fig llHHlHlllHlIHIIWIHIHHWNW“HHI!!!|4|I|!H| _‘-‘ ‘ “ .. 0“ i ' ‘8: in: v\}‘ {30" t r ‘ ‘ * . g 3 3 ‘ i_ ». ”M's i Migk‘ifigfifl; £33th gatvewi’iy L 4‘ 7" This is to certify that the thesis entitled The Role of Suspended Sediments in the Bioaccumulation of Polychlorinated Biphenyls by Fathead Minnows (Pimeghales promelas) presented by Paulette M. Queener has been accepted towards fulfillment of the requirements for Master of Science degree in Fisheries and Wildlife Major professor August 11, 1982 Date 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution 1”" '" V ] 'Ilnlmmlfllmiiflnfinnmmmll' 3 1293 10726 6847 L MSU LIBRARIES \— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. K 7‘31?“ ' 4" “ GHQ :9 :Nfliw V. . t: ' i? I ("f 1 1'7. .5 ‘T I 9 I. ‘7 __ __ v H.) ‘3 ,*-J '3‘- J1 THE ROLE OF SUSPENDED SEDIMENTS IN THE BIOACCUMULATION OF POLYCHLORINATED BIPHENYLS BY FATHEAD MINNOWS (PIMEPHALES PROMELAS) BY Paulette M. Queener A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1982 ABSTRACT THE ROLE OF SUSPENDED SEDIMENTS IN THE BIOACCUMULATION OF POLYCHLORINATED BIPHENYLS BY FATHEAD MINNOWS (PIMEPHALES PROMELAS) By Paulette M. Queener A proportional diluter system was used to deliver PCB-contaminated suspended sediment to four flow-through aquaria. Fathead minnows were exposed to suspended sediment in five experiments lasting six to 24 days, using three different soils. Two of the soils were contaminated in the lab with Aroclor 1254 and the other soil contained Aroclor 1242 as a result of environmental contamination. The soils contained 1.0, 5.5, and 12.696 organic matter. Fathead minnows exposed for 14 days to suspended sediment with 5.596 organic matter and contaminated with 100 ug/g of Aroclor 1254 accumulated over twice as much Aroclor 1254 as did fatheads exposed under the same conditions with a suspended sediment of 12.696 organic matter. Fathead minnows exposed for six days to suspended sediment with 196 organic matter environmentally contaminated with Aroclor 1242 accumulated high levels of Aroclor 1242. Also, the rates of uptake of Aroclor 1242 were high when compared to uptake of Aroclor 1254. Fish appeared to accumulate the lower chlorinated congeners, however, this phenomenon was not statistically significant. To my parents, Edith C. and Anderson B. Queener ii ACKNOWLEDGEMENTS My sincere thanks go to Ms. Terry Aiken for her untiring efforts in the laboratory, advice and friendship. A special thanks to Dr. Tom Lynch for his unofficial role as advisor and mentor. I would also like to extend my thanks to the members of my committee, Dr. Howard Johnson, Dr. Matthew Zabik and Dr. Frank D'Itri. A special thanks to Dr. Niles Kevern for relief pitching in the ninth inning. And thanks to Fred Lehle for assistance with graphics. iii + TABLE OF CONTENTS LISTOFTABLES . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . .' . INTRODUCTION AND OBJECTIVES . . LITERATURE REVIEW . . . . . . METHODS . . . . . . . . Experimental Plan . . . . . . . Sediments . . . . . . . . . . . . Fish . . . . . . . . . . . . . Water Supply . . . . . . . . . . Exposure Methods . . . . . . . . . Sampling . . . . . Analytical Methods . . . RESULTS........ Suspended Sediment and Water Isomer Distribution Analysis PCB Accumulation by Fish . . Effects on Accumulation by Organic Matter Bioconcentration . . . . . . Transfer Coefficients . . . . . . . . Experiment IV . . . . . . . . . . . ExperimentV'. . . . . . . . . . . DISCUSSION . . . . Distribution of Isomers . . Effects on Accumulation by Organic Matter Experiment IV . . . . . . . . Experiment V . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . 51 51 52 53 55 APPENDIX I o o o o o o o o o o o o o o o o 0 Statistical Analysis to Detect a Difference in PCB Concentration in Suspended Sediment between Experiments (u g/g) ' . . . , . iv APPENDIX 11 Statistical Analysis to Detect a Difference in PCB Concentration in Water Samples from Experiments II and III (11 8/2) , APPENDIX III Statistical Analysis to Detect a Difference in Suspended Sediment Concentration in Experiments II and III (mg/2.) - . APPENDIX IV Formula for the Statistical Comparison of Regression Slopes Assuming Heterogenous Variance APPENDIX V Statistical Analysis to Detect a Difference in PCB Concentration in th (u g/g) O O O O O O 0 O O O I O O O O O O I O 59 60 61 62 LIST OF TABLES 1 Characteristics of Soils and PCB Concentration of the Soil Fractions . 2 PCB Concentrations in Suspended Sediment 3 PCB Concentrations in the Water . . . 4 Suspended Sediment Concentrations . . . . . . 5 Comparison of Percent Peak Areas by Student T Test . . 6 PCB Concentrations in Fish Samples from the Final Sample in Each Emaimmt O, O O O O O O O O O O O O O O O O O O 7 Results of Approximate '1‘ Test on Slopes of Uptake Regression Lines fromExperimentsIIandIII . . . . . . . . . . . . 8 Comparison of Bioconcentration Values from Experiments II, III, IV and v 0 O I O O I O O O O I O O O O I O O O O O O O O O 9 Comparison of Transfer Coefficients for Experiments II, III, IV and V . Appendix 1 Statistical Analysis of PCB Concentration in Suspended Sediment betweenExperiments . . . . . . . . . . . . . . . 2 Statistical Analysis of PCB Concentration in Water Samples from Experimentsnandm . . . . . . . . . . . . . . 3 Statistical Analysis of Suspended Sediment Concentration in Experimentsllandm.... 4 Formula for the Statistical Comparison of Regression Slopes Assuming Heterogenous Variance . . . . . . . . . . . . . . 5 Statistical Analysis of PCB Concentration in Fish . vi 12 20 22 23 27 29 42 43 47 58 59 60 61 62 V LIST or FIGURES Sample Chromatographs of PCBs in Fish, Water, Soil and a Standard Aroclor 1254 Chromatograph Aroclor 1254 Uptake by Fathead Minnows (P Pimemales pro melas las) 1n Four Aquaria in Experiment I . Aroclor 1254 Uptake by Fathead Minnows (Pime ales romelas) in Aquarium No. 2 in Experiments H (Low Organic Soil} and III H1gh Organic Soil) 0 O O Aroclor 1254 Uptake by Fathead Minnows (Pime ales romelas) in Aquarium No. 3 in Experiments 11 (Low Organic SoiII and III H1gh ormic $fl). O O O O I O O O 0' O. O O O» O O C O Aroclor 1254 Uptake by Fathead Minnows (Pime hales romelas) in Aquarium No. 4 in Experiments 11 (Low Organic Soil and III High Owicsoil)................. Aroclor 1254 Uptake by Fathead Minnows (Pime hales romelas) in Aquarium No. 5 in Experiments II (Low Organic SO11 and III H1gh Organic Soil). . , , , 2, 5, 3' ,4'-Tetrachlorobiphenyl Uptake by Fathead