a multitrophic approach to monitoring the effects of metal ...paula spencer,* michelle f bowman, and...

A Multitrophic Approach to Monitoring the Effects of MetalMining in Otherwise Pristine and Ecologically Sensitive Riversin Northern CanadaPaula Spencer,* Michelle F Bowman, and Monique G Dube

Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5B3, Canada

(Received 30 October 2007; Accepted 16 March 2008)

ABSTRACTIt is not known if current chemical and biological monitoring methods are appropriate for assessing the impacts of

growing industrial development on ecologically sensitive northern waters. We used a multitrophic level approach to evaluate

current monitoring methods and to determine whether metal-mining activities had affected 2 otherwise pristine rivers that

flow into the South Nahanni River, Northwest Territories, a World Heritage Site. We compared upstream reference conditions

in the rivers to sites downstream and further downstream of mines. The endpoints we evaluated included concentrations of

metals in river water, sediments, and liver and flesh of slimy sculpin (Cottus cognatus); benthic algal and macroinvertebrate

abundance, richness, diversity, and community composition; and various slimy sculpin measures, our sentinel forage fish

species. Elevated concentrations of copper and iron in liver tissue of sculpin from the Flat River were associated with high

concentrations of mine-derived iron in river water and copper in sediments that were above national guidelines. In addition,

sites downstream of the mine on the Flat River had increased algal abundances and altered benthic macroinvertebrate

communities, whereas the sites downstream of the mine on Prairie Creek had increased benthic macroinvertebrate taxa

richness and improved sculpin condition. Biological differences in both rivers were consistent with mild enrichment of the

rivers downstream of current and historical mining activity. We recommend that monitoring in these northern rivers focus on

indicators in epilithon and benthic macroinvertebrate communities due to their responsiveness and as alternatives to lethal

fish sampling in habitats with low fish abundance. We also recommend monitoring of metal burdens in periphyton and

benthic invertebrates for assessment of exposure to mine effluent and causal association. Although the effects of mining

activities on riverine biota currently are limited, our results show that there is potential for effects to occur with proposed

growth in mining activities.

Keywords: Metal mining Benthos Cottus cognatus Environmental effects monitoring

INTRODUCTIONNorthern aquatic ecosystems are considered to be sensitive

to contaminants due to their high latitude and extremeclimate (Schindler and Smol 2006). The reduced ice-freeperiod results in low primary production, slow growth cycles,and reduced energy stores with invertebrates and fishcommonly reproducing every 2 y compared with annualreproduction in southern watersheds (Evans 2000; Evans etal. 2005; Chambers et al. 2006; Lumb et al. 2006). Slowerreproduction cycles leave northern species at an increasedvulnerability to stressors in the environment (Barrie et al.1992). Species diversity tends to be reduced due to the harshclimate, limited opportunity for species migration fromsouthern ecosystems, and short period of time since the lastglaciation (Evans et al. 2005). Most northern ecosystems areconsidered to be highly oligotrophic, with less productivityand species abundance than southern systems (Chambers etal. 2006; Schindler and Smol 2006).

Canada’s north currently is experiencing large growth inindustrial developments and the potential for continueddevelopment is extremely high. In 2005, the value of mineralproduction increased nearly 100% from the mineral produc-tion in 2002 (Natural Resources Canada 2005). As anindication of economic significance, diamond mining, which

occurs exclusively in the Northwest Territories and Nunavut,accounts for 15% of the global market with exports totalling$1.9 billion in 2006 (Mining Association of Canada 2007).Given the significance of industrial development in the north,there is a need to assess the sensitivity of these unique high-latitude ecosystems and to develop tools to track differencesover spatial and temporal scales (Lockhart et al. 1992;Mallory et al. 2006). Establishing a baseline of response toanthropogenic stressors is also critical because the advent ofclimate change and global warming undoubtedly will alter andpotentially magnify responses of aquatic systems to stress(Schindler and Smol 2006). The increased pressures onnorthern ecosystems and their unique characteristics relatedto climate, reduced productivity, and lower species diversityemphasize the importance of effective monitoring programsthat can be used for detecting changes in northern ecosystemsand to provide early warning signals for higher trophic levels.

Under the Fisheries Act, the 2002 Metal Mining EffluentRegulations require metal mines in Canada to conductenvironmental effects monitoring (EEM) to assess effectspotentially caused by their effluents. The EEM programfocuses particular attention on biological monitoring toevaluate whether or not environmental effects are occurring(Walker et al. 2003). An effect is a statistically significantresponse in at least one of the selected endpoints (e.g.,abundance) in comparisons between biological samples takenfrom exposure and reference areas. Canada’s EEM programprovides a nationally consistent monitoring approach for

* To whom correspondence may be addressed: [email protected]

Published on the Web 4/4/2008.

Integrated Environmental Assessment and Management — Volume 4, Number 3—pp. 327–343� 2008 SETAC 327

Orig

inalRese

arch

Review

assessing effects and its methods have influenced monitoringpractice in other arenas including monitoring under provincialpermits and environmental impact assessment (Kilgour et al.2006). However, methods may not be applicable in sensitiveenvironments with low productivity and species abundance(Munkittrick et al. 2000; Walker et al. 2003). Pristinenorthern ecosystems, such as the South Nahanni Watershed,have inherently low species abundance and diversity that mayimpact the utility of traditional monitoring endpoints.

The South Nahanni Watershed includes part of NahanniNational Park Reserve, a United Nations Educational,Scientific and Cultural Organization (UNESCO) WorldHeritage Site, and Canadian Heritage River, and is considereda sensitive northern environment with pristine aquaticenvironments (Halliwell and Catto 2003). Two industrialdevelopments are located on tributaries of the South NahanniRiver: Tungsten Mine and Prairie Creek Exploration Property.These tributaries are ideal for evaluating the applicability ofsouthern monitoring methods for use in northern systemsbecause they are otherwise pristine, and significant impactswere not apparent from previous water quality studies(Halliwell and Catto 2003; Lumb et al. 2006).

Multitrophic level monitoring allows for determination ofbiological interactions and assessment of aquatic responsepatterns in these ecosystems. In the summer of 2006, weconducted a multitrophic level effects monitoring program onboth the Flat River and Prairie Creek in the South NahanniWatershed. Sentinel fish populations, benthic macroinverte-brates, benthic algal communities, water quality, and sedi-ment quality were sampled at upstream reference areas andcompared to near-field and far-field areas exposed to mineeffluents. The objectives of this work were to documentcurrent aquatic conditions in these rivers and any effects dueto mining activities, assess current monitoring methods fornorthern environments, and make recommendations toimprove future effects monitoring in northern waters.

METHODS

Study sites

The South Nahanni River is located in the southwest cornerof the Northwest Territories (Figure 1). The low subarcticclimate in the Nahanni Plateau is characterized by coolsummers (mean temperature is 9 8C) and cold winters (meantemperature is �19.5 8C; Environment Canada 1991). Mostareas are less than 1372 m above sea level but mountainranges reach to over 1800 m (Halliwell and Catto 2003). Theterrain is underlain by Palaeozoic carbonates, and is incised bydeep and narrow valleys. Vegetation is sparse at higherelevations but open stands of black spruce with an understoryof dwarf birch, Labrador tea, lichen, and moss occur in valleysand at lower elevations (Environment Canada 1991). TheSouth Nahanni River flows in a southerly direction 540 kmfrom ice fields near the Yukon-NWT border through theMackenzie Mountains into the Liard River, which thenconverges with the Mackenzie River (Parks Canada 1984).The normal range of flows at the flow gauging stationupstream of Virginia Falls is 55 to 1500 m3/s (Halliwell andCatto 2003).

The South Nahanni River is a UNESCO World HeritageSite and a highly valued national park. Classified as aCanadian Heritage River, the South Nahanni River has waterquality considered to be nearly pristine (Halliwell and Catto

2003). Its unique landscapes and ecosystems only emphasizethe need to ensure that it is protected from potential stressesof development (Inland Waters Directorate 1991; Halliwelland Catto 2003).

The Flat River and Prairie Creek tributaries of the SouthNahanni River have both historical and current industrialactivity that could affect this sensitive environment. The FlatRiver flows into the Nahanni River below Virginia Falls andmay exhibit properties more similar to the Mackenzie River.Distinct water quality changes have been detected along theFlat River as a result of high natural variability in sedimentloads (Halliwell and Catto 2003). The presence of hot springsand mineral springs also increase the variability within the FlatRiver watershed. The Flat River valley is filled with glacial andfluvial deposits through which the river meanders in a well-defined flood plain (EBA Engineering 2002). Limestone,sandstone, and dolostones are the predominant geologicalforms underlying the Flat River (Parks Canada 1984) andserve the basis for tungsten mining in the area. Flow rates onthe Flat River average 247 to 900 m3/s (Halliwell and Catto2003), which is lower than flow rates in tributaries aboveVirginia Falls.

