2 AReview of the Use of Unionid Mussels as Biological Indicators of Ecosystem Health John H. VanHassel and Jerry L. Farris INTRODUCTION The currently heightened concern for freshwater mussel resourceshas broughtabout acritical need for accurately monitoring the status of theseorganisms. In addition, because of their sensitivity to many environmental perturbations, unionid mussels have been used as indicatorsofboth exposure to and effects of these perturbations. This review provides the historicalbackground of unionid mussel use in various biomonitoring applications and critically evaluates the effectivenessofthese applications. Recommendations for future directionsinmussel biomonitoring and for system-level management practices for the conservation of mussel resourcesare offered. An extensive search of publishedliterature and unpublished reports was made, with acut-off date of May 2003. This literature search was comprehensive, but not exhaustive, and is certainly representative of the work that has been conducted in this field. In all, over 700 publishedarticles, reports,and abstracts were examined for relevance to the scope of this review. In this review,the term“biomonitoring”isexpandedbeyond thenarrowdefinitionof “a continuing collection of data to establishwhether explicitlystatedquality conditions are being met” (Cairns and Smith 1994)toaddress the entire scope of applications of mussel measures that have been, or may be, used to evaluate and protect mussel populations.Previous reviewshave focused on specific aspects of mussel biomonitoring, includingbioaccumulation and toxicity of contaminants (Havlik and Marking 1987; Elder and Collins 1991; Naimo 1995), metal bioaccu- mulationand tissueresidue effects(Salazar 1997),and biomonitoring usingcaged mussels (Smolders et al. 2003).Inthe presentreview, we attempt to linkthe considerable literature dealing with the use of unionid musselsasbiological indicatorsofecosystemhealth. According to Cairns, McCormick, and Niederlehner (1993),the basictypesofbiological indicatorsare: early warning, diagnostic, and complianceindicators. Theeffectivenessofmussel toxicity testing as a regulatory compliance tool is dealt with in Chapter 5, whereas this chapter deals with the other applications: (1) the effectiveness of mussels as sentinels of environmental perturbations (i.e., the effectiveness of measurements applied to mussels for the purposeofdetermining that exposure to an environmental insult has occurred), and (2) the effectivenessofmussels as indicatorsofeco- logical integrity (i.e., the effectiveness of measurements applied to musselsfor the purposeof discriminating effects on spatial and temporal scales),including their use in impact assessment, waterbody status monitoring, and as recorders of environmental history. 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 19 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) REVIEW OF UNIONID MUSSEL BIOMONITORING LITERATURE M ONITORING OF M USSEL P OPULATIONS Collection Techniques During the past twentyyears, many studies have been conducted with field-collected musselsto measuretheir response to anthropogenic impacts. In undertaking these studies, avarietyofcollec- tion techniques have evolved to address the varietyofhabitat conditions encountered, as well as the particular needs of the study. Table 2.1provides asummary of the mostcommonly used methods, themostsuitableapplications foreach, andthe individualbiases inherenttoeachmethod. TABLE 2.1 Applications, Advantages,and Disadvantages of MusselCollection Methods Method Wadable Stream Large River Lake Quan- titative Quali- tative Advan- tages Disad- vantages References Brail XX X1,2,3 1,2,3 1,2,4–7,11,12, 14,16,17,20 Hand pick XX2,4 1,2,5 1–5,7,11,17,18 Dip net XXX2,4 85 Rake w/basket XXX2,4 84,5 Mechanical grab XXXX3,5,6,9 4,7,11,14,17 Skimmer dredge XXXX2,4 5,7,8 2,7,9,11,17,20 Diver transects XXXX42,4,5,8 1,2,4,7–12,14,17 Quadrats XXXXX5 4,5,6 1–8,10–12, 15,17,18 Shoreline shell collection XXXX1,2,3,6 92,7,11,13,17,19 Key to advantages 1Useful as an exploratory device Key to disadvantages 1Does not provide quantitative data 2Inexpensive 2Size selective 3Not time intensive 3Low catch rate 4Provides good qualitative 4Expensive information5Time intensive 5Provides good quantitative information 6Statistical confidence requires large sample size 6Provides supplemental 7Habitat intrusive information8Difficult to obtain quantitative samples 9Data unreliable, qualitativelyand quantitatively References cited: 1. Cawley (1993);2.Dunn (2000);3.Hornbach and Deneka (1996);4.Isom and Gooch (1986);5.Klemm et al. (1990);6.Kovalak, Dennis, and Bates (1986);7.Miller and Nelson (1983);8.Miller and Payne (1993);9.Miller, Whiting, and Wilcox (1989);10. Miller et al. (1993);11. Nelson (1982);12. Payne and Miller (1987a);13. Rothwell (1979); 14. Sickel, Chandler, and Pharris (1983);15. Smith et al. (2000);16. Sparks et al. (1990);17. Strayer and Smith 2003;18. Vaughn, Taylor, and Eberhard (1997);19. Watters (1993–1994); 20. Wilcox, Anderson,and Miller (1993). Freshwater BivalveEcotoxicology20 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) Discussions of each of these methods and their individual strengths and weaknesses can be found in Nelson (1982), Miller and Nelson (1983), and Isom and Gooch(1986). The consensus of several evaluations of the techniques listed in Table 2.1 was that crowfoot brailsare useful, deep-water, exploratory