o.‘ Izeéy : '32:. . that! 51.3 . I. v. .. .1 70:1 r 23:... I) r» :\«.1‘ .J‘nili... ,5... :3. v.1 16;...aav V. r. e I.- .21..:l . . 7 ii: . .3; .. V.(l 6...? .3 1 av... 13:1; $4141....zn‘ .l ulvit-AV 7..» .ti... :1.-. t». I..D.t.4r.€ 5. .lt: .‘..rt 1. ‘1...»1 .0 .A a 9.. $1 .I....o.i v: clv.;.\.vwro .9...I-L. :.vl...u »a.p. In t 1..) 5‘- r.:. .szi‘. :In... 1‘ \ranal. It]: I: a 5). 3‘! nil.‘ 'THFSIS‘ ANSTATE UIN VEHSITY LIB IIIIIIIIIIIIIIII III I III IIIIIIIIIIII 31293 00881 2772 This'is to certify that the thesis entitled The Role of Atmospheric Deposition of Contaminant Metals to the Great Lakes: Deduced from Sediment Cores presented by Adam W. Heft has been accepted towards fulfillment of the requirements for Master's Geological Sciences degree in WW/a Major professrg Datefi flPIB/C Q3 O-7639 MS U is an Affirmative Action/Equal Opportunity Institution ‘7 LIBRARY Michigan State University # PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or baton date due. DATE DUE DATE DUE DATE DUE UAR I 4 ‘995' ‘i ._, ‘— " usu Is An Affirmative Action/Equal Opportunity Institution czwmmpnwt _,__'——a#fi THE ROLE OF ATMOSPHERIC DEPOSITION OF CONTAMINANT METALS TO THE GREAT LAKES: DEDUCED FROM SEDIMENT CORES By Adam w. Heft A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 1 993 David T. Long, Advisor ABSTRACT THE ROLE OF ATMOSPHERIC DEPOSITION OF CONTAMINANT METALS TO THE GREAT LAKES: DEDUCED FROM SEDIMENT CORES BY Adam W. Heft In recent years, contamination of aqueous environments by atmospheric source pollution has become a major concern. Atmospheric deposition of organic contaminants has been documented to be beyond a doubt the most significant input process to the Great Lakes. It is the purpose of this study to determine if atmospheric deposition is the most significant input process for the trace metals As, Cd, and Pb. Sediment cores were collected from Lakes Michigan and Ontario. Sediment thus collected was subjected to a total metal extraction using a microwave-nitric acid digestion technique, and analyzed using graphite furnace atomic absorption spectroscopy. Results of this study indicate that background concentrations for all trace metals are within the range for uncontaminated soils. From 210Pb corrected data and comparisons of excess concentrations and atmospheric deposition rates, it was determined that atmospheric deposition is the most significant input process of As, Cd, and Pb to the Great Lakes Fiegion. This thesis is Dedicated to all those who live in and love the Great Lakes Region ..... ACKNOWLEDGEMENTS There are many people I would like to acknowledge for their help and/or support of this project. First, I am grateful to the United States Environmental Protection Agency and the Great Lakes Protection Fund for their grants to the Universities of Minnesota and Michigan State which funded and made this study possible. I would like to thank the Captain and crew of the Research Vessel Lake Guardian, from the decks of which the data used in this study were collected. I would also like to thank my thesis advisor, Dave Long, for his guidance in this project, and for his criticisms. I am grateful to the members of my thesis committee, Peggy Ostrom and Graham Larson for their comments, and suggestions regarding this work. I would like to thank the other "lab rats" (Joe McKee, Jane Matty, and Eric Roth) for their help in understanding the instruments and techniques necessary for this study. Thanks as well to Anton Spirenburg, Bill Sitarz, and Jon Kolak for all their help on the early phases of this project. A big thank you to my undergraduate thesis advisor Dave Matty for his encouragement to me to continue my education past my BS» yes, it was worth it. I am grateful to my parents and grandparents for all their support and understanding through the years. Finally, a special thank you to my fiancee Sue Sorensen for her patience and support during all the rough spots in this work. ii TABLE OF CONTENTS List of Tables ......................................................................................................................... vi List of Figures ........................................................................................................................ vii I. INTRODUCTION ............................................................................................................. 1 BACKGROUND .......................................................................................................... 3 STATEMENT OF THE PROBLEM ............................................................................... 7 HYPOTHESIS ............................................................................................................ 9 ll. PREVIOUS STUDIES ................................................................................................. 11 ATMOSPHERIC DEPOSITION .................................................................................. 11 METAL CYCLING ...................................................................................................... 15 III. METHODS .................................................................................................................. 26 SAMPLING ............................................................................................................... 26 CLEAN PROCEDURES ............................................................................................ 30 CHEMICAL EXTRACTION ......................................................................................... so QUALITY ASSURANCE PROCEDURES .................................................................... 32 Iv. RESULTS ................................................................................................................... 33 ARSENIC ................................................................................................................. 33 CADMIUM ................................................................................................................ 4o LEAD ....................................................................................................................... 47 v. DISCUSSION .............................................................................................................. 54 GENERAL OBSERVATIONS .................................................................................... 54 BACKGROUND CONCENTRATIONS ........................................................................ 55 SEDIMENT FOCUSING AND 210Pb DATING .............................................................. 64 EXCESS CONCENTRATIONS .................................................................................. 76 VI. CONCLUSIONS ......................................................................................................... 82 APPENDIX A ..................................................................................................................... 64 APPENDIX B ................................................................................................................... 108 APPENDIX c ................................................................................................................... 112 APPENDIX D ................................................................................................................... 130 REFERENCES ............................................................................................................... 135 iii LIST OF TABLES Tablet: ........................ Minimum, Maximum, and Excess As concentrations ........................... 39 Table 2: ........................ Minimum, Maximum, and Excess Cd concentrations ........................... 46 Table 3: ........................ Minimum, Maximum, and Excess Pb concentrations ........................... 53 Table 4: ........................ Compilation of As, Cd, and Pb data from the literature ......................... 56 Table 5: ........................ Background concentrations of sediments and soils ............................ 