Skip booklet index and go to page content

Metal Mining Technical Guidance for Environmental Effects Monitoring

Chapter 9

9. Alternative Monitoring Methods

9.1 Overview

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.3 Use of Caged Bivalves as an Alternative Monitoring Method

9.4 References

List of Tables

Table 9-1: Summary of artificial stream applications for assessing the effects of pulp and paper and mining effluents on aquatic ecosystems as required under Canadian environmental effects monitoring

Table 9-2: Fish mesocosm study effect indicators and endpoints and related statistical procedures

Table 9-3: Recommended response variables and suitable additional supporting information, and suggested statistical analysis for Fathead Minnow application in modular mesocosm systems

Table 9-4: Suggested taxa for use in caged bivalve studies for EEMs

Table 9-5: Differences noted between two species of mussels over 5-year study in the Burrard Inlet, Vancouver, British Columbia

Table 9-6: Differences noted between Unionoidea and Sphaeriidae

Table 9-7: Caged bivalve study effect indicators and endpoints and related statistical procedures

Table 9-8: An example of a field data sheet for recording survival and growth raw data

List of Figures

Figure 9-1: A) Large mesocosm system with streams situated on tables (Model I) used in the Athabasca River, Alberta. B) Large mobile mesocosm system with streams on 2 trailers (Model II) used in the Fraser River, British Columbia; the Saint John River, New Brunswick; and in Saint John Harbour, New Brunswick. C) Large mobile mesocosm system with streams on a single trailer (Model III) used in the Miramichi and Little rivers, New Brunswick; the Wapiti River, Alberta; and Junction Creek, Ontario.

Figure 9-2: A) Small microcosm system with streams situated on tables over mixing reservoirs used in the Thompson River, British Columbia. B) Modular mesocosm system with streams situated on tables over mixing reservoirs used in the Little River, New Brunswick; Junction Creek, Ontario; the Wabigoon River, Ontario; and Key Lake, Saskatchewan.

Figure 9-3: Schematic of large mesocosm trailer system (not to scale)

Figure 9-4: Photograph of a modular mesocosm parts diagram.

Figure 9-5: Modular mesocosm flow schematic.

Figure 9-6: Multitrophic Fathead Minnow reproductive bioassay and feeding barrier.

Figure 9-7: Site set-up for modular mesocosms

Figure 9-8: A) Factorial experimental design to investigate the importance of water vs. diet in responses of Fathead Minnow to metal mine effluent in modular mesocosms; B) Experimental design to investigate the influence of pH and natural organic matter (NOM) on Fathead Minnow responses after exposure to an MME mixture and a single metal in multitrophic modular mesocosms.

Figure 9-9: Factorial experimental design to investigate the effects of MME and historical sediment contamination in isolation and in combination on Fathead Minnow in modular mesocosms.

Figure 9-10: Mussel showing ripe mantle lobe

Figure 9-11: Reproductive cycle of Blue Mussels from British Columbia: A) Mantle energy stored in fall; B) Mantle reproductive content in spring

Figure 9-12: Mytilus spp. shell scars markings

Figure 9-13: Duplicate frame from a caged mussels exposure experiment

Figure 9-14: Modular mesocosm parts diagram

Figure 9-15: Modular mesocosm flow schematic

Figure 9-16: A) Mantle plug and tools necessary for its removal; B) Mantle plug after homogenization and ready for assessment

9. Alternative Monitoring Methods

9.1 Overview

At some mines, standard fish and benthic invertebrate community monitoring studies may not be appropriate. The reasons for this are site-specific, but the most common reasons are the presence of hazardous conditions (e.g., high flow velocity); unsuitable habitat for sampling; or the presence of confounding factors, such as other effluent discharges in the exposure area, that make it impossible to isolate any effects attributable to the effluent being monitored.

Where mines cannot design the fish or benthic invertebrate community surveys to resolve difficulties associated with confounding influences, they will provide a scientific rationale and justification, and propose cost-effective and technically feasible alternative monitoring methods within the study design. A number of alternative monitoring methods are recommended in this chapter.

Mines may choose other scientifically defensible methods, provided that the results can determine if the effluent is having effects on the fish population (growth, reproduction, condition and survival), fish tissue (mercury), or the benthic invertebrate community (benthic invertebrate density, taxa richness, the Simpson’s Evenness Index and the Bray-Curtis Index). Currently, recommended alternatives to the fish survey monitoring method are mesocosm (artificial stream) and caged bivalve studies. For benthic invertebrate community surveys, the recommended alternative monitoring method is the mesocosm study.

Other alternatives to the fish and benthos field surveys may also exist. Mines can suggest other alternative methods in their study design. New alternative methods will be evaluated by the Authorization Officer, with the support of the Technical Advisory Committee and the Environmental Effects Monitoring (EEM) Science Committee. In reviewing the suggested alternative, some specific design elements will be considered as essential to meeting the objective of the program, including environmental relevance and interpretable results that are scientifically defensible and manageable.

The objective of section 9.2 is to provide guidance on the study design and implementation of mesocosm studies as an alternative EEM method for assessing the effects of metal mine effluent on benthic invertebrates and fish. This guidance document is intended to provide information on recommended standards of good scientific practice available to meet the outlined EEM requirements. In 2002, the first guidance document for the use of artificial stream systems (mesocosms) was released. Guidance was updated in 2011 to reflect ongoing research and development for improvement of this alternative EEM method.

The objective of section 9.3 is to provide technical guidance for conducting controlled experiments using caged bivalves suspended in the water column to test for effects associated with industrial discharges and to compare measurements between exposure and reference areas. Caged bivalves are also an alternative to the fish survey and may be considered for mines where the fish survey has been unsuccessful or impractical in past phases of EEM, or where there may be study design issues, such as confounding influences, or safety concerns.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.1 Background Information on Artificial Stream Development and Application

Artificial streams are recommended as monitoring alternatives because years of research and development have demonstrated that, with respect to effluent effects, they can produce good-quality data that fit within the required regulatory context (Table 9-1). Since 1991, field-based artificial stream system studies have been conducted for assessing the effects of point-source effluents on aquatic ecosystems. Field-based artificial stream studies relevant to EEM applications were conducted in Canada 14 times in 8 years between 1993 and 2008 (Table 9-1). All of these studies, and development of the alternative method, were conducted as collaborative partnerships between industry, government, academia and consultants. All funding for the research was acquired through mechanisms independent of the EEM Program. Applications of this method are presented in detail below to provide a thorough understanding of the work that has been done to date. A summarized version with references can be found in Table 9-1. These references should be consulted if similar types of studies and experimental designs are being considered.

Top of Page

1991-1996, Northern River Basins Study (NRBS), Alberta

One of the first mesocosm applications assessed the effects of pulp mill effluent (PME) on benthic invertebrate and periphytic algal communities in the Athabasca River (Table 9-1) (Culp and Podemski 1996; Culp et al. 1996; Podemski and Culp 1996; Podemski 1999; Culp et al. 2001). Artificial streams were used to distinguish the effects of nutrients in whole mill effluent from contaminants, on the basis of directional differences in biological response. Specifically, moderate nutrient enrichment would increase primary and secondary productivity, whereas contaminant effects would reduce growth and reproduction and eventually result in mortality (Culp and Podemski 1996; Podemski and Culp 1996; Culp and Lowell 1998; Culp et al. 2001). To achieve this objective, 3 treatments were tested in the spring of 1993: control Athabasca River water, 1% (volume/volume [v/v]) treated PME, and 1% (v/v) nutrients (nitrogen + phosphorus) at levels measured in the PME. The hypothesis was that exposure to both the PME and nutrient treatments would result in nutrient enhancement effects on the benthic food web and that the PME and nutrient treatments would not differ. This would suggest that the effects of PME at levels found in the Athabasca River were due to nutrient enrichment rather than contaminant toxicity.

A large non-mobile artificial stream system was used near the pulp mill at Hinton, Alberta. The system consisted of 16 circular tanks or streams (0.9 m2 each) placed on tables (Model I, Figure 9‑1A). River water was pumped into each stream at a controlled rate, and effluents and nutrients were added to the treatment streams as previously published by Culp and Podemski (1996) and Podemski (1999). A standardized benthic community, endemic to the Athabasca River, was created in each stream and exposed to PME for 28 days. At the end of the exposure period, algal biomass, growth of mayfly (Ephemeroptera: Siphloneuridae, Baetidae) and stonefly (Plecotera: Capniidae) nymphs, and insect abundance, increased in the treatment streams relative to the reference treatment (Culp and Podemski 1996; Podemski and Culp 1996; Culp et al. 1996). In addition, these response variables did not differ between the 1% PME and the 1% nutrient treatments, supporting the hypothesis that the effects of PME on the benthic food web were attributable to nutrient enrichment.

Top of Page

1991-1998, Fraser River Action Plan (FRAP), British Columbia

The FRAP was conducted from 1991 to 1997 to determine the current state of health of the Fraser River Basin ecosystem, including assessment of 8 pulp and paper mill effluents (Gray and Tuominen 1998; McGreer and Belzer 1998).

Thompson River, Kamloops, British Columbia (1993-1994)

The Thompson River showed signs of nutrient enrichment due to the discharge of PME at the City of Kamloops. This problem has been investigated since the early 1970s, when excessive accumulations of periphytic algae occurred in the river downstream of the pulp mill (Federal-Provincial Thompson River Task Force 1976). Bothwell and Daley (1981), Bothwell (1985), Bothwell et al. (1992) and Bothwell and Culp (1993) illustrated how periphytic algae growth was enriched by bioavailable phosphorus discharged in PME.

Artificial streams were used to separate the interacting effects of nutrients and contaminants in PME on algae and benthic invertebrates (Table 9-1). The approach differed from the NRBS studies in that a dose-response design was employed with the expectation of observing nutrient effects at low effluent concentrations and contaminant effects at higher concentrations. In 1993 and 1994, periphytic algae and chironomids were exposed to a dilution series of PME (0.25‑10% [v/v]) (Dubé and Culp 1996; Culp and Lowell 1998). Smaller artificial streams were used for testing the effects of the PME on single insect species (Lowell et al. 1995, 1996) and simplified benthic food webs (Dubé and Culp 1996). The single-species approach focused the assessment of effects on key sentinel taxa, to improve our understanding of species-specific responses (Culp et al. 2000b).

The artificial stream system was set up on the banks of the Thompson River at Kamloops just upstream of the effluent outfall. The system included a water distribution system, treatment reservoirs for mixing the respective effluent dilutions with a continuous supply of river water, and small circular 0.33‑L streams (45 cm2 planar area) (Dubé 1995; Lowell et al. 1995) (Figure 9‑2A). Algae and chironomid larvae (Diptera: Orthocladiinae) from a reference area were placed into the streams, and changes in algae and chironomid biomass were measured after 2-3 weeks of effluent exposure (Dubé and Culp 1996). Dubé and Culp (1996) reported that algal biomass (chlorophyll a) increased in all effluent concentrations due to nutrient enrichment. Total chironomid biomass and individual weight were also enriched at low effluent concentrations (< 5%). At higher concentrations (5% and 10%) chironomid biomass decreased, possibly due to contaminant effects.

Top of Page

Table 9-1: Summary of artificial stream applications for assessing the effects of pulp and paper and mining effluents on aquatic ecosystems as required under Canadian environmental effects monitoring. (text description)
YearProgram1Effluent Type2Research ObjectiveLocationArtificial Stream SystemReferences
1993-1994NRBSPMETo determine the effect of nutrients and contaminants in PME on periphytic algae and benthic invertebrate communitiesAthabasca River, Hinton, ABModel I: Large non-mobile mesocosm. Benthos. Field study.Podemski (1999)
Culp and Podemski (1996)
Podemski and Culp (1996)
Culp et al. (1996)
Culp et al. (2001)
1993-1994FRAPPMETo assess the effects of increasing concentrations of PME (0.25–10%) on periphyton and chironomids (Diptera: Orthocladiinae)Thompson River, Kamloops, BCSmall microcosm system. Benthos. Field study.Dubé (1995)
Dubé and Culp (1996)
Dubé et al. (1997)
Lowell et al. (2000)
1993FRAPPMETo determine the food-dependent effects of PME on the mayfly species Blue-winged Olive (Ephemeroptera: Baetis tricaudatus)Thompson River, Kamloops, BCSmall microcosm system. Benthos. Field study.Lowell et al. (1995)
Lowell et al. (1996)
Culp et al. (1996)
Lowell et al. (2000)
1994FRAPPMETo determine the effects of PME (1% and 3%) on periphyton and benthic invertebrate communitiesFraser River, Prince George, BCModel II: Large mobile mesocosm. Benthos. Field study.Culp and Cash (1995)
Culp et al. (1996)
Culp et al. (2000a)
1997Pulp and Paper EEMPMETo assess the effects of PME (3%) on a small-bodied fish, Mummichog (Fundulus heteroclitus), in a marine environmentSaint John Harbour, Saint John, NBModel II: Large mobile mesocosm. Fish. Field study.Cash et al. (2003)
1997-1998Pulp and Paper EEMPMETo determine the effects of final PME, in-mill process streams, and a mill process change on a small-bodied fish, Mummichog, in an estuarine environmentSaint John River, Saint John, NBModel II: Large mobile mesocosm. Fish. Field study.Dubé (2000)
Dubé and MacLatchy (2000a)
Cash et al. (2003)
1999TSRIPMETo determine the effects of primary- and secondary-treated PME on a small-bodied fish, Mummichog, in an estuarine environmentMiramichi River, Miramichi, NBModel III: Large mobile mesocosm. Fish. Field study.Dubé et al. (2002)
2000TSRIMMETo determine the effects of MME (20%, 80%) on juvenile Atlantic Salmon (Salmo salar)Little River, Brunswick Mines, Miramichi, NBModel III: Large mobile mesocosm. Fish. Field study.Dubé et al. (2005)
To evaluate the individual and combined impacts of MSE and PME on Longnose Dace (Rhinichthys cataractae)Wapiti River, ABModel III: Large mobile mesocosm. Fish. Field study.Dubé et al. (2004)
To evaluate the individual and combined impacts of MSE and PME on the benthic food webWapiti River, Grande Prairie, ABModel III: Large mobile mesocosm. Benthos. Field study.Culp et al. (2004)
2001-2002IndustryMMETo assess effects of treated MMEs from three mines discharging to Junction Creek, Sudbury, on Creek Chub (Semotilus atromaculatus) and Pearl Dace (Semotilus margarita)Junction Creek, Sudbury, ONModel III: Large mobile mesocosm. Fish. Field study.Dubé et al. (2006)
2002IndustryMMETo evaluate the effects of MME (45%) on the partial life cycle of the chironomid Chironomus tentansJunction Creek, Sudbury, ONModular mesocosm system. Benthos. Field study.Hruska and Dubé (2004)
2003NSERCMMEComparison of a partial–life-cycle bioassay in artificial streams to a standard beaker bioassay, to assess effects of MME (45%) on the chironomid C. tentansJunction Creek, Sudbury, ONModular mesocosm system. Benthos. Lab study.Hruska and Dubé (2005)
PMETo determine effects of final PME (1%, 100%) and various process streams on the partial life cycle of Fathead Minnow (Pimephales promelas) under environmentally realistic conditions (i.e., ambient water and effluent quality)Terrace Bay, ONBioassay trailer. Fish. Field study.Rickwood et al.(2006a, 2006b)
MMETo develop a self-sustaining multitrophic bioassay, using C. tentans and Fathead Minnow (partial life cycle) to comparatively assess effects of water-borne vs. food- and water-borne exposure to MME (45%) on Fathead Minnow reproductionJunction Creek, Sudbury, ONModular mesocosm system. Multitrophic. Lab study.Rickwood et al. (2006c)
2005NSERC/ IndustryMMETo develop a self-sustaining multitrophic bioassay, using C. tentans and Fathead Minnow (partial life cycle) to comparatively assess effects of water-borne vs. food- and water-borne exposure to MME (45%) on Fathead Minnow reproductionJunction Creek, Sudbury, ONModular mesocosm system. Multitrophic. Field study.Rickwood et al. (2008)
PMETo assess effects of PME (20%, 40%, 60%) on Fathead Minnow (partial life cycle)Wabigoon River, Dryden, ONModular mesocosm system. Fish. Field study.Pollock et al.(2009)
MME and potential causative metalTo comparatively evaluate response patterns of Fathead Minnow (partial life cycle) to an MME mixture (100%, 25%, 5%) vs. selenium as selenate, using a multitrophic mesocosm bioassayUnknown Lake, Key Lake, SKModular mesocosm system. Fish. Lab studyPollock et al. (unpublished)
MMETo comparatively evaluate effluent (current discharge: 25%) vs. sediment (historical contamination) exposure pathways on Fathead Minnow (partial life cycle)Unknown Lake, Key Lake, SKModular mesocosm system. Fish. Field study.Driessnack et al. (unpublished)
MMETo assess the effects of three different MME discharges on Fathead Minnow (partial life cycle).Junction Creek, Sudbury, ONModular mesocosm system. Multitrophic. Field study.Ramilo et al. (unpublished)

1 NRBS: Northern River Basins Study; FRAP: Fraser River Action Plan; EEM: Environmental Effects Monitoring; TSRI: Toxic Substances Research Initiative; NSERC: Natural Sciences and Engineering Research Council of Canada.
2 PME: pulp mill effluent; MSE: municipal sewage effluent; MME: metal mine effluent.

Top of Page

In 1993, Lowell et al. (1995, 1996) conducted small-scale artificial stream experiments on the Thompson River in concert with those of Dubé and Culp (1996). Using the mayfly species Blue-winged Olive, the effects of PME (1% and 10% v/v) on survival, growth, moulting and morphological development were investigated under 2 feeding regimes (low and high). Effluent exposure significantly stimulated growth and development, with 20–50% increases in dry body weight relative to controls. Although moulting frequency increased with moderate effluent exposure (1%), higher exposure (10%) reduced moulting frequency, suggesting a contaminant-mediated mechanism (Lowell et al. 1996). These artificial stream results using mayflies as the sentinel species were consistent with the chironomid exposure experiments conducted by Dubé and Culp (1996), which showed an enrichment response at low PME concentrations and the appearance of inhibitory effects at higher concentrations.

In addition to consistency among artificial stream experiments, these results were consistent with field survey results (Culp and Lowell 1998). Long-term trend analysis showed that several families of stoneflies (Plecoptera), caddisflies (Trichoptera), and mayflies (Ephemeroptera) were more abundant in the years when the mill output of suspended solids and phosphorus was higher (Lowell et al. 1996, 2000). Field monitoring by Dubé et al. (1997) also showed that temporal and spatial patterns in water-column phosphorus, periphyton biomass and chironomid biomass (Diptera: Orthocladiinae) were consistent under normal mill operating conditions. The effects of the mill on the Thompson River benthic food web were restricted to nutrient enrichment. However, Dubé (1995) also observed that toxic effects of mill-related contaminants decreased chironomid densities in the Thompson River at low effluent exposure (far-field) sites in 1992 when the mill’s secondary effluent treatment system shut down.

Fraser River, Prince George, British Columbia (1994)

The effects of PME on benthic food webs using artificial streams were also examined in the Fraser River at Prince George, British Columbia, which received effluent from 4 pulp mills located within a 100-km stretch of the river (Culp and Lowell 1998). In 1994, benthic communities were exposed to 1% and 3% concentrations (v/v) of PME for 35 days to determine if nutrient enrichment effects occurred at low PME concentrations and toxic effects manifested at higher concentrations (Table 9-1) (Culp and Cash 1995; Culp et al. 2000a). The number of measured response variables increased in this study, and included bacterial number, periphyton biomass, composition, accumulation of target PME contaminants, and benthic invertebrate community structure. Community-level responses, in addition to species-specific responses, were measured to increase the ecological relevance of the study (Culp et al. 2000b).

The design of the large artificial stream system was modified to improve its flexibility, transportability and cost-effectiveness. The streams and tables were secured onto 2 mobile flatbed trailers (Culp et al. 1996) (Model II, Figure 9‑1B). In addition, each trailer was constructed with enclosed laboratory space for effluent and water header tanks, pump storage, and space for sample processing. Design and operation of the streams, including benthic community inoculation, flow rates and sampling protocols, were as previously described by Culp and Cash (1995), Culp and Lowell (1998), and Culp et al. (2000a).

Results from the Fraser River studies supported those from both the Thompson River and NRBS studies, illustrating that the effects of PME on the benthic food web were caused by nutrient enrichment. Culp et al. (2000a) reported that bacterial numbers, periphyton biomass and the biomass of dominant insect taxa (i.e., chironomids and stoneflies) increased with effluent exposure. Interestingly, although a dose-response relationship was observed for pulp mill contaminants (i.e., resin acids and chlorinated phenolics) measured in periphyton, these tissue burdens did not translate into a decrease in algal growth or a change in species richness. Results were also consistent with laboratory and field studies building a weight of evidence on the effects of PME on riverine benthos response patterns in the Fraser River (Culp et al. 2000a).

Top of Page

1997-1998, Industrial EEM Pilot Studies, New Brunswick

Studies conducted during the NRBS and FRAP illustrated the utility of employing artificial streams for assessing the effects of PME on bacteria, periphyton and benthic invertebrate communities. The systems provided a mechanism to measure responses of endemic biota to controlled effluent concentrations under ambient, environmentally relevant conditions of light, temperature and water quality (Culp et al. 1996). These demonstrated qualities also made artificial stream application an attractive alternative for assessing the effects of PME on fish (Courtenay et al. 1998; Parker and Smith 1997). Three industrial EEM pilot studies were conducted in marine and estuarine environments in 1997 and 1998 to develop artificial stream techniques for assessing PME effects on fish.

Saint John Harbour, Saint John, New Brunswick (1997)

The first pilot study was conducted in Saint John Harbour, New Brunswick, using the large mobile mesocosm system (Model II, Figure 9-1B) to assess the effects of a secondary-treated thermo-mechanical PME on a saltwater killifish, the Mummichog (Table 9-1) (Cash et al. 2003; Dubé et al. 2002). The mill discharged to a complex marine environment characterized by extreme tidal fluxes, historical sediment contamination, and the presence of other effluents (e.g., treated and untreated sewage, storm water, another pulp mill effluent, and oil refinery effluent).

The artificial stream system was situated on shore at the end of a breakwater. Receiving water, unexposed to PME, was pumped into each stream during each tidal exchange as described by Cash et al. (2003) and Dubé et al. (2002). Two treatment conditions were created: control receiving water and 3% effluent (v/v). The 3% effluent concentration represented the concentration found over the largest spatial extent in the receiving waters as determined by plume delineation studies. Effluent was dosed for 28 days into each 3% treatment stream in conjunction with receiving water exchanges simulating exposure conditions of the sentinel species (Mummichog) that remain in tidal pools during ebb and low tide (Kneib 1986). Juvenile fish (120 fish per treatment) and adult fish (60 per sex per treatment) were allocated to the control and to the 3% effluent treatments, and were fed daily using frozen brine shrimp at a rate of 3% total stream biomass. Mummichog was selected as the sentinel species because it is well-studied, endemic to Saint John Harbour, a suitable size to place into the streams, and sexually dimorphic for ease in controlling sex ratios (Kneib and Stiven 1978; Atz 1986; Scott and Scott 1988). In addition, juvenile growth rates are high enough to detect effluent-related effects over the exposure period at ambient study temperatures (Kneib and Stiven 1978). Response variables included effect endpoints congruous with the EEM wild fish survey (i.e., growth, gonad and liver size, condition factor) as well as additional physiological supporting endpoints (mixed-function oxygenase [MFO] induction, reproductive hormone levels) (Cash et al. 2003; Dubé et al. 2002).

This study provided information on mill-related effects for an endemic fish species. The rate of survival was close to 100% in all treatments, and effluent exposure did not affect growth or MFO activity (Cash et al. 2003). However, effluent exposure did significantly reduce gonad and liver size in males and increased production of some sex steroids in both sexes.

Saint John River, Saint John, New Brunswick (1997-1998)

In 1997 and 1998, Dubé (2000) used artificial streams to determine the effects of a PME on Mummichog in the Saint John River and to evaluate changes in final effluent quality associated with a mill process change (Table 9-1). This same large artificial stream system (Model II, Figure 9-1B) and sentinel species were used as above. This study differed with respect to the scope of the hypotheses tested, the type of mill process investigated (bleached-kraft chemical pulping process) and the type of receiving environment studied (estuarine).

In 1997, before the mill process change, adult Mummichog were exposed to final mill effluent (1% v/v) for 27 days (Dubé and MacLatchy 2000a). In 1998, after the mill process change, both adult and juvenile Mummichog were exposed to 3 concentrations of PME (0.5%, 1.0% and 5.0% v/v) for 30 days and 60 days. The large artificial stream system was situated on the Saint John River beyond the zone of effluent influence. Reference water was pumped continuously into each stream to simulate site-specific exposure conditions. Response variables included juvenile growth, adult organ size (liver, gonad), condition, MFO induction, and reproductive hormone levels.