Prairie Creek also flows through deposits of dolostones,limestone, and shale (Halliwell and Catto 2003) containingmineralized veins of zinc, lead, copper, and silver. PrairieCreek originates in and flows through upland and steepcanyon terrain that leads to reduced suspended sedimentscompared to other tributaries in the Nahanni watershed(Halliwell and Catto 2003). Due to the underlying geologicalformations, Prairie Creek precipitates carbonate and otherminerals as well as metals from natural mineral deposits(Indian and Northern Affairs 2001). Flow rates in PrairieCreek are lower than those in both the Flat River and theSouth Nahanni River, ranging from 0.5 m3/s in the wintermonths to 30 m3/s in the summer months (Inland WatersDirectorate 1991).

Tungsten Mine on the Flat River began operations in the1950s and continued to operate until 1986. Increased siltfrom direct discharge during this time remains in the riverdownstream of the mine. In 1986, the mine went into careand maintenance until it reopened in 2001. Tungsten Minecurrently is mining and milling tungsten ore, in the form ofshelite. Tailings are deposited into an exfiltration pond systemto remove the suspended sediments. The exfiltrate leachesthrough the soils and into the Flat River (Figure 2a)approximately 50 km upstream of the confluence with theSouth Nahanni River (Environment Canada 1991; Halliwelland Catto 2003). Limited information currently exists on theorientation and delineation of the exfiltrate into the FlatRiver, which complicates quantification of biological expo-sure to current discharges from the mine.

The Prairie Creek Advanced Exploration Program is locatedapproximately 18 km upstream from the confluence with theSouth Nahanni River (Environment Canada 1991; Halliwelland Catto 2003). The property initially was explored in the1950s and a mill complex and a tailings pond wereconstructed in the 1980s (Figure 2b). Due to financialdifficulties, mining and milling did not proceed at the PrairieCreek property and tailings were not generated. CanadianZinc Corporation took over the property in 1995 and hasbegun an advanced exploration program for base metals (i.e.,lead, zinc, copper, and silver). A portal was constructed at thesite for exploration and potential mining purposes. The portal

328 Integr Environ Assess Manag 4, 2008—P Spencer et al.

discharges water to a polishing pond and subsequently to acatchment pond. The water from the catchment pond thendischarges to Prairie Creek and has potential to be high inseveral metals (i.e., zinc, lead, and cadmium).

Control-impact monitoring surveys were conducted in thesummer of 2006 at reference, near-field (high exposure), andfar-field (low exposure) sites on the Flat River and PrairieCreek using methods outlined in the Metal Mining EffluentRegulations EEM Guidance Document (Environment Canada2002). In the absence of tracer studies that document thepath of flow from the exfiltration basins into the river, thelocation of the exfiltration basins at Tungsten Mine (Pond 3and Pond 4, Figure 2a) was considered the area of develop-ment activity. At Prairie Creek, the portal discharge intoHarrison Creek was considered the point source discharge(Figure 2b). From August 27 to September 2, 2006, sentinelfish populations, benthic macroinvertebrates, benthic algalcommunities, water quality, and sediment quality were

sampled at upstream reference areas located 3 km above thezone of influence and compared to near-field and far-fieldareas exposed to mine effluents. Far-field areas were locatedapproximately 2 km downstream from the near-field sites.

Water chemistry and effluent characterization

Dissolved oxygen, ammonia, conductivity, and pH weremeasured in situ, whereas samples for total metals, generalwater chemistry (conductivity, pH, total suspended solids),and nutrients (nitrate, total phosphorus) were analyzed in alaboratory. Dissolved oxygen was measured with a WTWDissolved Oxygen Pocket Meter (Model Oxi 330i, WTW,West Wareham, MA, USA). Ammonia was measured with aHanna HI 93733 Ammonia meter (Hanna InstrumentsCanada, Laval, QC, Canada). Conductivity and pH weremeasured with an Oakton 300 series meter (Oakton Instru-ments, Vernon Hills, IL, USA). Single grab water sampleswere collected from a location in stream at each reference,

Figure 1.Map showing the South Nahanni watershed located in the southwest corner of the Northwest Territories and southeast corner of the Yukon Territory.Nahanni National Park is located within the South Nahanni watershed.

Monitoring Rivers in Northern Canada—Integr Environ Assess Manag 4, 2008 329

near-field, and far-field site. Effluent samples were collected

from points of discharge at both Tungsten Mine (Ponds 3 and

4) and the Prairie Creek Exploration Property (Catchment

Pond) to characterize effluents. Water samples were kept at

48C until they were shipped on ice to Taiga Environmental

Laboratory in Yellowknife, NWT. Samples for total metals

were unfiltered and acidified and analyzed using the

inductively coupled plasma–mass spectrometric method.

Conductivity was analyzed using the radiometric method,

pH using electrometric methodology, and total suspended

solids using a gravimetric method. Samples for nitrate and

total phosphorus were unfiltered and analyzed by ion

chromatography and colorimetric determination, respectively.

Sediment samples

A sediment sample was collected at each sampling area

(reference, near-field, and far-field). Grab samples were

collected in 500-mL glass jars and analyzed for total metals,

particle size, and total organic carbon at Taiga Environmental

Laboratory, Yellowknife, NWT, Canada. Total metals and

Figure 2. (a) Aerial view of Tungsten Mine discharge: Tailings Ponds 3, 4, and 5 comprise an exfiltration system for discharge to the Flat River. Discharges to thetailings ponds include mine water, process water, and sewage treatment plant effluent. (b) Aerial view of the Prairie Creek Exploration Property. Mine waterdischarge from the portal main adit enters a polishing pond and catchment pond before release into Prairie Creek.


minerals were prepared according to US EnvironmentalProtection Agency (USEPA) 3050B methodology for aciddigestion and analyzed by inductively coupled plasma–massspectrometric method or inductively coupled plasma–atomicemission spectroscopy, with the exception of mercury, whichwas analyzed by cold vapor atomic absorption (USEPA 245.5method). Particle size analysis was conducted using Carter’smethod 47.3. Total organic carbon was analyzed using a drycombustion method.

Benthic algal sampling

Nine 15- to 25-cm diameter cobbles were selectedrandomly per reference, near-field, and far-field site at eachmine. For chlorophyll a analysis, 1 scraping of algae fromwithin a circular template (13.2 cm2) was taken from eachcobble. Three templates then were pooled (representing atotal sampled area of 39.6 cm2 pooled from 3 rocks), resultingin 3 replicate chlorophyll a samples per site. For algaltaxonomy, 1 circular template was taken from 3 rocks andpooled, resulting in a single taxonomy sample per site.Samples were placed in scintillation vials and preserved in5% buffered formalin with a final volume of 21 mL(taxonomy samples) or frozen in stream water (chlorophylla samples). For taxonomic analysis, 2-mL subsamples ofepilithon suspension were sonicated for 10 to 20 s using aSonifer Cell Disruptor (model w140; Findlay et al. 1999) andgravity settled for 24 h in an Utermohl chamber (PhycoTech,St. Joseph, MI, USA). Cells were identified, counted, andmeasured from random fields until 100 cells of the dominantspecies were found. Estimates of cell volume for each specieswere obtained by measurements of up to 50 cells of anindividual species and applying the geometric formula bestfitted to the shape of the cell (Rott 1981). Methods forflorometric analysis of chlorophyll a are outlined in Bowmanet al. (2005).

Benthic invertebrate sampling

Five replicate benthic invertebrate samples were collectedfrom each of the reference, near-field, and far-field sites,consistent with EEM methods. A U-net (0.101 m3) was usedto collect 3 subsamples that were pooled for each replicate fora total area of 0.303 m3 at each replicate station. Fiverandomly selected replicate stations were sampled withineach area to provide an area delineated estimate of benthicinvertebrate communities. Material was preserved in 95%ethanol and sent to Cordillera Consulting (Summerland, BC,Canada), where they were sorted, identified to the lowestpractical concentration, and counted. Following EEM guide-lines, family taxonomic concentration was used. Studiesindicate that changes in community composition and com-munity responses to disturbances are maintained whenspecies-level data are aggregated to family (e.g., Bowmanand Bailey 1997). In addition, the predictability of relation-ships between benthic communities and physical variables canbe stronger when family data are used relative to when specieslevel data are used (e.g., Reynoldson et al. 2001). The benthicinvertebrate endpoints assessed included total invertebratedensity (number/m2), taxa richness, Simpson’s Diversityindex, and the Bray-Curtis index. These benthic invertebrateendpoints were selected because they provide quantitativedata and important information on community structure andcomposition (Environment Canada 2002). Total invertebratedensity indicated the number of individuals in all families at

the selected sites. Taxon richness represented the number oftaxonomic families collected for each replicate. Simpson’sDiversity index takes into account abundance patterns as wellas taxonomic richness by calculating the proportion ofindividuals in each taxonomic family and their relativeabundance. The Bray-Curtis index is a dissimilarity index,which is a measure of the percentage of difference betweensites.