57 Table A1-1: .................. EPA#11 Location and Description .................................................... 84 Table A1-2: .................. EPA#18 Location and Description .................................................... 86 Table A1-3: .................. EPA#19 (M) Location and Description ............................................... 88 Table A1-4: .................. EPA#23 Location and Description .................................................... 89 Table A1-5: .................. EPA#27 Location and Description .................................................... 90 Table A1-6: .................. EPA#34 Location and Description .................................................... 91 Table A1-7: .................. EPA#4O Location and Description .................................................... 92 Table A1-8: .................. EPA#41b Location and Description .................................................. 94 Table A1-9: .................. EPA#19 (0) Location and Description ............................................... 95 Table A1—10: ................ EPA#25a Location and Description .................................................. 96 Table A1-11: ................ EPA#40a Location and Description .................................................. 97 Table A1-12: ................ EPA#648 Location and Description .................................................. 99 Table A1—13: ................ EPA#LG1 Location and Description ................................................ 101 Table A1-14: ................ EPA#LGZ Location and Description ................................................ 104 Table A1-15: ................ EPA#55 Location and Description .................................................. 106 Table A2—1: .................. QA Checks: AS replicate results ..................................................... 108 Table A2-2: .................. QA Checks: Cd replicate results ..................................................... 109 Table A2-3: .................. QA Checks: Pb replicate results ..................................................... 110 Table A2-4: .................. QA Checks: SRM results ............................................................... 111 Table A3-1: .................. EPA#11 trace metal and porosity data ............................................. 1 12 Table A3-2: .................. EPA#18 trace metal and porosity data ............................................. 1 13 Table A8-3: .................. EPA#19 (M) trace metal and porosity data ........................................ 1 14 Table A3-4: .................. EPA#23 trace metal and porosity data ............................................. 1 15 Table A3-S: .................. EPA#27 trace metal and porosity data ............................................. 1 16 Table A3-6: .................. EPA#34 trace metal and porosity data ............................................. 117 Table A3-7: .................. EPA#4O trace metal and porosity data ............................................. 1 18 Table A8-8: .................. EPA#41 b8 trace metal and porosity data ......................................... 1 19 iv Table A3-9: .................. EPA#19 (O) trace metal and porosity data ........................................ 120 Table A3-10: ................ EPA#25a trace metal and porosity data ........................................... 121 Table A3-11: ................ EPA#40a trace metal and porosity data ........................................... 122 Table A3-12: ................ EPA#64a trace metal and porosity data ........................................... 123 Table A3-13: ................ EPA#LGt trace metal and porosity data ........................................... 124 Table A3-14: ................ EPMLGZ trace metal and porosity data ........................................... 125 Table A3-15: ................ EPA#55 trace metal and porosity data ............................................. 126 Table A3-16: ................ Normalized data for As .................................................................... 127 Table A3-16: ................ Normalized data for Cd ................................................................... 128 Table A3-16: ................ Normalized data for Pb ................................................................... 129 LIST OF FIGURES Figure 1: ...................... Diagramatic representation of terms used in profiles ............................. 2 Figure 2: ...................... Map of Great Lakes Region (after Upchurch, 1972) ............................ 12 Figure 3: ...................... Diagram showing processes affecting metal cycling (after Matty, 1992) ............................................................................. 16 Figure 4: ...................... Diagrammatic representation of double walled complex (after Long, 1990) ............................................................................. 21 Figure 5: ...................... Lake Michigan sampling locations (after Eisenreich et al., 1990) .......... 27 Figure 6: ...................... Lake Ontario sampling locations (after Eisenreich et al., 1990) ............ 28 Figure 7: ...................... EPA#‘s: 11, 18. 19, and 23 As profiles ............................................. 34 Figure 8: ...................... EPA#‘s: 27, 34, 40, and 41b As profiles ........................................... 35 Figure 9: ...................... EPA#‘s: 19. 25a, 40a, and 64a As profiles ........................................ 36 Figure 10: .................... EPA#‘s: LG1, LGZ, and 55 As profiles .............................................. 37 Figure 11: .................... EPA#‘S: 11, 18, 19, and 23 Cd profiles ............................................. 41 Figure 12: .................... EPA#‘s: 27, 34, 40, and 41b Cd profiles ........................................... 42 Figure 13: .................... EPA#‘s: 19, 25a, 40a, and 64a Cd profiles ........................................ 43 Figure 14: .................... EPA#‘S: LG1, LG2, and 55 Cd profiles .............................................. 44 Figure 15: .................... lsslhf ............................................................................................... 45 Figure 16: .................... EPA#‘s: 11, 18, 19, and 23 Pb profiles ............................................. 48 Figure 17: .................... EPA#‘s: 27, 34, 40, and 41b Pb profiles ........................................... 49 Figure 18: .................... EPA#‘S: 19. 25a, 403, and 64a Pb profiles ........................................ 50 Figure 19: .................... EPA#‘S: LG1, L62, and 55 Pb profiles .............................................. 51 Figure 20: .................... slgogr ............................................................................................. 52 Figure 21: .................... Comparison of background concentrations of sediments and uncontaminated soils (from Connor and Shackette, 1975) ................... 58 Figure 22: .................... Background concentrations of As from Lakes Michigan and Ontario 60 Figure 23: .................... Background concentrations of Cd from Lakes Michigan and Ontario... 61 Figure 24: .................... Background concentrations of Pb from Lakes Michigan and Ontario... 62 Figure 25: .................... Comparison of As, Cd, and Pb profiles for EPA#25a .......................... 63 Figure 26: .................... EPA#11 profile showing correspondence of redox zones and background concentration peaks ....................................................... 65 Figure 27: .................... EPA#19 (M) profile showing correspondence of redox zones and background concentration peaks ....................................................... 66 Figure 28: .................... EPA#27 profile showing correspondence of redox zones and background concentration peaks ....................................................... 