devices, buthavelimitedusefulness in quantitative studies because of their low catch rate and selectivity against juvenile mussels and small species (Kovalak, Dennis, and Bates 1986; Payne and Miller 1987a; Cawley 1993; Wilcox,Anderson, and Miller 1993). Crowfoot brails consist of wood or iron bars that range from two to twentyfeet in length, depending upon the application. Suspended from the bar are several, approximately one- foot, lengths of chainorcord with “crowfoot”hooks attached and with the tips of the hooks beaded to improveretention of captured mussels. The brail is towed very slowlyalongthe bottom and captures mussels when the hook slidesbetween an animal’svalvesthatare slightlyagapefor filtering, eliciting areflexivemusclecontraction thatcausesthe valvestoclampdownonthe hook. Klemm et al. (1990) suggestatleast six 100-meter brail haulsper site for the collection of qualitative data. Handpicking (without the use of quadrats or transects) of musselsisrestricted to shallow watersbut can be an inexpensive method of obtaining qualitative information (Nelson 1982). Restricted visibility during hand picking can bias asample against juvenile or small species of mussels (Cawley 1993). Dip nets and rakes with baskets are alsorestricted to wadablewaters and are another inexpensive way to collect qualitative data. Sampling for quantifiable measures with such devices is difficult, and somewhat habitat limited, as most dip nets and rakes work best in fine- particle substrates with low-debris loads (Miller and Nelson 1983). Transect sampling in deep water (diving)orshallow water (snorkeling) has been used to collectquantitative data in certainappli- cations;however,because of the collector bias inherent to this method (e.g., Payne and Miller 1987a), we recommend that it be used quantitatively only under carefully defined circumstances. Quantitative samples have primarily been collected using mechanical grabs,skimmer dredges, and hand excavation of defined quadrats.Grab samplers, such as Ponar dredges, are generally not recommendedfor mussel studies because of the very high number of grabs required to obtain arepresentative sample and the correspondingly high level of required effort (Sickel, Chandler, and Pharris 1983; Isom and Gooch 1986). Skimmer dredges can offer acost-effective alternative to diving for deep-watersampling (Miller, Whiting, and Wilcox 1989; Wilcox,Anderson, and Miller 1993), but sampling is difficult to quantify, requires considerable equipment outlays,isdestructive to both habitat and mussels, and is effective primarily only on sand-silt substrates (Nelson 1982). For most applications, the method of choice for quantitative sampling is quadrats. Thequadrat size has varied considerably among studies,with the most common areas sampled being 0.25, 0.5, and 1.0m 2 (Salmon andGreen1983; Kovalak,Dennis, andBates1986; Miller andPayne 1988; Holland-Bartels 1990;Klemmetal. 1990;Amyot andDowning 1991;Goudreaueta1. 1993; Harris et al. 1993; Miller 1993). Aquadrat size of 1.0 m 2 or lessisrecommendedbecause of the necessity of replicate samples for quantitative data. Thedepth to which quadrats are excavated also varies but is typically in the range of 5–15 cm (Miller and Payne1988; Holland-Bartels 1990). The numberofreplicates needed to obtainstatistically useful data hasbeen addressed by Cawley (1993),Milleretal. (1993), andStrayer andSmith (2003) formussels in particular,and by Green (1979) for biological sampling in general. Depending upon the needs of specific studies, quadrat sampling programscan require up to 100quadrats per site (Klemm et al. 1990; Harris et al. 1993; Miller 1993). In decidingupon the qualitative versus quantitative needs of aproposed sampling program, insight from past studies can be drawn upon to offer somebasic guidelines.These guidelines, provided in Table 2.2,list measures that have been applied in various mussel investigations, the type of sampling requiredfor these measures, and the strengths and weaknesses of each measure. In general,good qualitative samplesprovide sufficient data for determining species richness, species diversity,evenness, community composition, relative abundance, andpresence/absence of rare species.Onthe otherhand, quantitativesamplingisrecommendedfor themeasurementof density,production, biomass,sizedemography,and recruitment, as well as forhypothesis AReview of the Use of Unionid Mussels as Biological Indicators of Ecosystem Health 21 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) TABLE 2.2 Applications, Advantages,and Disadvantages of Measures Used on Field-Collected Mussels MeasureApplications Data Needs Advantages Disadvantages References Species richness/ diversity/evenness 2,3,4 Qualitative 1,2,3,4 43,4,9,30,37–40,44,53 Relative abundance 2,3,4 Qualitative 1,2,3,4 49,13,22,25,29,37–39,44 Density/biomass 2,3,4 Quantitative 1,2,3,4 43,9,20,27,29,38–40, 43,44,49 Size demography1,2,3,4,5 Quantitative 1,2,3,4 42,9,22,27,37–39,43,44,51 Mortality 1,2,3,4 Quantitative 1,2,3 321,32,43,51 Growth 2,3,4 Nonspecific2(at times), 3,4 55,8,12,13,16,23,24,27,28, 31,33,35,36,40,43,48, 50,51,54 Condition indices 1,2,3,4 Nonspecific1,2,3 (at times) 2(at times), 326,32,33,40,45–47,51 Behavioral responses 1,2,3,4 Nonspecific1,2,3 1,3 1,52 Bacteriological parameters 1,2,3,4 Nonspecific1,2,3 18,52 Physiological para- meters (e.g., enzyme activity, blood chemistry) 1,2,3,4 Nonspecific4 26–8,10–12,14,15,17–19, 21,32–34,36,41,42 Key to applications 