In both studies, the survival rate was > 95% and fish in all treatments increased in biomass throughout the exposure period, showing an adequate feeding rate (Dubé 2000). Exposure to final effluent at 1% did not affect adult organ size (gonad or liver) in either study. However, to illustrate the responsiveness of Mummichog to PME and to support conclusions that adult fish were largely unaffected by exposure to environmentally relevant concentrations of PME at this mill, Dubé (2000) employed a dose-response study design in 1998. Exposure to a 5% concentration of PME for 60 days resulted in significant increases in liver size in both sexes and significant decreases in both the length and weight of juvenile fish (Cash et al. 2003; Dubé et al. 2002).

The artificial stream system was also used in this study to evaluate the effects of a mill process change on final effluent quality (Dubé and MacLatchy 2000a, Dubé et al. 2002). Changes in liver size, gonad size and condition in adult Mummichog were not observed between 1997 and 1998. However, patterns in reproductive hormone levels differed between years, showing significant depressions in plasma testosterone in both males and females in 1997 but not in 1998. Further investigation using toxicity tests (Dubé and MacLatchy 2000b) and laboratory exposures of Mummichog to mill process effluents (Dubé and MacLatchy 2001) in a weight-of-evidence approach confirmed that the process change removed acute toxicity of the final effluent and significantly reduced sublethal toxicity, including reducing reproductive effects on a local fish species.

Top of Page

1999-2001, Toxic Substances Research Initiative (TSRI)

Artificial stream development occurred in 4 main areas over this period: further development and optimization of technology design, use with other effluents (metal mining), use with other fish species, and use in cumulative effects bioassessment programs with multiple effluents (Table 9‑1).

Miramichi River, Miramichi, New Brunswick (1999)

This study evaluated the effects of primary and secondary bleached-kraft PME (1% v/v) on Mummichog after 23 days of exposure, using a redesigned, large artificial stream system (Table 9‑1) (Dubé et al. 2002). The system consisted of 16 circular tanks (0.42 m2) on a single trailer for improved transportability (Model III, Figure 9‑1C). Improved control over effluent dilution and dissolved oxygen levels was also attained, by redesigning the plumbing and adding an air-lift system (Cash et al. 2003). AMEC Earth & Environmental Ltd. (previously Washburn & Gillis Associates Ltd.) constructed and currently owns the system.

Adult survival was high in all treatments (> 90%) and effluents did not affect length, weight, condition, liver somatic index (LSI) or gonadosomatic index (GSI) after 23 days of effluent exposure. However, both sexes of Mummichog exposed to secondary-treated effluent showed significant, 5-fold depression in plasma testosterone concentrations compared to the control fish. These concentrations were also significantly depressed relative to levels measured in fish exposed to a 1% primary-treated effluent. These results suggest that secondary treatment of some bleached kraft pulp mill effluent may not remove the compounds responsible for depression of reproductive hormones in some fish.

Little River, Bathurst, New Brunswick (2000)

In this study, artificial stream techniques were applied to assess the effects of an MME. In 2000, artificial stream studies were conducted by Culp et al. (unpublished) to evaluate the effects of an MME on benthic invertebrate and algae communities. Dubé et al. (2005) concurrently evaluated MME effects on juvenile Atlantic Salmon through water-borne exposures as well as through exposure in a naturally cultured multitrophic-level food web (algae + benthic invertebrates + fish). Studies were conducted at a mine near Bathurst, New Brunswick. In the first study, the large (Model III) artificial stream system (Figure 9‑1C) was used to assess the effects of 20% and 80% (v/v) MME on salmon. The treatment levels for this study were selected to represent current effluent discharge (80%) into the Little River, New Brunswick, and predicted discharge levels upon mine closure (20%). The experiments consisted of 37 days of exposure, and response variables included growth, liver size, condition, metal tissue burdens, and stress variables including levels of muscle glycogen (Dubé et al. 2005). In the second set of studies, the modular stream system (Figure 9‑2B) was used to measure benthic invertebrate responses to 20% and 80% MME after 24 days of exposure. Response variables included changes in total invertebrate density, taxon richness, Simpson’s Diversity Index, Bray-Curtis Index and insect emergence (Culp et al. unpublished). In the third set of experiments, the modular stream system (Figure 9‑2B) was also used to expose a self-sustaining multitrophic-level food web to 20% and 80% concentrations of MME for 26 days (Dubé et al. 2005). In these multitrophic-level studies, young-of-the-year Slimy Sculpin (Cottus cognatus) were placed into streams that had been inoculated with algae and benthic invertebrate communities from a reference river. This permitted assessment of MME effects on fish using a more environmentally realistic pathway of contaminant exposure (i.e., through the food web as opposed to using an unexposed food source).

Wapiti River, Grande Prairie, Alberta (2001)

Mesocosms were used to separate out the confounding effects of a secondary-treated bleached-kraft PME from an MSE on survival, growth, condition and reproduction in adult and juvenile Longnose Dace (Dubé et al. 2004). Longnose Dace were exposed to the following treatments for 42 days: reference river water, PME (3%), PME (10%), MSE (1%), and MSE (1%) + PME (3%). The objective of the dose-response exposure to PME was to examine the response pattern to PME in isolation under low and high concentrations. The MSE and mixture treatments were representative of conditions upstream (MSE 1%) and downstream (MSE 1% + PME 3%) of the PME discharge in the Wapiti River. Results showed that 10% PME slightly reduced juvenile condition and altered some reproductive hormones in adults. Exposure to 3% PME slightly increased juvenile condition, suggesting nutrient enrichment at lower PME concentrations. No effects on survival, growth, liver size, gonad size, or stage of gonadal development were observed with PME exposure. MSE affected reproductive response variables such as male gonad size, female fecundity and some hormone levels in males and females. Hormonal changes after exposure to 10% PME were similar in magnitude to changes measured after exposure to 1% MSE. This study specifically examined the effects of water-borne exposure to PME and MSE on a forage fish after 42 days in a field-based mesocosm.

Culp et al. (2004) examined the cumulative effects of PME and MSE on benthic invertebrate and algal communities. Four treatments were established, as in the above study (i.e., control, 1% MSE, 3% PME, 1% MSE + 3% PME). Replicate benthic food webs were established across all treatments by inoculating each mesocosm stream with substratum, the associated microbes and algae, and invertebrates that were obtained from a reference area. Adult insects were collected from emergence traps placed over each stream every 2-3 days (Figure 9‑2B), while benthic invertebrates and algal biomass were sampled at the end of the experiment. The results indicate that both MSE and PME were a significant source of nutrients to the river. MSE appeared to be a primary source of nitrogen, while PME appeared to be an important source of phosphorus and carbon. Algal biomass increased with effluent exposure and was more strongly related to nitrogen than to phosphorus or carbon. Insect emergence data suggested a synergistic rather than additive effect of exposure to the 2 complex effluents (Culp et al. 2004).

Top of Page

2001-2008, Academic/Industry Partnership Applications

Junction Creek, Sudbury, Ontario (2001, 2002)

Junction Creek in Sudbury, Ontario, historically exposed to sediment contamination from decades of mining, receives 3 treated mine effluents, a municipal wastewater effluent, and several other nonpoint-source impacts. In 2001 and 2002, effects of treated MMEs from 3 different mining operations discharging to Junction Creek on 2 fish species--Creek Chub and Pearl Dace--were assessed (Dubé et al. 2006). Treatments tested for 35 to 41 days included reference water, MME #1 (30%), MME #2 (20%), and MME #3 (45%). In 2001, effects on chub included reduced survival (not statistically significant) and depressed testosterone levels. In 2002, chub and dace survival were reduced to less than 60% in MMEs #1 and #3. In addition, the total body weights of male and female dace were reduced after exposure to these same effluents. In 2001 and 2002, responses were most common to MMEs #1 and #3, with consistent increases in nickel, rubidium, strontium, iron, lithium, thallium, and selenium observed across treatment waters and body tissues. These studies identified changes in response variables for fish endemic to Junction Creek and after exposure to mine discharges independent of historical sediment contamination.

After several years of investigation of effluent effects on fish in water-borne exposures, there was a need to further develop the mesocosm systems for trophic-transfer applications. This was based on the fact that dietary pathways of exposure to contaminants are more environmentally relevant than exposures through the water alone (although the latter is certainly most common in aquatic toxicological research). In addition, results from the national EEM assessments for the pulp and paper EEM program were indicating that reproductive effects of effluents on fish were a dominant national response pattern. This suggested that development of the fish mesocosm method should focus on dietary exposure pathways as well as a more thorough evaluation of reproductive response variables.

In 2002, we developed an in situ life-cycle bioassay with the chironomid C. tentans in the modular artificial streams, to evaluate the effects of an MME under ambient environmental conditions in Junction Creek, Ontario (Hruska and Dubé 2004). The chironomidswere exposed throughout their life cycle to MME #3, which is the average effluent concentration measured in the creek. C. tentans in the effluent treatment exhibited reduced survival, total emergence, hatching success and increased time to emergence. This research showed how a life-cycle bioassay could be used in situ to assess MME effects on a benthic invertebrate. In addition, valuable information was obtained on C. tentans growth rates, hatchability and survival in mesocosms, which is information required for improved development of culture-based multitrophic-level mesocosm systems.

In 2003, development of the C. tentans mesocosm continued for assessment of MME (Hruska and Dubé 2005). The utility of this test was compared to an existing standard beaker life-cycle bioassay under laboratory conditions. C. tentans larvae were exposed to 45% (v/v) treated MME #3 from day 11 through hatching of the second generation. Response patterns were consistent between the 2 bioassays for hatching success and time to emergence but inconsistent for other variables. Significant effects were obtained for growth, survival, number of adults emerged, and number of eggs per egg case in the artificial stream bioassay but not in the beaker bioassay. Conversely, significant effects on sex ratio and number of egg cases per female were observed in the beaker bioassay but not in the artificial stream bioassay. These differences are believed to be a consequence of the number of organisms per replicate used in each bioassay, which results in a difference in statistical power. As a result, higher coefficients of variation and effects sizes were observed in the beaker bioassay relative to the artificial stream bioassay for almost all variables. These results provided evidence that the mesocosm approach was an effective tool for evaluating the effects of MME on life-cycle variables in C. tentans. It is recognized that the EEM program focuses on benthic invertebrate community structure and not individual benthic species. However, these studies were necessary to set the scientific basis for developing a culture-controlled multitrophic mesocosm, as well as to provide mesocosm options that might be of value when programs move into investigation-of-cause phases and may require more detailed information--especially in cases where benthic communities are dominated by chironomids.

The C. tentans mesocosm approach was valid to serve as the self-sustaining food base for a fish mesocosm. This would increase the relevance of the fish mesocosm to more natural exposure conditions wherein fish are exposed to effluents through both water and diet. This was a critical improvement, as many metals are known to affect fish through dietary pathways. Another improvement that was required for the fish mesocosm was to increase the relevance and significance of the response variables investigated.

Terrace Bay, Ontario (2003)

Exposure of fish to a contaminant through a partial-life-cycle experiment provides the opportunity to examine the direct effects of effluent on reproduction in adults as well as effects on offspring. In addition, standard EEM effect endpoints (condition, relative liver size, relative gonad size) can also be investigated. Fathead Minnow is a toxicological workhorse used to assess and screen contaminants worldwide for endocrine-disrupting substances. Short-term (7-day), medium-term (21‑day) and long-term (full life cycle) tests have been developed for Fathead Minnow. These tests provide an opportunity to directly assess effects on actual reproductive performance (number of eggs, size of eggs, number of spawning events) as well as more indirect measures such as gonad size. However, almost all of the studies in the literature using Fathead Minnow are water-borne exposures that allow for toxicant screening, but at the expense of greater environmental realism. The first objective was to assess if the 21‑day Fathead Minnow test could be implemented in the field with natural reference as dilution water for PME assessment, holding water temperature and photoperiod constant. The second objective was to link the chironomid C. tentans life-cycle bioassay with the partial-life-cycle bioassay of the Fathead Minnow to develop the multitrophic mesocosm system.

In 2003, a 21‑day Fathead Minnow test was implemented at a pulp mill in Terrace Bay, where reproductive effects on wild fish have been documented. The first objective was to determine the effects of PME on Fathead Minnow at 1% and 100% concentrations (Rickwood et al. 2006a). The second objective was to use the Fathead Minnow test to identify waste stream sources within the mill that affect reproductive indicators (Rickwood et al. 2006b). Various process streams were selected, characterized with respect to effluent chemistry and acute toxicity, and a subset were tested on-site with the bioassay. An enclosed mobile bioassay trailer (photo not shown) was set up on-site at a bleached-kraft mill for 60 days, allowing supply of both ambient water (Lake Superior, Canada) and final PME. This was not an outdoor, exposed mesocosm system, as the interest was in evaluating if the Fathead Minnow 21‑day bioassay could be used with ambient reference water and holding other factors (temperature and photoperiod) constant. The results demonstrated a stimulatory response pattern at 1% PME (e.g., increased egg production, cumulative spawning events) compared to the controls. In the 100% PME treatment, spawning was delayed, resulting in fewer eggs produced in the first 2 weeks of exposure. Exposure to 100% PME also resulted in ovipositor development in males and development of male secondary sex characteristics in females. The results for the second objective showed that both the combined mill effluent (before secondary treatment) and the combined alkaline stream (CALK) caused decreased spawning events (~ 55% for both streams) and decreased egg production (28% and 74%, respectively), and the CALK stream resulted in significant male ovipositor development. By comparing response patterns, the CALK stream was identified as a source of the compounds affecting reproductive indicators in Fathead Minnow at this mill.

Junction Creek, Sudbury, Ontario (2004-2005)

Development of the life-cycle bioassay in mesocosms with the chironomid C. tentans, and the 21‑day partial-life-cycle bioassay with Fathead Minnow using natural reference water as dilution water, established the foundation for a multitrophic mesocosm bioassay. The objective was to develop a self-sustaining trophic-transfer bioassay, using C. tentans and Fathead Minnow, that made it possible to assess the effects of water-borne (Fathead Minnow only) and food- and water-borne (trophic-transfer) exposure to MME. The reproductive performance of Fathead Minnow was assessed for 21 days under controlled laboratory conditions to obtain baseline data on various parameters, including egg production and hatching success (Rickwood et al. 2006c). Exposure to MME #3 (see above) was then conducted for a further 21 days in the laboratory. It was evident that reproductive output in both the water-only and the trophic-transfer system was reduced compared to controls. It was only in the trophic-transfer system that a significant reduction in larval hatching and an increase in deformities occurred after exposure to the MME. This would suggest that contaminated food was an exposure pathway for effects on offspring.

The multitrophic mesocosm was then taken out into the field for application in 2005 (Rickwood et al. 2008). The objectives were to assess (1) the effects of a mine effluent and municipal wastewater mixture on Fathead Minnow reproduction in an on-site artificial stream and (2) the importance of food (C. tentans) as a source of exposure using a trophic-transfer system. Exposures to the effluent mixture through the water significantly reduced egg production and spawning events. Exposure through food and water using the trophic-transfer system significantly increased egg production and spawning events. Embryos produced in the trophic-transfer system showed a similar hatching success, but also showed an increased incidence and severity of deformities after exposure to the mixture. It was concluded that the effects of the effluent mixture on Fathead Minnow were more apparent in water-borne exposures. Exposure through food and water may have reduced effluent toxicity, possibly due to increased nutrients and organic matter that may have reduced metal bioavailability.

Wabigoon River, Dryden, Ontario (2006)

This study investigated the link between PME and endocrine disruption in an attempt to explain the presence of intersex fish in the Wabigoon River, Ontario (Table 9‑1; Pollock et al. 2009). A field survey of the Wabigoon River near Dryden, Ontario, in the fall of 2000 found intersexed Walleye (Sander vitreus vitreus) with significantly altered hormone levels and reduced gonad size. The Wabigoon River receives discharge from a bleached-kraft pulp and paper mill and MSE. It also has historical wood-fibre mats contributing to extended periods of low dissolved oxygen under low-flow drought conditions. A partial-life-cycle test was conducted exposing Fathead Minnow to reference water and to 20%, 40% and 60% PME in field mesocosms. A field survey of Walleye in the Wabigoon River was also conducted. Testosterone decreased in males with increasing effluent concentration, and vitellogenin induction occurred in males exposed to 60% PME. These results did not reflect the magnitude of endocrine disruption seen in the wild fish survey. Several hypotheses have been proposed to explain these discrepancies. Specifically, evidence from published studies indicated that either hypoxia or MSE, alone or in combination with PME, may explain the discrepancy between the field experiment and the wild fish survey. Later studies at this site have examined the effects of low dissolved oxygen levels on Fathead Minnow as well as the interactive effects of low dissolved oxygen (6.0 milligrams per litre [mg/L] no-effect level) and PME (40% no-effect level) (Dubé unpublished).

Top of Page

2007-2009, Ongoing Mesocosm Development

The development of fish mesocosm applications has been consistent and ongoing, with improvements in the methodology with each application. The assessment of effects on small-bodied fish endemic to the system of interest, or assessment using a partial-life-cycle Fathead Minnow test in outdoor mesocosms (water-borne or trophic-transfer experimental designs), is now fairly straightforward. Future investigations are now focusing more on applications in the investigation of cause for the metal mining EEM program.

Junction Creek, Sudbury, Ontario (2008-2011)

Due to the complexities of aquatic ecosystems, the understanding of how metals and metal mixtures affect river food webs is limited. Furthermore, the assessment of effects is only the first step toward the mitigation of those effects and, ultimately, sustainable development. Understanding the causes of the effect (e.g., causative metals) and the factors that modify toxicity is the next step toward the investigation of solutions. Ongoing mesocosm research in Sudbury, Ontario, will i) confirm responses of Fathead Minnow to MME on-site using self-sustaining multitrophic-level bioassays; ii) contrast and compare minnow response patterns to whole effluent mixtures relative to effluent-equivalent doses of single metals of potential concern (copper [Cu], selenium [Se] and thallium [Tl]); iii) ascertain the relative importance of water and diet as the pathway of exposure causing toxicity of metals to Fathead Minnow; and iv) explore factors with the potential to modify toxicity (pH/alkalinity and natural organic matter, and diet quality and quantity) of effluent mixtures and dominant single metals (Cu, Se and Tl) to Fathead Minnow (Dubé et al. unpublished).

Key Lake, Saskatchewan (2007-2011)

This study is being conducted over 4 years, also using a combination of field-based and laboratory-based mesocosm studies (Driessnack et al., unpublished, Dubé et al. unpublished). The objective of the laboratory mesocosm study is to assess changes in an aquatic food chain (multitrophic mesocosm), including reproductive output of Fathead Minnow due to exposure to a uranium effluent. In addition, a comparative evaluation of Fathead Minnow response patterns to the effluent mixture vs. Se as selenate will be conducted. This experimental design isolated the contribution of Se from that of the effluent mixture. Results illustrated that the response patterns in egg production could not be explained by Se in the comparison between treatments and relative to controls. The objective of the field mesocosm study was to determine the relative and cumulative contribution of water-borne (current) vs. sediment-borne (historical) Se contamination on the reproductive success and survival of breeding Fathead Minnow and their offspring. Results showed that effects on fathead minnow were exclusively effluent mediated with insignificant contributions from contaminated sediments. The sediments tested in the study were of sand composition as it represented the largest sediment type in the Key Lake drainage. Further work is required to determine the significance of organic sediments to responses and in the context of their spatial and temporal distribution at the site.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.1 Background Information on Artificial Stream Development and Application

Figure 9-1: A) Large mesocosm system with streams situated on tables (Model I) used in the Athabasca River, Alberta. B) Large mobile mesocosm system with streams on 2 trailers (Model II) used in the Fraser River, British Columbia; the Saint John River, New Brunswick; and in Saint John Harbour, New Brunswick. C) Large mobile mesocosm system with streams on a single trailer (Model III) used in the Miramichi and Little rivers, New Brunswick; the Wapiti River, Alberta; and Junction Creek, Ontario.

Figure 9-1: A) Large mesocosm system with streams situated on tables (Model I) used in the Athabasca River, Alberta. B) Large mobile mesocosm system with streams on 2 trailers (Model II) used in the Fraser River, British Columbia; the Saint John River, New Brunswick; and in Saint John Harbour, New Brunswick. C) Large mobile mesocosm system with streams on a single trailer (Model III) used in the Miramichi and Little rivers, New Brunswick; the Wapiti River, Alberta; and Junction Creek, Ontario.

Top of Page

Figure 9-2: A) Small microcosm system with streams situated on tables over  mixing reservoirs used in the Thompson River, British Columbia. B) Modular  mesocosm system with streams situated on tables over mixing reservoirs used in  the Little River, New Brunswick; Junction Creek, Ontario; the Wabigoon River, Ontario; and Key Lake, Saskatchewan.

Figure 9-2: A) Small microcosm system with streams situated on tables over mixing reservoirs used in the Thompson River, British Columbia. B) Modular mesocosm system with streams situated on tables over mixing reservoirs used in the Little River, New Brunswick; Junction Creek, Ontario; the Wabigoon River, Ontario; and Key Lake, Saskatchewan.

Top of Page

Figure 9-3: Schematic of large mesocosm trailer system

Figure 9-3: Schematic of large mesocosm trailer system (not to scale).

Top of Page

Figure 9-4: Photograph of a modular mesocosm parts diagram

Figure 9-4: Photograph of a modular mesocosm parts diagram.

Top of Page

Figure 9-5: Modular mesocosm flow schematic

Figure 9-5: Modular mesocosm flow schematic.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.1 Background Information on Artificial Stream Development and Application

Figure 9-6: Multitrophic Fathead Minnow reproductive bioassay and feeding barrier

Figure 9-6: Multitrophic Fathead Minnow reproductive bioassay and feeding barrier.

Top of Page

Figure 9-7: Site set-up for modular mesocosms

Figure 9-7: Site set-up for modular mesocosms.

Top of Page

Figure 9-8: A) Factorial  experimental design to investigate the importance of water vs. diet in  responses of Fathead Minnow to metal mine effluent in modular mesocosms; B) Experimental design to investigate the influence of pH and natural organic  matter (NOM) on Fathead Minnow responses after exposure to an MME mixture and a  single metal in multitrophic modular mesocosms.

Figure 9-8: A) Factorial experimental design to investigate the importance of water vs. diet in responses of Fathead Minnow to metal mine effluent in modular mesocosms; B) Experimental design to investigate the influence of pH and natural organic matter (NOM) on Fathead Minnow responses after exposure to an MME mixture and a single metal in multitrophic modular mesocosms.

Top of Page

Figure 9-9: Factorial experimental design to investigate the effects of MME and  historical sediment contamination in isolation and in combination on Fathead  Minnow in modular mesocosms

Legend: RWRS: Reference Water and Reference Sediment; EWRS: Effluent Water and Reference Sediment; RWCS: Reference Water and Contaminated Sediment; EWCS: Effluent Water and Contaminated Sédiment

Figure 9-9: Factorial experimental design to investigate the effects of MME and historical sediment contamination in isolation and in combination on Fathead Minnow in modular mesocosms.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.2 Applicability within the EEM Programs

It should be emphasized that while mesocosms are a recommended monitoring alternative, their use should only be considered when field surveys cannot be designed to unequivocally answer the hypothesis, or simply cannot be conducted. Such situations include confounded receiving environments or where unsafe sampling conditions exist. Examples of confounded receiving environments include areas with the presence of historical effects, the absence of suitable reference areas for comparison to exposure areas, the presence of other effluents, and changes in relevant habitat types that cannot be factored out in the design of a field survey. Mesocosms can also be used for assessment of magnitude (dilution series) and for investigation of cause. The above-described case studies illustrate the different types of questions and experimental manipulations that can be applied to the different phases of the EEM program.

It may be possible to use the fish from a fish survey mesocosm study for the fish usability component of EEM as well, provided that the effect endpoint has the potential for responding over exposure periods typical of mesocosm studies. In this case, the effect endpoint and statistical procedures would be as outlined in Chapter 3.

There are advantages and disadvantages to using mesocosms, and these should be weighed against the advantages and disadvantages of other monitoring alternatives before the final selection of an approach.

Top of Page

9.2.3 Mesocosm Technology Suitability of Mesocosm Design

To maintain data quality and ensure consistent application of mesocosm technologies in EEM programs, a review of the physical design of the system used and the study design, including standard operating procedures, is required. This section outlines physical design guidance and experimental design recommendations for both a large trailer mesocosm system as well as for the smaller modular design.