67 vi Figure 29: .................... EPA#34 profile showing correspondence of redox zones and background concentration peaks ....................................................... 68 Figure 30: .................... EPA#40 profile showing correspondence of redox zones and background concentration peaks ....................................................... 69 Figure 31: .................... EPA#41 b profile showing correspondence of redox zones and background concentration peaks ....................................................... 70 Figure 32: .................... EPA#19 (0) profile showing correspondence of redox zones and background concentration peaks ....................................................... 71 Figure 33: .................... EPA#64a profile showing correspondence of redox zones and background concentration peaks ....................................................... 72 Figure 34: .................... Comparison of As excess concentrations .......................................... 77 Figure 35: .................... Comparison of Cd excess concentrations .......................................... 78 Figure 36: .................... Comparison of Pb excess concentrations .......................................... 79 Vii I. INTRODUCTION In recent years the fluxes of trace metals to the natural environment have increased dramatically. This increase has led to an upsurge of interest in atmospheric contamination processes by environmental scientists and governmental authorities alike (Barrie et al., 1987; Tessier and Campbdl, 1987). The most dramatic increase in trace metal concentrations was documented to have occurred during the period from 1957-74 (Rybak et al., 1989), although anthropogenic contamination has been significant since 1933 (Mueller et al., 1989). Christensen and Goetz (1987) indicate contamination in Southern Lake Michigan began in 1894. The contamination of air and water by persistent toxic substances is one of the most important environmental issues concerning the Great Lakes region (Arimoto, 1989). The goal of this study is to evaluate the Significance of atmospheric deposition of certain heavy metals to the Great Lakes region. In order to accomplish this, a certain process will be followed to isolate the effects of atmospheric deposition. That process is outlined below. However, some terms must be defined at this time. First, this study will examine the background (or natural) concentrations of trace metals found in the Great Lakes. The background concentration is the average value of all samples below (inclusive) the background depth. Background depth is the depth in a core where the concentration of a trace metal becomes a relatively constant, minimum concentration (calculated for each of the trace metals separately). The excess concentration is defined as the average value of all samples above the background depth, less the value of the background concentration for a given core. Figure 1 is a graphic representation of these terms. The inventory of a metal is somewhat Similar to the excess concentration. Inventory is simply the total amount of 1 Concentration In polo 10 ~- Background depth Depth so ~- (cm) 40 .. 50 I Background concentration 60 ‘- Figure 1: Diagrammatic representation of terms used in profiles the metal in the core less the background concentration. It also represents the anthropogenic loading of the metal. The next step is to use 210Pb data to establish a focusing factor. The focusing factor is used to account for variable wdimentation rates. The details of how the focusing factor is established are presented later. By accounting for variable sedimentation rates, data from different parts of the Great Lakes can be compared. This is accomplished by dividing the inventory of the metal in a core by the focusing factor. The corrected data may then be compared. If the results are similar, then atmospheric deposition is the significant process for depositing trace metals into the Great Lakes region. At this time. however, there is 21°Pb data for only three of the cores. This may not be enough to establish conclusive results, so another method will be used to support the findings of this method. The supporting method will be a comparison of excess concentrations. By calculating and comparing the excess concentrations of all the cores, regional trends may become apparent. These regional trends will be examined to see if there is any correspondence with the geographic location of the core to the proximity of atmospheric sources. Finally, the excess concentrations of the trace metals will be compared to atmospheric deposition rates of the metals which were compiled by Eisenreich and Strachan, (1992). If the excess concentrations of the metals show a correspondence to the atmospheric deposition rates, then this will indicate that atmospheric deposition is the most significant depositional process of trace metals into the Great Lakes region. BACKGROUND The Great Lakes are a precious natural resource which represent the largest freshwater lake system in the world. For the past 150 years, man's activities have had an ever increasing impact on the quality of the water in the Great Lakes. Anthropogenic activities have disrupted the natural cycles. lndiscriminant use of pesticides and other chemicals, the use of lead as a gasoline additive, and heavy industrialization of the Great Lakes region have all played a part in the contamination of these lakes. Mining and related activities, waste incineration, fossil fuel combustion, and the automobile industry have been the greatest anthropogenic culprits. Until the latter portion of this century, the Great Lakes were regarded as being large enough to be unaffected by man's activities. This attitude changed as numerous studies revealed the wope of damage done. Since that time, there have been attempts to stop, mediate, or reverse the effects of contamination to this vast ecosystem. The Great Lakes region is a heavily industrialized and densely populated area, which contains 20% of the US population and 60% of the Canadian pOpUlation (Long, 1992). Many of the people living in this area are living directly on the shore of one of the lakes or connecting channels, or only a short distance inland. As a result, these people are easily affected by contaminants released into these waters. It has been estimated by the US Government Accounting Office that in one year 89,000 pounds of lead, 1,900 pounds of PCBS, and 900 pounds of mercury are discharged Legallyinto the Great Lakes basin (Schoonover, 1992). These estimates do not include illegal discharges such as the 200,000 pounds of inorganic mercury discharged into the St. Clair river during the 1960s and 19703 (Wood, 1971; Hamdy and post, 1985; Annett et al., 1972), agricultural and urban runoff, or atmospheric deposition. Assuming a volume of 2.28x1016 L for the Great Lakes, legal concentrations in the Great Lakes would be approximately 1.77 parts per billion (ppb) for lead, .003 ppb for PCB's, and .002 ppb for mercury, assuming an even concentration throughout the entire Great Lakes system. All of these values are far below the health guidelines for these contaminants set up by the federal government. In reality, there are certain areas (proximate to discharge points) where the levels are much higher, and other areas which are at lower levels. The amount of contaminants entering the Great Lakes due to these discharges is clearly unable to account for the total concentration levels reported in various areas in the basin. Of the storage capacity of metals in Lake St. Clair, 62% of the 690 metric tons of Cd, and 39% of the 3200 metric tons of Pb are a direct result of anthropogenic activities (Rossmann, 1988). During the time period 1960-1973, there were government warnings and bans on fish consumption for this region. Other regions reported similar problems: high mercury levels were also reported in fish and waterfowl in