1Mussel populationstatus/characterization 2Impact assessment 3Monitoring of spatial/temporal changes 4Hypothesis testing 5Recruitment success Key to advantages 1Ease of use 2Data obtainable in the field 3Nondestructive 4High discriminatory power when quantitative data used Key to disadvantages 1Subjective or poorly defined 2Destructive 3Low discriminatory power 4Requires large, quantitative samples to provide adequate statistical power 5Requires long-term monitoring or labor-intensive thin- sectioning techniques References cited: 1. Amyot and Downing (1991);2.Anderson, Romano, and Pederson (1993);3.Arbuckle and Downing (2002);4.Bailey (1988);5.Beckvar et al. (2000);6.Berg et al. (1995);7.Blaise et al. (2002);8.Blaise et al. (2003);9.Cawley (1993);10. Cherry,Farris, and Belanger (1988);11. Couillard et al. (1995a); 12. Couillard et al. (1995a);13. Davies (1963); 14. Day, Metcalfe, and Batchelor (1990);15. Doyotte et al. (1997);16. Fischer et al. (1993);17. Fleming, Augspurger, and Alderman (1995);18. Gagne et al. (2002);19. Gardner, Miller, and Imlay (1981);20. Goudreau,Neves, and Sheehan (1993); 21. Haag et al. (1993);22. Harris et al. (1993);23. Haukioja and Hakala (1978);24. Hinch and Green (1989);25. Holland Bartels (1990);26. Hornbach et al. (1996);27. Houslet and Layzer (1997);28. Imlay (1982);29. Isom and Gooch (1986); 30. Klemmetal. (1990);31. Layzerand Madison(1997);32. Makela (1995);33. Makela et al. (1992);34. Malley, Huebner, and Donkersloot(1988);35. McCuaig and Green (1983);36. Milam and Farris (1998);37. Miller (1993);38. Miller and Payne (1988);39. Miller and Payne (1993);40. Miller et al. (1993);41. Naimo and Monroe (1999);42. Naimo et al. (1998);43. Negus (1966);44. Payne and Miller (1987a);45. Payne, Miller, and Lei (1995);46. Pekkarinen (1993);47. Roper and Hickey (1994); 48. Rothwell (1979);49. Russell-Hunter and Buckley (1983);50. Scott (1994);51.Sparks and Blodgett (1987);52. Sparks et al. (1990);53. Watters (1992);54. Yokley (1976). Freshwater BivalveEcotoxicology22 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) testing, impact assessment,and monitoring of spatial/temporal changes (Payne and Miller 1987a; Cawley 1993; Miller and Payne 1993; Miller et al. 1993; Vaughn, Taylor, and Eberhard 1997; Dunn 2000; Strayer and Smith 2003). Forsome applications, the success of the sampling program is specifically tied to the type of sampling used, whereasfor other measures, success can be achieved by avarietyofsampling approaches. For example, in determining the presence/absence of rare species,studies have determined that quantitative methods,such as quadrats, are limited because of the excessive number of samples neededtofind rare species.Qualitative methods such as timed searches,oracombined qualitative/quantitative approach, are recommended(Kovalak, Dennis,and Bates 1986; Payne and Miller 1987a; Vaughn, Taylor, and Eberhard 1997; Metcalfe-Smith et al. 2000; Smith et al. 2000, Smith, Villella, and Lemarie 2003; Strayer and Smith 2003). Conversely, most functional measures, such as thoselisted in the bottom half of Table 2.2,can be applied to a variety of qualitatively or quantitatively collected data, depending upon the needs of the study. Structural/Functional Indices Unionidmussels have long been known as important indicatorsofenvironmental perturbations. For example, Ortmann (1909) reported that unionids were the first organisms to disappearfrom streams subjected to pollution stress.However, the use of freshwater musselsfor biomonitoring purposes is amostly recent phenomenon. Although Wurtz (1956) and Ingram (1957) promoted their use as a pollution indicator organism in the 1950s, it was not until the 1970s that musselswere widely used for biomonitoring. Prior to that time, published studies of mussel biomonitoring were relatively few. Early environmental studies involving the use of mussel measurements addressed the effects of temperature and heated effluents(Grier 1920; Davies 1963; Negus 1966), sewage (Baker 1920; Shimek 1935), siltation (Ellis 1936), and impoundment (Bates 1962). Effects measures applied to unionid musselsfulfill mostofthe criteria provided by Widdowsand Donkin (1992): 1. They should be sensitive to environmental levels of pollutants and have alarge scope for response throughout the range from optimal to lethal conditions. 2. They shouldreflectaquantitative and predictablerelationship with toxic contaminants. 3. Theyshouldhavearelatively shortresponsetimesothatpollutionimpactcan be detected in its incipient stages. 4. The technique shouldbeapplicable to both laboratory and field studies to relate labora- tory-based, concentration-response relationships to field measurements of spatial and temporal changesinenvironmental quality. 5. They shouldprovide both an integratedresponse to the total pollutant load and insight into the underlying cause and mechanism of toxicity. 