Minnows (P Pimemales promelas) 1n Four Aquaria in Experiment IV PCB (as Aroclor 1242) Uptake by Fathead Minnows (P Pimephales promelast) 1n Four Aquaria in Experiment V (Shiawassee River lmen O O O O O I 0 vii 26 31 35 37 39 40 46 50 INTRODUCTION AND OBJECTIVES . An estimated 353.8 million kg of polychlorinated biphenyls (PCBs) have {entered the environment by various routes (Shea, 1973). Much of it has been directly dumped into aquatic environments by industries. It was reported that up to 14 kg a day were at one time being dumped into the Hudson River (Hellman, 1976). Atmospheric fallout, especially rainfall, is thought to be a large source of PCBs now entering aquatic environments (Murphy and Rzeszutko, 1978). A recent study estimates 4,800 kg of PCBs are deposited in Lake Michigan by rainfall in one year (Murphy and Rzeszutko, 1978). The resulting bioconcentration of these compounds by fish has caused serious problems for sport and commercial fisheries. Fish sampled from Lake Michigan in 1971 contained PCBs similar to Aroclor 1254 and ranging in concentrations from 2.7 ug/g in smelt to 15 ug/g in lake trout. PCB concentrations in fish sampled from Lake Superior on 1972-1973 ranged from 0.12 ug/g in sculpins to a high of 5.6 ug/g in a large lake trout (5.6 kg). However, most of the smaller trout contained less than 5 ug/g which is the tolerance level set by the U. S. Food and Drug Administration for human consumption (Veith, 1975; Veith et a1., 1977). As a result of the widespread environmental contamination by PCBs, much research has been conducted on the toxicity, accumulation and fate of PCBs in aquatic environments. ,, - ——.M r Environmental contamination of fish by PCBs has been most closely ) + k associated with highly industrialized areas. Veith (1975) in his survey of Lake Michigan, generally found higher concentrations of PCBs in fish from the 2 southern half of the lake which is surrounded by highly industrialized areas. Rivers with industrial sites also have a serious contamination problem. In 1972 fish from the Hudson River were averaging 200 ug/g and Ohio River and Allegheny River fish were averaging 100 u g/g (Stalling and Mayer, 1972). Fish, in both lakes and rivers, have been exposed to very high concentrations of PCBs in the water, however, since direct dumping of P033 is no longer allowed, water concentrations have been greatly reduced. The problem now is the reservoir of PCBs in the sediments leaching into the water providing a continuing source of contamination for aquatic biota. If the sediments are left undisturbed the PC38 (would slowly leach from the sediments into the water at low -levelsfi_(‘gremep, I 1914), but little is known about the availability of PCBs from a contaminated sediment which is resuspended in the water column. In such a case the suspended sediments would be in direct contact with fish, possibly increasing the uptake of PCBs. The object of this study is to investigate the availability of PCBs from resuspended sediments to fish. Specific objectives include: (1) the determination of the availability of PCBs sorbed to different soils to fish, (2) the determination of the effects of organic matter content and other soil parameters on the availability of P033 to the fish and (3) to test a sediment from an aquatic environment that has already been contaminated with P038, to compare with our artificially-contaminated soils. LITERATURE REVIEW Toxicity data for P033 varies a great deal depending on species of fish tested and chlorine content of the biphenyl. It also has been observed that toxicity is inversely related to the percentage of chlorine in the compound (Stalling and Mayer, 1972) and that direct exposure via the water is more toxic than dietary uptake (Stalling and Mayer, 1972; Gruger et al., 1975). Flow through tests yield 30-day LC50 values ranging from three to 433 ug/R. in three different species of fish for Aroclors 1242, 1248, 1254 and 1260 (Mayer et al., 1977). 4 to 106 depending on percent Bioconcentration values range from 10 chlorine content, the usual pattern being that higher chlorinated PCBs are concentrated to a greater extent than lower chlorinated isomers (Branson et al., 1975; DeFoe et al., 1978; Gruger et al., 1975; Mayer et al., 1977; Metcalf et al., 1975; Sanbord et al., 1975; Veith, 1975; Veith et al., 1977). In a feeding study conducted on juvenile coho salmon (Oncorhynchus M) comparing a tetrachlorobiphenyl with two hexachlorobiphenyls, it was found that after 108 days there was a significantly (9096 c.f.) higher accumulation of each hexachlorobiphenyl compared to the tetrachlorobiphenyl (Gruger et al., 197 5). In the same study, after a starvation period, the hexachlorobiphenyl concentrations in the juvenile coho increased while the concentration of the tetrachlorobiphenyl decreased. They concluded that there was a mobilization or a transformation of the chlorobiphenyls probably due to metabolization of the lower chlorinated biphenyls or the difference in the partition coefficients of the chlorobiphenyls in 4 aqueous media. This conclusion is supported by a model ecosystem study suggesting that Gambusia m can hydroxylate tetrachlorobiphenyls (Metcalf et al., 1975). Another study indicated that green sunfish (Lemmis cyanellus) were able to hydroxylate trichlorobiphenyls while tetra- and pentachlorobiphenyls were much less susceptible to hydroxylation (Sanborn et al., 1975). Other studies also have indicated a rapid depletion of lower-chlorinated biphenyl residues in fish (DeFoe et al., 1978; Mayer et al., 1977). Many organochlorine pollutants, including PCBs, readily adsorb to suspended particulate matter in the water or become associated with the sediments. Therefore, sediments play a major role in the translocation of PCBs in aquatic environments. This role, however, is not clearly defined, but much depends upon the physical, chemical and biological characteristics of the sediments involved. Among the characteristics of the sediment that can affect the adsorption and release of chlorinated hydrocarbons, and similar compounds, it is generally considered that organic matter content and particulate matter are the two most important. Other important factors to consider are pH of the soil and water solubility of the compound involved. The lipophilic nature of PCBs make them easily adsorbed to organic matter. A study with DDT demonstrated a close relationship between organic content and high distribution