Contaminants in Unionid Mussels from the Confluence of the Mississippi and Illinois Rivers Joan Esarey, David J. Soucek, Jeffrey M. Levengood, Robert J. Hudson, Wade Wimer, Richard S. Halbrook Illinois Natural History Survey Bulletin Volume 38, Article 5 November 2008 Institute for Natural Resource Sustainability, William Shilts, Executive Director Illinois Natural History Survey Brian D. Anderson, Director I-Building 1816 South Oak Street Champaign, Illinois 61820 217-333-6880 Citation: Esarey, J., D.J. Soucek, J.M. Levengood, R.J. Robertson, W. Wimer, and R.S. Halbrook. 2008. Contaminants in unionid mussels from the confluence of the Mississippi and Illinois rivers. Illinois Natural History Survey Bulletin 38(5):197–214. Editor: Charles Warwick US ISSN 0073-4918 US ISBN 1-882932-19-6 Printed by authority of the University of Illinois © 2008 P0217729—.75M—11-08 Printed with soy ink on recycled and recyclable paper. The University of Illinois will not engage in discrimination or harassment against any person because of race, color, religion, national origin, ancestry, age, marital status, disability, sexual orientation including gender identity, unfavorable discharge from the military or status as a protected veteran and will comply with all federal and state nondiscrimination, equal opportunity and affirmative action laws, orders and regulations. This nondiscrimination policy applies to admissions, employment, access to and treatment in University programs and activities. University complaint and grievance procedures provide employees and students with the means for the resolution of complaints that allege a violation of this Statement. Inquiries or complaints may be addressed to the Director and Assistant Chancellor, Office of Equal Opportunity and Access, 601 East John Street, Swanlund Administra- tion Building, (217) 333-0885, fax (217) 244-9136, TTY (217) 244-9850 or the Associate Provost and Director, Academic Human Resources, Henry Administration Building, (217) 333-6747, fax (217) 244-5584. For other University of Illinois information, contact University Directory Assistance at 333-1000. Contaminants in Unionid Mussels from the Confluence of the Mississippi and Illinois Rivers Joan Esarey1, David J. Soucek1, Jeffrey M. Levengood1, Robert J. Hudson2, Wade Wimer2, Richard S. Halbrook3 1. Illinois Natural History Survey 2. University of Illinois 3. Southern Illinois University Correspondence to: Jeffrey M. Levengood Division of Ecology and Conservation Science Illinois Natural History Survey 1816 S. Oak St. Champaign, IL 61820 271-333-6767 jleven@inhs.uiuc.edu Contents Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199 Mussels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199 Water and Sediment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201 Statistical Comparisons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .202 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206 Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 ii ACknowledgments Special thanks to the following: John Tucker and Charlie Warwick, Illinois Natural History Survey; Loretta Skowron, Mary LeFaivre, and Daniel Webb, Illinois State Water Survey; Danny Brown and San- dra Rackers, Missouri Department of Conservation; Richard Sparks, National Great Rivers Research and Education Center (NGRREC); Jon Talbott, Illinois Sustainable Technology Center; Rudiger Laufhutte and Marie Keel, University of Illinois at Urbana-Champaign Microanalytical Laboratory; Gary Dreher, Illinois State Geological Survey. This project was funded by the NGRREC. Contaminants in Unionid Mussels from the the Mississippi and Illinois RiversNovember 2008 197 ABSTRACT Unionid mussels were collected from three mussel beds near the confluence of the Mississippi and Illinois rivers in 2003 to evaluate concentrations of selected elements and organic compounds in three abundant species and to preliminarily investigate the relative contribution of these waterways to observed contaminant burdens. Copper (Cu), selenium (Se), and zinc (Zn) concentrations were higher and lead (Pb) concentrations were lower in Amblema plicata collected downstream of the confluence than in those collected upstream. Mean concentrations of nickel (Ni), total mercury (Hg), methylmercury (MeHg), Pb, and Zn varied by species. Concentrations of cadmium (Cd) decreased with age in A. plicata from two of three sites. Tissue concentrations of some elements, e.g., arsenic (As), Cd, Cu, Pb, Se, and Zn, were similar to or higher than those previously reported for unionid mussels from areas of contaminated sediment. Concentrations of Cd, Cu, and Zn in A. plicata were comparable to those collected from the Mississippi River approximately 450 and 900 km upstream from our study sites (Naimo et al. 1992). Although total Hg concentrations we observed were an order of magnitude lower than in that study, MeHg concentrations were above those associated with reductions in soft tissue mass in a study of Elliptio complanata (Salazar et al. 1995). A number of polychlorinated biphenyl (PCB) congeners were detected in A. plicata tissues, with 85% of detections occurring in mussels from downstream of the confluence. Con- centrations of individual PCB congeners were ≤33 ng/g ww and the maximum summed PCB congener concentration was 100.2 ng/g ww. Although few persistent pesticides were detected, β-hexachlorocyclohexane (HCH) was detected in each of the species collected from below the confluence of the two rivers, and in A. plicata collected above it on both the Mississippi and Il- linois rivers, at a maximum concentration of 103.5 ng/g ww. Aldrin, δ-HCH and dichlorodiphe- nyltrichloroethane (DDT) were detected in few of the specimens collected. The findings of this preliminary investigation suggest that unionid mussels from near the confluence of the Mississippi and Illinois rivers may be at risk of negative health effects of elevated exposure to certain environ- mental contaminants. Studies examining the health and productivity of unionid mussels from this area appear warranted. INTRODUCTION Freshwater mussel populations in the United States and Canada have declined dramati- cally over the past few decades. Two hundred thirteen of 297 known species are considered endangered, threatened, or of special concern (Williams et al. 1993, Cummings and Mayer 1992). Mussels are sensitive to environmen- tal perturbations and are often among the first organisms to become extirpated when an aquatic system is degraded. A variety of factors, including over-harvest, siltation, and pollution from coal-mining and other activities, have been linked to their decline (Ahlstedt and Tuberville 1997, Cummings and Mayer 1992, Bogan 1993, Lydeard et al. 2004). In the state of Illinois, where 79 species in the families Margaritiferidae and Unionidae (order Union- oida) may have historically occurred, 15 spe- cies have been extirpated, 19 are listed as state endangered, and 7 are considered threatened (Cummings and Mayer 1992). Bivalve mollusks have been used widely as indicators of environmental stress because they are sedentary and accumulate contami- nants but are not efficient in metabolizing chemicals compared to some other organisms (ASTM 2001). While filter feeding, they ingest contaminants that are in solution and sorbed to fine particulate organic matter, and they are ex- posed to sediment-bound contaminants through pedal feeding (Moore 1971, Cope et al. 2008). Bivalves are keystone species in some aquatic environments, often constituting a large portion of the biomass, and can influence numerous ecosystem functions (Vaughn and Hakenkamp 2001). In addition, they usually provide suf- ficient tissue for chemical analysis and exhibit measurable sublethal effects such as reduced growth, tissue condition, glycogen levels, shell length, DNA strand breakage, cellulolytic activ- Vol. 38 Art. 5198 Illinois Natural History Survey Bulletin ity, inhibited ciliary movement on the gills, and inhibited valve movement (ASTM 2001, Allen et al. 1996, Belanger et al. 1986, 1990, Black 1997, Doherty 1990, Farris et al. 1989, Naimo et al. 1998, Newton et al. 2001, Smith and Beauchamp 2000). Furthermore, immature stages are among the most sensitive of animals in acute laboratory bioassays (Cherry et al. 2002, Jacobson et al. 1993, 1997). The Mississippi River (MSR) basin covers a large portion of the central continental United States, which includes many cities and towns, industrial and mining areas, and a great expanse of land devoted