cambell, P.G.C., A.G. Lewis, P.M. Chapman, AA. Crowder, W.K. Fletcher, B. Imber, S.N. Luoma, P.M.Stokes and M. Winfrey. 1988. Biologically available metals in sediments. National Research Council ofCanada No. 27694. 298 pp.C.B.R. International. 1990. Metallothionein analyses. Report prepared for Limnotek Research andDevelopment Inc. C.B.R. reference number CBR 289. 8 pp.Chakoumakos, C., R.C. Russo, and R.V. Thurston. 1979. Toxicity of copper to cutthroat trout (Salmo clarki)under different condition of alkalinity, pH, and hardness. Am. Chem. Soc. 13: 213-219.Chan, K.M., W.S. Davidson, and G.L. Fletcher. 1989. Metallothionein messenger RNA: potential molecularindicator of metal exposure. In: J.O. Nriagu and J.S.S. Lakshminarayana (Eds.). Aquatic Toxicology andWater Quality Management. Wiley Series in Advances in Environmental Science and Technology. JohnWiley and Sons, Toronto. pp. 89-109.82Conte, F.P. 1969. Salt Secretion. In: W.S. Hoar and D.J. Randall (Eds). Fish Physiology Vol. 1: Excretion,Ionic Regulation, and Metabolism. pp. 241-292.Cowan, C.E., J.M. Zachara and C.T. Resch. 1991. Cadmium adsorption on iron oxides in the presence ofalkaline-earth elements. Environ. Sci. Technol. 25(3): 437-446.Cusimano, R.F., D.F. Brakke and GA. Chapman. 1986. Effects of pH on the toxicities of cadmium, copper,and zinc to steelhead trout (Salmo gairdneri Can. J. Fish. Aquat. Sci. 43: 1497-1503.Davison, W. and C. Woof. 1984. A study of the cycling of manganese and other elements in a seasonallyanoxic lake, Rostherne Mere, U.K. Water Res. 18: 727-740.Deniseger, J., L.J. Erickson, A. Austin, M. Rock and M.J.R. Clark. 1990. The effects of decreasing heavymetal concentrations on the biota of Buttle Lake, Vancouver Island, British Columbia. Water Res. 24(4):403-416.Eichenberger, E. 1986. The interrelation between essentiality and toxicity of metals in the aquatic ecosystem.In: H.E. Sigel (Ed.). Metal Ions in Biological Systems. Vol. 20: Concepts on Metal Ion Toxicity. MarcelDekker, N.Y. pp. 67-100.Eichorn, G.L. 1974. Ecological Toxicity Research. Plenum Press, New York.Eisler, R. and G.R.J. Gardner 1973. Acute toxicity to an estuarine teleost of mixtures of cadmium, copper andzinc salts. J. Fish Biol. 5: 131-142.E.P.S. 1987. Environment Canada. Mine and Mill Wastewater Treatment, Environmental ProtectionDirectorate, Conservation and Protection, Report EPS 2/MM/3, Ottawa.E.P.S. 1988. Metals in Water. Methods manual. 8 pp.Errington, J. and K.D. Ferguson. 1987. Acid mine drainage in British Columbia today and tomorrow.Proceedings of the Eleventh Annual British Columbia Mine Reclamation Symposium. April 8-10,1987.Campbell River, B.C. pp. 7-30.Evans, G.W. 1976. Zinc absorption and transport. In: A. Prassad and D. Oberlease (Eds.). Trace elementsin human health and disease. Vol. I. Academic Press, New York. pp. 181-187.Faust, S.D. and O.M. Aly. 1981. Chemistry of Natural Waters. Ann Arbor Science Publishers, Inc. AnnArbor, Michigan. 400 pp.Fish, W. and F.M.M. Morel. 1983. Characterization of organic copper-complexing agents released by Daphniamagna. Can. J. Fish. Aquat. Sci. 40: 1270-1277.Florence, T.M. 1977. Trace metal species in fresh waters. Water Res. 11: 681-687.Florence, T.M. and G.E. Batley. 1980. Chemical speciation in natural waters. C.R.C. Critical Reviews inAnalytical Chemistry. pp. 219-296.Florence, T.M. 1980. Speciation of zinc in natural waters. In: J.O. Nriagu (Ed.) Zinc in the Environment.Part I: Ecological Cycling. John Wiley and Sons, Toronto. 453 pp.83Florence, T.M. 1982. Review of trace elements in waters. Talanta. 29: 345-364.Florence, T.M., B.G. Lumsden and J.J. Fardy. 1984. Algae as indicators of copper speciation. In: C.J.M.Kramer and J.C. Duinker (Eds.). Complexation of Trace Metals in Natural Waters. Martinus Nijhoff/Dr.W. Junk Publishers. Boston. pp. 411-418.Forstner, U. 1987. Metal speciation in solid wastes - factors affecting mobility. In: L. Landner (Ed.)Speciation of Metals in Water, Sediment and Soil Systems. pp. 13-41.Frausto, JJ.R da Silva and RJ.P. Williams. 1976. The uptake of elements by biological systems. Struct. Bond.29: 67-121.Freeman, RA., and Everhart, W.H. 1971. Toxicity of aluminum hydroxide complexes in neutral and basicmedia to rainbow trout. Trans. Am. Fish. Soc. 100(4): 644-658.Gardiner, W.C. 1969. Rates and Mechanisms of Chemical Reactions. Benjamin, Menlo Park, CA.Garrels, R.M. and M.E. Thompson. 1960. Oxidation of pyrite by iron sulfate solutions. Am. J. Sci. 258A:57-67.Garrels, R.M. and M.E. Thompson. 1962. A chemical model for seawater at 25 °C and one atmosphere totalpressure. Am. J. Sci. 260: 57-66.Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold Co.New York. 320 pp.Guy, R.D. and C.L. Chakerabarti. 1976. Studies of metal-organic interactions in model systems pertaining tonatural waters. Can. J. Chem. 54: 2600-2611.Handy, R.D. and F.B. Eddy. 1990. The influence of starvation on waterborne zinc accumulation by rainbowtrout, Salmo gairdneri, at the onset of episodic exposure in neutral soft water. Water Res. 24(4): 521-527.Handy, R.D., F.B. Eddy, and G. Romain. 1989. In vitro evidence for the ionoregulatory role of rainbow troutmucus in acid, acid/aluminum and zinc toxicity. J. Fish Biol. 35: 737-747.Hanna, J.E. and M. Speyer. 1988. Effects of manganese on aquatic biota. In: Stokes, P.M, P.G.C. Campbell,W.H. Schroeder, C. Trick, R.L. France, KJ. Puckett, B. Lazerte, M. Speyer, J.E. Hanna and J. Donaldson(Eds.). Manganese in the Canadian Environment. N.R.C.C. No. 26193. pp. 81-91.Hardy, R.W. and KD. Shearer. 