coefficients, as derived from the Freundlich and Langmuir adsorption equations (Shin, 1970). It has been generally concluded that under field conditions organic matter is the single most important factor in the retention of pesticides in the soil (Weed and Weber, 1974). This conclusion is supported by data obtained from an investigation of southern Lake Michigan sediments in which t-DDT (DDT + DDE + DDD) and dieldrin were found to be related (r = 0.6 and 0.66, respectively) to organic carbon content of the 5 sediments (Leland et al., 1973). Also, it was generally observed that large accumulations of insecticides were associated with sediments containing high amounts of organic matter (Leland et al., 197 3). Bioactivity of chlorinated hydrocarbons is affected by organic content, especially in moist soils (Weber and Weed, 1974). Therefore, the organic content of suspended sediments will play an important role in the bioactivity of adsorbed chlorinated hydrocarbons. Specific studies indicate a similar relationship between PCBs and soil as seen with other chlorinated hydrocarbom. Adsorption studies with 2,4'-dichloro—, 2,2',5,5‘-tetrachloro- and 2,2',4,4',5,5'- hexachlorobiphenyls with humic acid, illite clay and Del Monte sand, show the greatest adsorption for hexachlorobiphenyl followed by tetrachloro- and dichlorobiphenyl and total adsorption of the di-, tetra— and hexachlorobiphenyls followed the series humic acid > illite clay > Del Monte sand (Haque and Schmedding, 1976). Therefore, adsorption generally increases with increasing chlorine number and organic content of the soils. Haque and Schmedding (1976) conclude that most of the transport of PCBs in aquatic systems will be by adsorption to particulates and that movement due to solubility will be negligible for most PCBs. In a later study it was concluded that both total organic carbon and particle size distribution were important factors in the partitioning of two tetrachlorobiphenyl isomers, but there was no correlation between organic content and the partitioning of PCB mixtures (Steen et al., 1978). This suggests that the adsorption effects of organic matter are modified by association with various particle sizes, especially clay. In Lake Michigan a correlation (r = 0.65) was found between clay material (< 2 u) and organic carbon which led to the conclusion, that the occurrence of t-DDT and dieldrin in the deep basins of southern Lake Michigan is due to transport by small particles (Leland et al., 197 3). In a study on the retention of DDT and methoxychlor on soils treated with 6 hydrogen peroxide to eliminate organic matter and on the same soils untreated, it was found that in treated and untreated soils the highest retention capabilities were shown by the clay fractions (Richardson and Epstein, 1971). Chlorinated hydrocarbon compounds have been associated with particulate matter in sediments. It has been determined that pesticides are associated with various particulate components as a function of particle size in water samples from Lake Erie (Pfister et al., 1969). The high degree of association between clay fractions and PCBs has been demonstrated in studies using percolating water to induce leaching of PCBs from soils (Tucker et al., 1975). It was discovered that the amount of leaching of PCBs was related to the clay content of the soils. When leaching did occur it included only the lower chlorinated, more water-soluble isomers from those soils with a low percent clay content. The greater water solubility of the lower chlorinated isomers is well documented. A solubility study found that two different batches of Aroclor 1254 had water solubilities of 2 to 3 mg/f. in fresh water, the latter having a greater concentration of higher chlorinated isomers (Zitko, 1970). Aroclor 1221, which is predominantly dichlorobiphenyl, was reported to have a water solubility of 5.0 mg in fresh water (Zitko, 1970). Water solubilities of Aroclor 1242 and 1016 in one study are reported at 0.34 mg/l and 0.42 mg/SL, respectively (Paris et al., 1978). A value for Aroclor 1016 was also reported at 0.225 to 0.250 mg/l (Tucker et al., 1975). The values cover a wide range and depend on the method used to solubilize the P038 in water. The suspendion of sediments in the water column can greatly influence the transport of sorbed PCBs. It has been noted that when water is added to soils it displaces sorbed chlorinated hydrocarbon pesticides on soils (Green, 1974). The increased surface area caused by resuspension of sediments increases the competition between water and PCBs for adsorption 7 sites and increases the amount of PCBs solubilized in water. In accordance with the solubility values it is the lower chlorinated isomers that are displaced first and become more readily available for direct absorption from the water by aquatic organisms. The pH of the suspended sediment-water system can influence the adsorption and release of organic compounds with the soil. Being essentially non-polar, PCBs are not greatly affected, however, there is a general trend of increased sorption with decreased pH due to enhanced hydrogen bonding (Shin, 197 0). METHODS I. Experimental Plan As described in the literature review, the dynamics of the relationship between PCBs, soil and water have been examined closely. Halter and Johnson (1977) combined these components with a system containing fish forming a model aquatic system to investigate the "dynamics" of Aroclor 1254. PCB desorption from sediments and accumulation by fish were studied under flow-through and static water conditions with undisturbed bottom sediments contaminated with Aroclor 1254. The experiments by Halter and Johnson were continued in this study using the same model aquatic system to study suspended sediments contaminated with Aroclor 1254 under flow-through water conditions. The purpose of this study was to verify the prediction that PCBs sorbed to sediments would be more available to fish if the sediments were suspended. Also that sediment characteristics, such as percent organic content and particle size distribution, were important factors in the desorption of PCBs from the suspended sediment and the accumulation of PCBs by the fish. The model aquatic system used throughout this study consisted of a proportional diluter system which continuously delivered contaminated suspended sediment to four aquaria where fish were exposed to four different concentrations of suspended