Regardless of the system used, general operation is the same. Replicated tanks or streams hold the biota (fish, benthic invertebrates), algae, and/or substrate of interest. A total of 5 to 8 replicate streams per treatment is considered adequate. The systems are not static but either completely flow-through (trailer mesocosm) or partially recirculating (modular mesocosm). Ideally, reference water and dilution water are collected from a reference site and delivered to a head tank. Treated effluent is collected daily, or no less frequently than weekly, and stored on-site in a head tank. Head tank liquids can be heated or chilled depending on the circumstances. Mixing tanks are used to mix reference water and effluent to the desired test concentration. Mixed water or “treatment water” is delivered from each mixing tank to the mesocosm streams at a flow rate to achieve a target turnover time (minimum of one turnover or complete stream volume exchange every 24 hours and up to 6 turnovers per 24 hours). The experiment is typically run for 30 to 65 days, with daily and weekly measurements taken of physical and chemical variables, flows, fish mortality, and reproduction (if eggs are collected daily, for example). At the end of the exposure period, fish are examined and supporting measures are taken.

Top of Page Description and General Operation of Large Trailer Mesocosm System

The large trailer mesocosm system, developed by the National Water Research Institute of Environment Canada, has been used to assess PME and MME effects on benthic food webs (benthic invertebrates and algae) (Culp and Podemski 1996; Culp et al. 1996, 2000a, 2001; Cash et al. 2003) and small-bodied fish (Dubé 2000; Dubé and MacLatchy 2000a; Cash et al. 2003; Dubé et al. 2002). This system has been used for benthic invertebrate community assessments as well as water-borne exposures with fish. The system can still be used, although most applications now use the modular mesocosm system described in the next section.

The large trailer stream system consists of 16 circular tanks or streams, with a surface area of 0.9 m2, placed in pairs on tables that are 74 cm high (figures 9‑1 and 9‑3). Water from a reference area is pumped into a 378‑L polyethylene head tank placed on a platform that is 1.2 m high, and gravity-fed through a system of pipes to the streams. Intake pumps such as a land-based pump (e.g., commercially available irrigation or pool pumps) or a submersible pump (0.5‑HP Hydro-Matic Model # SPD50‑H) can be used, depending upon volume requirements, pumping distance, and the head required. Pumping reference water from a mine water intake is also a possibility. Gate valves control water flow to individual streams and allow flow rate calibration for each stream. Water is delivered to each stream at a rate of 2 L/min, resulting in a total water requirement of 32 L/min for the 16‑stream system.

Water depth in the streams is maintained at 26.9 ± 0.1 cm (x-bar ± 1 SE) by an overflow drain that returns all wastewater to the river. The overflow drains are screened to prevent fish loss and limit emigration of insects from the streams. Each stream contains 227 L of water, resulting in a hydraulic residence time of approximately 2 hours. By increasing water residence time within the streams to 4 hours, the volume of effluents and reference water required during a study is minimized. Final determination of residence time for a particular system depends upon the size of the tank used and species requirements for oxygen and temperature. If mesocosm studies are conducted in the late autumn, the head tank and water delivery lines can be wrapped with heat tape and insulated to allow the system to be operated when freezing temperatures may occur (‑5°C). Shade cloth can be used to reduce solar heating in the summer.

The streams on the large mesocosm system are tanks, 107 cm in diameter, constructed of polyester fibreglass. Streams are placed on 8 tables that are 74 cm high, 2 to a table. The water outflow pipe passes through a standpipe and drains into pipes beneath. These drain pipes connect to a general outflow pipe for discharge to the river downstream of the water intake point.

For invertebrate applications, current velocity in each stream can be created using a propeller system (Podemski 1999). Water velocity in each stream near the water column midpoint is normally maintained at 20 cm/sec, although site-specific velocities should be the determining factor. Other current-generation mechanisms can be used, provided they produce the water velocities. In fish studies, current can also be generated using the propeller system, although spray bars attached to the water delivery system for each stream have also been used for species where water velocity is not a critical requirement.

Effluent is collected daily or every second day and stored in polyethylene containers. The point of effluent collection depends upon the study design. For the first monitoring study at a metal mine, for example, the final effluent that is representative of what is being discharged to the aquatic environment is the target source. Effluent treatments are delivered independently and continuously to individual streams by peristaltic pumps (Masterflex ® L/S Nema-type 13 wash-down controllers and cartridge pump heads). Effluent flow rates depend upon the site-specific, environmentally relevant concentrations to be tested. For example, if plume delineation studies in the field have determined effluent concentrations to be 1%, then an effluent flow rate into each stream is set at 20 ml/min (1% of 2 L/min). If there are 16 tanks on the trailer, which are allocated between 2 treatments (8 streams for control; 8 streams for 1% exposure), then the following effluent volume is required on a daily basis: 20 ml/minute x 60 minutes/hour x 24 hour/day = ~ 29 L/day/stream x 8 streams = ~ 231 L/day. This volume would fit into a small polyethylene container.

After water and effluent flows have been calibrated for each system, the individual tanks or streams are seeded with natural substrates, algae, benthic invertebrates and/or fish species endemic to the receiving environment being studied. Inoculation of biotic populations is described below for fish and invertebrates. Further details on the construction of this trailer system can be found in the literature referenced herein.

Top of Page Description and General Operation of Modular Mesocosm System

Each modular mesocosm unit or table consists of a shipping pallet, metal frame, wet table, up to 8 replicate circular polyethylene streams, a reservoir for holding the exposure solution, a manifold for equal flow distribution to the replicate streams, a blue Viking pump for flow delivery from the mixing tank to the reservoir, and an orange March pump for flow recirculation within the unit from the reservoir to the manifold/stream complex (figures 9‑4 and 9‑5).

Each mesocosm holds a total capacity of 185 L of water (85‑L reservoir, 82.4 L = 10.3 L/stream x 8 streams, ~ 15 L in manifold/hoses/pumps) and requires 3.0 amps to run without heating, cooling or aerating the water. Each replicate circular stream has a volume of 10.3 L. There is a central non-functional standpipe in each stream. The circular high-density polyethylene streams sit on top of a table that drains into an 85‑L dilution reservoir. Each stream requires a cover to keep fish in and other things (e.g., birds) out. A window screen or Nytex mesh cover is used, and secured with bungee cords. Each table has 8 polyethylene streams of 10.3 L each that are custom-made and moulded. The reservoirs are 85‑L polyethylene plastic totes that are also custom-made and moulded. The 8‑port manifold system was custom-designed to allow for equal flow distribution to each stream without requiring 8 separate pumps. There are 2 types/sizes of tubing in the mesocosm system to deliver water to and from the manifold and to the streams (internal diameter [ID] 3/8” and ID 3/4”).

An electronic metering pump (blue Viking/Pulsatron pump, Series E 240 GPD LEH 75A-PHC3-XXX) controls the water/effluent turnover times in the reservoir. The March “Series 3” Seal-less Magnetic Drive Centrifugal Pump (orange pump) controls the movement of water/effluent from the reservoir to the streams. Pumps are the most expensive part of this system and the most critical component of the study. An incorrect flow rate means incorrect dilutions to the test organisms.

Each mesocosm unit or table represents a treatment with 8 replicate streams per treatment. Note that it can and has been argued that the systems are pseudo-replicated at the level of the reservoir. This point has been successfully argued and defended through the peer-review publication process. For fish mesocosm application, streams can be fitted with feeding barriers for multitrophic studies. This mesh barrier will allow a benthic invertebrate culture to develop under treatment conditions, while controlling access for the fish above it (Figure 9‑6).

Researchers have experimented with many different forms of flow delivery, including in-line mixing pumps. Based on experience, and in the interest of keeping costs manageable, mixing the treatment solutions in a mix tank and delivery using a pump to each mesocosm reservoir has become the preferred option over that of in-line mixing of water + contaminant. Water is delivered from a mixing tank to a mesocosm reservoir using a blue Viking pump at a rate of 1-4 turnovers every 24 hours (Figure 9‑5). There is an overflow drain at the back of each reservoir and a baffle inside each reservoir to prevent short-circuiting of the inflow to the overflow drain.

The March pump is the recirculating pump. Flow moves from the March pump through the manifold, into the streams, overflows from each stream into the wet table, and then drains back into the reservoir through the hole in the table. The manifold must be level and free of air bubbles to operate effectively. All tubing must be the same length for the manifold to be able to pressurize equally and deliver water at an even flow rate to each of the streams. Water enters the stream along the length of water inlet tubes threaded through the stream wall. Because of the angle at which the water enters the stream, a slight circular current is created. Water fills the streams and then drains over the top, collects briefly on the tabletop, and is rerouted back down to the reservoir. Reservoirs are insulated using silver insulation sheets wrapped around the outside of the tank.

Top of Page

9.2.4 Considerations for Site Selection

Irrespective of the mesocosm used, there are some basic site requirements. The system is typically located at a reference area for access to reference water. In a freshwater riverine situation, reference water is pumped from upstream of the effluent outfall into the system and discharged back to the river downstream from the point of intake. This is the most straightforward scenario for mesocosm use. Distance to the reference water source is also a consideration for site selection, as extensive pumping requirements with respect to distance, or elevation (head), can exceed pump specifications for the rate required. Other site requirements include adequate space, site access, electrical power and security.

Access to power is one of the major site-selection requirements. The facility typically provides this, via a power line that the facility’s electricians install into the mesocosm’s power panel on either the trailer or the power pallet designed specifically for the modular mesocosm system. In remote areas, if power is not available, generators have been used to power the system, although this approach is not recommended. Due to maintenance and supervision needs, generator use is often not cost-effective.

During site set-up, the process consists of selecting the site, unloading equipment, and placing the head tanks and mixing tanks at the end opposite the power pallet (in the case of the modular system). Modular mesocosms are typically placed under a tent for shade, to reduce particulate deposition (dust), and for security purposes. The head tanks, mixing tanks and power pallet do not go under the tent (Figure 9‑7). The modular mesocosm or trailer is placed in a north-south direction perpendicular to sunset/sunrise. First, all equipment is placed, then power is connected, mesocosm tables are levelled, tubing is run from head and mixing tanks to streams, all tanks are filled, pumps are turned on and calibrated, tents are set up (for modular mesocosms), and then animals are added.

It is critical to calculate the electrical requirements of the system for the experimental design selected and to assess the site to determine adequate electrical availability. Power reliability at industrial sites is an important consideration.

Top of Page

9.2.5 Biological Monitoring Study Designs Overview

The objective of a study design is to outline what mesocosm and associated laboratory work is needed to complete the biological monitoring portion of the EEM study. Many of the study components are similar to those outlined in the various chapters of this document.

A study design is submitted to the Authorization Officer at least 6 months prior to the commencement of sampling for biological monitoring studies. The study design will include:

  • a site characterization;
  • a description of how the studies respecting the fish population and fish tissue will be conducted and how these studies will provide the information necessary to determine if the effluent has an effect on the fish population and fish tissue;
  • the fish species, sampling areas, and sample size selected;
  • A detailed timetable for conducting the mesocosm study;
  • a description of how the study respecting the benthic invertebrate community will be conducted and how this study will provide the information necessary to determine if the effluent has an effect on the benthic invertebrate community;
  • the dates and times that the samples will be collected for the biological monitoring;
  • the field and laboratory methods selected;
  • a description of the quality assurance and quality control (QA/QC) measures that will be implemented to ensure validity of the data that is collected; and
  • a summary of the results of any biological monitoring studies that were submitted previously.

Other recommended details of the study design may include:

  • defining the goals and expectations of the EEM study;
  • determining the overall approach, including stating the rationale for choosing an alternative, which may be based on previous monitoring results;
  • establishing statistical design criteria: development of hypotheses, selection of statistical methods, determination of data needs (statistical significance and power analysis);
  • developing operating plans and procedures: sampling procedures, laboratory analysis procedures, QA/QC procedures, data storage and retrieval, data analysis; and
  • describing a plan for data interpretation and program evaluation.

Top of Page Study Treatments

In mesocosm studies, the effect of effluent on fish is evaluated by comparing effect indicators (growth, reproduction, condition and survival) between fish held under control conditions (reference water) to those held in effluent. In mesocosm studies using benthic invertebrates, an effect is determined by comparing effect indicators (total benthic invertebrate density, taxa richness, evenness index and similarity index) between invertebrates in reference vs. exposure streams. It is crucial that mesocosm studies be designed to maximize the possibility of detecting effects if they are present. This includes selecting appropriate treatments, level of replication, sentinel species, response variables, and conducting the studies at the proper time of year.

Mesocosm studies provide for controlled experimental manipulation, with the added benefit of environmental relevance (natural water quality, photoperiod, water and air temperature). The flexibility in experimental design is one of the most significant and creative advantages of using mesocosms. Several study designs can be employed, depending upon site-specific requirements and the phase of the monitoring program (i.e., magnitude and geographic extent, and investigation of cause). In the simplest case, 2 treatments (control vs. effluent exposed) are compared. It is recommended that the environmentally relevant concentration of effluent be based upon a plume delineation study conducted during site characterization (Chapter 2); thus it should represent the effluent concentration in the high effluent exposure (near-field) area after complete mixing.

In some cases, additional treatments may be desired. Dose-response study designs where additional and higher effluent concentrations are used are helpful to confirm biota responsiveness and the absence of effects (see Dubé and MacLatchy 2000a; Culp et al. 2000a). Additional treatments may also be desired if more than one mine discharges into the same sampling area. Each discharge can represent an exposure treatment in the experimental design and the effects of each effluent can be examined in isolation or in combination. For example, if a mine discharges 3 effluents into the same receiver, effluent effects can be evaluated using the following treatments: control, effluent 1 (environmentally relevant concentration), effluent 2, effluent 3, effluent 1 + 2 + 3 (environmentally relevant concentration; see Table 9-1).

More recently, modular mesocosms have been used to assess water vs. dietary pathways of exposure for Fathead Minnow exposed to MME. This design was of value for investigating different metals in effluents and their causal contribution to fish response patterns. Fish were held in water (reference or effluent) and fed with the chironomid C. tentans cultured under either control water or effluent exposure conditions (Figure 9‑8A). These results were compared to those of concurrent multitrophic mesocosm treatments. Studies have also been conducted to evaluate the influence of different water chemistry variables in ameliorating the toxicity of a metal and MME on Fathead Minnows by focusing on the effects of increased pH and natural organic matter using multitrophic mesocosms (Figure 9‑8B).

Top of Page Replication

In modular mesocosm systems with benthos or fish, typically 5 to 8 replicates are used. In the trailer mesocosm system employing a control/impact design with 2 treatments (reference vs. exposure), 8 replicates will be used as there are 16 tanks on the mesocosm trailer. If additional treatments are preferred, then 5 replicate streams per treatment should be the minimum number of replicates for EEM studies, as this should provide adequate power for assessing effects when streams are the level of replication and when no previous monitoring data are available. This approach is consistent with the recommended method for determining the number of sampling stations, using statistical power in field survey designs (Chapters 3 and 4).

The unit of replication is less clear when the system is used to measure individual-based response variables in fish. The basis of the decision lies in the quantification of the importance of a tank or stream effect (i.e., an effect of one tank relative to another within the same treatment). In many laboratory studies where fish are held in aquaria, an assumption is made that there is no biological reason for tank differences within a treatment, and thus all individual fish measurements are pooled within a treatment (i.e., the variability attributable to tank effect is often ignored). For mesocosm studies it is suggested that a tank effect may exist, and the variation attributed to that effect requires some consideration in the statistical design. This is especially the case for longer-term exposures (e.g., 60 days). A nested analysis of variance (ANOVA) could be used in this example where the tank effect is nested within the fixed-effect factor of effluent treatment. If a tank effect exists, the level of replication should be at the stream level. If tank effects do not exist, fish can be selected as the unit of replication and pooled across streams within a treatment, resulting in significant increases in the number of replicates. For regression-based analyses (analysis of covariance [ANCOVA]), any tank differences would likely emerge as outliers in the analysis.

Top of Page Sample Sizes and the Role of Effect Size No Available Pre-existing (Historical) Data

Trailer Mesocosm System: In the trailer mesocosm, there is increased capacity to hold fish for longer periods than in the modular mesocosm system. For a standard fish population field survey, where there are no background monitoring data, the minimum sample size recommended is 20 sexually mature males and 20 sexually mature females of 2 fish species collected from each of the reference and exposure areas. If small-bodied fish species are chosen as one or both of the fish species, an additional 20 sexually immature fish should also be collected. The rationale for using 20 fish of each sex is that there is little change in the 95% confidence limits with increasing sample size beyond 20 fish. In trailer mesocosms, 15-20 sexually mature males and 15-20 sexually mature females of one small-bodied sentinel species are added to each tank. In addition, 20-30 juvenile fish are allocated to the same tank. These are recommended sample sizes that can be increased or decreased depending upon the species selected, the statistical power of the study design, and fish variability. If tanks within a treatment are selected as the unit of replication, 20 fish per sex per tank provides a mean with good precision (i.e., there is little change in the 95% confidence limits with increasing sample size beyond 20 fish). If tank effects do not exist and error terms are pooled so that fish become the unit of replication, 15-20 fish per sex per tank provides much higher replication per treatment. For example, a study design of control vs. 1% effluent with 8 streams per treatment, and 20 males, 20 females and 20 juveniles per tank, can result in 160 males, 160 females, and 160 juveniles per treatment if fish are the unit of replication.

Modular Mesocosm System: In modular mesocosm studies using benthic invertebrates or fish, sample size represents the number of streams replicated per treatment, as streams are the unit of replication. In a control-impact design, 8 streams are replicated per treatment. Each mesocosm table represents a treatment and thus this simplest design would only require 2 mesocosm tables for assessment. If additional treatments are required, a minimum of 5 replicates or samples are required for each treatment to address power requirements without pre-existing data (see below). Breeding pairs or trios can be used in each stream depending upon the objectives of the study. In the protocol developed by Ankley et al. (2001), a ratio of 2 males to 4 females (2M:4F) is used per replicate. However, this breeding ratio is too high for the mesocosm streams and certainly too many fish to sustain in a self-sustaining multitrophic mesocosm test over 21 days of exposure. Thus, the use of pair or trio (1M:2F) breeding is recommended. If breeding trios with Fathead Minnow are used in a modular mesocosm design, data from both females are assessed with a measure of central tendency keeping the level of replication at the stream level.

It is important to consider the objectives of the fish study when selecting which mesocosm system to use. While the trailer system provides for greater numbers of fish in each replicate, exposures are water-borne only and effect endpoints measured are condition, survival, organ size in adults and growth in juveniles, the latter as a substitute for size-at-age. The modular mesocosms have lower fish numbers per replicate but allow for multitrophic and/or water-borne exposures, and allow for investigations over a partial life cycle from breeding to offspring production over time in a repeat spawner such as Fathead Minnow. Reproductive variables, such as cumulative numbers of eggs produced per female per day, can be replicated over time, and distributions in egg production can be assessed for each treatment using distribution-based statistical tests such as the Kolmorogov-Smirnov test.

Top of Page Available Pre-existing (Historical) Data

If data are available, the sample sizes to measure a certain effect size in a parameter with a targeted level of statistical power can be calculated, because sample variability is known. Larger numbers of samples are needed if parameters are highly variable, if detection of small differences between reference and exposure streams (small effect size) is desired, or if a high level of power is required. To determine the sample sizes using pre-existing data, the effect size and the statistical power level that is acceptable for the decision-making process needs to be determined. The purpose of defining an effect size and power level is to determine when the sampling program is collecting adequate information to provide decision support. Sample sizes can be calculated using methods described in Chapters 3 and 4. Appendix 1 of Environment Canada (1997) provides a detailed discussion of power relationships, effect sizes and the benefits of reducing variability in terms of increasing power. It is recommended that α and β be set equally at 0.10 for power calculations.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.6 Fish Monitoring – Effects Assessment Study Design Species Selection

The most important factors when selecting fish species for mesocosm studies are environmental relevance, abundance for collection, size, availability of adults and juveniles, sexual dimorphism, spawning period, and sensitivity to effluent. The recommended method for assessing effluent effects in mesocosm studies conducted in the trailer mesocosm system is by monitoring pre-spawning adults (sexually mature fish) and juveniles of one small-bodied fish species (e.g., darters, minnows, sculpins) that is relevant to the receiving environment. Monitoring of adults provides for assessment of effluent effects on survival, reproduction (energy use) and condition (energy storage). Monitoring of juveniles provides for assessment of effluent effects on survival, energy use (i.e., growth as a size-at-age substitute) and energy storage (i.e., liver size and condition).

Species selection for mesocosm studies is restricted by size requirements (numbers of fish per tank decrease with increasing fish size) and sexual dimorphism (if males and females cannot be externally sexed, increased numbers should be considered to ensure adequate sample sizes). As such, mesocosms are best suited for use with small-bodied fish species. Sampled species should also be suitable for measuring the recommended variables. To date, fish species used in mesocosm studies include Mummichog, juvenile Atlantic Salmon, Slimy Sculpin, Creek Chub, Longnose Dace and Fathead Minnow (Table 9‑1).

The advantages and disadvantages of using small-bodied species in field surveys have been described by Gibbons et al. (1998a, 1998b). A small-bodied fish can be considered as a fish species that has a maximum length of 150 mm. Information on maximum growth of fish species can be found in the scientific literature, including Scott (1967), Scott and Crossman (1973), Fritz et al. (1975), Roberts (1988), Nelson and Paetz (1992), Jenkins and Burkhead (1993), Coad (1995), and Leblanc and Couillard (1995). Small-bodied fish species are usually more abundant, easier to capture, and more sedentary than larger-bodied fish species. Smaller home ranges are desirable, as it increases probability and consistency of effluent exposure compared to larger, more mobile, possibly migratory species.

There can be disadvantages to using small-bodied fish. Often, less is known about their basic biology, particularly their spawning habits, making it difficult to determine the best sample areas, times and methods. However, due to the significant amounts of data collected on small-bodied fish as part of EEM programs as well as many Canadian research studies, knowledge of small-bodied fish life-history strategies and basic biology has increased significantly (Munkittrick, University of New Brunswick, unpublished). Some species are multiple spawners (i.e., they produce more than one clutch of mature ova every year; see Heins and Rabito 1986; Burt et al. 1988; Paine 1990). This can be a disadvantage for field surveys because reproductive effort in these species is difficult to estimate from a single sample. Reproductive tissue can be turned over almost completely between clutches (i.e., most of the mass of ova in the ovary will be spawned and then a new clutch of mature ova will be developed). The number of clutches produced during the spawning season becomes the important reproductive variable and is difficult to estimate for an individual female in the field, even with frequent sampling. However, multiple spawners such as the Mummichog have been used successfully in mesocosm studies to evaluate effluent effects on EEM effect endpoints, including changes in gonad size (Dubé and MacLatchy 2000a; Cash et al. 2003; Dubé et al. 2001). An advantage of using controlled-exposure mesocosm studies is that the state of fractional spawners can be monitored throughout the exposure period.

The latest advances in mesocosm technology and application using the modular system has evolved around the use of Fathead Minnowin a 21- to 30-day partial-life-cycle exposure experiment (Table 9‑1). Mesocosm applications using the trailer system have been successful in assessing effluent effects. However, exposures are water-borne, which lessens environmental relevance. Water-borne exposure is standard practice in fish toxicological assessments with single contaminants, but given the nature of complex mixtures, dietary pathways of exposure should be considered. For example, numerous studies have investigated the importance of the trophic-transfer of metals (Ni et al. 2000; Chen et al. 2000; Mason et al. 2000; Xu and Wang 2002) in aquatic environments. Including dietary exposure in the mesocosm approach would therefore be an improvement. The ability to directly assess reproductive output (number of eggs produced, number of spawning events and offspring survival, hatching success, deformities) after exposure to effluent, in addition to the standard EEM effect endpoints, would also be desirable for more causal investigations.

Fathead Minnows have been used extensively as a toxicological workhorse in laboratory investigations, as they represent an ecologically significant part of the Cyprinidae family, they have been extensively tested, and a large database of knowledge exists regarding their culture and life cycles (Panter et al. 2002; Ankley et al. 2001; Jensen et al. 2001). They are also used in risk assessment and government/industry monitoring studies on an international scale (US EPA 1982, 1996, 1999, 2002; OECD 2001; Shaw et al. 1995a, 1995b). In addition, Fathead Minnows are small (average length of 6 cm and width of 1 cm), fractional, substrate spawners that, under specific conditions, can easily be manipulated in captivity to produce clutches of 50-150 eggs every 3-5 days. They are also an environmentally relevant species, as they are abundant in freshwater systems across Canada. Thus, the development of a mesocosm approach using Fathead Minnows to measure EEM effect endpoints and provide focus for direct evaluation of how reproductive output is affected by effluent mixtures would be highly useful in cases where field surveys cannot be conducted. Furthermore, self-sustaining (no external food source) mesocosms wherein fish and their diet are co-cultured, resulting in fish exposure through both dietary and water-borne pathways (i.e., a multitrophic mesocosm), would also be highly relevant.