1985. Effect of dietary calcium phosphate and zinc supplementation on wholebody zinc concentration of rainbow trout (Salmo gairdneri). Can. J. Fish. Aquat. Sci. 42: 181-184.Hardy, R.W., C.V. Sullivan, and A.M. Koziol. 1987. Absorption, body distribution, and excretion of dietaryzinc by rainbow trout (Salmo gairdneri). Fish Physiol. Biochem. 3(3): 133-143.Havas, M. and J.F. Jaworski. 1986. Aluminum in the Canadian Environment. National Research Council ofCanada No. 24759. 331 pp.Hering, J.G. and F.M.M. Morel. 1990. Kinetics of trace metal complexation: ligand-exchange reactions.Environ. Sci. Technol. 24: 242-252.84Herrmann, J 1987. Aluminum impact on freshwater invertebrates at low pH: a review. In: L. Landner (Ed.)Speciation of Metals in Water, Sediment and Soil Systems. pp. 157-178.Howarth, R.S. and J.B. Sprague. 1978. Copper lethality to rainbow trout in waters of various hardness andpH. Water Res. 12: 455-462.Hughes, G.M. 1972. Morphometrics of fish gills. Respir. Res. 14: 1-25.Hunt, D.T.E. 1987. Trace metal speciation and toxicity to aquatic organisms: a review. Water ResearchCentre, Medmenham, U.K. 51 pp.Imber, B., M.G. Robinson, and F. Pollehne. 1984. Complexation by diatom exudates in culture and in thefield. In: C.J.M. Kramer and J.C. Duinker (Eds.). Complexation of Trace Metals in Natural Waters.Martinus Nijhoff/Dr. W. Junk Publishers. Boston. pp. 429-440.Irving, H. and R.J.P. Williams. 1948. Order of stability of metal complexes. Nature 162: 746-747.Irving, H. and R.J.P. Williams. 1953. The stability of transition-metal complexes. J. Chem. Soc. pp. 3192-3210.Jardim, W.F. and H.E. Allen. 1984. Measurement of copper complexation by naturally occurring ligands. In:C.J.M. Kramer and J.C. Duinker (Eds.). Complexation of Trace Metals in Natural Waters. MartinusNijhoff/Dr. W. Junk Publishers. Boston. pp. 1-16.Jenne, EA. 1968. Controls of Mn, Fe, Co, Ni, Cu, and Zn concentrations in soil and water: the significantrole of hydrous Mn and Fe oxides. Adv. Chem. Ser. No. 73: 337-348.Jenne, EA. 1979. Chemical modelling - goals, problems, approaches, and priorities. In: EA. Jenne (Ed.).Chemical Modelling in Aqueous Systems. Speciation, Sorption, Solubility, and Kinetics. A.C.S. Symposium93: 3-24.Kagi, J.H.R. and B.L. Vallee. 1961. Metallothionein: a cadmium-and zinc-containing protein from equine renalcortex. J. Biol. Chem. 235: 3460-3468.Kagi, H.H.R. and M. Nordberg (Eds). 1979. Metallothionein. Proceedings of the 1st International Meetingon Metallothionein and Other Low Molecular Weight Metal-Binding Proteins. Boston: Birkhauser Verlag.Karin, M. 1985. Metallothioneins: Proteins in search of a function. Cell 41: 9-10.Kay, J., M.W. Brown, A. Cryer, J.F. deL.G. Solbe, D. Shurben, J.S. Garvery and D.G. Thomas. 1987.Metallothionein gene expression and cadmium toxicity in freshwater fish. Experientia Supplementum Vol.52. Metallothionein II. pp. 627-630.Khangarot, B.S. and P.K. Ray. 1989. Investigation of correlation between physiochemical properties of metalsand their toxicity to the water flea Daphnia magna Straus. Ecotox. Envir. Safety 18: 109-120.Kirschener, L.B. 1977. External charged layer and Na + regulation. In: Osmotic and Volume Regulation.Alfred Benzon Symposium XI. Academic Press, New York.Kito, H., T. Tazawa, Y. Ose, T. Sato and T. Ishikawa. 1982. Formation of metallothionein in fish. Comp.Biochem. Physiol. 73C: 129-134.85Klaverkamp, J.F., WA. Macdonald, D.A. Duncan, and R. Wagemann. 1984. Metallothionein and acclimationto heavy metals in fish: a review. In: V.W. Cairns, P.V. Hodson and J.O. Nriagu (eds.). ContaminantEffects of Fisheries. pp. 99-113. Adv. Env. Sci. Tech., 16. Wiley & Sons, Toronto.Knittel, M.D. 1980. Heavy metal stress and increased susceptibility of steelhead trout (Salmo gairdneri) toYersinia ruketi infection. In: J.G. Eaton, R.R. Parrish and A.C. Hendricks (Eds.). Aquatic Toxicology.A.S.T.M. Philadelphia, PA. STP 707. pp. 321-327.Koenings, J.P. 1976. In situ experiments on the dissolved and colloidal state of iron in an acid bog lake.Limnol. Oceanogr. 21(5): 674-683.Lauren, DJ. and D.G. McDonald. 1985. Effects of copper on branchial ionoregulation in the rainbow trout,Salmo gairdneri Richardson. Modulation by water hardness and pH. J. Comp. Physiol. B. 155: 635-644.Lauren, DJ. and D.G. McDonald. 1986. Influence of water hardness, pH, and alkalinity on the mechanismsof copper toxicity in juvenile rainbow trout, Salmo gairdneri. Can. J. Fish. Aquat. Sci. 43: 1488-1496.Lauren, DJ. and D.G. McDonald. 1987. Acclimation to copper by rainbow trout, Salmo gairdneri:biochemistry. Can. J. Fish. Aquat. Sci. 44: 105-111.LaZerte, B.D. and K. Burling. 1990. Manganese speciation in dilute waters of the precambrian shield, Canada.Water Res. 24(9): 1097-1101.Leckie, J.O. and JA. Davis. 1979. Aqueous environmental chemistry of copper. In: J.O. Nriagu (Ed.).Copper in the Environment. Part I. Ecological Cycling. John Wiley and Sons Inc. Toronto. pp. 89-121.Leckie, J.O. and R.O. James. 1975. Control mechanisms for trace metals in natural waters. In: AJ. Rubin(Ed.), Aqueous-Environmental Chemistry of Mets. Ann Arbor Science Publishers, Ann Arbor, Mich., pp.1-76.Leland, H.V. and J.S. Kuwabara. 1986. Trace metals. In: G.M. Rand and S.R. Petrocelli (Eds).Fundamentals of Aquatic Toxicology: Methods and Applications. Hemisphere Publishing Corp. New York.pp. 374-415.Lerman, A. and C.W. Childs. 1973. Metal-organic complexes in natural waters: control of distribution bythermodynamic, kinetic and physical factors. In: P.C. Singer (Ed.). Trace Metals and Metal-OrganicInteractions in Natural Waters. Ann Arbor Science Publishers Inc. Ann Arbor, Michigan. pp. 201-235.Ley, H.L., M.L. Failla, and D.S. Cherry. 