sediment. A fifth aquarium received water frpm the diluter that contained no suspended sediment or P035 and acted as a control. Five experiments were conducted using two different sediments and one soil which were selected for their differences in percent organic content. Two 9 of the soils were contaminated in the laboratory with PCBs prior to use in the diluter system and the third soil had been contaminated with P083 in the environment. In all experiments the three components of the system, fish, suspended sediment and water, were sampled and analyzed for PCB content. The first experiment, lasting 24 days, was conducted primarily to determine if the diluter system could deliver reproducible amounts of suspended sediments to the aquaria and to determine a reasonable time scale for future experiments. Suspended sediment levels (mg/ll. water) were monitored as well as PCB concentrations in suspended sediment, fish and water. The distribution of the various isomers present in Aroclor 1254 (as represented by different peaks on the chromatograph) was examined in fish, water and suspended sediment and compared to chromatographs of standard Aroclor 1254 solutions. The isomer distribution was monitored to determine if there was a selective retention of isomers by the suspended sediments and/or a selective uptake by the fish. The time span for the second experiment was shortened to 14 days since it was determined from the first experiment that 14 days was a sufficient amount of time for the fish to concentrate easily detectable levels of PCBs. The distribution of Aroclor 1254 isomers was examined as in the first experiment. The third experiment was conducted in the same manner as the second experiment, the only difference being the type of soil used. The third experiment utilized a soil with 12.696 organic matter (greenhouse soil) and was med as a direct comparison to the second experiment in which a soil with only 5.5% organic matter was used. Distribution of Aroclor 1254 isomers was examined as in experiments I and 11. Comparison of the second and third experiments was expected to show that PCBs adsorbed to a suspended sediment with a high organic matter content would be less available to fish by direct or indirect absorption than PCBs 10 adsorbed to a suspended sediment with a low organic matter content. The defintion of direct absorption, as used here, is the absorption of PCBs by fish as a result of contact with PCB-contaminated suspended sediment. In other words, the PCBs are absorbed directly from the suspended sediments by the fish. Indirect absorption is the uptake of PCBs by fish from water resulting from the PCB desorption from soil to water. The exact mechanism of uptake by fish is not known but is probably a combination of both direct and indirect absorption. A fourth experiment was conducted to monitor the movement of a single PCB isomer, 2,5,3',4'-tetrachlorobiphenyl. The soil used in this experiment contained 12.696 organic matter. The purpose of this experiment was to collect data on tetrachlorobiphenyl uptake by fish where the source of contamination was the suspended sediment. These data were to be compared with literature values of fish uptake of tetrachlorobiphenyl where the source of contamination was water with no suspended sediment present. The nature of the suspended sediment was expected to have a modifying affect on accumulation of tetrachlorobiphenyl by fish because of higher retention of PCBs on the high organic content soil. In the final experiment a soil was used which had been obtained from a stream that had been contaminated with PCBs. The soil contained only 196 organic matter and the experiment was intended to be used as a direct comparison to the second and third experiments in which soils containing 5.5 and 12.696 organic matter were used. However, a direct comparison of experiments was not possible for the following reasons. First, the soil used in the last experiment contained mostly Aroclor 1242 and in the previous experiments the soils had been dosed with Aroclor 1254. The higher percentage of tetrachlorobiphenyls in Aroclor 1242 increased the desorption from soil to water since tetrachlorobiphenyls are more water soluble than penta- and 11 hexachlorobiphenyls which are the predominant isomers in Aroclor 1254. Secondly, there was a limited amount of soil available in the last experiment which made it necessary to reduce the experiment to six days, whereas the other experiments that were intended for comparison lasted 14 days. While no direct comparisons could be made with previous experiments, this final experiment was valuable in providing data on the movement of PCBs in the environment under conditions where environmentally-contaminated sediments are redistributed through the water column. 11. Sediments Three different soils were used in the experiments (Table 1). In experiments I and II, the sediment used was obtained from a pond in Fenner Arboretum, Lansing, MI. This sediment was selected because the arborteum had not used any pesticides for several years, therefore reducing the chance of interfering compounds being present (especially other organo-chlorine compounds). The arboretum soil had a 5.596 total organic conent, a pH of 7.2 and the following percentages of fractions: sand, 8596; silt, 13%; and clay, 296. Organic content of the fractions was 5.296, 20.68% and 12.6596 for sand, silt and clay, respectively. The soil used in experiments III and IV was obtained from the Michigan State University Pesticide Research Center greenhouse and showed no PCB- interfering peaks upon analysis. The greenhouse soil had a total organic content of 12.6%, a pH of 7.7,. and sand, silt and clay percentages of 75, 8 and 17, l respectively. Percent organic content of the fractions was 9.91, 17.73 and 17.13 for sand, silt and clay, respectively. The soil used in experiment V was dredged from the south branch of the Shiawassee River, Livingston County, Michigan (approximately 153 m west of 12 .EE Sac. : :65 63 «o 66:66 a 5.3 865.6: 3630:— >306 .58 3.3. c 3 9. EE Sue. : «o 66:66 a 5:5 863.6: 3620:: game .696— ..o 5:. 2.5. c «o 66:66 a 565 863.6: 86:65 .653“ . 3.3 5 >30 3 S 8.6 s :6 8.3 mmé mm 65m «.5 :4 63m 88:33am . an: S E6 s a: 2.: a a... 5.2: 3.: 2. 6:em s.» o.§ 626.3626 . aqua a >66 3 3. 8.8 2 o a... 3.6m cud mm 6 :am «4. m.m 830.692 RR 3 mo: 3:690 “:86: 33. 6266.5 a: 3:690 new «:86: 6:029: 66,—. 266626 :8 a: :6 86288.80 :0: 65 26m :6 86683.85 A 638. 