'7 J“. ’3... . ..7 7. ;"o.:7..1 7 7~ "Y. 3W7 >~ ‘ ‘55- s; ‘ ‘ 4L 2“.) ~35t‘jw' s . .7“ 7' f7“ 7 _ ‘4 ;1‘L;v:‘.i7(.‘g 7 in“: Ynmv ». ! KKti—v" . “ a Jflgj‘r; u. .. . n. urn-.7117. ‘7 "" ‘ I"' ‘..‘:"‘:b. -'~ 2 .71, Tum, ”1 ) n 79-1 537:, . :13"! H . s .577“ . . Viv} ‘ v: 7" 7 M 3..“ .7 __ .... -- “.uL . m 3‘) , ' .::Q.,..g.s.am ’3 ': r r44- . 4- =3? r1.“ ur‘. ’¢r¢fl_ 7 : ;,. 7'»; 4:4“ug-JM 717;?" I", Luau-t :7 ., 7 fit? a}. 7.7 -z._ ‘ x 7 17?. ‘p 7:577h...u‘m .. .3743" " up». 7 7",! \l hm ' .M r: :15 7 7 7- .11 w U| [- r" Maw“ ”.47 3 7-r7-‘4, 'l 7. -‘«l :7“! \1. 7‘ 77.5%"- M» 7. 7 ‘3. w) -7’. I .. 7' _ 7 7 .... . . "7| , _ 7. . 7 ,. I 77‘. . . ' . . v 7 7-. v ,1 . 7. ~ .4; .‘I’. . .v 7 \ 7. 1- .7 7. , ',.’r‘,,,,‘ 17;,” . . . . '1713 w .71... _ ,-,. ,,7 7 . u: n 5‘, ~ . I. a.- , > .1 7 7.70;" ,- ".,,,.. . VJ - ‘ "~ "I!!! l ‘7": ‘1. - ,‘.;.,,..".-.N,.‘, a . . 7 ~ I7 . "I ‘ _v -; —, Q7. «7-7. n .-.,- .,7 , ‘ . , _ . v quill- ,u._., 7 .' - V'r I, . n .. 1 ' - . :v:”»o1:m«, . .7 7-.. .,J, .— .v:‘;~:\; 7.. . 7'7 ,,. . 7 {7.7' '1';.'.. 7. . ‘l‘j’;'>ll|‘l-€‘ ’ I}! ..,'.' 7 ”7.77:3: ’m 1 .s‘ “3“" f - #73:..4' -.,. ' 7' '7. in“ 53’" :<,',$7"."v“~-7 ‘ r I 1' 1,-7.1- 'f' 1': ‘7‘... ::' ’3 -r:l;§""I.'~‘ 2-77: .. 7 . 715‘. 7177-9777,, .7734. ,7 7:. 3.32:..~...:.1.:,.I, ”2.3....“ in a. I. a». —r 3’ 2-1 7 ‘0’ Q: 21:" ,,§ . 77 :,--7:,7.‘ 77 774,11” ”U? 7 m . f..m'r! n 771,-{7121‘ -. , . agrzt'y 711:5... , v‘ '4 m7? ‘7. /. ' jig/73“ “"n’u‘p ' W‘fi‘i.: ‘0‘.“ 1:27;: - rtv.~.._ . . llllllllllll This is to certify that the dissertation entitled Influence of Early Diagenesis on the Geochemical Cycling of Arsenic and Mercury Investigations in the Great Lakes and the Gulf of Maine presented by Jane M. Matty has been accepted towards fulfillment of the requirements for Ph.D. Geology degree in Mil/M Major profe@ Date 27 January 1992 MSUis an Affirmative Action/Equal Opportunity Institution 0- 12771 LlBRARY 1 {Michigan Stats 1 lhflwmsuy PLACE IN RETURN BOX to romovo this checkout from your record. TO AVOID FINES return on or baton data duo. DATE DUE DATE DUE DATE DUE ! 1 1 3., Lifljf 0 4 2094 l‘ | ETC—ll; J [:14 _f MSU Is An Affirmative ActionlEqual Opportunity Institution chS-nt w-~-—_ _ INFLUENCE OF EARLY DIAGENESIS ON THE GEOCHEMICAL CYCLING 0F ARSENIC AND MERCURY Investigations in the Great Lakes and the Gulf of Maine by Jane M. Matty A DISSERTATION Submitted to Michigan State University in pattial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Depattment of Geological Sciences 1992 ABSTRACT INFLUENCE OF EARLY DIAGENESIS ON THE GEOCHEMICAL CYCLING OF ARSENIC AND MERCURY Investigations in the Great Lakes and the Gulf of Maine by Jane M. Matty The geochemical cycles of arsenic and mercury in aquatic systems are strongly influenced by the association of these elements with particulate matter. In aquatic basins, arsenic and mercury are scavenged by particulate matter, which settles to the bottom, where it is subjected to the physical, chemical and biological processes of early diagenesis. The effects of these processes on arsenic and mercury were investigated in selected depositional basins of Lake Michigan, Lake Superior, and the Gulf of Maine. Sediment cores were collected and sectioned at 1 cm intervals, porewaters were separated by centrifuging, and sediments subjected to sequential chemical extractions. Porewaters and sediment leachates were analyzed for arsenic and mercury. Alkalinity, pH, and ferrous iron of porewaters, and the organic carbon content of sediments was also determined As sediment is buried, changes in the partitioning of mercury and arsenic among different phases of sediment occur, indicating that both elements are mobilized and repartitioned during early diagenesis in all of the sites examined. Concentration gradients of arsenic in porewaters indicate that there is a flux of arsenic from the sediments to the sediment-water interface via porewater at most sites. Concentration gradients of mercury in porewaters are more complicated than those for arsenic, but there are gradients suggesting some flux of mercury to the sediment-water interface at all of the Great Lakes sites, although not in the Gulf of Maine. The upward diffusive fluxes of mercury and arsenic released during early diagenesis are responsible for the observed repartitioning of these elements in buried sediments, and for the enrichment of surface sediments in these metals. Diagenetic enrichment of surface sediments is more efficient in freshwater than in the marine setting, and more effective for mercury than for arsenic. This enhances the potential bioavailability of these metals. Permanent burial of arsenic and mercury in sediments is governed by the formation of authigenic minerals, particularly sulfides, in the reduced zone of sediments. ACKNOWLEDGEMENTS I thank my advisor, Dr. David T. Long, for making this project possible, and for sharing his expertise and enthusiasm for geochemistry. I also appreciate the assistance of the other members of my thesis committee, Dr. Duncan Sibley, Dr. Michael Velbel, and Dr. Frank D'Itri. I would also like to thank: Tim Wilson and Joe McKee for teaching me the ropes and for help with sampling and lab work; the Captain and crew of the R/V Seward Johnson and DSRV Johnson-Sea-Link II for their assistance and cooperation; Steve Eisenreich, Deb Swackhamer, and Joel Baker, for making the cruises more interesting (both intellectually and otherwise); and especially Dave Matty, for his help and his patience. This project was funded by grants (to DTL) from the National Oceanic and Atmospheric Administration National Underseas Research Program, Michigan Sea Grant, and The Max and Victoria Dreyfus Foundation, Inc . TABLE OF CONTENTS List of Tables List of Figures I. INTRODUCTION MERCURY AND ARSENIC IN AQUATIC ENVIRONMENTS The scavenging process Role of particle cycling Microbial processes Effects of early diagenesis--- OBJECTIVES II. METHOD OF STUDY SAMPLING ............... Sample sites Sample collection -- Clean procedures Shipboard sample processing SEQUENTIAL EXTRACTIONS ANALYTICAL PROCEDURES III. RESULTS SITE CHARACTERIZATION Ferrous iron PARTITIONING OF ARSENIC Sediments - Porewater - - PARTITIONING 0F MERCURY Sediments --_--- Porewater IV. DISCUSSION DIAGENETIC PROCESSES: EVIDENCE AND EFFECTS Early diagenesis of arsenic Early diagenesis of mercury Fluxes to the sediment-water interface Implications for bioavailability ANALYSIS OF DIAGENETIC VARIABILITY Great lakes vs. Gulf of Maine Variability among Great Lakes sites. Potential for seasonal variations iv V. SUMMARY AND CONCLUSIONS- - ..... _ -92 ROLE OF DIAGENESIS 1N GEOCI-IEMICAL CYCLING 92 CONCLUSIONS 93 APPENDICES 1. METHODS ' -- - ......... 95 2. SAMPLE DATA ...... - 108 3. ANALYTICAL DATA 129 REFERENCES - ........................ - - 146 LIST OF TABLES Table 1. Summary of methods used for sequential chemical extractions ............................... 19 Table 2. Solubility product constants of some sulfide minerals at 25 'C (data from Faure, 1991) ______ - ________ - - ................... 67 Table 3. Parameters used for calculation of diffusive fluxes of arsenic ................................. 71 Table 4. Parameters used for calculation of sedimentation fluxes of arsenic ......................... 73 Table 5. Summary of site characteristics that influence early diagenesis ............................... 78 Table Al-l. Graphite furnace conditions for arsenic analyses. ............................................... 106 Table Al-2. HAS-200 conditions for mercury analyses ........................................................ 107 Table A2-l. Shipboard core descriptions ................................................................................ 109 Table A2-2. Sediment sample data ......................................................................................... 120 Table A3-l. Extractable arsenic data ...................................................................................... 130 Table A3-2. Extractable mercury data .................................................................................... 134 Table A3-3. Porewater data ..................................................................................................... 139 Table A3-4. Organic carbon data ............................................................................................ 144 Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure ‘7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figurezs. Figure26. Figurc27. Figure 28. LIST OF FIGURES Conceptual model of particle cycling and metal interactions in aquatic systems Locations of sampling sites in (a) the Great Lakes and (b) the Gulf of Maine ........ Ferrous iron in porewater pH of porewater Alkalinity of porewater Organic carbon content of sediments - Arsenic extracted from the EX fraction of sediment ...... - -- - - - Arsenic extracted from the WAS fraction of sediment ........................................... Arsenic extracted from the ER fraction of sediment ........... - Arsenic extracted from the MR fraction of sediment - - Arsenic extracted from the OX fraction of sediment ....... - -- Total extractable arsenic Arsenic in porewater ............ Mercury extracted from the BS fraction of sediment ........................................... Mercury extracted from the AS fraction of sediment ....... Mercury extracted from the OX fraction of sediment __ -- Total extractable mercury Mercury in porewater Summary of chemical data. Lake Michigan North Basin- 1 Summary of chemical data: Lake Michigan North Basin - 2 Summary of chemical data: Lake Superior Ile Parisienne Summary of chemical data: Lake Superior Caribou Basin Summary of chemical data: Gulf of Maine Murray Basin Relationship between sediment organic carbon content and (a) total extractable arsenic, and (b) total extractable mercury in sediment Cartoon of iodine recycling at the sediment-water interface (from Kennedy and Elderfield, 1987) Relationship between ferrous iron and dissolved mercury in porewater of reduced sediments, Great Lakes samples - Dissolved Pb, Fe, Mn, and Cd in porewater from the Laurentian Trough (from Gobeil and Silverberg, 1989) Enrichment factors for arsenic and mercury (from Table 5) vii 5 14 22 24 26 28 29 30 31 32 33 36 37 39 4o 41 43 44 - 46 47 43 49 50 54 63 66 68 8O Figure 29. Cores collected by submersible from Lake Superior Caribou Basin .................... 89 Figure Al-l. Alkalinity measurement apparatus for small-volume samples 98 Figure Al-2. Results of alkalinity measurement in air x5. nitrogen atmosphere test. ............ 99 Figure A1-3. Results of steady-state analysis for arsenic extractions - 101 I. INTRODUCTION The geochemical cycling of elements is a fundamental theme of geochemistry. Understanding the controls on geochemical cycling of elements allows geochemists to interpret elemental distribution pattems in rocks and other geologic materials, and thereby gain insight into geologic processes. An mtderstanding of geochemical cycles is also of substantial practical value. The geochemical cycles of elements have influenced the earth‘s surface environment throughout geologic time, and have themselves been altered by changes in that environment (Fame, 1991). Knowledge of the cycles of toxic elements, what controls them, and how they respond to perturbations is essential for environmental scientists researching such elements. One environmental issue which generates a good deal of public concern is the pollution of aquatic environments. In order to comprehend the problem fully and develop practical guidelines for the control and cleanup of toxic substances in aquatic systems, the behavior and fate of contaruinants must be understood This requires adequate knowbdge of the geochemical cycles of the contaminating elements. Mercury and arsenic are toxic; they are also ubiquitous in aquatic environments, with both natural and anthropogenic sources. Fish consumption in some areas is proscribed or limited because concentrations of mercury in fish are elevated, even when concentrations in water bodies are low. The concentrations of many contaminants (including mercury and arsenic) in water bodies are generally low due to efficient scavenging by particulate matter in the water column. Particles adsorb dissolved contaminants from the water cohrmn, eventually settle, and are incorporated into the bottom sediments. Their associated contaminants are thus removed from the water column. This process has been regarded as a "self-cleansing" mechanism for polluted aquatic systems (e.g. Fbrstner and Wittmann, 1983; Hart, 1982); however, there is substantial evidence indicating that this is not the complete cycle. The occurrence of elevated cmcentrations l of mercury in fish, relative to concentrations in water bodies, indicates that there is some process (or processes) at work facilitating the transfer of mercury from particulate matter to the biota. Scavenging and burial provide an adequate description of contaminant behavior only on long (i.e. geologic) time scales. On shorter time scales, the behavior of contaminants such as hydrophobic organic compounds and heavy metals has been linked to the dynamic behavior and short-term cycling of particulate matter in lakes and oceans (e.g. Baker and Eisenreich, 1989; Honeyman et al., 1988). It is the short term behavior that governs the bioavailability of contaminants, while the long term behavior controls the permanent removal of contaminants from aquatic ecosystems. Processes occurring at the sediment-water interface are of particular importance, as this geochemical bormdary has been found to exert the greatest control on the cycling of many elements in shallow aquatic systems such as lakes and coastal marine embayments (Santschi, 1988). This project is an investigation into the geochemical cycles of arsenic and mercury in aquatic environments (the Great Lakes and the Gulf of Maine). Mercury and Arsenic in Aquatic Environments Mercury and arsenic are introduced to aquatic environments from both natural and anthropogenic sources. The principal pathways are via the atmosphere (particularly for mercury) and weathering processes. The major natural sources of mercury and arsenic are sulfide ores and minerals. Anthropogenic sources of mercury and arsenic inchrde a variety of industrial and manufacturing processes, the burning of fossil fuels, and municipal sewage effluent. Because mercury andsome ofits compounds arehighly volatile,thereisaconstantfluxtothe atmosphere from ores, soils, and volcanic emissions. Mercury in the atmosphere is adsorbed by particulate matter and removed by rainfall or dry deposition, thus providing a flux to aquatic systems. The major pathway for the transport of arsenic to aquatic environments is by weathering processes, ratesofwhicheanbesubstantially acceleratedby antluopogenicactivities. Itisestimatedthatup t050%ofthe mercurycurrentlycycledthroughtheatmosphereis ofanthropogenic origin; the anthropogenic sources of arsenic are approximately 2.5 times the natural contribution from weathering (Faust and Aly, 1981; Moore and Ramamoorthy, 1984). THE SCAVENGING PROCESS Like other trace metals, mercury and arsenic are readily scavenged from the water column by particles (FOrstner, 1982; Fbrstner and Wittmann, 1983). Metals can enter an aquatic basin already adsorbed to particles, or they can be scavenged from the water column by suspenwd and settling particles within the basin. This represents a major pathway in the biogeochernical cycling oftracecontaminants (Hart, 1982). Theparticulatematterin aquaticsystemscanbethought ofas comprising two distinct fractions, an inert portion (consisting primarily of detrital silicate minerals) and a reactive-or hydromorphic—portion (including clay minerals, carbonates, sulfide minerals, hydrous iron and manganese oxides, and organic matter). The phases most important in scavenging dissolved metals from solution are fine-grained organic matter and iron-manganese oxides (Fbrstner and Witnnann, 1983). Dissolved contaminants scavenged from the water column are incorporated into the hydromorphic fraction, which is capable of taking up or releasing metals (Gibbs, 1977). Contaminants associated with the hydromorphic fraction are likely to be bioavailable and chemically reactive in aquatic and sedimentary environments (Ftlrstner and Wittmann, 1983; Allan, 1986). The removal of contaminants from the water column by scavenging is, however, neither complete nor permanent. The efficiency of scavenging by particles has been found to be related to the concentration of particles in water, the concentration of the element, the nature of the surfaces available, and the affinity of the element for the available surfaces (Salomons and Fbrstner, 1984; Honeyman et a1., 1988). The capacity for sorption of trace elements by particulate matter has been found to be limited by competition for sorption sites by major elements (Frenet, 1981; Rae and Aston, 1982; Fbrstner and Wittmann, 1983). It is also uncertain whether increases in the anthropogenic input of pollutants are being balanced by increased removal by the scavenging process (Sigg et aL, 1987). The presence of complexing agents in solution and the alteration of solid phase surfaces due to changes in redox conditions (dissolution of iron and manganese hydroxides) or pH (dissolution of carbonates and hydroxides; desorption of metals) have also been found to alter sorption capacities (Fbrstner and Wittmann, 1983). ROLE OF PARTICLE CYCLING Since mercury and arsenic are associated with particulate matter, their cycling is linked to the cycling of particles in aquatic systems. A conceptual model for the cycling of particles and their associatedmetalsinlakesandoceansisshowninFigure l. Particlefluxhasbeenfoundto control concentrations and residence times of particle-reactive elements in lakes (Santschi, 1984). Processes modifying particles in water bodies can influence the residence times of elements, (Santschi, 1984; Whitfield and Turner, 1987; Bacon and Rutgers van der Loeff, 1989), the bioavailability of contaminants (Elder, 1988), and the proportion of deposited metal that is retained in the sedimentary record (Shaw et al, 1990). Particulate matter in aquatic systems is derived from atmospheric deposition, river inputs, resuspension of bottom sediments, and biological production. Although distributed throughout the water column, particles are often concentrated in several distinct layers. The first is an upper nepheloid layer which develops at the thermocline where higher density water below slows the settling of particles (Rea et al., 1981). Below the nepheloid layer, a high concentration of particles can occur in the benthic nepheloid layer (BNL) which extends several meters upward from the bottom (e.g. Biscaye and Eittreim, 1974; Feely et al., 1974; Eadie et al., 1983). Below the BNL particles are present in the sediment column. In depositional basins and other areas where currents are minimal, particles also occur in a layer between the BNL and the sediment column: the "fluff“, or sediment boundary layer (SBL), where particles are in physical contact with one another, yet remain sufficiently diffuse to be resuspended very easily (Wilson et al., 1986; Sweerts et al., 1986). Significant compositional differences have been observed between layers, in both lacustrine and marine environments (Meade et al., 1975; Eadie, 1984; Eadie and Robbins, 1987). The composition of particle layers within the water column has been observed to vary spatially and change seasonally (Sandilands and Mudroch, 1983; Eadie and Robbins, 1987; Tsunogai and Uematsu, 1978). This can lead to seasonal and spatial variations in the distribution of particle- - associated metals. In addition, the chemical composition of suspended particles comprising these layers maydifferfrornthat ofactively settlingparticles inthewatercolumn (Eadieand Robbins, Atmospheric Deposition Biological-Production ;. 5mg ‘3' Organic tree. coz Figure 1. Conceptual model of particle cycling and metal interactions in aquatic systems. 1987; Honeyman et al., 1988). Masuzawa et al. (1989) have found settling particles to change composition as they settle into deeper waters. There is evidence that settling particles develop from populations of suspended particulate matter; this process has been linked to the transfer of particle-reactive pollutants and nutrients to bottom sediments in lakes (O'Melia, 1985; 1987). Significant processes believed to occur within the nepheloid and BNL include photosynthesis and respiration, precipitation and dissohrtion, adsorption and desorption, aggregation and disaggregation, and biological uptake and decomposition (Honeyman et al., 1988). These processes can result in the uptake or release ofmetals within the various particle layers, and therefore influence residence times of metals in aquatic systems. Such processes can significantly retard the permanent burial of scavenged chemical species (Csanady, 1986). For example, Peterson and Carpenter (1983) attribute arsenic enrichment in deep waters of an anoxic fjord to the release of arsenic from decomposing organic matter in deep waters. Many of these processes probably occur in the SBL as well, but because of sampling difficulties, the natrue of particles and processes within the SBL are largely unknown (Pedersen et al., 1986). McKee et al. (1989a) have demonstrated that the SBL is important in the cycling of trace elements in Lake Superior. Depositional basins, where fine-grained sediments are actively accumulating, are significant sites for the cycling of contaminants in lakes and oceans. Hydrodynamic processes result in the selective transport of fine-grained sediments to deep areas where currents are minimal, this process in known as sediment focusing (Hilton et al., 1986). Since contaminants are primarily associated with fine-grained particles (FOrstner and Wittmann, 1983), focusing results in the accurmrlation ofcontarninants primarilyinareaswherefine—grained wdirnents accumulate(Eadie and Robbins, 1987; Luring, 1975). Physical processes occurring near the sediment surface (within the SBL and upper sediment column) include bioturbation and resuspension; these processes can affect element cycling. Wave-induced resuspension has been documented in both lacustrine (Hikanson, 1982; Matty et al., 1987) and marine environments (Baker and Feely, 1978; Lampitt, 1985). Resuspension can also result from current activity (e.g. Johnson et al., 1984; Lampitt, 1985). Bioturbation results in the resuspension of bottom sediments (Nowell et al., 1981) and in the mixing of the upper layers of the sediment column (Aller, 1978; Forsmer and Wittrnann, 1983). These processes can result in the transfer of elements from the sediment to the water column in two ways. Resuspension (wave- or current-induwd, or via bioturbation) increases the residence time of particles in the water column. This increases the extent of alteration of and potential release of contaminants from particulate matter. This process has been shown to be responsible for the recycling of mercury-polluted sediments in the Wabigoon River system of northwestern Ontario (Allan, 1986). Resuspension and bioturbation can also release porewaters from the sediment column. Since porewaters are typically enriched in metal contaminants due to diagenetic reactions, this can result in a flux of dissolved metals to the overlying water (Fdrstner and Wittmann, 1983). Although the mobilization of contaminants from bottom sediments may be only afraction ofthe total amormtaccumulated, this may represent asubstantial environmental impact (Jennett et al., 1980). The accumulation of mercury by fish exposed to resuspended sediments (under simulated dredging conditions) has been documented (Seelye et al., 1982). MICROBIAL PROCESSES Microbially-mediated processes (in addition to those which drive early diagenesis) affect both mercury and arsenic in aquatic environments. These processes can alter residence times of mercury and arsenic in the water column, and increase bioavailability. The release of mercury from sediments has been linked to processes which generate volatile forms of mercury, most of which are microbially-mediated. Aerobic bacteria can oxidize HgS, producing soluble Hg“. This can then be converted to elemental mercury, methyl mercury, or dimethyl mercury via the detoxification mechanisms of other bacteria (Wood, 1974). Microorganisms can also degrade methyl mercury by reduction to elemental mercury (Spangler et al., 1973; Wood, 1974), and mmficacidshavealsobeenfmmdwproduceekmenmlmwryfiommeremicims (Albertset al., 1974). Elemental many and dimethyl mercury are volatile, and may be lost from the sediments; methyl mercury is readily taken up by organisms (Wood, 1974). Micmbialprowssescanalsoefi‘ectthereleaseofarserficfiomsediments bytheproduction of volatile methylated compounds (Wood, 1974; Faust et al., 1987; Sanders, 1985). However, Andreae (197 9) found no evidence for the biomethylation of arsenic in the interstitial waters of oxic or anoxic marine sediments, and Aggett and O'Brien (1985) found no methylated arsenic species in lake sediments where conditions should have favored their formation. Additional microbiological processes which occur in sediments can result in the oxidation of arsenite to arsenate by aerobic bacteria, the reduction of arsenate to arsenite, and the reduction of both arsenite and arsenate to volatile arsine (Faust et al., 1987). EFFECTS OF EARLY DIAGENESIS Early diagenetic reactions occurring in the upper layers of sediments can be important in the remobilization of heavy metals (Berner, 1976; 1980). Changes in particle surfaces and changes in metal speciation which occur during early diagenesis can remobilize bound metals (Shaw et al., 1990). As the sediments become buried, the continuing decay of organic matter lowers the redox potential of the sediment. Eventually, iron and manganese oxides begin to dissolve and elements are released The dissolved iron, manganese, and associated elements build up in the porewater and diffuse upward. When the iron and manganese reach oxygenated water they are reoxidized, precipitate as oxides, and scavenge some of the dissolved elements. Elements can continue to diffuse upward to where they can be taken up by biota or scavenged by iron-manganese oxides and organic material; this occurs throughout the sediment column, but principally in the uppermost layers of sediment and in the sediment boundary layer. Any dissolved element which difiusesoutofthesedimentcoltunncanbescavengedbyparticulatematterintheBNL,thus increasing the metal content of the upper layers of sediment and the SBLand producing metal concentration profiles which resemble the effects of anthropogenic input. This set of processes constitutes the redox cycles of iron and manganese, which have been shown to influence the behavior of several elements (e.g. Salomons and FOrstner, 1984; Balistrieri and Murray, 1986; McKee et al., 1989a; Belzile and Tessier, 1990). Much of the biogenic detritus (a major carrier of particle-associated elements) reaching the sea floor is degraded at the sediment-water interface (Gerringa, 1990). This indicates that the transportofmetalstosedimentsby settlingparticlesmay notdirectly contributetopermanent metalaccunmlationinthesedimentcolrunn; ithasbwnsuggestedthattheuptakeofmetalfrom porewaters may be the primary link between detrital flux and metal accumulation in sediments (Shaw et al., 1990). Analysis of the partitioning of elements among the various hydromorphic fractions of the bottomsediment, andintheinterstitialwatersandoverlyingwaters,canbeusedtodeducethe effects of diagenetic chemical changes on the associated elements (e.g. Takamatsu et al., 1985; Moore et al., 1988; Holm, 1988; Farmer and Lovell, 1986; Graybeal and Heath, 1984; Lennan and Brunskill, 1971; Jennett et aL, 1980; McKee et al., 1989a). Partitioning of elements among the hydromorphic phases is most usually defined operationally by the chemical methods used to extract the element from the sediment (Martin et al., 1987). Arsenic in sediments and porewaters appears to follow the diagenetic cycles of iron and manganese, although there is some controversy as to whether arsenic is adsorbed onto hydrous iron and manganese oxides or coprecipitated with them. Farmer and Lovell (1986) found substantial enrichmart of arsenic in the top few centimeters of sediment in Loch Lomond, Scorland, which could not be attributed to any anthropogenic source. Based on element concentrations in the sediments determined by selective extraction procedures, and on porewater profiles, they came to the following conclusions: [1] arsenic is associated with amorphous iron compounds in oxic surface sediments, where it is either adsorbed onto or coprecipitated with fenic oxides and hydroxides; [2] under reducing conditions lower in the sediment column, iron compounds are reduced and dissolved, releasing adsorbed arsenic (or accompanied by the reduction and solubilization of arsenic compounds); [3] both iron and arsenic migrate upward in the porewaters to the oxidized zone, where precipitation and adsorption (or coprecipitation) again takeplace; and [4] theseprocesses produce adiagenetic zone ofarsenic enrichmentnearthe surface of the wdiments. Holm (1988) found a similar association of arsenic with ferric oxide- hydroxide complexes in sediments. He determined that arsenate (AsO43‘) was adsorbed to the surface of these complexes in the same manner as phosphate ions. Aggett and Roberts (1986) determined that arsenate and phosphate ate co-precipitated with hydrous iron oxides in lake sediments rather than adsorbed onto existing surfaces. Moore et a]. (1988) found that arsenic 10 concentrations in porewaters of reservoir sediments were controlled by the solubility of iron and manganese oxyhydroxides in the oxidized zone and of metal sulfides in the reduced zone. Microbial sulfate reduction and decomposition of ferric oxide-hydroxides can also result in the release of arsenic from sediments (Holm, 1988). It has been suggested that mercury in sediments is not affected by diagenesis. For example, Rossmann (1986) concluded that mercury was not affected to any substantial degree by diagenesis inlakeSuperior. This was basedonastudy ofthetotalmereurycontentofsediments. Total metal profiles can resemble the effects of changing inputs (such as increased pollution) even when studies of partitioning among hydromorphic phases indicate diagenetic remobilization is responsible (e.g. McKee et al., 1989a). There is also some experimental evidence for the immobility of mercury in sediments: experiments lasting up to 6 months indicated no diagenetic release of mercury from sediments of a model marine ecosystem (Santschi et al., 1987). Six months, however, is not a long time relative to sedimentation and burial rates and remobilization of mercury may take longer. Other studies have found evidence for the diagenetic remobilization of mercury. In an investigation of the partitioning of mercury in the hydromorphic fractions of sediment from Lake Superior, Stnmk (1991) determined that most of the mercury was associated with the oxidizable (organic matter and sulfides) and base soluble (hurnic and fulvic acid) phases of the sediment, with lesser amounts in the acid soluble (iron and manganese oxide) phases. Concentration profiles of mercury in the base soluble and strongly acid soluble phases suggest that mercury is mobilized from borh phases by diagenetic reactions; however, the fate of the mercury released by such processes was not determined. A similar distribution of mercury among the hydromorphic phases of sediments from the Palos Verdes shelf was documented by Eganhouse et al. (1978). They determined that the enrichment of mercury in surficial sediments appearedtobeduetodiageneticreactions. Indirectevidence fordiageneticrernobilizationof mercmyhasbeendetectedintheAtlanticOceanbyGillandFitzgerald (l988),whoproposethe release of mercury from sediments by diagenetic reactions as the most reasonable explanation for elevated concentrations of meretuy in some ocean waters. Evidence for the diagenetic remobilization of mercury has also been found in fluvial (Jackson et al., 1982) and estuarine 11 environments (Lindberg and Harriss, 1974). Detailed porewater profiles are lacking from all of these studies. Bothner et al. (1980) found evidence for fluxes of dissolved mercury out of contaminated marine sediments under anoxic conditions in in situ bell jar experiments; they attribute these fluxes to the release of mercury following dissolution of iron and manganese oxides. Thesestudiesdescribed aboveindicatethatbotharsenicandmercurymaybereleasedfrorn sediments following burial due to early diagenesis. Although early diagenetic processes may recycle mercury and arsenic within the upper layers of sediment, remobilization processes are not efficient enough to prechrde the permanent burial of sediment-bound elements altogether. The proportion of an element which becomes pemtanently buried is a function of the diagenetic processes and the hydromorphic phase(s) sequestering the element. Phases which appear to be particularly important in the permanent burial of elements are refractory organic matter, sulfides, metastable iron and manganese oxides, and clays (Flirstner and Wittrnann, 1983). Objectives This project was designed to investigate basic controls on geochemical cycling of mercury and arsenic in aquatic environments. The goal was to identify the geochemical processes operating in the sediments, and to determine how these processes influence the cycling of mercury and arsenic. The hypothesis investigated was that as particles move from one layer to another (from the SBL to the sediment column, and with increasing depth in the sediments) toward permanent burial, the composition and chemical character ofthe particles change. The processes that cause these changes influence the cycling of mercury and arsenic, by sequestering these elements within the sediments, by releasing them from sediments, or by repartitioning them among different phases of the sediment. This was pursued by examining the distributions of mercury and arsenic among the waters and particulate matter of different types of aquatic environments. Changes in the composition, mercrrryarrdarserriccontengandpartitioningthatoccmbetween layerscanbeusedtoidentify 12 processes at work. By examining the behavior of two different elements that respond to different conditions and processes in different ways, and by studying these in several diverse settings that undergo different processes to different degrees, a more complete understanding of geochemical cycling should be obtained. Toward this end, two different elements were chosen (arsenic and mercury) and several different sample locations. Sample sites were selected so that variations in diagenetic processes might be observed on several scales, even when employing identical procedures andtechniques. Onecoastalmarineandseverallakesiteswerechosentoexamine differences in diagenesis between freshwater and marine envirmments. Within the freshwater environment, two different lakes were selected, and two different sites in each lake. Locations chosen for this study were selected depositional basins of Lake Michigan, Lake Superior, and the Gulf of Maine. 