to agriculture. As a result, the larger rivers composing this watershed carry considerable contaminant loads. Fish collected in the watershed often have elevated concen- trations of bioaccumulative contaminants and at many sites are in poor health (Schmitt et al. 2002). One of the largest lead mining areas in the world is located on the upper MSR, and Cd, chromium (Cr), Cu, Pb, and Hg are used extensively by industries along the river (Gar- barino et al. 1995). The heavy use of fertilizers can result in the introduction of large amounts of metals onto the landscape, part of which may be carried into the river by run-off. Addition- ally, sewage treatment plant effluents may introduce large quantities of metals such as Cd, Hg, and Pb into riverine systems. Consequent- ly, heavy metal concentrations in suspended or bed sediments have often been elevated and have exceeded pollution criteria (Garbarino et al. 1995). The Illinois Department of Public Health currently has fish consumption adviso- ries in effect for the MSR bordering Illinois due to elevated levels of PCBs and MeHg. Schmitt et al. (2002) reported that fish in the upper MSR subbasin commonly contained elevated concentrations of chlordane, dieldrin, PCBs, As, Pb, and Zn. Many towns and cities, along with defunct Zn smelters and PCB manufacturing plants, oil refineries, and other heavy industry are located in the Illinois River (ILR) drainage. Fish consumption advisories are in effect for portions of the river due to high concentrations of PCBs and MeHg. Groschen et al. (2000) reported that chlordane, dichlorodiphenyl- dichloroethane (DDD), dichlorodiphenyldi- chloroethylene (DDE), dieldrin, total PCB, 4 polycyclic aromatic hydrocarbons (PAHs), As, Cd, Cr, Cu, Pb, Hg, Se, and Zn were elevated in ILR bed sediments to concentrations at or above national benchmarks for protection of benthic life. Cadmium and Pb concentrations in sediments were at, or in some instances above, the national average (Groschen et al. 2000). Cadmium was highly elevated in fish tissue samples from these areas, in most cases exceeding concentrations in sediments and the national average. Chlordane, dieldrin, PCBs, as well as As, Pb, and Zn were elevated in fish tis- sues collected in the lower reaches of the river (Schmitt et al. 2002). Levengood (2003) found high concentrations of Cd in kidneys of wood ducks collected along the ILR, and suggested that breeding ducks inhabiting this system may be chronically exposed to Cd. Several surveys of the mussel popula- tions of the ILR and MSR have documented declines of sensitive species and episodic die-offs of localized populations (Blodgett and Sparks 1987). The mussel bed we sampled downstream of the confluence is considered a valuable resource due to the density and species richness of unionids (Miller et al. 1997) and is one of the few remaining beds on the upper MSR where Megalonaias nervosa, a commer- cially harvested species, is abundant (Ecologi- cal Specialists, Inc. 1999). A 1999 survey of Pool 26 found over 1,000 mussels representing 22 extant species, including 15 species with in- dividuals <3 years of age (Ecological Special- ists, Inc. 1999). Whereas PCBs and Hg and other metals are known to be especially prevalent in sedi- ments, water, and biota throughout much of the upper MSR sub-basin, no information exists on contaminants in the MSR and ILR confluence area (confluence) (but see www.umesc.usgs.gov for information on contaminants in the upper MSR), an area that may experience a mixing of contaminants from each of the rivers, as well as settling of water- and sediment-borne con- taminants in slack-water off the main channel following normal high flow and flood events. This study represented a preliminary investiga- tion of contaminant burdens in several tolerant, regionally abundant mussel species from near the confluence of the Illinois and Mississippi rivers. The overall objective of the study was to determine if concentrations of selected con- taminants were elevated in mussels from the confluence, and, if so, whether concentrations might be high enough to present potential risks to mussel health and survival. This information is required to determine whether the presence Contaminants in Unionid Mussels from the the Mississippi and Illinois RiversNovember 2008 199 of environmental contaminants is limiting mus- sel populations in this area and whether this would be expected to influence the restoration of mussel beds. METHODS Mussels Five regionally abundant mussel species, Am- blema plicata, Megalonaias nervosa, Quadrula quadrula, Obliquaria reflexa, and Obovaria olivaria were hand-collected from three mussel beds near the confluence of the Mississippi and Illinois rivers (Fig.1) in the autumn of 2003; a total of 109 individuals was collected (Table 1). We selected mussels of similar size to facilitate spatial comparisons among the populations of A. plicata. However, mussels used in this study were generally representative of the sizes encountered during sampling. Mussels were placed in cool- ers on wet ice until transported to the Illinois Natural History Survey Aquatic Ecotoxicology Laboratory, Champaign, Illinois, for process- ing and analyses Mussels were rinsed with deionized water to remove attached sediment or algae prior to measurements and tissue preparation. Shell length (anterior to posterior margin) and height (dorsal to ventral margin, excluding the hinge) were measured to the nearest 0.01 mm using digital calipers. Mussels were opened with a stainless steel knife and wet tissue was trans- ferred into pre-weighed, disposable, polysty- rene boats using a stainless steel spatula or a Teflon policeman. Whole wet tissue weights were recorded to the nearest 0.0001 g before homogenization in a blender. Two species, Q. quadrula and O. reflexa, required composite samples of three and four mussels, respectively, to provide sufficient tissue for processing. Five grams of the homogenate were transferred to scintillation vials and stored at –20 ºC prior to IL MSU MSD Illinois Missouri Mississippi River Illinois River Figure 1. Map showing Mississippi and Illinois rivers confluence area on the border between Illinois and Missouri. Dots indicate sampling sites: Mississippi River upstream of its confluence with the Illinois River (MSU), Illinois River at Swan Lake (IL), and Mississippi River downstream of the confluence (MSD). Vol. 38 Art. 5200 Illinois Natural History Survey Bulletin transfer to the Cooperative Wildlife Research Laboratory at Southern Illinois University for organic contaminant analysis. Following the removal of homogenized tissue for organochlo- rine, Hg, and MeHg analysis, measured volumes of deionized water were added to optimize homogenization for metals analysis. The weight of this addition was accounted for in calcula- tions of a wet:dry tissue ratio. All homogenized tissue was rinsed with deionized water into pre- weighed 250-ml glass beakers and oven dried at 80 ºC to a constant weight. Shells were also dried to a constant weight at 80 ºC. Dry weights were recorded and dry homogenates were im- mediately transferred to plastic bags and stored at –20 ºC. Tissues for Hg and MeHg analysis were dried under vacuum at room temperature for four to five hours, pulverized with mortar and pestle, and refrigerated at 4 ºC until digestion. A Tissue Condition Index (TCI) was calculated for each mussel (whole dry tissue weight / dry shell weight). Starrett’s (1971) and Whitney et al.’s (1997) studies of mussels on the Alton Reach of the Illi- nois River yielded aging criteria based on length or height for A. plicata, M. nervosa, Q. quadru- la, and O. reflexa. We used equations generated in both studies to determine ages for A. plicata, M. nervosa, Q. quadrula, and O. reflexa. Both length equations produced similar (i.e., not significantly different, p > 0.05) ages for each species. No aging information was available for O. olivaria in Pool 26, thus, ages for this species Table 1. Species and number of mussels collected at each site. Site Species n Mississippi at Cache Hollow Rd. (MSU) Amblema plicata 21 MRM 233.5 Obovaria olivaria 5 Obliquaria reflexa 22 Illinois River at mouth of Swan Lake A. plicata 17 (ILR) IRM 5.5 Mississippi south of Grafton (MSD) A. plicata 18 MRM 217 Quadrula quadrula 19 Megalonaias nervosa 7 were estimated by averaging ring counts made by two observers. Tissues dried at 80 ºC were prepared for elemental analysis (see below for Hg and Se de- termination methods) following