Ball Lake, Ontario (Annett et al., 1975). Johnson (1987) reported that most lakes in Ontario have anthropogenic loadings from 1.8 to 2.6 times the background levels, although Stephenson and Mackie (1988) reported that lakes within 20 km of Sudbury have much higher Cd concentrations than other lakes in central Ontario. Furthermore, the anthropogenic enrichment of As, Cd, and Pb are highest near industrial regions; whereas the enrichment of Hg is more widespread, which may be due to the higher volatility of Hg. The St. Clair River is just one example of a polluted channel in the Great Lakes region. The St. Mary's River, and the Detroit River are also heavily polluted. The fact that all of the connecting channels of the upper Great Lakes are polluted led the lntemational Joint Commission (IJC) of the US. and Canada to designate them (and other areas) as "areas of concern” (Nichols et al., 1991; Marsalek and N9, 1989). Calculations Show that the 1983 median values for worldwide emissions of trace metals into the atmosphere are as follows: AS 18.820 x103 kg/YI’; Cd 7,570 x103 kglyr; Hg 3.560 x103 kglyr; Pb 332,250 x103 kglyr. The emission of these trace metals into aquatic environments for 1983 was: As 9.4 x106 kglyr; Cd 9.4 x105 mm; Hg 4.6 x106 kg/yr; Pb 138 x106 kglyr (Nriagu and Pacyna, 1988). By preponderance of the evidence from these studies, it can be seen that man is the single most important factor in the biogeochemical cycling of trace metals. There are both organics and heavy metal contaminants in the Great Lakes. The organic contaminants (pesticides, chlorinated hydrocarbons, etc.) have no natural component in the environment. Because of this fact, it is relatively easy to discern locations where these organic contaminants tend to accumulate. Furthermore, it is also possible to determine which processes have an effect on the organics, and how great that effect is. Heavy metals, on the other hand,do have a natural component: crustal degassing, rock weathering, and volcanic emission (Glass et al., 1986). Because of these natural processes, heavy metals are commonly found in the natural environment (Matty, 1992; Edenbom et al., 1986). It is much more difficult to trace heavy metal behavior patterns in the aquatic environment. While there is a paucity of data on background metal levels, some do exist for Hg. According to Fleishcer (1970), most igneous and sedimentary rocks contain less than 200 ppb Hg. Furthermore, background levels of Hg in sediments range from .01 -.15 ppm (Jemelov and Asell, 1973) to .25 ppm reported in the Gulf of St. Lawrence (Zingde and Desai, 1981; Lowring and Bewers, 1978). Arsenic concentrations in soils are typically .140 mglkg and in sediments ~18=II5 mglkg (Farmer and Lovell, 1986). The difficulty in discerning anthropogenic contamination arises from the many different processes interacting on the metals and with each other in the lakes: and the fact that there is a natural background concentration whose magnitude is not always known (Arimoto, 1989). Furthermore, the natural metal concentrations entering the aquatic environment may change with changes in the chemistry of the source region (Prohic and Juracic, 1989). Part of the concern about environmental contamination stems from the fact that it is not usually known what amounts of the metals enter the Great Lakes due to anthropogenic activities. There are many minable deposits in the Great Lakes region: several gold deposits which have associations with mercury; deposits of zinc with cadmium; lead and arsenic deposits, and extensive deposits of iron, copper, and nickel. These deposits are all to be found within the Great Lakes basin area, and the Canadian shield rocks found therein. Besides these deposits, there are many areas where these metals were brought into the region by way of repeated glacial activity, especially during the Pleistocene. 7 STATEMENT OF THE PROBLEM Several metals, such as arsenic, cadmium, and lead, have long been recognized as a human health threat (T aymaz et al., 1984). These are nonessential elements for biological processes (Ozretic et al., 1990; Langston, 1982) and are toxic even in very low concentrations. The presence of these metals in humans and animals has been linked to cardiovascular disease, reproductive impalrrnents, brain damage, and various other problems (Furgesson, 1990: Nriagu, 1988; Long, 1992). For trace metals to have an impact on aquatic organisms, the metals must be in a form that is biologically available to them (Waldichuk, 1985); only in this form are they toxic (Nelson and Donkin, 1985). In many cases, trace metals are emitted to the environment in a form which is not harmful to organisms. There are many cases where mercury was emitted to the environment as elemental mercury and was transformed to methylmercury (D'ltri, 1992). Methylmercury is the most toxic species of mercury (Senaratne and Dissanayake, 1989; D'ltri, 1992). Mercury metal species often undergo some kind of transformation to a toxic form, caused by either biological (enzymatic) or nonbiological (chemical-physical) agents. A significant number of people have been affected by these metals, even as far back as Roman times, when trace metal concentrations in the environment were increased four to five times. This increase in available trace metals has even been linked to the fall of the Roman Empire through heavy metal poisoning (Urban et al., 1990). Increases of Pb have been confirmed in remote areas such as the large ice sheets, where only atmospheric deposition can account for the metals present. The Antarctic ice sheet has shown a 5x increase in Pb levels in the last 13,000 years (Boutron and Patterson, 1987). Furthermore, Ng and Patterson (1981) Showed a 300x increase in 3,000 year old ice from Greenland. It can be seen from past studies that the amounts of heavy metal contamination in the environment has been increasing to a dangerous level. These heavy metals are toxic in low concentrations to both humans and other organisms sharing our environment. As more people become aware of the dangers pom by contaminants, the more necessary it becomes to impose limits on anthropogenic emissions. Because of the problems caused by contamination, the International Joint Commission (lJC) has targeted the four metals As, Cd, Hg, and Pb; and 10 organic compounds: PCBs, benzo(a)pyrene, PAHS, HCB, mirex, dieldrin, HCHS, DDT, toxaphene, PCDDs and PCDFS, as critical contaminants in the Great Lakes region (Colbome et al., 1990; Eisenreich and Swackhamer, 1990; Eisenreich et al., 1990: Long, 1992). Some of these metals tend to bioaccumulate in fish and other aquatic organisms (this is indisputable for Hg), and, therefore, pose health risks to other organisms besides humans (Apsimon et al., 1990; Lindberg et al., 1987). Fish have been known to easily bioaccumulate metals and pesticides in high concentrations (T hommes et al., 1972; Seagran, 1970) which pose a threat not only to their own health, but to humans and predators such as herons and other birds. Since trace metals are not degradable, the threats they pose do not just go away over time. It can be demonstrated that some of the contaminants in the Great Lakes are not the direct result of anthropogenic activities in the Great Lakes region. Toxaphene, an organic compound which is on the UC critical contaminant list, was used only minimally in the Great Lakes Basin, and yet is found in high concentrations in Lake Michigan (D'ltri, 1992; Voldner and Schroeder, 1989). Toxaphene was used extensively in Southern states, and atmospheric wind currents brought it north. Toxaphene was applied extensively to cotton, and was used in the Great Lakes as a rough fish control (Rapaport and Eisenreich, 1986). The atmosphere is conceded to be the major source of organic contamination, at least for the international upper Great Lakes (Strachan, 1985: Swackhamer and Armstrong, 1986; Swackhamer et al., 1988). To a certain degree, this is probably the case for some, or all, of the