13 Highway 59 just north of 1-96). An Ekman dredge was used to sample the sediment in approximately one foot of water. This sediment had been contaminated with PCBs, and contained only 1.0% total organics and had a pH of 7.4. The sand fraction accounted for 89% of the soil, silt, 4%, and clay, 7%. Organic content of the fractions was as follows: sand, 2.38%; silt, 12.97% and clay, 25.50%. The percentages of sand, silt and clay for each soil were determined by the hydrometer method (Bouyoucous, 1927; 1928; 1929). All soils were filtered in a no. 60 standard sieve (opening size 0.0098 inches, approximately 0.25 mm) to remove large debris and most of the heavy sand to minimize settling in the diluter and aquaria. Soil fractions were separated by sieving with a no. 230 standard sieve (0.0025 inch mesh, approximately 0.06 mm) to remove the sand fraction. The remaining fraction, containing the silt and clay, was suspended in water (using the Bouyoucous method) and allowed to separate by settling out the heavier silt in 30 minutes. The silt fraction was all that had settled out in 30 minutes. The material remaining in suspension at the end of 30 minutes represented the clay-colloid fraction and was separated from water by centrifugation. The three fractions were analyzed for percent organic matter and PCB concentration. Total percent organic matter of the soil as a whole and percent organic matter for each fraction was determined by combustion in a muffle oven at 500°C for two hours. For experiments I, II, and III the soils were dosed with Aroclor 1254 at 100 u g/g soil. The Aroclor was dissolved in excess acetone and mixed with the dry soil in a shallow pan. The acetone was allowed to evaporate and then the pans were covered with aluminum foil to minimize volatilization of the PCBs. In experiment IV the soil was dosed at 50 ug/g soil with 2,5,3',4'- - ‘60-“- F... 14 tetrachlorobiphenyl in the same manner. In experiments II, III and IV the dosed soils were prewashed by allowing water to flow over them for 14 days, stopping the flow of water twice, stirring the sediment, allowing it to settle, then resuming the flow of water. This was done to eliminate any excess PCB that was not tightly bound to the soil therefore simulating a sediment contaminated in an aquatic environment. Before being introduced into the diluter system the soil was mixed with water to form a slurry which was constantly stirred in a reservoir. The slurry was pumped through 0.49 cm i.d. silicone tubing, via a flex pump, to the mixing box of the diluter. III. Fish Fathead minnows (Pimemales mmelas) were raised in outdoor ponds and were acclimated to well water in indoor holding tanks at least two months prior to each experiment. Aquaria were stocked with 30 fish each, in experiment I, and 20 fish each, in all remaining experiments. Fish were fed a diet of crushed Purina Trout Chow (residue-free) every day during acclimation and experiment I, but only two or three times during each remaining experiment, generally 25 g every four days. IV. Water Supply Water for the fish holding tanks and the diluter sysatem was supplied from a well. It was pumped into large storage tanks and then delivered to each system via PVC piping. The water temperature averaged 13°C. Total hardness was 300 mg/2. as CaC03, pH was 7.4 and dissolved oxygen averaged 8.6 mg/il. 15 V. Exposure Methods PCB-contaminated sediment was introduced into the mixing box of the proportional diluter where it was mixed with incoming well water. The suspended-sediment solution was delivered into separate diluter boxes from where it was simultaneously siphoned with well water into four separate aquaria, each receiving a different amount of suspended sediment. The control aquaria in the system received water without suspended sediment. The bottom of each aquaria was screened off (approximately 1 cm mesh) to prevent fish from directly contacting any contaminated soil which had settled out. Aquaria with a 50-1iter capacity were used initially, however, due to a high rate of settling of suspended material, smaller aquaria (35 liter) were med. The me of smaller aquaria decreased the total flushing time, therefore decreasing the amount of time the suspended material had to settle out and reducing the amount of sedimentation. Reducing sedimentation was important because non-suspended sediment could still release PCBs into the water. Therefore, by minimizing sedimentation it was assumed that most of the P088 in the water column were released by the suspended sediment. Also the fish spent much of their time at the bottom of the aquaria where high concentrations of PCBs released from sedimented soil could occur. VI. Sampling Fish, water and suspended sediment samples were taken on the same day. Each experiment was sampled four or five times at approximately three-day intervals. Three to five fish were sampled from each tank. Fish from the same tank were wrapped together in aluminum foil and immediately frozen. Water was sampled with a bent glass siphon tube so that the same depth could be sampled in each tank. Samples were then filtered in buchner funnels using Whatman GF/C glass microfiber paper (effective retention = 1.2 um) 16 (Campbell and Elliott, 197 5). The filters were pre-extracted in a hexane/acetone solution and dried to a constant weight in an oven at 110°C. Suspended sediment samples were collected on preweighed filters by filtering 500 or 1000 mi. of water. Two suspended sediment samples were taken for each tank. One was dried and weighed to determine suspended sediment concentration expressed as mg/IL. The other filter containing the sample was placed in a small Erlenmeyer flask and capped for later PCB analysis. Experiment I lasted 24 days with samples taken on days 2, 4, 8, 12 and 24. Experiments II and III lasted 14 days and the samples were taken on days 3, 6, 8, 11 and 14 in experiment 11 and days 3, 8, 10, 12 and 14 in experiment In. Experiment IV ran for five days with samples taken on days 1, 2, 3, and 5. Experiment V lasted six days with sampling done on days 1, 3, 5 and 6. VII. Analytical Methods Filtered water samples (1 liter) were extracted with three 100 mil. portions of hexane in a 2000 mi. separatory funnel. The combined extracts were filtered through anhydrous sodium sulfate then evaporated to approximately 1 mi. for gas chromatography analysis. The filters containing the suspended sediment samples were kept moist until