1983. Isolation and characterization of hepatic metallothionein fromrainbow trout (Salmo gairdneri). Comp. Biochem. Physiol. 74B(3): 507-513.Lipman, K.M., R. Dodelson and R.M. Hays. 1966. The surface charge of isolated toad bladder epithelial cells.J. Gen. Physiol. 49: 501-508.Lloyd, R. 1960. The toxicity of zinc sulphate to rainbow trout. Ann. Appl. Biol. 48: 84-94.Lovegrove, J.M. and B. Eddy. 1982. Uptake and accumulation of zinc in juvenile rainbow trout, Salmogairdneri. Env. Biol. Fish. 7(3): 285-289.Luoma, S.N. 1983. Bioavailability of trace metals to aquatic organisms - a review. Sci. Tot. Envir. 28: 1-22.86Maliuga, D.P. 1941. Doklady Akad. Nauk, U.S.S.R. 31(2): 145. Translation RJ-296 of the AssociatedTechnical Services, Inc., POB 271, East Orange, New Jersey, U.S.A.; Chem. Abstr. 37.1 (1943), 894.Malian, J. 1985. Fish gill structural changes induced by toxicants and other irritants: a statistical review. Can.J. Fish. Aquat. Sci. 42: 630-648.Mance, G. 1987. Pollution Threat of Heavy Metals in Aquatic Environments. Elsevier Press, New York, N.Y.362 pp.Martin, J.H., GA. Knauer and A.R. Flegal. 1980. Distribution of zinc in natural waters. In: J.O. Nriagu(Ed.). Zinc in the Environment Part I: Ecological Cycling. John Wiley and Sons, Toronto, 453 pp.Martin, R.B. 1986. Bioinorganic chemistry of metal ion toxicity. In: H.E. Sigel (Ed.). Metal Ions inBiological Systems. Vol. 20: Concepts on Metal Ion Toxicity. Marcel Dekker, N.Y. pp. 21-65.McCarter, JA., A.T. Matheson, M. Roch and R.W. Olafson. 1982. Chronic exposure of coho salmon tosublethal concentrations of copper - II. Distribution of copper between high- and low-molecular-weightproteins in liver cytosol and the possible role of metallothionein in detoxification. Comp. Biochem. Physiol.72C(1): 21-26.McCarter, JA. and M. Roch. 1984. Chronic exposure of coho salmon to sublethal concentrations of copper-III. Kinetics of metabolism of metallothionein. Comp. Biochem. Physiol. 77C(1): 83-87.McCarter, JA. and M. Roch. 1983. Hepatic metallothionein and resistance to copper in juvenile coho salmon.Comp. Biochem. Physiol. 74C(1): 133-137.McKnight, D.M., G.L. Feder, E.M. Thurman, R.L. Wershaw and J.C. Westall. 1983. Complexation of copperby aquatic humic substances from different environments. Sci. Tot. Envir. 28: 65-76.McWilliams, P.G. and W.T.W. Potts. 1978. The effects of pH and calcium concentrations on gill potentialsin the brown trout, Salmo trutta. J. Comp. Physiol. 126: 277-286.McWilliams, P.G. 1982. The effects of calcium on sodium fluxes in the brown trout, Salmo trutta, in neutraland acid water. J. Exp. Biol. 96: 439-442.Millar, T.G. and W.C. McKay. 1982. The effects of hardness, alkalinity and pH of test water on the toxicityof copper to rainbow trout (Salmo gairdneri). Water Res. 14: 129-133.Millward, C. 1991. Pers. comm. E.P.S. Metal Chemist, West Vancouver E.P.S. Laboratory.Moore, J.N., S.N. Luoma and D. Peters. 1991. Downstream effects of mine effluent on an intermontaneriparian system. Can. J. Fish. Aquat. Sci. 48: 222-232.Morin, KA. 1990. Acid drainage from mine walls: the Main Zone pit at Equity Silver Mines. BritishColumbia Acid Mine Drainage Task Force.Morrison, G.M.P. 1987. Approaches to metal speciation analysis in natural waters. In: L. Landner (Ed.)Speciation of Metals in Water, Sediment, and Soil Systems. pp. 55-74.Neubecker, TA. and H.E. Allen. 1983. The measurement of complexation capacity and conditional stabilityconstants for ligands in natural waters. Water Res. 17: 1-14.87Neville, C.M. 1985. Studies on the mechanism of toxicity of acid and aluminum to juvenile rainbow trout ina low ion environment. Poster presentation, Int. Symp. Acidic Precipitation, Sept. 15-20, Muskoka,Ontario.Nieboer, E. and D.H.S. Richardson. 1980. The replacement of the nondescript term 'heavy metals' by abiologically and chemically significant classification of metal ions. Envir. Poll. 1B: 3-26.Nix, J. and R. Ingols. 1981. Oxidized manganese from hypolimnetic water as a possible cause of troutmortality in hatcheries. Prog. Fish Cult. 43: 32-36.Nordstrom, D.K, L.N. Plummer, T.M.L. Wigley, T.J. Wolery, J.W. Ball. EA. Jenne, R.L. Bassett, DA. Crerar,T.M. Florence, B. Fritz, M. Hoffman, G.R. Holdren, Jr., G.M. Lafon, S.V. Mattigod, R.E. McDuff, F.Morel, M.M. Reddy, G.Sposito, J. Thrailkill. 1979. Comparison of computerized chemical models forequilibrium calculations in aqueous systems. In: EA. Jenne (Ed.). Chemical Modelling in AqueousSystems. A.C.S. Symposium Series 93. pp. 857-894.Nordstrom, D.K. and J.W. Ball. 1984. Chemical models, computer programs and metal complexation innatural waters. In: C.J.M. Kramer and J.C. Duinker (Eds.). Complexation of Trace Metals in NaturalWaters. Martinus Nijhoff/Dr. W. Junk Publishers. Boston. pp. 149-164.Norusis, M.J. 1988. SPSS/PC+ V2.0 Base Manual for the IBM PC/XT/AT and PS/2. SPSS Inc. Chicago.Nriagu, J.O. 1979. Global inventory of natural and anthropogenic emissions of trace metals to the atmosphere.Nature 279: 409-411.Nriagu, J.O. 1980. Production, uses, and properties of cadmium. In: J.O. Nriagu (Ed.) Cadmium in theEnvironment. Part I: Ecological Cycling. pp. 35-70.O'Donnel, J.R., B.M. Kaplan, and H.E. Allen. 1985. Bioavailability of trace metals in natural waters. In: R.D.Cardwell, R. Purdy, and R.C. Bahner (Eds.), Aquatic Toxicology and Hazard Assessment: SeventhSymposium, ASTM STP 854. American Society of Testing and Materials, Philadelphia. pp. 485-501.Olafson, R.W. 1987. Electrochemical quantification and characteristics of metallothionein. ExperientiaSupplementum 52: 329-333.Olafson, R.W. and R.G. Sim. 1979. An electrochemical approach to quantification and characterization ofmetallothioneins. Anal. Biochem. 