trace metals that will be dealt with in this study. Therefore, in order to begin to bring the levels of heavy metals in the lakes to an acceptable level, the background vs. atmospheric inputs and surficial vs. atmospheric inputs, as well as the processes affecting those inputs, must be understood. It, for example, the background (natural) concentrations are underestimated, legislation might be introduced which would require the metal concentrations to be at a lower level than the background, in effect legislating natural processes, an unsound principle. This study represents a phase of the understanding of trace metal behavior which is necessary to begin to control heavy metal levels in the Great Lakes region. HY POTHESIS The hypothesis that atmospheric deposition is the most important source for the trace metals As, Cd, and Pb in the Great Lakes is the basis of this research. If this hypothesis is correct, then similar atmospheric depositional loadings of metals can be expected for all regions of the Great Lakes. Studies by Eisenreich et al. (1992, 1990) showed this to be the case for selected trace organic compounds. There are three methods to determine if this hypothesis is correct. First, the atmosphere signal can be determined by backing the atmosphere signal out of the total Great Lakes signal. This is done by examining the trace metal signal in remote lakes and ombrotrophic peat bogs in the Great Lakes region. These locations (by definition) have only atmospheric sources of trace metals. This atmosphere signal can then be subtracted from the total signal in the Great Lakes cores to determine the relative significance of atmospheric deposition. This method is currently being worked on by Bill Sitarz at MSU, and will not be further addressed in this study. A second method to answer this question is to use 21oPb dates of the sediments, and. combined with inventories of the metals, establish a focusing factor. This focusing factor can be used to normalize the data throughout the Great Lakes region to account for things like variable sedimentation rates. The corrected data can be compared using this focusing factor: if all the data are similar, it indicates that atmospheric deposition is the dominant input process to the Great Lakes. if the data are not similar, it means that the study is inconclusive, and that the question 10 must be confirmed in another manner. This method will be used to a limited extent in this study because not all of the 21oPb data is available at this time. The final method used to answer this question is to compare the excess concentrations of the metals in the sediment profile. These excess concentration values can be used in the same way that 21oPb normalized data is compared. This can be used to establish regional patterns of each of the metals, and some information regarding atmospheric depositonal patterns may be derived. This method will be used extensively in this work due to the absence of other data. II. PREVIOUS STUDIES ATMOSPHERIC DEPOSITION The Laurentian Great Lakes are especially sensitive to atmospheric deposition because they have high surface to drainage basin area ratios (figure 2), are near and downwind of urban and industrial centers, and receive a major fraction of their hydrologic input by direct precipitation on the lake surface (Eisenreich et al., 1992). In the past 20 years, numerous studies have shown atmospheric deposition to be a significant source of pollution to the Great Lakes basin. Several studies (Levy and Moxim,1989; Eadie et al., 1984, Spencer and Sachs, 1970) noted that the background atmosphere chemistry varies seasonally: a late summer contamination increase due to US. emissions; and a smaller increase in the spring due to Asian emissions. Remoudake et al., (1991) attributed the variability of atmosphere aerosol concentrations on a daily and seasonal time scale to scavenging by precipitation, and not changes in the source regions. Many studies have documented that the atmosphere is the dominant input for organic contaminants to the Great Lakes (Kelly et al., 1991; Eisenreich et al., 1986; Evans, 1986; Eisenreich et al., 1984; Murphy et al., 1984; Doskey and Andren, 1981; Eisenreich et al., 1979). Urban et al., (1990) found that wet deposition of lead decreased from 1981 to 1983, and attributed this to decreased use of lead in gasoline. The use of lead as a gasoline additive and exhaust from automobiles has been the major source of lead to the atmosphere (Veron et al., 1987). Murphy and Rzeszutko (1977) reported PCB concentrations on Beaver Island in northern Lake Michigan to be the same as those in Chicago. This was interpreted to mean that sources to the atmosphere are diffuse and/or residence times in the atmosphere are long. Residence times 1 1 12 Extent of Great Lakes Basin 1 I l l I I W T I l 0 155.5 313 N Distance in miles Figure 2: Map of Great Lakes Region (after Upchurch, 1972) 13 for Hg in the atmosphere are about 6 to 90 days (Clarkson et al., 1984). This probably holds true for other trace metals besides Hg. Lake Simcoe, Ontario has an atmosphere input of 77% of the total inputs for Cd (Johnson and Nicholls, 1988). According to Nriagu (1986), atmospheric deposition is responsible for 60% of the Cd and 64% of the Pb in Lake Ontario (Coale and Flegal, 1 989). Atmospheric deposition begins with the generation of trace metals in a form which is conducive to transport by the atmosphere. This can occur two ways. The natural component, which is significant for Hg, but less so for other trace metals, is the result of re—emission of mainly metallic Hg vapor from soil, lakes and oceans (Brosset, 1982). The other way trace metals get into the atmosphere is due to combustion of fossil fuels, smelting or waste incineration which volatilizes the metal (Lyons et al., 1983; Andren and Strand, 1981; Annett et al., 1972 ). Globally, these anthropogenic processes are significant, as reported values of the release of Hg is at least 3 times higher than the amount released naturally (Annett and mm, 1973). Once released into the atmosphere, the trace metals are either associated with soot produced in the combustion process, or they exist as a vapor which is often sorbed to other particles in the atmosphere (Arimoto, 1989). Several studies have documented that ~90% of all Hg species emitted to the atmosphere by combustion processes exist as Hg" in the vapor state (Lindberg, 1987; Lindberg, 1980; Johnson and Braman, 1974). Of the metals associated with particulates, the water soluble spades are usually the most common form (Lum et al., 1987). If associated with particulates, there is some dependence on wind direction, which is important to the concentration of the trace metals (Brossett, 1982). Sanderson et al., (1985) showed that primary statistical analysis indicates concentrations of the metals are related to intensity and amount of precipitation, but not to wind direction. It has been shown (Buat-Menard and Duce, 1987), that there is no ”normal" background type of aerosol in the atmosphere. The type and size of these particles have considerable variability. The particles in the atmosphere exist in various sizes. The largest size particles (diameter larger than 22.5 pm) are produced by mechanical means, such as weathering of soils, 14 sea spray, pollen and spores, among others. The smallest particles (diameter smaller than .08 pm) are called aitken nuclei, and are produced by gas-to-particle conversion. These small particles make up most of all particulates in the atmosphere, but have little mass. Mid size particles (diameter .082 um) are produced by either coagulation of aitken nuclei or by gas-to-particle conversion. The mid size particles represent about 50% of the mass, and most of the surface area, of all atmosphere particulates (Bidleman, 1988). These mid size particles are the most significant phase for atmosphere transport. The size of the mid-size and smaller particles may be due to high temperature processes such as combustion (Jeffries