Fish collection and sampling

Previous work in the Nahanni National Park has indicatedthat slimy sculpin (Cottus cognatus) are the most abundantfish species and are likely the only possible sentinel species foreffects monitoring due to low abundance of other species(Sigma Resource Consultants 1978; EBA Engineering 2002;Halliwell and Catto 2003). Watershed studies conducted byMF Bowman et al. (University of Saskatchewan, unpublisheddata) in 2006 concurrent with the sampling described hereinalso confirmed that slimy sculpin were the only species thatcould be used as a sentinel based on monitoring guidanceunder the EEM Program (Environment Canada 2002).Although arctic grayling (Thymallus arcticus), bull trout(Salvelinus confluentus), and round whitefish (Prosopiumcylindraceum) were captured in each of the rivers, theirpresence was sporadic, and neither lethal nor nonlethalsampling protocols could be implemented as specified inthe EEM guidance document (20 males and 20 females forlethal sampling; approximately 100 fish for nonlethalsampling).

Slimy sculpin are a sedentary species and studies haveshown limited movement between sites located 200 m apartacross the Saint John River in New Brunswick, Canada (Grayet al. 2004, Galloway et al. 2005). Sites in this study werechosen (2000–3000 m apart) to minimize sculpin movementbetween sites. Sculpin were collected from upstream, near-field, and far-field sites with backpack electrofishers. Totaleffort to collect 40 fish was measured in electrofishingseconds. Electrofishing seconds for reference, near-field, andfar-field were 12480, 874, and 6510 s, respectively. Twentymales and 20 females collected at each sampling area weresacrificed according to lethal sampling protocols. Nonlethalsampling for this species also was conducted for comparativepurposes, although results are not presented here.

Fish were collected and stored alive temporarily in bucketsat each site. Upon completion of electrofishing, captured fishwere immediately measured for total length and total weight.Sculpin were transported back to the lab at the mine site andeuthanized in clove oil prior to dissection. Liver and gonadtissues were removed and weighed (to 0.001 g). Visualobservations or presence of parasites were noted duringdissections. Fish were analyzed for the standard EEM fishsurvey endpoints including liver weight, gonad weight, length,weight, condition estimates, age, fecundity, and egg size.

Total metal concentrations in slimy sculpin tissues wereinvestigated in equal numbers of adult males and femalesfrom each site. Flesh muscle and liver tissue were analyzedfrom fish collected in the Flat River. Metals were examinedonly in muscle in fish from Prairie Creek due to lower bodyweights of the fish captured at the reference, near-field, andfar-field sites. Lower body weights resulted in insufficientamounts of liver tissue. Mean sample weights of muscle andliver were 3.5 and 0.3 g, respectively. Samples were analyzedat a commercial laboratory (SRC Laboratories, Saskatoon, SK,


Canada) for a suite of 27 metals that were determined by

nitric and hydrochloric acid digestion and inductively coupledplasma–mass spectrometer or inductively coupled plasma–optical emission spectrometer following USEPA method200.3, except that HCl was used instead of H2O2.

Ovaries from each female were placed in 10% buffered

formalin. For fecundity estimates, 0.02 to 0.10 g of follicles(42–217 follicles) were removed from the middle of ovaries,separated, and photographed. Image analysis (Image-Pro Plus;Media Cybernetics, Washington, MD, USA) was used to

count and measure the mean diameter of vitellogenic follicles.Counts of vitellogenic follicles were made and fecundityextrapolated as the number of follicles per gram of carcassweight.

Slimy sculpin were aged by counting annuli on sagittalotoliths according to the methods of Gibbons et al. (1998)and modified so that otoliths from fish larger than 6 cm, orthose where annuli were unclear, were ground to the nucleusand charred over an alcohol flame and read. Briefly, sagittal

otoliths were removed surgically from each fish and stored inpaper envelopes before being read under water with incidentlight on a dissecting microscope. Whole otoliths were read 3times, charred otoliths twice, and Flat River otoliths were

read independently by 2 experienced otolith readers asfurther verification of assignment of ages.

Statistical analysis

One-way analysis of variance (ANOVA) with site (i.e.,reference, near-field, far-field) as the factor was conducted foralgal biomass, total benthic invertebrate density, invertebratetaxon richness, Simpson’s Diversity index, Bray-Curtis index,

and each of the fish tissue metals. Assumptions of ANOVAwere checked for each variable prior to conducting ANOVA.Normality was confirmed using the Shapiro-Wilkes test andequality of variances was compared using Levene’s test. Alltests were performed using Systat 11 (Systat Software, SanJose, CA, USA). For all analyses, p , 0.05 was consideredsignificant. Tukey’s post hoc test was used to determine sitedifferences when p , 0.05. Following EEM guidelines,condition, liver weight, gonad weight, and size-at-age wereanalyzed using analysis of covariance (ANCOVA) for all fish.For condition, length was used as a covariate against bodyweight. Liver and gonad weights were analyzed using bodyweight as a covariate. Length-at-age ANCOVA was used forgrowth comparisons. In female fish, ANCOVA was per-formed for egg size and fecundity using body weight as acovariate. Assumptions were confirmed by checking fornormality and equality of variances as described above.Assessment of linearity and interactions also was performedprior to ANCOVA. Mean age for all fish was analyzed usingANOVA after checking assumptions as described above.

RESULTS

Flat River

Concentrations of several metals were higher in the settlingponds compared to river concentrations and increased in thewater column downstream of the Tungsten Mine (Table 1).Concentrations of iron in Flat River water were 143% higherin near-field compared to reference sites, remained elevated atthe far-field site, and at both exposure sites were above theCanadian Council of Ministers of the Environment (CCME)

Table 1. Total metal concentrations in Flat River water (lg/L) and sediment (lg/g) for reference, near-field, and far-fieldsites; effluent in Ponds 3 and 4 (n ¼ 1 sample per site); and associated Canadian Council of Minister of the Environment

(CCME) guidelinesab

Parameter

Detection limit Upstream Near-field Far-field Pond 3

Water Sediment Water Sediment Water Sediment Water Sediment Water Sediment

Aluminum 0.60 20 35 15000 42 10900 53 11700 4500 19200

Arsenic 0.20 0.2 0.4 37.3 0.4 23.6 1.0 35.3 14.90 8.4

Cadmium 0.10 0.01 0.1 0.65 0.1 0.57 0.1 0.78 5.30 1.84

Chromium 0.30 0.5 0.6 27.3 0.3 16.4 15.6 17.5 19.60 21

Copper 0.30 1 1.0 25 1.1 750 2.20 710 13400 1280

Iron 50 100 143 42300 347 108000 326 88900 86100 69800

Lead 0.10 0.1 2.3 21.7 0.1 11.8 67.60 12 39.00 13

Manganese 0.10 10 8.5 556 24.9 463 23.40 441 362.00 2,180

Mercury 0.02 0.01 0.02 0.01 0.02 0.01 0.02 0.01 3.46 0.02

Nickel 0.10 0.5 3.6 52.3 2.9 29.5 2.50 32.7 22.10 20.4

Selenium 1.00 0.3 1.0 0.5 1.0 3.1 1.0 2.7 2.00 1.5

Tungsten 0.50 0.5 0.8 1.3 2.80 323 3.60 243 7,210.00 326

Uranium 0.10 0.5 0.76 1.7 0.98 1.7 0.90 2.4 3.10 1.3

Zinc 0.10 1.0 4.51 151 3.56 111 5.00 134 676.00 251

NA ¼ not available.a Water quality guidelines are provided from the CCME for the protection of freshwater aquatic life (CCME 2003).b Sediment quality guidelines are provided from the CCME Interim Sediment Quality Guidelines (CCME 2003).


Water Quality Guideline for the Protection of Aquatic Life(300 lg/L; Table 1). Iron was present in extremely highconcentrations in both Pond 3 and Pond 4. Copper showed anincreasing trend from reference site to the far-field site butremained below CCME water quality guidelines of 2 to 4 lg/L. Copper was elevated in water collected from both Pond 3and Pond 4. Aluminum and tungsten also demonstrated anincreasing trend from the reference site to the far-field site,with aluminum above CCME guidelines at all sites. Alumi-num and tungsten in Ponds 3 and 4 were orders of magnitudehigher in concentration compared to sampling sites in the FlatRiver. Few differences were found in cadmium, mercury, andselenium concentrations among sites, although both mercuryand cadmium were elevated above river concentrations inPonds 3 and 4. Selenium was elevated above river concen-trations in Pond 3 but not Pond 4. Zinc, chromium, and leaddecreased at the near-field site and increased above referenceconcentrations at the far-field site. Concentrations in Ponds 3and 4 were not elevated above concentrations in the far-fieldsite for chromium and lead but zinc concentrations in theponds were well above river concentrations. Manganese anduranium concentrations peaked at the near-field site and wereelevated above reference site concentrations at the far-fieldsite. Nickel demonstrated a decreasing trend from thereference site to the far-field site.