6. The biological response should have ecological relevance (i.e., is related to deleterious effects on growth,reproduction, or survivalofthe individual, population,and community). Several measures have been applied to determinethe status of aparticularmussel population, or to initially characterize apopulation. Of these, the mostuseful are size demography (i.e., analysis of age-size classes of apopulation)and physiological parameters.Moststructural measures (e.g., density, mortality, andcondition indices) tend to be insensitivetosmall or moderate changeswithin apopulationand arebiased by habitatinfluences. Size demography analyses provide species-specific information indicative of factorsaffecting specific segments of the popu- lation, such as removal of olderindividuals by commercial harvest, or lack of recruitment causedby environmental stress (Sparks and Blodgett 1987; Anderson, Romano, and Pederson 1993; Miller 1993; Houslet and Layzer 1997). Studies of mussel recruitment success require alarge, quantitative sizedemographydatabase (Millerand Payne1988). Anderson, Romano, andPederson(1993) demonstrated the use of annuli-length regression data to characterize mussel reproductive activity. AReview of the Use of Unionid Mussels as Biological Indicators of Ecosystem Health 23 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) Physiological measures provide ameans of documentation of mussel condition at the time of sample collection. Although such measures are notwelldeveloped for use on unionids, afew investigators have successfully usedthemand established their attractiveness forfuture appli- cationsrequiring increased sensitivityand rapidity overmoretraditionalmeasures(growth). Haag et al. (1993) applied avariety of physiological measures to astudy of zebramussel ( Dreissena polymorpha )fouling effectsonunionidsand determined that cellulolytic enzyme activity and glycogen content were very sensitive measures of mussel stress.Malley, Huebner, and Donkersloot (1988) detectedchanges in mussel blood-ioncomposition caused by alum addition to alake. Gardner,Miller, and Imlay (1981) found that tissue amino acidlevels were elevated in mussels from acid mine streams, whereas Day,Metcalfe,and Batchelor(1990) measured asimilar response to agricultural and urban runoff. Aresponse in gill antioxidant enzyme activity to acokingeffluent was documented by Doyotte et al. (1997).Fleming, Augspurger, and Alderman (1995) detected reducedcholinesteraseactivityinmussels at anddownstreamofasuspected pesticide-related mussel kill. Anumber of related studies of municipal effluentsdemonstrated measurableeffects on several enzymesystemsand othercellular and subcellular measures (Blaise et al. 2002, 2003; Gagneetal. 2002).Production of metallothionein in musselshas been related to gradients in cadmiumconcentration but not to concentrations of copper or zinc, possibly because of homeo- static control of the latter two elements (Couillard, Campbell, and Tessier 1993; Couillard et al. 1999). The samegroup of researchersproposed that metallothioneinaloneisinsufficient as an effects biomarker but has value in combination with information on metal partitioningtoother cytosolic ligands (termed “spillover”) (Giguere et al. 1999). Further researchonthese physiological measures is neededtoaddress the issues of reversibility of effects and links to long-term impli- cations for the health of individual mussels. All of the measures listed in Table 2.2 have been applied to impact assessment studies, moni- toring of spatial/temporal changes, and field hypothesis testing. Of these measures, density and growth have been themostsensitive andfrequently used measures. Davies (1963)found that mussels within athermal discharge attained greater maximum size than musselsoutside of the dischargeinfluence. Both Negus (1966) andRothwell (1979)measuredmorerapid growth in mussels residing in aheated effluent than in musselsoutside of the effluent influence. Yokley (1976) foundthatmussel growth rateswere much greater upstream of dredging compared to immediatelydownstream. Astudy of an industrial discharge by AEPSC/APCO (1987) found that mussels upstreamofthe discharge achieved greater length gains than did mussels within the discharge influence. Imlay (1982) also found decreasedgrowth rates in field-collected mussels under stress.Hinch and Green (1989) found significant differences in growth between reciprocally transplantedmusselsamong environmentally variable sites. Fischer et al. (1993)discriminated longitudinal differences in growth rates within species (see also Scott 1994). Goudreau, Neves, andSheehan (1993) used densitymeasurementstoshow depletion of musselsdownstreamof sewage treatment plants compared to upstream populations.Cherry, Farris, and Belanger (1988) found significantly depressed cellulolytic enzymeactivity in mussels caged for 30 days below an industrial effluent compared to thoseinupstream cages.Growth was not significantly affected over the 30-day period; however, studies with Corbicula fluminea have demonstrated that reductions of cellulolytic enzyme activity are associated with eventual reductions in growth rate (Farris et al. 1989). Similarly, Couillard et al. (1995b)measured reduced growth along with elevated metal- lothionein and cellular toxicity in mussels transplanted to acadmium-contaminated lake. Houslet and Layzer (1997) measured reduced growth,recruitment, and density in mussels downstream of a strip-mined area compared to upstream specimens. Although growth and density measurements have been successfully applied to quantitative field studies,bothmeasureshavedrawbacks(Roperand Hickey 1994;Layzerand Madison 1997). Becauseofthe patchy distribution of mussels, the ability to discriminatedifferences in density requiresvery large, quantitative sample sizes. Because of the slow growth of unionid mussels, direct measurement of growth in situ requireslong-term monitoring and, usually, caging or tagging Freshwater BivalveEcotoxicology24 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) of the mussels. Both caging and tagging have been used successfully in growth studies,and a standardized protocol for in situ growth studies has been developed (ASTM 2001). AEPSC/APCO (1987) used wire cages to hold mussels in situ over afive-monthperiod. Yokley (1976) used a tagging method in aone-year study. Several other studies have employedmarking of the shells for repeat-measurement studies (Sparks and Blodgett 1987; Hinch and Green 1989; Miller et al. 1993), and Miller and Nelson (1983) include adiscussion of how to mark shells.Walleretal. (1993a) describeagrid device that can be anchored to the substrate to aid in the recovery of musselsunder study. Growth rates can also be determined using measurements of annuliorlength-frequency histograms. Both techniques require large sample sizes, and the useoflength-frequencydata can be very problematic for any but the mostabundant species.For studies examining annulimeasure- ments, thin sectioning to allow the use of internal bands is highly recommended(Neves and Moyer 1988; Metcalfe-Smith and Green 1992). McCuaig and Green (1983) discuss the use of growth rate parameters in quantitativemusselstudies. Furtherdevelopment of physiological measuresfor mussels, such as various enzymeassays (e.g., cellulase, metallothionein, and cholinesterase) and tissue glycogen, is needed to provide rapid and less labor-intensive methods of quantitative field assessment, to supplement traditional measures andprovide awider array of toolsfor ecotoxicological investigations. Contaminant BodyBurdens Much of the monitoring that hasbeen undertaken with freshwater unionid musselshas involved the measurement of contaminant body burdens for various applications. Mussels have been found to be auseful organism for this purpose, and they possess several attributes that are characteristic of useful biomonitors (Phillips 1980; Green, Singh, and Bailey 1985): 1. Most species inhabitshallow, nearshore areasinlentic waters or riffle areas in lotic waters,which arethe most productiveand themostsusceptible to most types of pollution. 2. They are sedentary. 3. They are long-lived. 4. They are often quite abundant. 5. They are large enough to easily provide adequate tissue for contaminant analyses. 6. They are easy to sample. 7. Many species are very sensitive to avarietyofcontaminants, however,there are also tolerant species that can accumulate contaminants under fairly severe perturbations. 8. They have been demonstrated to accumulate avarietyofcontaminants to concentrations that correlate well with exposure concentrations and durations. 9. They provide arecord of exposure historyintheir shells. Most of the applications that have been developed for monitoring contaminantbody burdens in freshwater musselshave been based on the large body of work established using marinebivalves. Mix(1984) providedanextensive review of themarinebivalve biomonitoringliterature. The marine“mussel watch”program employsanextensivemonitoringsystemthathas provided useful biomonitoring data for manyyearstoassess the level of contamination in marine coastal areas (Goldberg 1975). Metcalfe-Smith (1994) stated that this program has been successful because of sound samplingand analyticalprotocols. Similar protocols are now beingapplied to freshwater mussel investigations. Table2.3 summarizes biomonitoring applications that have measured freshwater mussel contaminantbodyburdens.Avariety of sampling andexposureprotocols have been used, dependinguponthe purposeofthe study.Analyticaltechniqueshavevariedaswell, anda AReview of the Use of Unionid Mussels as Biological Indicators of Ecosystem Health 25 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) TABLE 2.3 ApplicationsUsing MusselCont aminant Body Burdens, andAdvantages of Using Mussels for Those Applications Applications Component Analyzed Advantages Disadvantages References Impact assessment 1,2,3 2,3,4,6 1,2 1,5,6,11,14,16, 18–22,24–26,28,29,35, 39,40,47,54,60,61 Pollution source identification 1,2,3 1,2,3,4,5,6,7 24,5,12,23,24,27,28,30–34,43,46–52, 54,56,57,60,63 Sentinel organism 1,2 1,3,4,5,6,7 2,3 2,3,6,13,14,23,24,36,38,42,44, 46–50,52–54,60,62,66,70,71 Bioavailability measure 1,2,3 1,2,6 25,7–9,12–15,17,22,25,26,29,33,34, 39–42,44–46,48,53,55,58,59,61, 65–69,71 Environmental history 32,4,5,6,7 2,4 10,23,24,37,64 Key to componentanalyzed1Whole body; 2Specific tissues/organs; 3Shell/shell layers Key to advantages 1Cost-effective, quantitative measure for this application 2Good exploratory tool for this application 3Mussels are often sensitive indicator organisms 4Exposures are integrated over time 5Mussels are typically long-lived 6Mussels are sedentary and amenable to caging 7Mussel shell layers offer potential to date results Key to disadvantages 1Does not provide aquantitative measure of impact 2Body burdens and uptake rate influenced by mussel size, age, growth rate, type of contaminant,long tissue retentiontimes for many contaminants, source water (for transplants), and environmental variations 3Selected species needs to be pollution-tolerant in some situations 4Analyticalrestraints and other confounding influences on results References cited: 1. Adams, Atchison,and Vetter (1981);2.Anderson (1977);3.Balogh (1988);4.Becker et al. (1992); 5. Beckvar et al. (2000);6.Bedford, Roelofs, and Zabik (1968);7.Brungs (1967);8.Campbell and Evans (1987);9.Campbell and Evans (1991);10. Carell et al. (1987);11. Cherry, Farris, and Belanger (1988);12. Couillard, Campbell, and Tessier (1993);13. Couillardetal. (1995a); 14.Couillard et al.