USEPA Sample Preparation Procedure for Spectrochemical Determination of Total Recoverable Elements in Biological Tissues (USEPA 1991). Approxi- mately 0.5 g of dry tissue was weighed into acid- washed 125-ml Erlenmeyer flasks. Digests were conducted using Fisher trace metal grade HNO3, HCl, and H2O2 (30%), and were transferred into 100-ml volumetric flasks, diluted to 100 ml with deionized water, and homogenized. One blank was prepared for every six to nine tissue digests. Prepared samples were stored in 500-ml Nalgene containers at 4 ºC prior to analysis for metals at the Illinois State Water Survey (ISWS) using inductively coupled argon plasma optical emission spectroscopy with a Thermo Elemen- tal Model 61E vacuum spectrometer. Analyses were performed using USEPA Method 200.7 (USEPA 1991). The majority (>70%) of the calibration blanks adhered to the applicable control limits of less than the method detection limit (MDL). In other instances, the amount observed was less than or equal to five times the MDL. Initial calibration verifications met control limits (+/– 15%) for analyses of each element and continuing calibration verifica- tions data for each element were in the defined acceptance range of +/– 15% differences in recovery. Duplicate sample analyses for each Contaminants in Unionid Mussels from the the Mississippi and Illinois RiversNovember 2008 201 metal were within control limits, with the rela- tive percent difference (RPD) being less than 20% when the analyte concentrations were at least 10 times the MDL. Selenium concentrations were determined at the Illinois Sustainable Technology Center at Champaign, Illinois, by inductively coupled plasma mass spectrometry using yttrium as an internal standard. Maximum error was estimated at +/– 20%, based on quality control parameters. Analytical duplicates were reproducible, ranging from 0%–11%. Analytical and matrix spikes recovered in the range of 51%–87% and SRM (DOLT-2) recovered at 65%. Low spike recover- ies were corrected using the method of standard additions. Tissues for total Hg analysis were digested using a modified procedure from EPA Method 245.5 (USEPA 1991). Pulverized tissue (0.1 to 0.3 g) was mixed with 5 mL concentrated nitric acid (Fisher Trace Metal grade) and heated for eight hours at 65 ºC. An aliquot of the acidic digest was diluted 1:4 with an ultra-low Hg, aqueous HOBr solution (10% v/v) in a borosili- cate glass vial with a fluoropolymer-lined cap (I-Chem Class 200) and then heated in an oven overnight at 60 ºC. After cooling, samples were filtered using individual, acid-rinsed, 0.45-µm PTFE syringe filters. Except for the heating steps, calibration standards were processed iden- tically to samples. Diluted digests were analyzed for Hg using a flow-injection system with detec- tion by cold vapor atomic fluorescence spectrom- etry (Shade and Hudson 2005). The detection limit of the method is ~1 pg Hg absolute with a SD of 1.1 ng g-1 for replicate analyses. Spike recoveries were 93%–98% and the analysis of DOLT-2 Certified Reference Material yielded 90% of the reference value. Methylmercury was extracted from mus- sels using an acid digestion followed by solvent extraction. We began by placing 0.1g of mussel tissue in 5 mL of 4M HNO3 solution and heating the solution for 1.5 hours in a laboratory oven at 70ºC. After heating, the acidic sample digests were placed in a sonicator for approximately one minute each to liquefy any remaining solid tissue. One mL of 30% hydroxylamine was added to each sample digest followed by shaking and then 1 mL of concentrated HCl was added, followed by shaking. Next, to separate the MeHg from the highly organic tissue digest, each sample digest was combined with a 5-mL aliquot of toluene and shaken for one hour. Then each acidic solution/toluene mixture was centri- fuged and 4 mL of toluene was removed. The MeHg in the toluene was then back- extracted into an aqueous thiourea solution (1% thiourea, 0.1% HCl, 0.2% acetic acid) by shaking the samples for an additional 15 min- utes. After centrifuging the samples again, 4 mL of the thiourea solution was removed and filtered. Individual acid-rinsed 0.45-µm PTFE syringe filters were used to filter the final sample solution. Detection for these mussel