Consistent with the increase in total iron in water fromreference relative to both near-field and far-field sites, iron insediment showed an increase of 155% from reference to near-field and remained elevated above reference concentrations atthe far-field site (Table 1). Copper in sediment increased2900% from reference to both near-field and far-field sites andwas above CCME Interim Sediment Quality guidelines atboth exposure sites. Tungsten in sediment increased 250%from reference to near-field sites and remained high at the far-

Table 1. Extended

Pond 4 CCME guidelines

Water Sediment Water Sediment

6,580 11000 5–10 NA

2.60 21.6 5 5.9

0.30 0.39 0.017 0.6

11.00 16.2 9.9 37

183.00 339 2–4 35.7

26,200 73400 300 NA

6.90 14.4 1–7 35

629.00 326 NA NA

4.07 0.01 0.026 0.17

8.50 28.5 25–150 NA

0.80 1.7 NA NA

3,760.00 238 1 NA

3.28 2.0 NA NA

31.60 106 30 123Table

2.Gen

eralwaterch

emistryresultsfortheFlatRiverandPrairie

Creek

colle

cted

atreference,nea

r-field,andfar-fieldsites

Parameter

Detection

limit

FlatRiver

upstream

FlatRiver

near-field

FlatRiver

far-field

Pond3

Pond4

Prairie

Creek

upstream

Prairie

Creek

near-field

Prairie

Creek

far-field

Catchment

pond

Nitrate

(mg/L)

0.01

0.02

0.03

0.03

2.27

0.32

0.09

0.10

0.08

—

Totalphosp

horous(lg/L)

0.01

,0.01

,0.01

,0.01

0.25

0.15

,0.01

,0.01

,0.01

0.2

Conductivity(lS/cm)

0.4

224

253

235

1090

983

452

459

460

826

pH

—7.99

7.98

88.02

8.21

8.44

8.46

8.42

8.3

Totalsusp

ended

solid

s(m

g/L)

3000

46

,3

911

973

,3

,3

,3

,3


field site. Tungsten also was present in high concentrations in

sediments of Pond 3 and Pond 4. Selenium in sediment

increased at the near-field and far-field sites and concen-

trations were higher than those observed in Ponds 3 and 4.

Uranium in sediment peaked at the near-field site and

decreased again at the far-field site but concentrations still

were above those observed at the reference site. All other

sediment metals decreased from the reference site to theexposure sites.

General chemistry results for water column samplesshowed that nitrate and conductivity were elevated in Pond3 and Pond 4 as well as slightly elevated at the near-field andfar-field sites (Table 2). Sediment composition samplescollected at the 3 sampling sites along the Flat River showedthat the amount of silt at exposure sites was higher than at

Figure 3. Algal endpoints collected in the Flat River and Prairie Creek at reference, near-field, and far-field sites: (a) mean (6 standard error [SE]) algal biomassreported as chlorophyll a (lg/cm2; analysis of variance [ANOVA] results for Flat River: n¼3, F¼22.355, p¼0.002; results for Prairie Creek: n¼3, F¼0.400, p¼0.687); (b) algal richness reported as number of families/cm2 (n¼1); (c) algal diversity reported as a diversity index (n¼1); (d) % A. minutissima. Bars with likeletters were not significantly different.


reference sites but these differences were not statisticallysignificant due to high variability in sediment compositionwithin sites (mean 6 standard deviation percent silt atreference, near-field, and far-field sites was 6.7% 6 8.1%,19.2% 6 10.0%, and 17.7 % 6 18.6%, respectively; 1-wayANOVA, p ¼ 0.30).

Algal abundance in the Flat River, measured as chlorophylla, increased significantly (p ¼ 0.002) from the reference siteto the far-field site (Figure 3). Mean chlorophyll a at thereference, near-field, and far-field sites were 0.46, 1.72, and4.58 lg/cm2

, respectively. Algal richness showed a decrease atthe near-field site and showed recovery at the far-field site butdid not return to reference. Diversity followed the trend ofalgal richness with a decrease at the near-field site andsubsequent increase at the far-field site. Decrease ofAchnanthes minutissima at the near-field site was substantial.

Algal communities in the Flat River showed distinctdifferences in taxonomic composition at the near-field sitecompared to the reference and far-field site (Figure 4). At thenear-field site, chlorophytes were not detected, diatomabundance decreased, and cyanophyte abundance increasedrelative to the reference site. At the far-field site, chlorophyteswere detected in the algal community and diatoms werepresent in a higher abundance than at the reference site. Incontrast, cyanophytes decreased below those observed at thereference and near-field sites.

The near-field site on the Flat River had a significantlyhigher benthic invertebrate density compared to the referenceand far-field sites (p ¼ 0.017; Figure 5). A statisticallysignificant difference in the Bray-Curtis index was observedbetween the reference and near-field sites as well as betweenthe reference and far-field sites (p , 0.001). No significantdifference was found in richness and diversity of benthicmacroinvertebrates between the reference and 2 exposuresites. The proportion of Diptera substantially increased at thenear-field site (Figure 6) with significant Chironomidaeproliferation. Diptera remained elevated above reference atthe far-field site. Mayflies decreased in density at the far-fieldsite. Caddisflies demonstrated an increasing trend from thereference site to the far-field site, whereas stoneflies showed asignificant decrease at the exposure sites compared to the

reference site, with the lowest density of stoneflies at the far-field site.

The condition of slimy sculpin in the Flat River, as definedas length against weight, showed a significant increase (p ¼0.002) at the far-field site in male fish compared to thereference and near-field site (Table 3). Condition also wassignificantly higher (p¼ 0.006) at the far-field site for femaleswhen compared to the near-field site but was not significantlydifferent from the reference site. Size-at-age analyses indi-cated an increase in growth at the far-field site compared tothe near-field and reference sites in female fish (p , 0.001).Male fish, however, showed a decrease in size-at-age at thenear-field site and an increase at the far-field site (p , 0.001).Female fish were older at the far-field site compared to thereference and near-field sites. Age of male fish did not varysignificantly among sites. Gonad weights and liver weights formale and female fish were not significantly different betweenthe 3 sites. Egg size significantly increased from the referencesite to the near-field site (p ¼ 0.001) but no significantdifference was observed in fecundity between the sites.

Concentrations of metals in liver tissue of Flat River fishindicated a significant accumulation of copper and iron in fishat the near-field site (Table 4) when compared to fish fromthe reference and far-field sites. Copper concentrations in thefish livers at the reference site were 61% lower than thosecollected at the near-field site (p¼ 0.002). Iron in liver tissuewas 275% higher in fish from the near-field site compared tothe reference site (p ¼ 0.001). Both copper and iron in livertissue decreased at the far-field site but remained abovereference site concentrations. Selenium in liver tissue waslower at the near-field and far-field sites (p ¼ 0.006).Cadmium concentrations in liver tissue showed a decreasingtrend from the reference site to the far-field site (p¼ 0.001).Arsenic concentrations in liver tissue decreased at the near-field site compared to the reference site (p ¼ 0.022) butincreased again at the far-field site above reference concen-trations.

Zinc (p¼ 0.004) and mercury (p¼ 0.035) showed a strongdecreasing trend in concentration from the reference site tothe far-field site. Arsenic (p ¼ 0.010) and nickel (p ¼ 0.031)concentrations in fish muscle decreased at the near-field site

Figure 4. Algal taxonomy for the Flat River and Prairie Creek at reference, near-field, and far-field sites. Algal community was compared by analyzing algal cellsper sample.


compared to the reference site but increased again at the far-

field site. Neither copper (p ¼ 0.210) nor iron (p ¼ 0.147)

increased significantly in muscle tissue collected at the near-

field site when compared to the reference site despite

accumulation of these 2 metals in liver tissue.

Prairie Creek

Total metals measured in the water column that occurred in

higher concentrations in the catchment pond than in Prairie

Creek included cadmium, lead, manganese, mercury, and zinc

(Table 5). Detection limits for metals measured in the

catchment pond were higher than detection limits for water

samples collected from reference, near-field, and far-field sites

as a result of the presence of salts (dissolved solids) above

0.2% in catchment pond water (Warnken 1999); samples with

high dissolved solids routinely are diluted in the lab due to

interference with the plasma spectrometer’s ability to

determine trace metal concentrations. Thus, several metals

were measured as nondetectable in the catchment pond yet

detection limits were generally higher than concentrations

Figure 5. Benthic invertebrate endpoints collected in the Flat River and Prairie Creek at reference, near-field, and far-field sites: (a) mean (6standard error [SE])invertebrate density reported as individuals/m2 (analysis of variance [ANOVA] results for Flat River: n¼5, F¼5.851, p¼0.017; results for Prairie Creek: n¼5, F¼1.572, p¼0.247); (b) mean (6 SE) invertebrate richness reported as number of families per sample (ANOVA results for Flat River: n¼5, F¼1.367, p¼0.292; forPrairie Creek: n¼ 5, F¼ 3.835, p¼ 0.052); (c) mean (6 SE) invertebrate diversity reported as Simpson’s Diversity index (ANOVA results for Flat River: n¼ 5, F¼2.065, p¼ 0.170; for Prairie Creek: n¼ 5, F¼ 2.065, p¼ 0.170; (d) mean (6 SE) community dissimilarity reported as Bray-Curtis index (ANOVA results for FlatRiver: n ¼ 5, F ¼ 163.34, p ¼ 0.000; for Prairie Creek: n ¼ 5, F ¼ 2.351, p¼ 0.138). Bars with like letters were not significantly different.