(1995b);15. Couillardetal. (1999);16. Czarnezki(1987); 17.Dobrowolskiand Skowronska (2002);18. Foster and Bates (1978);19. Gaglione and Ravera (1964);20. Gagne et al. (2002);21. Gardner and Skulberg (1965);22. Giguere et al. (1999);23. Green et al. (1989);24. Green, Singh, and Bailey (1985);25. Hayer and Pihan (1996);26. Hayer, Wagner, and Pihan (1996);27. Herve (1991);28. Herve et al. (1988);29. Herve et al. (1991);30. Hinch and Green (1989);31. Hinch and Stephenson (1987);32. Hinch, Bailey, and Green (1986);33. Kauss and Hamdy (1985);34. Kauss and Hamdy (1991);35. Klose and Potera (1984);36. Leard, Grantham, and Pessoney (1980); 37.Lingard, Evans, and Bourgoin (1992);38. Lopes et al. (1993); 39. Makela (1995);40. Makela, Lindstrom-Seppa, and Oikari (1992);41. Malley, Chang,and Hesslein(1989); 42. Malley, Stewart,and Hall(1996); 43.Manly and George(1977); 44.Marquenie (1985);45. Mathis andCummings (1973);46. Metcalfe andCharlton(1990); 47.Metcalfe andHayton (1989);48. Metcalfe-Smith (1994);49. Muncasteretal. (1989);50. Muncaster, Hebert,and Lazar(1990); 51.Naimo (1993);52. Pellinen et al. (1993);53. Perceval et al. (2002);54. Phillips (1980);55. Price and Knight (1978);56. Pugsley, Hebert, and McQuarrie (1988);57. Pugsley et al. (1985);58. Renaud et al. (1995);59. Renzoni and Bacci (1976);60. Salanki (1989);61. Schmitt et al. (1987);62. Servos et al. (1987);63. Smith, Green, and Lutz (1975);64. Sterrett and Saville (1974); 65. Stewart(1999);66. Storey and Edward (1989);67. Tessier et al. (1984);68. Tessier et al. (1993);69. Tevesz et al. (1989);70. Turick, Sexstone, and Bissonnette (1988);71. Wang et al. (1999). Freshwater BivalveEcotoxicology26 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) review of these will not be undertaken here;however, several good, standardized protocols exist (e.g., ASTM 1999). Formost applications, the analysisofthe whole tissue mass of the mussel is sufficient. Analysis of specific tissuesisappropriatefor certain typesofuptake studies,but is not recommended for routine applications because of the additional time necessary to separate the desired tissuesand the potential for analytical limitations related to the analysis of small amounts of tissue.Similarly, analysisofshell materialisnot recommended for mostapplications because of difficulties in digestion and analytical interferences. Several studies have used freshwater mussels as ameans of pollution sourceidentification and impact assessment. Mussels have been found to be particularly useful in the former application. Studies of body burdens in musselsdownstream of contaminantdischarge sources compared to uninfluenced areashave clearly shownelevated concentrations related to the discharge of both metals (Foster and Bates 1978; Czarnezki 1987; Schmitt et al. 1987; Cherry, Farris, and Belanger 1988; Salanki 1989; Couillard et al. 1995b; Makela 1995; Giguereetal. 1999; Beckvar et al. 2000; Gagne et al. 2002)and organics (Kauss and Hamdy 1985; Herve et al. 1988; Metcalfe and Hayton 1989; Metcalfe and Charlton 1990; Herveetal. 1991; Makela 1995; Hayer and Pihan 1996; Hayer, Wagner, and Pihan 1996). Mussels have also been avaluable tool for contaminant mapping of a waterbody for both metals (Adams,Atchison, and Vetter 1981; Pugsley, Hebert, and McQuarrie 1988; Becker et al. 1992; Naimo1993; Beckvar et al. 2000)and organics (Pugsley et al. 1985; Herve et al. 1988; Herve 1991; Kauss and Hamdy 1991). Although contaminant body burdens in musselseffectively reflectexposure to avarietyofpollutants, they have not been shown to quan- titativelymeasure impact. In order to demonstrate impact, body burdens of aparticularcontaminant would have to be unambiguously linked to significant impairment of some aspect of the organism’s life cycle (Stewartand Malley1997). This has not been done for freshwater mussels beyond simple association of tissue contaminant levels with effects on growth, etc. (Couillard et al. 1995a; Makela 1995; Beckvar et al. 2000; Gagne et al. 2002; Perceval et al. 2002). The same qualitiesthat makemussels useful for pollution sourceidentification also make them good sentinel organisms (Table 2.3). Mussels have been successfully used to monitor waterbodies for avariety of contaminants, including metals, organics, and fecal coliform bacteria (Bedford, Roelofs, and Zabik 1968; Leard,Grantham,and Pessoney 1980; Turick, Sexstone, and Bissonnette 1988; Muncaster et al. 1989; Salanki 1989; Lopes et al. 1992). However, alevelofdiscrimination is necessarytoensure that the most applicable monitor is well suited for the suspectedcontaminant of impact. Such route-specific uptake or effect is not always apparent with mussels and may, therefore, require the use of othermonitored organisms to demonstrate system impact. In comparative studies, Metcalfe and Hayton (1989) found leeches to be better than mussels for monitoring chlorophenols, and Balogh (1988) found that zooplankton were better biomonitors than mussels for metals from a sewage treatment plant. Renaud et al. (1995)attributed differential organochlorineuptakeby musselscomparedtolampreys to differencesinlipid composition betweenthe species. Areview by Elder and Collins (1991) stated that metal accumulation in mollusks is typicallyhigher than in fish. Studies of the bioavailability of various contaminants to freshwater mussels have shown that uptake of metalsislikely related to metal concentrations that are easily extractablefrom sediments or that exist in ionic form, and are not necessarilycorrelated to total water or sediment concen- trations (Tessier et al. 1984, 1993; Marquenie1985; Czarnezki 1987; Schmitt et al. 1987; Pugsley, Hebert, and McQuarrie 1988; Koenig and Metcalfe 1990; Campbell and Evans 1991; Elder and Collins 1991; Kauss and Hamdy1991; Couillard, Campbell, and Tessier 1993; Metcalfe-Smith 1994; Perceval et al. 2002). These studies have also helped to define thosefactorsthat tend to confoundthe interpretation of mussel bodyburdendata.Metals have been shown to increase, decrease, or nottovary with thesize of theanimal, depending upon themetal (Manly and George1977; Priceand Knight1978; Hinch andStephenson 1987;Greenetal. 