samples was achieved using our Hg/ TU Complex - Ion Chromatography - Flow Injection-CVAFS (Hg-Tu IC) system (Shade and Hudson 2005). A small portion (100 µL) of the aqueous back-extractant was analyzed by direct injection into the HgTU/IC system. Content of MeHg in samples was calculated from observed peak areas for unknowns using calibration with external standards and using a correction for fractional recovery (77% ± 2.6%) in samples of mussel tissue spiked with MeHg prior to digestion. Tissues were analyzed for organochlorine pesticides and PCBs following the Association of Official Analytical Chemists methods for organochlorine compounds using gel perme- ation chromatography clean-up (USEPA 1999). Samples were homogenized with a 50/50 mix of methylene chloride and cyclohexane and dried with anhydrous sodium sulfate at a ratio of 4:1 Na2SO4 to tissue. The pesticide fractions were collected and dried using roto-evapora- tion. Samples were reconstituted with 2 mL of iso-octane. A Hewlett-Packard 5890 II gas chromatograph equipped with a Ni-63 electron capture detectors and auto injectors was used to analyze the extracts. Fused silica megabore columns (DB-5 and/or DB-624) were used for the analyses. Helium was used as the car- rier gas and 5% argon-methane for make-up gas. Injection port and detector temperatures, respectively, were 250 ºC and 350 ºC. Dupli- cates and spiked samples were analyzed with samples for quality control. Water and Sediment Water samples were taken at approxi- mately mid-column and were placed in 1-L glass jars for organic contaminant analysis (n = 4 per site) or 1-L Nalgene bottles for metals, total carbon, hydrogen, nitrogen, and particle-size analyses (n = 1 per site). Sedi- ments were collected using a wide-mouth Vol. 38 Art. 5202 Illinois Natural History Survey Bulletin Nalgene container to gather the top 3 to 6 cm of sediment into a plastic 4-gallon tub for homogenization by stirring. Subsamples of the homogenate were transferred to 1-L glass jars for organic contaminant analysis (n = 4 per site) and 500-ml Nalgene containers for metals analysis, CHN analysis, and particle-size deter- mination (n = 1 per site). Water and sediment samples for organic chemical analysis were stored at –20 ºC. An aliquot was taken from the homoge- nized composite sample of the top few centime- ters of the mussel bed. Subsamples were dried to a constant weight at 105 ºC and ground to a powder with mortar and pestle. Total carbon, hydrogen, and nitrogen (CHN) analysis was conducted using a CHN/O/S Elemental Ana- lyzer CHN440. Samples were broken into their atomic components in an oxygen atmosphere at 980 °C forming CO2, H2O, and NxOy. These gases were carried to a detector by a helium gas stream. Samples were analyzed in duplicate. Recorded values were compared to a standard (Acetanilide, OAS), and calculations were made based on sample weight. Particle-size analyses were conducted by using a pipette extraction procedure described in Indorante et al. (1990). Percent carbon was unknown at the time of analysis; therefore, samples were not treated with 30% H2O2 to remove organic matter (a procedure that is typi- cally performed on samples with greater than 1% organic C). The resulting calculations for our ILR and MSD sediment composites theo- retically could have a 5 to 10% lower clay-size content and a 5 to 10% higher silt content due to this omission. Metals analyses for a composite sediment sample from each site were conducted by the Illinois Sustainable Technology Center. Prior to digestion, disaggregated sediment samples were shaken to homogenize them. Represen- tative portions of each sediment sample were used in a nitric and hydrofluoric acid micro- wave digestion procedure, equivalent to US EPA Method 3052 (USEPA 1999), to dissolve sample matrices into solution for total metals analysis. Results were obtained by inductively coupled plasma mass spectrometry using scandium, yttrium, and thorium as internal standards. Quality control parameters pose an uncertainty of no more than +/– 20 % for all metals except Al. Although Fe, Pb, and Zn were observed in digested blanks, the concen- trations were extremely low compared with sample concentration levels. Relative percent differences of digested