and Snyder, 1981). Atmospheric deposition can occur at any time after the formation of particulates of vapor in the atmosphere. Edgington and Robbins (1976) found the residence time for Pb in the atmosphere over Lake Michigan was approximately10 hours. The actual deposition process can be wet or dry. Dry deposition entails the settling of particulates out of the atmosphere. The size of the particulates determines the probable distance of travel before deposition occurs. Vapors containing trace metals (or organic contaminants) may react with other contaminants in the atmosphere and become sorbed to particulates or settle out on their own. Most particulates are of a small size, and, therefore, usually travel a long distance before they settle out of the atmosphere. Vapors are conducive to long range transport of the contaminant (especially Hg) and global dispersion. Migon et al., (1991) found that dry deposition accounts for about one third of contaminants deposited from the atmosphere. Wet deposition is the removal of particulates and vapor by some form of precipitation. The precipitation scavenges, or scrubs, these materials from the atmosphere (Andren and Strand, 1981). This process typically has a greater effect on the larger particulates than on the smaller ones. Wet deposition acts to drastically shorten the residence time of trace metals in the atmosphere, and is the main cause of short range transport of contaminants identified in other studies. Furthermore, wet deposition is responsible for bringing nearly all of the contaminants deposited atmospherically into the Great Lakes. Andren and Strand (1981) reported an input ratio 15 of 40:1 of wet to dry deposition for total organic carbon. Lindberg (1987) reported that precipitation scavenging is the major removal process for Hg vapor. The rate of contaminant removal from the atmosphere is ultimately dependant on the rate of pollutant attachment to the falling precipitation particles and to the precipitation flux at the ground (Scott, 1981). METAL CYCLING in the past 20 years, there have been an ever increasing number of studies focusing on the effects of contamination in the Great Lakes. Most of these studies are concemed with organic contaminants. There are two reasons for this. First, these contaminants are solely the result of man's activities: there are no natural sources for any of these contaminants. Second, once the contaminants are in the natural environment, it is easier to keep track of them: there are no background concentrations that might mask or confuse the concentrations. In addition, most of the physical properties of the organics must be known before they are allowed to be released into the natural environment. Simulating contaminant behavior by using computer modeling programs makes it easier to predict what will happen to the contaminant. Figure 3 shows a summary of the processes affecting the concentration and distribution of trace metals in aquatic environments. Metal studies are much more difficult because the physical properties may not be fully understood with regards to how the metal behaves in the environment. Furthermore, the metals have a natural component in the environment, and the “natural” (or background) levels are seldom known. Because of interactions of natural vs. anthropogenic contamination, the physical processes are more complex for metals than for organics. Because of this complexity, an in-depth discussion of all the processes involved in metal cycling is beyond the scope of this study. Therefore, this study will only briefly discuss the processes involved in metal cycling, and how those processes may impact this study. . '.'\' . , . f...” . l. - v I l . . h‘.‘ ' ~"'.;r'I-I-:i.--, '. ,3‘,f{",:"'j- .' 1' ' .~ . Nep e|0Id Metals " ' Sorbed Organic Deca CO2 Carbon nutrients metals released 'L Benihl ‘fii "S113"? : Reductive Disoluiion Metals Released Figure 3: Diagram showing processes affecting metal cycling (after Matty, 1992) 17 Once the metals enter the Great Lakes (whether by atmospheric deposition or by riverine input) they are subject to many processes before they become a "permanent” part of the sediment column. To remain consistent with the theory that atmospheric deposition is the most significant source of trace metals to the Great Lakes, these processes will be dealt with beginning at the lake surface and working towards the sediment. Upon first entering the lake, the contaminants enter what is known as the surface microlayer, or film. Deposition to this layer is by molecular diffusion and/or by particulate settling (Eisenreich, 1987; Armstrong and Elzennan, 1982; Slinn et al., 1978). Despite the fact that these microlayers are transitory, and even though the residence times of trace metals in them may be on the order of minutes (Eisenreich, 1982), they are important. It has been documented that trace metals accumulate to high concentrations in these microlayers. This is due to complexation with organic matter (Eisenreich, 1991; Ridgeway and Price, 1987; Santschi, 1984). Enrichment of trace metals in this layer indicates a significant input of atmospheric aerosols (Elzennan, 1982). Furthermore, metals entering by particulate settling may be released to the water column by partial dissolution of the particulates (Armstrong and Elzennan, 1982). The movement of trace metals into and out of this layer has been of concern in relation to their upward and downward fluxes and possible toxicity to biota (Elzennan, 1982). Concentration gradients can be affected by several processes operating in the water column. For instance, the magnitude and depth dependence of the diffusion coefficient can have a critical influence on the interpretation of observed profiles with regard to both the rate and location of the chemical reactions supporting concentration gradients (McDuff and Ellis, 1979). When dealing with thin layers (up to a few meters thickness), the timescale for diffusion across the layer is much less than the time for sinking, or advecticn, out of that layer (Denman and Gargett, 1983; Plait et al., 1982). Whereas for very thick layers, the timescale for diffusion across the layer greatly exceeds the time for sinking out of the layer (Small et al., Lande and Wood, 1987; Takahashi and Honjo, 1983). 18 Particles (and trace metals) sinking through the water column pass through the epilimnion. This is the portion of the lake warmed by the sun and above the Therrnocline. According to Coale and Flegal (1989). the residence times for Cd and Pb are 9 days and 4 days respectively. These short residence times of the dissolved phased reflect rapid scavenging of the trace metals by particulates in the water column. Suspended solids in the water column are important in controlling the water column concentrations of trace metals (Dolan and Bierman, 1982). Humic material, which is one of the dominant forms of organic material has a shonger affinity for Pb than for Cd (Campbell and Evans, 1987). This may account for the difference in residence times. Other factors affecting trace metal removal in the primary sedimentation process include the metal's solubility and the settleability of the insoluble forms (Kempton et al., 1987). Any free metal ions in the water column can become hydrated, which plays a major role in initiating adsorption to particulates (Jean and Bancroft, 1986). Of the particulates that the metals can adsorb onto or complex with, metallic oxides, organic matter, carbonates and clays are the most important components (Rapin et al., 1983; Forstner, 1982; Jenne, 1973). This sorption and complexation occurs in the water column, and acts to keep the concentration of metals (especially mercury) near background levels, except for near local points of discharge dimctly into the lake (Bubb et al., 1991a: 80, 1980). Adsorption is recognized as the main control on trace metal behavior (Long, 1991; Honeyman et al., 1988: Santschi, 1984; Eadie et al., 1984; Stumm and Morgan, 1981; and Balistreri et al., 1981). Due