1989; Naimo 1993; Balogh and Mastala 1994). Jeffree et al. (1993) eliminated the size-related variabilityin individual mussel metalconcentrations by regression againsttissuecalciumconcentrations. AReview of the Use of Unionid Mussels as Biological Indicators of Ecosystem Health 27 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) Organochlorinebody burdenshavealsobeenshowntoberelated to mussel size (Muncaster, Hebert, and Lazar 1990). Other biological factors that have been shown to influence contami- nant-bodyburdens include age, reproductivestatus, growth rate,tissuetype,species, lipid content,and the valveclosure response (Brungs1967; Smith, Green, andLutz1975; Renzoni and Bacci 1976; Foster and Bates 1978; Phillips 1980; Hinch and Stephenson 1987; Green et al. 1989; Tevesz et al. 1989; Metcalfe-Smith and Green 1992; Balogh and Mastala 1994; Renaud et al. 1995; Stewartand Malley 1997). Environmentalfactorsthat have been shown to influence contami- nant-body burdens include pH, temperature, humic matter, calcium, organic and inorganiccarbon, season, and interactions with other contaminants (Phillips 1980; Tessier et al. 1984, 1993; Camp- bell andEvans 1987, 1991; Hanna 1992;Metcalfe-Smith1994; Naimo1995; Stewart1999; Perceval et al. 2002). The sex of amussel has been shown to have relatively little influence on body burdens of both metals(Metcalfe-Smith1994)and organochlorines (Muncaster, Hebert, and Lazar 1990). Many applied studies of mussel contaminant body burdens have successfully used transplanted and/or caged mussels(see Chapter9). Transplantingisuseful forthe purposeofintroducing mussels for biomonitoring of areas where residentmussels do not exist or existinnumbers that are too low for effective sampling, and for the purposeofreducing experimental error by using mussels from aknown source or sources. Several biomonitoring studies have successfully used transplanted mussels (Schmitt et al. 1987; Cherry, Farris, and Belanger 1988; Hinch and Green 1989; Salanki 1989; Couillard et al. 1995a; 1995b; Makela 1995; Beckvar et al. 2000). The use of transplanted mussels can be advantageous because of the unknown exposure history of resident mussels and the long retention time for some contaminants (Marquenie1985); however,aconcern involving the use of transplants has been raised by Hinch and Green (1989) and Salanki (1989),who found that metal uptake by transplanted mussels is influenced by their source. Areciprocal trans- plantstudy by Englundand Heino (1996) found cage and source-relatedeffects on mussel valve movement measures to be minimal.Investigators should be aware of the potential for source-related effects when planning atransplant study. In particular, the selection of source locations, typesand frequencyofmeasurements, number of replicate measurements, and statistical analysis of resulting data all need to account for potential bias due to source. Many investigators alsochoose to cage mussels during biomonitoring, in order to (1) increase their chances of retrieval, (2) ensure originalmussels are collectedfor analysis, and (3) protect the mussels from predationand other outside influences. Disadvantages of cages include the potential for vandalism and the potential for fouling of the cages by debris. Many of the studies referenced in Table 2.3 used caged mussels, with exposure periodsranging from two days to one year. Muncaster, Hebert, and Lazar (1990) reported that cages,corrals, and leashes were all equally effective for biomonitoring studies. Salazar et al. (2002) developed acage for use in long-term in situ studies, and an ASTM (2001) method is available providing the appropriate methodology. Astandardized exposure period may not be advisable because of the demonstrated tendency of contaminants to fluctuate over time (Muncaster et al. 1989). Rather, exposure time should be determined by the study objectives, characteristics of the contaminantbeing measured, and the environmental conditions at the time of the study. Salazar (1997) recommended an exposure period of 60–90 days based on the time necessary to reach chemical equilibrium for most metals and hydrophobic chemicals. Freshwater mussels have received limited use to date as recorders of environmental history. Lingard, Evans, and Bourgoin(1992) detailed techniques for measuring metal concentrations in the shellnacre. It is recognizedthatmusselsrepresentapotentiallyvaluable toolinthisregard, especially in fluvial systemsand other areas where sediments do not often provide achronological record of contaminantconcentrations. Promising results based on the separation and elemental analysisofmussel shell layers as ahistoric record have been obtained (Sterrett and Saville1974; Carell et al. 1987, 1995), but further work is neededtoexplore the relationship betweencontami- nant levels in the environment and levels in shells (Stewartand Malley1997). Freshwater BivalveEcotoxicology28 4284X—CHAPTER 2—17/10/2006—11:41—JEBA—XML MODEL C–pp. 19–49 © 2007 by the Society of Environmental Toxicology and Chemistry (SETAC) [...]