Ankley et al. (2001) developed a short-term bioassay using Fathead Minnow that assesses reproduction as well as aspects of early development in a shorter time frame than traditional life-cycle assays. The time frame of the partial-life-cycle test (21 days) made it a suitable candidate for mesocosm use. A number of investigations have used the 21‑day bioassay to monitor effects of estrogenic (Harries et al. 2000; Ankley et al. 2001; Sohoni et al. 2001), androgenic (Ankley et al. 2003) and anti-androgenic (Jensen et al. 2004) compounds. These studies have predominantly focused on single contaminants that do not represent the complexity of effluents. A limited number of investigations into the effects of industrial effluents using the 21‑day bioassay have been conducted (Martel et al. 2003; Parrott 2005). However, these studies have used differences in methodology (e.g., number of independent replicates, number and type of variables measured) and were conducted in the laboratory under water-borne exposure conditions.

Development of the Fathead Minnow 21‑day bioassay for use in EEM mesocosm studies first required in situ use with natural receiving water, yet controlled temperatures and photoperiods (Rickwood et al. 2006a, 2006b). As the chironomid C. tentans life-cycle bioassay had been previously established for use in the modular mesocosm system (Hruska and Dubé 2004, 2005), the next step was to combine these tests into a multitrophic mesocosm test for use in the modular mesocosm system. This was first done in the lab with controlled water temperature (Rickwood et al. 2006c) and then moved to the field, with full testing under ambient temperatures and photoperiods and using ambient reference water (Pollock et al. 2009; Rickwood et al. 2008). The Fathead Minnows used in these studies have been laboratory cultured and transported to the site to control for fish quality, age, state of reproductive development and exposure (or lack thereof).

Top of Page Effect Indicators

Trailer Mesocosm System: The effect indicators measured in the trailer mesocosm study using fish are the same as those measured in a field survey. Effects on fish growth, reproduction, condition and survival are determined in answer to the question, “Has the fish population been modified by effluent?” Survival, growth and reproduction (energy use), and condition (energy storage) are measured to detect any effluent-related effect on the fish population. All of these measurements can be taken from fish exposed to effluents in mesocosm studies (Table 9‑2).

Growth is the change in size (weight or length) with time or age. Mesocosm studies over a longer exposure period (45-60 days) can assess effluent effects on juvenile growth, specifically changes in total body weight and length relative to control fish. In mesocosms, growth is measured in YOY (young of the year) or juvenile fish from the start of the study to the end, a duration of 6‑8 weeks depending upon the temperature during the study. The determination of growth effects over a short exposure period requires that studies be conducted during warm water temperatures and with a sentinel species that has a high growth rate. Kneib and Stiven (1978) have shown that juvenile growth rates of Mummichog, for example, are very high post-hatching (an increase of 15 mm in 2 months), which makes them ideal candidates for assessing the effects of effluent on growth. Other studies using YOY Slimy Sculpin have shown that growth effects can be measured after only 26 days of effluent exposure (Dubé et al. 2005). Although length and weight measurements are taken in adults, growth is unlikely during the short duration of the mesocosm studies.

Top of Page

Table 9-2: Fish mesocosm study effect indicators and endpoints and related statistical procedures (text description)
Effect IndicatorsEffect and Supporting EndpointsStatistical Procedure
  • *Change in size (weight and length) with time (end measurements compared to starting measurements)
  • Juvenile growth (change in length over time)
  • Juvenile growth (change in whole animal wet weight over time)
  • Adult body weight (whole)
  • Adult length
  • *Relative gonad size (gonad weight against body weight [adults])
  • Relative gonad size (gonad weight against length [adults])
  • Fecundity (number of eggs/female against body weight, length and/or age); n/a for juvenile fish and males
  • Relative egg size (mean egg weight against body weight)

Juvenile and Adult Condition

  • *Body weight relative to length (k)x
  • *Relative liver size (liver weight against body weight)
  • Liver weight against length

Juvenile and Adult Survival

  • *Percentage surviving at the end of the exposure period

© M. Dubé

* Mesocosm effect endpoints used for determining effects as designated by statistically significant differences between exposure and reference streams. Other supporting endpoints can be used to support analyses.

Top of Page

Table 9-3: Recommended response variables and suitable additional supporting information, and suggested statistical analysis for Fathead Minnow application in modular mesocosm systems (text description)
Type of ResponseResponse VariableDependent Variable (Y)Indepen-
dent Variable (X)
riate (X)
Statistics (Single Factor)Statistics (Two Factors)
Stage of experiment: Pre-exposure (for fish that met criteria)
Energy Use (Adults)Mean total egg production (number)Mean total number of eggs per breeding groupTreatmentn/aANOVA
Mean egg production (number)Mean number of eggs per female per dayTreatmentn/aANOVA
Cumulative eggs per female per day (number)Cumulative eggs per femaleDayn/aKolmogorov-SmirnovKolmogorov-Smirnov
Mean total spawning events (number)Mean total number of spawning events per breeding groupTreatmentn/aChi-SquareChi-Square
OffspringFertilization success (%)Number of eggs fertilized/ number of eggs laid x 100Treatmentn/aANOVA
Stage of experiment: Exposure
Adult SurvivalMean adult survival (%)Number of adults surviving at end/Number of adults at start x 100Treatmentn/aANOVA
Energy Storage (Adults)Condition (g/cm)Total body weight (log)n/aLength (log)ANCOVAANCOVA
Relative liver size (g)Liver weight (log)n/aTotal body weight (log)ANCOVAANCOVA
Liver weight (log)n/aLength (log)ANCOVAANCOVA
Relative egg size (µm)Mean egg size (log)n/aTotal body weight (log)ANCOVAANCOVA
Mean egg size (log)n/aLength (log)ANCOVAANCOVA
Energy Storage (Adults)
(If ANCOVAs cannot be done due to lack of spread on x axis)
Mean condition factor (%)Total body weight/ (length)^3 * 100Treatmentn/aANOVA
LSI (%)Liver weight/body weight * 100Treatmentn/aANOVA
GSI (%)Gonad weight/body weight * 100Treatmentn/aANOVA
Energy Use (Adults)Total body weight (g)n/aTreatmentn/aANOVA
Fork length (cm)n/aTreatmentn/aANOVA
Relative gonad size (g)Gonad weight (log)n/aTotal body weight (log)ANCOVAANCOVA
Gonad weight (log)n/aLength (log) ANCOVAANCOVA
Cumulative eggs/breeding group/day (number)Cumulative eggs/
breeding group
Mean total egg production/
day (number)
Mean total number of eggs/female/
Mean egg production/ day (number)Mean number of eggs/ female/dayTreatmentn/aANOVA
Cumulative spawning events/
breeding group/day (number)
Cumulative total number of spawning events/
breeding group
Mean total spawning events/day (number)Mean total number of spawning events/
Relative fecundity (number)Number of eggs/female (log)n/aTotal body weight (log)ANCOVAANCOVA
Number of eggs/female (log)n/aLength (log)ANCOVAANCOVA
Other Adult Reproductive ResponsesDevelopment of secondary sexual characteristicsPresence/
absence of tubercles; banding;
fin dot; fat pad
Rank of ovipositor sizeTreatmentn/aChi-SquareChi-Square
Reproductive hormonesMale/female testosterone (ng/g)Treatmentn/aANOVA
Female estrogen (ng/g)Treatmentn/aANOVA
Male vitellogenin (ng/g)Treatmentn/aANOVA
Gonadal histologyNumber and stage of cellsTreatmentn/aChi-SquareChi-Square
OffspringMean hatching success (%)Number of eggs hatched/
number of eggs fertilized x 100
Mean total deformities (%)Total number of deformities at hatch/total number of larvae x 100Treatmentn/aANOVA
Mean fertilization success (%)Number of eggs fertilized/
number of eggs laid x 100
Mean larval survival (%)Number of larvae survived after 5 days/total number of larvae x 100Treatmentn/aANOVA
Mean days to hatchMean number of days to hatchTreatmentn/aANOVA
Water ChemistryIn situ measurementsVariousTreatmentn/aANOVA
Lab analysisVariousTreatmentn/aANOVA
Fish Tissue Body BurdenLab analysisVariousTreatmentn/aANOVA
Fish Tissue - GonadsLab analysisVariousTreatmentn/aANOVA
Fish Tissue - GillsLab analysisVariousTreatmentn/aANOVA
Fish Tissue - LiverLab analysisVariousTreatmentn/aANOVA
Chironomid TissueLab analysisVariousTreatmentn/aANOVA
Algal TissueLab analysisVariousTreatmentn/aANOVA
Sediment AnalysisLab AnalysisVariousTreatmentn/aANOVA

© M. Dubé
Legend: LSI: Liver Somatic Index; GSI: Gonadosomatic Index.

Top of Page

Reproduction is expressed as reproductive effort, fecundity, egg size or gonad weight relative to body size. To date, mesocosm studies have examined changes in gonad size to assess effluent effects on reproductive function. However, measures of fecundity and egg size are easy to measure, if an appropriate sampling time is chosen. Ideally, exposure studies should commence 6-8 weeks prior to the spawning season in order to assess effluent effects on gonad weight.

During the mesocosm study, the physical state of the fish is also assessed. A visual estimation of physical malformations and lesions on the body surface, including eroded, frayed or hemorrhagic fins, parasites, or other physical deformations, is required.

Modular Mesocosm System: The response variables measured in the modular system, which has been used primarily with Fathead Minnow, are summarized along with suggested statistical analysis procedures in Table 9‑3. The variables shaded in grey are those recommended for use. Those variables that are not highlighted are variables that have been measured in different study designs and could be included to provide additional support. It is important to note that use of the standardized Ankley et al. (2001) 21‑day reproductive assay has been modified for improved application under the EEM program and specifically for use in mesocosms as an alternative to the fish survey. The methods are described below and remain to be validated and standardized by the EEM Science Committee. Mines proposing this alternative approach should expect to evaluate the response variables to be measured and should include additional supporting variables to adapt and improve this method for use in their EEM program.

Top of Page Timing and Duration of Effluent Exposure

The timing of the mesocosm study is important and is dependent upon the test species. Water temperatures affect the experimental design. If a study is being conducted in the spring when the water is cold, exposure times may be longer, especially if a response such as growth is being measured. Temperatures below -5°C are prohibitive to mesocosm operation due to freezing of water lines.

Mesocosm studies should strive to balance duration with cost-efficiency. Exposures of 21‑45 days are common for invertebrate community studies in either mesocosm system. Exposures of 21 days are common if the Fathead Minnow partial life cycle is being used as the sentinel species in the modular mesocosm system. If the trailer system is being used for a water-borne exposure, 30 days is common during summer and fall months, and has been used to measure changes in adult organ size (liver size, gonad size) and growth rates of YOY Slimy Sculpin as a result of effluent exposure.

All mesocosm studies should be conducted during normal industrial operations. Effluents should be representative of normal operating conditions. Mesocosm studies should not be conducted when effluent has not been discharged for long periods or during wastewater treatment upsets. Ensure any planned shutdowns are identified well in advance and studies are not affected.

Trailer Mesocosm System: Timing of the mesocosm study is important and dependent upon the spawning cycle of the fish species. See Chapter 3 for additional information on spawning cycles and seasonal sampling. Ideally, for spring spawners, studies will be conducted 6-8 weeks before spawning commences. For early spring spawners where it is impossible to study for this length of time prior to the spawning season, studies should be conducted as late in the year as possible to allow for gonadal senescence and recrudescence. For fall spawners, a spring or summer survey is appropriate. However, this may not apply to fish in which ova mature rapidly. For example, some late-spring-spawning minnows should be studied in early spring, rather than in fall, when ova may still be immature. Water temperatures are also a consideration for timing of studies where juvenile growth rates are higher during the summer months.

Modular Mesocosm System: The modular system, when used with Fathead Minnow, is on-site for a 10-14 day pre-exposure (length depends upon when fish meet appropriate baseline selection criteria) and a 21- to 30-day exposure period. Studies should be conducted from May to late September when water and air temperatures are between 15°C and 28°C.

Top of Page Fish Methods General Set-up – Trailer Mesocosm System

Once the trailer mesocosm system has been transported to the study site and reference water and power have been connected, substrate can be added to the streams and water flow rates can be calibrated. Depending upon the site-specific substrate characteristics, washed and crushed gravel, or sand, can be placed into each tank to a depth of 5 cm. In previous studies, large washed rocks were also added to each tank to serve as refuge for the fish.

Fish are collected from a reference area that is not exposed to the effluent being studied. Obviously, collection requires non-lethal sampling techniques including minnow traps, trap nets or electro-fishing with barrier nets (Portt et al. 2006). Fish are usually sorted in the field to ensure adequate numbers of juveniles and of adults of each sex (if sexually dimorphic), and to collect fish of similar size-classes. If the species chosen is not sexually dimorphic, an assumption is made that sex ratios in the field are equal and the fish should be randomly allocated to the tanks. Fish are transported back to the study site in containers with covers and under adequate aeration. If reference areas are not available, hatchery-reared fish can also be used if these fish are relevant to the study area.

At the study site, and preferably within the same day, fish are measured for length and weight, and randomly assigned to each stream. The precision of these measurements is as described in Chapter 3, with increased precision when using small-bodied species. The optimal numbers of fish per stream are 20 adults of each sex and 20-30 juveniles. To ensure random allocation of fish to streams, 16 aerated buckets with reference water are set up. Fish are measured and placed, one by one, until each bucket contains 1 fish. This procedure is repeated until all buckets have 5 fish each. Buckets are then randomly allocated to the streams and the procedure is repeated until target sample sizes are attained in the streams. It is essential that this procedure be followed to ensure that the largest fish, for example, are not all allocated to the first few streams.

There are 2 possible fish allocation scenarios. Both adults and juveniles can be allocated to the same tanks if juvenile growth rates will be based solely on starting and ending length and weight measurements, and adults are not cannibalistic. If juveniles will be sampled during the course of the study for growth (perhaps juveniles are tagged and repeated measurements are conducted), allocating adults and juveniles to different tanks (or different areas within a tank if these areas can be physically separated) would be recommended to minimize adult capture stress. Mesocosm streams are randomly assigned to each study treatment (control, effluent) and fish are randomly allocated to each stream. This ensures a randomized distribution to minimize stream differences due to factors such as stream position on the trailer.

After inoculation, fish are acclimated to artificial stream conditions for a minimum of 72 hours prior to effluent addition or until fish are feeding. Streams should be covered with netting to minimize fish loss due to escape or predation by birds. During the exposure period, which extends from 30 to 60 days, fish should be fed trout pellets at a rate of 4-6% total body weight per tank per day. The size of the pellets is dependent upon the size of the species used. In previous studies, trout pellets were crushed prior to feeding.

Top of Page General Set-up – Modular Mesocosm System

In the modular mesocosm system, each system consists of a table (one per treatment) holding 5-8 replicate (preferred), 10.3-L, circular, high-density polyethylene streams. The replicate streams sit on the table, which drains into an 85-L dilution reservoir as previously described. If the trophic-transfer system is being used, each stream consists of a sediment (pre-cleaned silica sand) culture of the chironomid C. tentans, a feeding barrier, spawning tile and one breeding group of Fathead Minnow (Figure 9‑6) (Rickwood et al. 2006c, 2008). The feeding barriers are used to control Fathead Minnow access to the C. tentans that have established in the substrate during the pre-exposure portion of the study. Each circular barrier contains a 1/8” (0.32 cm) square mesh screen with a pie-shaped opening of 1/10th of the stream area (approximately 71 cm2) that is turned every second day to dispense the appropriate amount of food (1 g C. tentans/day) (see below on culturing appropriate densities of C. tentans).

Pre-exposure Design: Six-month-old, naive Fathead Minnow are obtained from a reputable culture laboratory. Fish are typically transported to site by ground or air within 24 hours in oxygenated bags with water inside coolers. Upon arrival, containers are aerated and allowed sufficient time for acclimation to ambient temperatures. Reference water is slowly added to acclimate. Once acclimatized, fish should be placed into an appropriate holding tank to become reproductively stimulated prior to the pre-exposure breeding trial. This stimulation period should last for approximately 3 days.

The Fathead Minnow modular mesocosm method requires a pre-exposure and exposure trial. The pre-breeding trial normally consists of double the number of required breeding groups, which are bred in independent replicates in the absence of effluent to establish baseline reproductive performance. At the beginning of the breeding trial, total body weight (g), fork length (mm) and secondary sex characteristics are recorded. Secondary sex characteristics include banding, nuptial tubercles, dorsal pad and fin dot in males and ovipositor size in females (Parrot and Wood 2001). Female Fathead Minnow are size-matched, if possible, to within ± 25% of the male body length (Pollock et al. 2008). Each breeding pair is fed 0.5 g of frozen bloodworms twice daily throughout the pre-exposure period. Each day prior to feeding and recording of water quality, breeding tiles in each stream are checked for egg deposition. If breeding has occurred, eggs will be gently rolled off of the spawning tile into petri dishes with reference water and photographed for digital counting. At the end of the breeding trial, breeding groups will be selected for the exposure phase of the experiment. Breeding groups (pairs or trios, depending upon the study) are selected on the basis that there is 100% survival of all adults, that eggs are present in each replicate at least once in the immediately preceding 7 days, and that > 80% fertilization of eggs has occurred (OECD 2006; US EPA 2007). Breeding groups at both extremes (superior breeders and breeders with very few eggs) should be excluded from the study. The selected groups should be distributed throughout the mesocosm treatments and streams so as to minimize variance between the treatments. Statistical analyses are performed prior to final selection to ensure that there are no significant differences among and between treatments before effluent exposure commences. The unit of replication is n= 5-8 per treatment.

Trophic-transfer system: The trophic-transfer system and associated sediment cultures of the chironomid C. tentans are set up during the pre-exposure period. Target invertebrate densities in each stream are based on an optimal daily feeding amount of 1 g/breeding pair/day (Rickwood et al. 2006a, 2006b). Seven-day-old C. tentans larvae are shipped to the site once per week for 3 weeks to establish the cultures in the mesocosm streams before adding the fish. The larvae are obtained from a reputable culture supplier. C. tentans will have been exposed to the effluents for at least 7 days prior to the introduction of the fish, to ensure dietary exposure. In addition, this ensures that C. tentans will be at various life stages (egg, larvae, pupae, adult) within the streams to maintain a healthy breeding cycle. The number of invertebrates required to sustain the fish over 21 days in each stream is calculated based on the average number of C. tentans that emerge from one egg sac (~ 300) and by determining the number of third and fourth instars with a combined weight of 1 g (~ 50 3rd and 4th instars). Once it is known how many C. tentans weigh 1 g, the total number of C. tentans needed for the entire 21‑day exposure period will be calculated (50 C. tentans/g x 21 days = 1050 C. tentans or 350 7‑day-old larvae/stream/week for 3 weeks).

Once the larvae have arrived at the site, they are acclimated using treatment water in their individual containers, by adding 25% of the treatment water to each container 4 times in 12 hours. The larvae are acclimated to the ambient temperature by slowly lowering the water temperature in a water bath no faster than 1°C every 1-2 hours. Once the C. tentans are acclimated to the ambient temperature and effluent water, they are distributed among the artificial streams and fed Tetramin™ slurry (100 g Tetramin™ flakes to 1000 ml reference water) at a rate of 10 ml in the first week, 20 ml in the second, and 30 ml in the third and subsequent weeks, 3 times/week. It is recommended that 3 sediment cores (core sampler area approximately 9 cm2) be taken from each stream at the end of the pre-exposure period before fish placement so that invertebrate densities can be calculated.

Exposure Design: Once Fathead Minnow are allocated to the treatment streams, treatment solutions are delivered to the mesocosms as previously described, and the exposure period commences.

Top of Page Maintenance and Monitoring

Daily maintenance for the duration of the study includes calibration of flows to each reservoir to ensure target effluent dilutions are attained, cleaning of screens, and feeding of fish. Screens are placed over the water outlet pipes on each stream to prevent fish loss. These screens may require daily cleaning.

Daily monitoring requirements for each stream include recording of any fish mortality and monitoring of physical water variables, including temperature, dissolved oxygen, conductivity and pH. These measurements can be taken using YSI Inc. instruments or continuous recording equipment such as hydrolabs or thermisters. Dissolved oxygen is of critical importance, especially in studies using effluents high in organic content (e.g., pulp mill effluents). Dissolved oxygen levels should be maintained above 60% (minimum), and preferably above 80%, in all streams, using aeration if required.

For the modular mesocosm system, daily observations and measurements can include egg production, hatching success and larval survival (Table 9‑3). Breeding tiles are checked daily at the same time and before feeding or water quality measurements are taken. Eggs are removed from the tiles and photographed. A consistent sub-sample of eggs from each productive brood can be collected for future analysis and total egg production corrected for this removal. Remaining eggs are then rolled into an egg cup and placed in the appropriate aerated culture tubs with treatment water. Twenty-four hours after spawning, the eggs are photographed again to check fertilization success. They are then left undisturbed until all eggs are either hatched or dead (~ 4‑5 days). Once larvae have hatched, they are preserved in 10% formalin for latter enumeration and examination for deformities where the latter is a desired parameter for supporting assessments.

Top of Page Sampling and Analysis of Fish

At the end of the exposure period, fish are anaesthetized and fork length (mm), whole body weight (g) and secondary sexual characteristics are recorded. Fish are then euthanized by spinal severance, and gonads, liver and eviscerated (carcass) weights are recorded. All streams should be sampled on the same day and all fish from each stream should be sampled before progressing to the next stream.

The measurements required and level of precision are the same as those outlined in Chapter 3. Sample preparation, laboratory analysis and QA/QC procedures for mesocosm studies are the same as those required for field fish surveys.

Top of Page Data Assessment and Interpretation General Requirements

Similar to field studies, in mesocosm studies, data assessment and interpretation follow each monitoring or assessment phase. In data assessment and interpretation, the following questions are answered:

  • Is there an effect?
  • Is the effect mine-related?
  • Is the magnitude and extent of the effect known?
  • Is the mine-related cause of the effect known?

An overview of data analysis and interpretation for mesocosm studies is presented here.

Top of Page Statistical Analysis of Parameters

To determine whether there is an effect on fish exposed to effluents in mesocosm streams, statistical analyses of the data are conducted of the biological variables, as suggested in Table 9‑2 for the trailer mesocosm system and Table 9‑3 for the modular mesocosm system with Fathead Minnow. Table 9‑3 lists typical analytical approaches for two types of experimental designs: single-factor ANOVA (e.g., effect of effluent) or two-factor ANOVA (e.g., effect of effluent and contaminated sediment).

Sex differences in growth rate, body weight, condition factor, gonad size and liver size are common due to differences in overall energetic requirements between male and female fish. Therefore, for all parameters, sexes should be treated separately when estimating variability. Immature or juvenile fish should also be treated separately. The analyses that are suggested in tables 9‑2 and 9‑3 are preferred. However, other analyses for specific parameters can be conducted depending upon the variability of the data set. For example, in mesocosm studies, individuals of a similar size-class are selected for placement in the streams, resulting in reduced variability in parameters compared to those measured in field surveys. Thus the range over which regressions such as ANCOVAs are conducted for mesocosm data is narrow enough that ANOVAs on original organ weights (liver size, gonad size) and ratio metrics (LSI, GSI) or condition factors are justified in this instance.

Statistical considerations specific to mesocosm data sets are stated below. Additional reference can be found in the literature cited in Table 9‑1.

Top of Page Nested ANOVA Analyses

For any variable measured once on a whole replicate stream, the statistical design is a simple t‑test, comparing exposed and control treatments (control/impact design). The mesocosms or streams are replicates for the treatments. However, all biological and biochemical variables are measured on replicates at a lower level--either individual fish or composites of several fish within streams. A nested ANOVA can be used to analyze these variables. If the variance among replicate streams within a treatment (Error I) is large relative to the variance among fish or composite samples within streams (Error II), then the nested ANOVA is effectively a t-test comparing treatments, with the stream means as replicate observations. However, if Error Iis not large relative to Error II (e.g., p > 0.25), then the two error terms can be pooled to increase the power of the test comparing treatments (i.e., individual fish or composite samples rather than streams are used as replicates).

Top of Page ANCOVA Analyses

Most parameters are normally estimated using ANCOVA, as discussed in Chapter 8. This is often unnecessary in mesocosm studies, as mentioned above, unless limited fish availability prevented the standardization of size and age.

Top of Page Data QA/QC and Analysis

The importance of ensuring data quality cannot be overstated. There are basic requirements for study design, consistency of methods and measurements, and definition of protocols and procedures; these are outlined in detail in Chapter 8.

9.2 Use of Mesocosms as an Alternative Monitoring Method

9.2.7 Benthic Invertebrate Community Monitoring - Effects Assessment Study Design Species Selection

The benthic community that is established in the mesocosm streams is representative of that found in the reference field areas. Benthic samples are collected from the river and the entire community is inoculated into each mesocosm stream. Culp et al. (1996, 2001), Podemski (1999) and Culp et al. (2000a) have conducted studies that compared the community structure of benthic invertebrates in the mesocosm streams to that of field communities at reference areas. No ecologically significant differences in structure were observed, illustrating the effectiveness of the inoculation procedures and the suitability of mesocosms for testing effluent effects on environmentally relevant communities of benthic invertebrates.