Table

3.Mea

n(6

standard

deviation)forslim

ysculpin

metricsandtheresultsofanalysisofvariance

(ANOVA)andanalysisofco

variance

(ANCOVA)forea

chriver(sites

withthe

sameletterswerestatistically

equivalent).Mea

nagewasanalyzedusingANOVA.Weight(rep

resentedbyco

nditionfactor)

wasanalyzedusingANCOVA

withlength

asa

covariate.Gonadsize

(representedbyGonadosomaticindex)

andliver

size

(rep

resentedbyLiversomaticindex)wereanalyzedwithANCOVAusingbodyweightasaco

variate.

Slopes

amongsiteswerenotsignificantlydifferent(i.e.,p

.0.05)in

anyoftheANCOVAanalyses(FlatRiver

reference:Malesn¼19,femalesn¼26;FlatRiver

nea

r-field:Malesn

¼30,femalesn¼22;FlatRiver

far-field:Malesn¼23,femalesn¼22;Prairie

Creek

reference:Malesn¼8,femalesn¼17;Prairie

Creek

nea

r-field:Malesn¼18,femalesn¼14;

Prairie

Creek

far-field:Malesn¼

21,femalesn¼

10)

Sex

Parameter

FlatRiver

upstream

FlatRiver

near-field

FlatRiver

far-field

pva

lue

Prairie

Creek

upstream

Prairie

Creek

near-field

Prairie

Creek

far-field

pva

lue

Male

Mea

nage

3.342

60.138A

3.860

60.151A

3.848

60.223A

0.089

3.625

60.350A

4.324

60.261A

3.912

60.322A

0.343

Conditionfactor

0.991

60.021A

0.893

60.020A

1.015

60.024B

0.002

0.802

60.026A

0.909

60.020B

0.900

60.021B

0.008

Liversomaticindex

1.352

60.093A

1.356

60.097A

1.544

60.183A

0.897

1.348

60.111A

1.206

60.093A

1.762

60.168A

0.062

Gonadosomaticindex

1.665

60.152A

1.629

60.167A

1.695

60.194A

0.283

1.140

60.271A

0.880

60.162A

1.348

60.179A

0.327

Female

Mea

nage

3.192

60.206A

3.682

60.224A

4.727

60.315B

,0.001

3.50

60.174A

3.731

60.231A

4.800

60.539B

0.023

Conditionfactor

0.912

60.023A

0.859

60.027A

0.918

60.013B

0.006

0.768

60.024A

0.901

60.045B

0.904

60.043B

0.001

Liversomaticindex

1.801

60.143A

1.733

60.129A

2.295

60.110A

0.296

1.414

60.108A

1.733

60.240A

1.660

60.188A

0.160

Gonadosomaticindex

1.727

60.242A

2.383

60.290A

3.420

60.160A

0.172

1.247

60.183A

0.940

60.266A

1.232

60.285A

0.158


measured at any riverine sites. Also, changes in sampleviscosity and surface tension with samples containing highdissolved solids (especially those exceeding 1500 mg/L) orhigh acid concentrations can cause physical interferences.

The same detection limits were used for samples collectedat reference, near-field, and far-field sites, facilitating compar-isons across sites. Water quality results indicate that total zincincreased from 2.4 to 12.3 lg/L (400% increase) from thereference site to near-field site but remained below CCME

aquatic quality guidelines for the protection of aquatic life (30lg/L). Zinc concentrations decreased to 6.8 lg/L at the far-field site. Total aluminum concentrations also increased at thenear-field site (12.6 lg/L) compared to the reference site (7.9lg/L) but decreased at the far-field site (7.3 lg/L). All totalaluminum values fell within the range identified in theguidelines for protection of aquatic life. A slight increase intotal copper was seen at the near-field site but all copperconcentrations were well below aquatic life guidelines (2–4

Table 4. Tissue metal concentrations (lg/g) for the Flat River and Prairie Creek and one-way analysis of variance (ANOVA)results (means with the same letters were not significantly different). An ANOVA was performed for each river and eachtissue type (n ¼ 5 for Flat River flesh metals; n ¼ 5 for Prairie Creek flesh metals; and n ¼ 6 for Flat River liver metals)

Parameter

Flat River upstream Flat River near-field

Muscle Liver Muscle Liver

Aluminum 4.40 6 2.40 A 7.54 6 0.759 A 2.38 6 0.67 A 11.83 6 2.60 A

Arsenic 0.20 6 0.032 A 0.13 6 0.007 A 0.17 6 0.015 A 0.103 6 0.008 A

Cadmium 0.008 6 0.002 A 1.042 6 0.101 A 0.005 6 0.000 A 0.850 6 0.102 A

Chromium NM 0.050 6 0.000 A NM 0.175 6 0.072 A

Copper 0.578 6 0.081 A 3.880 6 0.174 A 0.553 6 0.210 A 6.233 6 0.508 B

Iron 11.58 6 3.873 A 60.2 6 9.162 A 16.33 6 2.667 A 225.5 6 22.381 B

Lead 0.017 6 0.004 A 0.021 6 0.007 A 0.012 6 0.003 A 0.043 6 0.019 A

Manganese 0.728 6 0.095 A 1.960 6 0.129 A 0.610 6 0.031 A 1.983 6 0.140 A

Mercury 0.112 6 0.033 A 0.001 6 0.001 A 0.048 6 0.032 A 0.003 6 0.002 A

Nickel 0.166 6 0.044 A 0.288 6 0.032 A 0.063 6 0.003 B 0.242 6 0.048 A

Selenium 0.820 6 0.058 A 6.80 6 0.636 A 0.850 6 0.034 A 4.12 6 0.566 B

Tungsten 6.660 6 4.836 A 100.0 6 62.780 A 4.600 6 1.455 A 61.667 6 9.358 A

Uranium 0.001 6 0.001 A 0.004 6 0.002 A 0.001 6 0.001 A 0.003 6 0.000 A

Zinc 17.4 6 1.166 A 38.60 6 3.847 A 13.0 6 0.365 B 40.67 6 1.944 A

NM ¼ not measured.

Figure 6. Benthic taxonomy for Flat River and Prairie Creek at reference, near-field, and far-field sites. Ephemeroptera, Plecoptera, Trichoptera, and Dipterawere compared using number of invertebrates/m2.


lg/L). Few differences in cyanide, nutrient concentrations, orgeneral water chemistry were noted among the sites on PrairieCreek (Table 2).

Arsenic in sediment was the only metal that increased inconcentration at the near-field site relative to the referenceand far-field sites, although values were slightly above theCCME interim sediment quality guideline at all locations(Table 5). Concentrations of cadmium and zinc in sedimentcollected at the 3 sites did not differ. All other metalsdemonstrated a decreasing trend from the reference site to thefar-field site.

Algal richness decreased at the exposure sites compared tothe reference site and diversity also showed a decrease at bothexposure sites (Figure 3). Abundance of A. mimutissimadecreased at the near-field site. A significant difference inchlorophyll a was not observed between the reference, near-field, and far-field sites. Only diatom species were detected atthe reference site (Figure 4). Diatom densities decreased atthe near-field site and increased again at the far-field site withdensities slightly below those measured at the reference site.Cyanophytes became dominant at the near-field site andsubsequently decreased again at the far-field site. Limiteddensities of chlorophytes were present at the near-field siteand were not detected at the far-field site.

Benthic macroinvertebrate richness increased significantly(p¼ 0.050) at both exposure sites compared to the referencesite in Prairie Creek (Figure 5). Benthic invertebrate densityand diversity were not significantly different betweenreference, near-field, and far-field sites, although there seemedto be a slight increasing trend from the reference site to thefar-field site. The Bray-Curtis index did not indicate signifi-cant dissimilarity between sites. An increasing trend was

noted in the abundance of mayflies and true flies from thereference to the far-field site (Figure 6).

Slimy sculpin condition, as defined as length against weight,showed a significant increase at both the near-field and far-field sites in male (p ¼ 0.008) and female (p ¼ 0.001) fishcompared to the reference site (Table 3). Size-at-age analyseson male fish indicated that fish had higher growth rates at thefar-field site compared with the near-field and reference sites(p¼0.043) but significant differences in growth were not seenin female fish. However, mean age analyses in femalesindicated fish at the far-field site were older (p¼ 0.023) thanfish at the reference and near-field site, but no differenceswere detected in mean age of male fish between the 3 sites.Gonad weights and liver weights for male and female fishwere not significantly different between the 3 sites. Fecunditydid not differ between the 3 sites but eggs were significantlylarger at the near-field site (p ¼ 0.041).

Metal concentrations in slimy sculpin muscle from PrairieCreek were not significantly different between the referencesite and the near-field site. At the far-field site, however,aluminum (p ¼ 0.024), arsenic (p ¼ 0.002), cadmium (p ¼0.044), copper (p¼ 0.032), iron (p¼ 0.018), and nickel (p ¼0.011) were significantly lower when compared to thereference and near-field exposure sites. Concentrations oflead, manganese mercury, selenium, and zinc did not differbetween sites.

A high number of parasites were detected in Prairie Creek(Table 6). Parasites were identified as Ligula intestinalis by T.Dick (University of Manitoba, Winnipeg, MB, Canada,unpublished data). Parasite body burdens and percentoccurrence did not differ between the 3 sites on PrairieCreek (Table 6).

Table 4. Extended

Flat River far-field p valuePrairie Creekupstream

Prairie Creeknear-field

Prairie Creekfar-field p value

Muscle Liver Muscle Liver Muscle Muscle Muscle Muscle

3.82 6 0.487 A 12.84 6 3.679 A 0.572 0.369 1.790 6 0.530 A 2.040 6 0.326 A 0.650 6 0.272 B 0.024

0.280 6 0.018 B 0.166 6 0.022 B 0.010 0.022 0.172 6 0.013 A 0.142 6 0.017 A 0.066 6 0.009 B 0.002

0.004 6 0.001 A 0.398 6 0.065 B 0.325 0.001 0.021 6 0.002 A 0.028 6 0.003 A 0.016 6 0.002 B 0.044

NM 0.710 6 0.576 A NM 0.226 NM NM NM NM

0.488 6 0.094 A 4.560 6 0.277 C 0.210 0.002 0.335 6 0.014 A 0.340 6 0.020 A 0.254 6 0.023 B 0.032

19.80 6 1.960 A 161.8 6 31.181 C 0.147 0.001 9.525 6 0.881 A 11.56 6 0.717 A 6.82 6 0.635 B 0.018

0.015 6 0.007 A 0.022 6 0.014 A 0.495 0.488 0.130 6 0.025 A 0.172 6 0.047 A 0.077 6 0.034 A 0.231

0.604 6 0.064 A 2.180 6 0.206 A 0.437 0.591 0.335 6 0.018 A 0.384 6 0.032 A 0.312 6 0.024 A 0.160

0.004 6 0.001 B 0.001 6 0.001 A 0.035 0.181 0.028 6 0.018 A 0.066 6 0.024 A 0.078 6 0.021 A 0.070

0.128 6 0.073 A 0.710 6 0.500 A 0.031 0.436 0.095 6 0.003 A 0.102 6 0.030 A 0.024 6 0.007 B 0.011

0.840 6 0.068 A 4.08 6 0.403 B 0.764 0.006 1.250 6 0.065 A 1.060 6 0.068 A 1.160 6 0.087 A 0.465

6.560 6 3.323 A 148.40 6 73.701 A 0.632 0.514 NM NM NM NM

0.001 6 0.001 A 0.004 6 0.002 A 0.327 0.595 NM NM NM NM

11.6 6 0.678 B 33.80 6 1.934 A 0.004 0.062 43.5 6 3.279 A 64.60 6 4.273 A 49.6 6 8.244 A 0.078


DISCUSSIONDifferences in the water and sediment chemistry, benthos,

and sculpin endpoints in both the Flat River and Prairie Creekdownstream of mining activity generally were indicative ofmild eutrophication, although there was also evidence ofcontaminant effects. Differences indicative of eutrophicationthat were observed downstream of the mine on the Flat Riverwere increased algal abundance and increased sculpin egg size.Elevated concentrations of iron, copper, and tungsten in riverwater and sediments that were above national guidelines;differences in algal and benthic macroinvertebrate communitycomposition; decreased size-at-age in male sculpin; andelevated concentrations of copper and iron in liver tissue ofsculpin were indicative of effects of mining activity on the FlatRiver. Similarly, evidence of mild eutrophication downstreamof mining activity in Prairie Creek included increased benthicmacroinvertebrate taxa richness and improved condition ofsculpin of both sexes along with increased egg size at the near-field site. In contrast, there were increases in water columnzinc and aluminum and in sediment arsenic, decreased algal

diversity, and differences in taxonomic composition betweenthe reference and near-field sites on Prairie Creek.

Water and sediment chemistry are inherently variable andthe interactions between chemical constituents with oneanother and biological processes are not well understood.Lack of conclusive patterns in water and sediment qualityresults for these rivers was expected due to the high degree ofmineralization, landscape features such as hot springs, andhighly variable flow rates (Halliwell and Catto 2003). Inaddition, higher sampling frequency typically is needed to getan accurate estimate of water and sediment chemistry(Markert et al. 2003). Although water and sediment qualityserve as useful indicators of direct effluent exposure, theability to predict biological effects may be limited in theseunique areas. Thus, monitoring studies which focus onbiological responses supported by water and sediment qualityfor exposure assessment are recommended.

The oligotrophic state of the rivers makes them sensitive tonutrient enrichment and other stressors (Kiffney and Clem-ents 1996; Vis et al. 1998). Proposed guidelines for theprotection of oligotrophic rivers in northern Alberta rangefrom 1.9 to 4.5 lg/cm2 chlorophyll a (Chambers et al. 2006).Concentrations of chlorophyll a in Prairie Creek wereconsistently below this range at all sites, whereas meanconcentrations of chlorophyll a in the Flat River were belowthis range at the reference site, within this range at the near-field site, and above this range at the far-field site. The trend ofincreasing algal abundance from the reference to the far-fieldsite in the Flat River could represent a mine-related enrich-ment or natural nutrient inputs along the gradient. Typically,enrichment due to a natural river continuum would occur overa far greater spatial distance among sites than what wassampled here. The natural range of algal abundance in these

Table 5. Metal concentrations in Prairie Creek water (lg/L) and sediment (lg/g) for reference, near-field, and far-field sites;effluent in Pond 3 and Pond 4; and associated CaNMdian Council of Minister of the Environment (CCME) guidelinesab

Parameter

Detection limit Upstream Near-field Far-fieldCatchment

pond CCME guidelines

Water Sediment Water Sediment Water Sediment Water Sediment Water Water Sediment

Aluminum 0.60 20 7.9 4820 12.6 3220 7.3 2620 ,20 5–10 NMc

Arsenic 0.20 0.2 0.2 6.4 0.3 7.6 0.3 6.4 ,50 5 5.9

Cadmium 0.05 0.01 0.05 1.24 0.06 1.42 0.05 0.88 3 0.017 0.6

Chromium 0.1 0.5 0.1 9.9 0.1 7.7 0.1 7.8 ,5 9.9 37

Copper 0.3 1 0.3 12 0.8 9 0.5 8 ,5 2–4 35.7

Iron 50 100 89 12300 89 8900 91 7970 15 300 NM

Lead 0.10 0.1 0.1 25.1 0.2 22.8 0.1 15.2 30 1–7 35

Manganese 0.10 10 0.2 229 0.4 205 0.3 179 234 NM NM

Mercury 0.02 0.01 0.02 0.06 0.02 0.03 0.02 0.04 0.07 0.026 0.17

Nickel 0.10 0.5 1.2 24.7 1.3 20.3 1.2 20.0 ,8 25–150 NM

Selenium 1.00 0.3 1.3 0.6 1.2 0.6 1.2 0.5 ,30 1 NM

Uranium 0.10 0.5 3.98 1.7 4.11 1.7 4.30 2.4 NM NM NM

Zinc 0.10 1.0 2.4 182 12.3 179 6.8 102 534 30 123a Water quality guidelines are provided from the CCME for the protection of freshwater aquatic life (CCME 2003).b Sediment quality guidelines are provided from the CCME Interim Sediment Quality Guidelines (CCME 2003).c NM ¼ not measured.

Table 6. Parasite loads for fish in Prairie Creek (% Occurrenceis calculated by comparing the number of parasite-infectedfish with the total number of fish captured at the reference,

near-field, and far-field sites)

Site % Occurrence % Body burden

Reference 25.5 10.386 6 1.653

Near-field 26.4 14.703 6 1.608

Far-field 18.8 12.713 6 3.451


rivers (i.e., 0.3–0.5 lg/cm2) is lower than the range reportedfor rivers in northern Alberta (Chambers et al. 2006).

Assessment of algal community composition also suggestsmine-related enrichment in the Flat River and Prairie Creekbut does not preclude possible effects of metal exposure.Decreased abundance of diatoms in general and of A.minutissima in particular, as well as reduced overall richnessand diversity, occurred at the near-field site in both rivers.Diatoms can be sensitive to metal pollution (Dixit et al. 1992;Genter and Lehman 2000; Gold et al. 2003) but also respondto changes in nutrient status in water bodies (Canter-Lundand Lund 1995; Markert et al. 2003). Achnanthes minutissimahas been shown to act as an indicator of metal exposure(Ivorra et al. 2002; Gold et al. 2003). In addition, increasedabundance of cyanophytes at the near-field site of PrairieCreek can occur naturally but may be associated withenrichment (Canter-Lund and Lund 1995). The results ofalgal biomass and taxonomy from both rivers suggest thatalgae are a useful monitoring tool for detecting differences inthese sensitive environments. In addition, previous researchhas indicated that algae accumulate metals and therefore canbe used to determine metal bioavailability and accumulationin aquatic ecosystems (Gurrieri 1998; Ivorra et al. 2002;Meylan et al. 2003).