extracted four times with 25 mi. portions of a 1:1 hexane/acetone mixture on a wrist shaker. To increase extraction efficiency dry soil samples were wetted thoroughly by shaking with water before extraction (Bellar and Lichtenburg, 1975; Richardson and Epstein, 1971). The extracts were combined in a separatory funnel and washed with 100 ml. distilled water. The wash water separated from the extracts was back extracted with 25 mt. hexane, which was added to the other extracts. The combined extracts were then washed with three 100 m2. portions of distilled water. The washed extract was then filtered 17 through anhydrous sodium sulfate in a fritted glass buchner funnel. The filtered extract was evaporated to approximately 1 ml. The florisil used for cleanup was prewashed with 50 mil. hexane then oven dried at 130°C. A 2 cm i.d. column was packed with a glass wool plug, 11.5 cm florisil and 2 cm. anhydrous sodium sulfate. The column was prewashed with 50 mt. hexane and the sample was added and rinsed with a few milliliters of hexane. The column was eluted with 100 mil 6% diethyl ether in hexane. The eluate was evaporated to a suitable volume for gas chromatography analysis. This procedure was slightly modified from that of Goerlitz and Law (1974). Environmental soil samples were more difficult to extract because of interfering substances. A column extraction technique was used where 50 g of the air-dried soil was placed in a 2 cm. i.d. chromatographic column and eluted with 250 ml. of a 1:1 hexane/acetone mixture. The eluate was washed with 300 mil. distilled water which was back-extracted with 20 mi. of 15% methylene chloride in hexane. The extract was added to the original eluate and the combined extracts were washed twice with 100 ml. distilled water. The washed extract was filtered through sodium sulfate and evaporated to approximately 1 m9. A florisil column as described previously was used for cleanup. The column was eluted with 100 mil. 6% diethyl ether in hexane and the eluate evaporated to approximately 10 mi. for gas chromatography analysis (U.S.E.P.A., 1977). Interference due to elemental sulfur was eliminated by treatment with mercury at this point by adding a small dr0p of mercury to the final solution and mixing well (Goerlitz and Law, 1971). The fish were thawed, blotted dry and weighed. Whole fish extractions were made in hexane with a tissuemizer. Alternate methocb were mortar and pestle grinding with soxhlet extracation in hexane or in a ground-glass tissue grinder with hexane. A total extraction volume of 100 ml was used, of which 10 1.8;} mi. was placed in pre-weighed pan, air-dried and weighed for percent fat analysis. The remaining extract was evaporated to approximately 5 mt. for column cleanup. A florisil column, as used for water and soil cleanup, was also used for fish samples and likewise eluted with 100 mi. 6% diethyl ether in hexane. The eluate was evaporated to between 1 and 10 ml. for gas chromatography analysis. .1, . Computer-assisted gas chromatographic analysis was used for PCB #5219550 : identification. The glass column was 1.8 m (6 feet) long and packed with 3% SE- 30 on 60-80 mesh Gaschrom Q. Detection was by electron capture with a Nickel-63 detector. Column temperature was 180°C. for Aroclor 1254 analysis and 160°C. for Aroclor 1242 analysis. Inlet temperature and detector . temperature were 225°C and 350°C., respectively. The detectable limit for J" Aroclor 1254 was approximately 0.5 mg/L The total peak area was used as a Lmeasure of total PCB. Specific peaks were not included in total peak area due to interference in the samples (DDE interference was confirmed by standard DDE injection on the gas chromatograph). These peaks were also excluded from )1. )2!) {.175 f) \individual peak areas as measured by the computer. I standard total peak area. Total peak area was calculated, as the sum of the 4.. The presence of Aroclor 1242 in the Shiawassee River sediment was confirmed by high pressure liquid chromatography using a C-18 reverse phase mode with Octadecyl Silane as stationary phase and methanol as the mobile phase. Flow was 1 cc/min., injection volume was 20 112., R = 0.20, wavelength of 235 nm. RESULTS 1. Suspended Sediment and Water In experiments I and 11, using Arboretum soil dosed at 100 ug/g with Aroclor 1254, suspended-sediment samples yielded levels of Aroclor 1254 higher than the dosage level. The same phenomenon occurred in experiment III with greenhouse soil closed at the same level. In the first three experiments Aroclor levels in the suspended sediment varied widely over the course of the experiment (Table 2) but there was no significant difference (Appendix 1, 90% c.f.). It also was noted in each experiment that some suspended material settled to the bottom of the aquaria while smaller, lighter particles remained suspended in the water column. When water samples were siphoned to obtain suspended-sediment samples, it was the smaller particles that made up the sample. Considering this, it was theorized that the smaller particles (i.e., silt and clay), which probably contained most of the organic matter, adsorbed most of the P088. This theory was confirmed by analyzing each fraction (sand, silt and clay) for percent organic content and PCB concentration (silt and clay fractions were combined for PCB analysis due to small sample sizes). In each soil sample analyzed, the silt and clay fractions were composed of more organic matter than the sand fraction and correspondingly had higher concentrations of PCB. This pattern was especially evident in the Shiawassee River sediment (Table 1). In experiment IV the greenhouse soil was dosed with 2,5,3',4'- tetrachlorobiphenyl at 50 ug/g, and the suspended-sediment samples yielded tetrachlorobiphenyl levels near the dosage level. The greenhouse (high organic) 19 20 News: 393: :3: RES 65: 2a 2a 2s a: :62 > 3-3 8-3 2-: 31.6 693: 8 S 2. S. :32 a $73 378 “5.3: 2:6: «as: a: «.2 a: .2: :52 E «2A.: «3.2: $23 :32 «was: a: 3m 2“ «S :82 : 3N1; RTE. 2.23 378 65: 3H 2: S: 3: 562 _ m w m m 36> 66:52 69:62 63.62% 626366 “5.5.695“ BB 3 36:563. 606535 :_ 3263:6300 mom .N 036,—. 