duplicates ranged from 1 to 14%. Recoveries of analytical spikes, ma- trix spikes, and a Standard Reference Material (SRM #2710, NIST Montana Soil) ranged from 76 to 126%. Aluminum was high in digested blanks and only recovered at 23% from the SRM. Samples and SRM Al concentrations were >1%, therefore values reported herein are minimums. Sediment was homogenized with so- dium sulfate and Soxhlet extracted with 50 mL hexane for 30 minutes. The extract was transferred to a round-bottom flask and reduced using rotary evaporation and solvent exchanged to hexane. Clean-up followed using a Florisil column to separate PCB fractions from organo- chlorine pesticides. Each fraction was reduced to 1 mL final volume for analysis by GC/ECD. Water samples were liquid-liquid extracted three times by adding 60 mL of methylene chloride (MC) to 1 L of water sample in a separatory funnel. The MC layer was de- canted and new MC added for each of the three extractions. The extract was then reduced using rotary evaporation, solvent-exchanged to hexane, and clean-up with Florisil for fraction separation as described for sediment. The same instrumentation and analytical methods were used as described for tissue extracts. Statistical Comparisons Elemental tissue concentrations and mor- phological parameters of A. plicata were com- pared spatially using one-way ANOVA (α = 0.05) followed by Bonferroni multiple compari- sons. The entire data set (n = 30) was analyzed for normality (Shapiro-Wilks); non-normal data were log10 transformed and reanalyzed. If data were still non-normal, we used a nonparametric test (Kruskal-Wallis, α = 0.05). Elemental tis- sue concentrations and morphological param- eters were compared among species collected from the MSR upstream (MSU) and down- stream (MSD) of its confluence with the ILR. Since interspecific elemental concentration comparisons were made using fewer samples (n = 5 per species), the same nonparametric test was used. In both spatial and interspecific comparisons, some elements were below the Method Detection Limit (MDL) in several mussel samples. For instances in which at least 40% of the data being analyzed were below Contaminants in Unionid Mussels from the the Mississippi and Illinois RiversNovember 2008 203 the MDL, means for each species are reported on the number of samples with concentrations above the MDL and are indicated as such. Statistical comparisons were not performed on these elements; however, descriptive statistics are provided. For instances in which <40% of the data were below the MDL, a value of one- half the MDL was used. Organic contaminants were detected in too few mussel tissue samples, sediment samples, and water samples to allow statistical comparisons. RESULTS Total carbon concentrations in bed sediment ranged from 1.0% at MSU to 1.8% at MSD (Table 2). Proportions of clay, silt, and sand were also variable among sites. Water quality characteristics including pH, conductivity, alka- linity, and hardness at the two MSR sites were quite similar, whereas ILR had higher values Table 2. Total organic carbon, hydrogen, and nitrogen content and texture characteristics of composite sediment samples from the Mississippi River upstream (MSU) and down- stream (MSD) of the river confluence, and the Illinois River by Swan Lake (ILR). Sediment characteristic MSU ILR MSD %C 0.95 1.645 1.79 %H 0.15 0.405 0.19 %N 0.045 0.135 0.075 % Clay 6.62 22.77 8.74 % Silt 16.66 74.47 42.23 % Sand 76.72 2.76 48.25 Soil Type loamy sand silt loam loam Table 3. Water quality characteristics of grab samples from the Mississippi River upstream (MSU) and downstream (MSD) of the river confluence and the Illinois River by Swan Lake (ILR). Site pH Conductivity µS/ cm Alkalinity mg/L CaCO3 Hardness mg/L CaCO3 MSU 7.8 404 160 180 ILR 7.7 748 200 272 MSD 8.1 441 162 190 for conductivity, alkalinity, and hardness (Table 3). With the exception of Cr in sediments at ILR and MSD, concentrations of elements in both media were relatively low (Table 4). Amblema pictata specimens were physically similar across sites in terms of mean length (anterior to posterior margin), mean height (dorsal to ventral margin), dry shell weight, and TCI (Table 5). Amblema pictata from the ILR had significantly higher wet tissue