11. METHOD OF STUDY The approach and general methodology used in this study are described here. Details of sampling, sample preparation, and analytical procedures are presented in Appendix 1. Sampling SAMPLE SITES Deep basins where fine-grained sediments are actively accumulating were chosen as sites for collection of samples. These areas represent locations where the majority of sediment-bound contaminants accumulate due to the process of sediment focusing. Locations of sample sites are shown in Figure 2. The Laurentian Great Lakes were chosen to represent the freshwater environment; three sites with different sedimentological and geochemical characteristics were selected: (1) the Caribou Basin of Lake Superior, which is 335 m deep, with a slow sedimentation rate and a well-defined redox zone within the sediment column; (2) the He Parisienne Basin in Lake Superior, which is 160 m deep, with a rapid sedimentation rate, and a weakly-defined redox zone within the sediment column; and (3) the North (Algoma) Basin of Lake Michigan, which is 200 m deep, with a high sedimentation rate, a high organic matter content, and a redox zone near the sediment-water interface. One depositional basin within the Gulf of Maine was included in this study: the Murray Basin. TheGulfofMainewas chosenasasuitable site forthisstudybecauseitis similartothe Laurentian Great Lakes in several important respects. Both the physical setting and the particle dynamics in the gulf resemble those of the Great Lakes; these similarities are discussed below. 13 l4 - 47' 47:. . b LMNB — 45' l "' 45: LMNB 2 ° Lake :. h ‘3 N Michigan ‘3 20' 315’ 7o 68‘ 66' l I @ Nova Scotia ‘4' - 44° 42' 42' Wilkinson ..'|. . sot. c.....|| l_l I l 70' 68' 66' Figure 2. Locations of sample sites in (a) the Great Lakes and (b) the Gulf of Maine. 15 Because of the partially enclosed nature of the gulf, most of the particles delivered to or generated in the gulf will remain there, eventually accumulating in the depositional basins (Spinrad, 1986). Nepheloid layers have been observed at the thermocline (about 25 m depth) and near the bottom (Spinrad, 1986), corresponding to those in the Great Lakes. A third nepheloid layer has also been observed, associated with the base of the Maine Intermediate Water (Spinrad, 1986). Seasonal influences on the concentration and distribution of suspended particulate matter, such as those observed in the Great lakes (Baker and Eiserueich, 1989), are pronounced ill the Gulf of Maine (Spencer and Sachs, 1970; Spinrad, 1986). . CirculationintheGulfofMaineisquitedifferentfromthatintheGreatlakesandmay affect the cycling of particulate matter and associated contaminants. The gulf is a relatively enclosed basin; the exchange of waters with the Atlantic Ocean is confined mostly to the Northwest Channel (Brooks, 1985). Oceanic water entering though the Northeast Channel is warmer and saltier than other water masses in the gulf. This forms the Maine Bottom Water (MBW), which flows into the deepest parts of the basins (Brooks, 1985). During the summer stratified period, the Maine Surface Water (MSW) and Maine Intermediate Water (MlW) overlie the MBW (Hopkins and Garfield, 1979; Brooks, 1985). These layers are less saline than the MBW (Brooks, 1985). The MIW is cooler than the MSW or MBW during summer stratification (Hopkins and Garfield, 1979). Density differences in the water column are mainly controlled by salinity rather than temperature (Brooks, 1985). This would account for the concentration of particulate matter which has been observed at the base of the MIW: particle settling is slowed at the interface with the denser, more saline MBW. During the winter, the MSW cools and is mixed with the MIW, forming a single water mass (Hopkins and Garfield, 1979). ContanfinantsentefingdleGulfofMainefiomflleAdanficOceanviameMBWmay accumulate withparticulatematterinthedeep basinsviasedimentfocusing. Forexample,Gill and Fitzgerald (1988) observed that concentrations of merwry in water samples from the Gulf of Maine were lower than in samples from the adjacent continental slope, and suggested that the gulf maybeasinkformerwryenteringfromtheAtlanticOcean. 16 SAMPLE COLLECTION Sampling for this project made use of a research ship, the R/V Seward Johnson (equipped with a gravity coring system and suitable laboratory space) and submersible, the DSRV Johnson- Sea-Ll'nk II (equipped with a mechanical arm and nepheloid/SBL sampling system, as described in McKee et al., 1989a). Samples taken included column waters, water and suspended material in the nepheloid, benthic nepheloid, and SBL layers, and bottom sediments with associated porewaters. Samples of benthic nepheloid and SBL were collected via the suction filtration apparatus designed for the submersible. Although some nepheloid and benthic nepheloid particulate matter was collected at each site, there was not enough to process for chemical analysis. Box cores (15 cm x 15 cm x 40 cm, stainless steel) and short cores (7.6 cm butyrate) were also collected from the submersible. In addition to samples collected from the submersible, long cores were collected by gravity coring from the surface ship. All of the cores used for pH, alkalinity, arsenic, and mercury analyses were taken by gravity coring procedures, and are designated "gc". CLEAN PROCEDURES Precautions were taken to prevent contamination from any of the sampling, processing, or analytical procedures. Details of clean procedures are described in Appendix 1. Samples for rneruuyandarsenicanalysismflycamemcontaawimmatefialwhidlhadbeulacid-cleanedand stored in plastic bags. Only distilled deionized water (DDW) was used for cleaning and sample processing. Gloves were worn at all times while handling samples, sample processing equipment, or sample containers. Care was taken to avoid airbome contamination and most shipboard sample processing was performed ill closed plastic glove-bags purged with nitrogen gas. Sample processing in the laboratory was performed within clean hoods supplied with filtered air (passed through a Class 100 filter). SHIPBOARD SAMPLE PROCESSING Sample containers and all sample-processing equipment were acid-cleaned before use (see Appendix 1). All samples collected for arsenic analysis, and samples fi'om the Gulf of Maine 17 collected for mercury analysis, were processed in an inert atmosphere (utilizing Nz-filled glove bags). Collection and processing of samples for mercury analysis was performed under oxidizing conditions (open to the atmosphere) at the Great Lakes sites; this was intended to prevent the loss of volatile reduced mercury (Strunk, 1991). Cores were stored at 4'C (approximate in situ temperature) and sectioned within a few hours of collection. The sections were transferred to acid-cleaned 50 mL polyallomer centrifuge tubes and centrifuged at 15,000 rpm (using a chilled centrifuge head to keep the temperature near 4'C) to separate the porewaters from the sediment. Following removal of porewater, sediment samples were stored frozen in the centrifuge tubes. Porewaters were removed from centrifuged samples by syringe, filtered through acid-cleaned 0.4 pm Nucleopore membrane filters, acidified to pH < 2 with sub-boiling distilled UltrexT“ nitric acid, and stored in acid-cleaned polyethylene bottles. Samples to be analyzed for mercury were also preserved with gold (chloroauric acid) and hermetically sealed following the prowdures of Moody et al. (197 6) as recommended by Gill and Fitzgerald (1987). All water sample bottles were sealed in plastic bags and stored in a cold room at 4’C. Procedural blanks were carried through all processing steps. pH and alkalinity were measured on one sediment core from each site, which was sectioned exposed to the atmosphere. pH was measured by inserting a spear-tip electrode (Orion Ross combination pH) into the wet sediment before removing each section. Alkalinity was measured in porewater samples using an apparatus designed for small-volume titrations; results were converted to mg/L HCO3'. Sequential Extractions Sequential chemical extractions were employed to examine the partitioning of mercury and arsenic among the hydromorphic phases of the sediment. In this procedure, samples are treated with a series of successively harsher chemicals to remove metals from the sediment. Metals are released inresponsetotl'lechangeinchemicalenvironmentproduced bytheextractant,so "phases" are really operationally defined Each extraction, however, is believed to affect primarily 18 one (or more) physical phase of the sediment, which responds to the extractant; thus each operationally-defined phase roughly corresponds to a physical portion of the sediment. Although there is some controversy surrounding the use of sequential chemical extractions to examine partitioning of metals in sediment (e.g., Rendell et al., 1980; Tipping etal., 1985; Nirel et al., 1985; Rapin et al., 1986; Kersten and Forstner, 1987 ; Kheboian and Bauer, 1987; boring and Rantala, 1988; Rauret et al., 1989; Papp et al., 1991), there is a general consensus that--as long as limitations are noted—useful insights into metal partitioning can be gained by use ofthis type of procedure (e.g., McKee et al., 1989a; Prohic and Kniewald, 1987; Martin et al., 1987; Boust et al., 1988; Belzile et al., 1989; Aggett and Roberts 1986; Salomons and Forstner, 1984; Santschi et al., 1987; El Ghobary and Latouche, 1986; Belzile and Tessier, 1990). The extraction solutions, conditions, and sediment phases theoretically affected are summarized in Table 1. All reagents used were analytical reagent grade, prepared with distilled deionized water. ARSENIC Arsenic was extracted from the hydromorphic phases of sediments following the procedures determined by McKee (1989a); these procedures were modified from Tessier et a1. (1979) and Gephart (1982) and are summarized in Table l. The duration of each extraction step was verified for arsenic by steady-state analysis (see Appendix 1). Samples were thawed in a refrigerator, but not dried. Aliquots were placed in acid-cleaned tared centrifuge tubes, weighed, and treated with (1) magnesium chloride solution to remove the exchangeable arsenic [EX fraction]; (2) sodium acetate/acetic acid to dissolve carbonates and remove arsenic associated with the weak-acid soluble phase [WAS fraction]; (3) hydroxylarnine hydrochloride in nitric acid to release arsenic associated with the easily reducible phases [ER fraction]; (4) hydroxylarnine hydrochloride in acetic acid to extract arsenic associated with the moderately reducible phases [MR fraction]; and (5) hydrogen peroxide and nitric acid, followed by ammonium acetate, to release arsenic associated with the oxidizable phases [OX fraction]. All processing was performed under an inert (N 7) atrncsphere until the final (oxidizing) step. Leachates were analyzed as described below. 19 Table 1 Summary of Methods Used for Sequential Chemical Extractions SEDIMENT CHEMICAL EXTRACTION EXTRACTION SUBSTRATE PHASE SOLUTION’ CONDITIONS A. Arsenic (from McKee, 1990) Clay Minerals Exchangeable 1.0 M MgClz, 7 pH 20°C, 1 hour EX V 10 mL Carbonates Weak-Acid Soluble 1.0 M NaAc, 5 pH 20°C, 5 hours WAS 10 ml. Mn Oxides Easily Reducible 0.1 M NHZOH-HCI 20°C, 1/2 hour ER in 0.01 N HNO3 25 mL Fe Oxides Moderately Reducible 0.04 M NH20H~HC1 90°C, 5 hours MR in 25% (v/v) HAc 20 mL Organics Oxidizable 30% H202, 2 pH, 8 mL 85°C, 5 hours & Sulfides OX 0.02 N HNO3, 3 mL that add 3.2 M Nl-[4Ac, 5 mL 20°C, 1 hour ii .(i (oi then add ‘ HZOtomakeZSmL B. Mercury (from Stmnk, 1991) Clay Minerals Exchangeable 10% KC] 20°C, 1 hour EX 15 mL Humic & Base Soluble 0.1 N NaOH 20°C, 30 hours Fulvic Acids BS 15 mL Fe & Mn Acid Soluble 1.0 N HCl 20°C, 6 hours Oxides AS 10 mL Organics Oxidizable 30% H202, 2 pH, 7 mL 50°C, 5 hours & Sulfides OX 0.02 N HNOg, 2 mL then add 4 mL 2.0 M NH4C1 in 20% HNO3 20°C, 1 hour then add H20 to make 25 mL a"Volumes optimized for 1.0 g sample. MERCURY Mercury was extracted from the hydromorphic fractions of the sediments using the selective chemical extraction procedures determined by Strunk (1991). Samples were prepared as for arsenic, then treawd with (1) potassium chloride to remove exchangeable mercury [EX fraction]; (2) sodium hydroxide to remove base-soluble mercury [BS fraction]; (3) hydrochloric acid to remove acid-soluble mercury [AS fraction]; and (4) hydrogen peroxide and nitric acid, followed by ammonium chloride in nitric acid to extract oxidizable mercury [OX fraction]. Leachates were analyzed immediately, as described below. Analytical Procedures ARSEMC Arsenic in liquid samples was analyzed by graphite fumace atomic absorption, utilizing a Perkin-Elmer Iceman/5100 with Zeeman background correction and autosarnpler. Stabilized temperature platform furnace (STPF) procedures were followed (see Appendix 1). Using ST'PF techniques, graphite furnace analyses are interference-free, and highly stable and repeatable (Beaty, 1988). Blanks and standards were prepared in extraction solutions (for leachates) or in distilled deionized water (for water samples). Each analysis was performed in triplicate. MERCURY Mercury in samples was analyzed by hydride-reduction/flow-injection, using a Perkin-Elmer Iceman/5100 with MHS/FIAS-200 equipped with autosarnpler. Preconcentration of mercury in water samples was performed by amalgamation onto gold using the Perkin-Elmer Amalgam System accessory. Blanks and standards were prepared in extraction solutions (for leachates) or ill distilled deionized water (for water samples). Each analysis was performed in triplicate. ORGANIC CARBON The organic carbon content of sediment samples was measured in splits of the core samples which were used for chemical extractions for arsenic, following the modified Walkley-Black titration procedtne of Gaudette et al. (1974). III. RESULTS Samples of sediment, porewater, and column water were collected at two sites in the North Basin of Lake Michigan, and at each of the other sites (see Figure 2). Preliminary shipboard descripu'ons of sediment cores are presented in Appendix 2 (Table A2-1). We were unable to collect adequate samples of SBL sediment from the Gulf of Maine (this layer was abesent at the time of sampling), or of particulate matter from the nepheloid or BNL at any of these sites (due to equipment problems) to perform chemical analyses. Results of chemical analyses are presented in Appendix 3. Site Characterization Supplemental data on dissolved iron in porewaters, organic carbon content of sediments, and the pH and alkalinity of porewaters were acquired to aid in characterizing each site. Iron, organic earbon, pH, and alkalinity can all be used to examine the extent and effects of early diagenesis in sediments. Changes in these parameters can be used to help identify the early diagenetic processes occurring and how they affect mercury and arsenic. FERROUS IRON Profiles of Fe(II) in porewater are shown in Figure 3. These data were provided by J.D.McKee (unpublished data). At all sites, dissolved ferrous iron is undetectable at the sediment-water interface, and concentrations increase in porewater at some depth below. This increase occurs verynearthesurfacein LakeMichigan andlle Parisienne, butmuch weperinthe sediments in Caribou Basin and the Gulf of Maine. At each of the lake sites, there is a narrow zone of lower iron concentration just below the the initial peak; below this the concentrations increase once again. 21 Depth in Core (cm) 22 Ferrous iron (mg/L) 0 1 2 3 O 1 o 'fi—‘m o ‘ ‘Q. % Lake \ ' ‘ Michigan North 20 - 2° ‘ Basin - 2 \ core 889cm I 40 . / 40 . / I so " Lake Michigan 3° J \ North Basin - 1 core 88904 80 80 O 1 2 4 5 6 0 0.5 0 at.» L L ‘ 0 1 09‘“: Lake Superior L». _ Caribou Basin 20 . / 2° ' - '3' core aagczs 40 . 1 4o . 60 ' Lake Superiok 60 r a/ lie Parisienne core Bagels 80 80 0 0.6 1 o s— a s’ 40 1 5° ‘ Gulf of Maine Hurray Basin core 89906 80 Figure 3. Ferrous iron in porewater. 23 Maximum values of Fe(Il) are highest in Ile Parisienne and lowest in the Caribou Basin. There is also much more iron in Lake Michigan site 1 than in site 2 samples. In the Gulf of Maine, Peal) is restricted to a layer, between about 10 and 20 cm depth. In all of the lake sites, concentrations are variable, but tend to continue increasing with depth in the sediment. Oxidation Potentials Profiles offerrous iron in porewater can be used to delineate the various redox zones in the sediment column: where ferrous iron is absent, sediments are more oxidizing, although 02 may be absent near the base of this zone (Bemer, 1980). This environment extends to a depth of 1 cm at Lake Michigan site 1 (LMNB-l), a depth of 3 cm at Lake Michigan site 2 (LMNB-Z), a depth of 2 cm at Ile Parisienne (LSIP), a depth of 18 cm at Caribou Basin (LSCB), and a depth of 9 cm at the Gulf of Maine site (GMMB). The redox horizon, where iron is reduced and iron oxides dissolve to produce ferrous iron, is indicated by a peak in porewater Fe (II) concentrations. This occurs at adepth of5 cmatLMNB-l, adepth of lOcmatLMNB-Z, adepth of7 cmatLSIP,a depth of 19 cm at LSCB, and at a depth of 15 cm at GMMB. Above the redox horizon, ferrous iron diffuses upward along the concentration gradient, is oxidized, and is precipitated as iron oxides. This constitutes the redox cycle of iron which has been found to influence the behavior of many metals (Forstner and Wittrnann, 1983). pH pH profiles of sediments are displayed in Figure 4. In the Lake Michigan cores, pH of surficial sediments is distinctly lower than in bottom waters, rapidly increases below the uppermost sediments, then draps off slightly with increasing depth At site 1 the pH rises to a maximum of 7.6 at 6 cm depth. This is significantly higher than its value in bottom waters (6.8). Atsite2theincreaseianistoamaximumof7.3,notmuchhigherthanthepH ofthebottom water (7.1). pH then decreases only slightly with depth, reaching a constant value of 7.0 at about 60 cm depth. Depth in Core (cm) pH of Porewater 5.5 5 5.5 7 7.5 8 5.5 5 5.5 7 7.5 l A l l a J o a A? o , .__..——__—__==_"°E——‘ 20 - I/ 20 - I I' \I so - so t I .0 I ). ' Lake Michigan 30 ‘ Lake Michigan North Basin - 1 North Basin . 2 core Bagci core 33097 80 BO 5 5.5 A 7 7.5 5.5 1‘ 7 7.15 8 0 snail—a - 0 . ‘fia - «E _" 20 4 .x" 20 . a/ Lake \. Superior > Caribou _ ‘0 “ ‘0 ‘ Basin \ I can my 60 ‘ ruin Superior 60 ‘ \ . Ila Parisienne \ core 850012 a 50 50 5.5 7 7.15 8 'I 20 - r3 x: A Overiyhgwaer 4O . .r a Porowdor 50 - out oi Maine Murray Basin core ”ch 50 Figure 4. pH of sediments. 25 In lake Superior samples, pH is much more variable with depth in the sediment cohrrm than in the Lake Michigan samples. The decrease in pH from overlying water to surface sediments is slightly less for Ile Parisienne, and much less of Caribou Basin than that seen in lake Michigan. There is no consistent trend with depth in either of the Lake Superior cores, although the fluctuations decrease somewhat below 20 cm depth in both cores. In sediment fromthe GulfofMaine, pH ofthe surficial sedirnentis lowerthanthatofthe bottom waters, and continues dropping to a depth of 4 cm. Values then remain fairly constant, with a slight increase with depth until 22 cm, then pH begins to decrease somewhat. The total degree of variability in pH in this marine sample is lower than that observed in Lake Superior or lake Michigan. ALKALINITY Profiles of porewater alkalinity are shown on Figure 5. In the lake Michigan samples, alkalinity shows a relatively rapid increase in the first few cm, then a slight but continued increase withdepth. Atsite 1 allvaluesinthesedirnentarehigherthaninthelake bottomwaters; atsite2 porewater alkalinity at the sediment surface is identical to that of the bottom water, but is higher at all subsequent depths in the sediment. In the lake Superior samples, porewater alkalinity shows an initial decrease below lake bortom water values, then a continuous increase in alkalinity with depth. AtIleParisiennetherateofincreaseinalkalinityisrapid belowtheminimumvalueat3cm depth, then slows with depth. In the Caribou Basin alkalinity increases slowly until a depth of approximately 30 cm, then increases more rapidly. Alkalinity in bottom waters and near-surface interstitial waters oflake Michigan is nearly three times as high as in lake Superior. The total increase in alkalinity with depth in the sediments is greater for the lake Superior samples. Gulf of Maine porewater alkalinity drops initially from the bottom water value, continues dropping slightly until a depth of 5 cm, then increases, drops sharply at 9 cm, then increases further with depth, to a maximum value of 433 mg/L HCO3'. Below a depth of ~15 cm, alkalinity '6 much higherthaninthe lakes, andthe total increasein alkalinityis substantially higher. Depth in Core (cm) Alkalinity of Porewater (mg/L HCOa' ) 100 150 200 250 0 " an %- \ 20 - 1 so a > I so i 50 . Lake Michigan North Basin - 1 core 85ch 100 0 so 100 150 F———II- r O ‘1 “We 20 . V. \I 40 4 <3. \I .o. ‘l 1. Lake Superior 80 ‘ lie Parisienne core 589cm 100 o 100 200 300 400 500 On 20- 40-1 Pi it: "P- R > Cult of Maine Murray Basin MW 100 .250 200 250 0 ‘ “Is-‘3 '7 20 ‘ ‘l .. 40 " \. ./ 80 ‘ g 80 .. Lake Michigan .11 North Basin - 2 .0 core BBgc‘I \. 100 0 so 100 150 0 20 " ‘ \I 40 r \\3 60 ‘ \.\ i 8 0 . Lake Superior Caribou Basin core Bagels 100 A Ovariyingwdet I Pore writer Figure 5. Alkalinity of porewater. 27 ORGANIC CARBON Profiles of organic carbon content are shown in Figure 6. In lake Michigan, values ill near- surface sediments are slightly greater than 3 % (w/w) organic carbon, decreasing fairly rapidly at first, then more slowly with depth Values seem to stabilize at about 2 % deep in the sediments. Thereis anexcursiontoover3 % organiccarbonat~10cmdepthatsite 1. Atsite l theorganic carbon content of the SBL is lower than that of the uppermost layers of the sediment cohlmn; at site2the valuesinSBL and surficial sedimentsarevery similar. ' In lake Superior, values oforganic carbon are highest in the SBL samples, and decrease rapidly in the sediment column. Values then fluctuate somewhat, and in both areas seem to stabilize at about 1.5 % at depth. Ile Parisienne has a lower organic carbon content in near-surface sediments than any of the other lake sites. Organic carbon in the Gulf of Maine site is slightly lower at the surface than deeper in the core, and there is very little variation in the organic carbon content with depth. The organic carbon content of Gulf of Maine sediments is lower than that of Lake Michigan sediments, and similar to that of more deeply buried lake Superior sediments. Partitioning of Arsenic SEDIMENTS Results of chemical extractions are displayed in Figures 7-11. In the SBL and uppermost layers of sediment, the moderately reducible (MR) and oxidizable (OX) phases sequester by far themost arsenic. Asburialdepth increases,thetotal amount ofarsenic extractedfromthe sediments decreases, and the proportion associated with the MR and OX phases decreases as well. Thedegree ofenridlmentofthesurficial layersinarsenicrelativetodeepersedimentsis greaterintheGreatIakesthanintheGulfofMaine. The relationship between concentrations ofarsenic in the SBL and in the uppermost core sediment varies between sites. In lake Michigan, concentrations are lower in the SBL than in the core top samples at site 1, and approximately equal at site 2. In lake Superior, the SBL is Depth in core (cm) Organic Carbon Content (weight percent) o 1 2 3 4 o 1 2 3 0 cut V I bfi o -_ a t I I _ g Lake Michigan .\/_ 25 _ 25 I North Basin - 2 L core 58gc11 3 n/ / so . .I so - i 75 b Lake Michigan 75 ' ! North Basin - 1 coroBSch 100 100 0 1 2 3 4 0 1 2 3 o b I i n I 0‘. I I ' .II §-I 25 - 25 ~ L—a. i n . 2 50 - /\'I so . 75 b . Lake Superior 75 I Lake Superior iie Parisienne Caribou Basin core 880c15 core Bagc25 100 100 0 1 2 3 4 ° '———ra=-II—r r A SBLsarmies 25 ' a Caeaanplos ‘r .\ I 50 - 75 I can or Maine Murray Basin coreSOch 100 Figure 6. Organic carbon content of sediment. 3:0. 053 C. EGO: Depth in Core (cm) 20 4O 80 100 20 4O 60 80 100 Arsenic in EX Fraction 0.3 o 0.1 0.2 " a : 3:» i ll I It ll II II " Lake Michigan North Basin - 1 ‘ coreaagca o 0.1 0.2 o 3 '1 A? ’I .J '\a in 1 ‘ Gulf oi Maine Murray Basin core 89963 29 rig/9 2O 40 60 BO 100 0.1 0.2 0.3 -l---J Lake Michigan North Basin - 2 core aagctt A SBLearrpiee a Coreeanpies Figure 7. Arsenic extracted from the EX fraction of sediment. Depth in Core (cm) 30 Arsenic in WAS Fraction (Carbonates) ins/9 dry weight 0 0.5 1 1.5 0 0.5 1 1.5 .a 20 i . 20 ‘ 'x. ‘- 7’ 40 - " 4o . I; ‘ A“ 60 " . 60 . r Lake Michigan '|_ Lake Michigan 80 ‘ North Basin - 1 80 i \_ North Basin . 2 core Bagca core sagctt 100 100 0 0.5 1 1.5 9 . 0.5 1 1.5 ' I. 20 . f 20 i"- \ '\ 40 - 4o - _> f . 60 - 1 60 q Lake Superior LakeSuperlor 3° ‘ ile Parisienne 3° ‘ Caribou Basin core 889015 00"! 880¢25 100 100 0 0.5 1 1.5 0 “—3: ‘ 20 i I 15- A SBLserrpies 4o - .II . c I 60 It Gulf of Maine 80 < Murray Basin core 89903 100 Figure 8. Arsenic extracted from the WAS fraction of sediment. 31 Arsenic in ER Fraction (Mn-oxides) Depth in Core (cm) pglg dry weight 0 1 2 3 4 O 1 2 3 o J ‘35-. o r — _"e.-I__. / 20 ‘ 20 .. K. I \s 40 t j. 40 . a’ - I 60 - If 5° .r Lake Michigan 80 / Lake Michigan 9° ‘ North Basin . 1 t I North Basin . 2 core sages core “go“ 100 100 ° ‘. ‘3 f ‘ ° . 1 z t 0 _l.=——I 0 '33: 20 L 20 .‘I \\ 40 - 4o . ./ 60 - so . a Lake Superior Lake Superior 80 ' iie Parisienne 80 . Caribou Basin core Bagcts core was 100 100 0 1 2 3 4 o _L J. l I L 20 Is ‘0 l A SBL armies j a Core ean'pies 60 Gulf of Maine 30 ‘ Murray Basin core 89903 100 Figure 9. Arsenic extracted from the ER fraction of sediment. Depth in Core (cm) 32 Arsenic in MR Fraction (Fe oxides) 119/9 dry weight 0 2 ‘4‘ 6 a 10 o 2 1 e a to o ’r_'__.'j_._:=._—-—o 0 a ,.———E. i 't 20 " 20 a. i " \. \- 40 " Y 40 " 60 ‘ 60 -' Lake Michigan Lake Michigan 80 - North Basin - 1 80 . North Basin - 2 core Bsgca core BBgc11 100 100 ° 2 4 6. 9 1° ‘1 . a 1 9 8 1° o. F.- 5. 0 20 ' 20 E. \ 401 40 ~./' 60 ‘ 60 1 Lake Superior Lake Superior go . lie Parisienne go . Caribou Basin core Bagcis core sauces 100 100 0 1 2 01-— 20* 40- i A SBLearmies a Core armies 60‘ Cult oi Maine .0. Murray Basin core BOgca 100 Figure 10. Arsenic extracted from the MR fraction of sediment. Depth in Core (cm) 33 Arsenic in OX Fraction (Organics & Sulfides) 119/9 dry weight 0 2 4 B 8 10 0 2 4 B 8 10 0 4w 0 . W . 20 - 'X 20 . 'f. \I 40 - “/2. 40 - / \s so . 4 so . a/ Lake Michigan > Lake Michigan so . North Basin - 1 so . > North Basin - 2 core sages core Bagctt 100 100 0 1 2 3 4 5 0 1 2 3 4 o L Q ml ‘ o ‘ }= —=fi 20 . 20 4O - 40 - so ‘ Lake Superior Lake Superior lie Parisienne Caribou Basin core aagcts core aagczs SO 60 0 0.2 0.4 0.8 ° ‘—$; t:- 20 1 a". is so ., } A SBLearrpies e Coreearrpies .0 ‘ Gulf of Maine Murray Basin core 899c3 80 Figure 11. Arsenic extracted from the OX fraction of sediment. 34 distinctly enriched in arsenic compared with the top-most core sediments; in both locations the total extractable arsenic in the SBL is nearly twice the total extractable arsenic in the uppermost core sample. At all of the sites, below the surficial enriched zone there is a layer where total extractable arsenic concentrations are at a minimum; the position of this layer corresponds to the redox zone. Total extractable arsenic concentrations increase again below this depth In most cases, there is a distinct secondary maximum below this minimum at the redox zone, then lower concentrations again, and some additional peaks in total extractable arsenic concentration. The exchangeable (EX) fraction is insignificant in sequestering arsenic in the lake sediments, but is a major phase holding arsenic in the gulf sediments. The proportion of arsenic extracted from the exchangeable fraction is small but not insignificant in oxidized sediments of the gulf, drOps to nearly zero in the redox zone, and is the major contributor of extractable arsenic from the reduced sediments (see Figure 7). The weak-acid soluble (WAS) fracu'on contributes a small amount of arsenic in the Great Lakes; not surprisingly, its contribution is larger in the Gulf of Maine, where carbonate sediments are more abundant (Figure 8). WAS arsenic is also more abundant in Lake Michigan than in lake Superior, where modern sediments do not contain carbonates. In the lakes, the easily reducible (ER) and MR fractions contribute comparable amounts of arsenic, except in the uppermost layer where MR contributes much more, and at a few locations deeper in the sediments where ER concentrations exceed MR concentrations (Figures 9 and 10). In the Gulf of Maine, the ER fraction consistently contributes less than the MR fraction. In Lake Michigan, the oxidizable (OX) fraction (Figure 11) is at least as important as either the ER or MR fractions in holding arsenic in sediments; in Lake Superior it is noticeably less important than these phases, contributing approximately the same amount of arsenic as does the WAS phase. In Gulf of Maine sediments, arsenic from the OX fraction is approximately equal to thatfromtheERandMRfractions. Total extractable arsenic (Figure 12) is highest in Lake Michigan sediments, with near-surface concentrations exceeding 10 ug/g (23 ug/g at site 1 and 11 ug/g at site 2), and declining to a 35 baseline level of about 3-4 ug/g. These values are similar to total arsenic concentrations reported for Lake Michigan by Mudroch et aL (1988): 5-15 ug/g in surficial sediments and 5-8 ug/g as a background level. In Lake Superior, total arsenic at the sediment surface shows considerable enrichment (to ~75 ug/g) at the Ile Parisienne site, but little enrichment (to ~2.2 ug/g) at the Caribou Basin site; baseline values seem to be about 1 ug/g. Overall total extractable arsenic is lower in Lake Superior than in Lake Michigan. In the Gulf of Maine, total extractable arsenic ' reaches amaximumofl.6 ug/gatthe sediment surface, showsasecondpeakof1.7 uglgatSfi cm depth, and shows a baseline value of about 1 ug/g. POREWATER Arsenic concentrations in porewater tend to be low near the sediment-water interface, and higher deeper in the sediment column (Figure 13). At the Caribou Basin site of Lake Superior, the low arsenic concentrations extend to about 13 cm below the sediment-water interface; at each of the other sites there is a distinct concentration gradient near the surface suggesting a flux of arsenic upward toward the sediment-water interface. Lake Michigan site 1 exhibits a nearly classical profile of porewater arsenic: increasing rapidly from a low concentration at the sediment-water interface to a high concentration at the redox zone, with the porewater maximum value occurring immediately below the surface zone of sediment enrichment. At other sites, the maximum porewater concentration occurs at a greater depth below the zone of surface enrichment, although in Lake Michigan site 2 and Ile Parisienne there are peaks (not the largest) immediately below the enriched layer. Multiple peaks in porewater profiles are evident in all of the locations, most notably in the LMNB-Z and [SIP sites. Although sediment concentrations of arsenic are low in the Gulf of Maine, porewater concentrations aremuch higherthan inthe lakes. Concentrations ofarsenic in sediments ofthe GreatLakesare ontheordcroflOOOtimestheporewaterarsenicconcentrations; intthulfof Maine, sediment concentrations are about 100 times the porewater values. Porewater arsenic in thegulfseemstoincreasecontinuously withdepthratherthanreachingamaximumvalueas 36 Total Extractable Arsenic rig/g dry weight 60-1 Lake Michigan Lake Michigan 20- %E f .9“: North Basin - 1 80 North Basin — 1 core sages ' core 88gc11 10 15 0 10 15 _ r‘ 0 L Depth in core (cm) i \ 40-1 /' Lake Superior Lake Superior lie Parisienne Caribou Basin J core 88gc16 core BSgc25 so 3 4 5 b SBL nannies I Core armies Gulf of Maine Murray Basin M39003 Figure 12. Total extractable arsenic. Depth in core (cm) 37 Arsenic in Porewater O 5 10 15 2O 0 -fiI-‘.‘El a \ 20 . f I / 't 40 - f I .i so . go . Lake Michigan North Basin - 1 core sages 100 o 5 10 15 20 0 4 . . 20 .. 40 - I/ i so . \ 30 q ' Lake Superior Ile Parisienne core SSgc15 100 50 °i 20 . 40s 60.. 80. O 10 20 30 40 amx Cult of Maine Murray Basin 007039903 100 (us/L) o s 10 15 20 0 .fi, 1 a a 20 . \- 40 . 50 r a; 80 - )‘ Lake Michigan North Basin - 2 core SSgctt too 0 s 10 ts 20 l 50 . 80 ‘ Lake Superior Caribou Basin core SSgc25 100 Figure 13. Arsenic in porewater. 38 appears to be the case in all of the lake sites. Below the redox zone, porewater shows an inverse relationship to total extractable arsenic. Partitioning of Mercury SEDIMENTS Results of chemical extractions for mercury are displayed in Figures 14-16. Mercury was not detected in the exchangeable fraction of any core. Most of the extractable mercury in all of these cores is associated with the oxidizable fraction, although the base-soluble and acid-soluble fractions contribute significant amounts of mercury in the uppermost layers of sediment. At all of the sites, total extractable mercury (Figure 17) is highly enriched‘in the uppermost layers of sediment, and concentrations decrease rapidly to a background level of approximately 20 ng/g. This is similar to the results of Strunk (1991) for mercury in sediments of Lake Superior. Profiles of mercury in the base soluble (BS) and acid soluble (AS) fractions are very similar to one another, particularly in the two Lake Michigan sites and in Ile Parisienne of Lake Superior (Figures 14 and 15). In both of these fractions, mercury is high near the sediment-water interface, and drops rapidly to very low values. The depth at which this occurs is identical for the two fractions in both Lake Michigan sites and in the Ile Parisienne samples, but in the Caribou Basin and Gulf of Maine cores, the depth at which concentrations drop to near-zero is somewhat deeper for the acid-soluble phase than for the base—soluble phase. Inallofthe lakesamples,mercuryintheoxidizable (OX) fractionincreasesfromthe sediment-water interface to a maximum near the surface, decreases to a minimum immediately below this enriched zone, then increases again to a second maximum before dropping to a relatively constant value (Figure 16). In Lake Michigan, the upper enriched layer is thicker, and moredistinctlyenrichedinmercurythaninlakeSuperior. Inalllakesites,thezoneofOX enrichment occurs directly below the enrichment in the AS and BS fractions. In the gulf, mercury in the OX fraction is enriched in the uppermost sample, and nearly constant below this depth. 39 Mercury in Base Soluble Fraction Depti in core (cm) 40 SO 20 4O 80 80 20 40 SO Figure 14. Mercury extracted from the BS fraction of sediment. (Humic & Fulvic Acids) ng/g dry weight o 10 20 30 40 so 0 ’ 10 20 so so so All: =_... o . ._.___.. 20 L 40 Lake Michigan 80 Lake Michigan North Basin-1 North Basin-2 core 88gc2 1 core 88ch 80 — 0 25 50 75 1 00 1 25 0 1 0 20 30 40 50 1 l j l o . J; 1 A P5 in 20 II II II II ll 40 ll Ii II 'i Lake Superior 80 “ Lake Superior " ile Parisienne " Caribou Basin n core SSgct? “ core SSgc22 ii ll 80 o 10 20 A SBLearrpiee I Coreaanpiee Gulf of Maine Murray Basin core 89908 Depth in core (cm) Mercury in Acid Soluble Fraction (Fe & Mn Oxides) "9’9 dry weight 0 1o 20 30 40 50 0 10 20 30 4o 50 L a L L I a a a a ‘l - k_ 20 20 40 40 Lake Michigan Lake Michigan 3° North Basin-1 5° North Basin-2 core SSch core 880$ 80 7 80 - 0 25 50 75 100 125 0 1o 20 o q l l L k 0 - l = i 20 20 Ii '1 ll 40 ll 40 I II “ Lake Superior Lake Superior 5° “ ile Parisienne 3° Caribw Basin 1! core 88gc17 ? core 889m I 80 80 0 10 20 i . 0 .I'd' 20 .J A SBLearrpioe 40 I Coreaanpies Gulf of Maine 3° Murray Basin core ”ch 80 . Figure 15. Mercury extracted from the AS fraction of sediment Depth in core (cm) 41 Mercury in Oxidizable Fraction (Organics & Suitides) nglg dry weight 0 20 40 SO SO 0 25 50 75 100 125 o . ‘ o._..___.j o ‘ g ' ‘ ' _ = ; I _ m " ‘7 20 d k 20 d to « // 4o - " Lake Michigan 80 a so .. Lake Michigan / North Basin-1 North Basin-2 > core SSgc2 core am SO 80 «L 0 10 20 30 40 50 0 26 60 75 100 o l L o ‘ E J l .I tI "La- :3- l’ d ., 20 .. 7 20 / 1 I 40 - \ 4O . ’ i .0 . Lake Superior 60 ‘ I Lake Superior lle Parisienne '\ Caribou Basin '\ core 880m 7 ' core SSgc22 SO 80 O 10 20 30 40 50 o l I l l :5 .l 20 ‘ ’) <3- 6 SBLsanpies e I Coreaarrpies 40 d 80 Gulf of Maine ' Murray Basin core 8Sgc8 SO Figure 16. Mercury extracted from the OX fraction of sediment. 42 Profiles of total extractable mercury (Figure 17) are very similar to OX profiles, with additional enrichment of surficial sediments reflecting contributions of the BS and AS phases. Total extractable mercury is highest in the SBL sample only in site 2 of Lake Michigan; at the other three lake sites the SBL has lower total merctn'y than the uppermost sediments. Lake Michigan site 2 also displays the most distinct sub-surface maximum, refbcting the highest maximum values in mercury from the OX fraction. Values of total mercury in sediments of Lake Michigan's depositional basins have been reported in the range 0.030 to 0.380 ug/g in surface sediments; values reported for Lake Superior range from 0.094 to 0.160 nglg in surface sediments, and 0.044 to 0.68 ug/g for "background" levels (Mudroch et al., 1988). Results from this study are comparable, although baseline concentrations of total extractable mercury are generally lower (~0.02 nglg). POREWATER Mercury porewater profiles in the lakes exhibit high concentrations near the surface, a zone of low/minimum concentration below this, then higher concentrations again at depth in the cores (Figure 18). These rrrinima in dissolved mercury do not correspond to the major enriched zones inthetotalextractablesedimentmercury; theydocorrespondtosecondarymaximainthe oxidizable fraction (see Figure 16), although the porewater minirna span a greater depth distribution. Mercury concentrations in the Lake Superior Caribou Basin core are much higherthan in the othercores; alsotheminimumconcentrationishigherthanthatintheotherlake locations. The Lake Superior profile more closely resembles the Lake Michigan profiles, although peaks below the minimum are higher in Ile Parisienne. In the Gulf of Maine core, dissolved mercury is uniformly low in porewaters, although slightly higher in the vicinity of the redox zone (near 15 cm depth) and lower near the sediment surface and deeper in the core. Depth in core (CM) 43 Total Extractable Mercury nglg dry weight 0 25 50 75 100 125 0 25 50 75 100 125 o ‘ L l l l o d J J L .l I - —P‘l 20 I 20 ‘ { so . 4O 4 80 ' Lake Michigan 50 Lake Michigan 1 I North Basin-1 ‘ North Basin-2 \ core 83902 core 88ch 80 I go ...a 0 25 50 75 100 125 150 175 0 25 50 75 100 o . A L l A l 1 o d l 1 i” 0 « 20 ' 2 'l I 40 - 40 < I) 50 Lake Superior 50 1 Lake Superior ‘ ile Parisienne ‘ \ Caribou Basin I‘. core 889c17 I core SSgc22 80 80 0 25 50 75 L 1 ° ' {Ii—'- 5 I 2 A SBL Sanpiee e’ I CoreSarmies 20 - f A Cult of Maine 3° '1 Murray Basin \1 core 89905 40 Figure 17. Total extractable mercin'y. Depth in Core (cm) Mercury in Porewater ~ ngIL 0 200 400 o 200 400 o 4g; . o E? . LORI n 20 Michigan 20 ., Lake Michigan North Basin-1 ' North Basin-2 \ core 88gc2 k core 889w I 40 - \ 4o . so 'l 60 at \- / \ 80 80 L o 200 400 500 800 0 200 400 500 800 o a -‘ in a 1 o __ a J I I 20 -' 20 ~ I 40 40 . / _)' \. 50 Lake Superior 80 . Lake Superior lie Parisienne Caribou Basin core 889c17 core 889c22 80 80 0 200 400 o a 1% 2"; <1 20 « e/ \- Gulf of Maine Murray Basin core 8Sge8 40 Figure 18. Mercury in porewater. IV. DISCUSSION Examination of the data shows several interesting results: - Changes in the partitioning of mercury and arsenic which accompany increasing burial depth indicate that both elements are mobilized and repartitioned during early diagenesis at all of the sites investigated. 0 Mercury and arsenic behave differently during early diagenesis. The repartitioning of mercury appears to be more extensive than is that of arsenic. The "enrichment" of mercury near the sediment-water interface relative to W sediments is also more prominent. - Concentration gradients of arsenic in porewaters indicate that there is a flux of arsenic from the sediments to the sediment-water interface via porewater at all of the sites except the Caribou Basin of Lake Superior. . Concentration profiles of mercury in porewaters are more complicated than those for arsenic, but there are gradients suggesting some flux of mercury toward the sediment-water interface at all ofthe Great Lakes sites. There does not appear to be a significant upward flux ofmercury from sediments of the Gulf of Maine. . The apparent upward diffusive fluxes of mercury and arsenic released during early diagenesis contribute to the enrichment of‘surface sediments in these metals. High concentration of mercury and arsenic exist in surface sediments even when porewater gradients are not distinct. Enrichment appcatstobcgtcatctinmcotcstukcssitcttitanmthcourromnm. . There is notable variability in diagenetic conditions among sites. Evidence for differences is provided by data for organic carbon, ferrous iron, pH and alkalinity of sediments and porewaters. These are discussed in detail in the following sections. To facilitate comparisons, summaries of geochemical data from each sample site are presented in Figures 19-23. 45 Depth in Core (cm) Lake Michigan North Basin - 1 rig/L 0 50 100 150 200 250 0 "0’0 0 25 50 75 100 125 0 ans 0 'lb ----:::::::... . ...;------ \mi ‘, Extractabie 20 4 20 J : hole ‘0‘ ‘0‘ --‘--'I. ‘gfi'POI'O ’ \wder 80 u 80 - ‘-~-‘ ngIL :: Arsenic "no" ' Core88 3 " 30 gr: 80 Core 88ch 0 1 2 3 0 1 2 3 4 a I l l L 1 L .2. . . M '- A SBL 20 q \ 2° ‘ ‘ I 00'. 4o ~i 40 . I 80 . 80 4 Organic Ferrous iron 0"50" C 88 [no/L W 90 ’0 0" 8C4 80 ‘ Core 88§c3 100 150 200 250 5.5 8 8.5 7 7.5 8 o _ a i t o t I t l a a4? 20 'l l_ 20 " A bottom water ‘0 . l— 40 . Alkalinity i 3 P" I 80 - 50 ‘i j .0 Core £8351 _ 80 Core 885:! _ Figure 19. Summary of chemical data: Lake Michigan North Basin - 1. Depth in Core (cm) 47 Lake Michigan North Basin - 2 0 15 20 O - . ‘ A ‘0 «I \ Total SBL emanate 2° ‘ (We) 20 1 4o 1 "3‘" 40 4 Porewater eo . pgli. eo . so 4 so 4 Core 88ch I Arsenic 100 100 O 1 2 1 2 3 4 ° La 4 M I Arc?“ Ferrous ' 20 . ‘ Organic iron 20 Carbon \- "'°’L Worm '/ 40 -I /I 40 a i 50 ‘ so .. I/ A 88L a Core so ' ' - so - .I Care 88ch 0 Core 88ch] 100 100 5.5 6.0 5.5 7.0 7.5 100 150 200 250 o (W o ’—1§_‘.. l A .. 'I — -._- A / . 7 at I' ? l— \a .- ‘0 1 pH y 40 ' Alkalinity '5. . Mm; .\ SO « ) 50 r ./ I SO - 1 80 i i \. Co 88 7 Car 88 7 100 n 8c ' 100 ' 8c Figure 20. Summary of chemical data: Lake Michigan North Basin - 2. Depth in Core (cm) Lake Superior Ile Parisienne O 150 300 450 600 ngIL o s 10 15 0 so too 150 zoo nglg 0 . ‘ ir— . {sis-:22- ssq.’;;1ii- --s::a:%~ 2° 1 Ton] 0"P0l'ml' Extraaable t' "9"- ue/a so - ,v 1‘ Arsenic 50 "' “0‘ l t Care 38861-5 Core 883c17 80 O 2 4 6 O 1 2 3 o “w o - __ M I I-r_____. T k. 20 .. 20 q A SBL so . 40 .. I Core Ferrous \ 0'9“." iron Carbon "‘0"- W 96 60 -I 60 ‘ Core 88ch 5 Core 883c15 80 80 5.0 5.5 7.0 7.5 O 50 100 150 o 4—-—-_.::==-‘- __‘ _ o . ” ‘ - A Bottom . a' water I 20 . ,/ 20 d \ . ”mm“ , I/ \I \I so - all \f so . r Alkalinity null-1003. \- V' / I Care 883::12 Core 88ch2 80 80 Figure 21. Summary of chemical data: Lake Superior Ile Parisienne. Depth in Core (cm) 49 Lake Superior Caribou Basin 20. 40-1 so 4 110/0 Arsenic Core 88ch5 80 0 0.5 1 o a Ferrous iron 2° *K "'°" 40 d _,¢H*”::>. I\ so - tr/ Core 888c25 80 0.5 7.0 7.5 8.0 o i—fi-ai ‘ _ 20 'l 40 'l .\ 9" I so . .2 \ '\. 80 Core 88ch9 250 500 750 25 50 75 I A l 1000 ngIL too 0010 ---I-..‘- '. --.-.0 o'- :' Porewater I nglL .9’ 5° ‘ t’ ‘. Mercury Core 883c22 80 0 1 2 3 s 0 a 1 i .I 20 s A SBL so . ./ I Core 30 ‘ Organic Carbon 90 Core 883c25 W 80 O 50 100 150 o i . ‘ A bottom 20 4 .EL, 1 \ 40 r \ Alkalinity "ION-H003 so . \} Core 883c19 BO Figure 22. Summary of chemical data: Lake Superior Caribou Basin. Depth in Core (cm) 10 20 30 40 50 50 Gulf of Maine Murray Basin 50 o 100 200 ngIL 5 0 so 100 nglg 10s 20- 301 4° ‘ Porewater I ng L Core 89gc8 50 . 10 i Organic H Carbon r Ferrous I” 20 - II iron .1 mgIL 30 ~ I 1 so « ./ Core 89gc3 \ 50 0 100 200 300 400 500 o a .-’a; a a 10 . h 1- Alkalinity. 1N Mme, 20 a 30 d \I 40 1 Core 89gc5 Core 393:5 50 Figure 23. Summary of chemical data: Gulf of Maine Murray Basin. 51 Diagenetic Processes: Evidence and Effects EARLY DIAGENESIS OF ARSENIC Theinfluence ofearlydiagenesisonarseniccanbedetemrined by exarniningprofilesof arsenic concentrations in sediments and porewaters. Profiles of total extractable arsenic in sediments, although produced in part by diagenetic processes, they do not reveal much about . these processes. Changes in partitioning of arsenic among the hydromorphic fractions of the sediment are also caused by early diagenesis, but these do provide evidence for the operation of individual processes. Oxidized Zone The redox cycling of iron is one of the principal influences on diagenesis of arsenic. There are three lines of evidence for this, as follows. First, total arsenic is enriched in the upper layers of the sediment column at all of the sites (Figure 12). This type of enrichment has been attributed to the adsorption of upward-diffusing porewater arsenic by iron oxides in the oxidized zone of sediments (Farmer and Lovell, 1986; Belzile, 1988; Belzile and Tessier, 1990). The enriched zone lies above the iron redox zone (defined by the appearance of dissolved ferrous iron in porewater) in all of the sites. Second, profiles of arsenic from the moderately reducible (MR) fraction of sediment (Figure 10) show distinct enrichment of the upper, oxidized layers of sediment. Since arsenic from this fractiorrisprirnarilyassociatedwithironoxides,dreeviderrtenrichnrenthrdreMRfiacfiorris consistent with the adsorption of upward-diffusing arsenic from porewater. .‘lhis supports the ideathatarsenicisassociatedwiththeredox cycling ofiron. Thecnrichmentofarsenicinthe easily reducible fraction is less distinct, suggesting that manganese oxides play a minor role in the redox cycling of arsenic. This is consistent with the results of other investigations (e.g. Aggett and Roberts, 1986). Third, porewater profiles of dissolved arsenic (Figure 13) support this explanation. At all of thesites,thereisageneralgradientofdissolved arsenicfromhigh valuesirrthereducedzoneto 52 lower values in the oxidized zone. Dissolved arsenic is produced during the reductive dissolution of iron oxides, and diffuses up along the concentration gradient toward the oxidized zone, where arsenic is removed from the porewater by adsorption onto solid phases. This is shown by the abruptdecreaseinporewaterarsenicwithintheenriclwdzone atallsites (except fortheCaribou Basin, wheretheredox zoneisquitedeepinthe sedimentcolurnn). IntheLakeMichigansites,m IlePafisiaine,andmdreGulfofMainedrepomwaterarsenicgradiaitisquitesteepjustbelowthe sediment-water interface, suggesting that significant diffusive fluxes out of the sediments are possible; this is discussed further below. At the Caribou Basin site, the gradient of porewater arsenic shows that diffusion should occur from the reducing sediments up to just above the redox zone (about 15 cm depth). This depth coincides with a slight enrichment of arsenic in the MR and oxidizable (OX) fractions. Above this, porewater arsenic concentrations are very low, suggesting nearly complete removal of dissolved arsenic from the porewater in and above the redox zone. Relatively high arsenic concentrations do exist in the surface sediments, even though porewater gradients indicate that redox cycling does not provide arsenic to the sediment surface. The degree of enrichment of surface sediments relative to "background" concentrations is much less at this site than is observed at other lake sites. Also, there are two depths (at ~25 cm and ~40 cm) where total extractable arsenicisequaltoconcentrations foundatthe sediment surface. Thesefacts suggestthaeinthe Caribou Basin, concentrations at the sediment surface are elevated due to some process other than redox cycling. Degradation oforganic matter also plays an important role in arsenic diagenesis. Above the redox zone, organic matter is first degraded aerobically. The decay of organic matter releases associated arsenic, contributing to the increase in porewater arsenic just below the sediment-water interface which occurs at all ofthe sites. In general, oxidizable arsenic profiles (Figure 11) closely resemble those oforganic carbon content (Figure 6). In the Great Lakes sites there is a distinctzone ofdecreasing organiccarboncontentimmediately belowthesedirnentwater interface, extending to a depth of several cm; there is a corresponding sharp decline in arsenic associatedwiththeOXfractioninallofthelakes sites. Asimilardecreaseinconcentrationsof 53 cadmium in near-surface sediments of the Laurentian Trough has been noted, and attributed to the aerobic degradation of organic matter (Gobeil et al., 1987; Gratton et al., 1990). There appears to be a strong link between arsenic and organic matter. This relationship is most evident in the lake Michigan sediments, as can be seen in a plot of arsenic concentration versus organic carbon content (Figure 24a). Additionally, there is a considerable decrease in oxidizable arsenic between the SBL and the tops of cores in lake Superior, suggesting that arsenic is lost from easily oxidizable organic matter that is largely decomposed before being buried in the sediment column. ThisdecreaseisnotnotedinlakeMichigan,wheretheorganiccarboncontentoftheSBLis similar to (LMNB-Z) or lower than (LMNB-l) that of the surface sediments. .This difference is discussed in more detail in the section on diagenetic variability. In the Gulfof Maine, there is also a sharp decrease in OX arsenic just below the sediment-water interface, even though no concomitant decrease in organic carbon content is observed. This suggests that the loss of arsenic from the OX phase is caused by some process other than simple degradation of organic matter. Other phases sequestering arsenic do not show such clear distribution patterns. Patterns are most evident in profiles from lake Michigan site 1, where exchangeable (EX), weak-acid soluble (WAS), and easily reducible (ER) fractions show enrichment in upper layers. The arsenic enrichment of these phases probably results from the uptake of arsenic from porewater—arsenic provided by the decay of organic matter and arsenic which diffused upward from the reduced sediments. The adsorption of arsenic from porewaters appears to be very effective within the oxidized layers of sediment: most of the arsenic released from decomposing organic matter is transferredtoothersolidphasesratherthan accumulatingintheporewater. ThefactthattheMR fractions of the SBL samples do not show additional enrichment over the upper layers of the core sediments also indicates that adsorption is efficient within these upper layers, removing upward- diffusing arsenic from porewaters before it reaches the SBL. Surrdby et al. (1986) found that diffusion of metals out of fiord sediments did not occur even though porewater concentrations were higher than those of overlying waters; they attributed this to fixation of metals by oxygen diffusing into the sediment. ' 25 (a) Extraetabie Arsenic vs. ' 20 _ 96 Organic Carbon 3 g 15 .. g 10 r I < e . D D I “ -5h#t. O o 0 ‘ °° . 0° - Q!” t & O 1 2 3 Organic Carbon (weight %) I D O O A LM-NB1 LM-NB2 LS-lP LS-CB GOM 200 (b) Extractable Mercury vs. 0 96 Organic Carbon A 150 " or E, '5 V a a 100 - _ n g e I DO 6 O) a O D 2 so - . ' 0 D O O O. D O aseeq.:’§§a§:a.&m Uas O 1 2 3 Organic Carbon (weight %) Figure 24. Relationship between sediment organic carbon content and (a) t0tal extractable arsenic, and (b) total extractable mercury in sediment. 55 The same patterns of arsenic distribution among the EX, WAS, and ER fractions are not as clear in Lake Michigan site 2, however, even though arsenic concentrations in SBL samples and ”background" concentrations of these fractions are similar for the two sites (Figures 7-9). Total arsenic profiles are similar, except for the degree of enrichment of the uppermost sediments, but comparisons of the EX, WAS, and ER fractions show lack of enrichment, and MR and OX fractions show less enrichment than is seen for LMNB-l. One possible explanation is that the surface enrichment at LMNB-2 has been obscured by bioturbation. The porewater profiles (Figures 19 and 20) are also quite different. At LMNB-l, dissolved arsenic increases rapidly belowthesediment-waterinterfacetoamaxirrmmat4cmdepthfiustbelowthezone ofsediment enrichment) then decreases slowly with depth, with few minor excursions from this general trend. The site 2 profile shows an initial peak just below the enriched sediment layer, then numerous higher concentration peaks at depth. This type of profile may be caused by bioturbation and bioirrigation (Belzile, 1988). These processes can also enhance the fluxes of dissolved metals out of the sediments (Gratton et al., 1990); this may account for the lesser degree of enrichment of surficial sediments at LMNB-Z relative to LMNB-l. Thus the differences in both sediment and porewater arsenic profiles may be caused by differences in the degree of bioturbation at the two sites. The Ile Parisienne site also shows some evidence of bioturbation (multiple peaks in porewater arsenic). Effects of bioturbation on arsenic diagenesis are discussed further below. In the Caribou Basin core, the peak in sediment-bound arsenic at ~25 cm depth is associated withastrongly enriched zmeintheMRphase (Figure 10) and slightenrichmentintheERand OXphases (Figures 9and11). Thishyerisflieredoxcruscalayerdistinctinappearancefrom overlying and underlying sediments, where a concentration of iron oxides is developed just above the iron reduction zone. This layer coincides with the top of a large gradient in porewater arsenic (see Figure 22), and may represent uptake of arsenic released from below. Just above this zone, at~20to24cmdepth,aresmallpeaksinferrous iron,dissolvedarsenic,andorganiccarbon (Figure 22). This layer consists of sediments that are mottled in appearance (see Table A2-1), and may represent a relict redox horizon. This may have developed due to relatively recent changes in conditions in this basin, such as variations in organic matter inputs which can alter the effective 56 depth of oxygen penetration (Pedersen et al., 1986); a similar explanation has been proposed to explain manganese profiles in the Caribou Basin (McKee et al., 1989b). This does not explain the increased organic carbon content of this zone, however. Jahnke et a1. (1989) noted the presence ofadecp ”reaction layer" which was enriched in organic carbon in many sites in suboxic sedimentsoftheeastemequatorialAtlantic ocean. This layerisbelievedtobearelictorganic—rich layer that is still decaying, possibly a turbidite deposit (Jahnke et al., 1989). To summarize, the elevated concentrations of arsenic in the upper layers of sediment are partly dawdefiammsedinenmbemgdeposhedcmminmomammicassodawdwimmganicmauer than do those that have been buried; partly due to the repartitioning of arsenic within the oxidind layers as organicrnatterdecays andthereleased arsenic is takenup by otherphases; andpartly due to fixation of arsenic provided from reduced sediments below via porewater fluxes. Reduced Sediments Below the redox zone other processes control arsenic distributions in sediment and porewater. Here, the marine environment is quite different from the freshwater setting. This is due in part to differences in mineralogy, and in part to differences in water chemistry. Partitioning of arsenic among the hydromorphic sediment phases is different below the redox zone in all of the sites. In the lake sediments, total extractable arsenic concentrations show a minimum just below the redox horizon, increase somewhat below this, becoming more or less constant with depth. Arsenic in the WAS fraction decreases slightly but steadily with increasing depth in sediments of the Lake Michigan sites. This can be explained by the slow dissolution of carbonate minerals as burial depth increases, which is consistent with pH profiles from lake Mchigan (Figure 4), that indicate buffering of pH, probably by carbonate mineral dissolution. There is also adeclineintheERarsenicconcenuationinthelakeMidrigansites. Thismay indicate confirming dissolution of manganese oxides with depth of burial. Alternatively, this may result from loss of amorphous iron oxides. There is some evidence that the easily reducible extraction can release metals from some amorphous iron oxides (Tipping et al., 1985) and there is ample evidence that iron oxides, although thermodynamically unstable, can persist well into the 57 reduced zone of sediments (e.g. Canfield, 1989; Wesrin et al., 1991), continuing to release sorbed metals as they slowly dissolve. Arsenicappears to be beingreleasedfromastrongly enriched layerat~40cmdepthinthe Lake Superior Caribou Basin core. This sample was no different in appearance from surrounding sediments (we Table A2-1). All of the extractable phases exhibit enrichment at this depth, and a large peak in porewater arsenic also occurs, with steep gradients above and below. Similar eruichedzoneswithcoincidingporewaterpeaks alsoocatrinthereduced sediments ofthe other lakesites(at~lOcminLMNB-1,at-40cminLWB—2, and at~20¢mdepthinLSIP). This iridicaesflratarsenicirrtliereducedzorreis notirnmobile, butcan betransferred betweendifferent phases and different depths in the sediment. In the Gulf of Maine, the EX and WAS phases become the dominant sequesterers of arsenic below the redox zone. The transfer of a large proportion of the extractable arsenic from the reducible phases which dominate in the oxidized zone to exchangeable sites may happen because the number of sorption sites is reduced via the dissolution of iron and manganese oxides and the decay of reactive organic matter. Arsenic in known to adsorb preferentially onto iron oxides over other substrates (Crecelius et al., 1975; Sadiq, 1990). In the marine environment, where porewater is of substantially higher ionic strength than in freshwater settings, there is more competition for sorption sites (Fdrstner and Wittrnann, 1983). Some experimental work on sorption capacities of SBL sediment from Lake Superior (1D. McKee, pers. comm.) support the ideaMsorpdmsitescanbecmmsatmfled,mddr3sitesmoxidesareprefembbw exchangeable sites. These experiments showed that small amounts of copper added to the sediment were adsorbed by oxides, but that when greater amounts were adM, copper was adsorbed by oxidesuptoacertainlimit,thenappearedintheWAS andEXphases. Theloss of sorption sites may also contribute to the high concentrations of arsenic in porewaters of this site. Authigenic mineral formation appears to influence dissolved arsenic profiles. Porewater arsenic concentrations in the lake sites do not increase continuously with depth, suggesting that concentrations maybelimitedby incorporationintooradsorptionoritoauthigenicphases. In general, below the redox zone, porewater arsenic profiles resemble those of ferrous iron in all of 58 the lake sites, but not in the Gulf of Maine (see Figures 19-23). Similar trends in the relationship between iron and arsenic profiles were formd by Belzile (1988) in sediments from sites of varying salinity in the Laurentian Trough. Peterson and Carpenter (1986), however, found evidence for removal of dissolved arsenic to solid phases in reduwd zones of marine but not lacustrine sites. In the Gulf of Maine site, iron sulfide formation appears to be responsible for removing iron quantitatively from porewaters; this is typical of marine environments where there is excess sulfide to convert reactive iron to FeSz (Bemer, 1980) or to an FeS precursor (Schoonen and Bames, 1991b).,Arsenic concentrations, however, just keep on increasing with burial depth. The firstlargepeakindissolved arsenic coincides withtheironpeak, indicatingrelease ofarsenic from dissolving iron oxides. Arsenic is depleted from the porewaters for a few cm below this horizon, but then begins increasing again whereas iron concentrations remain low. There is a general inverse relationship between profiles of solid-phase arsenic and dissolved arsenic below the iron peak (Frgme 23) indicating some relationship between sediments and porewater arsenic, but it is not clear which phase(s) may be involved. Belzile (1988) found that in marine sediments of the Iamentian Trough, pyrite formation played a significant role in controlling arsenic Concentrations in both sediment and porewater, arsenic was incorporated into growing pyrite crystals. However, profiles of dissolved arsenic in the seaward-most samples show continuous increase with depth, to 35 cm at least, well below the iron peak (Belzile, 1988), much like the Gulf of Maine core. Sadiq (1990) found that AsCIII) sulfides (realgar, AsS) were stable in anoxic marine settings where pH + pe < 4.5, whereas As(V) as R3(As04)2 was stable for pH + pe > 5. He concluded that arsenic sulfide formation should control porewater arsenic concentrations in sulfidic marine sediments by removing arsenite from porewater (Sadiq, 1990); similar conclusions were reached by Moore et al. (1988). Perhaps the formation ofarsenic sulfides occurs deeperinthe sediments than was sampledinthiscore, where more strongly reducing conditions develop and pH 4» pe can drop below 4.5. Since there appears to be sufficient sulfide available to remove iron from porewaters, and iron is available in much higlra'cawmu'afimmanusenieitseansmfikelythmwlfidecmcennafims arelirnitingthe formation of arsenic sulfides. 59 In Gulf of Maine sediments, arsenic in the oxidizable fraction (where arsenic contributed from sulfides should appear) is essentially constant below the redox zone (see Figure 11). The base of the dissolved iron peak (just below 20 cm depth) indicates where maximum pyrite formation is expected; it appears to have little effect on the profiles of OX arsenic or dissolved arsenic (Figure 23). It is possible that uptake of arsenic by pyrite is balanced by loss of arsenic from organic phases,sinoe the OX extraction procedure does not distinguish between organics and sulfides. EARLY DIAGENESIS OF MERCURY Data from this study reveal that diagenetic processes do affect mercury in sediments; in fact, the diagenetic recycling of mercury is more effective than is that of arsenic. The diagenetic behavior of mercury is different from that of arsenic in a number of respects. The porewater profilesofthetwornetalsarenotatallalike,indicatingthatdifferentprocessescontrolthe distributions of these two dissolved components. Changes in the partitioning of mercury among the solid phases are also different from that of arsenic. Oxidized Zone Above the redox zone, the base soluble (BS) and acid soluble (AS) fractions contain substantial amounts of mercury, as does the oxidizable (OX) fraction; below this zone nearly all ofthemercury is associatedwiththeOXfraction (Figures 14-16). 'Ihissuggeststhatmercuryis very efficiently removed from the AS and BS fractions by processes operating in the oxidized upper layer, and by the onset of reducing conditions. The redox cycling of iron exerts considerable influence on the behavior of mercury. 'lhe nearly complete loss of mercury from iron and manganese oxides (AS fraction, Figure 15) noted for all sites was also observed by Strunk (1991) and is different from the behavior shown by arsenic or by other metals (e.g. McKee et al., 1989a). Forbes et aL (1974) formd that mercury bonds on goethitesurfaceswerelessstablethanthoseofothermetalsadsorbedtooxides. Asiron oxides begin 'to dissolve under reducing conditions, the tenuously bound mercury may be readily 60 released This may account for mercury being removed almost completely from the AS fraction during early diagenesis. The degradation of organic matter also affects mercury. The extensive loss of mercury from theBS fraction(humiclfulvic acids; see Figure 14) may be due to the factthatthese compounds (or their bonds to mercury) are readily broken down during early diagenesis. Readily decomposableorganicrnatterhas beenfoundto eonstituteupto45% oforganicmatterdeposited in Lake Superior sediments (Klump et al., 1989). This loss of BS mercury is in contrast to the, results of Stnmk (1991) who found that the BS fraction contribuwd significant amounts of mercury to deeper sediments in several locations of Lake Superior. It is unlikely that the methods used are responsible for this discrepancy, because identical extraction procedures were employed. Inonecore,fromthelle Parisienne area, mercury was foundtobeessentially absentfromtheBS fraction below the redox zone (Strunk, 1991); perhaps this is a feature that varies spatially as a result of contrasts in organic matter inputs. Small-scale spatial variations in the nature of organic matter accumulating in depositional basins have been observed (Silverberg et al., 1985; Klump et al., 1989) and related to differences in diagenesis of metals (Iricanin et al., 1985; Gobeil et al., 1987). This is discussed in more detail in the section on diagenetic variability below. Mercury in the OX fraction of Lake Michigan samples increases with depth in the oxidizing zone, reaching a maximum at the depth corresponding to peak iron dissolution, then drops quickly to a low values before showing a secondary peak within the reduced sediment layer. The pattem of enrichment in the OX fraction occurring directly below the enrichment in the AS and BS fiactions suggests that mercury released by decay of humiclfulvic acids and by reduction of ital/manganese oxides is taken up by some component ofthe oxidizable fraction. The depth of dre upper enriched layer corresponds to the bottom of the zone ofrapid organic matter degradation shown by the organiccarbmpmfilesindrelakeMichiganmdfleParisiarnesitesCFigmes 19- 21). This firdicatesfiflasflemomreacfiveorgmkmwisdegr'aMmawrymkasedfimnit is accumulated by some other component of the oxidizable fraction. Below this zone, mercury in he OX fraction drops to a minimum, and organic carbon becomes more constant. 61 Wrthin the zone of rapid organic matter decomposition, porewater concentrations fluctuate, but in general show an initial increase below the sediment-water interface, followed by a decrease in concentration. Thisisconsistentwiththereleaseofmercm'yfromtheAS andBSphases and subsequent uptake by the OX phase within this zone. A plot of organic carbon x5. mercury content of sediments (Figure 24b) shows a conelation between organic carbon and total extractable mercury at high levels of organic carbon (particularly for the Lake Michigan sites). Thisreflectstheretention ofrnercuryintheupperlayers where organiccarbmcontentisgenerally highest A similar pattern was observed for cadmium in the Laurentian Trough (Gobeil et al., 1987); they attributed the loss of cadmium to aerobic oxidation of organic matter and estimated that 80% of the total cadmium flux to the sediments was returned to the water column via upward diffusion. Porewater profiles of suggest that fluxes of mercury out of the sediment may also be occurring; this is discussed in the section on fluxes below. Mercury has been found to form complexes with dissolved, colloidal, and particulate organic carbon (Falchuk et al., 1977; Cline et al., 1973; Mantoura et al., 1978). Hallberg (1982) found experimental evidence that chelating agents are concentrated in upper sediment layers above the redox layer, and suggested that they may react with heavy metals there, sweeping them out of the system before they have time to be fixed as sulfides. This would tend to keep mercury in the upper oxidized portions of the sediment, and/or to return it to the water cohrmn. Iindberg and Har'riss (l 974) also found a significant correlation between dissolved organic carbon and dissolved mercury in porewaters of estuarine sediments, and that this association decreased with increasing depth in the sediments. This association can explain the very efficient retention of mercury in the upper layers of sediment: mercury released by the acay of solid-phase labile organic matter is complexed by dissolved organic carbon. Such complexes may then be transferred to solid phases by flocculation (Cline et al., 197 3), by coagulation or aggregation of colloids (Morel and Gschwend, 1987), or by scavenging of colloids onto sediments (Santschi et al., 1 987). Similar behavior has been observed for iodine by Kennedy and Elderfield (1987). They found that the association of iodine with organic matter was responsible for retaining iodine near 62 the sediment-water interface of pelagic marine sediments. Iodine released from decomposing organic matter was rapidly removed onto reactive organic matter at the sediment surface (Figure 25). Retention of iodine in the sediments was found to depend on the presence of newly deposited reactive organic matter (Kennedy and Elderfield, 1987). It is possible that the same mechanism affects mercury cycling in sediments: variations in the reactive organic matter content of surficial sediments could influence mercury enrichments at the sediment-water interface. This is one possible explanation for the lower enrichment of the surface sediment of the Gulf of Maine. Organic carbon content is quite low, suggesting there may be less reactive organic nutter at this site than at the Great Lakes sites. Anorher possible explanation is the higher salinity of the marine environment. Lindberg and Harriss (1974) found that higher salinity resulted in , lower mercury-complexing capacity of dissolved organic matter in porewaters. This could also contribute to the lesser degree of enrichment of surficial sediments in the Gulf of Maine. Reduced Sediments Below the redox zone, mercury appears to be influenced by sulfide mineral formation. The smallersecmdarypeaksinOXmercury occurringinthereducedzoneofallthelakesitesmaybe due to uptake ofmercury into some sulfide phase. In the GulfofMaine the formation of sulfides in reduced sediments results in rmiformly low concentrations of dissolved mercury, but in the Great Lakes this leads to somewhat more complicated behavior. Although there is not much sulfate in lake waters, and sulfate reduction is considered to be a minor contributor to organic matter decomposition (Carlton et al., 1989), sulfate reduction does occur (T‘isue et al., 1988) and authigenic sulfide minerals have been identified in modern sediments of the Great Lakes (Dell, 1 972; Sly and Thomas, 1974). Several studies have found that sulfate reduction can be important in oligotrophic lake sediments with low organic matter input (Capone and Kiene, 1988, and references cited therein). Evidence fortheinfluenceofsulfides onmercury inthelakes is provided by mercury porewater profiles, as discussed below. First, porewater mercury profiles in the lakes (Figure 18) show rninima at depths which generally correspond to secondary maxima in oxidizable mercury (Figure 16; see also Figures 19- 22). This suggests that mercury released from dissolving iron and manganese oxides which 63 lodne enrichment at seawater-sediment interface increasing with time Sediment Water E Burial Dapth Figure 25. Cartoon of iodine cycling at the sediment-water interface (from Kennedy and Elderfield, 1987). 64 diffuses downward is taken up by some component of the oxidizable fraction. This occurs below the depth of rapid organic matter oxidation, where organic carbon levels are relatively low (see figrne 6). The other principle phase affected by the oxidizable extraction is sulfides. Sulfate reduction typically occurs in sedimarts below the zone of iron reduction (Bemer, 1976; 1980). In marine environments where sulfate reduction is the major process of organic carbon oxidation, iron tends to be removed from porewater by precipitation of pyrite (Capone and Kiare, 1988). Inthe lakes there is often more iron than sulfide, so formation of iron sulfide minerals will effectively remove sulfide, allowing excess Fe(II) to accumulate in the porewater. Thisexplainsthecontinuous increaseinferrousironwith depthinthelakesediments,incontrast to the narrow zone of ferrous iron in the Gulf of Maine core. In lakes, sulfate reduction is generally completed within a few cm of the sediment-water interface, and once sulfate is depleted the remaining sulfide is precipitated as highly insoluble FeS minerals (Bemer, 1980; 1985). In the zone of sulfate reduction and sulfide generation, mercury may be incorporated into HgS, or may be adsorbed onto FeS minerals (Hyland et al., 1990). Kuivila and Murray (1984) found that the depth where sulfate concentrations in lake sediments reached a backgrormd level (i.e. where sulfate reduction was essentially completed) corresponded tothedepth where achangein slope ofthe alkalinityprofile occurred,toaless rapidrateof increase in alkalinity. Examination of alkalinity profiles from the Great Lakes (Figure 5) shows such trends in alkalinity,withthechangeinslope occurringat~9cminLMNB-l,~6cmin LMNB-Z, ~13 cm in LSIP and ~46 cm in LSCB. These depths are all several cm below the apparent zone of iron reduction (see Figures 19-21), and may indicate the base of the zone of sulfate reduction. These depths correspond to depths where mercury in porewater starts dropping tominimum values, ie. tlretopofthernercury minimumzone,inLakeMichiganandLake Superior He Parisienne. This is consistent with the hypothesis that mercury is removed from porewater by the formation ofsulfideminerals,atleastinthesetlueesites. Belowthiszone,no sulfide forms, so dissolved mercury will not be removed from the porewater by this process. In the Caribou Basin, there is no distinct minimum in porewater mercury; it is possible that sulfide fomrationis limitedinthislocation. 65 Second, dissolved mercury profiles in the lake sites display a general inverse relationship with ferrous iron profiles below the redox zone (Figure 26). Mercury in porewater tends to increase initially below the sediment-water interface, then concentrations decrease in the zone where ferrous iron concentrations increase (Figures 1922). Then the iron concentrations drop off somewhat, and mercury concentrations peak. Below this, mercury concentrations decrease and irmconcentrations increase orrceagain. This inverserelatiorrshipis essentiallytlreopposite of that between ferrous iron and arsenic, and suggests that score type of competition between iron and nrercrrry for sulfide may exist. Thefonnatiorr ofpyrite requires moretlranjustthepresence ofreduwd iron andsulfideiorrs. Sclroonerr and Barnes (1991a,b) have found that the nucleation of pyrite is inhibimd under typical conditions of early diagenesis, and FeSZ forrrrs only after conversion involving several steps, from FeS precursors through Fe283 to FeS; . Morse and Comwell (1987) found that identifiable iron sulfides in anoxic marine sedirrrents were almost always pyrite; they suggest that if precursors are present they must be as coatings or as subnricron particles. Other studies suggest that iron monosulfides form first when the pH is near neutral, but pyrite fomrs first at pH values below 6.5 (Drever, 1988). Values of pH approach 6.5 in some samples of LSIP, but all other areas have pH > 6.5 in the reduced sediment (Figure 4). One explanation for this behavior is that HgS fomrs in the shallower depths where sulfate reduction first occurs. HgS is more insoluble than the various FeS minerals (based on values of solubility products; see Table 2). Therefore, as mercury sulfide precipitates, virtually all of the mercury supplied to the porewater in this zone may be removed. Because mercury is present at trace levels only, the formation of HgS does not remove all of the sulfide. Deeper in the sediments FeS is converted to Fe283, which is much more insoluble than HgS, so iron is removed from porewater (to some extent), sulfide is used up, and any released mercury could appear dissolved in porewater once again. The solubility of FeS; is sorrrewhat lower than that of HgS, but much closer in magnitude than either FeS or FeZS3. Extending this hypothesis, solubilities of other trace metal sulfides could be used to predict trends in porewater profiles. Lead sulfide solubility is close to those of iron monosulfides, so 200 f Lake Michigan 150 . North Basin ' -- -1 n -2 100 - 50 r 0 0 1 2 3 300 e Lake Superior Ile Parisienne 200 ' MercurytngfL) 100- 800 | Lake Superior I 600 - Caribou Basin 400 . 200 - 0 0.25 0.5 0.75 Ferrous Iron (man) Figure 26. Relationship between ferrous iron and dissolved mercury in porewater of reduced sediments, Great Lakes samples. 67 Table 2 Solubility product constants of some sulfide minerals at 25'C (from Faure, 1991) a-HgS cinnabar 53.0 CuZS chalcocite 48.5 Hg2S 54.8 PbS galena 17.5 FessS pyrrhotite 17.4 CdS greenocltite 27.0 I O]. FeS 16.2 a - ZnS sphalerite 24.7 Fe233 greigite 88.0 B - ZnS wurtzite 22.5 FeS; pyrite 42.5 a - NiS 19.4 FeS; marcasite 41.8 7 - NiS 26.6 lead would not form instead of sulfides of the more abundant iron, and lead profiles would more closely resemble those of Fe(II). Solubilities of cadmium, zinc, and nickel sulfides are slightly lower than those of FeS minerals, with a - NiS being closest to FeS. These metals should be affected by sulfide in the same manner as mercury (i.e. inversely related to Fe). Mercury sulfide is more insoluble than any of these sulfides, so should be able to precipitate even though mercury concentrations in porewater may be much lower than those of other metals. Lead in porewaters from the Iamentian Trough has been found to correspond to ferrous iron profiles (Gobeil and Silverberg, 1989), whereas cadmium shows an inverse relationship with iron: concentrations are high near the sediment surface, decrease to undetectable values below the redox zone, then increase again deep in the core (Figure 27; Gobeil et al., 1987; Gobeil and Silverberg, 1989). These data support the hypothesis that metal-sulfide formation affects trace metal concentrations in porewater . Anahemafiveexphnafimisflratadsorpfimmwuonsulfidemmemlsisconuoningdte concentrations of mercury. Sulfide minerals are excellent scavengers of divalent cations of mercury, lead, zinc, and cadmium (lean and Bancroft, 1986). Mercury adsorption onto FeS minerals may be controlling dissolved mercury in marine porewaters as well (Hyland et aL, -' DEPTH (out) cu tum 120 o 0.4 one 1.2 STATION 22 STATION 23 8TATION 24 to" too ' zoo ' zoo oMn Us”) Figure 27. Dissolved Pb, Fe, Mn, and Cd in rewaterfrom the lamentian Trough (from Gobeil and Silverberg, 198%; 69 1990). It is more difficult to resolve the inverse relationship between dissolved iron and dissolved mercury, however, if this is what is controlling porewater mercury concentrations. An alternative explanation for the mercury porewater profiles is that some volatile mercury species fomrs at this zone in the sediments. It is possible that mercury is reduwd to elemental merctuy, which is volatile and may escape from the sediments. It is also possible that dimethyl mercury could be fornwd, perhaps by microbial processes. FLUXES TO THE SEDIMENT-WATER INTERFACE As described above, there is evidence suggesting diffusive fluxes of arsenic and mercury may be occurring from porewaters into overlying waters in many of the sites investigated. Porewater gradients in arsenic at all sites but the Caribou Basin of Lake Superior, and mercury gradients in the lake sites indicate that fluxes out of the sediment are possible. Arsenic concentraticms in Lake Superior water have been found to be highest in deep waters; this has been attributed to arsenic regeneration from bottom sediments (Rossmann, 1986). Total mercury in epilirnnetic waters of Lake Superior has been reported to average 44 ng/L in the eastem portion of the lake (Rossmann, 1986). These values are much lower than uppermost porewater samples for Lake Superior, indicating that a flux of mercury from sediments might be possible. Rates of organic matter degradation have been found to increase with increasing sediment deposition rates (Johnson et al., 1982). This results in increased fluxes of nutrients to the water column (Johnson et al., 1982) and should result in more rapid cycling of diagenetically-cycled elements. Rates of bioturbation have also been related to rates of organic matter decomposition and general diagenetic recycling (Gratton et aL, 1990). Deeper water should allow more decomposition of organic matter before it reaches the sediments, reducing rates of degradation in the sediments (Klurnp et aL, 1989). Thus greatest fluxes of mercury and arsenic from the sediments would be expected in shallower waters where sedimentation rates, organic matter accunmlationrates,andbioturbationratesarethegreatest. In the following sections, the diffusive fluxes and sedimentation fluxes of arsenic and some of the consequences of these fluxes are estimated. These calculations are not performed for mercury 70 because of the manner in which mercury is recycled during early diagenesis. Since most of the mercury is released from sediments very near the sediment-water interface, and appears to be removed again almost immediately, fluxes calculated from porewater concentration gradients would not be meaningful. Difl‘usive Fluxes Vertical diffusive fluxes can be estimated using Fick's first law for one dimension: J = {a - Ds- (BC/32) wherel=thediffusion flux,¢=theporosity,D,isthediffusioncoefficient, and (BC/az)isthe concentration gradient (Bemer, 1980). To calculate the flux for arsenic, several assumptions must be made: 1) viscosity and charge coupling effects are negligible; 2) arsenate and arsenite anions are the only arsenic species present, and they have identical diffusional properties; 3) there is no solid-phase consumption of dissolved arsenic near the sediment-water interface; and 4) arsenicconcentrationgradientsarelinearsothataC/BzisequaltoAAs/Az(Peoersonand Carpenter, 1986). D8 is estimated from the diffusion coefficient for the arsenate anion at infinite dilution, Do, estimated for 4°C from the data of Li and Gregory (1974) by assuming a linear change with temperature between O'C and 25’C (Peterson and Carpenter, 1986), and using the relationship D8 = Do- 02 (Lerman, 1977) to approximate the effects of sediment tortuosity. Calculations of diffusive fluxes (Table 3) show as the flux of arsenic from the sediments at site 1 ofLakeMichiganismoredtantwicethatatsiteZ. The organiccarboncontentis approxirnatelythesarneinsurfacesedirnents ofthesesites,thesedirnentationratesarereportedly similar (Christensen and Chien, 1981), and bioturbation (based on porewater arsenic profiles) appearstobegreateratsiteZthanatsite 1; thustherelativevaluesofdiffusivefluxarethe opposite of what might be expecmd. This phenomenon could be explained if the concentration gradientatsiteZhasbeenreducedbymixingofporewaterswithmoredilutelakewaterdueto biotm‘bation and bioirrigation. Sedimentation rates atlle Parisienne arenearly fourtirnes those for Lake Michigan (see Table 4), which could account for the highest flux value in this location. Although organic carbon content is low (probably due to dilution by terrestrial inorganic sediments), the sedimentation rate and bioturbation rates are high enough that much sediment- 71 Table 3 Parameters Used for Calculation of Diffusive Fluxes of Arsenic Site o D, A: AAs J LMNB-l 0.93 134 4.0 7.5 x 103 0.234 LMNB-z 0.93 134 3.5 2.5 x 10-3 0.089 LSIP 0.93 134 2.5 5.1 x 10-3 0.254 GMMB 0.93 134 4.5 3.6 x 10-3 0.100 fl = porosity; D, = Do - (62, and D0 = 155 cmzlyr at 4'C (estimated from data of Li and Gregory, 1974); A2 = depth (cm) to first concentration peak; AAs = difference in dissolved arsenic concentration (pg/m3) from the sediment surface to depth Az; and J = diffusive flux (ttg/cm2 - yr). bound arsenicisreleasedwithintheupperfewcmofthe sedimentcolumn(Az isnearestthe sediment surface at this site), and much is able to diffuse up toward the sediment surface. These calculated fluxes represent estimates only, since several of the assumptions are not strictly true. Although arsenate is generally predominant in oxidized waters, arsenite can also be present, as can methylated arsenic species (Crecelius, 1975; Andreae, 1979; Huang et al., 1982; Peterson and Carpenter, 1986; Brannon and Patrick 1987). There also appears to be significant incorporation of dissolved arsenic into solid phases in the near-surface sediments. Concentration gradients are probably not linear, but may appear so due to the 1 cm sampling interval. More closely spud samples could reveal steeper concentration gradients. These calculations do not take into account the effects of bioturbation; however, Sweerts et a1. (1991) found that the relationship between D, and Do did not change much with porosity and that effects of biomrbation on predictability of D, were only significant in sediments with very high invertebrate populations. Several studies have formd that measmed fluxes of dissolved metals out of sediments do not agree with fluxes calculated from porewater profiles (e.g. Westerlund et al., 1986; Sundby et al., 1986; 72 Berelson et al., 1990). Fluxes of dissolved metals that are redox sensitive are strongly dependent on the flux of oxygen into the sediments across the sediment-water interface rather than on pore- water gradients alone (Sundby et al., 1986). Nonetheless, differences between fluxes calculated for the different sites may provide some useful information, and differences between upward difl'usive fluxes of dissolved species and downward fluxes of sediment-bound metals can provide estimates of the proportion that is recycled from the sediment column. This is described below. Sedimentation Flwces Fluxesofarsenicanivingatdresedimentsurfacecanbeestimatedasdreproductofthemass sedimentation rate and the arsenic concentration of the SBL, which represents freshly deposited material (McKee et al., 1989b): F, = R . [As], where F, is the sedimentation flux, R is the mass sedimentation rate, and [As]0 is the total extractable arsenic concentration in the SBL. Herrnanson and Christensen (1991) have determined mass sedimentation rates for northern Lake Michigan as 1.33 x 10'2 g/cmz- yr for LMNB-l (their site NLM-E) and 1.37 x 10'2 g/cm2 . yr for LMNB-Z (their site NLM-B). Sedimentation rates for northern Lake Michigan have also been reported as 8.8 x 10'3 g/cmz- yr (Christensen and Chien, 1981). Rates have been estimated for Lake Superior Ile Parisienne Basin as 7.0 x 10-2 g/cm2- yr (Kemp et al., 1978) and 6.5 x 102 g/cm2 - yr (Krezoski, 1989). Sedimentation rates are not available for the Gulf of Maine; 210% studies have been unsuccessful due to extensive biotin-bation in this area (Brower, 1984 cited in Hines et al., 1991). Calculated sedimentation fluxes are shown in Table 4. Downward fluxes are similar for the twoLakeMichigansites. Thesevalues arehigherthanthe sedirnentationflux forarsenicin northern Lake Michigan (0.088 ttg/cm2 - yr) calculated by Christensen and Chien (1981). If their sedimentation rate (8.8 mg/crn2 - yr) is used for these calculations, sedimentation fluxes of arsenic areclosertotheirresult. AmuchhigherfluxisformdforthelleParisiermesite; thisislargelya frmctionofthesedimentationrate,asarsenicconcentrationsintheSBLarefairlysimilarforall three sites. 73 MIL! Parameters Used for Calculation of Sedimentation Fluxes of Arsenic Site R [As], F, J J/F, LMNB-l 1.33 x 10’2 10.2 0.136 0.234 1.72 8.8 x 10'3 10.2 0.090 0.234 2.60 LMNB-2 1.37 x 10'2 11.8 0.162 0.089 0.55 8.8 x 10'3 11.8 0.104 0.089 0.86 LSIP 7.0 x 10'2 11.3 0.791 0.254 0.32 6.5 x 10’2 11.3 0.735 0.254 0.35 R = sedimentation rate in g/cmz- yr; [As], = total extractable arsenic in SBL sample (jig/g); F, = sedimentation flux in pig/cm2 - yr, J is the diffusive flux from Table 3. J/F, represents proportion recycled. The proportion of sedimented arsenic that is recycled by early diagenetic processes can be estimated as the ratio of the upward diffusive flux, J, to the downward sedimentation flux, F, (Table 4). These values appear to be quite high for Lake Michigan: 55 - 86 96 for site 2, and over 100 % for site 1. This last value suggests that sedimentation fluxes are underestimated, or diffusion fluxes are over-estimated. At this site, the SBL contained less than half the total extractable arsenic of the uppermost core sample. It is possible that this SBL sample is not representative of typical arsenic sedimentation fluxes at this site. If the concentration from the top of the sediment core, 22.8 ug/g, is used in the calculations, F, becomes 0.201 ug/cmZ - yr, and JR", becomes 1.16 for the lower sedimentation rate of Christensen and Chen (1981), still indicating greater than 100 96 recycling, suggesting some error in the calculations. For the higher sedimentation rate of Herrnanson and Christensen (1991), F, becomes 0.304 rig/cm2 - yr and J/F, becomes 0.77. This represents a more reasonable number, but it is probable that the arsenic concentration in the core top sample does not represent freshly deposited material. This concentration more likely results from enrichment by adsorption of upward diffusing arsenic, as discussed below. 74 In all three ofthese sites, the concentration ofarsenic in the SBL and the concentration in porewaters of reduced sediments are similar (~11 nglg and ~8 [lg/L, respectively), while the "background" concentration of arsenic in the sediment is higher in the Lake Michigan sites (~4 nglg) than in the Ile Parisienne site (<2 [lg/g). These facts suggest that recycling is more effective inthelleParisienne site,aswmfldbeexpectedfromdtehighersedimentafimrateandshanower water depth. Yet the calculations show the least proportion of arsenic recycled at this site. There are several potential sources of error in these calculations. Uncertainties associated with diffusive fluxes are described above. Sedimentation rates are commonly determined from 210Pb dates, and this method was used by Christensen and Chien (1981) and by Herrnanson and Christensen (1991). Lead is known to be mobilized during early diagenesis (McKee et al., 1989a,b) and recently, 21"Pb was found to be redistributed in lake sediments (Benoit and Hemond, 1991). The sedimentation rate for Ile Parisienne calculated by Kemp et al. (1978) was determined from the total mass of sediment deposited above the Ambrosia horizon (dated at 1890 in this region). This rate would be averaged over the entire time interval, but it is similar to the rate reported by Krezoski (1989), based on 210Pb data. Despite the many uncertainties, these calculations suggest that significant proportions of arsenic reaching the sediments can be recycbd by diagenetic processes. Role of Fluxes in Enrichment of Surface Sediments The influence of upward diffusive fluxes on enrichment of sediments near the sediment-water interface can be estimated by calculating the amount of arsenic contributed by the diffusive flux to the sediment arriving at the sediment-water interface. The only suitable site for this calculation is lake Michigan North Basin site 1. This site has an upward diffusive flux at the sediment-water interfaceand aconcentrationofarsenicindte SBLlowerthanthatintheuppermostsediment column. ThisaflowsdreSBLtobeusedasanesdmateofineonfingarsaficcmuentanddre uppermost sediment core sample to be used as the enriched layer resulting from adsorption of diirusing arsenic. The sedimentation rate is 8.8 mg/cm2 - yr, and the concentration of arsenic in the SBL sediment is 1.02 ug/g. Therefore, for one cm2 of lake bottom for one year, 8.8 mg of sediment accumulates, which contains a total of 0.090 [lg of arsenic. To this would be added 75 0.234 ug of arsenic from the diffusive flux over that square cm for 1 year. If all of this arsenic were sorbed by the 8.8 mg of sediment, the total mass of arsenic would become 0.324 ug, resulting in an arsenic concentration of 36.8 [lg/g. This is higher than the observed concentration of 22.8 [lg/g in the uppermost core sample, but indicates that enrichment of sediment by diffusive fluxes is certainly possible. The recycling of arsenic from the sediments, and adsorption of upward-diffusing arsenic by sediments near the sediment-water interface would tend to retain arsenic near the sediment surface. IMPLICATIONS FOR BIOAVAILABILIT'Y Oneirnportantconsequence oftheretention ofdeposited mercury and arsenicnearthe wdiment surface is that these elements remain available for uptake by benthic organisms for longer periods of time than if they were buried and removed from the sediment-water interface. The activities of benthic organisms themselves have been found to promote recycling near the sediment surface (e.g. Cross et aL, 1975; Aller, 1978), enhancing potential bioavailability. Recent newspaper reports (e.g. Lange, 1991) cite the widespread nature of mercury contamination in North American lakes and their fish populations. Mercury finds its way into the food chain primarily as methyl mercury (Stokes and Wren, 1987). Methyl mercury is formd in sediments by bacterial action (Wood, 1974), and the rates of production of methyl mercury have beenfoundtodependontherateofsupply ofdissolvedmercurytothemicrobes(Mikacetal., 1985; Olson and Cooper, 1974). This supply will depend on early diagenetic processes releasing mermrynearthesedimem-waterinterface. Thefactthatmostofdremercuryreachingthebottom sediments isretained nearthe surfaceenhances thechances formethylation and subsequententry into the food chain. Gill and Fitzgerald (1988) find evidence thatthe scavenging ofmeruiry by settling particles in the ocean is so effective that any mercury regenerated (for example, by organic matter decay in the benthic nepheloid layer) is quickly removed from solution, maintaining relatively low levels of dissolved merctn'y in the deep ocean. This would also tend to keep mercury near the sediment-water interface, enhancing its potential bioavailability to benthic organisms. These phenomena suggest that the problem of mercury contamination could persist 76 foralongtirne; aslongasevensmallarnounts ofmercury aredischargedtolakes(directly orvia atmospheric transport), relatively high concentrations of mercury will remain available to benthic organisms. Analysis of Diagenetic Variability Differences in diagenetic parameters were observed on four scales: differences between the lakes and the ocean, differences between the two lakes, differences between depositional basins within a lake, and differences between sample sites within a single depositional basin. Observed differences range from distinct to subtle, and there were also numerous similarities among these sites. Factors contributing to diagenetic variability include differences in water depth, differences ill rates of bioturbation, variations in the supply of reactive organic carbon, and variations in overall sedimentation rate, as well as differences in water chemistry and sediment mineralogy; these are discussed below. Initial evidence for the variability among diagenetic environments was provided by porewater alkalinity profiles from the Great Lakes sites (Figure 5). Alkalinity of interstitial waters is largely controlled by early diagenetic reactions (Ben-Yaakov, 1973; Seuss, 1979, Kuivila and Murray, 1984; Anderson et aL, 1986), so differences in alkalinity profiles should reflect differences in diagenetic processes. Although the relative contributions of individual diagenetic reactions to changes in the alkalinity of sediment interstitial waters in lacustrine environments are variable, changes in alkalinity and pH are generally attributable to the decomposition of organic matter, and indicatedreextentofearlydiagenesisinlakesediments. Additionalevidencefordiagenetic conditions atthedifferentsites isprovided bydataforferlous ironinporewaters andby the organic carbon content of the sediments. Diagenetic variability has been observed in the Laurentian Trough by Gobeil et a1. (1987). Theyfounddifferencesinironandmanganeseprofiles betweencloselyspacedcorescollectedat one site; thesedifferences wereregarded asrelative stretchingorcompression ofthe profiles. Thiswasattributedtovariationsinthedepthdistribution ofdiageneticreactionscausedby differences in rates oforganic matter input, oxygen consumption, and distribution ofbenthic 77 organisms (Gobeil et al., 1987). The variability displayed among the five sites of this study is wider, but can be largely explained by these same variables. Differences ill diagenetic processes result in variations in the behavior of mercury and arsenic undergoing early diagenesis. Variations insitecharacteristies andmercuryandarsenicconcentrationsaresununarizedinTableS. GREAT LAKES VS. GULF OF MAINE Althoughthedirnensions andenergy inputs oflakesandoceans ueverydifierencprowsses controlling biogeochemical cycles of elements are similar, so the two systems can be compared (Santschi, 1988). The marine environment is chemically quite different from the freshwater environment. Particularly important differences in terms of early diagenesis are: the presence of higher concentrations of sulfate in seawater, the higher pH and alkalinity of seawater, and the higher ionic strength of seawater. Higher concentrations of sulfate allow sulfate reduction to play a more important role in early diagenesis. In the Wilkinson Basin, which adjoins the Murray Basin in the Gulf of Maine (see Figme 2), sulfate reduction has been found to be the dominant biogeochemical process at depths below ~11 cminshortsedimentcores (Hines etaL, 1991). The GulfofMainecouldbe characterized as a sulfidic envirorunent, whereas the Great Lakes are non-sulfidic according to the classification of Bemer (1981). The more active sulfate reduction influences porewater profiles, particularly for mercury, and helps to retain elements in the sediments. Dissolved arsenic and mercury concentrations in porewaters are governed by different phenomena in the two settings. Below the redox zone, in the lakes, the profiles of porewater arsenic and ferrous iron are similar; the porewaterprofile ofarsenic in thegulfis not as closely related to ferrous iron. Similar results were observed by Belzile (1988) in the Laurentian Trough, an estuarine setting: profiles of arsenic and iron in porewater from the seaward-most site resemble the profiles from the Gulf of Maine, while profiles from the most freshwater-influenced site show moresimilaritytotheGreatlakessites. Thisisduetotheactiveremoval ofironfrornporewaters inreducedsedimentoftheGulfofMaine,mostlikelyduetorheformationofironsulfide minerals. Although sulfidesdo appeartobeformingintheGreatlakes sites,there is 78 .3. 035. EB... 623.8 we .3: Eggs—=68 \ 5.: ozmake 239v ans. 0 3.262 m< R. .Awu onE 003 383.2888 93 e5 52: 5033 one u 190$ 288599. n.m.m 25568 3262 E 3B5 980:. 3 3%.: 3:88 0330828 .89 we cease—.088 omflofi u 95 do. 28 .5 4mm 5 8B5 babe .5 333 028.8 035856 .82 we 53.5888 E2552: u .82 45869. 5 .598 35mg 2088 2303 u no m. can so: Scuba 9 Es soon u can seem .38: a s 9:3. 52. 8:85 9.3 as .28: 52.0350 23 canes—.5: Spa cog—22 3.3 new ASNEQwEV 8:: Eugene—60w .38.: 5 59v .683 Em C no l in he e; an; i n _ : 3N EEO wfi o— no i 3. N._ of and com 3 Wm cmm mow..— hdg a we— mm 2 ~.— ~— 3..— end 5 E. «Q Em..— vd N— m: mm m.m Wm ~_ 5 .m and 2 :3 new «$233 m.w 3 a: N: no 9m mm “Nd mmfi m m.m_ awn TmZE... mm 95 a: canoe .mm mom a: 8288 ram ease as. 589 mg m: m: m: 2* 3. 2 2 cos cos Seem can as? £85qu beam 353::— §F 8558820 new Co 9885 33H 79 much more iron than sulfide, so iron concentrations are controlled by other mineral equilibria. For mercury, concentrations are very low everywhere in the Gulf of Maine porewaters, due to precipitation of sulfides. In the Lakes, dissolved mercury concentrations are higher in reduced sediments, again due to the much lower sulfide content. The higher pH, alkalinity, and ionic strength influence the partitioning ofarsenic among the hydromorphic phases of the sediment. Ill the lakes, the easily reducible, moderately reducible, and oxidizable fractions are the major sequesterers of arsenic; in the Gulf the exchangeable and weak-acid soluble phases contain most of the arsenic. The presence of carbonate minerals in the Gulfof Maine sediments, which persist in deep sediments due to the higher pH and alkalinity of these waters, may contribute to the greater amount of arsenic in the WAS fraction. Below the redoxzone,theEXfractionoftheGulfsedimentscontainsthemostextractable arsenic; inthe lakes, any arsenic in the EX fraction disappears in the reduced zone. This sorption of arsenic onto exchangeable sites in the marine setting is probably related to the ”samration" of other sorption sites by more abundant elements, as discussed in the section on diagenesis of arsenic, above. The higher ionic strength of the marine envirorunent appears to affect both the partitioning of arsenic, and the total amount of arsenic that is sorbed to sediments. The higher arsenic porewater concentrations, and the lower ratio of maximum to background concentrations in sediments (Table 5) suggest that arsenic released by early diagenetic processes is not as readily taken up by other phasesintheGulfofMaineasitisintheGreatlakes. Mercury partitioning among sediment phases is similar in both environments, and does not seem to be influenced by the same factors as arsenic. The biggest difference observed in sediment-bound mercury isthatthemaxirrmmconcentrations atthesediment—waterinterfaceare higher in the lakes, although the average concentration in reduwd sediments is highest in the Gulf ofMaine (Tab1e5). Thisresultsinaratioofmaximmntobackglomldmercuryconcenuations that islowestintheGulfofMaine. Thesamecontrastin”enrichmentfactors" (ratio ofmaxirnumto background concentrations) is observed for arsenic; this is shown graphically in Figure 28. There are several possible explanations for greater "emichment factors” (Figure 28) observed in the freshwater sites than in the marine setting. As discussed earlier, concentrations of arsenic .. - 80 10‘ Arsenic Enrichment Factor 101093109011pr MOOJOW LMNB-1 LMNB-2 LSIP SITE Figure 28. Enrichment factors for arsenic and mercury (from Table 5). 81 and mercury near the sediment-water interface are higher than ill more deeply buried sediments due to several processes: higher concentrations in freshly deposited material, recycling within the upper oxidized sediment layers, and adsorption of dissolved arsenic and mercury diffusing up from deeper in the sediment column. One possible explanation is that the Murray Basin is not a site of active sediment focusing; cmanuadmofamedcmdnmnymelowumesedinemsmfacebwmmefieshnmtaialwhh higher concentrations is not accumulating. The absence of a distinctive SBL layer at this site supports this erosion/non-deposition idea. Organic carbon concentrations are very low at the top of the core, also suggesting this is possible. It is conceivable that erosive action has removed SBL and upper layer of sediment, with its more reactive organic matter. The basins of the Gulf of Maine are known to be accumulating sediment, however (Spinrad, 1986), suggesting this is not the best explanation. The profiles of mercury in sediment also indicate that sediments are actively accumulatinghere. Thereisanenriched layeratthesediment surface,withhighmercury concentrations in the base soluble and acid soluble phases, and in the sediment immediately below this, the oxidizable phase is a highly enriched in mercury. This is the same pattern as observed for the uppermost layers of sediment in the lakes, where distinct SBLs and higher concentrations of organic matter are present. If erosion was occurring or sediment had not been recently deposited, these characteristic surface layers would be removed as well. The low concentrations of organic carbon at the sediment surface may be due to dilution by terrigenous organic material, as noted for nearshore basins of Lake Superior by Klump et al. (1989). Evidence that the suspended sediment in deep waters of the Gulf of Maine is dominated by silicates (Spencer and Sachs, 1970) suggests significant dilution of organic matter is possible. Hines et a1. (1991) found evidence that organic matter deposition was lower ill the Wilkinson Basin than in other areas ofthe GulfofMaine. Mayer et a1. (1988) found that sedimentary organicmatterinbasins oftheGulfwithwaterdepthsexceeding70mwasdominantly refractory material. They alsofindevidencethatorganicmatterreachingdlebottomindeepwaters ofthe GulfofMaineispresentmainlyascoatingsonmineralgrains,andisthereforemoreresistantto microbial breakdown (Mayer et al., 1988). These findings indicate that recent sediment is 82 accumulating, and a lack of freshly deposited sediment cannot be the cause of the lower "enrichment" of arsenic and mercury in Gulf of Maine surface sediments. A second possibility is that changes in inputs of arsenic and mercury have varied among sites. Recent inputs of these elements worldwide have increased relative to historical background levels duetoanthropogenicactivities. Ifthisincreasewas smallerintheGulfthaninthelakes,the resulting "enrichment" in more recent sediments would be smaller as well. A third possible explanation for the lower degrees of enrichment in Gulf of Maine sediments is dratmeruuymdarsaricmeuansfarednwrecmnpletelymmedeeper,reduced sediments; they are recycled less effectively. This may be the case for mercury, which is removed from porewater by sulfide formation below the redox zone. This process appears to be so effective that very little dissolved mercury is available for diffusion up toward the surface. The fact that concentrations of mercury in reduced sediments are highest for the Gulf of Maine, even though concentrations in in surface sediments are lowest (Table 5), indicates that mercury is buried more efficiently in marine sediments than in the Lakes. . Arsenic is not affected in the same manner, however. Porewater arsenic concentrations are very high, producing a steep gradient, and potentially substantial diffusive fluxes, from the reduced zone toward the sediment surface. In the upper 50 cm ofsediment, at least, there is enough available arsenic in Gulf porewaters that it is taken up by the exchangeable sites. "Background" extractable arsenic concentrations are also lowest in the Gulf of Maine; the average background concentration of total extractable arsenic is about 60% of that in Lake Superior and only about20% ofthatinlakeMichigaMseeTableS). Thesefactssuggestthatarsenicrnaybe buriedmoreeffectivelyinthelakesitesthanintheocean. Arsenicthatistransferredintothe reducedsedimentsdoesnotseemtobeprecipitatedasasulfide(asdiscussedinthesectionon arsenic diagenesis, above) btrt it may be sorbed onto iron sulfide surfaces. Kornicker and Morse (1991) found that rates of sorption onto pyrite decreased with increasing ionic strength, btrt that rates of desorption were not affected by ionic strength. Both sorption and desorption reaction rates were found to increase with increasing pH. The higher pH and much higher ionic strength 83 of the marine environment would lead to less sorption onto pyrite, and faster desorption. This would reduce the metal retention capacity of these sediments relative to freshwater sediment. A fourth possible explanation for the lower enrichment factors in the Gulf of Maine is that mercury and arsenic released by diagenetic processes within the sediments diffuse upward, but are not readsorbed and retained within the sediment column as efficiently as in the lakes. Instead, they arereleasedtothe overlyingwatercohrmn. Thedearthofreactive organicmatterinupper layers of sediment which is typical for deep waters of the Gulf of Maine (Mayer et al., 1988) would reduce the number of sorption sites available. This would certainly contribute to the lower enrichnwntofarsenicandmerctnyinsurfacesedimentsoftheGulfofMainecomparedtothe Great Lakes. Itappears mostlikelythatsomeofthearsenicreleasedfromsedimentsintheGulfofMaineby diagenetic processes is returned to the water column, which is why enrichments are lower than thoseseenintheGreatLakes. Notallofthearserricisreleased,asthereissignificantuptakeof arsenicbyironandmanganese oxidesintheoxidizedlayers ofsediment; itisjustthatlessofthe recycled arsenic is retaimdin the sediment than is thecase forthe lakes. Since both maximum and background concentrations are lowest in the Gulf, it is also possible that less arsenic is being deposited here than in the Great lakes. For mercury, the lower enrichments appear to be caused by a combination of factors. Recyclingwithinthe surface layersismuchless intenseintheGulfofMainethanintheGreat Lakes, because ofthe lackofreactive organic mattertotakeupreleasedmercury. Thereis also apparently agreaterburialofmercury inthe sedirnentcolurnnduetotheuptake ofdissolved mercury by the formation of sulfide minerals. Thus, although maximum concentrations of mercury are lowest at this site, indicating concentrations ofmercury being deposited are lowest, theamountofmercuryhrfiedhtdeepersedimartsishighesnmdicamdbymehighestbackgrmnd values of sediment-bound mercury. Although arsenic andmercury are affectedunequally by differences betweenfreshwaterand rmrine sites, the net result is the same: lower enrichment in the Gulf of Maine. The consequences of this result are different for the two elements, however. Because arsenic is retained by marine 84 sediments less effectively, its residencetime in the watercolumnwillbe longerthan inlakes, all other things being equal The opposite is true for mercury, which is more effectively buried in marine wdiments. The low abundance of reactive organic matter at the wdiment-water interface is largely responsible for the lower enrichment in both cases. This may be the factor which Gulf of Maine sediments have in common with Caribou Basin sediments. In some ways, the Caribou Basin of lakeSuperiorismoreliketheGulfofMaineflranfiketheotherlakesites. Thissimilarityis discussed further below. VARIABILIT'Y AMONG GREAT LAKES SITES Lake Michigan vs. Lake Superior Differences between Lake Michigan and Lake Superior that can influence early diagenesis include differences in pH and alkalinity of lake waters and porewaters (Frgm'es 4 and 5), and the difference in carbonate mineral content of the sediments: Lake Michigan sediments contain carbonate minerals, while modern lake Superior sediments do not (Lineback et al., 1979; Dell, 1972). Other factors influencing diagenesis vary as much or more between the two lake Superior sites as between the two lakes, so are discussed later. Although the profiles of mercury in sediments and porewaters of the Great lakes sites vary, no consistent differences between the two lakes are evident. The "background" concentrations of sediment-bound mercury are quite similar in all four sites; perhaps this reflects dominantly atmospheric inputs of mercury which may have been relatively consistent throughout the region. According to Mudroch et al. (1988), the reported range of mercury concentrations in sediments of lake Superior falls within the range reported for lake Michigan sediments, also indicating no inter-lake differences. There are also no distinct trends observed in the partitioning of nrercury among hydromorphic phases of sediments in the two lakes. Therearesomedistinctdifferences betweenthetwolakesintermsofarsenicpartitioning. lakeMichiganshowsanoverallhigherconcentration ofarsenicinthesedirnentsthanlake Superior. This most likely reflects higher inputs to lake Michigan than to eastern Lake Superior. 85 There is more arsenic in the weak-acid soluble fraction of lake Michigan sediments (Figure 8), which is probably related to the presence of carbonate minerals in lake Michigan. Alargerproportion ofthesediment-bound arsenicis irrtlreoxidizablefiactiorrinlake Michigan; in lake Superior sediments the oxidizable fraction contains little arsenic below the uppermost sediment layers (Figure 11). This may in part reflect the higher organic content of deepersedimentsinlakeMichigan. AlthoughtheorganiccarboncontentoflakeSuperior mrficialsedhnarmiscmnpmblemMoflakeMichigmudwmganicwbmwntaumdepm remains above 2 96 in lake Michigan, but drops below 2 96 fairly quickly in lake Superior sediments (Figure 6). Sediment-bound arsenic also shows a more distinct correlation with sediment organic carbon content in lake Michigan (Figure 24a). In both Lake Michigan sites, the organic carbon content of the SBL is not much higher than that of core-top sedirrrents, whereas in both lake Superior sites there is a significant decrease in organic carbon from the SBL to the top ofthe sediment column (Figure 6). The same pattern is observed for arsenic in the oxidizable fraction of these sediments (Figure 11). This indicates that more organic carbon decays at the sediment-water interface in lake Superior, and most of the arsenic associated with organic rnatteris released. In contrast, less organic nratterdecays atthe sediment-water interface of lake Michigan, and more organic carbon, with more associated arsenic, is buried into deeper sediments. Johnson et aL (1982) found that the decay rate vs. accumulation rate of organic carbon in lake Superior was related to sedimentation rate. Similarly, Klump et a1. (1989) found that the decomposition of labile organic matter in areas of lake Superior with low sediment accunrulatiorr ratesoccurredlargelyinthewatercolumnand ardresedirnent-waterinterfaoe. YetllePar-isienne hasthehighest sedirnerrtationrateofallthesesites andtheCaribouBasinthelowest(theratefor LSIPisnearly 30timestherateforLSCB; Table 5), sosedimentationratescannot account for these observed differences between the two lakes. The different patterns displayed for organic carbon arrdbetweenarsenicarrdorgarficcarboninthetwolakesmaybeduetodifferarcesindre natme of the organic matter accumulating. Kemp and Johnston (1979) found the proportions of more reactive components of organic matter (amino acids, arrrino sugars, and carbohydrates) 86 varied among lakes Ontario, Erie, and Huron. It is possible that lakes Michigan and Superior accumulate different proportions of the various organic conrporrents as well, and that this accounts fortlredifferences in orgarriccarbon burial andirr organic-associated arsenic between thetwo lakes. ThelargestdifferencesmnmgaflflrelakesamplesarebetweenCafimeasin andtherest; in many respects Ile Parisienne more closely resembles lake Michigan than it does the Caribou Basin. This is discussed further below. Lake Superior: Ile Parisienne vs. Caribou Basin The Ile Parisienne Basin and the Caribou Basin of lake Superior are very different (see Table 5 and Figures 21-22). Ile Parisienne has a sedimentation rate nearly 30tirnes that ofCaribou Basin. The redox zone in LSIP is much nearer the sediment surface than in LSCB. LSIP has higher concentrations of mercury and arsenic in surficial sediments, hence much higher enrichment factors than the Caribou Basin. LSIP is the shallowest site sampled and LSCB is the deepest LSIP is quite near shore whereas LSCB is more distant. Klurnp et al. (1989) found that the transition from nearshore to deep basins in Lake Superior was accompanied by a decrease in the fiactiorr of madflyanosable organic matter deposited orrthelakebottom. They foundthat40% oforgarricmattertlratwas depositedirranearshore shallow bay was recycled, whereas only 15% was recycled in a deep basin; this was attributed to anhrcreaseintheextentofremineralizafionwithindrewatercolunm. Thusdiffererrcesinwater depthcaninfluerrcediagerresis by influerrcingthe arnormtoflabileorganicrrrattertlratreachestlre sediments and drives rapid early diagenesis. The nearshore bay with the highest proportion of orgarricrnatterrecycled was alsofourrdtohave the lowestsurfaceorgarriccarboncontent,dueto dilution by tenigenous inorganic matter (Klump et al., 1989). These trends are observed in the two lake Superior sites. The organic carbon content ofLSIP sediments is lower than in LSCB; this indicates dilution by inorganic matter. The organic content of LSIP sediment drops off faster than it does in LSCB, srggesting rapid decomposition of the organic matter that accumulates. Early diagenetic recycling 87 also appears to be much more effective in the lSIP site, as indicated by the much higher ”enrichment factors" for this site relative to all of the other sites. Higher fluxes of metabolizable organic carbon to sediments have been formd to result in increasedrates ofrelease ofremineralizedconstituentstosedimentporewaters sothatporewater gradients steepen (Klump et aL, 1989). This would be expected to affect rrrercury and arsenic associated with organic nratter as well as nutrients; therefore, recycling and porewater gradients oftheseelernentswouldbeexpectedtobesteeperinareas withhighfluxes ofreactiveorgarric matter. Thisis alogical explanation forthe observeddifferences between LSIP (with ahiglrrate of organic carbon accumulation) and LSCB (with a low rate or organic carbon acatmulation). Johnson et a]. (1982) fotmd arelationship between the decay rate oforganic carbon and the total sedimentation rate in lake Superior that was consistent with trends observed in marine pelagic sediments; this maybeindicatedalsoby the similarities betweentheCaribouBasin andtheGulf of Maine sediments. Therole ofreactive mgarucmaueratmesedimentstufaceinrecyclingmercmyinsediments can be observed in these sites. LSCB sediments with higher percent organic carbon, but lower toral organic carbon accumulation rates has less recycling of mercury (Figure 28). Values of mercuryinsmfacesedimentsarelowestofallthelakesites andvaluesirrreducedsedirnentsare highest of all lake sites, resulting in the lowest enrichment factor of all the lake sites. Davison (1985) found that the proximity of the redox boundary to the sediment-water interface greatly influenced rates of elerrrental recycling of iron and manganese. The redox boundaryismuchdeeperintheCaribouBasinthaninany oftheotherlakesites. Itisalsofairly deepintheGulfofMaine. Depthoftheredoxboundaryisdeterminedby sedirrrentatiorrrate, organic carbon accumulation rates, and oxygen diffusion rates (Davison, 1985). Organic matter reaching the bottom of lake Superior was found to have a relatively uniform stoichiometry for the reactive component, indicating a similar source in different parts of the lake (Klump et al., 1989); this could accountforthesimilarities inorganiccarbon behavioroftlresetwositesrelativetothe 88 Lake Michigan North Basin: 2 sites Therearesomedefinitedifferences betweensedimentcorestakeninthetwositesintheNorth Basin of lake Michigan. Profiles of both dissolved arsenic and dissolved iron show multiple subsurfacemaxirna; tlresemaybecausedby bioturbation. Thisseemstobeanirnportant differencebetweenthesetwo sites: site 1 has experienwdlittlebioturbationwhilesiteZseemsto be more extensively bioturbated. Hennanson and Christensen (1991) report evidence for sedimentmixinginsites closetobothLMNB-l andLMNB-Z. This supportsthenotionofpatchy distributions of organisms and bioturbation in deep benthic environments. Arsenic data show significant differences in enrichment patterns within near-surface sediments and inporewaterprofiles. There is greatertotalenrichrrrent ofextractable arsenic atthe surface at site 1 thanatthesurfaceinsiteZ. Almosttwiceasmuclrtotalexuactablearsenicisfomrdinthe uppermostsample at site 1; thisenrichmentis observedinallofthefractions. Thereis about twiceasmuch arsenicintheEX,WAS, andOX fractions,morethantwiceasmuchintheER fraction, andslightlylessthantwiceasrrmchintheMRfractionatsite l relativetosite2. TheER fraction actually shows a depletion in arsenic in near-surface sediments at site 2. The "backgroun " concentration of arsenic in porewater '6 similar at both sites, approximately 7 rig/L. However, the near-surface gradient of porewater arsenic is much steeper at site 1; values reach a nraxirnumofnearly lOug/Lwithins cmatsite1,thenstabilizewith aslightdeclirreinarserricat increasing depths. At site 2, concentrations increase steadily to a depth of about 35 cm, then stabilize. This suggestsamore sigrrificantfluxofarsenicfromthesedirnentatsite l thanatsite2, the opposite of what would be expected due to bioturbation. Lake Superior Caribou Basin: cores collected by submersible During the final submersible dive in the Caribou Basin of lake Superior (1988), it was noticed that in sorrre areas the reddish-colored redox layer (enriched in iron oxides) was visible at the sediment surface, whereas in other nearby areas it was not. Several shore ”prurch” cores were collected along an east-to-west transect across the basin to examine the extent of variability in depthtotlreredoxlayer. Generalizeddescriptions ofthesecores,takenwithin100mofone anOther, are shown in Figure 29. In the eastern-most core (Lb-2) two redox layers are evident at Depth In Core (cm) 10“ 15" 20- 25" 30" 35“ 40-4 89 SBL Mod Tan Chy - WN" Redox Tan cuym Layer mm Guy Clay Figure 29. Cores collected by submersible from lake Superior Caribou Basin. 90 a depth of about 30 cm. In the westem-most (Lb-6), two redox layers are present between 5 and 10 cm depth. In core LD—l no redox layer is observed; this core is 35 cm long. In LD-4, only about3 meastole—6,athinredox layerispresentatthesedirrrent surface. Theextentofthe variability observed was surprising. The variations may be due to erosion by strong winter currents, such as those observed in other deep areas of lake Superior by Flood (1989). ThesecmesmggemflmmesedinwnuaMdiagemficmsesmdeepbasimofhkesmna necessarily homogeneous, even across small areas. Therefore, conclusions about sediments and diagenetic processes based on one core may not be representative of the basin. POTENTIAL FOR SEASONAL VARIATIONS Seasonal variability in diagenetic parameters has been observed in nearshore marine errvironnrents, where it related to temperature-dependent rates of nricrobially mediated organic matter oxidation (Klump and Martens, 1989) or to temperature-related variations in the intensity of bioturbation (Martin and Sayles, 1987). In these areas temperature fluctuations can exceed 20‘C annually. Bottom water temperatures in deep basins of Lakes Michigan and Superior are fairly constant, sosuch factors aremrlikelytobeimportantinthesesites. Seasonalvariationsininputs ofsediment,organicmatter,andmetals,andinlakecirculationpatternsmaybeexpected, however (e.g. Pocklington and Tan, 1987). Annual ice-out and overturn events in lake Superior waters have been found to have a substantial impact on particle transport and the dynamies of particles, organic matter, and associated hydrophobic organic contaminants (Baker and Eisenreich, 1989). They noted pulses of inorganic particles input following spring ice-out, which were concentrated in shallow near- shoreareas. Thiscouldconuihmmdrehighsedhnenmfimmemdloworgmuccarbonwntart of sediments of Ile Parisienne Basin. Baker and Eisenreich (1989) also formd that settling of particles was enhanced during summer stratification due to coagulation and fecal pellet production, and that resuspension of benthic material was potentially great during fall overturn. Evidence for the presence of strong currents in lake Superior during the winter has been found (Flood, 1989). 91 Theseprocesses wouldbeexpectedto affectarsenicandmercury,andothermetalsaswell. Sedirrrentatiorr fluxes of metals would be greatest during summer months, and during fall and winter when the lakes are isothermal and currents may be active sediment-bound contaminants can beresuspendedfromthelakebottomsandreintroducedtocoltnnn waters,potentiallyincreasing residence times and bioavailability. Johnson (1991) found seasonal variations in dissolved metal concentrations in Georgian Bay, lake Hmon which were related to higher river inputs during spring. ShnflarseasonalvafiafionsnfightalsobeexpectedindreGulfofMahrebmfiubseasonal variation in the distribution of suspended inorganic particles has been observed (Spencer and Sachs, 1970). V. SUMMARY AND CONCLUSIONS Role of Diagenesis in Geochemical Cycling Early diagenesis exerts considerable influence on the geochemical cycles of arsenic and mercury in aquatic systems. Processes operating above the redox zone are important in determining potential bioavailability and recycling elements to the water column, whereas processes operating below the redox zone are important in fixing metals in the sediments, transferring them to the next reservoir and creating the historical sedimentary record. Arsenic and rrrercury display several similarities as they undergo early diagenesis. They are both present in higher concentrations in upper layers of the sediment. They are both released from the sediment by aerobic degradation of organic matter. They are both strongly influenced by iron redox cycles. And both elements are subject to transfer between solid phases above and below redox zone; this involves transport via porewater along concentration gradients. The presence of higher concentrations in upper layers of sediment indicate that much of what is buriedinthesedimentremainsnearthesedimentsmfaceorrenmrstothe sedimentsurface. This has profound consequences for the bioavailability of these elements. As long at they remain near the sediment-water interface, they are potentially bioavailable. Bioturbatiorr and bioirrigatiorr have been found to enhance diagenetic fluxes (Belzile, 1988); thus elements can be made most available in areas where there are more organism to ingest them. Mercury is very effectively retainednearthesediment—waterinterfacebyreactionswithlabile organicrrratter; thisresultsirr enrichments of rrrercury in surface sediments relative to deeper sediments that are approximately twiceas greatasenrichmerrts ofarsenic. Thetotalenrichmerrtofbotharsenicandmercuryirr surfacesedirnentsrelativetodeepersediments appearstodependmostorrthenature oftheorganic 92 93 matter accumulating at the sediment surface. Rapid accumulation of labile (reactive) organic matter promotes strong enrichment. The dissolution of iron oxides and the changes in redox conditions associated with this cycle are major influences on both arsenic and mercury (as well as many other trace elements). Profiles of sediment-bound arsenic and mercury show a minimum in concentration coincident witlrthebase ofredoxzorre. Thisispreserrttosorrredegreeinallphases atallsitesandirrdicates that much of the sediment-bound metal reaching the redox zone is rerrroved from sediments and transferred to the porewater. Some of this moves back up into the oxidized zone via diffusion, some is transferred into the reduced zone. Below the redox zone, arsenic and mercury continue to be released from the sediments due to the continuing decay of organic matter (which is predominant for mercury) and the continuing dissolution of iron and manganese oxides (which is predominant for arsenic). In the marine environment, iron oxides can be reduced by reaction with sulfide to form iron sulfide minerals, or can be reduced nricrobially to produce dissolved Fe(II). Other metals can be adsorbed onto or or coprecipitated with the forming iron sulfides (which mercury seems to do), or can be released to solution due to the reduction of the iron oxide (which As seems to do). The presence of sulfide in reduced sediments promotes burial of rrrercury, but not of arsenic. Conclusions The goal of this project was to identify the geochemical processes operating in the sedirrrents of the sites investigated, and to determine how these processes influence the cycling of rrrercury and arsenic. Several processes were identified: the aerobic decay of organic matter (which releases sorbed mercury and arsenic), the sorption of mercury and arsenic onto fresh organic matteratthesedirnent-waterinterface (which contributes toerrrichrnentofsurface sediments),the reductive dissolution of iron and manganese oxides (which releases arsaric and mercury at the redox zone, allowing them to diffuse upward toward the sediment surface or to be transferred to 94 other solid phases within the sediment column), and the formation of sulfide minerals (which can permanently fix mercury and arsenic in the sedirrrent column) are the most important. These processes are not new; their importance in early diagenesis and elemental cycling is well docurrrented. The identification of specific influences of individual processes (particularly for mercury), and the recognition of the roles played by variations in diagenetic envirorrrrrents are new. These results provide some meaningful insights into the role of early diagenesis in the geochemical cycling of arsenic and mercury in aquatic environments. APPENDICES APPENDIX 1 METHODS Sampling Procedures CLEAN PROCEDURES Water: Deionized water (mixed resin) was further purified by distillation in a Corning model AG-ll still. The distilled-deionized water (DDW) was stored in acid-cleaned polyethylene carboys until use. DDW was used in all cleaning, processing, and analysis steps. Sample containers: Bottles, centrifuge tubes, and syringes were cleaned before use by soaking in 10% H0 (analytical reagent grade) in a water bath maintained at 60°C for 12-24 hours, rinsing 4 times in DDW, soaking in DDW for 24 horns, then rinsing again in DDW and allowing to dry in a clean hood. Containers were then capped and sealed in plastic bags for transportation to sampling sites. Gloves were worn at all times while handling sample containers. Filters: 0.4 pm pore diameter Nucleopore polycarbonate membrane filters used for filtration of water samples were cleaned by soaking in 10% HCl at room temperature for 24 hours, then rinsed 4 times in DDW, soaked in DDW for 24 hours, rinsed again in DDW, then stored in DDW in acid-cleaned polyethylene containers until use. Frlters were handled with acid-cleaned plastic forceps. Sample processing equipment: All other equipment (spatulas, scoops, etc.) used in sample processing was cleaned by soaking in 10% HCl at room temperature for 2 12 hours and rinsing 4 times in DDW. Acid-cleaned equipment was stored in plastic bags. 95 CORE SAMPLES All of the cores used for analyses were gravity cores, retrieved in 7.5 cm diameter butyrate core liners using a Bentlros gravity corer deployed from the R/V Seward Johnson. Cores were capped with plastic cups, stored upright in a cold room at 4'C (approximate in situ tenrperature), and sectiorwd within 2-3 hours of collection. Sediment was extruded using a hydraulic extrusion device; this can be done in a nitrogen-filled glove bag as required (for mercury, arsenic, and iron samples). As each section was extruded, the outer portion of the sediment which had been in contact withthecoretrrbe was removed. Samples were 1 crntlrick (neerthetoporthe core) or greater (toward the base of the core) slices of sediment, which were irmnediately transferred to acid-cleaned 50 mL polyallomer centrifuge tubes. SBL SAMPLES Sediment boundary layer (SB L) samples were collecmd via the submersible Johnson-Sea-Unk II. The mechanical arm was waved gently to suspend SBL sediment, which could then be pumped through Tygon tubing attached to the mechanical arm, through filter paper held on teflon- coated filter holders. At the surface, the filters were removed, and SBL sediments washed from the filter; this wash water was then removed by centrifugation in acid-cleaned 50 mL polyallomer centrifuge tubes. SBL samples were stored frozen. SAMPLE PROCESSING Sediment samples were centrifuged for 15 minutes at 15,000 rpm (using a chilled centrifuge head to keep the temperature near 4'C) to separate the porewaters from the sediment. In a nitrogen-filled glove bag, porewaters were removed from centrifuged samples by syringe, filtered through acid-cleaned 0.4 pm Nucleopore nrembrane filters, then acidified to pH < 2 with sub- boilirrg distilled Ultrexm nitric acid, and stored in acid-cleaned polyethylene bottles. Water sarnplestobeanalyzedformercurywerealsopreservedwidrgold(aschloroatuicacid,suchdrat 10 ng Au was added to each mL of sanrple) and hermetically sealed (using a wrench), following the prmdures of Moody et a1. (1976). All water sarrrple bottles were sealed in plastic bags and stored in a cold room maintained at 4‘C. Following removal of porewater, sediment samples 97 were stored frozen in the centrifuge tubes, which were placed in plastic bags. Sediment and water samples were transported to the laboratory packed in coolers with dry ice. pH and Alkalinity pH and alkalinity were determined for one sediment core from each site, which was sectioned in air. pH was measured by inserting a spear-tip electrode (Orion Ross combination pH) into the wet sediment before removing each section. The electrode was calibrated with pH ‘7 and pH 4 standard solutions, and the calibration verified with pH 7 standard every few meastuements. Recalibration was performed as necessary. Sections were processed as described above, and porewater separated for alkalinity analysis. Alkalinity was measured in 3 mL aliquots of porewater samples, which were titrated with 0.017 N HNO3 to an endpoint of 4.5 pH (Great Lakes) or 4.2 pH (Gulf of Maine). Appropriate endpoints were detemrined by examination of titration curves for several samples at each site. Titrations were performed using an apparatus designed for small-volume titratiorrs (Figure Al-l). Acid was added using a Brinkrnann digital micro-dispenser, in 25 pl. increments (lake Superior samples) or 50 pl. (Lake Michigan and Gulf of Maine sanrples). Volume increments of acid added were calibrated by titration of a 0.01639 N Na2'c03 standard solution prior to each series of titratiorrs. pH was measured with an Orion semi-micro gel-filled combination electrode (calibrawd as described above). Results were converted to mg/l. HCOg'. SamplesfiomacominlakeSupefiorwaetestedtodaanfinewhemeralkalhutymeasumdin air was affected by iron oxidation (which can consume alkalinity). One sample was collected from the oxidized zone, one from the redox horizon, and one from the reduced zone, all under N2. Alkalinity was determined immediately in a nitrogen-filled glove bag. Sarrrples were then removedandexposedtoair,andalkalinitymeasmedagainonasecond aliquotoftlresanrple. Thesesarnesarnples werethenallowedtosit,opentothe atmosphere, for7hours; alkalinitywas measured once more, using a third aliquot. The results of this test are shown in Figure Al-2. Althoughthere seemstobeasliglrtreductionalkalinity withtime,thismaybewithintheerrorof 98 Auto- Dlepeneer pH electrode negnettc StIr Plate Figure Al-l. Alkalinity measurement apparatus for small-volume samples. 4—-— Depth ln core Alkalinity (mg/L H005) 10 20 30 40 f Oxidzed Zone Redox Zone I Nitrogenatmosphere a Air 9 Air,7houtslater Reduced Zone Figure A1-2. Results of alkalinity measurement in air ya. nitrogen atmosphere test. 100 the rrrethod. The least amount of variation is seen in the sample from the reduced zone, where effects of iron oxidation would be seen. In this sample, there is no difference between alkalinity measuredinNzandthatinairsoon afterward. This suggests thattlreeffects ofiron oxidation are insignificant at porewater iron levels encountered, and alkalirrities determined in air are valid. Sequential Chemical Extractions ARSENIC Arsenic was extracted from sediment samples following the procedures developed by McKee et al (1989) from the rrrethods of Gephart (1982), Gupta and Chen (1975), and Tessier et a1. (1979). Steady-state analysis was performed to verify that these procedures and reaction times were suitable for arsenic (see Figure A1-3). All processing steps that involved opening sample containers were done in an inert atmosphere (N2 bag), until the final oxidizing step. Following each extraction step, leachate was separated form sedirrrent by centrifuging at 15,000 rpm for 20 nrin. Between extractions, the sediment was washed by adding 10 mL of distilled deionized water (DDW). This was mixed into the sediment with a vortex mixer, then samples were centrifuged to separate the water, which was removed by pipetting. Sediment samples were then treated with the subsequent extraction procedure, either immediately or following overnight storage in a refrigerator. Samples were thawed in a refrigerator for 3-7 days. In a Nz-filled glove bag, sample tubes were opened and a portion transferred to acid-cleaned, pre-weighed labelled centrifuge tubes. An additional portion was transferred to small plastic bottle for determination of the dry/wet weight ratio. Centrifuge tubes containing sediment subsamples for extractions were then re-weiglwd to determine the weight of the wet sediment; this was later converted to dry weight equivalent using the dry/wet weight ratios. 1. Exchangeable fraction (EX) 10 mL of 1.0 M magnesium chloride (at 7 pH) were added to each sample. Sample tubes were placed on a wrist-action shaker for 1 hour, then centrifuged. Leachate was transferred into acid-cleaned plastic bottles, and sediment was rinsed. Arsenic Concentration (mall. in solution) 101 SteadyoState Analyele Arsenic Extractions Weak Acid Soluble Fraction Eaelly Reducible Fraction 200 ‘ 40 - 150 4 so ‘ /' 100 i 20W ._._._- I so . 10 . / 0 U I V V I o 1 I V U I 2 3 4 5 6 7 0 1 5 3 0 45 60 75 Tina (hang) The (names) Noderately Fleduclble Frectlon Oxidizable Fraction 20d - J 150 ‘ _ _ - ———-_ 400 \3 : - 100 ‘ J 200 i 50 o ‘1 r r o 1 T 3 4 5 6 7 3 4 5 6 Tina (liars) fine W8) Figure Al-3. Results of steady-state analysis for arsenic extractions. 102 2. Weak-acid soluble fraction (WAS) 10 mL of 1.0 M sodium acetate (pH adjusted to 5.0 with acetic acid) was added to each sample. Tubes were placed on the shaker for 5 hours, then centrifuged Leachate was transferred into acid-cleaned plastic bottles and sediment was rinsed. 3. Easily reducible fraction (ER) 25 mL of 0.10 M NHZOH-HCL in 0.010 M HN03 was added to each sample, which was placed on the shaker for 30 nrin, then centrifuged. Leachate was transferred into acid-cleaned plastic bottles and sediment was rinsed. 4. Moderately reducible fiaction (MR): 20 mL of 0.040 M hydroxylamine hydrochloride in 25% (v/v) acetic acid was added to each sample. Samples were placed in a water bath maintained at 90°C for 5 hours. Samples were agitated approximately every 30 nrin. Samples were centrifuged, leachate was transfened into acid-cleaned plastic bottles, and sediment was rinsed 5. Oxidizablefi‘oction (0X): 3 mL of 0.020 M HNO3 was added to each sample, then a total of 8 mL of 30% H202 (with pH adjusted to 2 using I-INO3) was added. The peroxide was added in 500 uL aliquots to prevent bubbling-over of samples. Samples were plawd in a water bath maintained at 85°C for 5 hours and were agitated approximately every 30 min. Dining this step, samples were exposed to the atmosphere since bottle caps had to be left unscrewed during heating. After 5 hours, samples were placed on the shaker to cool, then 5 mL of 3.2 M ammonium acetate was added to each sample, and shaking continued for 1 hour. Leachates were then pipetted into 25 mL Class A volumetric flasks and diluted to 25 mL with DDW. Leachates were then transferred to acid-cleaned plastic bottles for storage. Sediment was washed. MERCURY Mercury was extracwd from sediment samples following the sequential chemical extraction procedures developed by Strunk (1991). All processing steps that involved opening sample containers was done in an inert atmosphere (N; bag). Following each extraction step, leachate was separated from sediment by centrifuging at 15,000 rpm for 20 minutes. Between extractions, the sediment was washed by adding 10 mL of distilled deionized water (DDW). This was mixed into the sediment with a vortex mixer, then samples were centrifuged to separate the water, which was removed by pipetting. Sediment samples were then treated with the subsequent extraction procedure, either irrrrnediately or following overnight storage in a refrigerator. 103 Frozen samples were thawed in a refrigerator for 3-4 days. In a Nz-filled glove bag, sample tubes were opened and a portion transferred to acid-cleaned pie-weighed labelbd centrifuge tubes. An additional portion was transfened to small plastic bottle for determination of the dry/wet weight ratio. Centrifuge tubes containing sediment subsanrples for extractions were then re- weighed to determine the weight of the wet sediment; this was later converted to dry weight equivalent using the dry/wet weight ratios. 1. Exchangeable fraction (EX): 10 mL of 10% (w/v) KCl was added to each sample and samples were placed on a wrist-action shaker for 1 hour. Samples were then centrifuged. Leachate was transfened into acid-cleaned plastic sarrrple containers and analyzed immediately (as described below). Sediment was rinsed as described above. 2. Base Soluble fraction (BS): 15 mL of 0.10 N NaOH was added to each sample. Samples were placed on the shaker for 30 hours, then centrifuged. Leachate was transferred into acid-cleaned plastic sample containers and analyzed immediately. Sediment was rinsed. 3. Acid Soluble fraction (AS): 10 mL of 1.0 N HCl was added to each sample. Samples were plawd on the shaker for 6 hours, then centrifuged. Leachate was transferred into acid- cleaned plastic sample containers and analyzed immediately. Sediment was rinsed. 4. Oxidizable fraction (0X ): 2 mL of 0.020 M HNO3 was added to each sample, then a total of 7 mL of 30% H202 (with pH adjusted to 2 using HNO3) was added. The peroxide was added in 1 mL aliquots to prevent bubbling-over of samples. Samples were placed in a water bath maintained at 50°C for 5 hours and were agitated approximately every 30 min. During this step, samples were exposed to air since bottle caps had to be left unscrewed during heating. After 5 hours, samples were placed on the shaker to cool, then 4 mL of 2.0 M ammonium chloride in 20% HNO3 was added to each sample, and shaking continued for 1 hour. Leachates were then pipetted into 25 mL Class A volumetric flasks and diluted to 25 mL with DDW. Leachates were then transferred to acid-cleaned plastic sample containers and analyzed. Sediment was washed. 104 Chemical Analysis ARSENIC Arsenic was analyzed in leachates and porewaters by graphite furnace atomic absorption using stabilized temperature platform furnace (STPF) conditions. This involves use of pyrolitic-coated graphite tubes, L'vov platforms, a cooldown step in the furnace program prior to atomization of the sample, maximum-power atorrrization, internal gas stop during atomization, fast spectrometer electronics, peak area measurement, baseline offset correction, matrix modification, and Zeeman effect background conection (Beaty, 1988). Several studies have shown this set of conditions to provide superior results (e.g. Grobenski et al., 1984; Desaulrriers et al., 1985; Letourneau et aL, 1987). Calibratiorrs were performed using blanks and standards prepared in extraction solutions for the leachates or in acidified DDW for porewaters. Each of these matrices required slightly different programs for optimum results; optimum conditions were determined prior to each set of analyses. Most of the leachate solutions required dilution with DDW prior to analysis; this was done automatically by the autosampler according to defined ratios. Sample volumes (or sample + diluent volumes) were 20 11L. Nickel nitrate was used as a matrix modifier and was added automatically by the autosampler (5 pl. of 0.068 M Ni(NO3)2 solution, to give 0.02 mg Ni per 20 itL sample). Temperature programs and dilution factors are summarized on Table A1-1. MERCURY Mercury was analyzed in leachates by flow-injection [hydride-generation technique using the Perkin-Elmer Zeeman-5100 atomic absorption (AA) spectrometer with MHS / FIAS 200 system. This method utilizes flow injection technology to determine mercury concentrations in small- volume samples. Peristaltic pumps transport sample to a mixing manifold where it reacts with sodium borohydride. Mercury is reduced to elemental rrrercury vapor, which is transferred by argongas toaheatedquartzcellintheAAlightpath, where absorbanceisread. Thismethoduses less concentrated reagents than other hydride reduction nrethods, reducing potential contamination. The carrier solution was 1.0 % (v/v) HCl prepared using Optima" HCl and 105 DDW. Reduction solution was 0.2 % (w/v) NaBH4 prepared with Aldrich reagent-grade NaBI-I4 in 0.05 % (w/v) NaOH and DDW. Samples were also acidified to 1 % HQ with Optima” HCl. Mercury in porewater samples was analyzed by FlAS-MHS with the addition of the Perkin- Elmer Amalgam System attachment which arnalgarnates mercury from samples onto a gold- platinum gauze, allowing lower concentrations of mercury to be detected Mercury vapor generated in the mixing manifold is transferred to the gold in the Amalgam. Upon heating. mercuryisreleasedatonce,whereuponitiscarriedby argongastotheheamdquartzcellinthe AA light path, where absorbance is read. Conditions for analysis of mercury in waters and sediment leachates are summarized in Table A1 -2. Organic Carbon Determinations The organic carbon content was measured in splits of the core samples used for chemical extraction for arsenic, using the modified Walkley-Black procedure of Gaudette et a1. (1974). Samples were dried at 60'C, ground with a mortar and pestle, dried again, and 0.2 to 0.5 g was weighed out. The sample was oxidized with potassium dichromate (1.0 N) and sulfuric acid (concentrated) for 30 rrrinutes. Phosphoric acid (85%) and sodium fluoride were added The dichromate remaining after oxidation of the organic matter was titrated with ferrous ammonium sulfate (0.5 N), using diphenylarrrine as an indicator. The percent organic carbon in each sample was calculated as: % OC = 10 (l-T/S) [(1.0) (0.003)/W] 100 where 10 = the volume of K2Cr207 added T = the volume of Fe(NI-l4)2(SO4)2 required to titrate the sample S = the volume of Fe(NH4)2(SO4)2 required to titrate the blank 1.0 = the normality of the K2CrzO-, .003=themass(g)of1meqofcarbon W=themassofthesample,and 100 is to convert to percent. Samples were run in batches of 6, with a blank prepared for each batch. Table Al-l. Graphite furnace conditions for arsenic analyses. 106 Fraction: Porewater EX WAS ER MR OX Dilution none 1to 3 Hot none HM 1101 Dry Temp 120 130 130 130 130 130 Step Ramp 1 1 1 1 1 1 Time 60 85 70 90 75 75 Thermal Temp 1400 900 1500 1500 900 1200 Pretreatment Ramp 1 1 5 4 1 1 Time 40 30 25 35 20 10 Cool Temp 20 20 20 20 20 20 Down Ramp 1 1 1 1 1 1 Time 15 15 15 15 15 15 Atomlzation Temp 2600 2500 2400 2300 2600 2500 Ramp 0 0 0 0 0 0 Time 5 5 5 5 5 5 Clean Temp 2600 2600 2600 2600 2600 2600 Out Ranrp 1 1 1 1 1 1 Time 5 5 7 7 5 5 Temperatures in °C Sample (orsamplei-diuentwoltune-ZOpL NiN03 matrix modifier added (0.02 mg Ni per 20 11L sample) Dilution indicates sample to diluent ratio; diluent is DDW Ramp and hold Times in seconds 107 Table Al-2. FlAS-200 conditions for mercury analyses. Porewater EX BS AS OX Method amalgam normal amalgam normal normal Dilution - . 1/10 . . Acid added 3% HCI 3% HCI 3% 1101' -1 -1 ' add acid immedimely before analysis tsolution sufficiently acidic Amalgamatoatan time speed (rpm) valve other (809 pump 1 pump 2 position events Steps: prefill 15 100 40 till - fill 15 100 40 till dr, soon lnieet 25 0 120 inject upon flush 10 0 40 till argon heat 15 0 40 till had. read cool 10 0 40 till air, upon return 1 0 0 till - Fiun using short reaction coil, and with glass fiber only in drying time. Cell temperattre 100°C Sample volume 1000 ttL HAS. flagrant time speed (rpm) valve other (see) pump 1 pump 2 position events Steps: prehll 20 100 120 till - till 15 100 120 illl - inject 20 0 120 Inject read return 1 0 120 till - Argon flow 50 mlJmin Run using long reaction coil Cell temperature 150°C Sample volume 500 at APPENDIX 2 SAMPLE DATA Table A2-1 presents shipboard core descriptions. Cores are listed in order by core number, from 1988gc1 through 1989gc8. Locations and water depths are also given for each core. SBL sample locations and water depths appear at the end of the table. Table A2-2 presents data for the sediment sub-samples used for the chemical extractions and those for determination of wet/dry weight ratios. Core and SBL sample data for arsenic analysis appear first, followed by core and SBL sample data for mercury analysis. 108 109 Winn: Core Sample Depth cm Description 19889c1 pH, Alkalinity 44° 46.19' N. 86° 43.37‘ W, depth = 242 m (792 it) 1 0 - 1 light brown flocculent sediment 2 1 - 2 ...same, turning slightly greyer 3 2 - 3 dark grey, somewhat firmer 4 3 - 4 dark grey with black streaks and light brown streaks 5 4 - 5 black bands (1 mm thick 10p & bottom) In pale grey 6 5 - 6 ...same 7 6 - 7 top 1/2 cm grey, bottom 112 cm black. odor sulfide? 8 7 - 8 ...same 9 8 - 9 ...same 10 9 - 10 ...same. distinctly banded. bottom very dark 11 10 - 11 darker, 3 black bands with grey/brown between; drier 12 11 - 13 sulfide odor. organic matter 13 - 19 altematlng dark greylblack and light grey bands 13 19 - 21 light grey with one 3 mm black band 21 - 31 top 5 cm banded then 5 cm grey clay 14 31 - 33 banded dark/grey 33 - 43 4 dark bands ~2 mm thick a few cm apart 15 43 - 45 streaky < 1 mm black bands in light grey/tan mud 45 - 55 top 4.5 cm tan. 2 cm banded, 3.5 cm tan 16 55 - 57 light tan/grey with one ~1 mm thick black streak 57 - 67 patchy rather than banded (black on tan) 17 67 - 69 ...same 69 - 79 ~2 cm dark. 4 cm tan, 1.5 cm dark. rest tan with streaks 18 79 - 81 ~1 cm thick black band at top. tan below, quite solid 81 - 91 alternating black 8 tan layers ~2 cm thick: sulfide odor 19 91 - 93 very iinn with thin bands of black in tan 20 93 - 96 mostly dark: seems lalrly dry 19889c2 Mercury 44° 46.19' N, 86° 43.37' W. depth = 242 m (792 it) 1 0 - 1 light greenish brown flocculent sediment 2 1 - 2 same color, somewhat firmer 3 2 - 3 ...same 4 3 - 4 somewhat greyer; worm (7) 5 4 - 5 ...same, timer 6 5 - 6 ...same. with ~1 mm thick black band 7 6 - 7 lighter ~2 mm with dark streaks; dry area at bottom 8 7 - 8 seems more solid, but gets weeter again at bottom 9 8 - 9 gray color, no dark streaks/layers 10 9 - 10 same appearance. slightly lirrner; wetter at bottom 11 10 - 11 same color with small black streak; middle lirmer 12 11 - 12 ...same. llnner atbase 110 Table A2-1 Continued Core Sample Depth (cm) Description 1988gc2 12 - 14 grey with dark grey streaks 13 14 - 16 top: darker grey streaks, base: lighter with black band 16 - 24 top 3 cm dark grey. 3,5 cm tan/grey. 1.5 cm dark grey 14 24 - 26 light with dark streaks; ~2 mm black layer in center 26 - 34 alternating layers dark grey and light tan/grey wl black 15 34 - 36 tan/grey with black band at top. dark streaks at bottom 36 - 44 alternating 2-3 cm layers tan/grey and dark greylblack 16 44 - 46 light tan/grey 46 - 54 ...same, with some mottling, sand at botorn 17 54 - 56 light tan/grey lirm clay wl black streaks; pocket of sand 56 - 64 very dry, firm layers dark gray; water oozes lrom cracks 18 64 — 66 color slightly lighter. no dark streaks 66 - 74 ...same 19 74 - 76 ...same 20 76 - 78 ...same 19889c3 Arsenic 44° 45.98' N. 86° 42.94' W, depth = 254 m (832 ft) 1 0 - 1 Flocculant, dark brown, very soupy 2 1 - 2.5 Soupy, tan becoming grey below 3 2.5-3.5 Dark grey, drier 4 3.5-4.5 Dark grey. moist (wetter than above) 5 4.5-5.5 Dark grey becoming tan 6 5.5-6.5 Tan 8 grey 7 6.5-7.5 Tan 8 black layers. becoming drier 8 7.5-8.5 Tan with many black streaks 9 8.5-9.5 Tan with black streaks, 1-2 mm top black band 10 9510.5 Tan with black streaks 11 10.5-11.5 Tan with black streaks. drier 12 16.5-18.5 Homogeneous tan 8 grey, lew black streaks. moist 13 23.5-25.5 Tan 8 grey with few black streaks 14 30.5-32.5 Tan 8 grey with black streaks 15 37.5-39.5 Tan 8 grey with black streaks 16 44546.5 Tan 8 grey with black sneaks. moist 17 51.5-53.5 Tan 8 grey with black streaks 18 585-605 Tan with dark streaks which are chunky 1988904 Iron 44° 45.98' N, 86° 42.94' W. depth 2 254 m (832 ll) 1 0 - 1 liocculenl sediment with some overlying water 2 1-2 talrlysolid(iortopoicore) 3 2 - 3 grey mud 4 3 - 3.3 hard dry layer ~3 m (not like redox), soupy below 111 Table A2-1 Continued Core Sample Depth (cm) Description 19889c4 5 3.3 - 4.3 soupy clay 6 4.3 - 5.3 ...same 7 5.3 - 6.3 tan with darker (organic?) layers 8 6.3 - 7.3 ...same 9 7.3 - 8.8 ...same 10 8.8 - 10.1 ...same 11 10.1 -11.1 ...same. becoming firmer 12 11.1 - 12.1 ...same 12.1 -22 ...same 13 22 - 24 ...same 24 - 34 ...same 14 34 - 36 ...same 36 - 46 ..same, with hair-like things 15 46 - 48 thick black streaks 48 - 74 16 74 - 76 quite cohesive clay. mostly tan. some darker clay 1988907 pH & Alkalinity 44° 28.42‘ N, 86° 45.07‘ W. depth = 275 m (900 it) 1 0 - 1 tan/grey goo with some overlying water 2 1 - 2 tan soupy mud 3 2 - 3 ...same 4 3 - 4 becoming grey 5 4 - 5 dark grey, very dark at bottom 6 5 - 6 dark grey becoming black 7 6 - 7 thin layer of dark on top. mostly tan clay 8 7 - 8 tan with black specks. thin streaks 9 8 - 9 ...same 10 9 - 10 tan mud. still moist 11 10 - 11 tan with a few grey streaks, becoming firmer 12 11 - 12 ...same 12 - 17 ...same 13 17 - 19 ...same 19 - 24 ...same 14 24 - 26 ...same 15 30 - 32 ...same 32 - 36 ...same 16 36 - 38 ...same 38 - 42 ...same 17 42 - 44 ...same 44 - 48 same with darker streaks near bottom 18 48 - 50 same with very faint dark streaks 112 Table A2-1 Continued Core Sample Depth (cm) Description 1 988907 50 - 54 same with black layer at base 19 54 - 56 alternating grey/tan bands 56 - 60 alternating greyhlack bands 20 60 - 62 grey with diffuse black bands 62 - 66 ...same 21 66 - 68 grey with few black bands 68 - 72 grey with ~2 om black band at ~70 cm 22 72 - 74 ...same 74 - 78 several black bands ~0.5 cm thick 23 78 - 80 ...same 80 - 84 ...same 24 84 - 86 ...same, a little drier 86 - 90 ...same 90 - 92 grey 8 tan with black streaks 92 - 96 ...same 96 - 98 ...same 1988gc9 Mercury 44° 28.42' N, 86° 45.07‘ W, depth = 275 m (900 ft) 4 0 -1 Greenish-tan flocculant material with overlying water 2 1 - 2 Same. becoming greyer at base 3 2 - 3 Grey-tan, a little firmer (but still soupy) 4 3 - 4 same 5 4 - 5 ...same. but with dark greylblak at bottom ~1I2 cm 6 5 - 6 dark grey with discontinuous black streaks (~1 mm) 7 6 - 7 same 8 7 - 8 same. with a pebble: lighter grey/tan at bottom 9 8 - 9 light grey/tan with dark grey streaks 10 9 - 10 same 11 10 - 11 light greyl‘tan with small black. things (~.5 cm x 1 mm) 12 11 - 12 same. without black things 13 12 - 14 light tan/grey with few fmall dark streaks 14 18 - 20 light tan/grey with few small discontinuous black streaks 15 24 - 26 light tan/grey with dark streaks 16 30 - 32 same. dark streaks rare 17 36 - 38 same 18 42 - 44 same 19 48 - 50 same. maybe more dark streaks 20 54 - 56 same. dark streaks rare 21 60 - 62 same 22 66 - 68 light tan/grey with ~ 5 mm thick dark streaks 23 72 - 74 alternating tan bands and dark/black 24 78 - 80 same. more dark than light 113 Table A2-1 Continued Core Sample Depth (cm) Description 19889010 Iron 44° 28.42' N, 86° 45.07‘ W, depth = 275 m (900 ft) 1 0-1 2 1 -2 3 2 - 3 4 3-4 5 4-5 6 5-7 7 7-9 8 9 - 11 See core 1988909 for description. 9 11 - 13 10 13- 15 11 15- 17 12 27-29 13 39-41 14 51-53 15 63-65 16 75-77 19889011 Arsenic 44° 28.42“ N, 86° 45.07' W, depth = 275 m (900 ft) 1 0 -1 2 1-2 3 2 - 3 See core 1988909 for description. 4 3-4 5 4-5 6 5-6 7 6-7 8 7-8 9 8-9 10 9- 10 11 10-11 12 11-13 13 17-19 14 23-25 15 29-31 16 35-37 17 41-43 18 47-49 19 53-55 20 59-61 21 65-67 22 71-73 23 77-79 24 83-85 114 Table A2-1 Continued Core Sample Depth (cm) Description 19889012 pH & Alkalinity 46° 44.84' N. 84°46.96' W. Mh = 122 m (400 it) 1 0 - 1 light brown flocculent mud 2 1 - 2 top same. less soupy; orange, cake-like redox at base 3 2 - 2.5 solid cake-like material with sand. greyer mud at base 4 2.5 - 325 alternating tanlbrown mud, some sand 5 325 - 4.25 top firm cakellke redox; below grey mud. sulfide odor 6 425 - 525 top 5011 mud with black streaks. than ten. sandy mud 7 5.25 - 6.25 alternating dark/black and light tan layers with sand 8 625 - 725 similar. tan with 1 black band near top 9 725 - 825 tan with thin black streaks 10 825 - 10 ...same 11 10 - 12 ...same 12 12 - 14 ...same 13 14 - 16 ...same 16 - 20 ...same 14 20 - 22 ...same 22 - 26 ...same, becoming slightly more cohesive 15 26 - 28 ...same 28 - 32 ...same, contains a pebble ~ 1 cm diameter 16 32 - 24 ...same 34 - 38 ...same 17 38 - 40 ...same 40 - 44 ...same 18 44 - 46 ...same 46 - 50 ...same 19 50 - 52 ...same 52 - 56 ...same 20 56 - 58 ...same 58 - 62 ...same 21 62 - 64 ...same 64 - 68 ...same 22 68 - 70 ...same 19889015 Arsenic 46° 44.84' N, 84°46.96‘ W, depth = 122 m (400 it) 1 0 - 1 Tan fiocculent clay 2 1 - 2 Redox layer 3 2 - 3 Grey clay 4 3 - 4 Dry tan clay 5 4 - 5 Grey on top. changing to tan 6 5 - 6 Tan clay 7 6 - 8 same 8 8 - 10 .. same 9 10 - 12 same 10 12 - 14 .. same 11 14 - 16 .. same 12 16 - 18 .. same 13 28 - 30 .. same 14 40 - 42 .. same 115 Table A2-1 Continued Core Sample Depth (cm) Description 19889015 15 52 - 64 same 16 74 - 76 same 19889017 Mercury 46° 45.15' N, 84°47.02'w, depth = 122 m (400 ft) 1 0 - 1 brown flooculent mud: redox at base 2 1 - 2 redox. becoming grey-brown at base 3 2 - 3 grey. some redox near center of core 4 3 - 4 grey becoming tan with black layers 5 4 - 5 tan with black layers, changing to tan. less firm 6 5 - 6 tan 7 6 - 7 tan with dark grey streaks 8 7 - 8 tan with a little grey 9 8 - 9 tan 10 9 - 10 tan 11 10 - 11 tan 12 11 - 12 tan with black streaks near base 13 12 - 14 tan 14 14- 16 tan with dark 15 16 - 18 tan with occassional thin dark streaks 16 18 - 20 ...same 20 - 25 ...same 17 25 - 27 ...same 27 - 32 ...same 18 32 - 34 ...same 34 - 39 ...same 19 39 - 41 ...same 41 - 46 ...same 20 46 - 48 ...same 48 - 53 ...same 21 53 - 55 ...same 55 - 60 ...same 22 60 - 62 ...same 62 - 67 ...same 23 67 - 69 ...same 69 - 74 ...same 24 74 - 76 ...same: ~1 mm thick sand layer in middle 19889019 pH& Alkalinity 47" 22.26' N. 86° 58.03' W. depth = 330 m (1082 11) 1 0 - 1 brown flocculent mud 2 1 - 2 same. changing to tan firmer clay 3 2 - 3 tan clay 4 3 - 4 ...same 5 4 - 5 ...same 6 5 - 6 ...same 7 6 - 7 ...same with lighter colored blotches 8 7 - 8 ...same 9 8 - 9 ...same 10 9 - 10 ...same 11 10 - 11 tan changing to lighter orange-tan 116 Table A2-1 Continued Core Sample Depth (cm) Description 1988901 9 12 11 - 12 hard orange redox layer ~4 mm, tan wired streaks below 13 12 - 13 tan clay with small reddish streaks and light tan blotches 14 13 - 14 ...same 15 14 - 15.5 ...same 16 15.5 - 16.5 ...same 17 16.5 - 18 ....same red streaks larger 18 19 - 20 ...same, with thin discontinuous redox bands 19 20 - 22 ...same; redox discrete layers s 1 mm thick 20 22 - 24 same with firmer clay streaked with grey 21 24 - 26 ....same changing to greyer clay 26 - 31 ...same. greyer, wetter at base 22 31 - 33 grey day with small black streaks 33 - 38 ...same 23 38 - 40 ...same 40 - 45 ...same 24 45 - 57 ...same 57 - 52 ...same 25 52 - 54 ...same 54 - 59 ...same 26 59 - 61 ...same 61 - 66 ...same 27 66 - 68 ...same 68 - 73 ...same 28 73 - 75 ...same 19889022 Mercury 47° 22.26‘ N, 86° 58.03' W, depth = 330 m (1082 ft) 1 0 - 1 brown flocculent on top; lighter brown clay beneath 2 1 - 2 tan clay 3 2 - 3 ...same 4 3 - 4 ...same. start to see red streaks at bottom 5 4 - 5 ...same 6 5 - 6 ...same. number of red streakslblotches increasing 7 6 - 7 ...same with orange redox layer ~2 mm thick in center 8 7 - 8 top ~ 1 mm = tan clay with v. thin redox; grey clay below 9 8 - 9 grey clay 10 9 - 10 ...same 11 10 - 11 ...same 12 11 - 12 ...same 13 12 - 14 ...same 14 14 - 16 ...same 15 16- 18 ...same. with some tan spots 16 189 - 20 ...same 20 - 25 ...same 17 25 - 27 ...same 27 - 32 ....same tan spots disappear toward bottom 18 32 - 34 plain grey clay 34 - 49 19 39 - 41 ...same 41 - 46 ...same 20 46 - 48 ...same 48 - 53 ...same 117 Table A2-1 Continued Core Sample Depth (cm) Description 19889022 21 53 - 55 ...same 55 - 60 ...same 22 60 - 62 ...same 62 - 67 ...same 23 67 - 69 ...same 69 - 74 ...same 24 74 - 76 ...same 19889025 Arsenic 47° 21 .96' N. 86° 58.01' W, depth = 305 m (1000 ft) 1 0 - 1 Dark brown flocculent mud on top. then tan soupy clay 2 1 - 2 Tan clay, with some flocculent material 3 2 - 3 Tan 4 3 - 4 Tan clay 5 4 - 5 same 6 5 - 6 same 7 6 - 8 same 8 8 - 10 same 9 10 - 12 Tan. becoming mottled with lighter clay 10 12 - 13 Tan 11 13 - 14 Ten with dark streaks 12 14 - 15 Tan with dark streaks 13 15 - 16 Tan with darker streak 14 16 - 17 Darker tan with orange areas 15 17 - 18 Tan, seems to be getting darker 16 18 - 19 same 17 19 - 20 same 18 20 - 21 Mottled dark/light tan 19 21 - 22 Darker tan with light tan mottling 20 22 - 24 same 21 24 - 26 Top cm = solid, dry dark tan. then wet grey clay below 22 31 - 33 Wet grey clay 23 38 - 40 same 24 45 - 47 Lighter colored. very wet grey clay 1989903 Arsenic 42° 26.53' N, 69° 45.97' W. depth 2 284 m (930 ft) 1 0 - 1 Light brown flocculent 2 1 - 2 same 3 2 - 3 Light brown with darker streaks 4 3 - 4 same 5 4 - 5 Ught brown, slightly firmer 6 5 - 6 same 7 6 - 7 Ught brown. pudding-like. still burrows 8 7 - 8 Ught brown, burrowed 9 8 - 9 ...same 118 Table 112-1 Continued Core Sample Depth (cm) Description 1989903 10 9 - 10 same 11 10 - 11 same 12 11 - 13 same 13 13 - 15 Same, without burrows 14 15 - 17 same 15 17- 19 Same, firmer clay 16 19 - 21 same 17 21 - 23 same 18 23 - 25 same 19 25 - 28 same 20 28 - 31 same 21 31 - 34 same 22 34 - 39 same 23 39 - 44 same 24 44 - 49 same 1989905 pH & Alkalinity 42° 26.48' N, 69° 46.29' W. depth = 284 m (930 ft) 1 0 - 1 rusty brown 2 1 - 2 brown 3 2 - 3 somewhat greyer 4 3 - 4 ...same 5 4 - 5 ...same 6 5 - 6 ...same 7 6 - 7 ...same with dark spots 8 7 - 8 ...same 9 8 - 9 olive clay 10 9 - 10 ..same 11 1o - 11 ...same 12 11 - 13 ...same 13 13 - 15 ...same 14 15 - 17 ...same 15 17 - 19 ...same 16 19 - 21 ....same becoming very sticky 17 21 - 23 ...same 18 23 - 25 ...same 19 25 - 38 ...same 20 28 - 31 ...same 21 31 - 34 ...same 22 34 - 37 ...same 23 37 - 40 ...same 24 40 - 43 ...same 119 Table A2-1 Continued Core Sample Depth (cm) Description 1989906 Iron 42° 26.50‘ N. 69° 46.29' W. depth = 284 m (930 it) 1 0 - 1 2 1 - 2 3 2 - 3 Squeezer core: similar to 1988 905 4 3 - 4 5 4 - 5 6 5 - 6 7 6 - 7 8 7 - 8 9 8 - 9 10 9 - 10 11 10 - 12 12 12 - 14 13 14 - 16 14 16 - 18 15 18 - 20 16 20 - 22 17 22 - 24 18 24 - 26 19 26 - 30 20 30 - 34 1989908 Mercury 42° 26.50' N, 69° 46.29' W. depth = 284 m (930 it) 1 0 - 1 wet brown mud 2 1 - 2 brown mud, becoming somewhat greyer; worm 3 2 - 3 ...same 4 3 - 4 ...same 5 4 - 5 ...same 6 5 - 6 ...same with burrow 7 6 - 7 ...same 8 7 - 8 ...same, with burrows 9 8 - 9 ...same 10 9 - 10 ...same, becoming sticky 11 10 - 11 ...same 12 11 - 13 ...same. with creature: ~0.5 cm long, hard 13 13 - 15 ...same. becoming stlckier 14 15 - 17 ...same 15 17 - 19 ...same 16 19 - 21 ...same. with black worrny thing 17 21 - 23 ...same 18 23 - 25 ...same 19 25 - 28 ...same 20 28 - 31 ...same 21 31 - 34 ...same 22 34 -37 ...same 120 2.9.... 8de 83.3 ..me.... ~39... mm L. - m . .mmmm. m..mm.m 2..de mmmmmd mama. mm m . m m 2.8m. «8:3 .836 mmmmmd «83.. m... m . v m 958.. «8.3 853 54.3 .mowm. mm c . m e mmvmm. 281m 2936 89...... «88.. md m - N m 8.3.. «88... 88a... «muomd momma. m.. m .. m 288.. mmmmm.~ 2.8mm mmmmmd .mmmg m.o . - o . ..oummm. mfiNm. .mmmm.~ 8. .m... .88.. 823 mmm m.om.m.mm m. 8m... 38$ 88m... 8.8.. «83.. mam m.mm.m..m .. ..83... 3.8.. 3. .m.o 88m. 893 m.me mmvmi m. m.mmm.o 3m... 23m... .836 .38.. mmm mmmmxm m. mmmmmd 8.8.. mmmsed .336 .58.. m..m m.~m.m.8 4. 8mm... memmmw. Sam... 88... SSE mew m.m~.m.m~ m. 883 wummm. 88.8 «88.. mmomod m... m.m..m.m. m. .336 Some. 98...... mm.m~.. mafia o. .. m....m.o. .. .vmvmd mmmmm. 88mm 8mm: .88.. ad. m.o..m.m o. mm.~m.o Ema... Emu-o 3R3 imvm. om m.m.m.m m 0886 88.... 88.8 .88... mmomod om m.m.m.\. m mBmmd .5... 2.346 «Emma gamed 3. msmm . 89m... mmmmm. memmed «836 8...... cm m.m.m.m m ~88... om.B.~ 3.8.3 833 mmmmo.~ om m.m.m.v m 8.8... 884.. 5.33 85m... «8.8.. o... m....mm 4 it; 88... mums... 883 «82... cm m.m.m.~ m vuovmd .m .3... 2.8.3 4823 mmma. m.. ma . . u summed .23.. 88m... m.mmm.o mmu .m.. m.o . - o . mommmm. 283. 95.3 mESwl 2.9 E903 mEEd1 2.2m Eu .8 29:3 £96; an .. £93 8; 623.. £92. to sue; a; 58.. :82 .95: snow: 4 9.8 8.95m 5.685 $5.598 b082, 83.5w 28 88 oaeem .5558 «.3 oz... 121 meoeee Snood mwoooo oohmw. 8.3.. o... w . . o. o oeomod oooeo.e 8. .oo ooeoo. o. . . ..m o.o o. - o o Rom ..o oooeoe Nooooo Noon ... Noon... Q. o - o n eommmd 8.3m umRod eoeoe. oomood m.m o . m m 833. moon ..m 3236 3.8.. 083.. m.e m . e m oooeod meeoe.e «mound ooeeoo mmoom. md e . o e ooeooN oomeoe ehmooo oeum ... whoa. md o . N o mooemd e .ewee ewohmo m .n 56 oo.~.o.. m.. w .. . m .oone. o.omm.m m.mmm.o mm.~m.o mkooo mo . - o . m.ommmm. ow. .~.w 88 ..e mo .86 .mome. om Sod o.eo mo - no em 883.. mom. ..m om.om.o emomo. onmo ..m ooh on - b. no emeo ..m oeewoe mmeseo women... oooood own on . E. «N o .8 ..N oEoe co .om.o momma. . noeond coo no . mo .N .oeom. eoo .06 x. .mneo .oooo. oooneo o.oo .o - om ow 88o. «Rmmd .eemeo menmm. monoud o.em mm - mm o. 8% ... owomod om .oeo e .me.... no .RN ode oe . me o. .me.... moomod a .093 ooommd moemm. ode me . ... .. hoo.e.. oomeod nu .oeo moo.e.. mooeod 98 no . mm o. omen... one .o.o moaned ooooo. mReNN o.oo ..m - mm m. oeomm. eoe .m.o oneeed mooooo oomwwd o.e~ mm - ow e. ooooe. .owoo.o ooomeo o.~oo.o he .8. o.o. o. - s. o. .83.. Noewod omo.e.o oooeN. mmohod od. 9 . .. .. N . em .09. ohm. ..o oo.oe.o mnoooo mnoeo. mo. .. . o. .. momma... ..oooe o .ooeo oomooo oe Bod mo o. . o o. me {m.. eowohd Noomeo on .o ... muoned mo o . o o omnme. homo..." ooooeo eoomno .83.. m.. o . n o ..ommmm. 253 2.36 2.... Emma. mEEm medal Eu .8 mas—mm £92. an .. £92. 5; $23.. sue; to £92. a; :58 :82 .322... 58m... * 28 8.95m gfim 5.2.5580 H.953 856mm 28 835:8 N.N< 03m... 122 8.88 88.... .89... 8.8.. .888 m. .w 8 . .u m. m 58.. 8.8... 8.2.... 898.. 888 m8 .w . 8 m. .88.. ...e....» 88.... . .8 ... 8.38 m... 8 . o. .. meeoeu 38m... «$8... momma. 8.88 mm. a. . m. m. .888 884... 8.2.... 88... 8.88 m... a. . .. m. 8mm... 28»... 88m... 8.8... 88... mm. s. . m. e. 8.88 988... 88.... 88... 8.88 mm. m. . m. a. Boemd .28... 2.2m... .58.. 88.8 m..: m. . ... m. .83.. 883 8.8... 3.3.. 888.. mm. e. . a. .. 88m. 888... m3.m.o .88.. 823 me. n. . u. o. 8.3.... 0.83 8.8... m .m..... 84min o. .. u. . o. m 888.. «8...... 88m... 88.... mmmeoe. o8 o. . m m .88.. 8.2.8 .8. .m... a... .m.. :83 .... o . m . 88.. 2.83 m8.m.o 8.8... 88.... mm m . m m 883 8.8... 89.... 8.2.... mommm. m... m . v m .88.. 8.488 88.... 88.... 8.8... m.» e . m e 8.8... 88.... «mam... Emmmd 88o. m.~ o . u a mmmeo. .888 .823 8.5... 8mm ... m.. u . . m 388... «8&8 38.3 888... .88... m.o . . o . magmama. mas-m .8. ..m 88m... .mnm... mmmm ..m cm. on . ... m. 8.8.." 88 ..m 88m... 8.. ... 88 ..m 38 ..m . mm m. 8888 Swim 82m... .88.. ~88...» o. .v we - 3 ... 8%.... 88m... 8mm... 8mm... .8... ..m 98 8 . 8 m. ~88... t . ....m 88m... 8...... 8.8... o... m. . m. m. 5.8..” 8.8... mm...m.o 8mm... 8.88 cm. m. - ... .. .888 8......m 88m... m.o~m.. 883 3. e. . u. o. m.umoam. mESu 252m 0.5. Emma. mESLm «Ema Eu .8 macaw .3 an .. 292.. e; $25 2%; a... £92. a; :68 :82 .88.... .....om... .. 28 ..an5 5.695 5.2.5.28 3963 8358 28 booSEoo N..N< ohm. 123 6: EN Nm 566 66¢ 56 mm 5 5 mommNN mNm 3 - .6 N N566; «N356 6336 6686; ..nnvoN m6N 5 - 6N 6N EcNmN 6856 K896 56666 «when; 98 6N . mN a. .N EQN 6336 66696 335.? Nmmva oéN mN . MN 9 N pomnN 3866 3.686 Nohmvé NmwaN 6.NN 0N - N t. 6N FBN 6N 566 66 .96 N866... NvamN 66N PN . up or mvmoo. P 662.66 3636 6536.? SN» fiN 66? 6F . 2 mp mmNmmé 65666 66596 665666 6636.. 66.. t . 9 3 66654 3366 869.6 65a; magmN 6.3 9 - or up moomNN pNovoé 3636 366 fir mNmoNN ON? 9 - 3 NF 366mg 66 P666 65:6 99 F... 66mmvN m6.. 3 . or Z 3 was; 633.6 333.. R866 3N2; 66 or - 6 6F 86 K; NomNod 83.16 3636 mNmnmé m6 6 . 6 6 tommN N836 6 536 6 SP 2 PmmovN mg. 6 . x. 6 6668; v ENQN N 596 66666.? 8N66N m6 h . 6 n oNONoN 66 F 56 62.96 3 EN; ...KNEN m6 6 . m o 3NNN... 3 FN hm N886 3636 8666; me m .. c m 8666... macNoé 69.96 wow 56 mNNomé m6 c . n v 635... 8590 R696 6336 83m; m.N a . N m m E 36 3665.6 386.6 66356 £56 6; N - F N mmmaNé «NomoN N396 9516 N 31.66 66 P . 6 F 9332 movad 6N8?“ N836 PNN 5; opmomN 66v nv - mt «N 3666; SNNQN 699.6 869.6 mm FNNP 6.8 6v . no «N 6665... 63666 3.6 36 “Nov 2 NnmoNN 6.NN no . 5 N 6566... EbNod 3656 63mm: mammoN o.mN 6N - «N N 683; 36th 963.6 msoNvé mums F.» 66N eN . NN 6N maouco¢P wESm 953 0.8. 59103 2:26 «Ema Eu .8 295w E 5 .. 59m; 5; 525 £92; an 5%; a; :58 :8: .98.... EB. * 28 moEEmw congm 8.855580 bozo; 8.6an 200 8:36:00 N.N< 032 124 $8... .88... 8a.? 8.8.. «$3.« 98 3 . ..m .. .88.. «..«...» 283 .88.. .3«..« on? 8 . .... o. 293.. .88... «$93 .88.. ..«: ..« 98 8 . 3 m. 8...... «««o..« .836 8.3.. «.33 o.m« o« . ..« c. 3.8.. .«83 883 «83.. 83«.« 9m. 0. - v. «. v.89. «8. ..« .886 «.8... .53.. m... «. . .. «. 98.... «$8... $83. cg... ..«vo.« m..: .. - o. .. 88... 8««o.« «.83 85.... «85.. 3 o. . a c. 8... ... 33v.« .893 88c... 28.... m.» a . o a £83 «.«.«.« 8.23 83a... <88. m.. o . . a 8.8.. .o«« ..« 833 «..mmd «83.. no . - o . .«8... «3%... «836 88a... ..«... ..« 3 o - m o 83... «5.3 8««v.o 92.... .89... m... m . .. m «83.. 3«.«.« ««8...o $83. 9.8.. m.., .. . n v 38 ... 8mm...« .39... .o..m.o c.«.«.. m.« o . « a «.83 m«..... 80:8 8.2... 8o... m.. « . . « 3.33 m««...« .223 «2%... «8.«.. we . . o . «uncom. backs. mm«« ... 88¢.« $8.3 «8R... o. ..m... o.o m .o. 2.» =2“. m««8.. .38.« mm .93 .83... 88.... 9o .8. 2... .5... «3:3 .mov«.« .c..«.o «.83 m .m..; .3 .8. 2.» 5... .82... «.8. «8.... 8.v«.o 3.3.. .3 8m. 2.» .sa :8...» 8S5 «.83 88m. 883” no. 2. - 3. ..« v8.3 «5%... «88d «.59. .88.« m. ... 3. - on ..« mm«o..« «8.3 «$85 .88.. 883 m8 8 - 3 «« 8 «o 8. mega mEmfl 2.9 5% «E967 «Ea .5 Eu 295w E 5 .. 59°; 5; 82.... 5%: to £92. a; 58.. :82 .98.... ...SIP . 28 3.8mm Euafim 8.358580 5.53 835w 28 8:538 N.N< 63m. 125 8.8.. .6««..6 .«8.6 6.8.6.. 8.6 ..« 6.6. 8 - 6. .« 8.86.. 8.8.. 8.6.6 8.6.6 688.. 6.6. .. . «. 8 888.. 8.8.6 .886 8.8 ... «.888 6.8 8 . 8 «« o.m««.« 868.. ..66.6 6.8«.. .86..« 6. .8 8 . 8 .« 88..« 88... .6 . .66 8.6 ... .8««.« 6.8 8 . 8 8 8.8.. 888.6 88.6 .8 ... 68...« 6.6. 8 . 6. 6. 88... .888 6. 8.6 86666 88..« 66. .. . «. 6. 668... 868.6 88.6 .««.«.. 88.88 6.8 8 . 8 .. 868.. «6 .666 6886 88.6... «86 ..6 6. 8 8 . 8 o. 88... 6. .8.« 8.86 .. . . ... .8...« 6.8 8 . .« 6. 6.8... 88..« 683.6 8.86 8.6... 6.6. 8 . 6. .. .8... .. .86 .68.6 66866 «6.8.. 6.6. .. . «. 6. 868.. ..8..« «.686 ..686 66.8.. m. .. «. . .. «. 8.8.. 8.6.«8 «8.6 ...86 688.. 6.6. .. . 6. .. .8.6.. o... ... 88.6 8866 .88.. 8.6 6. . 6 6. .88.. 6.6.«6 88.6 .886 .8... m6 6 . 6 6 8.8.. 888.6 6.66.6 8.686 .88.. m.. 6 . . 6 8.8.. 8.86 88.6 «.866 86.6.. 8.6 . . o . 8.8.. .886 88.6 ..866 2.8.. .8 o . m o 8.8.... 8866 88.6 6.8.6 8« ... m.. m - . m 6668.. .888 ...86 8.8.66 888.. m6 . . a . 8.8.. .8..6.« 38.6 .6 .86 .886 m.« 6 . « 6 886.6 66888 6886 8....6 668... m.. « - . « .886 .. .6 ..« 6.8.6 «.8.6 .88 ... m6 . . 6 . 888. 886.. 88..« 8.86 688.. 88 ..« 6... 6. . o. 8 83.6.. 8.8 ..m 8.86 8..«.. 6.86.« 6.8. m. - .. 6. 8.86.« 6886 8.86 8.66.. .88.. 6.8 8 . .6 6. 8:86 .658“ 6:6. Ea; 858m 8% Eu .8 99:3 E .6 .. £68; .6; 82.... 56.63 an 56.6; .6; 568 :8: .88.... 58W. . 28 8.658 £685 5.2.5.266 .6683 .2688 28 b03538 N.N< 2cm. 126 .668. 8.6.6 «666.6 .6....6 «.666. 6.. « . . « ..666. 666.... 866.6 8.666 .6666. 6.6 . . 6 . 86686. ..«.6. 8686 68.6.6 .68«.. .6686 6.6. 6. . .. .« 6...6.« 886.6 .6866 .66. ... 88..« 6.8 66 - .6 6« 66666.« 86.66 68.66 6686. . .866. 6. .6 8 . 8 «« .88.. ..«...« .6866 . .68.. 8.86 6.6 8 . 66 .« 6.8... 68666 6.866 .66«... 8.666 6... 6. . 6. 8 66666.. 86666 8. .66 66666.. 66866 6.6. .. - 66 6. 666 .6. 86666 866.6 .68«.. 66666.« 6.66 .6 . «6 6. .86... 6.86.6 666666 6.63.. 8858 6.3 R . 6N .. 8.66.. 8.666 .66...6 6666... «.6...« 6.6. 8 - 6. 6. 8.66.. .«6666 66.86 6666... 6.8.6 6... 6. - 6. 6. ....6. 8.6.6 «8.6 6.6.«.. ..«66.« 6.6. 6. - .. .. 6.6.6.. 88«... 66.8.6 88«.. 66«6«.« 66. .. . «. 6. .86... 688.6 «6.66.6 «8.66 6.6.6.. 6. .. «. - .. «. .66.... 88.6 .886 66666.6 66.66.. 6.6. .. . 6. .. «...6. 8. .6.« 6.6.6.6 66666.6 6.86.. 6.6 6. . 6 6. 6666... 6.«...« .6 .666 666 .66 ...8. 6.6 6 . 6 6 .8666 .6686 66666.6 886.. 88..« 6.. 6 - . 6 .86... 88..« .. .666 6666... 88..« 6.6 . - 6 . 6. . .6. 8.86 66666.6 6686.. .6.66.« 6.6 6 . 6 6 66.6..« 88«... .6686 .88.. 6686.. 6.. 6 . . 6 6666... 66.86 66.86 66666.. 66666.« 6.6 . - 6 . 66 .66.. 666.66 ...666 . .68.. 6. .6«.« 6.« 6 . « 6 6.8... 86666 6.6.6.6 .686. 886.. 6.. « .. « «.686 866... 6.8.6 ..«666 8. .6. 6.6 . - 6 . ..66666. 6666 6566 28. Emma 68%| 686.5 .8 Eu 6386.6 .3 an .. £6.62 .6; .62.... £68; an 56.6; .6; £666 68.2 .68: camel . 28 8.6.86 8.8.5 686558.66 ..e.... 8.8.66 28 832:8 N.N< 2cm... 127 668 ... ..8..« 66....6 66.86 6.86.. 6.6 . - 6 . 8666.. 686«.« .«.6.6 688.6 «668.. 6.« 6 . « 6 .8666 «386 88.6 «...66 666.6. 6.. « . . « 6.6666 66«66.« 6.6...6 6.6666 88«.. 6.6 . . 6 . 60066 6.666.« 8.86 .66.66 6.68.. 68.66 6.6. 6. - .. .« ..«66. .6666.« 68666 6.66... . .8 ..« 6.8 66 - .6 6« 88..« .6.66.. 6.66.6 .6.66.. 6.86.« 6. .6 8 . 8 «« .866. 88.6 .«.«66 8.666 6 .86. . 6.6 8 . 66 .« .666... .6 .86 «6.6..6 .6.66.. «68 ..« 6... 6. . 6. 8 .. .6 ..« 8.8.. 66866 6.6.6.6 «.6... 6.6. .. . 66 6. 668 ..« 6.68.. .886 .86... 6.66.6 6.66 .6. «6 6. 88... 68.66 ..«6.6 «.«.66 8.66.. 6.8 .« . 6« .. .«68.« 6.686 «..6.6 688.. 666.66 6.6. 8 - 6. 6. 868.. 88.6 .6 .6.6 886 «.«8. 6... 6. . 6. 6. 6.6.... 86666 86.6 8.666 .6666. . 6.6. 6. . .. .. 6 .6... . 8.666 . .8.6 6.6666 886. . 6.6. .. . «. 6. 86.6.. ..866 6.8.6 6.86.6 8666.. 6. .. «. . .. «. ..6.6.« 886.6 ..866 .6.66.. 6.«6..« 6.6. .. . 6. .. 6.6.... ...6.6 «6....6 886.. 8..6.« 6.6 6. . 6 6. 86.... .886 6.6.6.6 8.8.. 6.666.« 6.6 6 . 6 6 ..«66. «6 .«.6 .6686 6666.6 .66.... 6.. 6 . . 6 .6.66.. 86.«6 88.6 6.866 688. . 6.6 . . 6 . 86.6. . 886.6 . .8.6 668.6 .«.66.. 6.6 6 . 6 6 66 .666 6. .66.. 86.66 6.6666 6..66.. 6.. 6 . . 6 886.. 88.6 6.. .66 8.86 86.... 6.6 . . 6 . 6.666.. ..«666 8....6 6. ..66 6.86.« 6.« 6 - « 6 ««o6666. 668$ 8.83 9.6. Ema; 656a 66.651 .8 E6 29:66 E a. .. 56.6; .6: .62... £65; a. £6.63 .6; £68 :82 .68.... £68! . 28 8.6866 8.885 6865.528 .582. 8.6566 28 boa—Eta N.N< 036.. 128 66.6.. . .6.66.6 6.6666 66.666 . .6..6 6.6 6 .6. 666 666.... .666..6 ..6...6 .666..6 66.66.. 6.6 .66. 666 .6 .6..6 .6666.6 666.6.6 6 .. .66 66666. . 6.6 .66. 2.6 66666.6 6.6.6.. 6666.6 .666..6 66..6.6 6.6 666. 666 626.666 466 666.6.6 66.666 .6866 666.6. .66666 6.66 .6 . .6 66 66666. . 6666.6 66666.6 666.6. . .6.666 6.66 .6 . .6 .6 66. .66 666.... 66.6.6 6666... ..66..6 6.66 .6 . 66 66 6.6.6.. ...666 6666.6 6666... .6..6.6 6.66 66 . 66 6. 6666... .6..6.6 666.6.6 6... ... 6666.6 6.6 66 . 66 6. 6.66... 6. ..66 6666.6 .6.66.. 66..6.6 6.66 66 . .6 .. 66666.. 666.6.6 66.6..6 .6666. 66.666 6.66 .6 . 6. 6. 6.66... 66666.6 6666.6 6.66... 6.6666 6.6. 6. - .. 6. .666... 666 .66 66 .6..6 66666.. 6.6666 6.6. .. - 6. .. 66666.. 66.6..6 .666..6 6..66.. 66666.6 6... 6. . 6. 6. 6...6.. 666.6.6 6..6..6 666.6. 666.6.6 6.6. 6. . .. 6. 36.6.. 6. .6 ..6 06 .6..6 666. ... 3.6666 6.6. .. . o. .. 6666.6 .66.6.. 6.66.6 6666... .6..6.6 6.6 6. . 6 6. 666.6. . 6.6666 6.6...6 .6666. . 66666.6 6.6 6 - 6 6 6.66 ... 6.6666 66.6..6 .6666. . .66 .66 6.. 6 . . 6 .666... ....66 6.66.6 .6666. 6.6.6.6 6.6 . . 6 . .6666. . .6666.6 6666.6 .666 .. . 66666.6 6.6 6 . 6 6 .6.666 66..6.. ..66..6 6..66.. .666..6 6.. 6 . . 6 60066 66.6.6 6.66.6 6.66. .66.6.. 6.66661 66.6.6 .66 .66 6.66666 5 .66 a 66.6; .6; .6266 6.6.6; 66 66.6; .6; .6666 68.2 .6626. 6.666.. . 28 626.666 62886.6 62.66.6688 .6983 66.9666 860 3:538 N.N< 036. APPENDIX 3 ANALYTICAL DATA Results of chemical analysw of arsenic extraction solutions are presented in Table A3-l. Corresponding sediment concentrations are also presented. These were calculated from solution concentrations and dry/wet sediment ratios (see Appendix 2) according to the following formula: solution concentration (pg/L) x solution volume (L) sediment concentration (jag/g) = wet sample mass (g) x dry/wet mass ratio Total extractable arsenic is calculated as the sum of the sediment concentrations for all of the extracted phases. Data appears in core number order, with SBL data at the end. Results of chemical analysm of mercury extraction solutions are prwented in Table A3-2. Sedimentconcentrations weredetemtinedinthesamemannerasforarsenicconcentrations. Data appears in core number order, with SBL data at the end. Results of porewater analyses are reported in Table A3-3. These include arsenic and mercury concentrations, ferrous iron concentrations, and valuw of pH and alkalinity. Data appears in site order, with all data for a given site on the same page. OrganiccarboncontentsofsedimemsamplcsarercportedinTableM-4. Dataisprwentedin site order. 129 130 66.6 8.. ..6. «66.. 6.6. ..66 6.66 666.6 6.6. 6.6 .66 6.. 6 .6.. 66.6 ..8 866 6.6. 8.6 «.6 8«.6 ..«6 66 66 6.6 . 8.. 6.6 ..6. 666.6 6.6. .86 6.6 8.6 666 6.6 6... 6.6 6 .6.. .66 ..66 666.6 6..« 68.6 6.6 8.6 6.6. 66 ...6 6.. 6 6... .66 6..« 6.6.6 ..66 86.6 ..6. .666 6.8 6.6 .66 6.6 . 6.6 66.. 6.66 .6... ..8 ..66 ..66 .6..6 6. .. .6... 6.6 6.« 6 66.6 6.6 ..66. 6.6.6 «.6.. 6.6.6 6.6. 66.6 6.8 6.6 6... 6.. « 6.... 6... 6.8. 6.6.6 6.8 .666 ..« .86 6.8 .86 6.6 6.6 . ..8666. «..« 86 6.6 666.. 6.66 8.6 ..6. 6.«.6 ...6 6.6 66 6.66 6. 6..« 66.6 6.66 .... ..8 68.6 ..6 666.6 6.66 6... 66 6.66 .. 66.6 .... «.6. «66.. 6.«. .666 6.6. 6.«.6 6.«« 6.6 66 6.6. 6. .66 6.6 «.6.. .66.. «.6. .6..6 ...« .6..6 6.66 6.6 66 6.8 6. .«.6 8.6 6.66 .8. «.«6 6.6.. ...6 86.6 «. .6 6.6 6... 6. .6 .. .66 8.. .. .6 68.. 6.6. 66.6 6. .6 «8.6 ...6 6.6 6.6 6..« 6. .... 6... «.6. 8.6 6.8 6.6.. ..6. 6.6.6 6.6« 6... 6... 6... «. 66.6 «6.. 6.8 6666 ...6 686 «.6. 6.66 «66 6.6 6... 6... .. .6.. .... «.66 6.6.6 6.66 .«6. ..6. 86.6 6.8 6.6 6.6 6.6. 6. .66 «..« 6.6 86.6 6.8 «...« 6.8 686 6.6 66 66 6.6 6 6..« ...6 «..« 66.6 6.66 «...6 6.6. .««.6 ..6. 6... 6.6 6.6 6 8.6 66.6 6.6. «66.6 «.66 «66.6 6. .« 866 .6« 6.6 6... 6.. . .6.« 86 6.66 ....6 6.8 86.6 6.6. 6.6.6 ..8 ...6 66 6.6 6 66.6 66.6 6.6 6.6.6 6.«. ..66 «.6. 666.6 6.8 6.6 66 6.6 6 8.6 .6.. ..6. .6... «.6. 666.. 6.6. 666.6 6.6. ..66 6... 6.. . 66.6 .6.« 6... 68.. 6.6.. 666.. 6... .6... «.66 .«.6 6.6 6.6 6 8.6. 66.6 6..«. 666.. 6.66. «66.6 6.6. 666.6 6.8 66.6 8.6 6.. « 66.«« 8.6 6.«6. 86.6 6666 6.6.6 6.8 .66.. 6... «8.6 8.6 6.6 . 66686. in 6.6% 666le? 661%! 66671666166... .8 6.6.68 2 .25. 86 x0 .68 x0 68 6.2 .68 6.2 86 66 .68 66 .8 6<>> .666 6<>> .8 x6 .68 8 6.86 6. 28 6686.66 .66 66.6666. 6.6 .6.66.. 66.6; ...6 66586 s 6.66 5 86.66868 628.6... 86 .666 66666.66 6.66666. 6_ 668866 6666666668 66.66.66. 6.6m Sun 2:06.; 036.06....m TG< 0.96... 131 00.0 05.. NN. . 0... 50.0 05. 0 ¢0. . N5. 0 .0. 0 55. N 00N ¢00 0¢.5 0¢.N ¢0.N 00.N ¢¢N 00.N ¢00 .5. N 55. 0 0.. 0 00. N 05. ¢ 0¢. ¢ 00.0 00. ¢ 00. 0 N0. N 0.0.. 6< ..<..0.. 066 x0 ....66 x0 066 m2 ...6... m2 066 mm. ...6.6 mm 066 0<>> 6.66 05$ 066 xw ...—66 xm ...an ¢.0 0..0 0. .0 00. 0 0.. 0 0..0 0.. 0 00. 0 0.. 0 ¢0. 0 00.0 00.0 00.0 0... 0... 00.0 0... 00.0 00.. 50.0 00.. 50. N 00. 0 ¢N. N 05.. 50.N 00N 50.. 0... 0.0.. N. .N 0. 0. 0. 0N 0.5 5.00 0... 0.0. 5. N. 0. 0N 0.¢ 5.5 ..0 0.00 0.50 0.0 . . 0.05 5. 00. 0. 0¢ 0. 50. 00¢ 5.00 ¢. .0. ¢.N0 0.00. ¢00. ..0N . 0.00. 0.¢¢. 0.50 000.0 N000 50¢.0 0N¢. 0 00N. 0 000.0 00¢.0 .N0. 0 00¢. 0 ¢0N.. ¢¢0.. N000 N000 0N0. 0 000. 0 055.0 .000 005. 0 00N. . ¢500 0.... 000.N 5N0. 0 00¢ . 0¢.. . 000. 0 0.0. . ¢00. . .00. 0 5. .5 0. 00 0.00 0. N0 ¢. .0 ¢00 5.00 0. 00 5. 00 0. 00. N. .0N 0.00 0.03. 0..0 0. ¢.. ¢. 00 N. .0 0..0 0.0.. 0.00 5.05 ¢.¢¢. 0. 00 0. 00. 0. 00 ¢. 50 0. N0. 0. 00 0.00 0.0. 0 000. 0 00¢.0 N0. .0 55.0 ¢N00 55.0 500.0 .¢0.. 5N0. 0 000. 0 ¢.0. 0 000.0 500.0 0.0.0 0N0.0 ¢0¢0 0¢0. 0 000. 0 0¢0. 0 000.. 005.0 5.0.0 000. . 0¢0. . N000 000.0 N0¢0 0.0.. 5.5¢ 0.0 . . N. .5 0.N0 0.5¢ 0. .N 00¢ 0. 0N 0.0. . .00. .. 00 ..0 0.0. 5.0N 5.50 0.00 ..0N 0. 0N 5. 00 0.0N 5.5¢ 0. .0 ¢. 0¢ 0. 00 0. .0 0. 00 0.¢0 0.00 0.0N ...6: 00.. 0 00N. 0 50. .0 0N..0 ¢N..0 NN.0 00.0 5N. .0 000. 0 .0. .0 ¢0.0 0N0.0 .000 .0N0 0N..0 .0..0 0¢. .0 00N. 0 0NNO .0N0 ¢0N. 0 00N. 0 000.0 00¢. 0 ¢0¢. 0 00N. 0 00N0 ¢¢00 0.0.. 06.38. ..0< 6365 00¢ 5. 50 .. ¢0 0.00 ..00 ..50 0. .¢ 0.N¢ N. .¢ 0. 00 0. 0¢ 0. 5 ¢. 0 . ¢.¢¢ 0. .0 N.NN 0. .0 0.N0 ¢.50 0.0N N0¢ 0.N¢ 50¢ 0.N5 0. .0 ..00 0. .¢ 0.¢¢ 00¢ .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... 000.0 000.0 .0... 000.0 .0... 500.0 .0... .0... .0... .06 6.66 6. .06 6.5. 6. .06 6.6. .. .06 6.6. 6. .06 6... 6 .06 6.6 6 .06 6.5 5 .06 6.6 6 .06 6.6 6 .06 6.6 v .06 6.6 6 .06 , 6.. 6 .06 6.6 . 6 .66666. .06 6.66 66 .06 6.65 66 .06 6.6... 66 .06 6.66 .6 .06 6.66 66 .06 6.66 6. .06 6.6.. 6. 6.6. 6.6.. 5. 6.6. 6.66 6. .06 6.66 6. 66.6. 6.66 v. .06 6.6. 6. 66... 6.6. 6. .06 6.6. .. .06 6.6 6. .06 6.6 6 . 36666. 6.6 6.6.666 66.50 132 00. 0 00. N .N. . 00.0 N0.N 05.0 05.0 ¢5. 0 00. 0 00.0 00.0 50. 0 00. 0 00. 0 00. 0 00. 0 50. 0 0¢. 0 00.0 0¢0 00.0 0... 50.. N0.N 00.. 0N.. .0.. .r I 6< ._<...O._. 066 x0 6.66 x0 066 m2 6.66 m2 066 mm ....66 mm 066 0<>> ...—66 0<>> 066 xm 6.66 xm .0060 N .0 00.0 N00 .N0 N00 .N0 ¢.0 0..0 N.0 00.0 ...0 00.0 50.0 0..0 ¢.0 ¢.0 0..0 0..0 0..0 0..0 0..0 0N0 0¢0 ¢50 0..0 0..0 ¢.0 ) i 0.0. 5. 0. 0. 0N 0. 0. 0. .N ..0. 0. 0. 0. 5 5... 0. 0 ..0 0.5 0.5 0.5 0.0 ..0 0.5 ¢.0 5.5 N0. .. .. 5.0. 0.0N ..0N ¢.NN 0. .N N0. 0¢..0 .5.. . ¢00. 0 ¢00. 0 005.. 500.0 050.0 5¢¢0 .5.0 .0.0 0000 0.N0 00.0 ¢0N .0 0.0.0 N000 00N0 00N0 00N0 0NNO N000 5N00 ¢000 ¢0N.. 00¢.0 00¢.0 000.0 I ). 5.0N 0.05 0.¢N 0.¢0 0.N¢. 00¢ ¢.¢0 0.00 0. .N 0.0N 0.¢0 0.0N 0.0N 0.NN 0.0N ..NN ..0N 0.0. 5.0. 0.NN 0.0N ¢.0N ..5¢ 0.00 0.¢5 ..N0 0.N5 00N .0 0N0.0 0 .¢0 N0 . .0 ¢0N .0 050.0 00.0 0. .0 000.0 3.00 .000 0.00 000.0 ¢000 000.0 N500 000.0 .000 000.0 000.0 .000 05.0 .000 50.0 00N0 ¢0¢0 0¢¢0 ) | .. .0 0. 5N ¢. 0N 0.5 5. 0. .. 5 0... 0. 5 0.0 0.0 0.0 ..¢ ..0 0.0 N.¢ N.¢ 0.N 0.N 0.0 ¢.N N.0 5.0 0. 0. 0. 5 00¢ .. .0 0.5¢ .000 N00 .0 ¢0 . .0 000.0 000.0 000.0 000.0 ¢000 000.0 .000 000.0 0No0 0Nc0 0No0 0No0 0N0.0 5.0.0 0N0.0 0N0.0 0.0.0 500.0 000.0 000.0 0. .0 N0. .0 N0 . .0 00.0 .... «00:8. ..0< 6.065 ¢.0 ¢. 0¢ 0. 0N 0.0 0.0 N. . . 0. 0. 0. 0 0.0 0.0 0.0 0.0 N0 0.0 0.0 0.0 N0 ..¢ ¢.¢ 0.0 0.0 5.0 0.0 0.0. 0.N0 ¢.0¢ 00¢ ) .0... .0... .0... .0... .0... .0... .0... .0... .0... .0.: .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... .0... 66 6.6. v6 66 6.66 66 66 6.66 66 66 6.66 .6 66 6.66 66 66 6. .6 6. 66 6. 66 6. 66 6.6. .. 66 6.6. 6. 66 6... 6. 66 6.6. v. 66 6.6. 6. 66 66. 6. 66 6.6. .. 66 6.6. 6. 66 6. .. 6 66 6.6 6 66 6... 6 66 6.6 6 66 6... 6 66 6.6 v 66 6.6 6 66 6.. 6 66 6.6 . 6666666. 66 6.6.. 6. 66 6.66 6. 66 6. .6 6. 6.66666. .1 ...6 6266 6 6.66 133 666 66.6 6..... 6.6.. 666 666.6 6. 6. 666.6 6.66 6.66 66.6 6.6 6 .6. 2.6 .6... 66... 66.6. 66.6 .666 66.6 6. . 6.6.6 ..66 .666 6666 6.6 .66. 2.6 66. .. 66.6 6.666 666... 6.666 666.6 6. .. 666.6 6.66 666.6 66.66 6.6 .66. 2.6 6.6. 6.6 6. .6. 666... 6.66. 6.6.6 6.6. .666 ..66 66.6 66.6. 6.6 666. 2.6 8.66.66 ..66 66.6 6.. 6 6. .6 666.6 6.3 «66.6 ..6. 66.6 6.66 666.6 66.... 6.6. «6 66.6 6. .6 6.6. .666 6.66 .6..6 6. .. 6.6 6.66 .66. 6 66.66 6. .6 66 6.6 v. .6 6.6. 666.6 6.66 6.. .6 «.6. 6:6 6 .... 66. 6 6.6.. 6. 66 66 66.6 6.6 6.6. 666.6 ...66 666. 6 ..6. 6... .6 6. 66 66. .6 66.66 6.66 .6 6.6 6.6 6.6. 66.6 6.6. .... .6 6.6. .66. 6 6 66 666.6 66.66 6.66 66 66.6 6.6 ..6. 666.6 6.66 ...6. 6 6.. 66.6 6.6.. ...6 6.. 3. 6.66 6. 66.6 6.6 6.6. «...6 6.66 «66. 6 6.6 666.6 6. .6 6.6. 6 66. ..6 6. .6 6. ...6 6.6 6.6. 666.6 3.6 66. .6 ..6 666.6 6.66 666. 6 66. 6.. 6. 66 .. 66.6 6.6 6. .6 66.6 .66 66.6 6.66 .666 6.66 6.6. 6 3. .. 6.66 6. 6.6 6.6 6.6. 66.6 6.6. 6. .6 6.6 6.6.6 6.66 666. 6 6..66 6.6. 6. 6.6 6.6 6.6. 6.6 6.6. 63.6 6.6. 6.6.6 ..66 66. .6 66.66 6.6. ... 66.6 6.6 6.6. 66.6 6.6. 6. .6 6.. .666 666 .666 66.66 6.... 6. 66.6 ...6 ..6. .6..6 ...66 6. ..6 ...6. 66.6 66. 66.6 66.66 6.6. 6. «6.6 6.6 6.6. 6.6.6 ..66 666. 6 6.6 66.6 6.66 666.6 66. .6 6.6. .. 66.6 6.6 6.6. .6..6 6.6. 666. 6 ..¢ 66.6 6.66 666.6 66.6. 6.6 6. 66.6 ...6 6. .. 366 6... .66. 6 6.6 66.6 6. .6 666.6 8.. 6.6 6 .66 6.6 ..66 .666 6.66 666.6 6.6 6. .6 6.66 36.6 66.. 6... 6 6.6 ...6 6.6 .666 66. 66.6 ..6 6.6.6 6.6 6.6.6 66.6 6.6 . ..66 6.6 6.6. 6.6.6 6.6 6.6.6 6.6 .666 6.66 666.6 66.6. 6.6 6 66.. ...6 6.66 666.6 .6.. 66.6 6... 666.6 6.66 66.6 66.6. 6... 6 .6.. ..66 6.... 66.6 6.66 .6..6 «.66 36.6 ..66 :66 66.6. 6.6 6 66.. 66.6 6.6. 66.6 6.66 666.6 6. .6 6.6.6 ..66 6.6.6 66. .. 6.6 6 66.6 ...6 ...... .666 6.3 3.6 ...6. 6. ..6 6.66 666.6 66.6 6.. 6 66.. 66.6 6.6. 66.6 6..... 666.6 6.6. 666.6 ..66 6. .6 66.6. 6.6 . 666666. mam mam mam 00.. ”..m: 00.. ...0: a: a. j: ...6 6.6.666 6.. .25. 666 x0 6.66 x6 666 6s. 6.66 6.2 666 66 6.66 66 66... 65> 6.66 65> .66 x6 566 .6 £666 6 28 .0668. .6.. 6.66. 134 6.66 66.3 66.6 .66. 66.. 66... ...6 6.6 6 66.66 ..6.66 6... 66.3. 63 66.6. 66.6 6.. 6 .6..6 6666 66.. 66.6. .6.. 666. ..66 6.6 . 838. 6.6. 6.6. 6.6 66.6 66.6 66 66 6.... 66 66.66 66.6 66.. 66 66 66 66 6.6. 6. ...6 ...6 66.6 66 66 66 66 6.66 6. .66. .66. 666 66 66 66 66 6.66 .. 66.6. 66... 6... 66.6 6.6 66.. ...6 6.6.. 6. «6.66 66.66 66.6 .6.. 66.6 66 66 6.66 6. 6.... 66.6 66.6 66.6 ..66 66 66 6.66 .. 66. .. 666. .66 66.6 66.6 66.6 66.6 6.6. 6. 66.6 66... .66 ...6 ...6 6.6 66.6 6. .. 6. 66.. 66.. 66.6 66.. 66.6 .66 66.6 6.6. .. 66.6 66.6 66.6 66.6 6.6 66 66 6.6 6. 66.6 .66 6.6 .66 66.6 66 66 6.6 6 66... 66.6 6.6 .6.. 666 66 66 6.. 6 R66 6666 66.. 3... 66.6 66 66 6.6 .. 66.66 66.66 66.6 8.. 6.6 .... ...6 6.6 6 66.6. 3.66 .66 6.6. 6... 66.. 66.6 6.. 6 6...6.. 66.66 66.6 6.6. 6... .66 66.6 6.6 6 8.66 66.66 66.6 66.66 .66 6.. .. 66.6 6.6 6 66.66. 8.66 66.. 66.66 66.6 66.6.. ...6 6.. 6 66.6.. 66. .6 .6.. .6.: 66... ...6. 66.6 6.6 . 666666. Ijlufilllualjn a a ma .8 6.6.666 6: 2.6,. 666 x0 6.8 xo 666 6.. 6.8 6< .66 66 6.8 66 £666 .. 266 66.66.66 .66 66.66.66. 66 .6.66.. 6.6.6; .60. 665.666 6. 6.6.. 6. 62.6.6868 66.66.66. 666 ...66 68.5.66 6.66662 6. 66.8.60 66.66.6868 66.80... 6.66 6.66 .6626: 6.668866 6.6.. 6.66.. 135 66.66 66.6 66.6 66.6 66.6 8.6. 66.6 6.6 6 .6.66 6.6. 66.6 66.6. 66.. 66.66 6.6 6.6 6 66.66 .66. .6.. 66.66 66.6 66.66 66.6 6.. 6 66.66. 66.6. 66.6 66.66 66.6 66.6.. 66.6 6.6 . 6.66666. 66.6 .66 66.6 6.6 66.6 66 66 6.66 66 66.6 66.6 66.6 66.6 .66 66 66 6.66 66 66.6 66.6 66.6 66.6 66.6 66 66 6.66 66 66.6 66.6 66.6 66.6 66.6 66 66 6.6 .6 66.6 66.6 66.6 66.6 .66 66 66 6.66 66 66.6. 66.6. 66.6 6.6 66.6 66 66 6.66 6. 66.6. 66.6. 66.6 66.6 .66 66 66 6.66 6. 66.6. 66.6. 66.6 66.6 .66 66 66 6.66 6. 66.6. 66.6. 66.6 66.6 .66 66 66 6. .6 6. 66.6. 66.6. 66.6 66.6 66.6 66 66 6.66 6. 66.66 66. .6 66.6 66.6 66.6 66 66 6.6. 6. 66.66 66.66 66.. 66.6 .66 66 66 6.6. 6. 6.6. 66.6. 66.6 66.6 66.6 66 66 6... 6. 66.6. 66.6. 66.6 6.6 66.6 66 66 6.6. .. 66.6. 66.6. 66.6 66.6 .66 66 66 6.6 6. .66. 66.6. 66.6 66.6 .66 66 66 6.6 6 66.66 66.66 66.. 66.6 .66 66 66 6.6 6 66.66 .6.66 66.6 66.6 66.6 66 66 6.6 6 6.66. 66.66. 66.6 66.6 66.6 66 66 6.6 6 66.66 6..66 66.6 66.6 66.6 66 66 6.6 6 66.66 66.66 .66 66.6 66.6 66 66 6.6 6 666666. Jam“ jam: man my mm: mm: em: .5 29:66 6: .260. 666 xo .68 x0 666 6< 6..66 w< 666 66 £66 66 £666 6 28 66.686 6.6< 6.666 136 .66. 66.6. 66.. 66.6 66.6 66 66 6.6 6 66.66 66.6. 6... 6.6 66.6 66.6 66.6 6.6 6 66.66 66.6. 66.. 66.6 66.. 66.6 66.6 6.6 6 66.66 66.66 66.. 66.6 66.6 6.6 66.6 6.. 6 66. .6 6..66 66.6 66.6. 66.6 66.66 66.6 6.6 . 6666666. 86. 66... 66.6 66.6 .66 66 66 6.66 66 66.6 66.6 66.6 66 66 66 66 6.66 66 66.6 66.6 66.6 66 66 66 66 6. .6 66 66.6 .66 66.6 66.6 66.6 66 66 6.66 .6 6.6 66.6 66.6 66.6 66.6 66 66 6.66 66 66.6. 66.6. .66 66.6 .66 66 66 6.66 6. 66.6 66.6 66.6 66.6 66.6 66 66 6.66 6. 66.6 66.6 66.6 66 66 66 66 6.66 6. 66.6 66.6 66.6 .66 66.6 66 66 6.6. 6. 66... 66.6. 66.6 66.6 6.6 66 66 6.6. 6. 66.6. 66.6. 66.6 66.6 66.6 66 66 6.6. 6. 66.6. 66.6. 8.. 66.6 66.6 66 66 6.6. 6. 66.6 66.6 66.6 6... 6.6 66 66 6... 6. 66.6. 66.6 66.6 6... 6.6 66 66 6.6. .. .66 66.6 66.6 6.6 66.6 66 66 6.6 6. 6.6 66.6 66.6 .66 66.6 66 66 6.6 6 66.6 66.6 66.6 66.6 .66 66 66 6.6 6 66.6 66.6 66.6 66.6 .66 66 66 6.6 6 66.6 66.6 66.6 66.6 66.6 66.6 66.6 6.6 6 66.6 66.6 66.6 66.6 66.6 .6.. 66.6 6.6 6 6.66666. .Iqulallflmlllmfll: .664 .8 6.6.666 6: .2606 666 x0 6..66 xo 666 6< 6.66 66 666 66 6..66 66 .6666 6 28 66.686 6.6< 6.666 137 66. .6 .66. 66.6 66.6 66.6 66 66 6.6 6 6..66 6...6 66.6 .66 6.6 66 66 6.6 6 66.66 66.6. 66.6 66.6 66.6 66 66 6.6 6 6..66 66.6. 66.6 66.6 66.6 66.6 66.6 6.6 6 66. .6 66.6. 66.6 66.6 66.6 66.6 6.6 6.. 6 66. .6 66.66 66.. 66.6. 66.. 66.6. 6... 6.6 . 666666. 66.66 66.66 66.6 66.6 6.6 66 66 6.66 66 66.6. 66.6. 66.6 66 66 66 66 6.66 66 66... 66... 6... 66 66 66 6... 6.66 66 66.6. 66.6. 66.. 66.6 66.6 66 66 6.66 .6 6.6. .66. 6... 66.6 66.6 66 66 6.66 66 66.6. 66.6. 6... 66.6 66.6 66 66 6. .6 6. 6.6. 6.6. .6.. 66 66 66 66 6.66 6. 66.6. 66.6. 6... 66 66 66 66 6.6. 6. 66.6. 66.6. 66.6 66.6 .6.6 66 66 6.6. 6. 66.66 66.66 66.. 66 66 66 66 6.6. 6. 66.66 66.66 66.. 66 66 66 66 6.6. 6. 66.66 66.66 66.6 .66 66.6 66 66 6.6. 6. 66.6. 6.6. 66.. 66.6 .66 66 66 6.6. 6. 6.6. 66.6. 66.. 6.6 66.6 66 66 6.6. .. 6.6. 6.6. 66.. 66 66 66 66 6.6. 6. 66.6. 66.6. 66.. 66.6 66.6 66 66 6. .. 6 66.66 66.66 66.. 66.6 66.6 66 66 6.6 6 66.66 66.6. 66.. .66 .6.. 66 66 6.6 6 66.66 66.6. 66.. 6.6 6... 66 66 6.6 6 6666666. Ilmwflllllmmmlj 666 666 $6 .66 6.6.666 6: .2606 666 x0 6..66 xo 666 6< 6..66 6< 666 66 £66 66 .6666 6 6.8 66.68. 6.66. 6366 138 66.66 6..66 66.. 66.6. 66.. 6..6. 6... 6.6 6.6. 2.6 .6.66 66.6. 66.6 66.66 66.6 66.6. .... 6.6 666. 666 66.6.. .6.66 8.. 66.66 66.6 66.66 66.. 6.6 .66. 6.6 6..66 66.66 66.6 66. .6 6.6 66.6 66.6 6.6 666. 2.6 66.6566 .56 .66. .66. 66.. 66 66 66 66 6.66 66 6..6. 6..6. 66.6 66 66 66 66 6.66 .6 6.6 6.6 66.6 66 66 66 66 6.66 66 .66. 6..6. 66.. 66.6 6.6 66 66 6.66 6. 66.6. 66.6. 66.. 66.6 6.6 66 66 6.66 6. 66.6. .66. 66.. 6.6 66.6 66 66 6.66 6. 66. .6 66.66 66.. 66.6 66.6 66 66 6.66 6. 66.6. 66.6. 6... 66.6 6.6 66 66 6.6. 6. 66.6. 66.6. 66.6 66.. 6.6 66 66 6.6. 6. 66.6. 6..6. 66.. 66.6 66.6 66 66 6.6. 6. 66.6. 66.6. 66.. 6... 6.6 66 66 6.6. 6. 66.6. 6..6. 66.6 66.. 6.6 66 66 6.6. .. 66.66 66.6. 66.6 66.. 6.6 66 66 6.6 6. 66.66 .66. .66 .6.. 6.6 66 66 6.6 6 66.6. 66.6. 66.6 66.6 66.6 66 66 6.6 6 66.6. 66.6. 66.. 66.6 66.6 66 66 6.6 6 666666. :Iflnlllmwfluj 66.. ...6 6.6.666 6: .66.6.. 666 x6 6..66 x6 666 6< 6..66 6... 666 66 6..66 66 .6666 6 6.66 66.68. 66.6 6.666 139 66. 66.6 6.66 66 6..6 6.66 66 666 6..6 6.66 6. ...6 6.66 6. 666 6..6 6.66 6. .66 6.66 6. 6.66 6. 66. 66.6 6.66 6. 6..6 6.66 6. 6.6 6.66 6. 66. 66.6 6.66 6. 66.6 6.66 6. 6.6 6.66 6. 666 66.6 6.66 6. 66.6 6.66 6. ..6 6.66 6. 66. 66.6 6.66 6. 66.6 6.66 6. .66 6.66 6. 6.6 6..6 6. 66. 66.6 6.66 6. 66.6 6.66 6. .66 6.6. 6. 6.6 6.66 6. 66. 66.6 6.6. 6. 66.. 6.66 6. 66 6... 6. 6.6 6.6. 6. 66. 66.6 6.6. .. 66.6 6... .. 66.6 6.6. .. 6.6 6... .. 66. 66.6 6.6 6. 66.6 6.6. 6. 66.6 6.6 6. 6.6 6.6. 6. 66. 66.6 6.6 6 6... 6.6 6 66.6 6.6 6 6.6 6.6 6 66. 66.6 6.6 6 66.6 ..6 6 ...6 6.6 6 ..6 6.6 6 66. 66.6 6.6 6 66.. 6.6 6 6..6 6.6 6 6.6 6.6 6 66. 66.6 6.6 6 66.6 6.6 6 ...6 6.6 6 6.6 6.6 6 66. 66.6 6.6 6 66.. 6.6 6 66.6 6.6 6 6.6 6.6 6 66. 66.6 6.6 6 66.. 6.6 6 66.6 6.6 6 6.6 6.6 6 66. 66.6 6.6 6 66.6 6.6 6 66.6 6.6 6 6.6 6.6 6 66. 8.6 6.. 6 66.6 6.. 6 6..6 6.. 6 6.6 6.. 6 66. 66.6 6.6 . 6.6 6.6 . 66.6 6.6 . ..6 6.6 . .6666 66666 66666 66666 6 - 6.666 6.32 666666.666 663 666.. 66.6 6.6 6.6.666 +6.6 .66 665.66 66.. .66 6.6.666 66.. .66 6.6.666 6.66662 ...6 6.666 6 6.66 6..66 .6666 6 6.8 6.. 6.666 6 6.66 66 .6666 6 6.66 6.666643“... 6666636666 666666. 6.66666 2990926 5: n .U... 6.66 62636.66 .6.6< 6.66... 140 KON 00K 0.00 0N K06 00K 0.00 «N 0 P .0 0.0K VN 0.0 0.0 VN 00N 00K 0.N0 0N 00. 0 0.0K 0N 0K 0.0K 0N 00N 00K 0.0K NN 00. 0 0K0 N 0.0 0. NK NN 00N 00K 0. 0K ..N 3.0 0. ..0 ..N 0K 0. 00 N N0 .. 00K 0. v0 0N N 6.0 0. 00 0N 0.0 0.00 0N NON 0 F K 0.00 0.. 3.0 0. 0v 0' 0K 0.¢0 0.. 00 .. 00K 0.N0 0.. 0...0 0. 0v 0p 6.0 0.0V 0.. K06 0.. K 0.0V K.. 0...0 0K0 KP FK 0.N¢ Kw 0K6 00K 0.00 0.. ON... 0.0K 0.. ..0.0 0..0 06 0.06 0.00 06 00.. 0N.K 0.¢0 m.. 0N... 0. v0 0.. .0.... 0.0N 0.. 0. 0 0. 00 06 00.. 00K 0.0N 3 0K0 0. N0 .66 .0.: 0.0.. 3. 0.0 0. VN 3. 0K .. 0... K 0.0N 06 00. 6 0.0V 06 .0.: 0.06 0.. ...K 0.06 0.. 0K .. 0N. K 0.0.. Np 00.0 0.0N N.. 60.0 0. ..F NP N.0 0.NF N .. 00.. 00K 0.N? .... 00.0 0.0.. ..w .0.: 0.0.. 3 0.0 0. 06 .... 0K6 06K 0. .... 0.. 00.0 0.3. 0.. No.0 0.0 06 N.¢ 0. 0 06 00.. 00K 0. 0 0 00.0 0N.. 0 .0.: 0.0 0 0.0 0.0 0 0K.. 00K 0.0 0 00.0 0.06 0 No.0 0K 0 ¢.0 0K 0 0K 6 0.. K 0.0 K K00 0.0 K N00 0.0 K 0.N 0 .0 K 0K6 00. K 0.0 0 00.0 0.0 0 00.0 0.0 0 K.0 0.0 0 0K 6 00K 0.6 0 0 6.0 06 0 00.0 0.6 m 0.6 0.6 0 00.. 0N. K 0.0 V 00.0 0.0 V 00.0 0.0 v 0.0 0.0 v 0K.. 00. K 0.N 0 .0.... 0.N 0 00.0 0.N 0 0.0 0.N 0 006 00K 0.6 N .0.: 0.6 N 00.0 0.6 N 0.0 0... N 006 00.0 m .0 .. .0.: 0.0 .. 00.0 0.0 .. 0.N 0.0 p K0000 0 60000 00000 .. 60000 6 - 6.666 6662 666666.66 663 600: ...6... ...6 an6w ..mci. .66 2.5.60 .86 56 6.66.60 em: .8 6.6.660 6.6.66.2 ...6 6.666 6 6.66 6..66 .6666 6. 6.8 6... 6.666 6 6.66 6< .6666 6 28 Q6503 63% 3 «3.0.34 .338» 0.0< 030K 141 .6... 6.66 66 00.0 0.00 00 0.0. 00.0 0.00 «N .0... 0. 60 Na K 6 6 00.0 0.00 .0 00.0 0.60 60 K6 6 00.0 0.00 00 .0... 0.K6 00 K0 6 00.0 0. 60 06 .0... 0.06 06 0.00 00.0 0.06 0. 00.0 0.00 06 0. .0 00.0 0.00 K. K0.0 0.00 K 6 0.00 00.0 0.00 06 06.0 0.00 0. 60.0 0.06 06 6.0 0.0K 0. 0.00 00.0 0.K0 0 . 00.0 0.00 0. 00.0 0.K6 06 ..0 0.00 06 6.66 66.6 6. .6 6. 66.6 6. .6 6. 66.6 6.6. 6. 6.6 6. .6 6. ..00 0 K 0.06 06 00.0 0.00 0 . 60.0 0.06 06 0.0 0.00 06 K.0K 00K 0.06 0. 00.0 0.K— 06 60.0 0. .6 0.. 6.06 0.K. NP 6. .6 66.6 6... .. 66.6 6.6. .. 66.6 6.6. .. 6.6. 6.6. .. 0. .0 0K0 6.0 0. 0 6.6 0.06 06 6 ..0 0.0 06 0.K 0.06 06 0.06 00.0 0.K 0 06.0 0. 6 6 0 0 6.0 0.0 0 0.06 0. ... 0 0.06 0 K 0.0 0 0K.N 0.0 0 .0.... 0.K 0 0.K 0.0 0 0.06 00.0 0.0 K 00.6 0.K K 0 ..0 0.0 K 0.6 0.K K 0.00 00K 0.6 0 00.0 0.0 0 00.0 0.0 0 6.0 0.0 0 ..66 66.6 6.6 6 66.6 6.6 6 6..6 6.6 6 6.6 6.6 6 6. .6 66.6 6.6 6 66.6 6.6 6 6..6 6.6 6 6.6 6.6 6 0. 60 00.0 0.0 0 0 6.0 0.0 0 66.0 0.0 0 0.K 0.0 0 6.66 66.6 6.. 6 .6.: 6.. 6 6..6 6.. 6 6.6 6.. 6 6.66 66.6 6.6 . .6... 6.6 . 6..6 6.6 . 6.. 6.6 . «66000 0 66000 K5000 03000 6.666.666.6666 60 .. 66.66.6560 6663 666.. 66.6 .66 6.6.666 66.6 .66 6.6.666 .6666 .66 6.6.666 .66.. .66 6.6.666 6.666.... ...6 .6666 6 6.8 6..66 ...666 6 6.8 6... .6666 6 6.66 6< 6.666 6 6.8 03:66:40.0“. 6663666660 0:666: «.5660 66.6.8. 6.6< 6.666 142 .6. 66... 6.66 66 66.6 6.66 66 66.66666..6..«..6... 8. 66.6 6.66 66 66.6 6.66 66 66.6 6.6. 66 6.6 6.66 66 6.66 66.. 6.66 66 66.6 6.66 66 66.6 6.66 66 6.6. 6.66 66 6.66 66.. 6.66 66 66.6 6.66 66 66.6 6..6 66 6... 6.66 66 ..66 66.6 6.66 .6 66.6 6.66 .6 66.6 6.66 .6 ..6... 6.66 .6 6.66 66.. 6.66 66 6..6 6.66 66 .66 6...6 66 6.6 6.66 66 6.66 66.. 6..6 6. 6..6 6..6 6. 66.6 6.66 6. 6.6 6..6 6. 6.66 66.6 6.6. 6. 6..6 6.66 6. 66.6 6.66 6. .6.... 6.66 6. 6.66 66.. 6.6. 6.. 66.6 6.6. 6. R6 6.66 6. 6.6 6.6. .. 6.66 66... 6.6. 6. 6..6 6.6. 6. 66.6 6.6. 6. 6.6 6.6. 6. 6.66 66.. 6.6. 6. .6... 6... 6. 66.6 6.... 6. 6.. 6.6. 6. 6.66 66... 6.6. 6. .6... 6.6. 6. 66.6 6.6. 6. .6.... 6.6. 6. 6.66 66.. 6.6. 6. .6... 6.6. 6. 6..6 6.6. 6. 6.. 6.6. 6. ..66 66.. 6... 6. .6... 6.6. 6. 6.6.6 6... 6. 6.6 6.6. 6. 6.66 66.. 6.6. .. .6... 6.6. .. 6..6 6.6. .. 6.. 6.6. .. 6.66 66.6 6.6 6. .6... 6.6. 6. 6..6 6.6 6. 6.6 6.6. 6. 6.66 66.6 6.6 6 .6... 6... 6 6..6 6.6 6 6.6 6... 6 6.66 66... 6.6 6 .6... 6.6 6 66.6 6.. 6 6.6 6.6 6 ..66 66.6 6.6 6. .6... 6.6 ... 66.6 6.6 6 ..6 6... 6 6..6 66.. 6.6 6 .6... 6.6 6 66.6 6.6 6 6.6 6.6 6 6.6 66.. 6.6 6 .6... 6.6 6 6..6 6.6 6 6.6 6.6 6 6.66 66.6 6.6 6 .6... 6.6 6 66.6 6.6 6 6.6 6.6 6 ..66 66.. 6.6 6 .6... 6.6 6 66.6 6.6 6 6.6 6.6 6 6.66 66.6 6.. 6 .6... 6.. 6 66.6 6.. 6 6.6 6.. 6 6.66 66.6 6.6 . .6... 6.6 . .66 6.6 . 6.6 6.6 . 6.6666 666666 666666 666666 2.55 6.66.5.0 - 66.5360 63 as ...6 6.66.66 ...6... .66 6.6.666 .666} .66 6.6.666 a: ...6 6.66.66 6.6.66.2 :6 .6666 6 6.8 6666 6.666 ... 6.8 6: .6666 6 6.8 6.6 .6666 ... 6.8 6666346100 066363660 66666": 0.66660 .3608» 0.m< 030K 143 88 83 3v 3 3... 3e 3 8m «3 38 8 «.8 3e 3 8.. 83 3» «N .3. 3 3. m8 3 so as 3... a ...... a 33 3» & So a.» 33 8 8... 33 8 3. 23 8 3 83 3a 2 83. 33 2 .3. 33 2 .3 8.8 8 .3 83 3.3 9 £3. 93 2 .3. 33 S 8: c3 2 8w. 9:. 33 2 83. 33 t 2:. 3m 2 3. 33 2 8w 3.5 28 2 at... 98 m: mod 98 2 v.2 98 2 3 NE 3: 2 35 3. 9 m3. 3. m. 3. 3. 9 8P 9.3 3. 2 83. 3. 3 N3. 3.: 3 33 3: 3 8. 83 3. 2 23 3: 9 86 3. 9 3. 93 9 a: v3 3. m. 83. 3: NF 8... 3. N. 3. 3. up a: 83 3. z 83 3. : .3. 3. : S 3. 2 2: E 3 2 ~35 3 2 B... 3 S E. 3 2 8. m3 3 a £3. 3 a «3. 3 m 3 3 a 8. 5... 3 o 83 3 m 8... 3 o 3. 3 a 2: 83 3 h as... 3 3 N3. 3 h o... 3 h 8? as 3 o 33. 3 o 3... 3 o «.... 3 c we 93 3 m 93. 3 m 3... 3 m 3 3 m 8, 83 3 v as... 3 e .3. 3 c 3 3 v 8. a3 3 a 23° 3 a v3. 3 a 3 3 a 8. 92 3 m to... 3 u .3. 3 N 3 3 w 3 m3 3 P K3. 3 F «3. 3 P 3 3 F moms cums 823 83.. £83m .353: - wfiBEKabau 80: ...3. .8 23.8 ..m... .8 23.8 ..m: .8 23.8 ..m... .5 23.8 2.2.9.2 In 580 .. 28 ex“. 580 .. 2.8 2.. :58 .. 2.8 2 £80 * 2.8 Egg gum Hana: «aqua .338» m.m< $32. 144 ...... 8.. .. 3 ...... .. ...... 8.. ... 3 83 n o... ...... o 3 .3 o .... 8... o 3 ..3 m .... 8... h 3 .V 3 8... ... 3 8... a a . .88 3 ..3 m 3 8... u 532 3 8... .. 3 ...e ......88. 595.: 3 8... 3 2.52.8 o... «a... 28-32:. 9.3 3 8.. m .... 3 3... .8388. 3:88 3.. 83 ... o... ..3 28.28.. 9.3 3.. 8.. t 3.. .3 o. ...... 83 ..m 3.. ..3 m. ...... ....u 8 3.. 83 ... 3. «..m 8 33 m3 ... ...8 ....N .m 3. 8.. u. ...... ....m 8 3. ...... .. .....m .3 ... ...... ...... ... ...... a3 ... .... m3 .. 3.. :3 t 3 83 m ...8 ..3 ... .... 83 s ...8 ..3 .... 3 ..3 m ...8 83 ... ...m 83 m ...... ..3 a. o... ..3 v 3. ..3 m. 3 .3 .. . . .88 3. R3 .. .... ....o u 5...: 3 N3 ... 3 .3 .88»... 52:2: 3 N3 .....88 «82...: o... 83 ..mw...mzs.._ 9.3 ....o. .....8 co : .. 98 9.8 ....o. 5...... oo 2. .. 98 9.8 58 8.3.. 2:8... in... 28... 145 3.. 8.. 8 Q8 .3 .w 3.. 8.. 8 c8 8.. 8 3.. 8.. 8 3m 8.. a. 3» 8.. .N 3.8 83 ... 33 3.. 8 3. 2.. 3. 3m 8.. a. 3. ..3 3. ...8 .N. o. 3. .N.. 3. 3m 8.. ... 3. 3.. ... 98 8.. o. 3. .3 a. 3. 8.. m. 3. 2.. N. 3. 8.. ... 3. 8.. .. ...... 8.. a. 3. 8.. c. 3. .3 m. o... 8.. a 3. .3 .. 3 8.. m 3 8.. o. 3 8.. n 3 8.. m 3 m3 3 3 8.. o 3 83 m 3 8.. 3 3 83 .. 3 8.. o 3 23 a 596 3 2... m 528 3 ..3 m 33:8 3 .N.. v >93: 3 8... .8088. 3.8%.... 3 .3 a 9i: 3 8... .3803 9.3 3 a... N .o .3 3 8.. .888. 33 8.. 3. 3m 8.. m. 3.. 8... 3 o... 8.. v. 3.. 8.. 8 98 .3 a. 3.” 8.. 8-888 83 3. 8.. «$.88 “=3 4.5. .....8 oo 8 .. 98 9a :5. 5%.. oo .... .. 98 9.... .338. Y3 98.. REFERENCES REFERENCES Aggett J. and O'Brien GA. (1985) Detailed model for the mobility of arsenic in lacustrine sediments based on measurements in Lake Ohakuri. Environ. Sci. Technol. 19: 231-238. Aggett J. and Roberts LS. (1986) Insight into the mechanism of accumulation of arsenate and phosphate in hydro lake sediments by measuring the rate of dissolution with ethylenediamine- tetraacetic acid. Environ. Sci. Technol. 20: 183-186. Alberts J.J., Schindler J.E., Miller R.W. and Nutter DE. (1974) Elemental mercury evolution mediated by humic acid. Science 184: 895-897. Allan RJ. (1986) The Role of Particulate matter in the Fate of Contaminants in Aquatic Ecosystems. Inland Waters Directorate, National Water Research Institute, Canada Center for Inland Waters, Scientific Series No. 142. Aller R.C. (1978) The effects of animal-sediment interactions on geochemical processes near the sediment-water interface. In: Estuarine Interactions (ed. M.L. Wiley), p. 157-172. Wiley, New York. Anderson L.G., Hall 9.0.1., Iverfeldt A, Rutgers van der Loeff M., Sundby B. and Westerlund S.F.G. (1986) Benthic respiration measured by total carbonate production. Limnol. Oceanogr. 31: 319-329. Andreae MD. (1979) Arsenic speciation in seawater and interstitial water: Influence of biological- chemical interactions on the chemistry of a trace element. Limnol. Oceanogr. 24: 440-452. Bacon MP. and Rutgers van der Loeff MM. (1989) Removal of thorium-234 by scavenging in the bottom nepheloid layer of the ocean. Earth Planet. Sci. Lett. 92: 157-164. Baker ET. and Feely RA (1978) Chemistry of oceanic particulate matter and sediments: implications for bottom sediment resuspension. Science 200: 533-535. Baker LE. and Eisenreich SJ. (1989) PCBs and PAHs as tracers of particle dynamics in large lakes. J. Great Lakes Res. 15: 84-103. Balistrieri LS. and Murray J.W. (1986) The surface chemistry of sediments from the Panama Basin: the influence of Mn oxides on metal sorption. Geochim. Cosmochim. Acta 50: 2235- 2243. Beaty RD. (1988) Concepts, Instrumentation, and Techniques in Atomic Absorption Spectroscopy. Perkin-Elmer. 78 p. Belzile N. (1988) The fate of arsenic in sediments of the Laurentian Trough. Geochim. Cosmochim. Acta 52: 2293-2302. Belzile N., Lecomte P. and Tessier A. (1989) Testing readsorption of trace elements during partial chemical extractions of bottom sediments. Environ. Sci. Technol. 23: 1015- 1020. 146 147 Belzile N. and Tessier A. (1990) Interactions between arsenic and iron oxyhydroxides in lacustrine sediments. Geochim. Cosmochim. Acta 54: 103-109. Benoit G. and Hemond H.F. (1991) Evidence for diffusive redistribution of 210Pb in lake sediments. Geochim. Cosmochim. Acta 55: 1963-1975. Ben-Yaakov S. (1973) pH buffering of pore water of recent anoxic marine sediments. Limnol. Oceanogr. 18: 86-94. Berelson W.M., Hammond D.E., O'Neill D., Xu X.-M., Chin C. and Zukin J. (1990) Benthic fluxes and pore water studies from sediments of the central equatorial north Pacific: Nutrient diagenesis. Geochim. Cosmochim. Acta 54: 3001-3012. Bemer RA. (1976) The benthic boundary layer from the viewpoint of a geochemist. In: The Benthic Boundary Layer (ed. LN. McCave), p. 33-55. Plenum Press, New York. Berner RA. (1980) Early Diagenesis. Princeton University Press, Princeton, NJ. 241 p. Berner RA. (1981) A new geochemical classification of sedimentary environments. J. Sed. Petrol. 51: 359-365. Berner RA. (1985) Sulphate reduction, organic matter decomposition and pyrite formation. Phil. Trans. Royal Soc. London A 315: 25-38. Biscaye RE. and Eittreirn S.L. (1974) Variations in benthic boundary layer phenomena: nepheloid layer in the North American Basin. In: Suspended Solids in Water (ed. RJ. Gibbs). P. 227-260. Plenum Press, New York. Bothner M.H., Jahnke R.A., Peterson ML. and Carpenter R. (1980) Rate of mercury loss from contaminated estuarine sediments. Geochim. Cosmochim. Acta 44: 273-285. Boust D., Carpenter M.S.N. and Joron LL. (1988) Investigation of authigenic and diagenetic processes by chemical leaching of pelagic sediments from the Cape Verde abyssal plain. Chem. Geol. 68: 69-87. Brannon 1M. and Patrick W.H. Jr. (1987) Fixation, transformation, and mobilization of arsenic in sediments. Environ. Sci. Technol. 21: 450-459. Brooks DA (1985) Vernal circulation in the Gulf of Maine. J. Geophys. Res. 90: 4687-4705. Canfield DE. (1989) Reactive iron in marine sediments. Geochim. Cosmochim. Acta 53: 619- 632. Capone D.G. and Kiene RP. (1988) Comparison of microbial dynamics in marine and fresh- water sediments: Contrasts in anaerobic carbon catabolism. Limnol. Oceanogr. 33: 725-749. Carlton R.G., Walker 6.8., Klug MJ. and Wetzel R.G. (1989) Relative values of oxygen, nitrate, and sulfate to terminal microbial processes in the sediments of Lake Superior. J. Great Lakes Res. 15: 133-140. Cline J.T., Hillson LB. and Upchurch SB. (1973) Mercury mobilization as an organic complex. Proc. 16th Corrf. Great Lakes Res., p. 233-242. Christensen ER. and Chien N.-K. (1981) Fluxes of arsenic, lead, zinc, and cadmium to Green Bay and Lake Michigan sediments. Environ. Sci. Technol. 15: 553-558. 148 Crecelius EA. (1975) The geochemical cycle of arsenic in Lake Washington and its relation to other elements. Limnol. Oceanogr. 20: 441-451. Crecelius EA, Bothner MB. and Carpenter R. (1975) Geochemistries of arsenic, antimony, mercury, & related elements in sediments of Puget Sound. Environ. Sci. Technol. 9: 325-333 Cross P.A., Willis J .N., Hardy L.H., Jones N.Y. and Lewis J .M. (1975) Role of juvenile fish in cycling of Mn, Fe, Cu, and Zn in a coastal-plain estuary. In: Estuarine Research, v. 1 (ed. LE. Cronin), p. 45-63. Academic Press, New York. Csanady GT. (1986) Mass transfer to and from small particles in the sea. Limnol. Oceanogr. 31: 237-248. Davison W. (1985) Conceptual models for transport at a redox boundary. In: Chemical Processes in Lakes (ed. W. Stumm), p. 31-53. Wiley-Interscience, New York. Dell C.I. (1972) The origin and characteristics of Lake Superior sediments. Proc. 15th Corrf. Great Lakes Res., p. 361-370. Desaulniers J.A.H., Sturgeon R.E. and Bennan 8.8. (1985) Atomic absorption determination of trace metals in marine sediments and biological tissues using a stabilized temperature platform furnace. Atomic Spectroscopy 6: 125-127. Drever 1.1. (1988) The Geochemistry of Natural Waters, 2d ed. Prentice Hall, Englewood Cliffs, NJ. 437 p. Eadie BL (1984) Distributions of polycyclic aromatic hydrocarbons in the Great Lakes. In: Toxic Contaminants in the Great Lakes (ed. 10. Nriagu and MS. Simmons), p. 195-211. Wiley, New York. Eadie B.J., Rice GP. and Frez W.A. (1983) The role of the benthic boundary in the cycling of PCBs in the Great Lakes. In: Physical Behavior of PCBs in the Great Lakes (ed. D.Mackay, S. Paterson, SJ. Eisenreich, and MS. Simmons). P. 213-228. Ann Arbor Science, Ann Arbor, MI. Eadie BJ. and Robbins J.A. (1987) The role of particulate matter in the movement of contaminants in the Great Lakes. In: Sources and Fates of Aquatic Pollutants (ed. R.A. Hites and SJ. Eisenreich), p. 319-364. Eganhouse R.P., Young DR. and Johnson J .N. (1978) Geochemistry of mercury in Palos Verdes sediments. Environ. Sci. Technol. 12: 1151-1157. Elder J .F. (1988) Metal biogeochemistry in surface-water systems—A review of principles and concepts. U.S. Geol. Survey Circ. 1013. 43p. E1 Ghobary H. and Latouche C. (1986) A comparative study of the partitioning of certain metals in sediments from four nearshore environments of the Aquitaine coast (southwest France). In Sediments and Water Interactions (ed. P.G. Sly), p. 105-124. Springer-Verlag, New York. Falchuk KR, Boldwater LJ. and Vallee B.L. (1977) The biochemistry and toxicology of mercury. In: The Chemistry of Mercury (ed. C.A. McAuliffe), p. 261-283. The Macmillan Press Ltd., London. Farmer LG. and Lovell MA. (1986) Natural enrichment of arsenic in Loch Lomond sediments. Geochim. Cosmochim. Acta 50: 2059-2067. 149 Fame G. (1991) Principles and Applications of Inorganic Geochemistry. Macmillan Publishing Company, New York. 626 p. Faust SD. and Aly O.M. (1981) Chemistry of Natural Waters. Ann Arbor Science (Butterworth), Boston. 400 p. Faust S.D., Winka AJ. and Belton T. (1987) An assessment of chemical and biological significance of arsenical species in the Maurice River drainage basin (NJ). Part 1. Distribution in water and river and lake sediments. J. Environ. Sci. Health A22: 209-237. Feely R.A., Sullivan L. and Sackette WM. (1974) Ught-scattering measurements and chemical analysis of suspended matter in the near-bottom nepheloid layer of the Gulf of Mexico. In: Suspended Solids in Water (ed. RJ. Gibbs), p. 281-294. Plenum Press, New York. Flood R.D. ( 1989) Submersible studies of crment-modified bottom topography in Lake Superior. J. Great Lakes Res. 15: 3-14. Forbes E.A., Possner AM. and Quirk J .P. (1974) The specific adsorption of inorganic Hg(II) species and Co ([11) complex ions on goethite. J. Colloid Interface Sci. 49: 403-408. Forstner U. (1982) Accumulative phases for heavy metals in lirnnic sediments. Hydrobiologia 91: 269-284. Forstner U. and Wittrnann G.T.W. (1983) Metal Pollution in the Aquatic Environment, 2d rev. ed. Springer-Verlag, New York. 468 p. Frenet M. ( 1981) The distribution of mercury, cadmium, and lead between water and suspended matter in the Loire Estuary as a frmction of hydrological regime. Water Res. 15: 1343-1350. Gaudette H.E., Flight W.R., Toner L. and Folger D.W. (1974) An inexpensive titration method for the determination of organic carbon in recent sediments. Jour. Sed. Petrol. 44: 249-253. Gephart CJ. (1982) Relative importance of iron-oxide, manganese-oxide, and organic material in the adsorption of chromium in natural water-sediment systems. MS. Thesis, Michigan State University. 124 p. Geninga L.J.A. (1990) Aerobic degradation of organic matter and the mobility of Cu, Cd, Ni, Pb, Zn, Fe and Mn in marine sediment slurries. Mar. Chem. 29: 355-374. Gibbs RJ. (1977) Transport phases of ttansition metals in the Amazon and Yukon Rivers. Geol. Soc. America Bull. 88: 829-843. Gill GA. and Fitzgerald W.F. (1987) Picomolar mercury measurements in seawater and other materials using stannous chloride reduction and two-stage gold amalgamation with gas phase detection. Mar. Chem. 21): 227-243. Gill GA. and Fitzgerald W.F. (1988) Vertical mercury distributions in the oceans. Geochim. Cosmochim. Acta 52: 1719-1728. Gobeil C., Silverberg N., Sundby B. and Cossa D. (1987) Cadmium diagenesis in Laurentian Trough sediments. Geochim. Cosmochim. Acta 51: 589-596. Gobeil C. and Silverberg N. (1989) Early diagenesis of lead in Laurentian Trough sediments. Geochim. Cosmochim. Acta 53: 1889-1895. 150 Gratton Y., Edenbom H.M., Silverberg N. and Sundby B. (1990) A mathematical model for manganese diagenesis in bioturbated sediments. Am. Jour. Sci. 290: 246-262. Graybeal AL. and Heath GR. (1984) Remobilization of transition metals in surficial pelagic sediments from the eastern Pacific. Geochim. Cosmochim. Acta 48: 965-975. Grobenski A., Lehmann R., Radziuk B. and Vcellkopf U. (1984) Determination of trace metals in seawater using Zeeman graphite furnace AAS. Atomic Spectroscopy 5: 87-90. Gupta SK. and Chen KY. (1975) Partitioning oftrace metals in selective chemical fractions of near-shore sediments. Environ. Lett. 10: 129-158. Hikanson L. (1982) Lake bottom dynamics and morphometry: the dynamic ratio. Water Resour. Res. 18: 1444-1450. Hallberg RD. (1982) Diagenetic and environmental effects on heavy-metal distribution in sediments: a hypothesis with an illustration from the baltic sea. In: Interactions between sediments and fresh water (ed. B.L. Golterrnan), p. 305-316. Junk PubL, The Hague. Hart B.T. (1982) Uptake of trace metals by sediments and suspended particulates: a review. Hydrobiolobia 91: 299-313. Herrnanson MH. and Christensen ER. (1991) Recent sedimentation in Lake Michigan. J. Great Lakes Res. 17: 33-50. Hilton 1, Long F.J., Chapman LS. and Lishman J .P. (1986) Iron mineralogy in sediments. A Mossbauer study. Geochim. Cosmochim. Acta 50: 2147-2151. Hines M.E., Bazylinski D.A., Tugel LB. and Lyons W.B. (1991) Anaerobic microbial biogeochemistry in sediments from 2 basins in the Gulf of Maine: evidence for iron and manganese reduction. Est. Coastal Shelf Sci. 32: 313-324. Holm N.G. (1988) Arsenic regeneration from estuarine sediments of the Bothnian Bay, Sweden. Chem. Geol. 68: 89-98. Honeyman B.D., Balistrieri LS. and Murray J.W. (1988) Oceanic trace metal scavenging: the importance of particle concentration. Deep-Sea Res. 35: 227-246. Hopkins T.S. and Garfield N. III (1979) Gulf of Maine Intermediate Water. J. Marine Res. 37: 103-139. Huang P.M., Oscarson D.W., Liaw WK. and Hammer UT. (1982) Dynamics and mechanisms of arsenite oxidation by freshwater lake sediments. Hydrobiologia 91: 315-322. Hyland M.M., Jean GE. and Bancroft GM. (1990) XPS and AES studies of Hg(11) sorption and desorption reactions on sulphide minerals. Geochim. Cosmochim. Acta 54: 1957 -1967. Iricanin N. ,Trefry 11.1., Metz S., Trocine R. P. and Vetter T. W. (1985) Seasonal and spatial variati‘ons of interstitial water manganese and iron in Mississippi Delta sediments. Abstract. Eos 6 1307. Jackson T.A., Parks J.W., Jones P.D., Woychuk R.N., Sutton LA. and Hollinger JD. (1982) Dissolved and suspended mercury species in the Wabigoon River (Ontario, Canada): seasonal and regional variations. Hydrobiologia 92: 473-487. 151 Jahnke R.A., Emerson S.R., Reimers C.E., Schuffert J ., Ruttenberg K. and Archer D. (1989) Benthic recycling of biogenic debris in the eastern tropical Atlantic Ocean. Geochim. Cosmochim. Acta 53: 2947-2960. Jean GE. and Bancroft GM. (1986) Heavy metal adsorption by sulphide mineral surfaces. Geochim. Cosmochim. Acta 50: 1455-1463. Jennet J .C., Effler S.W. and Wixson B.G. (1980) Mobilization and toxicological aspects of sedimentary contaminants. In: Contaminants and Sediments, vol. 1 (ed. R.A. Baker), p. 429-444. Ann Arbor Science, Ann Arbor, MI. Johnson MG. (1991) Temporal trends in metal concentrations in rivers, coastal-zone waters, and sediments of northern and eastern Georgian Bay. J. Great Lakes. Res. 17: 241-254. Johnson T.C., Evans IE and Eisenreich SJ. (1982) Toral organic carbon in Lake Superior sediments: Comparisons with hernipelagic and pelagic marine envircnrnents. Limnol. Oceanogr. 27: 481-491. Johnson T.C., Halfman J.D., Busch W.H. and Flood RD. (1984) Effects of bottom currents 3nd fish on sedimentation in a deep-water, lacustrine environment. Geol. Soc. Amer. Bull. 5: 1425-1436. Kemp A.L.W., Dell Cl. and Harper NS. (1978) Sedimentation rates and a sediment budget for Lake Superior. J. Great Lakes Res. 4: 276-287. Kemp A.L.W. and Johnston L.M. (197 9) Diagenesis of organic matter in the sediments of Lakes Ontario, Erie, and Huron. J. Great Lakes Res. 5: 1-10. Kennedy H.A. and Elderfield H. (1987) Iodine diagenesis in pelagic deep-sea sediments. Geochim. Cosmochim. Acta 51: 2489-2504. Kersten M. and Forsurer U. (1987) Effect of sample pretreatment on the reliability of solid speciation data of heavy metals - implications for the study of early diagenetic processes. Mar. Chem. 22: 299-312. Kheboian C. and Bauer CF. (1987) Accruacy of selective extraction procedtu'es for metal speciation in model aquatic sediments. Anal. Chem. 59: 1417-1423. Klump J.V., Paddock R., Remsen C.C., Fitzgerald S., Boraas M. and Anderson P. (1989) Vafiafimsinsedimtaccmmflafimmesandmefhmofhbflemgmucmauermeastan Lake Superior basins. J. Great Lakes Res. 15: 104-122. Klurnp J .V. and Martens C.S. (1989) The seasonality of nutrient regeneration in an organic-rich coastal sediment: Kinetic modeling of changing pore-water nutrient and sulfate distributions. Limnol. Oceanogr. 34: 559-577. Konricker W.A. and Morse J .W. (1991) Interactions of divalent cations with the surface of pyrite. Geochim. Cosmochim. Acta 55: 2159-2171. Krezoski J.R. (1989) Sediment reworking and transport in eastern Lake Superior: in situ rare earth element tracer studies. J. Great Lakes Res. 15: 26-33. Kuivila KM. and Murray J .W. (1984) Organic matter diagenesis in freshwater sediments: The alkflhutyandtmflcozbalmcemdnemmpromwdmmdnsedinentsoflakeWaslfingtm Limnol. Oceanogr. 29: 1218-1230. 152 Lampitt, RS. (1985) Evidence for the seasonal deposition of detritus to the deep-sea floor and its subsequent resuspension. Deep-Sea Res. 32: 885-897. Lange T. (1991) Enviroscope: Mercury plagues North American lakes. Ocala Star-Banner, Saturday October 5, 1991, p. 21). Ocala, FL. Lerman A. (1977) Migrationalprocesses and chemical reactions in interstitial waters. In: The Sea, v. 6 Marine Modeling (ed. E.D. Goldberg, LN. McCave, J .J . O'Brien and J.H. Steele), p. 695-738. Wiley, New York. Lennan A. and Brunskill GJ. (1971) Migration ofmajor constituents from lake sediments into lake water and its bearing on lake water composition. Limnol. Oceanogr. 16: 880-890. Letoumeau V.A., Joshi B.M. and Butler LC. (1987) Comparison between Zeeman and continuumbackgroundcorrection forgraphitefumaceAAS onenvircnmentalsamples. Atomic Spectroscopy 8: 145-149. Li Y.-H. and Gregory S. (1974) Diffusion of ions in sea water and in deep-sea sediments. Geochim. Cosmochim. Acta 38: 703-714. Lindberg SE. and Harriss R.C. (1974) Mercury-organic matter associations in estuarine sediments and interstitial water. Environ. Sci. Technol. 8: 459-462. Lineback J .A., Dell CI and Gross D.L. (1979) Glacial and postglacial sediments in Lakes Superior and Michigan. Geol. Soc. Amer. Bull. 90: 781-791. Loring DH. (1975) Mercury in the sediments of the Gulf of St. Lawrence. Can. J. Earth Sci. 12: 1219-1237. Loring DH. and Rantala RT.T. (1988) An intercelibration exercise for trace metals in marine sediments. Mar. Chem. 24: 13-28. Mantoura R.F.C., Dickison A. and Riley J.R. (1978) The complexation of metals with humic materials in natural waters. Estuarine Coastal Mar. Sci. 6: 387-408. Martin J .M., Nirel P. and Thomas AJ. (1987) Sequential extraction techniques: promised and problems. Mar. Chem. 22: 313-341. Martin W.R. and Sayles FL. (1987) Seasonal cycles of particle and solute transport processes in nearshore sediments: Wm and 234mm disequilibrium at a site in Buzzards Bay, MA. Geochim. Cosmochim. Acta 51: 927-943. Masuzawa T., Noriki S., Kurosaki T., Tsunogai S. and Koyama M. (1989) Compositional change of settling particles with water depth in the Japan Sea. Mar. Chem. 27 : 61-87. Matty J. M. Anderson J. B. and Dunbar R. B. (1987) Suspended sediment transport, sedirnentatin, and resuspension in Lake Houston, Texas: implications for water quality. Environ. Geol. Water Sci. 10:175-186. Mayer L.M., Macko S. A. and Cammen L. (1988) Provenance, concentrations and nature of organic nitrogen in the Gulf of Maine. Mar. Chem. 25° 291-304. McKee J .D. (1990) Geochemical Cycling of Heavy Metals in Proundal Sediments of Lake Superior. MS. Thesis, Michigan State University, E. Lansing, MI. 85 p. 153 McKee J.D., Wilson T.P., Long D.T. and Owen R.M. (1989a) Geochemical partitioning of Pb, Zn, Cu, Fe, and Mn across the sediment-water interface, Caribou sub-basin, Lake Superior. J. Great Lakes Res. 15: 46-58. McKee J .D., Wilson T.P., Long D.T. and Owen R.M. (1989b) Pore-water profiles and early diagenesis of Mn, Zn, Cu, and Pb in sediments from the Caribou sub-basin, Lake Superior. J. Great Lakes Res. 15: 68-83. Meade R.H., Sachs P.L., Manheirn F.T., Hathaway J .C. and Spencer D.W. (1975) Sources of suspended matter in waters of the Middle Atlantic Bight. J. Sed. Petrol. 45: 171-188. Mikac N., Picer M, Stegnar P. and Tusek-Znidaric M. (1985) Mercury distribution in a polluted marine area, ratio oftotal mercm'y, methyl mercm'y and selenium in sediments, mussels, and fish. Water Res. 19: 1387-1392. Moody J.R., Paulsen PJ., Rains TC and Rock 111.. (1976) The preparation and certification of trace mercury in water standard reference materials. National Bureau of Standards Spec. Pub. 422: 267-273. Moore J.N., Ficklin W.H. and Johns C. (1988) Partitioning of arsenic and metals in reducing sulfidic sediments. Environ. Sci. Technol. 22: 432-437. Moore J .W. and Ramamoorthy S. (1984) Heavy Metals in Natural Waters. Springer-Verlag, New York. 268 p . Morel F.M.M. and Gschwend PM. (1987) The role of colloids in the partitioning of solutes in natural waters. In: Aquatic Surface Chemistry (ed. W. Stumm), p. 405-422. Wiley- Interscience, New York. Morse J.W. and Comwell J.C. (1987) Analysis and distribution of iron sulfide minerals in recent anoxic marine sediments. Mar. Chem. 22: 55-69. Mudroch A., Sarazin L. and Lomas T. (1988) Summary of surface and background goncentrations of selected elements in the Great Lakes sediments. J. Great lakes Res. 4: 241-251. Nirel P., Thomas AJ. and Martin J.M. (1985) A critical evaluation of sequential extraction techniques. In: Speciation of Fusion and Activation Products in the Environment (ed. R.A. Bulman and J .R. Cooper), p. 19-26. Elsevier, New York Nowell A.R.M., Jumars RA. and Eckrnan J .E. (1981) Effects of biological activity on the entrainment of marine sediments. Mar. Geol. 42: 133-153. O'Melia CR. (1985) The influence of coagulation and sedimentation on the fate of particles, associated pollutants, and nutrients in lakes. In: Chemical Processes in Lakes (ed. W. Stumm), p. 207-224. Wiley-Interscience, New York. O'Melia CR. (1987) Particle-particle interactions. In: Aquatic Surface Chemistry (ed. W. Stumm), p. 385-403. Wiley-Interscie , New York. Olson B.H. and Cooper R.C. (1974) In situ methylation of mercury in estuarine sediment. Nature 252: 682-683. Papp C.S.E., Filipek, L.H. and Smith KS. (1991) Selectivity and effectiveness of extractants used to release metals associated with organic matter. Appl. Geochem. 6: 349-353. 154 Pedersen T.P., Vogel LS. and Southon J .R. (1986) Copper and manganese in hemipelagic m? at 21°N, East Pacific Rise: Diagenetic contrasts. Geochim. Cosmochim. Acta 50: 19- 1. Peterson ML. and Carpenter R. (1983) Biogeochemical processes affecting total arsenic and arsenic species distributions in an intermittently anoxic fjord. Mar. Chem. 12: 295-321. Peterson ML. and Carpenter R. (1986) Arsenic distributions in porewaters and sediments of Puget Sound, Lake Washington, the Washington coast and Saanich Inlet, B.C. Geochim. Cosmochim. Acta 50: 353-369. Pocklington R. and Tan RC. (1987) Seasonal and annual variations in the organic matter contributed by the St. Lawrence River to the Gulf of St. Lawrence. Geochim. Cosmochim Acta 51: 2579-2586. Prohic E. and Kniewald G. (1987) Heavy metal distribution in recent sediments of the Krka River Estuary - an exmple of sequential extraction analysis. Mar. Chem. 22: 279-297. Rae J.E. and Aston SR. (1982) The role of suspended solids in the estuarine geochemistry of mercury. Water Res. 16: 649-654. Rapin F., Tessier A., Campbell P.G.C and Carignan R. (1986) Potential artifacts in the determination of metal partitioning in sediments by a sequential extraction procedure. Environ. Sci. Technol. 20: 836-840. Ramet G., Rubio R. and Lopez-sanchez J .F. (1989) Optimization of Tessier procedure for metal solid speciation in river sediments. Intern. J. Environ. Anal. Chem. 36: 69-83. Rea D.K., Owen RM. and Meyers RA. (1981) Sedimentary processes in the Great Lakes. Rev. Geophys. Space Phys. 19: 635-648. Rendell P.S., Batley GE. and Cameron M. (1980) Adsorption as a control of metal concentrations in sediment extracts. Environ. Sci. Technol. 14: 314-318. Rossmann R. (1986) Trace metal concentrations in the offshore waters and sediments of Lake Superior. Great Lakes Research Division Special Report N o. 121. Sadiq M. (1990) Arsenic chanistry in marine environments: acornpariton between theoretical and field observations. Mar. Chem. 31: 285-297. Salomons W. and Fbrstner U. (1984) Metals in the Hydrocycle. Springer-Verlag, New York. 349 p. Sanders P.G. (1985) Arsenic geochemistry in Chesapeake Bay: dependence upon anthropogenic inputs and phytoplankton species composition. Mar. Chem. 17: 329-340. Sandilands R.G. and Mucroch A. (1983) Nepheloid layer in Lake Ontario. J. Great Lakes Res. 9: 190-200. Santschi P.H. (1984) Particle flux and trace metal residence time in natural waters. Limnol. Oceanogr. 29: 1100-1108. Santschi P.H. (1988) Factors controlling the biogeochemical cycles of trace elements in fresh and coastal marine waters as revealed by artificial radioisotopes. Limnol. Oceanogr. 33: 848-866. 155 Santschi P.H., Amdurer M., Adler D., O'Hara R, Li Y.-H. and Doering P. (1987) Relative mobility of radioactive trace elements across the sediment-water interface in the MERL model ecosystems of Narragansett Bay. J. Marine Res. 45: 1007-1048. Schoonen M.A.A. and Barnes H.L. (1991a) Reactions forming pyrite and marcasite from solution: I. Nucleation of FeSZ below 100°C. Geochim. Cosmochim. Acta 55: 1495-1504. Schoonen M. AA. and Barnes H. L. (1991b) Reactions formingcpyrite and marcasite from solution: II. Via FeS precursors below 100°C. Geochim. osmochim. Acta 55° 1505-1514. Seelye J.G., Hesselberg RI and Mac MJ. (1982) Accumulation by fish of contaminants released from dredged sediments. Environ. Sci. Technol. 16: 459—464. Seuss E. (1979) Mineral phases formed in anoxic sediments by microbial decomposition of organic matter. Geochim. Cosmochim. Acta 43: 339-352. Shaw T.J., Gieskes J.M. and Jahnke RA. (1990) Early diagenesis in differing depositional environments: The response of transition metals in pore water. Geochim. Cosmochim. Acta 54: 1233-1246. Sigg L., Surrrn M. and Kistler D. (1987) Vertical transport of heavy metals by settling particles in Lake Zurich. Limnol. Oceanogr. 32: 112-130. Silverberg N., Gobeil C,. and Edenbom H.M. (1985) On-station versus seasonal variations in early diagenesis parameters in sediments of the Laurentian Trough. Abstract. Eos 66: 1307. Sly P.G. and Thomas R.L. (1974) Review of geological research as it relates to an understanding of Great Lake limnology. J. Fish. Res. Board Canada 31: 795-825. Spangler W.J., Spigarelli J.H., Rose J.M. and Miller H.M. (1973) Methylrnercury: bacterial degradation in lake sediments. Science 180: 192-193. Spencer D.W. and Sachs PL. (1970) Some aspects of the distribution, chemistry, and mineralogy of suspended matter in the Gulf of Maine. Mar. Geol. 9: 117-136. Spinrad R.W. (1986) An optical study of the water rrrasses of the Gulf of Maine. J. Geophys. Res. 91C: 1007-1018. Stokes RM. and Wren CD. (1987) Bioaccumulation of mercury by aquatic biota in hydroelectric reservoir: a review and consideration of mechanisms. In: Lead, Mercury, Cadmium, and Arsenic in the Environment (ed. T.C. Hutchinson and K.M. Meema), p. 255-277. Wiley, New York. Strunk J.L. (1991) The extraction of mercury from sediment and the geochemical partitioning of mercury in sediments from Lake Superior. MS. Thesis, Michigan State University, E. Lansing, MI. Stmdby B., Anderson L.G., Hall P.O.J., Iverfeldt A., Rutgers van der Loeff MM. and Westerlund S.F.G. (1986) The effect of oxygen on release and uptake of cobalt, manganese, ilrgrsrsand phosphate at the sediment-water interface. Geochim. Cosmochim. Acta 50: 1281- Sweerts J .P Rudd J. W. M. and Kelly C. A. (1986) Metabolic activities in flocculant Stu-face sediments and underlying sandy littoral sediments. Limnol. Oceanogr. 31° 330-338. 156 Sweerts J.-P. R.A., Kelly C.A., Rudd J.W.M., Hesslein R. and Cappenberg TE. (1991) Similarity of whole-sedirnent molecular diffusion coefficients in freshwater sediments of low and high porosity. Limnol. Oceanogr. 36: 335-342. Takarnatsu T., Kawashirna M. and Koyama M. (1985) The role of Mn2-i-rich hydrous manganese oxide in the accumulation of arsenic in lake sediments. Water Res. 19: 1029-1032. Tessier A., Campbell P.G.C. and Bisson M. (1979) Sequential extraction procedlne for the speciation of particulate trace metals. Anal. Chem. 51: 844-851. Tipping E., Hetherington N.B., Hilton J., Thompson D.W., Bowles E. and Hamilton-Taylor J. (1985) Arfifacmmtheuseofselecfiveclemicalemacfimmdemnnhwdisuihrfimsofnuah between oxides of manganese and iron. Analytical Chem. 57: 1944-1946. Time T., Edgington D.N. and Scils CA. (1988) Sulfate reduction in sediment interstitial fluids in Lakes Michigan and Eric. J. Great Lakes Res. 14: 14.22. Tsunogai S. and Uematsu M. (1978) Particulate manganese, iron, and aluminum in coastal water, Funka Bay, Japan. Geochem. low. 12: 39-46. Wesrin P., thener P., Giovanoli R. and Stumrn W. (1991) Early diagenetic influences on iron transformations in a fresh-water lake sediment. Chem. Geol. 90: 233-252. Westerlund S.F.G., Anderson L.G., Hall P.O.J., Iverfeldt A., Rutgers van der Loeff MM. and Sundby B. (1986) Benthic fluxes of cadmium, copper, nickel, zinc and lead in the coastal environment. Geochim. Cosmochim. Acta 50: 1289-1296. Whitfield M. and Turner DR. (1987) The role of particles in regulating the composition of seawater. In: Aquatic Surface Chemistry (ed. W. Stmnm), p. 457-493. Wiley-Interscience, New York. Wilson T.P., Long D.T. and Owen RM. (1986) The water-sediment interface via manned submersible. Abstract. International Assoc. Great lakes Res., 29th Conference, p. 52. Wood J.M. (1974) Biological cycles for toxic elements in the environment. Science 183: 1049- _ 1052. "‘illllllllllllllf