Top of Page Effect Endpoints

All the effect endpoints (total benthic invertebrate density, taxon richness, Simpson’s Evenness Index, similarity index [Bray-Curtis]) used to assess effluent effects on benthic invertebrates in EEM field surveys can be measured in mesocosm studies. Diversity, taxon density, proportion or presence/absence are recommended to allow for the interpretation of effects (Chapter 4). These effect endpoints are summary metrics selected to encompass the range of responses that may result from effluent, including changes in productivity, species composition and biodiversity. Many other benthic invertebrate descriptive metrics are available in the literature (for a review see Resh and McElravy 1993) and may be used, if applicable, on a site-specific basis to aid in the interpretation of effects determined with the effect endpoints listed above.

Top of Page Timing and Duration of Effluent Exposure

Mesocosm studies using benthic invertebrates can be conducted at any time from early spring to late fall. The primary limiting factor is temperature, with air temperatures below -5°C prohibitive to mesocosm operation due to freezing of water lines. Studies should be conducted during periods when field communities are under maximal effluent exposure for improved environmental realism. Studies should also be conducted at the time of year when the benthic invertebrate diversity is highest and water temperatures are conducive to growth. If historical data exist, it would be useful to examine the data and, if appropriate, conduct the study during similar periods so that the studies can be compared. Subsequent monitoring should also be conducted during similar periods of the year to be comparable.

The duration of the mesocosm studies for benthic invertebrate programs is typically 30-45 days, including a 7- to 12–day inoculation period for primary producers and a 25- to 30-day effluent exposure period. Effluents are collected daily or every second day during the studies and should be representative of normal industrial operating conditions.

Top of Page Benthic Invertebrate Methods Mesocosm Trailer – General Set-up

General set-up requirements specified in sections and also apply to benthic invertebrate studies. The recommended total flow of reference water or of dilution water and effluent into each stream is 2 L/minute. This flow rate results in a 2‑hour volume turnover time for each stream (stream volume 227 L). Stream volume turnover rates of 4 hours have been used to reduce effluent requirements. However, longer turnover rates are not recommended due to the effects on stream temperature and dissolved oxygen. For invertebrate applications, current velocity in each stream can be created by a belt-driven propeller system. Current in each stream is established based on site-specific conditions but is normally set between 10 and 20 cm/second.

Prior to placement of substrate into the streams, five sampling bags (0.1 m2, 500‑mm mesh) are installed on the stream bottom. These bags are lifted at the end of the experiment, resulting in 5 sub-samples per stream. A standardized benthic environment is created in each stream to simulate the dominant environmentally relevant habitat type. To date, mesocosm studies using benthic invertebrates have focused on riverine habitats where riffle substrate is dominant. The bottom of each stream and the sampling bags are covered with an 8‑cm layer of washed gravel (stones of 1-2 cm in diameter). The gravel is then left to colonize for a 7- to 12–day period to allow sufficient algal and microbial growth. Only water delivery to the mesocosm streams occurs (2 L/minute) during this colonization period. The duration of the colonization period is site-specific and depends upon temperature and colonization of algae from the river into the system.

Following the algal colonization period, existing benthic communities are transplanted to the stream mesocosms. The transplantation protocol is determined on a site-specific basis. The following is an example of inoculation protocol for a riverine, riffle habitat with a large cobble/gravel-dominated substratum (Podemski 1999; Podemski and Culp 1996; Culp and Cash 1995; Culp et al. 1996) (Table 9‑1):

  1. Large cobbles (surface area ~ 535 cm2) are randomly selected from the river at the reference area for placement into the mesocosm streams with their associated periphyton and invertebrate biota. These cobbles provide additional substrate and also stock the streams with a natural community of periphyton and benthic invertebrates.
  2. During collection, the stones are enclosed with a 0.1-m2 U-net (500-mm mesh) (Scrimgeour et al. 1993), carefully lifted from the stream bed, and placed into a container with river water. In addition, the gravel substratum beneath the stones is gently disturbed (to a depth of 5 cm) to collect any invertebrates under and around the base of the stone. The container is carefully transported to the mesocosm so as not to dislodge the periphyton and invertebrates associated with the stone. The cobbles are randomly placed in the artificial streams, oriented to the water flow in a manner similar to their original orientation in the natural environment. This process continues until the appropriate density (e.g., 10 large cobbles/artificial stream) of large cobbles is reached, based on natural substratum composition. Other samplers may be used to inoculate the streams with benthic communities, as outlined in Chapter 4. The objective, however, is to establish ambient densities in the mesocosm streams. To determine the number of river samples to place in the streams, the mesocosm stream area (0.9 m2) is divided by the area of the selected sampler. This random allocation of sub-samples from pooled invertebrate-community samples ensures that the initial invertebrate composition is similar among streams and limits the amount of variability in community composition among mesocosms that can be attributed to the pattern of species introduction (Wrona et al. 1982).
  3. An additional series of benthic samples (biota only) are collected from the reference area and are pooled into a common container to estimate initial composition and density. Sub-samples are then removed and randomly inoculated into each artificial stream until the density of invertebrates approximates ambient densities.

Standardization of inoculation protocols is important because the sequence through which species from a common-source pool are added to mesocosms can produce large differences in community structure; these dissimilarities are unrelated to the intended study treatments (Drake et al. 1996). The test community includes multiple trophic levels constructed from random samples taken from the study river. Consequently, measurements of community-level variables, such as species composition, community production/respiration, and decomposition, are possible. These community-level variables integrate both the direct effects of stressors and, importantly, indirect ecological effects that cannot be simulated in single-species or simple food-chain systems (e.g., competition-mediated shifts in community structure, or biomagnification) (Carlisle 2000).

Following biotic inoculation, the mesocosm system is left to acclimatize for 24-48 hours (water flow only) prior to commencement of effluent delivery to the treatment streams.

Top of Page Modular Mesocosm – General Set-up

The modular system that was described above for fish is also commonly used with benthic invertebrates. The mesocosm consists of wet tables (one per treatment) upon which partial flow-through streams are placed, and below which a reservoir containing treatment water for circulation to the streams is located. River water is pumped to a head tank, then distributed to each wet table reservoir by pumps (Culp et al. 2004). Water and effluent are pumped through distribution manifolds to the replicate artificial streams. Treated effluents are delivered to the mesocosm system daily or by truck every 2-3 days. The hydraulic residence time of each table reservoir is typically 1 hour for benthic invertebrate studies, while residence time in the circular artificial streams is about 4-5 minutes. This turnover time can vary depending upon effluent and reference water availability. Water velocity in the modular streams can be produced with paddle wheels that generate velocities of 11-12 cm/second; these velocities are typical of the substrate-water interface in rivers. Insect emergence traps are placed over each stream, and the wet tables are covered by a shade canopy that reduces light levels by approximately 60% to better simulate light levels at the river substratum.

The artificial streams are designed to simulate typical riffle communities of reference areas. Benthic food webs are established across all treatments and replicates by inoculating each stream with substratum extracted from a reference area not influenced by effluents. The stream bed substrate is handled carefully so that the associated microbes and algal biota remain intact. Using these techniques, algal growth in all streams is sufficient for invertebrate inoculation in less than 7 days. Similar benthic invertebrate communities are established in all stream mesocosms by inoculating each stream with biota from the reference area upstream. The area sampled establishes initial invertebrate densities of ~ 1.2-1.4 times ambient levels in the mesocosms to adjust for the possibility of initial handling mortality. Invertebrate communities are allowed to acclimate to the experimental conditions for 24 hours before the effluent dose is applied.

Top of Page Maintenance and Monitoring

Trailer Mesocosm System: Maintenance and monitoring is as described for the fish in section

Modular Mesocosm System: Daily maintenance of the systems includes regular calibration of all pumps and delivery systems to ensure target delivery volumes and current velocity are achieved. In addition, drain screens and the algae on inner stream walls are frequently brushed to prevent fouling of the streams.

Weekly grab samples of effluent, reference water and the treatments are collected and analyzed for general chemistry, nutrients and metals. Adult insects are collected from emergence traps each day with an aspirator and preserved in 80% ethanol for later identification to the family level.

Top of Page Sampling and Analysis of Benthic Invertebrates

For benthic invertebrate sampling, streams are also sampled in a random selection to minimize differences due to the time of sampling among replicate streams within the same treatment.

Trailer Mesocosm System: Invertebrate samples are collected by lifting the sampling bags within each stream, which results in 5 sub-samples per stream. Sub-samples are then washed through a 500‑mm mesh sieve. Field sieving is required immediately after sample retrieval and before preservation. The recommendation for sieve and/or mesh size for all freshwater mesocosm applications is 500 mm. In freshwater, macroinvertebrates are defined as those retained by mesh sizes of 200-500 µm (Slack et al. 1973; Weber 1973; Wiederholm 1980; Suess 1982), although immature life stages of some taxa may be smaller and some adult life stages may be larger. Note that these mesh sizes are applicable to all equipment used in the field and laboratory (i.e., both the Nytex mesh on the benthic samplers and sieving apparatus). In some site-specific circumstances it may be desirable for the samples to be screened for smaller organisms by using a smaller sieve size. Some examples of situations where the use of a smaller mesh size (less than 500 µm) may be appropriate include the following:

  1. for comparative purposes if historical benthic surveys for the system under investigation utilized smaller mesh sizes; or
  2. if sampling needs to be conducted, for logistical reasons, at times when organisms are very small; Rees (1984) and Barber and Kevern (1974) provide information on seasonal effects of mesh size.

In these aforementioned cases, it is highly recommended that a stack of screens be used that minimally have the mandatory sieve sizes, and then any other smaller sizes that are appropriate. This procedure simultaneously allows site-specific concerns to be addressed and fulfills the EEM objectives by allowing for national or regional comparisons to be conducted on the standardized mesh sizes.

Modular Mesocosm System: Benthic invertebrates are collected at the end of the experiment by washing the entire contents of each stream through a 250‑mm sieve and preserving the samples in a 10% formalin solution. In the laboratory, benthic invertebrate samples are sorted under 12x magnification, identified to the family level, and enumerated.

All summary statistics and descriptive metrics should be calculated and reported at the family level for submission in the EEM interpretative reports (see section 4.6.2 of Chapter 4). Organisms that cannot be identified to the desired level of taxonomic precision should be reported as a separate category in the fundamental data set. It is recommended that investigators use taxonomic keys appropriate to the geographic region of study; a detailed list of taxonomic references for various groups of freshwater organisms is provided in Chapter 4.

Top of Page Data Assessment and Interpretation

Use of mesocosms as a monitoring alternative for assessing effluent effects on the benthic invertebrate community largely follows the guidance of Chapter 8 regarding data assessment and interpretation in field surveys, as the effect endpoints measured to assess effects are the same. The only difference in data assessment is that replication per treatment in mesocosm studies is n = 8 for control/impact designs rather than the n = 5 as recommended for field surveys.

During the effects assessment, a significant difference between reference and exposure areas in any of the following effect endpoints is to be interpreted as an effect on the benthic invertebrate community: total benthic invertebrate density, taxa richness, evenness Index, and similarity index (MMER Schedule 5, section 1).

Diversity, taxon density, proportion or presence/absence are recommended to aid in the interpretation of effects. Calculation of these metrics is described in detail in the benthic invertebrate community survey (Chapter 4). Details on recommended statistical analyses, data QA/QC, and reporting requirements are as outlined for field surveys, with the qualification that a mesocosm treatment level is equivalent to an exposure area and a replicate artificial stream is equivalent to a field replicate station.

Top of Page

9.2.8 Supporting Measurements Water Chemistry Parameters

In mesocosm studies, it is essential that stream differences in physical water chemistry be minimized as much as possible to ensure effluent-related effects are not confounded. Streams are monitored daily for temperature, dissolved oxygen, pH, and input flow for water and effluent. Current velocity should be measured at the start and end of each study, especially for benthic invertebrate studies. The distribution of water velocities in the streams is characterized using one of many brands of current meter. In previous studies, mean velocity in the mesocosms (above stones) (x = 0.26 ± 0.01 m/second, n = 150) was similar to water velocity measured above stones at a similar water depth in the field (water velocity x = 0.26 ± 0.01 m/second, water depth x = 24.8 ± 0.72 cm, n = 30) while the study was being conducted (Podemski 1999).

Water temperature can be measured using continuous temperature data loggers placed in a stream and the head tank. Temperatures in the head tank reflect the temperature of incoming river water. In previous studies, a comparison of data from these 2 thermographic locations indicated that the 2‑hour hydraulic residence time in the streams resulted in slight heating or cooling of water in the streams depending upon ambient air temperatures (Culp and Podemski 1996). For example, over a 3‑day period, the streams were cooler at night and warmer during the day as compared to the incoming river water. The maximum instantaneous difference between water temperature in the river and the streams was < 5°C.

Top of Page Supporting Water Quality Parameters

In addition to assessment of daily changes in physical water chemistry among all of the mesocosm streams, water samples should also be collected on a less frequent basis for analysis of general chemical parameters. It is recommended that samples be collected from the reference water head tank, the full-strength effluent, and each mesocosm stream upon completion of the study. Samples should be analyzed for parameters as outlined in Chapter 5. Often this information proves invaluable to confirming effluent dilutions in the treatment streams. For example, in freshwater systems exposed to pulp mill effluents, sodium (Na) is a relatively conservative ion that appears in high concentrations. By comparing sodium ion (Na+) concentrations in the reference and exposure streams, the concentration of effluent in the streams can be verified.

9.3 Use of Caged Bivalves as an Alternative Monitoring Method

9.3.1 Introduction

A description of methods for caged bivalve studies is provided in this section, which includes detailed guidance on:

  • background and general approach of caged bivalve studies;
  • species selection;
  • study design;
  • variables to be measured;
  • methods for implementing the study;
  • data analyses;
  • tissue concentrations; and
  • reporting requirements.

Top of Page

9.3.2 Background

In October 2000, the National EEM Science Committee recommended the use of caged bivalve studies as a scientifically defensible alternative approach to a wild fish survey if it is not practical or technically feasible to conduct the fish survey. The Metal Mining Fish Subgroup also recommended the use of caged bivalves as an alternative method during the metal mining EEM multi-stakeholder consultation process. For more information see Courtenay et al. (1998), Andrews and Parker (1999), and Applied Biomonitoring (2000).

In November 2000, the American Society for Testing and Materials (ASTM) approved a method for conducting environmental studies using caged bivalves (Salazar and Salazar 2000). The ASTM method serves as the basis for this technical guidance on conducting caged bivalve studies. A number of other studies using caged and wild bivalves were also considered in developing this specific guidance for applying this approach within the framework of the EEM program. The format of this guidance follows that used for the fish survey. Bivalves, such as oysters and mussels, have been used in Mussel Watch programs since the mid-1970s to monitor trends in chemical contamination and assess the effects of human activities on coastal and estuarine areas. Mussel Watch programs began in the United States (Goldberg et al. 1978) and have since become international in scope (Jernelov 1996). The following are some of the reasons why bivalves are suitable test species:

  • bivalves are relatively non-mobile, such that exposure to contaminants is assured and is representative of the exposure area;
  • bivalves are abundant in many marine, estuarine and freshwater environments, and are relatively easy to handle and sample year-round;
  • the biology of many shellfish species is well known and considerable research has been conducted regarding effects on shellfish of exposure to various anthropogenic and natural environmental stressors;
  • several bivalve species have been shown to readily accumulate many chemicals from a variety of pathways (water, sediment, food) and show sublethal effects associated with exposure;
  • bivalve growth is relatively easy to measure and has been shown to be as sensitive or more sensitive than mortality in other standard test species such as Daphnia, Fathead Minnow and Rainbow Trout (see Salazar and Salazar 2000); and
  • bivalves are an important fisheries resource, with both the Atlantic and Pacific regions having commercially valuable shellfish aquaculture industries as well as commercial and recreational shellfish harvests.

Top of Page

9.3.3 Approach

The caged bivalve approach provides a number of advantages to the investigator in conducting a monitoring program (Crane et al. 2007). These include experimental control and realism, use of organisms naturally found in the study environment, and known exposure period. By using caged bivalves rather than resident populations, the variability in biological measurements can be reduced by using individuals of similar size and environmental history, thereby increasing the discriminating power of the test (Crane et al. 2007; Salazar and Salazar 1995). A considerable number of caged bivalve studies have been conducted in Canada and the United States, as well as other countries (see Salazar in Stewart and Malley 1997; St-Jean et al. 2003, 2005; Crane et al. 2007).

The effect indicators for caged bivalve studies in EEM are survival, growth, condition, reproduction and energy storage. A tissue analysis (mercury) may be required, and caged bivalves can be used to meet this requirement. Other chemicals or metals may be used to assess bioaccumulation to aid in interpreting results or for use during investigation of cause.

One of the difficulties associated with caged bivalve exposures in the EEM program is related to the difficulties in comparing between responses obtained through the adult fish survey and the caged bivalve exposures. The difficulty lies with the following assumptions:

  1. Mussels used in caged bivalve studies generally originate from clean areas or reference sites, while fish in the adult fish survey are generally long-term residents; therefore, their responses cannot be expected to be the same.
  2. At most sites in Canada, the reproductive cycle of mussels (Blue Mussels [Mytilus edulis]) spans a minimum of 9 months, whereas gametes produced in the spring are derived from energy (mostly glycogen) accumulated in the fall; therefore, a 60- or 90‑day exposure in the spring will have difficulties capturing effluent effects on reproductive effort.
  3. In the mussels, the same organ is used for energy storage and reproduction. In the fall, the mantle (Figure 9‑10) is mostly composed of energy (glycogen), which will be used to develop the eggs in the spring. Figure 9‑11 represents the cycle between energy and eggs in a population of Blue Mussels from British Columbia.

Top of Page

Figure 9-10: Mussel showing ripe mantle lobe

Figure 9-10: Mussel showing ripe mantle lobe.

Top of Page

Figure 9-11: Reproductive cycle of Blue Mussels from British Columbia: A) Mantle energy  stored in fall; B) Mantle reproductive content in spring

Figure 9-11: Reproductive cycle of Blue Mussels from British Columbia: A) Mantle energy stored in fall; B) Mantle reproductive content in spring.

Notes: Numbers on the axis represent months: from February (1) to November (10); A) Glycogen is expressed as mg/g; B) Reproduction is expressed as volumetric fraction of gametes VFG.

Therefore, in order to maximize caged bivalve results and comparability to the adult fish survey, and also facilitate the interpretation of results in terms of reproduction and energy, exposure of adults should occur from the onset of energy accumulation in the fall until the release of gametes in the spring.

Top of Page

9.3.4 Species Selection

Many bivalve species have been used for assessing chemical bioavailability or effects in marine, estuarine and freshwater environments. Ideally, species that have wide geographic distributions should be used so that test results can be compared across studies. Species selection for caged bivalve studies should be made carefully and should consider the biology of the species and local conditions, such as:

  • Are conditions at the exposure and reference areas similar to the natural habitat of the species in terms of tolerance limits for natural factors such as temperature, salinity, dissolved oxygen and pH? Is the species naturally present in the area under evaluation?
  • Is there documentation indicating that the species can accumulate and/or be sensitive to the contaminants of concern?
  • Is the life history of the species well known in terms of spawning cycle and life-stage requirements?
  • Does the species have threatened or endangered status?
  • Is an abundant supply of the species readily available?
  • Is the species easy to handle in the field?

For individual mussels, care should be taken in regards to the following:

  • Are specimens’ shells abnormally thick?
  • Does the shell have signs of worms? (holes in the shell are often a telltale sign)
  • Is the shell cracked?
  • Does the mussel have a slow-closing valve reflex?

Top of Page Most Commonly Used Taxa

The species most commonly used in field bioassays in Canadian waters and considered relevant for use in the EEM programs are described below and listed in Table 9‑4. The temperature and salinity tolerance limits for each species are provided, along with information on age at maturity, spawning periods and general distribution within Canada. Other species, such as marine clams (Mya arenaria, Macoma balthica) and scallops, may also be suitable for some applications. However, the environmental requirements and sensitivity of alternative bivalve test species should be established before they are used in an EEM program. For species-specific needs, Salazar and Salazar (2000) described the salinity, temperature and general distribution of several species of bivalve in Canada.

Table 9-4: Suggested taxa for use in caged bivalve studies for EEMs (text description)
Species and ReferenceTemperature Range (°C)Salinity Range (parts per thousand)Reproductive InformationGeneral Distribution in Canada
Marine and Estuarine Bivalves
Mytilus edulis
(Blue Mussel)
(Freeman et al. 1994; Grout and Levings 2000; Mucklow 1996; Stewart 1994; Salazar and Salazar 2000; Newell 1989; Toro et al. 2002)
-1.5 to 255 to 33Most energy utilized for spawning at length greater than 3.5 cm, or roughly 2.5-4 years old. Generally an abrupt spawning on the East Coast: no more than 3 weeks, between mid June and mid July. But spawning may vary among populations; some low-level spawning throughout year, mostly in area of anthropogenic influence or repeat spawners; first in early summer, second in the fall, mostly on the West Coast.Atlantic coast
Mytilus trossulus
(Bay Mussel, or Foolish Mussel)
(Freeman et al. 1994; Salazar and Salazar 2000; Skidmore and Chew 1985; Toro et al. 2002)
0 to 294 to 33Most energy utilized for spawning at lengths greater than 3.5 cm. Spawning generally spans 12-13 weeks from June to September.Atlantic and Pacific coasts
Crassostrea gigas
(Pacific Oyster)
(Waldock et al.1996)
4 to 2425 to 35Spawning July to AugustPacific coast
Freshwater Bivalves
Elliptio complanata
(Eastern Elliptio)
(Beckvar et al. 2000; Day et al. 1990; Hinch et al. 1989; McMahon 1991; Metcalfe-Smith et al. 1996)
0 to 300 to 3Age at maturity 6-12 years. Spawning occurs mostly June to July; some May to September.Eastern Canada
Pyganodon (Anodonta) grandis
(Common Floater Clam)
(Clarke 1973; Couillard et al. 1995a, 1995b; Malley et al. 1996)
0 to 300 to 3Spawns mostly April to May, some to AugustInterior and eastern Canada
Anodonta kennerlyi
(Western Floater Clam)
(Clarke 1981; Stewart and Malley 1997; Williams et al. 1993)
0 to 30FreshwaterSpawning begins in early August.Alberta and British Columbia
Sphaeriid clams
(e.g., Musculium securis, Sphaerium rhomboideum, Sphaerium striatinum)
(Hornbach et al. 1982; Mackie 1978a, 1978b; Mackie and Flippance 1983; Mackie et al. 1974; Stephenson and Mackie 1981.)
10 to 25*
(*optimal growth range)
FreshwaterLife cycle generally 1 year; life histories of many species are well documented; reproductive effort can be quantified.Widely distributed in Canada

Format modified from Salazar and Salazar 2000

Top of Page Marine and Estuarine Bivalves

The Pacific Oyster is a species that has been used in transplant studies in marine and estuarine studies (Waldock et al. 1996). Its shells are usually more difficult to measure because of their irregular shape and protrusions.

Mytilus have been used extensively in the ongoing International Mussel Watch Project to monitor trends of chemical contamination and assess the effects of human activities on coastal and estuarine areas in North America and around the world (O’Connor 1992; Jernelov 1996). Blue Mussels (Mytilus edulis) and Bay Mussels (Mytilus trossulus) are found on the Atlantic Canadian coast, whereas mostly M. trossulus is found on the Pacific Canadian coast. These two species may be easily confused where they co-occur on the Atlantic coast (Freeman et al. 1994; Mucklow 1996), and since their biology and reproductive cycles differ, species identification is essential. Mytilus spp. are often referred to as the M. edulis complex, in recognition of biochemical differences (Varvio et al. 1988), and may comprise the species M. edulis, M. galloprovincialis and M. trossulus. Several fundamental differences have been observed between M. edulis and M. trossulus, including gamete incompatibility (Rawson et al. 2003), temporal separation and duration of spawning in Atlantic mussels (Toro et al. 2002), and total egg production and size (Toro et al. 2002). Bay Mussels have smaller eggs, with longer spawning times and less gamete production than Atlantic mussels. The growth rate for M. edulis is faster than M. trossulus on the east coast of Canada (Penney et al. 2002). Work carried out on caged mussels on the Pacific and Atlantic coasts has confirmed differences in growth and reproductive cycles between the species, showing M. trossulus to be more sensitive, smaller, and producing fewer and smaller eggs (Metro Vancouver, unpublished data). Table 9‑5 lists several differences noted between the two species over five years of analysis.