... 3,6,8,17 ,22 ,24 26 ,29 –31,36,37,47,49,50, 58, 62, 66,67,71–73,75–77,80,84 Adult 1 ,2, 3,4 5 2, 3,15,54,67 Juvenile 56 Adult 1,4 3,4 10 Adult 1,3 2, 4 10,13,14,16 ,27 ,28 ,54,65 Adult 1 4,5 18 ,27 29 ,48,59–61,69 Adult 4 2, 4,5 1 ,2, 9 ,27 ,55,57,64, 72 Adult 2, 4 (shell) 1,3 2, 48,57 Adult 1,3,4 5 39,46,74 Adult 2, 4,5 1,7,18 ,21 ,27 ,29 ,46,55,63,65,68,70 1 Sensitive Key to disadvantages 1 Insensitive 2 Ease of use 2 High variability... Methods Used for Freshwater Mussels Life Stage Adult Test Length References Artificial stream Beaker, culture dish Static-renewal tanks Static or static-renewal Flow-thru or artificial stream In situ glass vial Static or static-renewal 30 days 25 min–6 days 12 20 days 1–10 days 10–37 days 30,36 12, 20 ,21 ,24 ,34,36,37,40,43,80, 82 30,36 5,13,33,36,40–43,51, 52, 77 2, 36,55,56 7 days 18 h–7 months Flow-thru tanks... 5 5 1 ,2 5 5 2, 5 Transformation success Encystment success Snap rate Bioaccumulation Siphoning activity Growth Enzyme activity Blood ion concentration Metabolic rates Shell/condition index Ciliary activity Other physiological Key to advantages References 11,19 ,20 ,23 , 32, 34,35,38,43,81,83 4, 12, 34,38,40, 42 44,51, 52, 56,78, 82 5,10,11,16 ,21 ,33,43,45,48,53,54, 72, 79,80 34,35,38,41 34,35 11 ,20 ,34,35 83 2 4 3,6,8,17 ,22 ,24 26 ,29 –31,36,37,47,49,50,... months 44,81 1,3,5,7–10,15,18,19 ,22 ,23 ,29 ,30,43,45–47, 50,53,57–59,61–64,68–76,78,79,83 4,6,14 ,25 28 ,31, 32, 35,38,48,54,57,60,65–67 11, 12, 16,17,39 References cited: 1 Aldridge, Payne, and Miller (1987); 2 Allran et al (20 02) ; 3 Baker and Hornbach (1997); 4 Balogh and Salanki (1984); 5 Barfield, Clem, and Farris (1997); 6 Bartsch et al (20 00); 7 Birdsall, Kukor, and Cheney (20 01); 8 Black et al (1996); 9... al (20 03); 82 Warren, Klaine, and Finley (1995); 83 Weinstein (20 01); 84 Winter (1996) © 20 07 by the Society of Environmental Toxicology and Chemistry (SETAC) 428 4X CHAPTER 2 17/10 /20 06—11:41—JEBA—XML MODEL C – pp 19–49 32 Freshwater Bivalve Ecotoxicology Adult unionid mussels were highly resistant in short-term exposures to metals (Cherry, Farris, and Neves 1991), acidic pH (Makela and Oikari 19 92) ,... conducting in-situ field bioassays with caged bivalves, E 21 2 2- 0 2, 20 03 Annual Book of ASTM Standards, Vol 11.05, ASTM, Philadelphia, PA, 20 01 Bailey, R C., Correlations between species richness and exposure: Freshwater mollusks and macrophytes, Hydrobiologia, 1 62, 183–191, 1988 Baker, F C., The effects of sewage and other pollution on animal life of rivers and streams, Trans Ill State Acad Sci., 13, 27 1 27 9,... of a freshwater mussel, Parreysia rugosa (Gmelin), J Ecotoxicol Environ Monit., 1, 23 0 23 3, 1991 Renaud, C B., Kaiser, K L E., Comba, M E., and Metcalfe-Smith, J L., Comparison between lamprey ammocoetes and bivalve molluscs as biomonitors of organochlorine contaminants, Can J Fish Aquat Sci., 52, 27 6 28 2, 1995 © 20 07 by the Society of Environmental Toxicology and Chemistry (SETAC) 428 4X CHAPTER 2 17/10 /20 06—11:41—JEBA—XML... 160–169, 20 02 Blaise, C., Gagne, F., Salazar, M., Salazar, S., Trottier, S., and Hansen, P.-D., Experimentally-induced feminisation of freshwater mussels after long term exposure to a municipal effluent, Fresenius Environ Bull., 12( 8), 865–870, 20 03 Brungs, W A., Distribution of Cobalt 60, Zinc 65, Strontium 85, and Cesium 137 in a Freshwater Pond, U.S Public Health Service, 999-RH -2 4 , Washington, DC, p 52, ... (20 03); 81 Warren, Klaine, and Finley (1995); 82 Weinstein (20 01); 83 Winter (1996) © 20 07 by the Society of Environmental Toxicology and Chemistry (SETAC) 428 4X CHAPTER 2 17/10 /20 06—11:41—JEBA—XML MODEL C – pp 19–49 30 Freshwater Bivalve Ecotoxicology the focus of their research This trend is readily apparent when examining the range of test lengths listed in Table 2. 4 There is little doubt that freshwater. .. cage for long-term, in-situ tests with freshwater and marine bivalves, In 29 th Annual Aquatic Toxicity Workshop Proceedings, Oct 20 23 , Whistler, British Columbia, Canada, 20 02, Canadian Technical Report of Fisheries and Aquatic Sciences 24 38, pp 34– 42 Salmon, A and Green, R H., Environmental determinants of unionid clam distribution in the Middle Thames River, Ontario, Can J Zool., 61, 8 32 838, 1983 . References Impact assessment 1 ,2, 3 2, 3,4,6 1 ,2 1,5,6,11,14,16, 18 22 ,24 26 ,28 ,29 ,35, 39,40,47,54,60,61 Pollution source identification 1 ,2, 3 1 ,2, 3,4,5,6,7 24 ,5, 12, 23 ,24 ,27 ,28 ,30–34,43,46– 52, 54,56,57,60,63 Sentinel organism 1 ,2. 24 ,5, 12, 23 ,24 ,27 ,28 ,30–34,43,46– 52, 54,56,57,60,63 Sentinel organism 1 ,2 1,3,4,5,6,7 2, 3 2, 3,6,13,14 ,23 ,24 ,36,38, 42, 44, 46–50, 52 54,60, 62, 66,70,71 Bioavailability measure 1 ,2, 3 1 ,2, 6 25 ,7–9, 12 15,17 ,22 ,25 ,26 ,29 ,33,34, 39– 42, 44–46,48,53,55,58,59,61, 65–69,71 Environmental history 32, 4,5,6,7. 1 ,2, 3,4 49,13 ,22 ,25 ,29 ,37–39,44 Density/biomass 2, 3,4 Quantitative 1 ,2, 3,4 43,9 ,20 ,27 ,29 ,38–40, 43,44,49 Size demography1 ,2, 3,4,5 Quantitative 1 ,2, 3,4 42, 9 ,22 ,27 ,37–39,43,44,51 Mortality 1 ,2, 3,4