weights than did specimens from both of the MSR sites (p<0.02). However, because other morpho- metrics and, therefore, calculated mean ages, were not significantly different, mussels were considered to be of the same cohort and useful for spatial comparisons of contaminant loads. Ambleme pictata, M. nervosa, and Q. quadrula collected at MSD varied by size and age (p<0.001, Table 5). Vol. 38 Art. 5204 Illinois Natural History Survey Bulletin Table 4. Priority element concentrations in composite sediment (mg/kg dw) and water (mg/L) samples from the Mississippi River upstream (MSU) and downstream (MSD) of the confluence and the Illinois River by Swan Lake (ILR). MSU ILR MSD Element sediment water sediment water sediment water Ag 0.19 - 0.22 - 0.15 - As - <0.037 - <0.037 - <0.037 Be - <0.002 - <0.002 - <0.002 Cd 0.16 <0.003 0.40 <0.003 0.25 <0.003 Cr 27 <0.004 53 0.01 43 <0.004 Cu 8 <0.003 21 <0.003 12 <0.003 Ni - 0.01 - 0.02 - 0.01 Pb 15 <0.028 24 <0.028 19 <0.028 Sb - <0.021 - <0.021 - <0.021 Se <0.02 <0.034 <0.02 <0.034 <0.02 <0.034 Zn 33 0.01 83 0.00 48 0.00 - indicates not analyzed Table 5. Mean ± sd length, height, weight, moisture content, TCI2, and calculated age (per Starrett 1971 and Whitney et al. 1997) of mussels collected during 2003 from three sites near the confluence of the Mississippi and Illinois rivers. Collection sites were Mississippi River upstream (MSU) and downstream (MSD) of the confluence and the Illinois River by Swan Lake (ILR). Numbers in parentheses are sample sizes. MSU ILR MSD Amblema plicata (10) A. plicata (10) A. plicata (10) Megalonaias nervosa (5) Quadrula quadrula (14)3 Length (mm) 99.5 ± 5.9 A 104 ± 4 A 100 ± 5 A 143 ± 17 60.8 ± 5.5 Height (mm)1 75.7 ± 3.5 A 77.5 ± 2.7 A 76.3 ± 4.5 A 103.4 ± 7.9 54 ± 4.2 Wet tissue weight (g) 38.17 ± 5.6 B 47.4 ± 10.2 A 38.7 ± 5.9 A 91.3 ± 24.9 34.8 ± 3.6 % water 0.82 ± 0.01 0.83 ± 0.02 0.82 ± 0.01 0.84 ± 0.02 0.82 ± 0.01 Dry shell weight (g) 169 ± 34 A 196 ± 29 A 177 ± 33 A 335 ± 95 154 ± 19 TCI2 4.2 ± 0.6 A 4.2 ± 0.8 A 3.9 ± 0.4 A 4.5 ± 1 4.2 ± 0.3 Age (yr) 16.5 ± 2.5 A 18.2 ± 2.0 A 16.8 ± 2.9 A,a 12.9 ± 3.9b 9.6 ± 1.0c Concentrations of Zn did not vary with age in A. plicata, M. nervosa, Q. quadrula or O. reflexa. However, we did observe a significant relationship between tissue Cd and age in A. plicata from both MSD (R2= 0.71, P= 0.002 ) and ILR (R2= 0.37, P= 0.08), though not MSU (R2= 0.09, P= 0.41); in both significant cases, this relationship was nega- tive. The mean concentrations of Cu (P= 0.03), Se (P< 0.001), and Zn (p< 0.001) were highest, and Pb (P= 0.001), and Cd (nominally) were lowest in A. plicata from MSD (Table 6). Mean concentrations of Cu, Pb, and Se were intermediate and Zn was lowest in mussels from ILR. Total Hg concentra- tions were low and did not differ among sites for A. plicata; however, a significant difference was observed for MeHg (P= 0.01), with the highest mean concentration observed at MSU (Table 6). Amblema plicata tissue from MSD contained significantly higher concentrations of Zn (P= 0.01) compared to M. nervosa, which in turn had higher concentrations of both Pb (P= 0.02) and Zn than Q. quadrula (Table 7). Quadrula quadrula was characterized by higher total Hg concentrations (P= 0.03), as compared to A. plicata at MSD (Table 7). Different capital letters indicate means for A. plicata are significantly different among sites (p< 0.05). Different lower case letters indicate significantly different means among species at a site (p < 0.05). 1 Height does not include hinge. 2 Tissue Condition Index = dry tissue weight / dry shell weight 3 Q. quadrula TCI, wet tissue weight, and % water measurements are based on five composites of three mussels each (one composite has two mussels). Contaminants in Unionid Mussels from the the Mississippi and Illinois RiversNovember 2008 205 Table 6. Mean ± sd concentrations* (mg/kg dry weight) of EPA priority elements and methylmercury in tissues of Amblema plicata collected from near the confluence of the Mississippi and Illinois rivers. Locations are Mis- sissippi River upstream (MSU) and downstream (MSD) of the confluence and the Illinois River by Swan Lake (ILR). MDL = method detection limit (mg/kg d) Element MDL MSU ILR MSD (n= 10) (n= 10) (n= 10) As 7.08 10.4 (1)