21 soil contained 10% more silt and clay combined than did the Arborteum (low organic) soil. The combined silt and clay fractions of the high organic soil contained a higher level of PCB than did the combined silt and clay fractions of the low organic soil. It was speculated that more of the greenhouse soil would stay in suspension due to the higher percentage of silt and clay, therefore allowing more opportunity for desorption of PCB from soil to water resulting in high PCB concentrations in the water. However, there was no significant difference in PCB concentration in suspended sediment between experiments II and III (Appendix 1, 9096 c.f.). Also water concentrations of PCB were compared for experiments II and III and no significant difference (9096 c.f.) was found except in aquarium number 3 (corresponding aquaria in each experiment were compared, see Table 3 and Appendix 2). Therefore the higher percentage of silt and clay did not greatly affect the suspended sediment and water concentrations of PCB. In all experiments PCB concentrations in the water fluctuated in response to the total amount of suspended sediment present. As the suspended sediment concentration increased, the PCB concentration in the water increased and decreased when suspended sediment decreased. In experiment In using the high-organic greenhouse soil, the suspended sediment levels were slightly lower than in experiment 11 using the low-organic, Arboretum soil (Table 4). However, there was no significant difference (90% c.f.) in suspended sediment concentrations between corresponding aquaria in the two experiments (Appendix 3). Correspondingly, the PCB concentrations in the water were lower in experiment 111 than in experiment II. The small difference between PCB concentrations in the water samples of the two experiments was felt to be due to the difference in suspended sediment concentration but there was no signficant difference (9096 c.f.) in the concentrations of PCB in the water samples of the two experiments. 22 .E n «A v 83:53 32:9: .333 ©9823 * 313 «.24 3:42. v.3..." 31. swam 3 3 3 3 3 :82 > «5-; «54.... 33.: ~23"... a mafia a... N... a... ea c 5m: 2 SE... 3...“... 3...”... 33.... 3;... «as; m... I. 3. a... a... :82 E 3...; SI... BIZ. 3E... a «yam S. 3. a... a... o :82 a 9:... BIZ. fig... $4-: a omega 3 a... 5 3 a :82 _ m o hr . i N 35.30 33> 895:2 #5832 E:_.3:o< Botmzfim «58:09am .m mass 3m 3 ‘.332 of 5 3233:0280 mom 23 «3 SA E4: 92.: a omfim 3 S «a 2: a :32 > m-” I To 2-fi Tc ems; e m a. E a :82 >— N: mg 3-2 3...: a mass: 3 2 2. S a :82 5 2-2 3-2 2-: 3-3 a; «use a S E 3 m 5o: a 3...: 8-2 SEN «Si 2 mafia 2 2 a: 2: e :32 _ m e a in 3.550 83> .2552 69:52 83.33:» EoSmSSm «582on Qua: 9.2355280 «soc—Bow ooocoamzm .v 033. f” l‘. E _ + / E‘.l;’/'.‘ré;()~ \ 3:1 The concentration of tetrachlorobiphenyl in water samples of experiment IV, using greenhouse soil dosed at 50 u g/g, was similar to Aroclor concentrations in experiments 11 and 111 (Table 3) even though the suspended sediment concentrations in experiment IV were much lower than in the previous two experiments (Table 4). This was probably due to the relatively high water solubility of the tetrachlorobiphenyl compared to the penta— and hexachlorobiphenyls that constitute most of the Aroclor 1254 mixture. II. Isomer Distribution Analysis Statistical comparison of average percent peak areas between each of the parameters (fish, water suspended sediment) and standard chromatograms revealed a pattern of distribution between the three parameters that. can be seen generally in Figure 1. Peaks 1 and 2 are mixtures of tetrachlorobiphenyls, 3, 4, 5 and 6 are mixtures of pentachlorobiphenyls, and 7, 8 and 9 are mixtures of hexachlorobiphenyls. Peaks. 1, 2, 3, 4, 6, 7, 8 and 9 were used for comparison when no interference was present. Chromatograms of samples in experiment I after 24 days were selected randomly and compared with an equal number of standard chromatograms (usually from the same day that the samples were analyzed on the G.C. because G.C. response varied from day to day). Table 5 gives the results of the calculated 'student t' values compared with the 'critical t' value at the 90% confidence level. In the suspended sediment, peaks 2 and 3 (tetra— and pentachlorobiphenyl) were significantly lower in percent area than standard peaks 2 and 3 and percent area of peak 8 (hexachlorobiphenyl) was higher than in the standards. For peaks 1, 4, 6, 7 and 9, there was no significant difference in percent peak area when compared with standard peak area. 25 Figure 1. Sample Chromatographs of PCBs in Fish, Water, Soil and a Standard Aroclor 1254 Chromatograph 26 S TANDARD WATER W 1234 6 7 8 9 PEAK N0. 1234 6 7 8 9 max 30 4455 5 6 6 6 CHLORINE no. 4655 S 6 6 6 cumulus so SOIL FISH r"... UK M--- all -W 1234. 6 7 8 u PEAK NO. 123/. 6 78 9 PEAK N0. 1.1.55 5 6 6 6 CHLORINE no. 1.455 5 6 6 6 CHLORINE .vo. 27 flaconanSanxo: O flzcocnfioBEoficoa Ehcozafioaofioah$am 2... ..... 8.. .2. .2. 2... 2... 2... was: 2... 2... 2.. ...... a... 2.. 2... 2... 8.8.5.. 288.... 8.. a... .22 2.2. 2.2 2.2 ...... a... :8... 8.885 .2. 2... .... S... E... 2... 2.. a... 8.5.5.. 288.... a... 2... 2.2 8.... 8.2 8.... 2... 2.. :52 8.... 2.... 2.... 2... 2.. 2.... 2.. 2... S... 8.8. .2. 2... 2... 2.. 2.. 2.. 2.... 2... 8.5.5.. 2885 2... a... an... 2.... 8.... 2.2 ..... 2.. 85.. 8.885 .2... a... s... 2... ...... 2.. S .N 2... 8.5.5.. 8.885 2.2 2.. S... 2.... .22 2.2 5.. 8.2 :82 8...: a... 2.. 2.. 2.... .2. 2.. 2.. 2... 2.8. 2.... 2 .. 2 .. 8.. 2.. 2 .N 2... 2... 8.5.5.. 888% 8... 2... 2.2 8.2 2.2 8.2 8.. 2... 852 8.885 2.... 2.. 2. .. 8.. .m .. 2.. 2... p... 8.5.5.. 2886 .8588 8... S... 2.2 2.... .22 .2... 2... 2... :82 8.8%.... 6.. ow 6.. no a. a” an a. 2...; on»... 8282 .8. 8.8.8... 5358 8.... 2...". 8. .e u .. .23 2...; 28...... .8. . p.885 5 82.. .8. 2.88.. .o 8885.... .... 2...... 