Table 9-5: Differences noted between two species of mussels over 5‑year study in the Burrard Inlet, Vancouver, British Columbia (text description)
Mytilus edulisMytilus trossulus
Faster growthSlower growth
Higher survivalLower survival
More eggs / larger eggsSmaller eggs / fewer eggs
Reproduction over 2-3 weeksReproduction over 9-14 weeks
Egg production more clearly separated from energy storageEgg production less separated from energy storage
Less susceptible to leukemiaMore susceptible to leukemia
Overall better suited for monitoringOverall not bestsuited for monitoring
Not always present (West Coast)Sometimes co-occur (East Coast)

Top of Page Freshwater Bivalves

Freshwater unionid, or freshwater mussels, have been used in a number of caged studies to examine water column and sediment exposures. Unionid bivalves (such as Elliptio complanata, Anodonta kennerlyi and Pyganodon grandis [formerly Anodonta grandis]) or sphaeriid clams might be considered suitable for assessing differences in the survival, growth, condition and reproductive rates of bivalves in freshwater receiving environments. The following authors have discussions on the life cycles of these species: Mackie (1978b), Sandusky et al. (1979), Stephenson and Mackie (1981), and Stewart and Malley (1997). Freshwater mussels are bivalves belonging to the super-family Unionoidea and comprise one of the most endangered groups of organisms in North America (Wolfe et al. 2009). The unionids are notable in that their glochidia require incubation in a vertebrate host for survival to adulthood. Glochidia are the parasitic larval stage of unionid mussels that attach to the fish host after release from the adult mussel. Glochidia remain on the host fish until metamorphosis is completed, the duration of which is dependent on water temperature. The glochidia of different genera are released at different times of the year (Bauer 1994). The various taxa of the genus Pisidium may be too small for practical handling and, in addition, are taxonomically difficult for the non-specialist. Additional research may be required to demonstrate the utility of sphaeriid bivalves for use in EEM. Freshwater bivalves do not have fused gonads as marine mussels do, but have distinct gonads. Freshwater bivalves also display distinct seasonal cycles in tissue biochemical content, related mostly to the reproductive cycle. As with the marine and estuarine bivalves, proteins, glycogen and lipids content are maximal during gonad development and gametogenesis, and minimal during glochidial release. As with their marine counterparts, glycogen presents the most variation (Jadhav and Lomte 1982). Table 9‑6 lists the differences between Unionoidea and Sphaeriidae.

Table 9-6: Differences noted between Unionoidea and Sphaeriidae (text description)
Fast growth until maturity; then slowSlower growth; bivalve can be very small
Life span: < 6 to > 100 yearsLife span: 1-4 years
Fecundity: 200 000 to 17 000 000 eggs per female; small eggs3-24 eggs per female; large eggs
Reproduction: one per yearReproduction: 3 per year, sometimes continuous
Egg production more clearly separate from energy storageEgg production less separate from energy storage
Age at maturity: 6-12 yearsAge at maturity: 0.2-1 year
Less suited for studies: gonochoristic, long-lived, interoparous, often rare, difficult to collect, complicated life cycles (parasitic stage)Better suited for studies: greater abundance, ease of collection, ease of maintenance, relatively simple life cycle, short life span

Top of Page Source

An important consideration will be whether or not to use farmed mussels or native local mussels. Depending on the parameter being measured, there are advantages and disadvantages to both; the choice should be made in consideration of the circumstances outlined below.

Bivalves should be obtained from a commercial grower when the parameter being measured is growth or chemical accumulation. Native mussels should be used for reproductive and energy parameters, preferably using a natural gradient or control impact design if the exposure period is short, or commercial mussels when the exposure period is longer. Growth should be measured using juveniles obtained from an aquaculture facility. Bivalves obtained from a commercial facility have an environmental history that is well known, and assurances of being uncontaminated are greater than for animals collected in the wild. In any study, some form of species identification should be confirmed, as results may vary significantly between species, as outlined in Table 9‑6. All individuals used in a caged bivalve study should be from the same population for the same parameter. If wild populations are the only possible source, they must be collected from an uncontaminated area. Epiphytic growth on bivalve shells should be removed gently by hand or with a soft brush or scraper. Collection permits for field-collected or transplanted bivalves are required by Fisheries and Oceans Canada and may be required by some local or provincial agencies. In addition, if cages present a potential obstruction to navigation, a permit may be required from the Canadian Coast Guard and a Notice to Mariners may be required. The permitting process should be considered early in the study planning process. It may take several weeks or months to process the necessary permits, as there may be a need for an environmental assessment as required by the Canadian Environmental Assessment Agency.

Top of Page Species Identification

Mussels can be identified by two methods: allozyme analysis or by morphometric measurement combined with statistical analysis (McDonald et al. 1991; Mallet and Carver 1995). For the morphometric measurements, empty mussel shells are scraped to remove any remaining tissues and dried for 4‑5 hours (60°C). A minimum of five shell characteristics (listed below) should be measured using a stereo microscope (6.4x magnification):

  • the length of the anterior adductor muscle scar;
  • the length of the hinge plate;
  • the distance between the anterior edge of the posterior adductor muscle scar and shell margin;
  • the distance between the ventral edge of the posterior adductor muscle scar and ventral shell margin; and
  • the distance between the pallial line and ventral shell margin midway along the shell (Figure 9‑12).

Three additional shell characteristics should be measured with callipers:

  • shell length;
  • shell width; and
  • shell height.

Each characteristic should be standardized using log10 and divided by the log10 shell length. These morphometric variables (log-transformed and length-standardized as appropriate) should then be multiplied by their raw canonical coefficients and summed to generate a canonical variate for each individual (Mallet and Carver 1995). Mytilus edulis typically has a longer anterior adductor mussel scar, a longer hinge plate, and a greater shell height than Mytilus trossulus,resulting in positive values of the standardized canonical coefficients.

Figure 9-12: Mytilus spp. shell scars markings

Figure 9-12: Mytilus spp. shell scars markings.

Note: 1) length of anterior adductor muscle scar; 2) length of hinge plate; 3) length of posterior adductor muscle scar; 4) distance between posterior edge of posterior adductor muscle scar and posterior shell margin; 5) distance between ventral edge of posterior adductor muscle scar and ventral shell margin; 6) shell width; 7) shell height.

Top of Page Size and Age

All bivalves used in a caged study should belong to the same age class and be as uniform as possible in size. Typically, juvenile mussels intended for growth measurement should not have more than 5 mm difference in length at the onset of the exposure. This minimizes the number of individuals required to achieve adequate power. Juveniles are the best candidates for this parameter, as most of their energy is directed toward growth. Bivalves, and mussels in particular, have an inverse relationship between energy directed toward growth and energy directed toward reproduction. Juvenile mussels will expend most of their energy on growth with little input toward reproduction, and the ratio between growth and reproduction will slowly shift as most energy is directed to reproduction and little to growth in adults. Typically, mussels can start producing some gametes as young as 1 year old, and by 3 or 4 years old most of their energy will be directed toward reproduction and they will grow at a much slower rate. For energy and reproductive output measurement, mussels older than 3 years of age are recommended (generally at least 4 cm in length), while juveniles (between 2 and 2.5 cm) are recommended for growth.

Mussels can be aged following a combination of the techniques described in Ramon and Richardson (1992) and Sejr et al. (2002). The technique is based on annual growth bands and has been validated using a mark-and-recapture approach (Sejr et al. 2002). An experienced biologist should perform this task.

Top of Page Number of Organisms

A power analysis should be used to determine the minimum number of bivalves needed to detect a specified “effect size.” It is recommended that studies be designed to detect a 20% difference in growth. Data sets to aid in predicting the number of required organisms may be available through Environment Canada. However, when the range in mussel length is an average of 5 mm, 100 mussels will be sufficient to achieve power and to ensure an adequate number of mussels survive the exposure.

The number of animals required per cage for the growth measurements will depend on the study design (e.g., number of cages per station; see more discussion on this in the following sections), species used, age of animals, variability in response to the station, and growth conditions at the station. The number of bivalves required to fill the cages will depend on study design in terms of:

  • number of areas (e.g., exposure area plus number of reference areas);
  • number of stations per area;
  • number of cages per mooring (if the study is designed to address varying depths in the water column); and
  • recommended bivalves per cage.

In addition, if tissue samples are required for tissue analysis, consideration must be given to the number of animals required to obtain a sufficient sample size for chemical analysis.

Top of Page Handling and Holding Conditions

Salazar and Salazar (2000) provide detailed guidance on handling and holding conditions for bivalves, and this is summarized below. Test organisms should be handled as little as possible and should be deployed as soon as possible after collection. When handling is necessary, it should be done carefully, gently and quickly so that the bivalves are not needlessly stressed. Bivalves should be kept in well-aerated, clean-flowing water as long as possible between collection, sorting and deployment. If transporting bivalves for extended periods, keep them moist and cool by placing them in a cooler with frozen gel packs or wet ice (wet ice at the bottom of the cooler). Use seaweed or cloth towels to keep bivalves separated from the gel pack or wet ice. Newspaper should be avoided, as it contains ink that can be toxic to mussels.

Top of Page

9.3.5 Study Design

Determining the appropriate study design is critical if the results are going to be meaningful. The study design for caged bivalves should include a number of components:

  • sampling design;
  • area and station selection;
  • replication of cage stations and cages per station;
  • timing and duration;
  • modifying or confounding factors;
  • supplementary measures;
  • cage design; and
  • mooring systems.

Top of Page Sampling Design

There are six main sampling designs that are recommended for caged mussel marine and freshwater assessments:

  • control-impact (C-I) design;
  • multiple control-impact (MC-I) design;
  • simple gradient (SG) design;
  • radial gradient (RG) design;
  • multiple gradient (MG) design; and
  • control–simple gradient (C-SG) design.

The C-I and MC-I designs are used to determine the magnitude of difference between homogeneous exposed and unexposed areas, while the SG, RG and MG designs examine changes in an effect along an effluent gradient. The C-I and MC-I designs are used when there are few qualitative levels of exposure. It is suggested that multiple reference areas be used rather than increasing sample sizes in one reference station. The SG, RG and MG designs may be used when there are many quantitative levels of exposure (Paine 2000); they may also be useful in discriminating among effects from sources other than metal mine effluent. Guidance on selecting the appropriate design is provided in Chapter 2.

For example, in estuaries with complex tidal regimes and mixing regimes, an MC-I approach may be the most appropriate. An SG approach may be applicable to a river receiving environment where flow is unidirectional.

C-I, C-SG and MC-I designs require some level of replication within the control and impact areas. C-SG is a combination between C-I and SG, which is sometimes useful when more than one reference site is desired. In this case, a control, or reference station, is added to a simple gradient, typically when the very low effluent exposure (far-far field) area does not have conditions similar to the exposed area in terms of depth or other important biotic factors. Replication at each station may not be needed for gradient designs, although an appropriate number of stations are needed in order to discriminate between spatial patterns related to effluent discharge and other spatial patterns in the environment. Replication is discussed below in section Ultimately, it is the responsibility of study designers to develop site-specific study designs that are scientifically defensible, robust and suitably sensitive.

Top of Page Area and Station Selection

Chapter 2 provides guidance for the selection of multiple reference locations for a variety of receiving environments and is applicable to caged mussel studies. Reference areas should be as similar as possible to study areas in terms of the water’s:

  • depth;
  • hydrodynamic conditions;
  • temperature;
  • salinity;
  • dissolved oxygen concentration; and
  • food availability and quality.

Multiple reference stations may help to identify natural differences and variability among uncontaminated areas. Often, a mixture of simple gradient and multiple control areas, such as C‑SG, allows for a more robust study.

The exposure area is defined by plume characteristics. Plume delineation, as described in Chapter 2, should provide sufficient information to model average effluent concentrations with distance from source and to identify the degree of vertical mixing in the water column. This information will assist in the selection of stations and depth of deployment in the water column. More guidance is provided in the sections below.

Top of Page Replication Number of Cage Stations

For the exposure area, stations and depths of cage deployment may be chosen to represent a gradient of exposure to the effluent plume. The deployment stations of bivalve cages should be given careful consideration for a number of potentially interfering factors, such as the following:

  • In estuarine and marine situations, effluent may be positively buoyant, resulting in a thin layer of “freshwater” effluent floating on denser saline water.
  • In tidal situations, the mixing behaviour of effluent may be quite complex, resulting in low confidence about the average exposure concentration to which bivalves may be exposed. Tidal situations may require consideration of exposure area stations that are both “upstream” and “downstream” of the outfall.
  • In river discharge situations, effluent plumes may be quite long and narrow, meaning that there is little or no opportunity to meaningfully replicate cage stations within (i.e., across the long axis of) the plume.

For these and other reasons, it might be more meaningful to simply evaluate “distance from outfall” rather than effluent concentrations when designing caged bivalve studies. This approach would remain consistent with one of the objectives of EEM, which is to evaluate the magnitude and geographic extent of effects that may be related to the effluent discharge.

The number of replicate stations must be determined to address the sampling design and EEM objectives. The number of replicate stations and sub-samples within replicate stations are determined by power analysis. The allocation and distribution of replicate stations is dependent upon the sampling design. Guidance on the use of power analysis is provided in Chapter 8. Readers are encouraged to ensure that no pseudo-replication occurs, but rather true replication. Hurlbert (1984) offers more guidance on the subject.

Top of Page Applying Replication to Study Design

Several parameters are suggested for caged bivalve studies (survival, growth [change in length or wet weight], soft tissue fresh weight, condition, reproduction and energy storage). In order to assess these parameters on a set of bivalves exposed at different locations in the field, many different configurations of cages and bivalves are possible. For EEM studies, it may be advantageous to deploy replicate cages containing multiple bivalves at each station, and to consider only the average performance within each cage. This approach confers the additional advantage of providing redundancy in case one or more cages are lost, and may simplify the construction and deployment of cages. So long as the statistical hypothesis testing is confined to evaluating whether or not there are significant differences between areas (without uniquely attributing effects to effluent exposure), the replication and statistical analysis is valid.

As general guidance, it is suggested that cages containing 20 animals per cage be deployed for the survival measurements and at least 5 cages be deployed at each station to evaluate growth. These cages should be deployed on individual moorings, not 5 cages on one mooring, since the mooring is the most appropriate unit of statistical replication. However, practitioners should be encouraged to explore the power and robustness of potential study designs, using synthetic data (or empirical data where available) as an integral part of the study design process to determine the minimum number of animals needed.

Top of Page Timing and Duration

Timing of the studies should be such that:

  • it coincides with a high growth period in natural populations so that growth is maximized and differences in growth rate among treatments are more measurable; and
  • it does not coincide with a spawning period if the test bivalves are adults.

For growth, survival and chemical accumulation, the duration of exposure should be 60-90 days (see discussion in Salazar and Salazar 2000). This should provide sufficient time for effects on survival and growth to be manifested. However, a minimum of 9 months may be needed to measure energy or reproduction; typically, on the East Coast, cages should be deployed in the summer to ensure that the energy accumulated in the fall (energy reading) is from those sites and that egg production reflects any potential effects of the effluent. Although 3 samplings are required (deployment, energy and reproduction), the cost is not generally prohibitive, as, unlike growth and survival, no measurements are needed prior to deployment.

Top of Page Modifying or Confounding Factors and Supplementary Measures

Results of caged bivalve studies will depend, at least in part, on natural factors such as temperature, food supply, other physicochemical properties of the test environments, species selected, condition of test organisms, exposure method, and handling of test organisms. Exposure and reference areas should be as similar as possible with respect to the factors listed below, to minimize confounding differences. It may be useful to measure some of the factors in order to assist in interpreting results. These factors may include life cycle, behaviour, temperature, lack of acclimation, current speed, salinity, fouling, chemical concentration and food availability. These are discussed further in Salazar and Salazar (2000).

9.3 Use of Caged Bivalves as an Alternative Monitoring Method

9.3.6 Cage Designs

For growth and survival studies, cages with individual compartments are suggested so that individual bivalves can be tracked through the study. The mesh size should be maximized to allow maximum water flow but small enough to contain the test animals. Individuals are assigned to compartments such that survival, growth and condition can be tracked in each bivalve. Individual test organisms are placed in the mesh bags and separated by using a plastic cable tie or other suitable tie. Sufficient space should be allowed in each compartment to permit test animals to grow during the exposure period.

For reproduction and energy measurements, mussels do not need to be pre-measured; therefore, compartments are not necessary. However, they should still be exposed to relatively uniform conditions, and stringing mussels into socks, clumped in 3-4 individuals, is recommended (Figure 9‑13). This significantly reduces the level of effort needed for this design.

A variety of cage designs are described in Salazar and Salazar (2000). A flat (i.e., two-dimensional) cage design, as shown in Figure 9‑13, is suggested, as it is a convenient unit to work with. Polyvinyl chloride (PVC) tubing is a convenient material to use for constructing cages. PVC should be water-supply grade obtained from a high-quality source and soaked for at least 24 hours in flowing fresh or seawater before use to remove water-soluble and volatile chemicals. Alternative materials are described in Salazar and Salazar (2000), section 9.

Final cage dimensions depend on the size of the test organism and the number of organisms per cage. Typical bags sizes for species like mussels and clams are 10-15 cm in diameter with 5‑mm mesh size. Each mesh bag should be long enough to accommodate the desired number of bivalves per bag, plus sufficient material for attachment to the PVC frame. For freshwater clam deployments, a wide variety of cage designs have been used by investigators. It is the responsibility of the study designer to ensure that cage designs are appropriate to the test species and receiving environment.

Top of Page Mooring Systems

The cages can be suspended in the water column by attaching them to mooring lines that have an anchor or weight on one end (e.g., iron chain links) and a surface or subsurface buoy attached to the other end (Figures 9-14, 9-15). Salazar and Salazar (2000) discuss the factors that should be considered for deployment of cages.

Figure 9-13: Duplicate frame from a caged mussels exposure experiment

Figure 9-13: Duplicate frame from a caged mussels exposure experiment.

Top of Page

Figure 9-14: Modular mesocosm parts diagram

Figure 9-14: Modular mesocosm parts diagram.

Top of Page

Figure 9-15: Modular mesocosm flow schematic

Figure 9-15: Modular mesocosm flow schematic.

Top of Page

9.3.7 Methods for Test Initiation, Cage Deployment and Retrieval and Test Termination Test Initiation

The first step is to sort all bivalves into the desired size range(s). By selecting bivalves within a narrow size/age range, the investigator can be relatively confident that the individuals will have similar growth potential. Determining the age of wild bivalves may be difficult if not impossible for some species; length is therefore used to obtain individuals with similar growth rates. Commercial growers can often provide bivalves of a known age. The size range selected for test organisms depends on the species, organism supply, and target age. Test organisms should be selected within a narrow size range, and should be kept cool and moist during the sorting stage to prevent stress, damage or death (see section ). For growth and survival studies, shell lengths, widths, heights and whole-animal wet weight (WAWW) are recorded for each animal and the animals assigned to individual holding containers (e.g., ice cube tray) that are labelled according to their specific location within the cage. This allows measurements of each individual test organism to be repeated at test termination. Animals in a sacrificed sub-sample (statistically appropriate number) of test organisms will be measured for shell length, WAWW, tissue weight and shell weight. For other parameters, mussels of similar size do not need to be measured; they can be placed in the socks, preferably in clumps of 3-4 individuals, as this mimics their natural clumping behaviour and minimizes stress.

For quality control purposes, approximately 20% of the measurements should be repeated and recorded by a different investigator. It is best to use an electronic spreadsheet for recording all measurements in order to reduce transcription errors. It is possible to electronically link measurement tools (e.g., vernier calipers and analytical balance) with a spreadsheet to provide an additional quality control measure.

All test organisms needed to fill the cages, plus a sub-sample for destructive sampling (tissue weight), should be sorted and measured prior to distribution among socks and cages. Test organisms are then distributed among individual mussel socks before they are attached to cages. An example of a labelling and distribution scheme is provided in Salazar and Salazar (2000), section 11.

Top of Page Cage Deployment

Ensure that an appropriate vessel is chartered to assure safe transport and deployment of all cages, moorings and floats. Moorings and floats may be attached to the cages on the vessel while en route to the site. When on-site, cage location should be identified using a global positioning system (GPS) or other reliable method (e.g., nearby onshore reference points). During the exposure period, the cages may be inspected (e.g., by divers) to check for presence, damage and fouling.

Top of Page Cage Retrieval and Test Termination

Cages should be retrieved using various location identifiers (e.g., GPS, depth sounders) and a grappling hook if a retrieval rope was used for each cage.

Details on test termination are provided in Salazar and Salazar (2000), section 11, and outlined here. At the processing site, bivalves from all bags for each cage should be processed together. For growth and survival measurements, it is essential that the order and orientation of each bivalve be maintained during all of the end-of-test measurements. Individuals can be removed from mesh socks and placed in labelled individual containers (e.g., ice cube trays) to facilitate measurement.

Five to 10 minutes prior to making length and weight measurements, set the tray(s) into a tub containing clean water. Individuals that float indicate that air is trapped between the valves of the shell. When all animals in the tray(s) have opened their valves slightly and are no longer floating (5‑10 minutes), length and weight measurements may begin. Start with WAWW and shell length measurements, then carefully shuck each individual and blot the tissue before taking tissue fresh-weight measurements. Note that if tissues are to be analyzed for chemical parameters, care must be taken to not introduce any contaminants from the blotting material. The Mussel Watch protocol (Gulf of Maine Council, 1997) is best suited to dissection for the purpose of chemical analysis.

Top of Page

9.3.8 Effect Indicators

The effect indicators to be measured in caged bivalve studies for EEM studies are survival, growth, condition and reproduction, as described in more detail below. Table 9‑7 shows the effect and supporting endpoints used for a caged bivalve study. The statistical procedures are also listed and described in more detail in section 9.3.10. An example of a reporting format for recording survival and growth raw data and endpoints is provided in Table 9‑8.

Table 9-7: Caged bivalve study effect indicators and endpoints and related statistical procedures (text description)
Effect IndicatorsEffect and Supporting EndpointsStatistical Procedure
  • Change in size (weight and length) over time (end measurements compared to starting measurements)
  • WAWW
  • Shell length and width
  • Soft tissue fresh weight
ANOVA (regression analysis for gradient designs)
  • *Mantle somatic index (MSI)† (similar to the GSI) (gonad weight against body weight)
  • *Weight (whole-animal dry weight, dry shell or soft tissue weight) related to shell length
  • Soft tissue weight related to shell weight
  • Soft tissue weight related to shell volume
  • *Percentage of individual animals alive per cage at the end of the exposure period
  • Length frequency analysis

* Caged bivalve effect endpoints used for determining effects. Other supporting endpoints can be used to support analyses.
† Currently, guidance is only provided on the relative mantle somatic index (similar to the gonad index [GSI]). See section

Top of Page Survival

Survival is not a particularly sensitive indicator of effects in caged bivalves, but it is an important parameter to monitor. Survival can be easily determined and quantified, although it is possible to have some individuals missing at the test end due to shell decomposition. Bivalves are dead if they are gaping open and do not close their shells when touched or tapped. Survival is expressed as percent of individual animals alive per cage at the end of the exposure period.

Top of Page Growth

Growth is a measure of energy use and is a sensitive indicator of effects that is easy to measure. Several types of growth measurements should be made; measuring only one could provide misleading results. Growth measurements, with their expected accuracy, are measured at test initiation and termination as outlined below:

± 0.001 g
Shell length
± 0.01 mm
Shell width
± 0.01 mm
Shell height
± 0.01 mm

Growth can be expressed in a number of ways:

  • absolute growth = absolute change in value from test initiation to test termination;
  • growth rate = absolute change in value per unit of time, typically using one week as the time increment; or
  • relative growth = (final weight – initial weight) / initial weight; relative growth may be used when there is a significant difference among cages in initial weights; relative growth is expressed as a proportion and therefore an arcsin square root transformation of relative growth values may be appropriate prior to applying statistics. Green (1979) provides useful advice on this transformation, noting that it is not usually required for proportion data in the range of 0.3-0.7, and that while it may not always help, it probably does no harm, either.

Use the most appropriate expression of growth to suit the study design and site-specific characteristics.

Top of Page Condition

Condition is a measure of how an animal stores its energy, and it can be measured in both adults and juvenile mussels. There is more than one option that may be considered for calculating condition, as described below. Note that some of these methods require measurement of variables that are in addition to those outlined in section The most appropriate method to calculate condition is left to the discretion of the investigators.