Differences in the community composition of benthicmacroinvertebrates at exposure sites in the Flat River arelikely a result of tailings that historically were depositeddirectly into the river, whereas increased taxa richness inPrairie Creek is consistent with nutrient enrichment.Although the practice of direct tailings discharge to the FlatRiver was halted by 1970, tailings are still visible in the FlatRiver and alter substrate composition in the area (EBAEngineering 2002). A higher proportion of silt is present atboth exposure sites, which would provide a more desirablebenthic habitat for dipterans such as chironomids (Courtneyand Clements 2002). The substantial decrease in stoneflies atthe near-field and far-field sites in the Flat River mayrepresent a sensitivity to mine effluent or differences insediment composition. Plecoptera, or stoneflies, have beenshown to be pollution intolerant in many previous studies(Hodkinson and Jackson 2005).

Unlike in the Flat River, benthic macroinvertebrate richnessin Prairie Creek increased at exposure sites but communitycomposition was comparable to the composition at referencesites. The substrate composition among sites in Prairie Creekdoes not differ, as seen in the Flat River; no deposition oftailings has occurred during the 30 y of exploration or careand maintenance phases at the mine. In highly oligotrophicecosystems, very little nutrient input is required to produce aresponse (Vis et al. 1998; Hill et al. 2000; Bowman et al.2005). It is likely that small increases in nutrient or mineralinputs have resulted in increased taxa richness of benthicmacroinvertebrates at the exposure sites in Prairie Creek.Benthic macroinvertebrate densities are lower in Prairie Creekthan in the Flat River and both rivers have lower densitiesthan systems further south (Benke and Cushing 2005).

Fish community assessments confirmed that slimy sculpinwere the only potential sentinel species that could be used formonitoring in the South Nahanni watershed. Slimy sculpinare found throughout watersheds in the Northwest Territories(Burr and Page 1998) and are present downstream of most, ifnot all, major development in the Northwest Territories andNunavut (Azimuth Consulting Group 2005; DBCM 2005,

2006; DDMI 2005; Golder Associates 2005a, 2005b, 2006;Whitford 2006). They are a small species of fish with a shortlife span and limited mobility due to the absence of a swimbladder (Gray et al. 2004), which allows comparison ofeffects and exposure to contaminants (Munkittrick et al.2000). Sculpin have been shown to be a good sentinel speciesin several monitoring studies (Gibbons et al. 1998; Tetreaultet al. 2003; Galloway et al. 2005; Gray and Munkittrick2005). Limited work has been done on slimy sculpin innorthern ecosystems, and their applicability to effectsmonitoring (DDMI 2005; Lumb et al. 2006).

Analysis of liver tissue but not the flesh of sculpin trackedexposure to metals in the environment. An increase in bothcopper and iron in the liver tissue of slimy sculpincorresponded to high copper and iron in the effluent of theexfiltration ponds, as well as elevated copper in Flat Riversediments and iron in river water. Tungsten, which increasedin both water and sediment at the exposure sites in the FlatRiver, significantly decreased in liver tissue, suggesting thattungsten is not highly accumulated in slimy sculpin. Becauseof the considerably smaller fish size at Prairie Creek, onlyresults for flesh metals could be obtained. An exposurepattern was much more difficult to detect in fish collectedfrom Prairie Creek, presumably a consequence of usingconcentrations of metals in flesh rather than liver. However,the mine on Prairie Creek currently does not discharge a mineeffluent per se, but a mine-water portal discharge thatcurrently is treated in their polishing pond and catchmentpond. Also, there has been no historical discharge of tailings tothe Prairie Creek system.

Although there was some evidence of metal accumulationin fish tissue, increased fish condition at both exposure siteson Prairie Creek and at the far-field site on the Flat Riverprovides further evidence of an enrichment effect. In bothrivers, the absence of differences in liver and gonad weightsand increased condition suggests that mine dischargescurrently do not affect slimy sculpin adversely. The consistentpresence of parasites at the reference and exposure sites onPrairie Creek indicated that parasitic infection was not amine-influenced effect but rather due to other ecologicalprocesses.

CONCLUSIONBiological differences downstream of mines on the Flat

River and Prairie Creek were consistent with mild enrichmentof these highly oligotrophic systems. The most responsiveendpoints were different in each river, reinforcing the need forconsidering site-specific factors in the design of monitoringprograms. Responses to mining activity occurred in eachtrophic level. Variable water quality and sediment qualityresults were best analyzed in conjunction with biologicaldifferences. Analysis of metal concentrations in benthic algaland/or macroinvertebrate tissue is recommended for expo-sure assessment and causal association. The focus of furthermonitoring programs in sensitive pristine riverine areasexposed to mining activity should include a multitrophicapproach, including adequately replicated benthic algae andmacroinvertebrate community samples combined with waterand sediment quality and tissue metal burdens. Given the lowabundance of fish species in these ecosystems, the effects oflethal monitoring programs on fish populations should beconsidered seriously. Although the biological effects of miningactivities appear to be limited at present, evidence of


environmental metal concentrations above national guidelinesand associated metal accumulation in the liver tissue ofsculpin suggest that there is serious potential for increasedmining activity to affect biological communities in northernrivers. Further, evidence of mild eutrophication, which couldbe exacerbated with the advent of climate change andassociated factors (e.g., flow alterations, increased ice-freeperiods), suggests the need for on-going assessment. Thisresearch should serve as a baseline assessment for evaluationof future changes in sensitive tributaries of the South NahanniRiver.

Acknowledgment—We acknowledge the financial and in-kind support of the Northern Ecosystem Initiative ofEnvironment Canada, Indian and Northern Affairs, NaturalSciences and Engineering Research Council of CanadaResearch Chair Program, and the Canadian Foundation forInnovation (Dube), Canadian Zinc Corporation, NorthAmerican Tungsten Corporation, and Parks Canada. Wethank A. Squires, M. Pollock, J. Inkster, D. West, E. Pietroniro(University of Saskatchewan), T. Searson (Indian and North-ern Affairs), and T. Abercrombie (EBA Engineering Con-sultants) for their assistance.

REFERENCESAzimuth Consulting Group. 2005. Environmental Effects Monitoring (EEM).

Nunavut (NT): Teck Cominco Polaris Mine for Azimuth Consulting Group.

Barrie LA, Gregor D, Hargrave B, Lake R, Muir D, Shearer R, Tracey B, Bidleman T.

1992. Arctic Contaminants: Sources, occurrence and pathways. Sci Total

Environ 122:1–74.

Benke AC, Cushing CE. 2005. Rivers of North America. Boston (MA): Elsevier/

Academic.

Bowman MF, Bailey RC. 1997. Does taxonomic resolution affect the multivariate

description of the structure of benthic macroinvertebrate communities? Can J

Fish Aquat Sci 54:1802–1807.

Bowman MF, Chambers PA, Schindler DW. 2005. Epilithic abundance in relation to

anthropogenic changes in phosphorus bioavailability and limitation in

mountain rivers. Can J Fish Aquat Sci 62:174–184.

Burr BM, Page LM. 1998. A field guide to freshwater fishes: North America North

of Mexico. 2nd ed. New York (NY): Houghton Mifflin.

Canter-Lund H, Lund JWG. 1995. Freshwater algae: Their microscopic world

explored. Bristol (UK): Biopress.

[CCME] Canadian Council of Ministers of the Environment. 2003. Canadian water

quality guidelines for the protection of aquatic life. Hull (QC): Environment

Canada.

Chambers PA, Culp JM, Glozier NE, Cash KJ, Wrona FJ, Norton L. 2006. Northern

rivers ecosystem initiative: Nutrients and dissolved oxygen—issues and

impacts. Env Monit Assess 113:117–141.

Courtney LA, Clements WH. 2002. Assessing the influence of water and

substratum quality on benthic macroinvertebrate communities in a metal-

polluted stream: an experimental approach. Freshwater Biol 47:1766–1778.

[DBCM] De Beers Canada Mining. 2005. Snap Lake Diamond Project proposed

discharge limits. Submitted to MVLWB. Yellowknife (NT): Contract report.

[DDMI] Diavik Diamond Mines. 2005. Ammonia discussion paper. Submitted to

MVLWB. Yellowknife (NT): Contract report.

[DDMI] Diavik Diamond Mines. 2006. Round whitefish for toxicity testing.

Submitted to MVLWB. Yellowknife (NT): Contract report.

Dixit SS, Smol JP, Kingston JC, Charles DF. 1992. Diatoms: Powerful indicators of

environmental change. Environ Sci Technol 26:22–33.

Dube M. 2003. Cumulative effects assessment in Canada: A regional framework

for aquatic ecosystems. Environ Impact Assess Rev 23:723–745.

EBA Engineering. 2002. Fish and benthos inventory for the Flat River Tungsten, NT,

fall 2002. Vancouver (BC): EBA Engineering Consultants for North American

Tungsten.

Environment Canada. 1991. Protecting the waters of the Nahanni National Park

Reserve, Northwest Territories. Yellowknife (NT): Inland Waters Directorate and

Canadian Parks Service Report TR91-I/NAH.

Environment Canada. 2002. Metal mining guidance document for aquatic

environmental effects. http://www.ec.gc.ca/eem/English/MetalMining/

Guidance/default.cfm. Accessed January 2006.

Evans MS. 2000. The large lake ecosystems of northern Canada. Aquatic

Ecosystem Health Management 3:65–79.

Evans MS, Muir D, Lockhart WL, Stern G, Ryan M, Roach P. 2005. Persistent

organic pollutants and metals in the freshwater biota of the Canadian

subarctic and arctic: An overview. Sci Total Environ 351–352:94–147.

Findlay DL, Kasian SEM, Turner MT, Stainton MP. 1999. Responses of

phytoplankton and epilithon during acidification and early recovery of a lake.

Freshwater Biol 42:159–175.

Galloway BJ, Munkittrick KR, Currie S, Gray MA, Curry RA, Wood CS. 2005.

Examination of the responses of slimy sculpin (Cottus cognatus) and white

sucker (Catostomus commersoni) collected on the Saint John River (Canada)

downstream of pulp mill, paper mill, and sewage discharges. Environ Toxicol

Chem 22:2898–2907.

Genter RB, Lehman RM. 2000. Metal toxicity inferred from algal population

density, heterotrophic substrate use, and fatty acid profile in a small stream.

Environ Toxicol Chem 19:869–878.

Gibbons WN, Munkittrick KR, McMaster ME, Taylor WD. 1998. Monitoring

aquatic environments receiving industrial effluents using small fish species 1:

Response of spoonhead sculpin (Cottus ricei) downstream of a bleached-kraft

pulp mill. Environ Toxicol Chem 17:2227–2237.

Gold C, Feurtet-Mazel A, Coste M, Boudou A. 2003. Impacts of Cd and Zn on the

development of periphytic diatom communities in artificial streams located

along a river pollution gradient. Arch Environ Contam Toxicol 44:189–197.

Golder Associates. 2005a. Con Mine environmental effects monitoring, cycle 1,

final interpretive report. Yellowknife (NT): Golder Associates for Miramar Giant

Mine.

Golder Associates. 2005b. Giant Mine environmental effects monitoring, cycle 1,

final interpretive report. Yellowknife (NT): Golder Associates for Miramar Giant

Mine.

Golder Associates. 2006. Lupin Gold Mine environmental effects monitoring, cycle

1, final interpretive report. Yellowknife (NT): Golder Associates for Kinross

Gold Corporation.

Gray MA, Cunjak RA, Munkittrick KR. 2004. Site fidelity of slimy sculpin (Cottus

cognatus): Insights from a stable carbon and nitrogen analysis. Can J Fish

Aquat Sci 61:1717–1722.

Gray MA, Curry AR, Munkittrick KR. 2002. Nonlethal sampling methods for

assessing environmental impacts using a small-bodied sentinel fish species.

Water Qual Res J Can 37:195–211.

Gray MA, Munkittrick KR. 2005. An effects-based assessment of slimy sculpin

(Cottus cognatus) populations in agricultural regions of northwestern New

Brunswick. Water Qual Res J Can 40:16–27.

Gurrieri JT. 1998. Distribution of metals in water and sediment and effects on

aquatic biota in the upper Stillwater River basin, Montana. Journal of

Geochemical Exploration 64:83–100.

Halliwell DR, Catto S. 2003. How and why is aquatic quality changing at Nahanni

National Park Reserve, NWT, Canada? Environ Monit Assess 88:243–281.

Hill BH, Willingham WT, Parrish LP, McFarland BH. 2000. Periphyton community

responses to elevated metal concentrations in a Rocky Mountain stream.

Hydrobiologia 428:161–169.

Hodkinson ID, Jackson JK. 2005. Terrestrial and aquatic invertebrates as

bioindicators for environmental monitoring, with particular reference to

mountain streams. Environ Manag 35:649–666.

Indian and Northern Affairs Canada. 2001. Exploration overview 2001 North-

west Territories minerals, exploration and geological investigations. Yellow-

knife, NT.

Inland Waters Directorate. 1991. Protecting the waters of the Nahanni National

Park Reserve, Northwest Territories. Yellowknife (NT): Environment Canada.

Contract report TR91-I/NAH.

Ivorra N, Hettelaar J, Kraak M, Sabater S, Admiraal W. 2002. Responses of biofilms

to combined nutrient and metal exposure. Environ Toxicol Chem 21:626–632.

Kiffney PM, Clements WH. 1996. Effects of metals on stream macroinvertebrate

assemblages from different altitudes. Ecol Appl 6:472–481.

Kilgour BW, Dube MG, Hedley K, Portt CB, Munkittrick KR. 2006. Aquatic

environmental-effects monitoring guidance for environmental assessment

practitioners. Environ Monit Assess 130:423–436.


Lockhart WL, Wagemann R, Tracey B, Sutherland D, Thomas DJ. 1992. Presence

and implications of chemical contaminants in the freshwaters of the Canadian

Arctic. Sci Total Environ 122:1–74.

Lumb A, Halliwell D, Sharma T. 2006. Application of CCME Water Quality Index to

monitor water quality: A case of the Mackenzie River Basin, Canada. Environ

Monit Assess 113:411–429.

Mallory ML, Ogilivie C, Gilchrist HG. 2006. A review of Northern Ecosystem

Initiative in Arctic Canada: Facilitating arctic ecosystem research through

traditional and novel approaches. Environ Monit Assess 113:19–29.

Markert BA, Breure AM, Zechmeister HG. 2003. Bioindicators and biomonitors:

Principles, concepts, and application. Boston (MA): Elsevier.

Meylan S, Behra R, Sigg L. 2003. Accumulation of copper and zinc in periphyton in

response to dynamic variations of metal speciation in freshwater. Environ Sci

Technol 37:5204–5212.

Mining Association of Canada. 2007. Facts and Figures: A report on the State of

the Canadian Mining Industry. p 72. Ottawa (ON): Adhawk Communications.

Munkittrick KR, McMaster ME, Van Der Kraak G, Portt C, Gibbons WN, Farwell A,

Gray M. 2000. Development of methods for effects-driven cumulative effects

assessment using fish populations: Moose River project. Pensacola (FL):

Society of Environmental Toxicology and Chemistry (SETAC).

Natural Resources Canada. 2005. Canadian minerals yearbook. www.nrcan.gc.ca.

Accessed 19 July 2006.

Parks Canada. 1984. Nahanni National Park Reserve resource description and

analysis. Winnipeg, MB.

Reynoldson TB, Rosenberg DM, Resh VH. 2001. Comparison of models predicting

invertebrate assemblages for biomonitoring in the Fraser River catchment,

British Columbia. Can J Fish Aquat Sci 58:1395–1410.

Rott E. 1981. Some results from phytoplankton counting intercalibrations. Swiss J

Hydrol 43:43–62.

Schindler DW, Smol JP. 2006. Cumulative effects of climate warming and other

human activities on freshwaters of arctic and subarctic North America. Ambio

35:160–168.

Sigma Resource Consultants, Jemmet JA. 1978. The Effects of Heavy Metals and

Sewage on the Biota of the Flat River, NWT. Produced for Tungsten Mine.

Vancouver, BC. Contract report.

Tetreault GR, McMaster ME, Dixon DG, Parrott JL. 2003. Physiological and

biochemical responses of Ontario Slimy Sculpin (Cottus cognatus) to sediment

from the Athabasca Oil Sands area. Water Qual Res J Can 38:361–377.

Vis C, Hudon C, Cattaneo A, Pinel-Alloul B. 1998. Periphyton as an indicator of

water quality in the St. Lawrence River (Quebec, Canada). Environ Pollut

101:13–24.

Walker SL, Ribey SC, Trudel L, Porter E. 2003. Canadian environmental effects

monitoring: Experiences with pulp and paper and metal mining regulatory

programs. Environ Monit Assess 88:311–326.

Warnken KW, Gill GA, Wen LS, Griffin LL, Santchi PH. 1999. Trace metal analysis of

natural water by ICP-MS with on-line preconcentration and ultrasonic

nebulization. J Anal At Spectrom 14:247–252.

Whitford J. 2006. Metal Mining Environmental Effects Monitoring Nanisivik Mine.

Calgary (AB): Jacques Whitford for CanZinco.


a multitrophic approach to monitoring the effects of metal ...paula spencer,* michelle f bowman, and...

Documents