28 In water samples peak 1 (tetrachlorobiphenyl) represented a significantly higher percent area, while peaks 4, 6, 7 and 8 (penta— and hexachlorobiphenyls) were significantly lower in percent peak area compared to standards. Peaks 2, 3 and 9 (tetra-, penta-and hexachlorobiphenyls respectively) showed no significant difference. In fish samples the only significant difference was found in peak 6 (pentachlorobiphenyl) which was slightly higher in percent peak area than the corresponding standards. No difference was found for peaks 1, 2, 3, 4, 7, 8 and 9. III. PCB Accumulation by Fish As expected the highest suspended sediment loads resulted in the highest PCB concentration in the fish. The highest suspended sediment loads were received by aquarium 2 followed by aquaria 3, 5 and 4. In experiments IV and V the order was reversed between aquaria 4 and 5, but in all experiments the two aquaria (4 and 5) were very similar (Table 4). Likewise, the concentration of PCBs in fish were highest in fish from aquarium 2 followed, in decreasing concentration, by fish in aquaria 3, 4 and 5 (Table 6). The PCB concentration values in fish were compared using the last sample day values. Figure 2 shows the general pattern of PCB accumulation by fish in the four different suspended sediment loads in experiment I. By day 2 they had all accumulated approximately the same amount of PCB. By day 4 fish from aquaria 3, 4 and 5 were similar in PCB concentration but fish in aquarium 2 were beginning to accumulate higher loads of PCB. By the end of the experiment fish in aquaria 4 and 5 were still increasing their concentrations of PCBs but there was not much difference between. the two groups. Fish from aquarium 3 were not sampled after day 8 because of fish mortality due to entanglement in the siphon tube. Fish in aquarium 2 had accumulated approximately twice as much 29 ..-.. .. ..-... 2-.. ..-2 .. .28.. ... .... 2 2 .. 8o... > 2.-.... ...-.... «...-.... ...-.... .. 28.. ... .... .... 2... .. 8m... >. ...-.... ...-.... ...-.... ...-.... ...-.... 28.. .... .... ...... .... .... :8... ... 2.. 2-.. ..-2 2-2 2.2-.. 28.. 2 2 2 2 ... :8... .. ...-2 2-2 ...-... ..2-2 .. 28.. .. 2.. 2 2. .. =8... . m . n 2 .9380 2...; 8.23,. .6252 532.39.. 3032.53 25......on AME ... 2.05.3.2... noun.— 5 2.38m 3:... 0... 8o... 2.355 cm... ... 22.955050 mum .. 28... 30 Figure 2. Aroclor 1254 Uptake by Fathead Minnows (Pimephales promelas) in Four Aquaria in Experiment I 31 mu cw an an .N ON a. o— h— @— ._._._._._._._._._. w>¢o 2.. m— v— m— N— _._.—._ ME: :2 m a h o w v m N n . —. _. ._._._._._._._._._ D 0 _x mm c‘qd_m..‘q.nvl‘ + e 23E¢o zu u:_» m r . H EoEtoaxm Eu EoEtoaxm Tm Y! I OI. Te OZ ‘HSIJ N1 #931 8013086 'O/Ofl 40 Figure 6. Aroclor 1254 Uptake by Fathead Minnows (Pimephales romelas) in Aquarium No. 5 in Experiments 11 (Low Organic Soil} and III (High Organic Soil) ’ 41 Experiment II on Experiment III + .J — 11l'TIW‘I'l'lfi‘l‘l‘T'T‘l‘T'Tfi_ Z 9972 a as 92 92 t2 :2 oz 919! n 2! m '0/on ‘HSIJ NI vSZI 8013086 TIME IN DRYS 42 Table 7. Results of Approximate t' Test on Slopes of Uptake Regression Lines from Experiments II and III Experiment Aquarium .. Number Number Slope t' Critical t' V II 2 5.12 III 2 0.56 8.555 3.182 2.89 11 3 2.63 III 3 0.49 6.544 2.776 3.47 H 4 1'34 3.623 2.776 3.53 III 4 0.34 H 5 0.98 III 5 0.24 3.071 2.776 3.84 t a /2,v a =0.05TABLE 7 43 25.5 25.2 25.5 25.5 m 25.5 25.: 25.: 25.5 5 25.5 25.2 25.2 25.2. m. 25.5 25.3 25...: 25.5 N > >_ E a eon—5.2 23:52 Hose—mafia 53.233» 2530 33 .83: 335335 255.5 :2: 255.5 :22 69.9—50.5 0255520 om»... mom 3:590 26.— 8323:... > .25 >~ .E .= 555263 Ea...“ $3.; 3355:8585 no 52.39200 .m 2an 44 Comparison of the transfer coefficients in Table 9 shows transfer coefficient values in experiment 11, two to six times higher than in experiment - 1]]. Again, the data indicate that the difference in PCB accumulation by fish is due to more than just PCB concentrations in the water medium. VII. Experiment IV In this experiment the high organic greenhouse soil was dosed with 2,5,3',4'— tetrachlorobiphenyl at 50 ug/g. Results show that suspended sediment levels were lower than in any of the other experiments (Table 4). Water concentrations of the tetrachlorobiphenyl were also lower than the PCB concentration in the water of other experiments (Table 3). Subsequently, PCB uptake by fish was the lowest of all the experiments (Table 6 and Figure 7). Comparing bioconcentration values for experiments HI and IV in Table 8 (both experiments with high organic soil), show that they are similar, even though experiment IV ran less than half the time of experiment III. A better comparison is with transfer coefficients which are time independent and take into account the rate of uptake. Transfer coefficients in experiment IV are much higher than those in experiment III (approximately two to four times, see Table 9) and more closely resemble transfer coefficients in experiment 11 (low organic soil). So even though overall concentrations were lower in experiment IV the fish accumulated relatively high levels of tetrachlorobiphenyls. VIII. Experiment V There was no statistically significant difference (90% c.f.) between Aroclor 1242 concentrations in the Shiawassee River suspended sediment used in experiment V and Aroclor 1254 concentrations in suspended sediments used in experiments II and 111 (Appendix 1). The Aroclor 1242 concentratiom did average 45 Figure 7. 2,5,3',4'-Tetrachlorobiphenyl Uptake by Fathead Minnows (Pimephales promelas) in Four Aquaria in Experiment IV 46 (\l ._. —LO L—d’ B —!”3 El SIN N... NV... m cmNN N... 3.: 0 RN: «:0 3.: N SN 0.: «N; N >— 03. m... «N... m 5mm «.0 3.: 0 :8 m6 3.: N mam a... 3.: N E No: 5.: 8.: m NNNN 0.: 3.: 0 SN” a... NoN N 83 a... N26 N : 2:22:80 3B 3 .883 5 00.30: 580083“ 80:52 80:52 8.88.? :03038080 m0m No 0005 838:3. 2882092.“ > 0:: >_ .E .= 2:05.892— ..8 3:22:30 88:95. «0 82.80800 .N 030:. 48 higher than the Aroclor 1254 concentrations but there was a great deal of fluctuation (Table 2). Suspended sediment concentrations were similar in experiment V to those in experiments 11 and III with the exception of aquaria 2 which received very high -loads of suspended sediment due to a mechanical failure in the diluter system (Table 4). Other than that, suspended sediment conditions and PCB concentrations in the sediment were simlar between experiments II and 111 (using Aroclor 1254) and experiment V (Aroclor 1242). Water concentrations of PCB were much higher in experiment V than in experiments 11 and III (three to ten times, see Table 3). A higher concentration of PCBs in the water was expected due to the low organic content of the Shiawasee River sediment (1.096) and the higher content of tetrachlorobiphenyls in Aroclor 1242 making it more water soluble. Therefore it was assumed that fish would be able to accumulate more PCBs, however, this was not the case (Figure 8). Bioconcentration values were lower in experiment V (Table 8) than in any of the other experiments. This was not too surprising because of the shorter duration of experiment V. Transfer coefficients, on the other hand, take into account the time factor but these too are lower in experiment V than in any of the other experiments (Table 9). 49 Figure 8. PCB (as Aroclor 1242) Uptake by Fathead Minnows (Pime hales rome las) in Four Aquaria in Experiment V (Sh1awass ee R1verSe Se1ment) 50 w>¢o z_ m:: x m EDE