to the large number of studies which involve adsorption, it is known what kind of an effect most of the adsorption controls have, but not how they all interact. The lower the pH, the higher the concentration of trace metals in the water column (Masscheleyn et al., 1991 ; Johnson, 1991; Lodenius and Autio, 1989; Stephenson and Mackie, 1988: Di Toro et al., 1985). More basic conditions tend to decrease the solubility of trace metals, and force them to become sorbed onto particles in the water column or sediment-water interface. The pH of the water column has been shown to have a diurnal cycle of variability due to effects produced by photosynthesis. Trace metal concentrations, have a similar cycle which lags a few 19 hours behind the pH cycle; the highest concentrations of As in the water column occurs just after the pH is the lowest (Fuller and Davis, 1989). The redox conditions (which can be controlled biologically as well as physically) are also important. Under oxidizing conditions, trace metals are not very soluble, and tend to be partitioned with particulates. Under reducing conditions, trace metals are more soluble, and concentrations in the water column increase significantly (Masscheleyn et al., 1991). The ionic strength of the water also plays a part in trace metal chemistry. Metal-clay sorption is known to decrease significantly as ionic strength increases (Di Toro et al., 1985). This may also be true for other types of substrate materials. Particle size and composition have an effect on trace metal concentrations. Brook and Moore (1988) reported that concentrations of As, Cd, and Pb in a Montana stream generally increase with decreasing particle size. Metals are usually enriched in the smaller silt/clay fractions of the sediment (Bubb et al., 1991b,1991c) Complexing of trace metals is a process that is similar to adsorption, and is also known as scavenging. The difference between adsorption and complexing is that adsorption is a single bonding of the trace metal to another substance, and complexing is a multiple bonding, usually to an organic substance. Most commonly, these organics are humic or fulvic materials (Eisenreich, 1991; Long, 1991; Himer et al., 1990: Davis, 1984; Frimmel et al., 1984; Elderfield, 1981: Reuter and Perdue, 1977; Andren and Harriss, 1975). This multiple bonding creates a stronger cohesive force than adsorption alone, and, as a result, trace metals are usually partitioned with organic materials and share their fate. Some materials, such as organic matter or iron and manganese oxides, have a scavenging ability far out of proportion to their abundance in the environment (T essier and Campbell, 1987). Phases most important in scavenging dissolved metals are fine grained organic matter and Fean oxides. The fine grained material is selectively removed to the deepest regions of depositional basins by sediment focusing, and the oxides continue to scavenge dissolved metals as they move through the water column (Long, 1989; Santschi, 1984). In one study, iron and 20 manganese oxides were found to scavenge metals to 50% of the total heavy metal content (Feijtel at al., 1988). It should be noted, however, that not all metals tend to complex with the same ligands (organics), nor is the degree (strength) of bonding the same. Cd adsorption is not significamly affected by the presence of organic material, due to the weak complex formation with organic ligands (Davis, 1984). Furthermore, the ligand concentration and ligand type will influence the degree and type of complexing taking place (Santschi, 1984). Complexing can also occur with water itself. Due to the polar nature of the water molecule, bonds can form with either positively or negatively charged species (Long, 1991). Bonding usually occurs with six water molecules. The water molecule can be broken down, and the trace metal may be complexed with five water molecules and one OH'. Breaking more water molecules will increase the number of OH', and, therefore, increase the pH of the system. The water molecule may first be bound to another inorganic substance; so, when trace metals complex with the water molecules, they are bound to an inorganic substance as well. This creates what is called the double layer complex, where H20, H30"', and OH' can be bound to a solid substance (figure 4). Consequently, both positive and negative species can be bound to the substrate. The rate of adsorption/complexing is usually rapid, but depends on several factors. Concentrations of the trace metals and ligands, type and size of ligands, and residence times of the trace metals and ligands in each reservoir, temperature, pH, and ionic strength are all important in determining the rate of adsorption (Frimmel et al., 1984; Balistrieri et al., 1980). However, Nyffeler et al. (1986) indicates that uptake of (radioactive) trace metals by suspended particles in natural aquatic systems is often slow, and the time constraints for scavenging are of the same order of magnitude as the residence times of particles in the water column. While there is some variability in the complexation process between the different trace metals and their prefered ligands, the complexation reactions with ligands in solutions and on solid surfaces are essential features of the biogeochemical cycling of trace metals (Comans and Van Dijk, 1988). 21 Charged particulate ”Double Layer" Figure 4: Diagrammatic representation of Double Layer complex (after Long, 1990) 22 Particulates in the water column settle to the sediment. carrying trace metals with them. The nepheloid layer, however, has a tendency to have a much higher density of particulates in it. The nepheloid layer is the region of the water body in which there is a marked increase in the suspended particulate matter. These particulates have been shown to have median diameters 20% smaller than the particulates found in surface waters (Baker et al., 1985), and are composed primarily of silica. calcite and organic matter (Mudroch and Mudroch, 1992). As a result of all of these particulates, there are highly elevated concentrations of trace metals to be found in the nepheloid layer. Sandilands and Mudroch (1983) found that the nepheloid layer in Lake Ontario had Pb concentrations similar to those found in the 0-1 cm layer of the sediments. The nepheloid later is important for accumulating, recycling, and transporting contaminants in Lake Ontario (Mudroch and Mudroch, 1992). There is a second zone in the water column which has an abnormally high particulate concentration. Located just above the sediment, it is called the benthic nepheloid layer. This layer is also higher in trace metals than the surficial sediments (Cahill and Shimp, 1984). The action of biota in the lake or sediment is an important but often neglected factor in trace metal behavior (Long, 1991). The fact that there are many ways that biota can effect the environment may be part of the reason for this omission. Microorganisms such as bacteria have evolved enzymes capable of changing the oxidation state of elements (Jackson et al., 1982; Wood, 1973; Jemelov and Assell, 1973; lverson et al., 1973). Wilhelmy and Flegal (1991) found that high concentrations of trace metals are often associated with high nutrient levels in the water column. Bioturbation of the sediments can exchange or mix particles in the water (Santschi, 1984). Mixing caused by benthic biota can drastically change the input recordiof contaminants in sediments. Christensen and Klein (1991) proposed a method for the "unmixing' of the sediment input records for areas where bioturbation is a significant problem. Eadie et al. (1984), Luoma and Davis (1983), and Shafer and Armstrong (1990) recognized the effects of grazing, filter feeding, and fecal pelletization on water chemistry. Biological materials are an important carrier phase for 23 trace metals because they are rich in organic matter, and allow easy sorption of the metals. The settling of material has been shown to be an important carrier phase for trace metals (Sigg et al., 1987; Landing and Feely, 1982). The effects reported from biological Uptake by organisms such as fish may not represent true concentrations of metals in the organism's habitat, but may be due to interactions with the food chain and the environment (Glass, 1973). The Uptake of trace metals by biota in the aquatic environment is known to be related to the activity of the free aquo metal ion and may be treated as a series of complexation reactions (Comans and Van Dijk, 1988). Bioturbation is not the only means by which the lake bottom may be disturbed or reworked. Bottom currents, which can be due to density, poor lake stratification, or temperature differences, can be especially strong during storm events, and have a significant effect on sediment profiles (Bennett, 1987; Flood and Johnson, 1984; Johnson et al., 1984). Furthermore, in places where the lakes are shallow, especially near shipping channels, freighters and other boats may chum up sediments, bringing buried metals back into the system. The process of resuspension plays an important role in the transfer of particles from the sediment to the water column. Several studies have clearly documented the importance of this process. Bennett (1987) has shown that if there were no resuspension and no source of sediment to Lake Michigan, the lake would be clear of suspended matter in about two months. In reality, there is about two years worth of sediment suspended in the water column. Walsh et al., (1988) found that the total particulate fluxes at the bottom of a lake were greater than those found in the mid-water column. Other studies Show that the exponential increase of suspended matter in the water column as the bottom is reached indicates resuspension of bottom sediments (Walsh et al., 1988; Aggett and O'Brien, 1985; Eadie et al., 1984; Chambers and Eadie, 1981; Spencer and Sachs, 1970). Resuspension is affected by stratification of the water column. When there is no stratification, resuspension of particles is most extreme (Aggett and O'Brien, 1985; Eadie et al., 1984). Because of the processes reworking the sediment, the sediment itself may become a source of contamination to the overlying water (Officer and Lynch, 1989; Salomans et al., 1987). 24 According to the Great Lakes Water Quality Board, this is occurring in 38 of 42 Great Lakes areas of concern (T heis et al., 1988). Organic material often tends to coat the surfaces of other inorganic materials such as Fean oxides and oxyhydroxides (Chen and Deng, 1989: Davis, 1984; Lion of al., 1982). These oxides and oxyhydroxides are thought to play an important role in trace metal cycling in sediments. Burial of sediments containing such oxides and oxyhydroxides lead to reductive dissolution below the redox interface (Belzile et al., 1989; McKee et al., 1989). Little is known about the extent to which organic material covers the surfaces of inorganic material in the natural environment. Davis (1982) illustrated that the amount of organic material adsorbed was influenced by the chemical nature of the surface; so hydrous oxides, which are basic, tend to adsorb greater amounts of natural organic matter than those with acidic surfaces such as silica (Davis, 1984). This means that Fe and Mn oxides are important to the processes controlling adsorption of trace metals (Jackson et al., 1982; Lion of al., 1982). As these oxides usually tend to accumulate at the redox boundary in the sediment, there is often a high concentration of trace metals there as well. It is also believed that the fate of some trace metals may be determined by processes related to organic matter diagenesis occurring at the sediment-water interface (McKee, 1987; Pedersen et al., 1986; Klinkhammer, 1980). Trace metals are released from particles in the sediment during burial and early diagenesis. During burial, organic matter decays and the redox state and pH of the wdimentary environment changes. There are several processes that affect particle-bound trace metals in various diagenetic environments in aquatic systems: the formation and reduction of Fe and Mn oxides (Rezabek, 1988: Comwell, 1986; Laxen and Chandler, 1983: Cerling and Turner, 1982; 1981; Balzer, 1982; Davison et al., 1982; Chapnick et al., 1982; Tipping et al., 1981; Davison, 1979; Davison and Heany, 1978; Anthony, 1977; Robbins and Callender, 1975); the decay of organic matter (So, 1980; Lerman, 1979; Bemer, 1972); the reduction of sulfate and the formation of sulfides such as pyrite, galena, and sphalerite (Jean and Bancroft,1986; Pyzik, 1981; 25 Jenne, 1973; Bemer, 1972, 1967); and the formation and dissolution of carbonates and Fe concretions (Effler, 1984; Treese et al., 1981; Dean and Gorham, 1976). In large lakes, the above processes have a signigicant effect on the remobllization of particle-bound trace metals (Long, 1989; Allan, 1986; Kosov, 1986; Rossmann, 1986; Salomons and Forstner, 1984; Rea et al., 1981; Johnson and Eisenreich, 1979; Sly and Thomas, 1974). The flux of trace metals (especially Hg) from the sediments will be higher if the bottom waters are anoxic (Bothner et al., 1980). The effect of early diagenesis on most heavy metals is unclear, and, in most cases is obscured by the strong anthropogenic signature of the total hydromorphic profile (McKee et al., 1 989) It should be further noted that trace metals are not irreversibly fixed on particles, but can be released in response to changes in the aquatic environment they are part of (Comans and Van Dijk, 1988). Therefore the term ”permanent sink” for trace metals is something of a misnomer. The metals will reenter the environment at an accelerated rate if changes in conditions in the environment occur, but, regardless, will eventually reenter the environment. The residence time for mercury in sediments is on the order of millions of years (Clarkson et al., 1984), and the other trace metals may have similar residence times. Ill. METHODS SAMPLING Collection of the sediment cores used in this study was done during September of 1991, and in August of 1992. Both trips involved the use of the USEPA Research Vessel Lake Guardian. Sediment cores were collected in Lakes Michigan and Ontario in several different locations (figures 5 and 6). A total of 15 cores were collected for this study; 8 from Lake Michigan, and 7 from Lake Ontario. Most of these cores are located within or on the edges of depositional basins in these lakes. The specific site locations and descriptions of the cores are given in tables A1-A15 of Appendix A. A stainless steel box coring device (30cm x 30cm x 70cm) was lowered from the Research Vessel Lake Guardian to the sediment and retrieved. This device was lowered by cable to the lake bottom, slowly enough so that the sediment was not disturbed by this action: evidenced by the lack of suspended matter in the water above the sediment, and the presence of ”fluff" found on the sediment surface of two cores. Fluff is the material often found at the sediment-water interface. It is typically very difficult to collect and easily disturbed, and appears as a nebulous, fluffy material. Subcores of this sediment block were then taken. Five 3" PVC core tubes were inserted into the sediment under vacuum to avoid compaction of the sediment. The cores were inserted about 5 cm away from the sides of the box core to avoid the disruption of the sediment caused by the box coring device. Once all core tubes were inserted into the box core, the bottom was opened, and rubber stoppers were inserted into the bottom of the core tubes. The top of the core tubes were sealed with polyethylene caps to prevent contamination of the sediment core. 26 27 Green Bay EPA I Ludington EPA 27 I EPA 23 I Milwaukee _ N I EPA 18 EPA 19 I I J l I j EPA 11 ' ' T T P o 25 so 75 100 Distance in miles Chicago Figure 5: Lake Michigan sampling locations (after Eisenreich et al., 1990) 28 83. .._w 5 fiecomm bag 2268. @5353 25:0 9.3 no 2:9“. mo__E E 8:920 09 me -I h n n n . J q 4 . Binocu— , \ , mw