Weight (whole-animal dry weight, dry shell or soft tissue weight) related to shell length: This is analogous to the Fulton Condition Index (Ricker 1975; Anderson and Neumann 1996) used in fisheries biology. This relationship may be characterized according to a conventional formula for a straight line (e.g., in Mackie and Flippance 1983), with slope (C) and intercept (b):

log W = b + (C × log L)

High values of C imply that a bivalve has a relatively high tissue weight at a given length, whereas low values may indicate that an animal is not obtaining sufficient food or is experiencing chronic stress that prevents it from thriving. This method of characterizing condition is suitable for assessing condition in wild bivalves where shell length is expected to be quite variable. However, since animals used for caged bivalve studies are screened for uniform length at test initiation, this method will not be reliable. This method may also not be suitable for bivalves because shell length and tissue weight are influenced by different factors in the environment (Salazar, personal communication).

Soft tissue weight related to shell weight: This method of characterizing condition uses soft tissue weight and shell weight. An ANOVA can be conducted or, more simply, soft tissue weight can be divided by shell weight. Grout and Levings (2000) measured the condition of Blue Mussels as the ratio of tissue weight to shell weight. They found that condition distinguished caged mussels in a high survival zone (condition index 1.10 to 1.42) from caged mussels in a low survival zone (condition index 0.82 to 0.96).

Soft tissue weight related to shell volume: This method was used by Mucklow (1996; based on Seed 1968) to calculate condition by dividing soft tissue dry weight by shell volume, measured as length x width x height. In a study on wild Blue Mussel populations, Mucklow (1996) concluded that seasonal patterns in condition index were variable and influenced by a number of natural factors, including food availability and physiological energy requirements.

Top of Page Energy Measure

As seen in Figure 9‑11, energy accumulation also occurs in the mantle, and has an annual cycle. Generally, Blue Mussels will reach their maximal energy content in late fall. When the mantle is at that stage, most of the weight consists of stored glycogen that will be used later for reproduction. Prior to dissection, mussels should be assessed for WAWW and shell measurements (length, weight, height and internal scarring). Mantle lobes should be separated from the body and weighed (mantle wet weight), after which the weight of the remainder of the body should be added to determine the body wet weight. Samples should be dried at 55°C until a constant weight is reached (approximately 2-3 days). Mantle dry weight and body dry weight should both be measured, from which the LSI-like measure can be derived. The bivalve LSI consists of the ratio of the dry weight of the mantle to the dry weight of the animal.

Figure 9-16: A) Mantle plug and tools necessary for its removal; B) Mantle plug after homogenization and ready for assessment

Figure 9-16: A) Mantle plug and tools necessary for its removal; B) Mantle plug after homogenization and ready for assessment.

Top of Page Reproductive Effort

Recent research has found that the MSI for mussels, similar to the GSI in fish, can be used to determine reproductive investment. Most of the gametes produced by the mussel are stored in the mantle lobes. The following is a summary of steps to determine the MSI of mussels:

  1. Measure the mussel length, width, height and the whole animal weight.
  2. Dissect each mussel.
  3. Determine the sex.
  4. Remove and weigh each mantle lobe and determine the body mass of each individual.
  5. If female: one mantle lobe should be used for dry weight and calculation of GSI ratio (ratio of body weight [minus gonads] to gonad). If male, both lobes are used. Female calculation should be extrapolated for both lobes.
  6. The second lobe should be used to assess reproductive effort, through egg measurement and count.
  7. Calculate the MSI (ratio of body weight [minus gonads] to gonad).

The MSI should be determined when 90% of the mantle lobe consists of gonads. There are numerous factors to consider that will affect the time of spawning: water depth, seasonal temperature, response patterns, and prevalence of different species. Boudreau et al. (in preparation), and St-Jean et al. (2008) conducted studies using MSI on the Atlantic and Pacific coasts of Canada, using different species. They found that Bay Mussels from the Pacific coast did not have a characteristic peak in gamete production preceding spawning, unlike the Blue Mussels from the Atlantic coast. Boudreau et al. (in preparation) also assessed reproductive effort by calculating and weighing the number of eggs in each mantle.

For more detailed information on the MSI, please refer to St-Jean (2003), or contact a regional EEM program coordinator for further information and complete techniques.

Briefly, the supplemental technique consists of the following (Fig. 9-16):

  1. Weigh gonad lobe on analytical balance.
  2. Extract a small plug of tissue using the straws provided.
  3. Weigh the plug.
  4. Mince the plug.
  5. Using the microscope ocular grid at a magnification of 400x, measure 20 eggs (measurements are of the whole egg and nucleus).
  6. Count representative sub-samples in a counting chamber.
  7. Repeat Step 6, 2 more times, for a total of 3 counts.
Table 9-8: An example of a field data sheet for recording survival and growth raw data.*

Caged Bivalve Field Data Sheet (text description)

Date received
Date processed
Chemical analysis required?
Date shipped
Sample weight (chemistry)

Length (L)
Height (H)
Width (W)
Mantle wet weight (WW)
Whole -Animal Dry Weight
Mantle dry weight (DW)
Dry Shell Weight

Number of mussels alive at end of test _________ Percent survival _________

Site Conditions:
(°C )
Dissolved Oxygen
(parts per thousand)
Current Velocity
Current Direction

Field Staff: _____________________________________________


* Data sheet provided by SSJ Environmental Limited

Top of Page

9.3.9 Quality Assurance and Quality Control

All work should be performed by suitably qualified and trained staff (biologists and technicians). Where contractors are used they should be selected for their specialist expertise. All fieldwork should be carried out following standard operating procedures to ensure overall consistency and that appropriate procedures are followed. All field and laboratory measurements should be made using properly calibrated instruments. All field data should be recorded using standard forms to ensure that all of the required data are collected in a reproducible and standardized format.

Replicate measures should be taken on 20% of all measures in order to verify the accuracy and reproducibility of both field measurements and laboratory analyses. For mussels in socks, this represents at least two mussels per sock.

In data analysis, the first step should be the screening of the data for outliers. A rapid way to screen for outliers is to create scatter plots of pairs of variables with 95% confidence ellipses superimposed. Potential outlier data points can then be identified as those that lie outside the confidence ellipses. Outliers can be the result of a number of causes, including data entry or transcription errors. Where outliers are detected, the data records should be reviewed in order to isolate and if possible correct the source of a potential error. Where no such identification is possible, the analysis should be performed both with and without the outlier, in order to evaluate the influence that the outlier exerts on the results of the data analysis.

Statistical data will be examined to evaluate the degree to which they conform to the underlying assumptions of the analysis (such as normality and homogeneity of variance, or equality of slopes in ANCOVA). Where appropriate, transformations may be applied in order to lessen the magnitude of violations of the underlying assumptions.

Top of Page

9.3.10 Data Analyses

Data analyses and interpretation of the results should be appropriate to the study design. Caged mussel studies for EEM are designed to determine if there are significant effects on biota in the vicinity of effluent outfalls. This can be accomplished by using a C-I (= reference − exposure) or a gradient (= regression) type of design. Statistical procedures appropriate to each effect indicator (i.e., survival, growth, reproduction, energy and condition) are summarized in Table 9‑7.

An ANCOVA should be performed to test the GSI (dry gonad weight), condition (dry body weight) with covariates to remove influences including dry body weight for GSI, and shell length for condition. Where an interaction between treatment and covariate precludes the use of an ANCOVA, stratified subsets of the covariate should be compared in a single-factor ANOVA. When the control groups are not significantly different, they should be grouped for the analyses. However, if a significant difference is found between groups, all controls should be included in the analyses. A Tukey multiple-comparison test can be employed when significant differences are found among groups. Non-normality (probability plot) or heteroscedasticity (Fmax test) that cannot be resolved by appropriate data transformations should be followed by non-parametric analysis using the Kruskal-Wallis test, followed by a Noether multiple-comparison test (Scherrer 1984; Zar 1999). Probit analysis can be employed for survival. The level of significance should be set at p < 0.05, and back-transformed means should be accompanied by their 95% confidence interval.

The first step is to generate summary statistics for each parameter (i.e., WAWW, shell length, soft tissue fresh weight) and each cage and station.

The second step is to determine whether there are significant differences among replicate cages for each of the parameters measured before deployment (if they were not measured post-distribution and before deployment). This involves assessing the data for normality and homogeneity of variances.

The final step is to use the appropriate statistical test for the study design. In general, ANOVA and multiple-comparison tests are used for hypothesis testing and comparison among stations. For C-I and MC-I designs using ANOVA and ANCOVA procedures, detailed guidance is provided in Chapter 8 of this guidance document. If statistical differences are found, a multiple-range test, or its non-parametric counterpart, can be used to determine which stations are different from the others. Linear and multiple-regression analyses (using variables regressed on distance) may generally be used to establish relationships among variables along exposure gradients. For ANCOVA analyses of condition, the covariates will be dictated by the condition formula that is used.

Top of Page

9.3.11 Mercury in Fish Tissue

If a mine is required to measure mercury in fish tissue, caged bivalves may be considered. However, there are some precautions that should be taken with this approach, including the following:

  • Duration of exposure should be chosen such that mercury is likely to be accumulated to detectable levels in bivalve tissue.
  • Consider whether bivalves would be harvested (commercially or recreationally) for human consumption in the area.
  • Ensure that the number of bivalves included in the design is sufficient to obtain enough tissue for analysis.

There may be mines for which caged bivalves would not be considered suitable for measuring uptake of mercury. Any proposal to use caged bivalves for mercury measurements would require approval by the Regional Authorization Officer.

Top of Page

9.3.12 Reporting

QA/QC considerations for caged bivalve studies should follow those outlined for the fish survey in Chapter 3 of this guidance document. QA/QC measures apply to the following components of caged bivalve studies:

  • study design;
  • field sampling;
  • sample processing / laboratory analysis;
  • data analysis; and
  • reporting.

9.4 References

Anderson RO, Neumann RM. 1996. Length, weight, and associated structural indices. In Murphy BR, Willis DW, editors. Fisheries techniques. 2nd edition. Bethesda (MD): American Fisheries Society. p. 447-482.

Andrews S, Parker R. 1999. Draft Report (II). An environmental quality evaluation of Pictou Harbour, Nova Scotia, using caged bivalves; Mytilus edulis. Prepared by Environmental Protection Branch of Environment Canada, Dartmouth (NS) for the Pictou Harbour Environmental Protection Project, New Glasgow (NS).

Ankley GT, Jensen KM, Kahl MD, Korte JJ, Makynen EA. 2001. Description and evaluation of a short-term reproduction test with the fathead minnow (Pimephales promelas). Environ Toxicol Chem 20:1276-1290.

Ankley GT, Jensen KM, Makynen EA, Kahl MD, Korte JJ, Hornung MW, Henry TR, Denny JS, Leino RL, Wilson VS, et al. 2003. Effects of the androgenic growth promoter 17-beta-trenbolone on fecundity and reproductive endocrinology of the fathead minnow. Environ Toxicol Chem 22:1350-1360.

Applied Biomonitoring. 2000. Caged Mussel Pilot Study Port Alice Mill, Vancouver Island. EEM Program. Environment Canada, Environmental Protection Branch, Pacific and Yukon Region. Regional Manuscript Report MS 00-01.

Atz JW. 1986. Fundulus heteroclitus in the laboratory: a history. Amer Zool 26:111-120.

Top of Page

Barber WE, Kevern NR. 1974. Seasonal variation of sieving efficiency in a lotic habitat. Freshwat Biol 4:293-300.

Bauer G. 1994. The adaptive value of offspring size among freshwater mussels (Bivalvia; Unionoidea). J Anim Ecol 63:933–944.

Beckvar N, Salazar S, Salazar M, Finkelstein K. 2000. An in situ assessment of mercury contamination in the Sudbury River, Massachusetts, using transplanted freshwater mussels (Elliptio complanata). Can J Fish Aquat Sci 57:1103-1112.

Bothwell ML. 1985. Phosphorus limitation of lotic periphyton growth rates: an intersite comparison using continuous-flow troughs (Thompson River system, British Columbia). Limnol Oceanogr 30:527-542.

Bothwell ML, Daley RJ. 1981. Determination of phosphorus sufficiency and growth rates of benthic algae in the Thompson River, BC, using experimental flowing troughs. IWD Regional Report. Vancouver (BC): Inland Waters Directorate, Environment Canada.

Bothwell ML, Derksen G, Nordin RN, Culp JM. 1992. Nutrient and grazer control of algal biomass in the Thompson River, British Columbia: a case history of water quality management, InRobarts RD, Bothwell ML, editors. Aquatic ecosystems in semi-arid regions: implications for resource management. (National Hydrology Research Institute Symposium Series 7). Saskatoon (SK): Environment Canada. p. 253-266.

Bothwell ML, Culp JM. 1993. Sensitivity of the Thompson River to phosphorus: studies on trophic dynamics. NHRI Contribution No. 93006.Saskatoon (SK): National Hydrology Research Institute, Environment Canada.

Burt A, Kramer DL, Nakasuru K, Spry C. 1988. The tempo of reproduction in Hyphessobrycon pulchripinnis (Characidae), with a discussion on the biology of “multiple spawning”. Environ Biol Fish 22:15-27.

Top of Page

Carlisle DM. 2000. Bioenergetic food webs as a means of linking toxicological effects across scales of ecological organization. J Aquat Eco Stress Rec 7:155-165.

Cash KJ, Culp JM, Dubé MG, Lowell RB, Glozier NE, Brua RB. 2003. Integrating mesocosm experiments with field and laboratory studies to generate weight-of-evidence risk assessments for ecosystem health. Aquatic Ecosystem Health and Management. 6: 177-183.

Chen CY, Stemberger RS, Klaue B, Blum JD, Pickhardt PC, Folt CL. 2000. Accumulation of heavy metals in food web components across a gradient of lakes. Limnol Oceanogr 45:1525-1536.

Clarke A. 1973. The freshwater molluscs of the Canadian Interior Basin. Malacologia 13:1-509.

Clarke A. 1981. The freshwater molluscs of Canada. Ottawa (ON): National Museums of Canada.

Coad BW. 1995. Encyclopedia of Canadian fishes. Waterdown (ON): Canada Museum of Nature and Canadian Sportfishing Productions Inc.

Couillard Y, Campbell PGC, Tessier A, Pellerin-Massicotte J, Auclair JC. 1995a. Field transplantation of a freshwater bivalve, Pyganodon grandis, across a metal contamination gradient. I. Temporal changes in metallothionein and metal (Cd, Cu, and Zn) concentrations in soft tissues Can J Fish Aquat Sci 52:690-702.

Couillard Y, Campbell PGC, Tessier A, Pellerin-Massicotte J, Auclair JC. 1995b. Field transplantation of a freshwater bivalve, Pyganodon grandis, across a metal contamination gradient. II. Metallothionein response to Cd and Zn exposure, evidence for cytotoxicity, and links to effects at higher levels of biological organization. Can J Fish Aquat Sci 52:703-715.

Courtenay SC, Parker WR, Rawn GP. 1998. Proceedings of a workshop to assess alternatives to the fish survey component of the environmental effects monitoring program for Canadian pulp and paper mills. (Canadian Technical Report of Fisheries and Aquatic Sciences 2233).

Crane M, Burton GA, Culp JM, Greenberg MS, Munkittrick KR, Ribeiro R, Salazar M, St-Jean S. 2007. In situ approaches for stressor and effect diagnosis. Integr Environ Assess Manag 3:234-245.

Culp JM, Cash KJ. 1995. Potential impacts of effluent on the Thompson and upper Fraser Rivers. DOE FRAP 1995-10. Vancouver (BC): Fraser River Action Plan, Environmental Conservation Branch, Environment Canada.

Culp JM, Podemski CL. 1996. Design and application of a novel stream microcosm system for assessing effluent impacts to large rivers. In Servos MR, Munkittrick KR, Carey JH, Van Der Kraak GJ, editors. Environmental fate and effects of pulp and paper mill effluents. Delray Beach (FL): St. Lucie Press. p. 549-555.

Culp JM, Podemski CL, Cash KJ, Lowell RB. 1996. Utility of field-based artificial streams for assessing effluent effects on riverine ecosystems. J Aquat Ecosystem Health 5:117-124.

Culp JM, Lowell RB. 1998. Pulp mill effluent impacts on benthic communities and selected fish species in the Fraser River Basin. In Gray C, Tuominen T, editors. Health of the Fraser River aquatic ecosystem, Vol. II. DOE FRAP 1998-11. Ottawa (ON): Environment Canada. p. 13-34.

Culp JM, Lowell RB, Cash KJ. 2000a. Integrating mesocosm experiments with field and laboratory studies to generate weight-of-evidence risk assessments for large rivers. Environ Toxicol Chem 19:1167-1173.

Culp JM, Podemski CL, Cash KJ, Lowell RB. 2000b. A research strategy for using stream microcosms in ecotoxicology: integrating experiments at different levels of biological organization with field data. J Aquat Eco Stress Rec 7:167-176.

Culp JM, Podemski CL, Cash KJ. 2001. Interactive effects of nutrients and contaminates from pulp mill effluents on riverine benthos. J Aquat Eco Stress Rec 8(1):67-75.

Culp JM, Glozier NE, Cash KJ, Dubé MG, Waiser M, MacLatchy DL, Brua B, Firth B, Wilson G. 2004. Cumulative effects investigation of pulp mill and sewage effluent impacts on benthic food webs: a mesocosm example. In Borton DL, Hall TJ, Fisher RP and Thomas JF, editors. Pulp and paper mill effluent environmental fate and effects, proceedings of the Fifth International Conference on Fate and Effects of Pulp and Paper Mill Effluents, June 1-4, 2003, Seattle, Washington, USA. Lancaster (PA): DEStech Publications. p. 464-472.

Top of Page

Day KE, Metcalfe JL, Batchelor SP. 1990. Changes in intracellular free amino acids in tissues of the caged mussel (Elliptio complanata), exposed to contaminated environments. Arch Environ Cont Toxicol 19:816-827.

Drake JA, Huxel GR, Hewitt CL. 1996. Microcosms as models for generating and testing community theory. Ecology 77:670-677.

Dubé MG. 1995. Effects of bleached kraft pulp mill effluent on periphyton and chironomid abundance in the Thompson River, British Columbia [master’s thesis]. Saskatoon (SK): University of Saskatchewan.

Dubé MG. 2000. Sources of contaminants in the bleached kraft pulping process and their effect on the mummichog (Fundulus heteroclitus) [dissertation]. Saint John (NB): University of New Brunswick.

Dubé MG, Culp JM. 1996. Growth responses of periphyton and chironomids exposed to biologically treated bleached kraft pulp mill effluent. Environ Toxicol Chem 15:2019-2027.

Dubé MG, Culp JM, Scrimgeour GJ. 1997. Nutrient limitation and herbivory: processes influenced by bleached kraft pulp mill effluent. Can J Fish Aquat Sci 54:2584-2595.

Dubé MG, MacLatchy DL. 2000a. Endocrine responses of an estuarine fish Fundulus heteroclitus to final effluent from a bleached kraft pulp mill before and after installation of reverse osmosis treatment of a waste stream. Environ Toxicol Chem 19:2788-2796.

Dubé MG, MacLatchy DL. 2000b. Reverse osmosis treatment of condensates from a bleached kraft pulp mill: effects on acute and chronic toxicity of process streams and final effluent. InRuoppa M, Paasivirta J, Lehtinen K-J, Ruonala S, editors. Proceedings of the Fourth International Conference on Environmental Impacts of the Pulp and Paper Industry. Helsinki (FI). p. 270-276.

Dubé MG, MacLatchy DL. 2001. Identification and treatment of a waste stream at a bleached kraft pulp mill that depresses a sex steroid in the mummichog (Fundulus heteroclitus). Environ Toxicol Chem 20:985-995.

Dubé MG, Culp JM, Cash KJ, Glozier NE, MacLatchy DL, Podemski CL, Lowell RB. 2001. Artificial streams for environmental effects monitoring (EEM): development and application in Canada over the past decade. Wat Qual Res J Can 37(1):155-180.

Dubé MG, Culp JC, MacLatchy DL, Gillis G, Parker R, Courtenay S, Gilman G. 2002. Utility of mobile, field-based artificial streams for assessing effects of pulp mill effluents on fish in the Canadian Environmental Effects Monitoring (EEM) program. J Aquat Eco Stress Rec 9:85-102.

Dubé MG, MacLatchy DL, Firth BK, Culp JM, Glozier NE, Cash KJ. 2004. Using mesocosms to explore confounding factors influencing longnose dace (Rhinichthys cataractae) responses to kraft mill effluent on the Wapiti River, AB, Canada. In Borton DL, Hall TJ, Fisher RP, Thomas JF, editors. Pulp and paper mill effluent environmental fate and effects, Proceedings of the Fifth International Conference on Fate and Effects of Pulp and Paper Mill Effluents, June 1-4, 2003, Seattle, Washington, USA. Lancaster (PA): DEStech Publications. p. 481-490.

Dubé MG, MacLatchy DL, Kieffer JD, Glozier NE, Culp JM, Cash KJ. 2005. Effects of metal mining effluent on Atlantic salmon (Salmo salar) and slimy sculpin (Cottus cognatus): using mesocosms to assess existing effects and predict future consequences. Sci Total Environ 343:135-154.

Dubé MG, MacLatchy DL, Hruska KA, Glozier NE. 2006. Assessing the responses of creek chub (Semotilus atromaculatus) and pearl dace (Semotilus margarita) to metal mine effluents using in situ artificial streams in Sudbury, Ontario, Canada. Environ Toxicol Chem. 25(1):18-28.

Top of Page

Environment Canada. 1997. Fish Survey Working Group Final Report. Recommendations from Cycle 1 Review. Ottawa (ON): National EEM Office, Science Policy and Environmental Quality Branch, Environment Canada. EEM/1997/6.

Federal-Provincial Thompson River Task Force. 1976. Sources and effects of algal growth, colour, foaming and fish tainting in the Thompson River system. Victoria (BC): Environmental Protection Division, British Columbia Ministry of Environment, Lands and Parks.

Freeman KR, PerryKL, DiBaccoTG, Scarratt DJ. 1994. Observations on two Mytilid species from a Nova Scotian mussel farm. (Canadian Technical Report of Fisheries and Aquatic Sciences 1969). Halifax (NS): Fisheries and Oceans Canada.

Fritz ES, Meredith WH, Lotrich VA. 1975. Fall and winter movements and activity level of the mummichog Fundulus heteroclitus, in a tidal creek. Chesapeake Sci. 16:211-215.

Top of Page

Gibbons WN, Munkittrick KR, Taylor WD. 1998a Suitability of small fish species for monitoring the effects of pulp mill effluent on fish. 1. Response of spoonhead sculpin exposed to bleached kraft mill effluent. Environ Toxicol Chem 7(11):2227-2237.

Gibbons WN, Munkittrick KR, Taylor WD. 1998b. Suitability of small fish species for monitoring the effects of pulp mill effluent on fish. 2. Comparison of trout perch responses to white sucker exposed to bleached kraft mill effluent. Environ Toxicol Chem 17(11):2238-2245.

Goldberg ED, Bowen VT, Farrington JW, Harvey G, Martin JH, Parker PL, Risebrough RW, Robertson W, Schneider E, Gamble E. 1978. The Mussel Watch. Environ Conserv 5(2):101-125.

Gray C, Tuominen T. 1998. Health of the Fraser River aquatic ecosystem, Vol. I. DOE FRAP 1998-11. Vancouver (BC): Environment Canada.

Green RH. 1979. Sampling design and statistical methods for environmental biologists. Toronto (ON): John Wiley & Sons.

Grout JA, Levings CD. 2000. Effects of acid mine drainage from an abandoned copper mine, Britannia Mines, Howe Sound, British Columbia, Canada, on transplanted blue mussels (Mytilus edulis). Mar Environ Res 51(3):265-288.

Gulf of Maine Council on the Marine Environment. 1997. Gulfwatch Project: standard procedures: filed and laboratory. Gulfwatch implementation period 1993-2001. August 1997.

Top of Page

Harries JE, Runnals T, Hill E, Harris CA, Maddix S, Sumpter JP, Tyler CR. 2000. Development of a reproductive performance test for endocrine disrupting chemicals using pair breeding fathead minnows (Pimephales promelas). Environ Sci Technol34:3003-3011.

Heins DC, Rabito FG Jr. 1986. Spawning performance in North American minnows: direct evidence of the occurrence of multiple clutches in the genus Notropis. J Fish Biol 28:343-357.

Hinch SG, Kelly LJ, Green RH.1989. Morphological variation of Elliptio complanata (Bivalvia: Unionidae) in differing substrata of soft-water lakes exposed to acidic deposition. Can J Zool 67:1895-1899.

Hornbach DJ, Wissing TH, Burky AJ. 1982. Life-history characteristics of a stream population of the freshwater clam Sphaerium striatinum Lamarck (Bivalvia: Pisidiidae). Can J Zool. 60:249‑260.

Hruska K, Dubé MG. 2004. Using artificial streams to assess the effects of metal-mining effluent on the life-cycle of the freshwater midge (Chironomus tentans) in situ. Environ Toxicol Chem 23(11):2709-2718.