100: 343-351.Olsson, P.E. and C. Haux 1985. Rainbow trout metallothionein. Inorganica Chimica Acta 107: 67-71.Olsson, P., A. Larsson, A. Maage, C. Haux, K. Bonham, M. Zafarullah, and L. Gedamu. 1989. Induction ofmetallothionein synthesis in rainbow trout, Salmo gairdneri, during long-term exposure to waterbornecadmium. Fish Physiol. Biochem. 6(4): 221-229.Olsson, P., C. Haux and L. Forlin. 1987. Variations in hepatic metallothionein, zinc and copper levels duringan annual reproductive cycle in rainbow trout, Salmo gairdneri. Fish Physiol. Biochem. 3(1): 39-47.Pagenkopf, G.K. 1978. Introduction to natural water chemistry. Marcel Dekker, Inc. New York, New York.271 pp.88Pagenkopf, G.K. 1980. Zinc speciation and toxicity to fish. In: J.O. Nriagu (Ed.), Zinc in the Environment.Part H: Health Effects. John Wiley and Sons, Toronto, Ontario. 480 pp.Pagenkopf, G.K. 1983. Gill surface model for trace-metal toxicity to fishes: role of complexation, pH, andwater hardness. Envir. Sci. Technol. 17(6): 342-347.Pankow, J.F. and J.J. Morgan. 1981. Kinetics for the aquatic environment. Environ. Sci. Technol. 15(11):1306-1313.Part, P. and RA.C. Lock. 1983. Diffusion of calcium, cadmium and mercury in a mucous solution fromrainbow trout. Comp. Biochem. Physiol. 76C(2): 259-263.Part, P. and 0. Svanberg. 1981. Uptake of cadmium in perfused rainbow trout (Salmo gairdneri) gills. Can.J. Fish. Aquat. Sci. 38: 917-924.Part, P. 1986. Testing the bioavailability of metals in natural waters. In: L. Landner (Ed.). Speciation ofMetals in Water, Sediment and Soil Systems. Springer-Verlag, New York. pp. 115-126.Part, P., 0. Svanberg and A. Kiessling. 1985. The availability of cadmium to perfused rainbow trout gills indifferent water qualities. Water Res. 19(4): 427-434.Part, P. and G. Wikmark. 1984. The influence of some complexing agents (EDTA and citrate) on the uptakeof cadmium in perfused rainbow trout gills. Aquat. Toxicol. 5: 277-289.Patrick, F.M. and M.W. Loutit. 1978. Passage of metals to freshwater fish from their food. Water Res. 12:395-398.Patterson, R.J. 1986. Equity Silver Mines Ltd. Environmental Report 1980 - 1985. Sections A and B. 253PP.Patterson, R.J. 1987. Environmental and reclamation measures, Equity Silver mines Ltd. Proceedings of theEleventh Annual British Columbia Mine Reclamation Symposium. April 8-10, 1987. Campbell River, B.C.pp. 117-138.Perrin, C.J. 1991. A stream community responses to a gradient of acid mine drainage additions. Reportsubmitted to B.C. Ministry of Mines, Energy, and Petroleum. 44 pp.Pearson, R.G. 1968a. Hard and soft acids and bases, HSAB, part I: fundamental principles. J. Chem. Educ.45(9): 581-587.Pearson, R.G. 1968b. Hard and soft acids and bases, HSAB, part II: underlying theories. J. Chem. Educ.45(10): 643-648.Pierson, KB. 1985. Isolation and partial characterization of a non-thionein, zinc binding protein from the liverof rainbow trout (Salmo gairdneri). Comp. Biochem. Physiol. Vol. 80C(2): 299-304.Perdue, E.M. 1979. Solution thermochemistry of humic substances. In: EA. Jenne (Ed.). Chemical Modellingin Aqueous Systems: Speciation, Sorption, Solubility, and Kinetics. A.C.S. Symposium Series 93. pp. 99-114.89Petering, D.H. and BA. Fowler. 1986. Discussion summary. Roles of metallothionein and related proteinsin metal metabolism and toxicity: problems and perspectives. Envir. Health Persp. 65: 217-224.Price-Haughey, J., K. Bonham and L. Gedamu. 1986. Heavy metal induced gene expression in fish and fishcell lines. Envir. Health Persp. 65: 141-147.Price-Haughey, J. and L. Gedamu. 1987. Heavy metal induced protein synthesis in fish cell lines. In: Secondinternational meeting on metallothionein and other low molecular weight metal-binding proteins. Zurich,Aug. 21-24, 1985. pp. 465-469.Randhawa, N.S. and F.E. Broadbent. 1965. Soil organic matter metal complexes at different pH values. SoilSci. 99: 67-83.Raspor, B.H.W. Nurnberg, P. Valenta, and M. Branica. 1984. Significance of dissolved humic substances forheavy metal speciation in natural waters. In: C.J.M. Kramer and J.C. Duinker (Eds.). Complexation ofTrace Metals in Natural Waters. Martinus Nijhoff/Dr. W. Junk Publishers. Boston. pp. 317-328.Reid, S.D. 1989. Metal-gill surface interactions in rainbow trout, Oncorhynchus mykiss. Ph.D. thesis,McMaster University, Hamilton, Ont.Reichert, W.L., DA. Federighi and D.C. Malins. 1979. Uptake and metabolism of lead and cadmium in cohosalmon (Oncorhynchus kisutch). Comp. Biochem. Physiol. 63C: 229-234.Remington, D. 1989. Memo to File Re: Goosly Lake Update. M.O.E. Report PE-4475 55 pp.Reuter, J.H. and E.M. Perdue. 1977. Importance of heavy metal organic matter interactions in natural waters.Geochim. Cosmochim. Acta. 41: 325-334.Roberts, K.S., A. Cryer, J. Kay and J.F. deL.G. Solbe. 1979. A high-molecular-weight cadmium-bindingfraction isolated from the liver cytosol of trout exposed to environmentally relevant concentrations of themetal. Biochem. Soc. Trans. 7: 650-651.Robertson, A.M. 1987. Alternative acid mine drainage abatement measures. Proceedings of the EleventhAnnual British Columbia Mine Reclamation Symposium. April 8-10, 1987. Campbell River, B.C. pp. 7-30.Roch, M., P. Noonan and JA. McCarter. 1985. Determination of no effect levels of heavy metals for rainbowtrout using hepatic metallothionein. Water Res. 20(6): 771-774.Roch, M. JA. McCarter, A.T. Matheson, M.J.R. Clark, and R.W.Olafson. 1982. Hepatic metallothionein inrainbow trout (Salmo gairdneri) as an indicator of metal pollution in the Campbell River System. Can.J. Fish. Aquat. Sci. Vol 39(12):1596-1601.Roch, M. and JA. McCarter. 1984b. Metallothionein induction, growth, and survival of chinook salmonexposed to zinc, copper, and cadmium. Bull. Environ. Contam. Toxicol. 32:478-485.Roch, M. and JA. McCarter. 1984a. Hepatic metallothionein production and resistance to heavy metals byrainbow trout (Salmo gairdneri) - II. Held in a series of contaminated lakes.Roth, M. and JA. McCarter. 1986. Survival and hepatic metallothionein in developing rainbow trout exposedto a mixture of zinc, copper and cadmium. Bull. Environ. Contam. Toxicol. 36:168-175.90Shannon, R.D. 1976. Revised effective ionic radii and systematic studies of interatomic distances in halidesand chalcogenides. Acta Crystallogr. A. 32: 751.Shears, MA. and G.L. Fletcher. 1983. Regulation of Zn 2+ uptake from the gastrointestinal tract of a marineteleost, the winter flounder (Pseudopleuronectes ameticanus). Can. J. Fish. Aquat. Sci. 40(Suppl. 2): 197-205.Sholkovitz, E.R. 1990. Redox-related geochemistry in lakes; alkali metals, alkaline earth elements, and 137Cs.In: W. Stumm (Ed.). Chemical Processes in Lakes. John Wiley and Sons, Toronto. pp. 119-142.Silver, M. 1989. Analytical techniques for research on the abatement of bacterial acid generation in the pyritictailings. Canmet Report 86-3E. Energy, Mines and Resources, Canada. 24 pp.Singer, P.C. and W. Stumm. 1970. Acidic mine drainage: the rate determining step. Science 167: 1121-1123.Singh, S.M. and P.N. Ferns. 1978. Accumulation of heavy metals in rainbow trout Salmo gairdneri(Richardson) maintained on a diet containing activated sewage sludge. J. Fish Biol. 13: 277-286.Singh, S.K. and V. Subramanian. 1983. Hydrous Fe and Mn oxides -scavengers of heavy metals in the aquaticenvironment. C.R.C. Crit. Rev. Envir. Cont. 14(1): 33-90.Skidmore, J.F. 1964. Toxicity of zinc compounds to aquatic animals, with special reference to fish. Quart.Rev. Biol. 39: 227-248.Skidmore, J.F. and P.WA. Tovell. 1972. Toxic effects of zinc sulphate on the gills of rainbow trout. WaterRes. 6: 217-230.Smith, R.W. and EA. Jenne. 1991. Recalculation, evaluation, and prediction of surface complexation constantsfor metal adsorption on iron and manganese oxides. Environ. Sci. Technol. 25: 525-531.Spear, PA. and R.C. Pierce. 1979. Copper in the aquatic environment: chemistry, distribution and toxicology.National Research Council of Canada No. 16454. 227 pp.Spear, PA. 1981. Zinc in the aquatic environment: chemistry, distribution and toxicology. National ResearchCouncil of Canada #17589. 143 pp.Spry, DJ. and C.M. Wood. 1985. Ion flux rates, acid-base status, and blood gases in rainbow trout, Salmogairdneri, exposed to toxic zinc in natural soft water. Can. J. Fish. Aquat. Sci. 42: 1332-1341.Stiff, MJ. 1971. The chemical states of copper in polluted fresh water and a scheme of analysis todifferentiate them. Water Res. 5: 585-599.Stockner, J.G. and K.S. Shortreed. 1985. Whole lake fertilization experiments in coastal British Columbialakes: empirical relationships between nutrient inputs and phytoplankton biomass and productions. Can.J. Fish. Aquat. Sci. 42: 649-658.Stokes, P.M. 1979. Copper accumulations in freshwater biota. In J.O. Nriagu (Ed.) Copper in theEnvironment, Part I: Ecological Cycling. John Wiley and Sons Inc. Toronto. pp. 357-381.Stone, H. and J. Ovemell. 1985. Non-metallothionein cadmium binding proteins. Comp. Biochem. Physiol.80C(1): 9-14.91Stone, A.T. and J.J. Morgan. 1990. Kinetics of chemical transformations in the environment. In: W. Stumm(Ed.). Aquatic Chemical Kinetics: Reaction Rates of Processes in Natural Waters. John Wiley and Sons,Toronto. pp. 1-41.Stumm, W. and J.J. Morgan. 1981. Aquatic Chemistry. An Introduction Emphasizing Chemical Equilibriain Natural Waters. 2nd Ed. John Wiley and Sons, Toronto. 780 pp.Stumm, W. 1990. Aquatic Chemical Kinetics: Reaction Rates of Processes in Natural Waters. John Wileyand Sons, Toronto. 545 pp.Sunda, W.G. and JA.M. Lewis. 1978. Effect of complexation by natural organic ligands on the toxicity ofcopper to a unicellular alga, Monochrysis lutheri. Limnol. Oceanogr. 23(5): 870-876.Suzuki, K.T. 1987. Quantification and characterization of metallothioneins in tissues of lower vertebrates andinvertebrates. Experientia Supp. Vol. 52. Metallothionein II. pp. 265-272.Thompson, J.A.J. and R.P. Cosson. 1984. An improved electrochemical method for the quantification ofmetallothioneins in marine organisms. Mar. Envir. Res. 11: 137-152.Thomas, D.G., M.W. Brown, D. Shurben, J.F. Del G. Solbe, A. Cryer, and J.Kay. 1985. A comparison of thesequestration of cadmium and zinc in the tissues of rainbow trout (Salmo gairdneri) following exposure tothe metals singly or in combination. Comp. Biochem. Physiol. Vol. 82(C)(1): 55-62.Thomas, D.G., J.F.deL.G. Solbe, J. Kay and A. Cryer. 1983. Environmental cadmium is not sequestered bymetallothionein in rainbow trout. Biochem. Biophys. Res. Comm. 110(2): 584-592.Waldichuck, M. 1974. Some biological concerns in heavy metal pollution. In: Pollution and Physiology ofMarine Organisms. Academic Press, New York. pp. 1-57.Weast, R.C., M. J. Astle and W.H. Beyer. 1986. CRC Handbook of Chemistry and Physics. CRC Press, Inc.Boca Raton, Florida.Webb, M. 1987. Toxicological significance of metallothionein. Experientia Supplementum Vol. 52.Metallothionein II. pp. 109-134Weber, W.J. and H.S. Poselt. 1974. Equilibrium models and precipitation reactions for cadmium. In: A.J.Rubin (Ed.) Aqueous Environment Chemistry of Metals. Ann Arbor Science Publishers, Ann Arbor,Mich. 360 pp.Wekel, J.C., K.D. Shearer and C.R. Houle. 1983. High zinc supplementation of rainbow trout diets. Prog.Fish. Cult. 45(3): 144-147.Wetzel, R.G. 1983. Limnology. Second Ed. Saunders College Publishing. Toronto.Wilkes, B. 1987. Consequences of unregulated release of raw acid mine drainage into the Bulldey River,British Columbia. Proceedings of the Eleventh Annual British Columbia Mine Reclamation Symposium.April 8-10, 1987. Campbell River, B.C. pp. 139-152.Williams, R.J.P. and J.D. Hale. 1966. The classification of acceptors and donors in inorganic reactions.Struct. Bond. 1: 249-259.92Williams, D R and J.P. Geisy Jr. 1978. Relative importance of food and water sources to cadmium uptakeby Gambusia affinis. Envir. Res. 16: 326-332.Wilson, D.E. and P. Kinney 1977. Effects of polymeric charge variations on the proton-metal ion equilibriaof humic materials. Limnol. Oceanogr. 22: 281-289.Winge, D.R., R. Premakumar and K.V. Rajagopalan. 1978. Studies on the zinc content of Cd-inducedthionein. Arch. Biochem. Biophys. 188(2): 466-475.Winge, D., J. Kranso and A.V. Colucci. 1973. Cadmium accumulation in the rat liver. Correlation betweenbound metal and pathology. In: W.G. Hoekstra, J.W. Suttie, H.E. Ganther and W. Mertz (Eds.). TraceElement Metabolism in Animals Vol. 2. University Park Press, Baltimore. pp. 500-501.Winge, D.R. and M. Brouwer. 1986. Discussion summary. Techniques and problems in metal-binding proteinchemistry and implications for proteins in nonmammalian organisms. Environ. Health Persp. 65: 211-214.Wollast, R. 1990. Rate and mechanism of dissolution of carbonates in the system CaCO 3-MgCO3. In: W.Stumm (Ed.). Aquatic Chemical Kinetics: Reaction Rates on Processes in Natural Waters. John Wileyand Sons Inc. Toronto. pp. 431-445.93APPENDIX 1Water Chemistry Data for Mesocosm Study94Table 25. Raw water chemistry data for the Equity Silver mine mesocosm study.Hc = Ca + Mg Hardness, Ht= Total HardnessDate Trough Treat Cond pH Alk^Al As B Ba Be Ca Cd Co Cr Cu^Fe K Mg Mn Na Ni P Sb Se Si Sr Zn Hc Ht====+====1====+====2====+====3====+====4====+====5====+====6====+====7====+====8====+====g====+====0====+====1====+====2====+====3====+===.4====+900816 5 1:1000 . . . 300 49 10 19 0.9 18.9 0.09 4 4 29 790 1.9 5.6 36 3.5 19 200 49 49 8.13 193 25 70.3 73.2900816 15 1:1000 . . 400 49 9 20 0.9 20.6 0.2 4 4 41 1100 5 5.9 56 3.4 19 99 49 49 8.35 202 34 75.6 79.6900823 5 1:1000 178 . 38.9 500 49 9 20 0.9 20.4 0.3 4 4 56 1360 1.9 6.1 78 3.6 19 99 49 80 8.13 205 104 75.9 80.8900823 15 1:1000 174 38.1 240 49 9 20 0.9 20 0.2 4 4 25 682 1.9 5.8 32 3.4 19 99 49 49 8.09 205 35 73.8 76.3900827 5 1:1000 . 5.78 .900827 15 1:1000 6.49 . . . . .900830 5 1:1000 177 5.94 41.7 470 49 9 18 0.9 22.7 0.3 7 4 49 1340 1.9 6.4 61 3.6 19 99 49 49 7.56 194 4 82.7 87.3900830 15 1:1000 174 7.03 34.8 260 49 20 19 0.9 24.1 0.2 4 4 28 772 3 6.4 35 3.6 19 99 49 49 7.84 200 1 86.7 89.3900906 5 1:1000 166 6.11 34.7 400 49 9 19 0.9 15.8 0.3 6 4 41 1110 6 6 57 3.8 19 99 49 49 8.2 199 23 64.3 68.2900906 15 1:1000 164 6.69 41.7 200 49 60 17 0.9 13.4 0.3 4 4 23 489 2 5.6 46 3.7 19 99 49 49 7.5 185 24 56.2 58.2900816 3 1:10000 . . . 210 49 9 19 0.9 19.9 0.09 4 4 21 400 1.9 5.6 29 3.3 19 100 49 49 8.18 194 19 72.9 74.7900816 2 1:10000 . . 49 49 9 20 0.9 22.1 0.4 4 4 36 959 1.9 6.5 133 3.7 19 99 49 49 8.59 212 75 82.1 84.6900823 2 1:10000 175 . 40.3 60 49 9 19 0.9 18.8 0.09 4 4 13 190 2 5.5 8 3.4 19 99 49 49 7.86 195 43 69.7 70.6900823 3 1:10000 172 41.7 90 49 9 21 0.9 21.3 0.09 4 4 9 265 4 6 11 3.4 19 99 49 49 8.41 212 39 77.6 78.8900827 3 1:10000 7 .900827 2 1:10000 6.77 . .900830 2 1:10000 174 6.97 34.8 80 49 9 19 0.9 22.5 0.2 4 4 8 212 1.9 6.2 10 3.5 19 99 49 49 7.65 194 1 81.4 82.4900830 3 1:10000 174 7.25 8.69 70 50 9 19 0.9 23.2 0.2 4 4 6 151 1.9 6.3 7 3.5 19 99 49 49 7.79 198 1 83.7 84.4900906 2 1:10000 158 6.64 45.2 50 49 20 19 0.9 16 0.2 4 4 6 167 6 5.8 10 3.6 19 99 49 49 8.29 201 7 63.9 64.7900906 3 1:10000 158 7.07 45.2 80 49 10 19 0.9 15.2 0.1 4 4 5 122 3 5.6 7 3.6 19 99 49 49 8.04 193 4 61.3 62900816 9 1:25 . 14000 70 9 20 2 34.4 16 68 4 1440 25900 1.9 15.2 1680 9.9 180 300 49 49 9.91 291 1070 149 254900816 4 1:25 . 14700 110 9 21 2 38.2 14 79 4 1480 28100 1.9 16 1810 9.7 180 300 49 49 10.4 307 1120 161 274900816 13 1:25 15200 110 9 19 2 34.1 15 65 4 1570 31600 4 15.6 1820 10.3 190 500 49 49 9.54 285 1160 149 270900823 13 1:25 375 0 12700 120 9 19 2 33.3 14 37 6 1330 23800 1.9 14.3 1560 10.1 170 300 49 49 9.24 287 995 142 239900823 4 1:25 582 0 19000 240 9 19 3 40.7 22 63 4 1990 43000 1.9 18.7 2300 13.2 240 600 49 49 10.2 332 1570 178 336900823 9 1:25 523 0 16400 260 9 20 3 39.7 18 51 9 1700 35900 1.9 17 2.05 11.5 210 400 49 49 9.97 321 1290 169 303900827 9 1:25 . 3.4900827 13 1:25 . 3.32900827 4 1:25 3.05900830 9 1:25 928 3.5 215000 270 41.6 17 196 6 1610 31300 1.9 17.6 1850 132 190 400 49 49 329.17 1160 2176900830 13 1:25 928 3.44 0 14200 220 9 17 2 41.7 18 190 4 1510 31700 4 17 1800 12.2 190 400 49 49 9.05 294 1120 174 291900830 4 1:25 896 3.44 0 13000 290 9 18 2 38.4 16 185 7 1400 31600 1.9 15.9 1590 11.7 170 500 49 49 8.86 284 990 161 273900906 4 1:25 655 4.09 0 12300 240 9 21 2 28.2 14 142 4 1260 27900 7 14.9 1630 11.5 150 300 49 49 9.98 306 886 132 234900906 13 1:25 982 3.48 0 18600 280 9 19 3 32.4 26 236 4 1910 43400 3 19.3 2410 15.8 250 500 49 49 10.5 350 1530 161 318900906 9 1:25 644 3.59 0 11000 170 20 19 2 24 13 112 4 1160 1870 1.9 13.7 1430 11.4 150 200 49 49 9.1 281 876 116 197900816 10 1:250 . 49 49 9 21 0.9 22.7 0.4 8 4 40 1130 4 6.7 160 3.8 19 99 49 49 8.69 215 91 84.3 87.2900816 1 1:250 . 90 49 9 20 0.9 19.8 0.09 4 4 8 214 1.9 5.6 12 3.3 9 99 49 49 8.28 196 10 72.5 73.5900823 1 1:250 189 32 49 49 9 21 0.9 22 0.8 7 4 44 737 3 6.7 162 4 19 99 49 49 8.27 216 145 82.6 8595Table 25 (continued). Raw water chemistry data for the Equity Silver mine mesocosm study.Hc = Ca Mg Hardness, Ht= Total HardnessDate Trough Treat Cond 1pH Alk^Al As B Ba Be Ca Cd Co Cr Cu^Fe K Mg Mn Na Ni P Sb Se Si Sr Zn Hc Ht====+====1====+====2====+====3====+.4====+====5====+====6====+====7.+====8====+====9====+====0====+====1====+====2====+====5====+====4====+900823 10 1:250 188 . 32.5 190 49 9 21 0.9 23.7^0.9 4 4 81 1410 1.9 7.1 227 4.1 19 99 49 49 8.41 223 162 88.3 92.5900827 1 1:250 5.79 .900827 10 1:250 5.26 .900830 10 1:250 206 5.88 17.4 220 5018 0.9 24.9^0.9 213 984 7.5 227 4.6 20 99 490i1 8349 7.65 96.6900830 1 1:250 191 5.83 17.4 80 49 9 18 0.9^24^0.8 10 4 59 583 1.9 7.2 180 4.3 20 99 49 49^7.6 204 78 89.6 91.7900906 1 1:250 188 5.63 38.2 80 49 9 18 0.9 17.5^0.7 9 4 57 379 7 6.8 192 4.5 19 99 49 49 8.06 224 101 71.5 73.4900906 10 1:250 175 5.94 27.8 70 60 9 18 0.9 14.5^0.5 4 4 30 339 1.9 6.1 113 4.2 19 99 49 49 7.68 193 61 61.4 62.8900816 7 1:2500 140 49 9 21 0.9 21.8 0.09 4 4 12 351 2 6 16 3.3 19 99 49 49 8.73 210 10 79 80.4900816 16 1:2500 . . 50 49 9 20 0.9 21.5 0.09 4 4 4 112 3 5.8 7 3.2 19 99 49 49 8.64 208 4 77.8 78.4900823 7 1:2500 173 . 55.2 180 49 9 21 0.9 21.8 0.09 4 4 18 488 2 6 23 3.4 19 99 49 49 8.51 213 33 79.3 81.2900823 16 1:2500 174 41 160 49 9 21 0.9 21.2^0.2 4 4 17 458 1.9 6 22 3.5 19 99 49 49 8.43 214 30 77.4 79.2900827 16 1:2500 . 6.71 .900827 7 1:2500 6.72900830 16 1:2500 170 7.07 38.2 166 49 20 19 0.9 23.5^0.14 44 15 1.9 6.3 22 3.5 19 99 4919860 7.73 1 84.5 86.2900830 7 1:2500 134 7.18 38.2 170 60 9 18 0.9 22.5 0.09 4 4 20 51 1.9 6.2 23 3.5 19 99 49 49 7.57 193 1 81.7 83.5900906 7 1:2500 162 6.86 41.7 100 49 9 18 0.9 15.1^0.1 4 4 10.8 303 7 5.6 14 3.5 19 99 49 49 7.91 191 4 61 62.2900906 16 1:2500 162 6.71 41.7 130 49 9 18 0.9 13.9^0.2 4 4 11.2 367 2 5.6 18 3.5 19 99 49 49 7.71 189 11 57.6^59900816 14 1:25000 . 49 49 10 20 0.9 19.7 0.09 4 4 5 95 6 5.5 7 3.2 19 99 49 49 8.17 196 6 72 72.6900816 12 1:25000 . . 140 49 9 21 0.9 21.6 0.09 4 8 15 246 8 5.9 19 3.2 19 100 49 49 8.58 207 13 78.2 79.5900823 14 1:25000 174 . 41.7 60 49 9 20 0.9 20.7^0.1 4 4 7.5 151 1.9 5.8 9 3.2 19 99 49 49^8.1 205 22 75.4 76.2900823 12 1:25000 173 . 44.4 49 49 9 22 0.9 22.5^0.1 4 4 3.9 125 1.9 6.1 5 3.3 19 99 49 49 8.65 219 18 81.5 82.1900827 14 1:25000 . 6.43 .900827 12 1:25000 6.94 . . .^.900830 14 1:25000 168 7.03 60 49 ; 17 0.9 21.i0.0;.615.2 1.9 5.9161 ;9 3.4 99 70 7.31 186 1 78.2 78.9900830 12 1:25000 169 7.41 41.7 50 49 9 17 0.9 21.9 0.09 4 4 3.3 82 1.9 5.9 3 3.3 19 99 49 49^7.3 186 1 78.7 79.2900906 12 1:25000 156 7.19 45.2 49 49 9 18 0.9 13.5^0.1 4 4 2.6 52 1.9 5.4 3 3.6 19 99 49 49^7.7 187 1 56.2 56.6900906 14 1:25000 158 6.69 41.7 90 49 50 18 0.9 13.8^0.1 4 4 10.8 159 3 5.5 16 3.4 19 99 49 49^7.7 188 11 57.1^58900816 11 control . . . 49 49 9 21 0.9 22.5 0.09 4 4 2 27 5 6 3 3.2 19 99 49 49 8.88 214 6 81.2 81.6900816 8 control 60 49 9 21 0.9 21.6 0.09 4 4 6.1 119 3 6 9 3.4 19 99 49 49 8.77 211 14 78.6 79.3900816 6 control . . 49 49 10 21 0.9 21.7 0.09 4 4 1.6 40 1.9 6 3 3.4 19 99 49 49^8.8 210 5 78.7 79.1900823 8 control 176 . 44.4 49 49 9 20 0.9^20 0.09 4 4 2.4 81 1.9 5.6 5 3.3 19 99 49 49 8.06 201 31 73.2 73.8900823 11 control 174 . 43.1 49 49 9 20 0.9 20.1 0.09 4 4 2.6 57 1.9 5.6 3 3.2 19 99 49 49 8.04 200 11 73.3 73.7900823 6 control 173 . 53.6 49 49 9 20 0.9 19.6 0.09 4 4 2.2 50 1.9 5.6 3 3.4 19 99 49 49^8.1 201 27 72.1 72.5900827 11 control . 6.72 . .900827 8 control . 7.26 .900827 6 control . 6.75 . . . . .^.^.^. . .^.900830 6 control 170 7.1 41.7 49 90 9 18 0.9 21.9 0.09 4 4 2.1 58 3 ; 2 3.4 19 99 49 80 7.42 189 1 79.2 79.6900830 11 control 169 7.27 45.2 49 49 10 18 0.9 21.9 0.09 4 4 3.9 61 1.9 5.9 3 3.3 19 99 49 49 7.36 187 1 78.8 79.2900830 8 control 169 7.57 41.7 49 49 9 18 0.9 21.4 0.09 4 4 1.5 32 1.9 6 0.9 3.5 19 99 49 49 7.45 190 1 78 78.3900906 11 control 155 6.9 46.9 49 49 9 17 0.9 13.2 0.09 4 4 1.7 35 1.9 5.3 2 3.6 19 99 49 49 7.55 183 1 54.9 55.2900906 8 control 155 7.4 41.7 49 49 30 19 0.9 15.6 0.09 4 4 2.4 47 4 5.9 3 3.8 19 99 49 60 8.28 199 3 63.1 63.5900906 6 control 155 6.75 41.7 49 70 9 19 0.9 16.2^0.1 4 4 1.8 53 5 5.9 3 3.7 19 99 49 49 8.44 201 1 64.8 65.296APPENDIX 2Fish Tissue Chemistry Data for Mesocosm Study97Table 26. Raw data from fish tissue analyses. dwt= change in weight (grams), Metal=metallothionein concentrationTrough TreatmentDateofDeathChangein weight(grams)Metallo-thioneinAmoLigGillCuA9/9'MuscleCuA9/9LiverCuAD/DGillZn119/9MuscleZnA9/9LiverZnA9/9LiverCdAD/DGillCdAD/D6 Control 900905 13.63 7.45 44.5 555 30.64 399 7.116 Control 900907 1.54 4.42 3.71 42.3 383 57.5 468 4.94 2.296 Control 900907 0.77 7.82 3.58 33.2 361 50.1 367 4.29 1.216 Control 900907 1.76 4.96 3.49 32.7 319 46.87 405 2.64 1.646 Control 900907 1.14 9.18 2.51 26.3 453 52.45 693 7.09 1.856 Control 900907 . 1.07 4.69 2.7 30.6 349 62.27 40 3.83 2.038 Control 900907 -1.15 3.11 5.58 3.3 43.9 387 62.91 1956 21.61 1.618 Control 900907 5.31 2.28 5.64 4.18 48.3 276 74.04 1713 18.1 08 Control 900907 0.45 1.74 4.37 4.24 21.1 381 55.65 261 9.83 1.148 Control 900907 1.68 7.55 3.63 44.5 608 50.78 795 7.58 08 Control 900907 4.05 2.15 9.07 3.94 29.1 359 65.42 709 12.68 0.688 Control 900907 -2.42 1.31 6.13 3.79 24.7 389 54.12 834 11.67 0.7211 Control 900907 4.06 1.22 15.31 2.58 34.3 670 167.26 455 7.13 1.8611 Control 900907 2.95 1.89 5.86 1.96 20.1 325 47.05 332 4.96 1.4611 Control 900907 3.19 1.95 14.52 4.63 28.4 802 84.19 419 4.67 1.0411 Control 900907 1.95 1.18 4.11 2.69 36.3 354 69.24 582 15.61 1.0411 Control 900907 3.51 1.72 6.87 3.04 38.8 711 89.55 434 7.85 1.8211 Control 900907 -0.76 1.02 6.43 2.82 13.3 232 47.39 391 6.59 0.5912 1:25000 900827 1.06 10.1 4.62 57.3 526 114.05 223 3.48 2.6112 1:25000 900907 0.63 15.69 3.13 12.7 464 76.05 324 3.86 0.6412 1:25000 900907 2.93 10.28 2.18 39.9 434 83.84 629 15.94 1.0112 1:25000 900907 1.2 9.35 2.52 111.7 453 72.84 1960 29.93 2.6412 1:25000 900907 2.02 12.48 2.54 39.8 681 59.66 2296 16.97 1.4712 1:25000 900907 . 1.23 8.73 3.42 26.4 381 84.86 531 3.86 1.5914 1:25000 900907 -2.21 1.23 2.97 2.72 23.4 536 46.79 468 8.22 0.3314 1:25000 900907 1.98 1.56 5.4 4.25 54.1 422 73.74 813 9.57 1.3714 1:25000 900907 1.45 2.31 4.74 2.33 50.0 521 82.11 635 8.34 3.2214 1:25000 900907 -3.72 1.27 8.74 3.1 16.8 352 57.41 408 8.5 1.2714 1:25000 900907 1.52 1.53 4.49 2.54 43.0 349 61.73 655 6.58 0.914 1:25000 900907 1.07 1.44 2.72 2.77 34.7 246 53.43 1411 15.28 1.552 1:10000 900907 2.66 1.11 8.03 2.49 26.3 543 54.72 176 3.91 1.892 1:10000 900907 3.35 1.75 7.51 4.9 56.1 310 60.79 316 5.51 2.32 1:10000 900907 1.04 1.59 7.45 3.54 12.3 292 62.55 498 3.31 2.882 1:10000 900907 -2.78 1.07 9.92 3 15.6 358 54.82 174 1.17 1.762 1:10000 900907 -1.24 1.47 3.88 5.39 26.6 302 48.98 324 0.25 1.62 1:10000 900907 4.19 1.25 9.84 3.13 36.9 329 59.83 329 4.17 2.0298Table 26 (continued). Raw data from fish tissue analyses. dwt= change in weight (grams), Metal=metallothionein concentrationTrough TreatmentDateofDeathChangein weight(grams)Metallo-thioneinAmol/gGillCuAg/gMuscleCuAg/gLiverCuAg/gGillZnAg/gMuscleZnAg/gLiverZnAg/gLiverCdAg/gGillCdAg/g3 1:10000 900907 0.05 1.92 7.9 2.21 31.5 357 84.54 381 1.64 1.523 1:10000 900907 1.42 2.01 5.53 3.16 16.3 457 66.51 278 2.74 1.23 1:10000 900907 2.67 2.83 4.06 4.45 50.9 241 51.87 513 2.61 13 1:10000 900907 0.97 2.44 7.03 2.97 35.2 395 51.89 577 2.61 1.523 1:10000 900907 1.68 1.5 7.71 2.63 39.7 473 69.28 515 2.633 1:10000 900907 0.64 1.85 4.88 3.75 20.2 452 63.9 340 1.68 1.537 1:2500 900907 3.16 1.56 9.21 3.27 39.0 276 55.88 643 11.15 2.917 1:2500 900907 1.83 1.09 3.39 2.37 43.9 439 80.45 849 11.56 07 1:2500 900907 1.72 2.09 8.52 3.28 42.0 617 70.93 954 13.8 1.0816 1:2500 900826 . 11.78 2.14 85.1 288 48.98 1416 8.4116 1:2500 900826 2.92 13.48 2.63 54.8 296 67.73 193 6.22 1.4416 1:2500 900827 -0.18 3.28 15.66 2.54 37.1 326 48.84 105 1.42 2.9516 1:2500 900827 0.81 . 23.1 3.93 63.8 258 71.11 428 12.53 1.4616 1:2500 900829 21.82 6.9 49.3 326 75.57 40 4.49 1.6816 1:2500 900907 2.65 1.19 8.83 2.63 37.3 362 37.2 403 3.28 1.915 1:1000 900827 0.51 103.38 5.32 29.9 769 152 479 6.87 3.045 1:1000 900828 0.79 86.24 3.61 16.4 283 39.47 425 6.46 1.99. Abbots 900827 1.64 5.02 3.38 26.3 398 41.89 5 2.22 1.36Abbots 900827 0.9 3.83 2.89 86.5 440 60.55 587 4.09 1.57Abbots 900827 3.64 3.69 3.08 61.2 360 66.3 14 3.43 2.03Abbots 900827 1.25 5.98 3.75 93.3 301 47.68 566 9.55 1.06Abbots 900827 1.71 4.6 3.95 64.4 436 36.23 18 3.99 1.64Abbots 900827 1.71 5.38 2.79 24.0 405 41.6 3 4.48 1.81Abbots 900827 1.74 3.14 32.4 48.36 6 6.52Abbots 900827 1.38 4.66 4.08 54.8 366 60.69 17 3.61 1.8'4'Abbots 900827 1.61 6.84 4.57 12.3 285 36.29 14 2.83 1.72Goosly 900907 . 5.05 3.07 304 56.99 2.52Goosly 900907 3.75 1.58 12.5 134 27.52 11 1.47 1.57Goosly 900907 6.67 2.51 . 182 48.77 2.62. Goosly 900907 7.57 2.49 206 56.84 . 2.66Goosly 900907 4.17 3.09 4.8 143 32.99 11 5.11 1.63Goosly 900907 . 7.42 3.01 49.9 135 26.86 13 7.31 3.41Goosly 900910 0.61 2.86 1.42 9.3 110 29.63 5 0.99 1.23Goosly 900910 1.37 4.02 2.25 20.9 108 25.34 19 4.88 1.75Goosly 900910 1.17 3.1 2.88 24.9 116 42.56 18 3.99 0.66Goosly 900910 1.46 4.02 1.48 3.6 192 16 2.89 1.84Goosly 900910 1.75 3.95 3.68 25.1 140 40.65 26 3.7 0.7Goosly 900910 3.1 4.34 1.43 53.4 177 28.91 18 2.11 1.7Goosly 900910 2.03 4.84 2.62 12.3 170 41.9 20 3.28 0.84Goosly 900910 2.34 4.95 3.36 43.2 170 24 4.51 1.41Goosly 900910 1.16 4.56 3.1 22.9 146 29.65 16 3.89 1.37Goosly 900910 1.02 3.37 2.78 3.8 159 48.85 10 1.79 1.2399Table 26 (continued). Raw data from fish tissue analyses.^dwt= change in weight (grams), Metal=metallothionein concentrationTrough^TreatmentDate^Changeof in weightDeath^(grams)Metallo-thioneinAmot/9GillCuAg/gMuscleCulig/gLiverCugg/gGillZnit9/9MuscleZn$01/9LiverZn119/9LiverCdAR/gGillCd149/9Lu 900905 1.42 2.22 18.6 33.6 12 3.53Lu 900905 0.53 5.1 3.46 5.6 262 27.27 9 2.23 0.72Lu 900905 2.1 8.51 3.03 17.0 168 32.54 11 3.48 0.71Lu 900905 1.05 6.25 4.53 8.3 288 31.31 14 3.13 2.18Lu 900905 0.92 6.68 3.07 14.9 194 31.92 11 3.7 0.3Lu 900905 0.97 5.22 2.51 16.5 270 26.97 7 2.95 0.87Lu 900905 0.97 5.99 3.1 18.4 162 29.11 8 1.82 0.47Lu 900905 1.16 5.77 2.59 26.0 265 32.79 21 2.16 1.38Lu 900905 3.11 8.39 2.62 40.4 221 31.45 6 5.86 1NAME:^CBIOGRAPHICAL INFORMATION MAILING ADDRESS:3 D7 - 301,. Oder^41- 5t4.-^10 din I/a-vieov vor, $C,^7^le I.,.PLACE AND DATE OF BIRTH:^PA^/V..3-c^gio.aEDUCATION (Colleges and Universities attended, dates, and degrees):^LA Woe., ei (,//off^/qgd -/9ez(/'1' 114-4 a^(ggit^g^l3.5. 120^dwie-4-1,0, ;y^61.^7g8 6 /1^. .0,74.40POSITIONS HELD:PUBLICATIONS (if necessary, use a second sheet):AWARDS:Complete one biographical form for each copy of a thesis presentedto the Special Collections Division, University Library.0E.5


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items