Hruska K, Dubé M. 2005. Comparison of a partial life cycle bioassay in artificial streams to a standard beaker bioassay to assess effects of metal mine effluent on Chironomus tentans. Environ Toxicol Chem 24(9):2325-2335.

Hurlbert SH. 1984. Pseudoreplcation and the design of ecological field experiments. Ecol Monog 54:187-211.

Top of Page

Jadhav ML, Lomte VS. 1982. Seasonal variation in biochemical composition of the freshwater bivalve, Lamellidens corrianus. Rivista di Idrobiologia. 21:1-7.

Jenkin, RE, Burkhead NM. 1993. Freshwater fishes of Virginia. Bethesda (MD): American Fisheries Society.

Jensen KM, Korte JJ, Kahl MD, Pasha MS, Ankley GT. 2001. Aspects of basic reproductive biology and endocrinology in the fathead minnow (Pimephales promelas). Comp Biochem Physiol Part C. 128:127-141.

Jensen KM, Kahl MD, Makynen EA, Korte JJ, Leino RL, Butterworth BC, Ankley GT. 2004. Characterization of responses to the antiandrogen flutamide in a short-term reproduction assay with the fathead minnow. Aquat Toxicol70:99-110.

Jernelov A. 1996. The International Mussel Watch: A global assessment of environmental levels of chemical contaminants. Sci Tot Environ 188 (Suppl. 1): 37-44.

Kneib RT. 1986. The role of Fundulus heteroclitus in salt marsh trophic dynamics. Amer Zool26:259-269.

Kneib RT, Stiven AE. 1978. Growth, reproduction, and feeding of Fundulus heteroclitus (L.) on a North Carolina salt marsh. J Exp Mar Biol Ecol31:121-140.

Top of Page

Leblanc J, Couillard CM. 1995. Description de la période de reproduction d’un poisson sentinelle: le choquemort (Fundulus heteroclitus) de l’estuaire de la Miramichi. (Rapport technique canadien des sciences halieutiques et aquatiques 2057). Mont-Joli (QC): Fisheries and Oceans Canada.

Lowell RB, Culp JM, Wrona FJ. 1995. Stimulation of increased short-term growth and development of mayflies by pulp mill effluent. Environ Toxicol Chem 14:1529-1541.

Lowell RB, Culp JM, Wrona FJ, Bothwell ML. 1996. Effects of pulp mill effluent on benthic freshwater invertebrates: food availability and stimulation of increased growth and development. In Servos MR, Munkittrick KR, Carey JH, Van Der Kraak GJ, editors. Environmental fate and effects of pulp and paper mill effluents. Delay Beach (FL): St. Lucie Press. p. 525-532.

Lowell RB, Culp JM, Dubé MG. 2000. A weight-of-evidence approach for northern river risk assessment: integrating the effects of multiple stressors. Environ Toxicol Chem 19:1182-1190.

Top of Page

Mackie GL. 1978a. Shell structure in freshwater Sphaeriaceae. (Bivalvia: Heterondonta). Can J Zool 56:1-6.

Mackie GL. 1978b. Are sphaeriid clams ovoviviparous or viviparous? Nautilus 92:145-147.

Mackie GL, Quadri SU, Clarke AH. 1974. Development of brood sacs in Musculium securis Bivalvia: Sphaeriidae. Nautilus 88:109-111.

Mackie GL, Flippance LA. 1983. Growth dynamics in Sphaerium rhomboideum (Bivalvia: Pisidiidae). Can J Zool 61:868-873.

Mallet AL, Carver CE. 1995. Comparative growth and survival patterns of Mytilus trossulus and Mytilus edulis in Atlantic Canada. Can J Fish Aquat Sci 52:1873-1880.

Malley DF, Stewart AR, Hall BD. 1996. Uptake of methyl mercury by the floater mussel, Pyganodon grandis (Bivalvia, Unionidae), caged in a flooded wetland. Environ Toxicol Chem 15:928-936.

Martel P, Kovacs T, Voss R. 2003. Survey of pulp and paper mill effluents for their potential to affect fish reproduction. Proceedings of the Fifth International Conference on the Fate and Effects of Pulp and Paper Mill Effluents, Seattle, Washington, USA, June 1-4, 2003. p. 78-91.

Mason RP, Laporte JM, Andres S. 2000. Factors controlling the bioaccumulation of mercury, methylmercury, arsenic, selenium and cadmium by freshwater invertebrates and fish. Arch Environ Contam Toxicol 38:283-297.

McDonald JH, Seed R, Koehn RK. 1991. Allozymes and morphometric characteristics of three species of Mytilus in the northern and southern hemispheres. Mar Biol 11:323-333.

McGreer ER, Belzer W. 1998. Contaminant sources. In Gray C, Tuominen T, editors. Health of the Fraser River aquatic ecosystem, Vol. I. DOE FRAP 1998-11. Ottawa (ON): Environment Canada. p. 7-22.

McMahon RF. 1991. Mollusca: bivalvia. In Thorp JH, Covich AP, editors. 1991. Ecology and classification of North American freshwater invertebrates. San Diego (CA): Academic Press. p. 315-399.

Metcalfe-Smith JL, Green RH, Grapentine LC. 1996. Influence of biological factors on concentrations of metals in the tissues of freshwater mussels (Elliptio complanata and Lampsilis radiata radiata) from the St. Lawrence River. Can J Fish Aquat Sci 53:205-219.

Mucklow LC. 1996. Effects of season and species on physiological condition and contaminant burdens in mussels (Mytilus edulis L. and Mytilus trossulus G.): implications for mussel watch programs [master’s thesis]. Halifax (NS): School for Resource and Environmental Studies, Dalhousie University.

Nelson JS, Paetz MJ. 1992. The fishes of Alberta. Calgary (AB): University of Calgary Press.

Newell RIE. 1989. Species profile of the blue mussel, Mytilus edulis. US Fish Wildl Ser Biol Rep 82(11):40.

Ni IH, Wang WX, Tam YK. 2000. Transfer of Cd, Cr and Zn from zooplankton prey to mudskipper Periophthalmus cantonensis and glassy Ambassis urotaenia fishes. Mar Ecol-Prog Ser 194:203-210.

Top of Page

O’Connor. 1992. Recent trends in coastal environmental quality: Results from the first five years of the NOAA Mussel Watch Project. Washington (DC): U.S. Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), and National Ocean Service.

[OECD] Organisation for Economic Co-operation and Development. 2001. Detailed review paper: appraisal of test methods for sex hormone disrupting chemicals. (OECD Series on testing and assessment 21). ENV/JM/MONO(2002)8.

[OECD] Organisation for Economic Co-operation and Development. 2006. Report of the validation of the 21-day fish screening assay for the detection of endocrine substances (Phase 1B). (OECD Series on testing and assessment 61). ENV/JM/MONO(2006)29.

Paine MD. 1990. Life history tactics of darters (Percidae: Etheostomatiini) and their relationship with body size, reproductive behaviour, latitude and rarity. J Fish Biol 37:473-488.

Paine MD. 2000. EEM Design. Short-Course presented at the 27th Annual Aquatic Toxicity Workshop, October 1-4th, St. John’s, NF.

Panter GH, Hutchinson TH, Lange R, Lye CM, Sumpter JP, Zerulla M, Tyler CR. 2002. Utility of a juvenile fathead minnow screening assay for detecting (anti-)estrogenic substances. Environ Toxicol Chem 21(2):319-326.

Parker R, Smith N. 1997. A synopsis of the results of environmental effects monitoring studies at 19 pulp and paper mills in Atlantic Canada. In Proceedings of the Third International Conference on Environmental Effects of Pulp and Paper Mill Effluents. Rotorua, NZ. p. 432-441.

Parrott JL. 2005. Overview of methodology and endpoints in fathead minnow lifecycle tests assessing pulp and paper mill effluents. Water Qual Res Can40:334-346.

Parrott JL, Wood CS. 2001. Fathead minnow lifecycle tests for detection of endocrine disrupting substances in effluents. Water Qual Res J Can37:651-667.

Penney RW, Hart MJ, Templeman N. 2002. Comparative growth of cultured blue mussels, Mytilus edulis, M. trossulus and their hybrids, in naturally occurring mixed-species stocks. Aquat Res 33:693-702.

Podemski CL. 1999. Ecological effects of a bleached kraft pulp mill effluent on benthic biota of the Athabasca River [dissertation]. Saskatook (SK): University of Saskatchewan.

Podemski CL, Culp JM. 1996. Nutrient and contaminant effects of bleached kraft mill effluent on benthic algae and insects of the Athabasca River. In Servos MR, Munkittrick KR, Carey JH, Van Der Kraak GJ, editors. Environmental fate and effects of pulp and paper mill effluents. Delay Beach (FL): St. Lucie Press. p. 571-580.

Pollock MS, Fisher SE, Squires AJ, Pollock RJ, Chivers DP. Dubé MG. 2008. Relative body size influences breeding propensity in fathead minnows: implication for ecotoxicology testing procedures. Water Qual Res J Can 43(4):257-264.

Pollock MS, Dubé MG, LaPorte J, Clarke L, Schryer R. 2009. Investigating the link between pulp mill effluent and endocrine disruption: attempts to explain the presence of male and female hermaphrodites in the Wabigoon River, Ontario. Environ Toxicol Chem 29(4):952-965.

Portt CB, Coker GA, Ming DL, Randall RG. 2006. A review of fish sampling methods commonly used in Canadian freshwater habitats. (Canadian Technical Report of Fisheries and Aquatic Sciences 2604). Cat. No.Fs97-6/2604E. Ottawa (ON): Fisheries and Ocean Canada.

Top of Page

Ramón M, Richardson CA. 1992. Age determination and shell growth of Chamelea gallina (Bivalvia: Veneridae) in the western Mediterranean. Mar Ecol Prog Ser 89:15-23.

Rawson PD, Slaughter C, Yund PO. 2003. Patterns of gamete incompatibility between the blue mussels Mytilus edulis and M. trossulus. Mar Biol 143:317-325.

Rees HL. 1984. A note on mesh selection and sampling efficiency in benthic studies. Mar Poll Bull 15:225-229.

Resh VH, McElravy EP. 1993. Contemporary quantitative approaches to biomonitoring using benthic macroinvertebrates. In Rosenberg DM, Resh VH, editors. Freshwater biomonitoring and benthic invertebrates. New York (NY): Chapman & Hall. p. 159-194.

Ricker WE. 1975. Computation and interpretation of biological statistics of fish populations. Bulletin 191. Ottawa (ON): Department of the Environment and Fisheries and Marine Service.

Rickwood CJ, Dubé MD, Hewitt LM, Kovacs T, Parrott JL, MacLatchy DL. 2006a. Use of paired fathead minnow (Pimephales promelas) reproductive test: Part I: Assessing biological effects of final bleached kraft pulp mill effluent using a mobile bioassay trailer system. Environ Toxicol Chem 25:191-201.

Rickwood CJ, Dubé MD, Hewitt LM, Kovacs T, Parrott JL, MacLatchy DL. 2006b. Use of paired fathead minnow (Pimephales promelas) reproductive test: Part II Source identification of biological effects at a bleached kraft pulp mill. Environ Toxicol Chem 25:202-211.

Rickwood CJ, Dube MG, Weber L, Driedger K, Janz DM. 2006c. Assessing effects of metal mining effluent on fathead minnow (Pimephales promelas) reproduction in a trophic-transfer system. Environ Sci Technol 40:6489-6497.

Rickwood CJ, Dubé MG, Weber LP, Lux S, Janz DM. 2008. Assessing effects of a mining and municipal sewage effluent mixture on fathead minnow (Pimephales promelas) reproduction using a novel, field-based trophic-transfer artificial stream. Aquat Toxicol 86(2):272-286.

Roberts WE. 1988. The sculpins of Alberta. Alberta Naturalist 18:121-127.

Top of Page

Salazar MH, Salazar SM. 1995. In situ bioassays using transplanted mussels: I. Estimating chemical exposure and bioeffects with bioaccumulation and growth. In Hughes JS, Biddinger GR, Mones E, editors. Environmental toxicology and risk assessment – Third Volume. Philadelphia (PA): American Society for Testing and Materials. p. 216-241.

Salazar M, Salazar S. 2000. Standard guide for conducting in-situ field bioassays with caged marine, estuarine and freshwater bivalves. (Draft dated November 17, 2000). Philadelphia (PA): American Society for Testing and Materials.

Sandusky MJ, Sparks RE, Paparo AA. 1979. Investigations of declines in fingernail clam (Musculium transversum) populations in the Illinois River and Pool 19 of the Mississippi River. Bull Amer Malacol Union 1979:11-16.

Scherrer B. 1984. Biostatistique. Québec (QC): Gaëtan Morin.

Scott WB. 1967. Freshwater fishes of eastern Canada. Toronto (ON): University of Toronto Press.

Scott WB, Crossman EJ. 1973. Freshwater fishes of Canada. (Fisheries Research Board of Canada Bulletin 184). Ottawa (ON): Fisheries Research Board of Canada.

Scott WB, Scott MG. 1988. Atlantic fishes of Canada. Can Bull Fish Aquat Sci219:314-316.
Scrimgeour GJ, Culp JM, Glozier NE. 1993. An improved technique for sampling lotic invertebrates. Hydrobiologia 254:65-71.

Seed R. 1968. Factors influencing shell shape in Mytilus edulis L. J Mar Biol Ass UK 48:561‑584.

Sejr M, Jensen KT, Rysgaard S. 2002. Annual growth bands in the bivalve Hiatella arctica validated by a mark-recapture study in NE Greenland. Polar Biol 25:794-796.

Shaw JL, Maund SJ, Marshall SJ, Hill IR. 1995a. Fathead minnow (Pimephales promelas) reproduction in outdoor microcosms: an assessment of the ecological effects of fish density. Environ Toxicol Chem 14(10):1763-1772.

Shaw JL, Maund SJ, Hill IR. 1995b. Fathead minnow reproduction in outdoor microcosms: a comparison to bluegill sunfish reproduction in large mesocosms. Environ Toxicol Chem 14(10):1753-1762.

Skidmore DA, Chew KK. 1985. Mussel aquaculture in Puget Sound. Seattle (WA): Washington Sea Grant Report. WSG 85-4.

Slack KV, Averett RC, Geeson PE, Lipscomb RG. 1973. Methods for collection and analysis of aquatic biological and microbiological samples. In Techniques of water-resources investigations of the United States Geological Survey, Chapter 4A, Book 5. Washington (DC): U.S. Department of the Interior, Geological Survey. p. 1-165.

Sohoni P, Typer CR, Hurd K, Caunter J, Hetheridge M, Williams T, Evans M, Toy R, Gargas M,

Sumpter JP. 2001. Reproductive effects of long-term exposure to bisphenol a in the fathead minnow (Pimephales promelas). Environ Sci Technol35:2917-2925.

Stephenson M, Mackie GL. 1981. Effects of 2, 4-D treatment on growth and survival of fingernail clams (Bivalvia: Pisidiidae) in artificial pond ecosystems. Bull Amer Malacol Union 1981:19-22.

Stewart, PL. 1994. Environmental requirements of the Blue Mussel (Mytilus edulis) in eastern Canada and its response to human impacts. Canadian Technical Report of Fisheries and Aquatic Science, 2004.

Stewart R, Malley DF. 1997. Technical evaluation of molluscs as a biomonitoring tool for the Canadian mining industry. Aquatic Effects Technology Evaluation (AETE) Program Project 2.3.1. Sponsored by Canada Center for Mineral and Energy Technology (CANMET) and the Mining Association of Canada (MAC).

St-Jean SD. 2003. Supplemental techniques for cage bivalve. National Water Research Institute Contribution number: 03-422.

St-Jean SD, Courtenay SC, Parker RW. 2003. Immunomodulation in blue mussels (Mytilus edulis) exposed to a pulp and paper mill effluent in Eastern Canada. Wat Qual Res J Can 38(4)647-666.

St-Jean SD, Stephens RE, Courtenay SC, Reinisch CL. 2005. Detecting p53 family proteins in hemocytic leukemia cells of Mytilus edulis from Pictou Harbour, Nova Scotia. Can J Fish Aquat Sci 62:2055-2066.

St-Jean S, Courtenay S, Coomber C, Bishay F. 2008. Responses to environmental stressors of wild vs. caged mussels. EEM Science Symposium, Gatineau, Quebec. 29-30 April 2008.

Suess MJ. 1982. Examination of water for pollution control. A reference handbook. Vol. 3. Biological, bacteriological and virological examination. Oxford (UK):Permagon Press.

Top of Page

Toro JE, Thompson RJ, Innes DJ. 2002. Reproductive isolation and reproductive output in two sympatric mussels species (Mytilus edulis, M. trossulus) and their hybrids from Newfoundland. Mar Biol 141:897-909.

[US EPA] United States Environmental Protection Agency. 1982. User’s guide for conducting life-cycle chronic toxicity tests with fathead minnows (Pimephales promelas). Duluth (MN): Environmental Research Laboratory. U.S. EPA/600/8-81/011.

[US EPA] United States Environmental Protection Agency. 1996. Prehatching development of the fathead minnow Pimephales promelas. Washington (DC). EPA/600/R-96/079.

[US EPA]. 1999. Screening level reproduction assay with the fathead minnow (Pimephales promelas) Duluth (MN): EPA-NEERL-DUL 3544.

[US EPA] United States Environmental Protection Agency. 2002. Comparative evaluation of fathead minnow assays for detecting endocrine disrupting chemicals. Washington (DC): Endocrine Disruptor Screening Program. EPA/68/W-01/023.

[US EPA] United States Environmental Protection Agency. 2007. Short term test method for assessing the reproductive toxicity of endrocrine-disrupting chemicals using the fathead minnow (Pimephales promelas). Report No. 600R01067.

Varvio S-L, Koehn RK, Vkinola R. 1988. Evolutionary genetics of the Mytilus edulis complex in the North Atlantic region. Mar Biol 98:51-60.

Waldock MJ, Thain JE, Waite ME. 1996. An assessment of the value of shell thickening in Crassostrea gigas as an indicator of exposure to tributyltin. In Champ MA, Seligman PF, editors. Tributyltin: Environmental Fate and Effects. London (UK): Chapman and Hall. p. 219-237.

Weber CI. 1973. Biological field and laboratory methods for measuring the quality of surface waters and effluents. Cincinnati (OH): U.S. Environmental Protection Agency. EPA-670/4-73-001.

Wiederholm T. 1980. Use of benthos in lake monitoring. J Wat Poll Cont Fed 52:537-547.

Williams JD, Warren ML Jr, Cummings KS, Harris JL, Neves RJ. 1993. Conservation status of freshwater mussels of the United States and Canada. Fisheries 18(9):6-22.

Wolfe BA, Burkhard MJ, Leavell S, Weiss RB, Kuehnl K, Valentine H, Watters GT. 2009. Health and stress assessment in freshwater mussels following translocation to a captive setting. In Proceedings of the International Symposium of the Freshwater Mollusk Conservation Society, 19-24 April 2009, Baltimore, Maryland. p. 37.

Wrona FJ, Culp JM, Davies RW. 1982. Macroinvertebrate subsampling: a simplified apparatus and approach. Can J Fish Aquat Sci 39:1051-1054.

Xu Y, Wang WX. 2002. Exposure and potential food chain transfer factor of Cd, Se and Zn in marine fish Lutjanus argentimaculatus. Mar Ecol-Prog Ser 238:173-186.

Zar JH. 1999. Biostatistical analysis. Fourth edition. Upper Saddle River (NJ): Prentice Hall.

Figures and Tables

Table 9-1 provides a summary of artificial stream applications for assessing the effects of pulp and paper and mining effluents on aquatic ecosystems as required under Canadian environmental effects monitoring. Summary information includes year, program, effluent type, research objective, location, artificial stream system, and references.

Return to the table

Figure 9-1 is a series of 3 photographs. Image A shows a large mesocosm system with streams situated on tables (Model I) used in the Athabasca River, Alberta. Image B displays a large mobile mesocosm system with streams on 2 trailers (Model II) used in the Fraser River, British Columbia; the Saint John River, New Brunswick; and in Saint John Harbour, New Brunswick. Image C shows a large mobile mesocosm system with streams on a single trailer (Model III) used in the Miramichi and Little rivers, New Brunswick; the Wapiti River, Alberta; and Junction Creek, Ontario.

Return to the figure

Figure 9-2 is a series of two photographs. Image A shows a small microcosm system with streams situated on tables over mixing reservoirs used in the Thompson River, British Columbia. Image B displays a modular mesocosm system with streams situated on tables over mixing reservoirs used in the Little River, New Brunswick; Junction Creek, Ontario; the Wabigoon River, Ontario; and Key Lake, Saskatchewan.

Return to the figure

Figure 9-3 is a schematic representation of a large mesocosm trailer system.

Return to the figure

Figure 9-4 is a photograph of a modular mesocosm set-up.

Return to the figure

Figure 9-5 is a schematic representation of a modular mesocosm flow.

Return to the figure

Figure 9-6 is a photograph of a multitrophic fathead minnow reproductive bioassay. A feeding barrier allows a benthic invertebrate culture to develop under treatment conditions, while controlling access for the fish above it.

Return to the figure

Figure 9-7 is a series of three photographs that illustrate the site set-up for modular mesocosms.

Return to the figure

Figure 9-8 is a conceptual model made up of two images: image A and image B. Image A shows the factorial experimental design to investigate the importance of water vs. diet in responses of Fathead Minnow to metal mine effluent in modular mesocosms. Image B displays the experimental design to investigate the influence of pH and natural organic matter (NOM) on Fathead Minnow responses after exposure to an MME mixture and a single metal in multitrophic modular mesocosms.

Return to the figure

Figure 9-9 is a conceptual model illustrating a factorial experimental design to investigate the effects of MME and historical sediment contamination in isolation and in combination on Fathead Minnow in modular mesocosms.

Return to the figure

Table 9-2 outlines the fish mesocosm study effect indicators, endpoints and related statistical procedures. Primary effect indicators include growth, reproduction, condition, and survival.

Return to the table

Table 9-3 provides the recommended response variables and suitable additional supporting information, and suggested statistical analysis for Fathead Minnow application in modular mesocosm systems. Primary information provided includes the type of response, response variable, dependent variable, independent variable, covariate, single factor statistics, and two factor statistics.

Return to the table

Figure 9-10 is a photograph of a mussel showing a ripe mantle lobe.

Return to the figure

Figure 9-11 shows two graphs illustrating the reproductive cycle of Blue Mussels from British Columbia (graphs A and B). Graph A plots the mantle glycogen concentration (mg/g) as a measure of mantle energy, with relation to the date (February to October). Graph B plots the volumetric fraction of gametes as a measure of mantle reproductive content with relation to the date (February to November).

Return to the figure

Table 9-4 outlines the suggested taxa for use in caged bivalve studies for EEMs. Primary information includes species and reference; temperature range (in Celsius); salinity range (in parts per thousand); reproductive information; and the general distribution in Canada.

Return to the table

Table 9-5 provides the differences noted between two species of mussels over 5 year study in the Burrard Inlet, Vancouver, British Columbia. Primary information includes differences in growth, survival, fecundity, reproduction, egg production related to energy storage, susceptibility to leukemia, and suitability for monitoring.

Return to the table

Table 9-6 displays the differences noted between the unionoidea and sphaeriidae species. Primary information includes differences in growth, life span, fecundity, reproduction, egg production related to energy storage, age at maturiry, and suitability for studies.

Return to the table

Figure 9-12 illustrates shell scars markings of the species mytilus. The markings are identified by numbers. Number 1 shows the length of the anterior adductor muscle scar; number 2 points to the length of the hinge plate; number 3 displays the length of the posterior adductor muscle scar; number 4 shows the distance between the posterior edge of the posterior adductor muscle scar and the posterior shell margin; number 5 displays the distance between the ventral edge of the posterior adductor muscle scar and the ventral shell margin; number 6 shows the shell width; and number 7 shows the shell height.

Return to the figure

Figure 9-13 is a photograph illustrating a duplicate frame from a caged mussels exposure experiment. The PVC frame, the aluminum H frame, and a cluster of mussels in socks are identified in the image.

Return to the figure

Figure 9-14 is a diagram of modular mesocosm parts.

Return to the figure

Figure 9-15 is a schematic representation of a modular mesocosm flow. The image offers both a vertical view and a plane view of the flow.

Return to the figure

Table 9-7 outlines the effect indicators, endpoints, and related statistical procedures of a caged bivalve study. Primary effect indicators include growth, reproduction, condition, and survival.

Return to the table

Figure 9-16 is a series of two photographs. Image A shows a mantle plug and tools necessary for its removal, while image B shows a mantle plug after homogenization that is ready for assessment.

Return to the figure

Table 9-8 provides an example of a field data sheet for recording survival and growth raw data. Information required includes animal number, length, height, width, WAWW, mantle wet weight, animal dry weight, mantle dry weight, and dry shell weight.

Return to the table

Date modified: