.1 v a 4 ‘L‘ 5113‘ a r' rkflh‘ 2.; a. f 5‘... ( .«rv - A M <_~m'~‘~a \\ r ,e‘d a n‘ k . h .v‘ ”‘11; A‘ “3‘ 'IC.- - . ”‘3, . ‘3' I, ‘ ' H. .m . ~», 9 ..I I ' I x. r: “7' l I 3.1;. § ' § .: ‘ _ our. :5, ;r $1.. A‘ : . ‘3‘" . '1 t- (on :V: ; "(I j... ‘1' . J . {v . {fl if?“ a 4 rr .JJ . ”“5: U ..‘ .3.» If? .... .rn_ , . -r"t‘:.;—;VI ‘11 -I " __ M. .o- a ‘V .ov '1‘".-‘.‘ .- “- Earmaje m9“ "‘ ' W ._ .. ../- ..‘ 'I. ‘u ‘ l\\\\l\l\m‘lllif\lll\“\llfl~l”“l\“\l M 5 3 (a 60 ’5? / LIBRARY Michigan State University This is to certify that the thesis entitled Geochemical Cycling of Heavy Metals In Profundal Sediments of Lake Superior presented by Joseph Duane McKeexg has been accepted towards fulfillment of the requirements for M-Sc- degree mm ‘ {912mg} fi’fl Major professor Date (Q/NUU 8Q 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES man on or before dde due. DATE DUE DATE DUE DATE DUE mas FEE! 1—— 97 mm :1 1 1:200 TL MSU II An Ailirmdivo Action/Equal Opportunity Institution GEOCHEMIGAL CYCLING OF HEAVY METALS IN PROPUNDAL SEDIMENTS OF LAKE SUPERIOR BY Joseph Duane McKee A THESIS submitted to Michigan state University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 1990 l "ti/V”? ABSTRACT GEOCHEMICAL CYCLING OP HEAVY METALS IN PROFUNDAL SEDIMENTS OF LAKE SUPERIOR BY Joseph Duane McKee The geochemical profiles of Fe, Mn, Cu, Pb and, Zn in sediments of Lake Superior were measured in interfacial sediments and in the sediment column, in the Caribou sub— basin. The profiles were examined by studying the partitioning of the 'metals among solution reactive or hydromorphic phases (as operationally defined by sequential chemical extractions). Mn, Cu, and Pb were measured in pore waters from the same site. Pb, Cu, and Zn are enriched in the surface sediments. Mn and Fe are concentrated at between 5 and 10 cm. depth in the sediment. Each of the metals is uniquely partitioned among the sediment phases and the partitioning changes from the interfacial sediment to the sediment column and with depth. Pore water profiles of Mn, Cu, and Pb show evidence for post-depositional mobility. Diagenetic modeling suggests Mn pore water and sediment profiles are linked, but the profiles are not at steady state. Flux estimates for Cu and Pb show that a significant fraction of these metals could be diffusing out of the sediment column due to remobilization from the sediments. Copyright by Joseph Duane McKee 1990 ACKNOWLEDGEMENTS I would like to thank: -Dr. David Long for all his help and encouragement as my advisor, -Dr. Michael Velbel and Dr. Michael Klug, the members of my faculty commitee, -Dr. Timothy Wilson for helping me get started and keep going. -All the people, too numerous to mention, who instigated and participated in the National Undersea Research Program in the Great Lakes, and have made it such a rewarding experience for all involved, -and finally Becky McKee, for moral support. iii TABLE OF CONTENTS page 1 Introduction... ....... ...... ......................... 1 1.1 Statement of the Problem 1 1.2 The Study Site 5 2 Methods..................................... ..... ....9 2.1 Sampling Techniques 9 2.2 Sample Preparation and Analysis 12 3 Discussion I: Metal Partitioning in the Sediments...16 3.1 General Results 16 3.1.1 Manganese 21 3.1.2 Iron 22 3.1.3 Zinc 22 3.1.4 Copper 23 3.1.5 Lead 24 3.2 Interpretation of Partitioning 25 4 Discussion II: Pore Water Results and Diagenetic MOdeISOOOOOOOOO0.00.0000...0.0.00.0000034 4.1 Manganese 34 4.1.1 Results 34 4.1.2 Theory of Manganese Diagenesis 36 4.1.3 Diagenetic Model for Manganese 38 4.1.4 Application of the Model 40 4.1.5 Discussion of the Modelling Results 48 4.2 Copper and Lead 51 4.2.1 Results 51 4.2.2 Flux Model for Copper and Lead 55 5 Conclusions.........................................61 5.1 Summary and Conclusions 61 5.2 Recommendations for Further Work 65 Appendix A: Site 1388 Data 66 Appendix B: Steady State Experiments 73 Bibliography 81 iv LIST OE TABLES TABLE 1. General description of core from site 1388 taken at 47°22.31'N 86°57.59'w at a water depth of 335 m. 2. Selective chemical attack method for a 1 gram sample, modified from Gephart (1982). 3. Comparison of trace element data (dry weight of sediments) from site 5 of Kemp et al. (1978a) for the Caribou sub basin and site 1388 this study. Values are for top sediments and lower sediments where concentrations become essentially constant. See text for locations and methods of determination. 4. Parameters used for fitting data from 1388 to diagenetic model of Burdidge and Gieskes (1983). 5. Parameters used in the calculation of the fluxes of Cu and Pb. A porosity of 0.93 is assumed. Concentrations are in pg/g, fluxes are in pg/cm2*yr. page 14 29 42 57 6. Per cent of metal remobilized from surface sediments for various sedimentation rates at site 1388. R is in g/cm2*yr. % remobilized =(Fd/Fs)*100%. A. Porosity and organic carbon concentrations at site 1388. B. Iron in sediment fractions at site 1388. Concentrations in pg/g of dry sediment. C. Manganese in sediment fractions at site 1388. Concentrations in pg/g of dry sediment. D. Copper in sediment fractions at site 1388. Concentrations in pg/g of dry sediment. E. Lead in sediment fractions at site 1388. Concentrations in pg/g of dry sediment. F. zinc in sediment fractions at site 1388. Concentrations in pg/g of dry sediment. G. Pore water metals at site 1388. H. Steady state experiments. vi 59 66 67 68 69 7O 71 72 '74 LIST OF FIGURES FIGURE 1. Location of coring and dive site (site 1388) in the Caribou Sub-Basin. 2.a. Lower work platform of the submersible JOHNSON SEA-LINK-II (JSL) showing arrangement of the sample containers, suction pump, and front compartment carrying scientists. b. vacuum/filtration device on JSL. The vacuum nozzle is gently waved above the sediment bottom, resuspending the interfacial sediments; The sediments are vacuumed through a glass fiber filter held in a teflon filter mount. 3. pH of the sediments at site 1388. The open symbol at zero depth is the overlying water. 4. Corg of the sediment at site 1388, (weight % of dry sediment). The open symbol at zero depth is the fluff sample. vii page 10 17 17 5. Concentration (Log) of a. Mn and b. Fe in chemical 18 phases of sediments at site 1388. Zero depth is the fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). 6. Concentration (Loglo) of Zn in chemical phases of 19 sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). 7. Concentration (Loglo) of Cu in chemical phases of 19 sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals), Ex (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). viii 8. Concentration (Loglo) of Pb in chemical phases of sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), 0X (oxidizable). 9. Mn pore-water and sediment concentration profiles. Values are mg/l and weight % of dry sediment respectively. ER (easily reducible phase), MR (moderately reducible phase), TM (sum of ER and MR phases), and PW (pore water). The open pore water symbol at x=0 is for Mn in the overlying water. 10 a) General zones of Mn diagenesis from model of Burdidge and Gieskes (1983) and Klinkhammer (1980). b) Hypothetical pore-water and sediment concentration profiles of Mn during steady state diagenesis. ix 20 35 37 11. Results of fit of Mn pore water profile to BC model for high linear sedimentation rate (0.017 cm/yr) with L1 = 9 cm, L2 = 12 cm, and parameters from Table 3. a) Measured pore water profile and modelled profile, k ox 4.18*1o'5 yr'1 and kred 3.18ir10'3 yr'l. b. IMeasured sediment. profile and predicted profile from 86 model. 12. Results of fit of Mn pore water profile to BC model for low linear sedimentation rate (0.0055 cm/yr) with L1 = 9 cm, L2 = 12 cm, and parameters from Table 3. a) Measured pore water profile and modelled profile, kox = 13.2 yr'1 and k red 2.76%10"3 yr'l. b) IMeasured sediment. profile and jpredicted. profile from BG ‘model, dashed line is sediment profile predicted from kox = 149 yr'l, kred 7.751':10'4 yr'l, and w = .0015 cm/yr. Pore water profile resulting is the same as a). 13. Results of fit of Mn sediment profile to BC model for high linear sedimentation rate (0.017cm/yr) with L1 = 4.5 cm, L2 = 7.5 cm, and the parameters from Table 3. a) Measured sediment profile and modelled profile, kox = 156 yr'1 and kred = 3.58*10'2. b) Measured pore water profile and predicted profile from BG model. 44 44 47 14. Results of fit of Mn sediment profile to BG model for low linear sedimentation rate (0.0055 cm/yr) with L1 = 4.5 cm, L2 = 7.5 cm, and parameters from Table 3. a) Measured sediment profile and modelled profile, kox =156 yr"1 and kred = 1,15*1o'2 yr'l. b) Measured pore water profile and predicted profile from BC model. 15. Pore-water and sediment concentration profiles of a) Cu, and b) Pb. Values are in units of pg/l and ug/g of dry sediment respectively. OX (oxidizable phase); MR (moderately reducible phase): TM (sum of WAS, EX, ER, MR and 0X phases): and PW (pore water). The open pore water symbol at x=0 is the concentration in the bottom water. A. Steady state experiment on exchangeable (EX) fraction. Arrow is time recommended by Gephart (1982). B. Steady state experiment on weakly acid soluble (WAS) fraction. Arrow is time recommended by Gephart (1982). C. Steady state experiment on easily reducible (ER) fraction. Arrow is time recommended by Gephart (1982). xi 47 52 76 76 78 D. Steady state experiments on moderately reducible 79 (MR) fraction. Arrow is time recommended by Gephart (1982) . xii CHAPTER 1 INTRODUCTION 1.1 STATEMENT OF THE PROBLEM This thesis examines the diagenetic mobility of Mn, Fe, Zn, Cu, and Pb in sediments and waters of Lake Superior. Geochemical partitioning of the metals in the interfacial sediment and in the underlying sediments and (pore water profiles of Mn, Cu, and Pb were measured. To determine if the pore water and sediment concentration profiles of Mn were in a steady state relationship, the measured Mn profiles were compared to theoretical Mn profiles predicted by the steady state diagenetic model of Burdidge and Gieskes (1983). Flux calculations similar to the those of Callender and Bowser (1980) were used to examine the net migration of Cu and Pb across the sediment-water interface. The effect of not including metal concentrations in the interfacial sediments on the flux estimates was also evaluated. Particles settling through the water column can scavenge metals from lake waters and transport them to the bottom sediments. Biogenic-organic particles are thought to play the dominant role in scavenging metals throughout the water column, but in deeper water Mn and Fe cycling due to REDOX reactions may be important (Sigg 1985). Similar processes are thought to deplete seawater of heavy metals 1 2 (Balistrieri and Murray 1986) and enrich surface sediments in heavy metals (Klinkhammer et al. 1982). Where high metal concentrations occur in surficial lake sediments, the concentrations typically decrease with depth to some "background" value (Leland et al. 1973) . These sediment-concentration profiles have been attributed to anthropogenic inputs (Allen 1986: Kemp et al. 1978a), and used to constrain the history of the chemical loading to lakes (Edgington and Robbins 1976: Stumm and Baccini 1978). However, all metal enrichments in surficial sediments may not be due to anthropogenic additions. The effects of diagenesis on trace metal concentrations in sediments can be misinterpreted as the results of anthropogenic inputs (Cornwell 1986). For example, Robbins and Callender (1975) demonstrated that the high Mn concentrations in surficial sediments of Lake Michigan are the result of processes associated with early diagenesis of the sediments rather than with anthropogenic inputs. They also suggested that early' diagenesis might lead to 'the enrichment. of other metals in surficial sediments. Carignan and Nriagu (1985) have shown that post-depositional migration of metals during early ,diagenesis is an important mechanism for the accumulation of Cu and Ni in acid-lake sediments. Diagenesis is the chemical, biochemical, and physical changes that occur during the burial of sediments. Early diagenesis refers to these changes in the upper few cm of the sediments (Berner 1980). Metals are known to be 3 sequestered by different phases of lacustrine sediments such as iron and manganese oxides, organic matter, and clays (Forstner 1982: Filipek and Owen 1979). The partitioning of the metals among these phases in sediments can change during diagenesis. For example, as Fe and Mn oxides are reduced and organic matter decays, metals associated with these phases can be released to solution or resequestered on other phases of the sediment (Klinkhammer et a1. 1982: Pedersen et a1. 1986). Except for Mn and perhaps Fe, the effect of early diagenesis on metal behavior in the sediments of the Great Lakes is not well understood (Robbins and Callender 1975: Kemp et al. 1978a: Rea et al. 1981). An understanding of the effect of early diagenetic processes on metal cycling at the sediment-water interface is important in predicting the fate of pollutants in and the long-term health of the Great Lakes. Modeling of diagenetic processes relies on an accurate chemical characterization of the sediment-water interface. However, sampling the interfacial sediments with gravity-driven devices is hampered by the "bow wave" created as the device approaches the bottom, dispersing the surface material before it is sampled. Thus, representative interfacial sediment may not be collected (Berner 1980: Heggie et a1. 1986) resulting in incorrect diagenetic models. Soutar and Johnson (1981), using a specially designed box corer, were able to collect interfacial sediments in a marine renvironment. The sediments ‘were. rich in organics 4 compared to underlying sediments. Murray et al. ( 1984) , Balistrieri and Murray (1984) and Balistrieri and Murray (1986) used the manned submersible ALVIN to collect interfacial sediments at two sites in the eastern equatorial Pacific Ocean. They also found the sediments to be enriched in organics and in heavy metals compared to the sediments below. Klinkhammer et a1. (1982) showed that interfacial sediments must be included in diagenetic models if accurate results are to be obtained. Under the sponsorship of the .National Underseas Research Program (NURP) of the National Oceanic and Atmospheric Administration (NOAA) , The manned submersible JOHNSON-SEA-LINK-II (JSL), was used to locate and sample a ’transition zone' of interfacial sediment approximately 0.5 to 1.5 cm thick in Lake Superior (Wilson et a1. 1986: McKee et a1. 1986). This sediment, the benthic-sediment layer, is referred to as the "fluff" because of its in situ appearance as observed from the submersible and seems to be the true interfacial sediments in the lake. 1.2 THE STUDY SITE The sampling site (NURP-UCAP dive site 1388) in the Caribou sub-basin was 47° 22.31' N and 860 57.59’ W, at a water depth of 335 m. The samples were collected during the 1986 dives of the JSL in the Caribou sub-basin, Lake Superior. The Caribou sub-basin occupies approximately 13,000 km2 in the eastern part of Lake Superior, Figure 1. The sediment is mainly clay. The post—glacial sediment is approximately 1.6 meters thick with the surface sediments comparatively rich in organic carbon (z3% C (Thomas and Dell 1978: org) Kemp et al. 1978b). The present-day sedimentation rate in the sub-basin, based on the Ambrosia horizon, is 0.12 mm/yr, which is not significantly different from its average post- glacial sedimentation rate of 0.17 mm/yr (Kemp et al. 1978b). This is in contrast to the other sub-basins in Lake Superior which have higher present-day sedimentation rates compared to their average post-glacial rates (Kemp et al. 1978b). A possible explanation for the difference in present-day and average post-glacial sedimentation rates of the Caribou sub-basin. and the other sub-basins in Lake Superior is discussed below. Table 1 gives a visual description of the core and associated interfacial sediment made during sectioning of a core from site 1388 in the Caribou sub basin of Lake Superior. In general, the basinal 49: 48. _ g/s ...... '0. _ . jCAQIeo gsuaj‘eAsm 47- - Figure 1. Location of coring and dive site (site 1388) in the Caribou Sub-Basin. 7 TABLE 1. General description of core from site 1388 taken at 47° 22.31' N 36° 57.59' w at a water depth of 335 m. Sample Thickness Average Comments number Depth (cm) F1 - 0 dark grey to chocolate brown 1 1 cm 0.5 dark grey 2 1 cm 1.5 dark grey to brown clay transition 3 1 cm 2.5 brown clay 4 1 cm 3.5 brown clay 5 1 cm 4.5 brown clay 6 1 cm 5.5 transition into brown layer, sediment begins to get firmer 7 1 cm 6.5 brown layer, very firm, darker streaks 8 1 cm 7.5 brown layer, dark brown streaks, turning to reddish, grey clay 9 1 cm 9 tan-gray clay, black streaks may be sulfides,sti11 firm 10 1 cm 10 transition to softer grey clay 11 1 cm 11 soft grey clay 12 1 cm 12 soft grey clay 13 2 cm 13.5 soft grey clay 14 2 cm 15.5 firm grey clay 15 3 cm 18 firm grey clay 16 3 cm 21 firm grey clay 17 3 cm 24 firm grey clay 18 3 cm 27 firm grey clay 19 4 cm 30.5 firm grey clay 1) the submersible collected interfacial sediment assigned a depth of 0 cm on the sediment profiles was 8 sediments of Lake Superior are similar to marine hemipelagic sediments (Johnson et al. 1982). The top 3 cm the sediments are a mix of clay and organic material and change from a chocolate brown color interfacial sediment to dark grey to brown. The clay remains brown and watery until about 6 cm: where the sediments become a dark brown and firm. This coloration and consistency continues until about 10 cm where the sediments become grey and soft. Below 15 cm the sediment remains grey but then becomes firmer. The brown sediments above 6 cm appear to be oxidized while the sediments below 10 cm appear to be reduced and occasionally have a slight odor of H28. Within the zone of firm sediments from 6 to 10 cm, dark brown steaks, reddish streaks, and at the base, black streaks occur. This zone appears to be the transition from the reduced sediments below and the oxidized sediments above. The highest sediment concentrations of Fe and Mn occur within this zone, which will be referred to as the REDOX layer. CHAPTER 2 METHODS 2.1 SAMPLING TECHNIQUES The JOHNSON-SEA-LINK-II (JSL) is a research submersible capable of operating at depths up to 3,000 feet. The vessel is approximately 27 feet long with a forward plexiglass sphere and an aft aluminum compartment each carrying two persons. The advantages of using a manned submersible for this work, which became evident during the 1985 and 1986 cruises, are that: (1) the researcher can properly locate a sampling area and observe and document the sample collection, (2) undisturbed box cores can be taken and (3) representative interfacial sediments can be collected. The ability of the JSL to vacuum material out of the water column and off the lake bottom was utilized to sample the fluff and was done as follows. The JSL has a lower platform which contains a hydraulically driven device for rotating 12 clear plexiglass containers (25.4 cm dia x 25.4 cm) under a suction head. The suction head is coupled to a hose connected along a mechanical arm, Figure 2. The flow rates and volumes of water through the hose are metered. 10 MECHANICAL ARM 3 HOSE —’-'i A. NOZZLE LowER» O 0 WORK PLATFORM O 0‘ SAMPLE~Q O CONTAINER FRONT COMPARTMENT THRU- HULL SAMPLING UNFILTEREO 9 -. / __. \ FILTERED WATER UNE / wATER PUMP 3 ”NE FLOWMETER FILTER / MOUNT 3.: NOZZLE CONTAINER B. Figure 2. a. Lower work platform of the submersible JOHNSON— SEA-LINK-II (JSL) showing arrangement of the sample containers, suction pump, and front compartment carrying scientists. b. Vacuum/filtration device on JSL. The vacuum nozzle is gently waved above the sediment bottom, resuspending the interfacial sediments. The sediments are vacuumed through a glass fiber filter held in a teflon filter mount. 11 Each container was outfitted with a filtration system, Steve Eisenreich and Joel Baker (1989, in press). The 20.3 cm diameter filter holders are patterned after the Millipore Corporation stainless disk holders #3165. The holder is constructed from anodized aluminum, coated with teflon and holds either membrane or glass-fiber filters. Glass-fiber filters were used in this study. After much experimentation during the 1985 and 1986 cruises, interfacial sediment samples were collected without apparent contamination from the sediments below. This was accomplished by slowly waving the suction hose over the surface of the sediment by means of the mechanical arm. By holding the nozzle pointed up, approximately 18 cm above the surface of the sediment, the interfacial sediment was gently resuspended and captured in the filter assembly. Approximately 10 minutes of pumping was required to collect a one-gram sample. At a dive site, two to three samples were collected and combined for analysis. Gravity cores (7.6 cm dia. butyrate) were taken from the R/V Seward Johnson. In addition, box cores (30.5 cm x: ’30.5 cm x 30.5 cm, Al frame) and punch cores (7.6 cm dia. butyrate) were taken at the site using the mechanical arm of the JSL submersible. Lake water was collected from approximately 1 m above the sediment surface by using the submersible’s mechanical arm to extend tygon tubing into the water column, away from the submersible. The tubing was routed into the submersible, 12 where the sample was immediately processed and measured for pH, temperature and Eh. This sample was measured for the same components as the pore water samples and was used to estimate the composition of the bottom water of the lake. 2.2 SAMPLE PREPARATION AND ANALYSIS The interfacial sediment samples were stored frozen. The cores were stored at 4°C (approximate in situ temperature) and sectioned within 2 hours of sampling under N2 at room temperature (~15°C) . The sediment/water slices were cooled to 4°C and then separated by high speed centrifuging (15,000 rpm) using a centrifuge head chilled to -25°C to keep the temperatures as close as possible to in situ conditions (Lyons et a1. 1984: Robbins and Gusting 1976) . Pore water and overlying lake water were filtered through 0.45 In filters, acidified to pH's < 2 with Ultrextm HNO3, and refrigerated until analysis. Sediments were stored frozen at -25°C until analysis. Both the interfacial sediments and the sediment samples were dried (covered) at room temperature before analysis. The partitioning of the metals in the sediments is determined by a series of sequential-chemical extractions of the air-dried sediment samples. The procedure used in this study attempts to extract only the hydromorphic phases of the sediment; that is, the phases that can interact with aqueous solutions to either take up or release metals, this excludes metals in most silicate minerals Tessier et a1. (1979). Selective chemical extractions on the sediments were 13 performed to gain insight as to what solid phases host trace metals in the sediments. The assumption in using this technique is that the various hydromorphic phases have different resistances to chemical attack. The sediment is reacted with successively harsher chemicals. After each reaction, the leachate is separated from the sediment and analyzed for metals. The sediment remaining after the chemical attacks is the detrital fraction of the sediment. Total metal concentration of the sediment would be the sum of metals in the hydromorphic phases and the detrital fraction. The extraction procedure used in this study, Table 2, is from Gephart (1982), and combines the methods of Tessier et a1. (1979) and Gupta and Chen (1975) who calibrated the extractions against known minerals. Steady-state experiments were done to confirm that the reaction times used by Gephart (1982) for five gram samples would be appropriate for the one gram sample size used in this study, Appendix B. The reaction times used did not differ from those of Gephart ( 1982) . Because the attacks may not be entirely selective for a phase (e.g. Fe-oxides), the results can only be interpreted in terms of operationaly defined phases. These phases are identified by the nature of the reaction between the sediment and the chemical rather than as a discrete mineral(s) (Long' and. Gephart 1982). For' example, metals apparently associated with Fe oxides are referred to as 14 TABLE 2. Selective chemical attack method for a 1 gram sample, modified from Gephart (1982) Hypothesized Chemical Reagents Reaction sediment fraction response of used Time attacked sediment to extraction loosely adsorbed exchangeable 1M MgClZpH 7 1 hr such as on clays [EX] room temp, 8 ml carbonates some weakly acid 1M NaOAc pH 5, 5 hr hydroxides soluble with HOAc [WAS] room temp, 8 ml Mn oxides easily 0.1M NH OH’HCl 0.5 hr reducible 0.02M 0 [ER] room temp, 25 m1 Fe oxides moderately 0.04M NHZOH'HCI 6 hr reducible in 25% (v/v) HOAc [MR] 96°C, 20 m1 Organic and oxidizable 25% H 0 pH 2 5 hr sulfides [OX] with N53, 85°C 1M NH OAc, 5 ml 1 hr disti led HZO,5 ml room temperature 15 metals associated with the moderately-reducing phase of the sediment, Table 2. Lake water, pore water, and leachate from the chemical extractions, were analyzed for Mn, Cu, and Pb by flame or graphite furnace atomic absorption spectrophotometry (Perkin-Elmer 360 with HGA 220 graphite furnace and AS 40 auto-sampler). Standard. matrices were prepared from the chemical reagents used in the selective chemical extractions. Standards and blanks for the sediment pore waters were made in double distilled H20. Analytical blanks were carried through all procedures. Coefficient of variation for the atomic absorption analyses of metals was estimated to be less than 10% from measurement of replicate aliquots of the extracts. The pH profile of the sediment was measured from a second gravity core, sectioned in air, taken at site 1388. The pH of each sediment layer was measured prior to its removal from the core by inserting a Ross spear tip electrode (Orion #816300 coupled to an Orion 407A pH meter) into the layer. Total organic carbon of the sediment was measured by wet oxidation with KZCrO7/HZSO4 following the methodology of Gaudette et al. (1974). CHAPTER 3 DISCUSSION I: METAL PARTITIONING IN THE SEDIMENTS 3.1 GENERAL RESULTS The pH is highest in the bottom water (7.40) but rapidly decreases with depth, Figure 3. pH's vary between 5.55 and 6.00 in the oxidized zone and become a nearly constant 5.95 below the REDOX layer. Organic carbon is about 3% in the interfacial sediment and decreases to about 1.5% with depth, Figure 4, with the lowest concentrations in the REDOX layer. Figures 5 a and b and Figures 6, 7, and 8 show the partitioning of Mn, Fe, Zn, Cu, and Pb respectively in the sediment. The concentrations are expressed as ppm sediment dry weight. The abbreviations (e.g. EX) refer to the hydromorphic phases as given in Table 1. TM refers to the sum of the metal concentrations in all of the phases. A 10910 concentration scale is used. The interfacial sediment sample hereafter designated 1.8., is assigned the zero depth. Zinc, copper, and lead exhibit typical anthropogenic concentration profiles in the sediment. These profiles have metal concentrations that are highest near the surface and 16 17 pH 11:1; AIALJALALILLAAI'AAJAlnjljtllllll11111 8" I—- ~ SEUIHENT _ ~ 0 _ "‘ i- Ui 7 d .— t’. a: __ c " :2 D -‘ < p. I. ~ '3 ~ _. '- F- “ I- “ F. S— I— 1"" ‘j‘Ul‘j—‘rvl'11‘1‘1"l""""'ijfj' -5 0 S 10 15 20 25 30 35 Depth in centimeters Figure 3. pH of the sediments at site 1388. The open symbol at zero depth is the overlying water. ORGANIC CARBON ILJLIIJLLIIlljlllllllllllll1111411111141l 4 — I—- \o : SEUIMENT : o .. _ a _. a 3 _ _ C “ .. o ... _. .D _.. 0’1 L Lu " (J 2" F- ._ U "‘ < r— o 2 I P “ il- I: .. _ o -——1 ~— cn 1 .. t (- .. o q “ 1 F. D — I Uh 7 'TT 1 I Y‘ V l' I 17" U I YT—trrtmt" I rYIYfYTr -S 0 5 10 15 20 25 30 35 Depth in centimeters Figure 4. Corg of the sediment at site 1388, (weight % of dry sediment). The open symbol at zero depth is the fluff sample. 18 A‘ MANGANESE 11111 ILLJILLLLIAIILlljlllllllllllllLLlll 5.3—: E— : SEDIMENT E 3 "3': A '5 u : ./‘\. E. E 3'3: .H/ : g 5 3:1 _/./ ' MR i 823d 2 *~—;;;;;;_- : g E 3 ER : u 1-3‘: C‘ '7 /--SAS : aw : q~ _ _ o 0.31V\’&V:/\//f\_//EXF _I d \/ : - .7 0 11'1' '1'1'UPUUI‘U‘U‘"U'UIUUUVIUU‘IIVVU'I -S D 5 10 15 20 25 3D 35 Depth in centimeters IRDN llllL llllljlllllllllJlllllllllljllllllll SEDIMENT U! 0..) illllillllillllilllliLlJli -. ' %\:: 2 C X WATER ’:§i7 on 23 Illlllllllliill[IIIIIIIIT 0.. TM A N H Log concentration HAS VIII UV‘IUIY‘I'VIII‘U[IVYUITTYTIUYTUIUUITT D S 10 15 20 25 3D 35 Depth in centimeters C3 I U1 Figure 5. Concentration (Log 0) of a. Mn and b. Fe in chemical phases of sediments at site 1388. Zero depth is the fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). 19 ZINC 11“; LALJlALLAlMLAlAAAAJJJAAJJAL1141111 2.31 T : SEUIMENT : (C) 1.8-: :_ .. I: I +4 -I o 1.31 L b q 3 C "i : U . m I 8 ‘ *' ~ U 0.3-: § :— )- U? 3 r 0 ~ " 4-0.21 L _‘ P- r- 3 I- .1 +- ‘D.7 — L 1" -5 D 5 10 15 20 25 3D 35 Depth in centimeters Figure 6. Concentration (Loglo) of Zn in chemical phases of sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). COPPER 23 51111 1111111lllllllllllllllJiLlJlll111111 . P—- : SEUIMENT ; C .2 :— 0 1.8 _ ,- ‘H : ................. : 8 1 3 41 DX :_ L ‘ : —————————— _ a : 5 "MR : a 0.8 1: I— - a - s 3 2 ER 2 8 0.3'3 :7 (TI :1 \ I E .3 -D.2'E \“AS :- - "\, C -0.7_" \EX #- filjitj IIIUII1UUIIUIrIITTTlI‘UYIYI‘IWIUYNT‘I -5 D 5 10 15 20 25 30 35 Depth in centimeters Figure 7. Concentration (Loglo) of Cu in chemical phases of sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals), EX (exchangeable phase), WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). 20 LEAD i.lJll lllLlJllliljllillllilllllllllilllii 2 .. :— : SEDIMENT : C —--l >— 0 106 d ’— -r-1 ._ .— ‘5 1 2 ‘ “ :3 ' - ~ ._ (I _ C _. _ w 0.8 -‘ E3 :- U 2 << - C _, 3 ._ 8 0.4-j __ 8° 3 ~ _J 0.0 A : -O.4— Iiilr IIITITIITIIIIIIIT‘IIIIIIIITWIIIIIIII -5 D 5 ID 15 20 25 30 35 Depth in centimeters Figure 8. Concentration (Loglo) of Pb in chemical phases of sediments at site 1388. Zero depth is fluff sample. TM (total hydromorphic metals) , EX (exchangeable phase) , WAS (weak acid soluble), ER (easily reducible), MR (moderately reducing phase), OX (oxidizable). 21 decrease to a constant value at depth. The changes in the relative importance of the phases in sequestering metals with depth are examined. 3.1.1 Manganese The REDOX layer (:4 cm thick) has the highest concentration of Mn, with the oxidized sediments above the layer containing more Mn than the reduced sediments below. This is typical of Mn profiles in many environments (Pedersen et a1. 1986: Callender and Bowser 1980: Burdidge and Gieskes 1983). The concentration of Mn in the OX phase of the LS. was not determined and was estimated to be approximately 50 ppm. Except in the REDOX layer, this is a typical Mn concentration in the OX phase of the sediments, Figure 5a. Since the OX phase is not as important as the MR and ER phases in sequestering Mn, significant enrichments of Mn in the OX phase of the 1.8. would have to occur to affect the total Mn concentration in the LS, Thus, Mn does not appear to be enriched in the 1.8. The relative partitioning of Mn among the phases changes with depth. In the LS. Mn is partitioned in the order: easily reducible (ER) z moderately reducible (MR) >> oxidizable (OX) z weak acid soluble (WAS) >> exchangeable (EX). Below the 1.8., but above the REDOX layer the order and magnitudes are, ER > MR >> OX > EX > WAS and below the REDOX layer, MR > ER > OX >> WAS > EX. The ER phase is the most important in sequestering Mn above the REDOX layer and 22 is expected, since the ER phase is mainly Mn-oxides. Below the REDOX layer MR becomes the dominant phase. 3.1.2 Iran The MR phase (Fe-oxides) essentially dominates the sequestering of Fe throughout the care. There is no change in the relative order (MR >> OX >> ER) and little changes in the magnitude among the phases, Figure 5b, as a function of depth. Between the 1.8. and the underlying sediments, only the relationship between the EX and ER phases changes. The EX phase sequesters more Fe in the LS. than the ER phase while below the I.S., the order is reversed. Iron concentration is highest in the REDOX layer where it is most enriched in the lower two cm. Iron occurs in a narrower band and at a lower concentrations than Mn. This is in contrast to earlier observations (Kemp et al. 1978a) and is addressed below. Concentrations of Fe in the MR phase below the REDOX layer are only slightly lower than those above the zone. There is little change in concentration above and below the REDOX layer in the other phases. 3.1.3 Zinc Total Zn concentrations show a typical "anthropogenic" profile, Figure 6. The MR, OX, WAS, and ER phases have similar profiles with enrichment factors for the 1.8. over the lower' sediments. of 3.9x, 1.8x, 5.6x Iand 7x, respectively. Enrichment factors were calculated as the metal concentration in the 1.8. divided by its concentration in. the sediment. at a Idepth where concentration becomes 23 essentially constant. Except for the WAS phase, the relative importance of the phases sequestering Zn does not change between the LS. and the sediments below. However, the relative magnitudes cf the concentrations of Zn in the different phases do change. In addition, the partitioning of Zn among the phases is somewhat sensitive to the REDOX state of the system. In the I.S., Zn is partitioned in the order, MR >> WAS z OX > ER >> EX. Below the REDOX layer the relative partitioning is the same, but the magnitudes change to MR > OX >>> ER > WAS > EX. The difference in Zn concentrations of Zn between the OX and ER phases is much less above the REDOX layer than it is below. The lesser importance of the ER phase in sequestering Zn below the REDOX layer is consistent with the lower concentrations of Mn oxides below the REDOX layer, Figure 5a. There is a slight enrichment of Zn in the REDOX layer for the phase ER and below the layer for the phases MR and OX. Since Cor is relatively depleted throughout the REDOX 9 layer Figure 4, the enrichment of Zn in this phase below the layer may indicate its presence as a sulfide mineral. The procedure used to extract metals in the oxidizable phase attacks sulfides as well as organic matter, Table 1. 3.1.4 Copper Total Cu exhibits an anthropogenic profile, but only the phases OX and MR show this type of profile, Figure 7. Although Cu in the ER, EX, and WAS phases is strongly depleted by about 3 cm into the sediment, by 5 cm, Cu in 24 these phases increases to values equal to those in the 1.8.. Copper' is highly' enriched in 'the I.S., with enrichment factors of 7.66x and 2.8x for the OX and MR. phases, respectively. In the 1.8. the relative order and magnitudes of the phases sequestering Cu are OX >> MR >>> WAS >> ER > EX. In the deeper sediments partitioning changes to OX >> MR >>> ER >>> WAS >>> EX. Copper concentrations in the MR phase in the REDOX layer are slightly higher than in the surrounding sediments. Total Cu concentrations do not show an enrichment in the REDOX layer because the Cu in the OX, ER, and WAS phases is relatively depleted in the layer. The importance of the OX phases in sequestering Cu is demonstrated in the REDOX layer, where Cu concentrations mimic Cor concentrations, 9 Figure 4 and Figure 7. 3.1.5 Lead Total Pb and. the phases MR, OX, EX, and ER. have anthropogenic profiles, Figure 8. The enrichments of the MR and OX phases in the 1.8. over the lower sediment are 20x, and 19x, respectively. Lead was not detected in the EX and ER phases below 3cm. In the I.S., the phase order and magnitudes are, MR z OX >>> EX z WAS >>> ER. In the lower sediments the order and magnitudes are generally MR > OX to OX > MR > WAS. Pb appears to be somewhat depleted in the REDOX layer, but relatively enriched below the layer in the OX: phase. Since total organic carbon does not show' an 25 enrichment at this depth, this could indicate Pb sulfide formation. 3.2 INTERPRETATION OF PARTITIONING The changes with depth in the relative importance of the phases which sequestered metals are interpreted to indicate that Pb, Cu and Zn undergo remobilization during early diagenesis. Total hydromorphic metal concentrations as shown in Figures 5 through 8, do not reveal the effect of early diagenesis on the heavy metals as clearly as the study of the individual phases. For example, the effect of the REDOX layer on the Pb and Cu concentrations would not be apparent from just the total hydromorphic metal plots. In most cases the concentrations of the metals in the LS. appear to be continue the trends of their concentrations in the sediments directly below the I.S.. For some phases (the MR and OX of Pb and the OX for Cu, for example), the decrease in the log concentration of the metal with depth can be modeled as a simple exponential decline, Figures 7 and 8. This would suggest that the 1.8. is not particularly unique , but only has higher metal concentrations than the sediments below the LS. . Such a sediment concentration profile could be caused by the exponential addition of new metal at the surface or exponential removal of the metal at depth during early diagenesis. It is possible that increases in anthropogenic inputs are largely responsible for the shape of the sediment concentration profiles of the metals. However, the 26 partitioning changes suggest that early diagenesis is also affecting the profiles. Thus, absolute shape of the profiles and elemental abundances may not be an exact record of the timing and rates of anthropogenic inputs. The 1.8. may have a unique chemistry compared to the deeper sediments. For example, Zn concentrations in the 1.8. for the phases MR, OX, WAS and ER and total hydromorphic Zn cannot be modeled as an exponential extrapolation of Zn concentrations in the sediments below the LS. , Figure 6. The 1.8. is more enriched in Zn than might be predicted from the sediment. Copper in the MR phase of the LS. appears to be deficient based on a simple extrapolation of Cu in the MR phase of the sediments below the I.S., Figure 7. We have observed this "deficiency" of Cu in the 1.8. compared to the bottom sediments in other cores (Wilson et a1. 1986). It may indicate the uptake of Cu by the MR phase during burial. Partitioning differences between the LS. and the sediment for other phases sequestering Pb, Zn, and Cu also suggest that the 1.8. has some unique chemical properties, Figures 6, 7 and 8. This is similar to what has been found in marine environments in which the interfacial sediment or "transition zone" was reported to have unique chemical properties (Balistrieri and Murray 1984: Pedersen et al. 1986). The REDOX layer has not accumulated significant excesses of Cu, Zn and Pb as might be expected by the known 27 adsorptive properties of Fe and Mn oxides (Leckie et a1. 1980: Jenne 1968). In contrast, the moderately and easily reducible phases (Fe-Mn oxides) in the oxic sediments above the REDOX layer have accumulated a large concentration of these metals, Figure 6. Iron and manganese oxides, however, are not enriched in the oxic sediments, Figure 5. The scavenging ability of the oxides is indicated in the surface sediments, but apparently not at depth. The absence of significant enrichments Zn, Cu, and Pb in the REDOX layer may be due to a lack of a significant source of these metals to the layer. For example, the Fe and Mn oxides in the surficial sediments occur where detrital organic matter is decaying. Metals associated with the organic matter can be released to the pore waters, as has been shown for marine sediments (Pedersen et al. 1986) . These metals would then be available for uptake by the oxides. The REDOX layer, on the other hand, occurs below such releases and hence is not significantly metal enriched. The slight changes in the partitioning of Pb, Cu and Zn across the REDOX layer seen in Figures 6 through 8 do suggest that the layer is slightly affecting the metal profiles and that some of the remobilized metal is interacting with the REDOX layer. The concentration profiles for Cu, Pb, Fe, and Mn from this study are generally similar in shape to those found in an earlier study by Kemp et al. (1978a), but there are some differences. Kemp et al. (1978a) measured heavy metal 28 concentrations in sediments from a gravity core (their core 5) from the Caribou sub-basin. The location was 44° 32.5' N 87°0.0' W (water depth of 313 m) which is sufficiently close to our study site to invite comparison. However, because of the differences in methodologies and locations between our study and theirs, caution must be used when comparing the concentrations and profiles. Kemp et al. (1978a) used hot aqua-regia to extract metals from the sediments. Such an extraction may be more of a measure of the total metal content of the sediment because components such as clays are attacked (Male 1977). In contrast, the chemicals used in the present study were chosen to extract only the hydromorphic phases of the sediment, as defined earlier. Since metals in lattice structures of silicates, for example, are not considered readily available to solution, the hydromorphic phases would be better indicators of the behavior of labile metals in natural systems than total metal concentrations. In addition, it is the metals in the hydromorphic phases which are most likely to interact with biota (Tessier et al. 1979). Kemp et al. (1978a) reported higher metal concentrations than those found in this study, which suggests that either their extraction scheme was more complete or that other phases were attacked. In addition, our data show that Zn exhibits an anthropogenic profile, Figure 6a, and Table 3, which was not observed by Kemp et 29 al. (1978a) . This suggests the sediment-concentration profile of Zn was obscured by the hot aqua-regia extraction. Although our sediment profiles for Mn and Fe are the same shape as found by Kemp et al. (1978a), the concentrations are different. Iron concentrations are significantly greater at site 5 than site 1388 and its concentration is more enriched in the REDOX layer, as compared to the sediment above and below, at site 1388 than at site 5. The difference in Mn concentrations between sites is not as large as it is for Fe. Site 5 still has the higher concentrations, but the relative enrichments of Mn in the REDOX layer in the two sites are similar, however. Finally, Fe concentrations are generally greater than Mn at both sites. However, at site 1388, Mn in the REDOX layer is greater than Fe. The discrepancy in Mn and Fe concentrations reported for site 1388 and site 5 may be attributed to the differences in analytical techniques already discussed. Iron is mainly in the detrital phase of the sediment (Gephart 1982) so the hot aqua-regia extraction would extract significantly more Fe than the chemicals used in the present study. The small difference in Mn concentrations between the 30 TABLE 3. Comparison of trace element data (dry weight of sediment) from site 5 of Kemp et al. (1978a) for the Caribou sub basin and site 1388 this study. Values are for top sediments and lower sediments where concentrations become essentially constant. See text for locations and methods of determination. Fe Mn % % Zone 5 1388 5 1358 Tap 5.76 .66 0.15 0.0722 REDOX1 8 1.7 5 2.2 Lower 5.56 .65 0.05 0.021 Pb Cu Zn ppm ppm Ppm Zone 5 1588 5 1555 5 1385 Top 74.9 74.5 141 161 172 135 Lower 20.5 4.2 84 35.2 137 40.3 factor3 3.7 18 1.7 6 1.3 3.4 1. Highest concentration in REDOX layer. Values for 5 estimated from figures presented by Kemp et al. (1978a). 2. Mn in oxidizable phase estimated at 50 ppm dry weight of sediment. 3. Enrichment factor calculated as the concentration in top- most sediment sampled divided by the concentration in the lower sediment where concentrations become essentially constant. 31 two sites may’ be due to this metal's presence in the hydromorphic phases (Gephart 1982) so that both analytical techniques would extract similar quantities of Mn. The sizes and chemical stratigraphies of the REDOX layers at both sites are similar. At site 1388, the REDOX layer is approximately eight cm below the sediment surface, four cm thick, and enriched in Mn throughout and Fe in the bottom two cm. Kemp et al. (1978a) reported a REDOX layer 13 cm deep, three cm thick, with Fe in the bottom one cm. If the sampling and analytical procedures of the present study and of Kemp et al. (1978a) yield results that are representative of the natural environment, these comparisons may indicate the natural chemical variability of the environment and not temporal changes, since changes in the solid phases of the REDOX layer are thought to occur on the order of hundreds of years (Froelich et al. 1979). Table 3 compares selected heavy metal data from Kemp et al. (1978a) to the data from this study. The chemical data compared are from the top-most sediments sampled in both studies, the lower sediments where the metal concentrations become essentially constant, and the REDOX layer. The relative enrichments of Zn, Cu, and Pb in the surface sediments over the lower sediments are higher in the present study, in which the I.S. was known to be sampled, than in the study by Kemp et al. (1978a). For Pb, the difference is significant. The analytical techniques used in the studies are again the most probable cause of the 32 difference. Since the analyses of the surface sediments at both two sites have yielded similar metal concentrations, the lesser extraction of metal from the bottom sediments at site 1388 results in the apparently higher enrichments at that site. The similarity in metal concentrations in the top sediments at both sites is curious and may indicate the unique chemistry of the I.S.. Lead concentrations are almost identical, while Cu is somewhat greater at 1388 and Zn at 5. If an analogy can be made to the bottom sediments, the top sediments at site 5 should contain significantly more metal than those at site 1388. Since this is not the case, we suggest that it is the presence of the I.S. , enriched in metals, in our samples that causes the apparent similarity between top sediment values in the two studies. Due to their sampling method, Kemp et al. (1978a) may not have sampled the I.S.. According to the figures from Kemp et al. (1978a), Pb and Cu in the top sediments are enriched over the lower sediments in only the top 1 cm. In the present Study the enrichment is in at least the top 3 cm. The higher enrichments and thicker surface enrichment zone in our core may further indicate that the I.S. zone was not sampled in the earlier study. Interestingly, if the top 1 to 1.5 cm of sediment was not sampled in the work of Kemp et al. (1978 a and b), their estimate of the sedimentation rate for the Caribou sub-basin could be incorrect. Adding the I.S. 33 thickness would increase the sedimentation rate, perhaps bringing it more in line with the trend of present-day and average post-glacial sedimentation rates in the other sub- basins of Lake Superior (Kemp et al. 1978b). CHAPTER 4 DISCUSSION II: PORE WATER RESULTS AND DIAGENETIC MODELS In summarizing the results of the pore water and sediment concentration profiles only the dominant hydromorphic sediment phases are shown. This allows a linear scale to used for both the concentration profiles rather than the loglo scale used in chapter 3. The I.S. and the overlying lake water samples are assigned the zero depth. The total concentration profile (TM) is the sum of metal concentration in all of the hydromorphic phases and is the same as the TM profile presented in chapter 3. 4.1 MANGANESE 4.1.1 Results Manganese is contained predominantly in the easily reducible (ER) and the moderately reducible (MR) fractions of the sediment, as operationally defined by the selective chemical extractions, Figure 9. The orange and brown sediment (REDOX layer) from the 7 to 10 cm. depth contains almost 30 times more Mn than the overlying oxic sediment layers and up to 200 times more Mn than the underlying grey sediment. Manganese concentration in the pore waters, Figure 34 35 SEDIMENT Mn wt‘% DEPTH 0 1 2 cm Irj‘lfilj'I'l"“IIUYYIIIIII WATER SEDIMENT I 3 . — : I 28 F . - E - _ : TM.... ' ; - MR --- . _ PW I q IJIJIJILALIJJIIIIilllll1‘l 0 0.25 0.50 PURE WATER Mn mg/l Figure 9. Mn pore-water and sediment concentration profiles. Values are mg/l and weight % of dry sediment respectively. ER (easily reducible phase), MR (moderately reducible phase), TM (sum of ER and MR phases), and PW (pore water). The open pore water symbol at x=0 is for Mn in the overlying water. 36 9, is generally less than 0.02 mg/l, above 10 cm, but increases to 0.58 mg/l below 10 cm. There is a small concentration peak (0.06 mg/l) at 4.5 cm. The increase in pore water Mn begins at a depth where Mn in the sediment has decreased to its lowest value. The pore water and sediment concentration profiles appear to be diagenetically linked. 4.1.2 Theory of manganese diagenesis The Mn sediment profile does not reflect anthopogenic input, but rather"the post. depositional mobility' of Mn (Davison 1982). Post-depositional mobility of Mn in sediments occurs when the Mn oxides formed in the shallower oxic sediments are buried and dissolve because of the lower pe’s in the deeper sediments (Froelich et al. 1979). The released Mn2+ diffuses upward, becomes oxidized, and is redeposited as Mn oxides. The zone of Mn oxide deposition can be in the water column, at the sediment-water interface, or in the sediment column as is the case at site 1388. Iron goes through a similar REDOX cycle (Drever 1988). The loci of Mn-oxide precipitation is determined by a balance between the downward flux of 02 and the upward flux of Mn2+. Steady state conditions will exist and Mn will build up in the sediments when the upward flux of Mn2+ is twice the downward flux of 02 (Froelich et al. 1979). During this steady-state condition, the pore water and sediment concentration profiles of Mn should resemble the idealized relationship shown in Figure 10. A. 8. WATER COLUMN WATER COLUMN IMENT-WATER INTERFACE _ OXIDIZED ZONE (Mn+2 -0) XILI ............................................ n—I'v SED‘MENT Mn OXIDATION ZONE PROFILE (Nn+2 —>MnOx) ngZ .............. REDOX BOUNDARY ............ .- Mn REDUCTION ZONE °. (M00; —>Mn+2) : sz3 ...................................... c1 PORE WATER...§ EQUILIBRIUM ZONE PROFILE Nh1-D> Figure 10. a) General zones of Mn diagenesis from model of Burdidge and Gieskes Hypothetical pore-water and sediment concentration profiles 37 (1983) and Klinkhammer of Mn during steady state diagenesis. (1980)., 38 4.1.3 Diagenetic model for manganese Burdidge and Gieskes (1983) developed a quantitative model (BG) that constrains both sediment and pore water distributions of Mn through coupled diagenetic equations. A schematic depiction of their model is shown in Figure 10, and is similar to the model discussed by Klinkhammer (1980). From the sediment-water interface to a depth L1, the sediments are oxidized and pore water Mn is held low due to the low solubility of Mn oxides. Oxidation of upward +2 occurs from L1 to L2 and the concentration of diffusing Mn Mn in the sediment increases within this depth range. Dissolved Mn also increases in this range because of the removal of oxygen. Depth L2 marks the peak Mn concentration in the sediment profile and represents the boundary between the zone of oxidation above and reduction below. Solid Mn oxide dissolves in the depth range L2 to L3 because of the lower pe’s. As a result the concentration of Mn in the pore water increases to its maximum value. Below depth L3, Mn concentration in 'the sediment profile remains at. a low constant value. Manganese in the pore water below depth 13 can remains constant with its concentration determined by equilibrium with a mineral phase other than Mn oxides. The BC model only considers the Mn oxidation and reduction zones, Figure 10. Pore ‘water and sediment Mn concentrations in the oxidized and equilibrium zones are held constant. The model also assumes that: 39 1. Steady-state diagenesis is occurring; 2. Migration of dissolved Mn2+ occurs only by molecular diffusion following Fick's laws: 3. Porosity and diffusion coefficients are constant over the depth range considered: and 4. The sedimentation rate and the supply of reducible Mn to the sediment are constant. The equations describing steady-state Mn diagenesis in the oxidation and the reduction zones are: for the oxidation zone (L1 5 x 5 L2) pore waters, 03*62an/6x2 ’ w*6an/6x - k°x*an = 0: for the oxidation zone sediments, -w*6MnS/6x + (6/1-¢)*kox*Mn = 0: P for the reduction zone (L2 5 x 5 L3) pore waters, DB*62an/8x2 - w*8an/8x + (1-¢/¢)*kred*uns =0: for the reduction zone sediments, -w*6Mns/6x - kred*Mns = 0: where DB = bulk sediment diffusion coefficient (cmZ/yr) = Do*¢2 (Lerman 1975) an = Mn in pore water (pg/cm3) Mns = Mn in sediment (pg/cm3) Mns° = Mn at sediment-water interface (pg/cm3) w = sedimentation rate (cm/yr) kox = lst order rate constant for oxidation (yr-1) red lst order rate constant for reduction (yr'l) ¢ = porosity x = depth 40 with the boundary conditions (an)ox(L1) = o (Hns)ox(L1) = (Mns)° (an)0x(L2) (unp) red(L2) (ms) ox(L2) = (Hus) red(L2) [6(an)ox/8X](L2) = [6(an)red/6XJ(L2) as x -- o, an remains finite. The subscripts 5x and reg refer to concentrations in the oxidation and the reduction zones, respectively, Figure 10. The solution (an)ox (an’red (uns)ox (unsired where to this set of equations is: = A*sinh[a(x-L1)] (1) = G - H eXPI-fl(x-L2)J (2) = (Mns)°*cosh[a(x-L1) 1 (3) = E*eXP[-fi(X-L2)] (4) = J(kox/DB) = kred/w = L2 - L1 (width of oxidation zone in cm) = [(Mns)°*(1-¢)*Wl/[a*¢*03] = (Mns) °*cosh[aLox] = A*(sinh[a*Lox] + (a/fi)*cosh[a*Lox]) = (E*W2*(1-¢))/(Rred*DB*¢) 41 4.1.4 Application of the model Although the measured sediment and pore water profiles of Mn appear to be diagenetically linked in Figure 9, their relationship is not exactly that expected from the steady state model, Figure 10. For example, the depths (L1) of the beginning of the concentration increase in the sediment and the pore water profiles are not equal. In addition, the shape of the measured pore water profile in the reduction and equilibrium zones does not resemble the steady state model. The sediment profile, on the other hand, exhibits a shape which is consistent with its formation under steady state conditions. The maximum concentration is a nearly symmetrical 'spike’. The Mn concentration in the sediment is low and constant above and below the maxima, with slightly higher concentrations above the maxima than below. Pedersen et al. (1986) used the BG model to examine the non steady state nature of Mn diagenesis in hemipelagic marine sediments from the East Pacific Rise. Pedersen et al. (1986) fit their measured pore water data to the 86 model to determine the rate constants kred and kox‘ Using these rate constants and equations (3) and (4) , the sediment concentration profile representing a steady state diagenetic relationship to the pore water profile was predicted . They compared the predicted sediment profile with the measured profile to make interpretations on the steady state nature of manganese diagenesis. 42 Table 4. Parameters used for fitting data from 1388 to diagenetic model of Burdidge and Gieskes (1983). Parameter Value (bulk sediment 78.5 cmz/yr (1) diffusion coefficient) 3 (sedimentation rate) 0.017 and 0.0055 cm/yr (2) ¢ (porosity) 0.83 (3) ps (dry density 2.65 g/cm3 (4) of sediment) (Mns)° (sediment Mn 17.9*102 pg/cm3 (5) at x=o) (1) D? was calculated from the data of Li and Gregory (1974 (2) w is the sedimentation rate from site L-42 of Kemp et al. (1978a). This location is very close (z 15 km. N) of site 1388 (this paper). 0.017 cm/yr is the average post- glacial sedimentation rate, 0.0055 cm/yr is the recent sedimentation rate at the same site. Both values were calculated from the relation _w= R/[(1-¢)*ps] Berner (1980), where R is the sediment flux rate in, mass/(area*time). (3) 6 is the average measured porosity of the 4.5 - 11 cm interval, range 0.89 - 0.79 (McKee, unpublished data). (4) after oJohnson et al. (1982). (5) (Mn) was estimated from the sediment profile of reducibIeo Mn, (ER+MR fractions of the sediment). Measured units for Mn were weig gpt %, which were converted to pg/cm3 by multiplying by ps *10 43 To constrain the cause of the apparent non steady state conditions of Mn diagenesis at our study site, the approach of Pedersen et al. (1986) was taken. The parameters used in fitting the data to the BG model are summarized in Table 4. The best fit of the measured pore water profile to the model was determined by nonlinear regression. Two sediment flux (mass sedimentation) rates (R) are given for the study site by Kemp et al. (1978b), 25 g/m2*yr for the recent rate and 80 g/m2*yr for the average post glacial rate. These sediment flux rates represent sedimentation rates (w) of 0.017 cm/yr and 0.0055 cm/yr, respectively in Table 4. Both rates were used in the diagenetic modeling. Based on the idealized pore water curve as shown in Figure 10, depths L1 and L2 were assigned 9 cm and 12 cm, respectively. Depth L1 was chosen at the beginning of the concentration increase in the pore water. Assigning depth L2 is more difficult and was chosen to approximate the inflection point of the pore water curve, similar to the methods of Pedersen et al. (1986) and Burdidge and Gieskes (1983). In the initial modeling however, the depth L2 was allowed to vary along with kred and kox' producing best fit values of L2 ranging between 11 cm to 13.75 cm. Therefore, 12 cm appears to be a reasonable depth for L2 and in the solutions for the two sedimentation rates, depths L1 and L2 were held invariant for consistency. Figures 11 and 12 show the Mn sediment profiles predicted from fitting the pore water profile to the BG 44 PURE WATER Mn mg/l SEDIMENT Mn wt 34. DEPTH O 0.5 l DEPTH D l 2 cm Cry-woa—I—i-v-vovw—I—u—u—W, 1‘r'v—0‘Yl cm r—I—vw-w-rl v11 .1 rs v '1' 11"!“v m '71 I ’ 1 1 I WATER ‘ wn'f? 0 ‘9 0 I“; ILn *3 E SEDIMENT a SEDIMEHT: ,0 .O 1 b .. 1 10»El 101:“ ‘ .1' ‘ I l 20 . 20 T, T ’ 1 IT 1 ’T T 30 f 30 :- MEASURED I 1' A B i PREDICTED— T . . 11.11.11.111.“ Figure ‘11. Results of fit of Mn pore water profile to BC model for high linear sedimentation rate (0.017 cm/yr) with L1 = 9 cm, L2 = 12 cm, and parameters from Table 3. a) Measured porelwater profile and mode311ed_ 1profile, kox = 4.18*10 yr and kre = 3.18*10 . . Measured sediment profile and predicted profile from BG model. PORE WATER Mn mg/l SEDIMENT Mn wt % DEPTH 0 D. 5 l DEPTH 0 1 2 cm Ptvv.v1vvvv7"'*1"'vtd CTN ’vanpnqan.”41.,1 WATER ‘ : WATER 1 0 v 0 'I'. SEDIMENT SEDIMENTI It. 1 : r L - i 10.. 1 —'1_— 10: _ i , —L2—— . ““:::=— J 3 » I 20 - 20 , .1 30 - 30 MEASURED. ‘ MEASURED I] A , MODEL — . B ’ PREDICTED—. ‘ .5 . .I ........ I. . . 1.1.. 0 Mullins mouAWAJJ Figure 12. Results of fit of Mn pore water prorile to BC model for low linear sedimentation rate (0.0055 cm/yr) with L1 = 9 cm, L2 = 12 cm, and parameters from Table 3. a) Mea ured pore water profile rapd modelled profile, kox = 13.2 yr and k = 2. 76*10 . b) Measured sediment profile and predicEea profile fromy BG model, dashed line is sediment profile predicted from k0 = 149 yr 1,k = 7. 75*10 , and w = .0015 cm/yr. Pore water profiTed resulting is the same as a). 45 model for the Ihigh sedimentation (0.017 cm/yr) and low sedimentation (0.0055) cm/yr rate, respectively. In neither case does the 36 model properly define the shape of the pore water curve at depth. After an initial rapid rise, there is a change in slope and the measured Mn increases linearly with depth. The 86 model, predicts a curvilinear increase. The fit between the measured profile and the predicted profile is good in the oxidation zone, but degrades in the reduction zone. An identical mismatch between the measured Mn pore water profile and modelled pore water profile was also found by Pedersen et a1. (1986). The 86 model is based on the hypothesis by Froelich et al. (1979) that once steady state conditions are reached, the peak concentration of Mn remains constant. The 86 model, therefore, should predict the maximum and total mass concentration of the Mn in the spike. The concentration of Mn in the spike is not predicted by the use of either sedimentation rate. The high sedimentation rate does not predicts a zone of Mn build up, while the low rate predicts a fraction (<1/3) of the build up found. In addition, the predicted peak using the low sedimentation rate, Figure 12 occurs 4.5 cm below the depth of the measured peak. Pedersen et a1. (1986) also found the predicted Mn concentration peak to occur below the measured peak, but the concentration was higher in the predicted peak than in the measured peak. A cause of the lack of agreement between the model predictions and measured profiles is the potential error in 46 assigning depths for L1 and L2 from the pore water profiles. Differing from the Mn sediment profile in the Pedersen et a1. (1986) study, at site 1388 the Mn sediment profile is well defined and approximates the idealized profile shown in Figure 10. Values for L1 and L2 can easily be assigned from this profile. Therefore, we fit the sediment profile to the 36 model to determine what constraints could be placed on the pore water profile, a reverse of the Pedersen et a1. (1986) approach. Calculations were again made using both sedimentation rates, only in this case equations (1) and (2) were used to predict the pore water profiles. Only the sum of Mn concentrations in the MR and ER phases were used in the modeling because these phases are the phases most likely to participate in the REDOX reactions . Values for L1 and L2 were assigned 4.5 cm and 7.5 cm, respectively. Depth L1 is the depth at which the increase in sediment Mn begins, depth L2 is the center of the peak. Figures 13 and 14 show the pore water profiles predicted by the fit of the sediment concentration profiles to the BG model for the high sedimentation and low sedimentation rates, respectively. For both rates, the BG model fits the measured sediment profiles well. In neither case, however does the predicted pore water profile match the measured profile in concentrations or shape. The predicted pore water concentrations are much higher than those measured, although they are somewhat closer to the measured concentrations when the low sedimentation rate is 47 SEDIHENT Mn wt 7. PURE NATER f-‘n Pas/l DEPTH D 1 2 DEPTHE] 1 2 3 4 C m :rYT‘Y‘TTT‘T'TT‘ITWW‘TTT‘r—TY 1171 r1} C m :T“ " *‘3‘1"‘*“"7"""’*"”’""‘.’1 u J E WATER 3 , wan; 0 :‘G + D g -... _. :3 SEDIMENT< : 3:01.— 57 . ' ‘ I _ K. 4 —L1—- .- C "~\\. . I\\~\_ * ',_..-—-—-—-——'—='— . “L2”— .0 ‘ --—\_\ , « ). I 10:{ 1 10 g T j I 4 C 1 ’ I I 3 . 20 : 1 20 . j u C MEASUREDI 3 ’ ' 30 'j 30 ’ I ‘ _ MDDEL —- : » MEAS:?EC? I 1 P “22751—2 A' 11111nxnilinnillnlLLljinl-1 B. .ii.l....RE.:.ii.iLlth Figure 13. Results of fit of Mn sediment profile to BG model for high linear sedimentation rate (0.017cm/yr) with L1 = 4.5 cm, L2 = 7.5 cm, and the parameters from Table 3. a) Measured sediment profile and modelled profile, kox = 156 yr" and kred = 3.58*10' . b) Measured pore water profile and predicted profile from 86 model. SEDIMENT Mn wt % PDRE WATER .er ”TE/'1 DEPTH D l 2 DEPTH D D. 5 1 1. E Cm hIIIIIIIIYIT‘I’IY1YTYYI‘I‘UYIIJ Cm *1 'j ' ' I ' V V V . ' ,4 p 4 > 1 ; WATER f 1 «473R 0 7’6 4 D G I: SEDIMENT; ;: ggggygur . i . I #- .. —L]"_ .- . ' ‘ ' a J 10 ’ '1 10 r.- I 4 > ' I 1 I I . I i 1 ‘ c . I . . 4 20 _ 1 20 . 1 i . 4 1 . 1 : MEASURED I 1 . . 30 ,,. 1 30 F . 1 - MODEL "‘1 ; MEASUQED I ; . s L 9229::755 —-: A' 1111114111111111111111111 Bo L14.1_1..L-l_1_1_1_1_'....--1._1_4._.l.. Figure 14. Results of fit of Mn sediment profile to 86 model for low linear sedimentation rate (0.0055 cm/yr) with L1 = 4.5 cm, 12 = 7.5 cm, and parameters from. Table 3. a) Measured sediment pro£dle_apd modelled profile, kox =156 yr- and k ed = 1.16*10 yr . b) Measured pore water profile and pre§1cted profile from BC model. 48 used. The model predicts that the steady state concentrations of Mn in the pore water should be reached at about 9 cm with a rapid rise to the steady state value beginning at about 5.5 cm. The measured pore water values do not appear to approach steady state until about 30 cm and with the concentration increase beginning at about 10.5 cm or 5 cm below the predicted depth. 4.1.5 Discussion of modelling results In summary, Mn diagenesis does not match the steady state model at the study site in the Caribou sub basin of Lake Superior. The Mn pore water profile appears to be displaced downward from its thoretical steady state relationship with the Mn sediment profile. The perched REDOX layer may reflect a relict Mn depositional horizon, with Mn deposition now taking place below it. Further evidence for non steady state. diagenesis is indicated by' the slight increase in pore water Mn at 4.5 cm depth, Figure 9, which should not occur if the sediment above the REDOX layer is oxidized, as indicated by its visual appearance. The concentration of Mn in the pore water at 4.5 cm could result in a downward flux of Mn to the REDOX layer causing, in part, the build up of Mn in the layer. Such a Mn flux is not considered in the 86 model. If important, this flux would cause the pore water fit to the 86 model to under estimate the concentrations in the sediment profile and the sediment profile fit to over estimate the pore water concentrations. Although this was found to be the case in 49 these calculations, the downward flux would supply less than 1/5 of the total flux of Mn to the REDOX layer. Therefore, the downward flux is the not the cause of the non steady state condition, but only an indication of it. The timing and cause of the decoupling of the pore water profile from the sediment profile is unclear. However, the pronounced shape of the concentration peak for Mn and apparent lack of visual evidence of a new depositional zone below the REDOX layer, suggest that the shift from steady state has been recent. The shift probably represents an increase in the 02 flux to the REDOX layer, a decrease in the Mn2+ flux, or both. This can be demonstrated by varying the sedimentation rate on the model predictions. The present sedimentation rate in the Caribou sub basin is not well constrained. The sedimentation rate cited by Kemp et al. (1978b) may have been underestimated if the true surface sediments were not sampled. Additionally, their rate is based on the depth in the sediment of the last occurrence of Ambrosia pollen. Since Ambrosia was only detected in the first one cm section taken, the actual depth of the Ambrosia within this cm section is unknown and thus the sedimentation rate could be over estimated. The formation of the Mn sediment concentration profile under a sedimentation rate lower than reported is consistent with the observation that decreasing the sedimentation rate, increases the Mn concentration in the REDOX layer predicted by the 86 model, Figures 11 and 12. Therefore, the 86 model 50 was used to calculate the sedimentation rate needed to produce the peak concentration of Mn in the study core, assuming present day 02 and Mn2+ fluxes. The result of this calculation is a sedimentation rate of 0.0015 cm/yr, Figure 12. Even though the highest concentration of Mn is matched using this sedimentation rate, the total mass of Mn in the REDOX layer is over estimated, as evidenced by the greater width of the predicted REDOX layer than the measured layer. To be able to correctly model the peak at this low sedimentation rate, the oxidation and reduction rates would have to be increased from their present values. Thus, these calculations demonstrate that changes in sedimentation rate alone cannot account for the apparent non steady state Mn diagenesis. Also the present oxidation and reduction rates of Mn can not account for the Mn concentration at this site if the REDOX layer was produced under steady state conditions. The 02 and Mn2+ fluxes must have changed in a direction which has slowed the oxidation rate in recent times. Since a sedimentation rate of 0.0015 cm/yr is rather low compared to rates in other depositional areas of Lake Superior (Johnson et al. 1982), a higher sedimentation rate, such as reported by Kemp et al. (1978b), might be expected for the Caribou sub basin. This implies that the oxidation and reduction rates during the formation of the concentration profile of Mn might have been significantly higher than at present. 51 Pedersen et al. (1986) interpreted the downward displacement of the Mn pore water profile with respect to its steady state relationship with the sediment profile to increased penetration of 02 into the sediment. They attributed the increased 02 supply to be due to decreased productivity in the water column causing less Corg sedimentation and therefore less oxygen consumption at the sediment water interface. Lower concentrations of organic carbon in surface sediments compared to deeper sediment was interpreted to be evidence for the decreases productivity. My profile and other profiles for organic carbon in Lake Superior do not support recent decreases in productivity (Johnson et al 1982). Other factors causing changes in the 02 and Mn2+ fluxes must be considered. However, with the data available, the cause of the apparent increase of 02 in pore water below the REDOX layer at site 1388, or the permanency of the increase, can not be determined. 4.2 COPPER AND LEAD 4.2.1 Results The sediment profiles of Cu and Pb exhibit apparent anthropogenic signatures in which their total hydromorphic concentrations (TM) are greatest near the surface and decrease with depth to a relatively constant value, Figure 15. Both metals are significantly enriched in the I.S. layer over their' concentrations in. ‘deeper sediments. The moderately reducible (MR) and oxidizable phases account for 52 SEDIMENT Cu pg/g SEDIMENT Pb .pg/g DEPTH D ‘7D 140 210 DEPTH D SD 100 cm J ...... , ...... , ...... . cm . .. I ...,....,....,J E WATER 3 E 3 Eli—P. - j 0, ; SEDIMENT} 10 3 u. { 10 :3 p . E ’ . E i 20 I 1 20 :3 . L . > . ' rl. ' 1 30 E 3 3D ETTM 3 3 3 3' x 'Pw 1 ; : :MB . A. l iiiiii l 111111 1141114.‘ B. pliJlllllAJllljlllllll‘ D 10 2D 30 D S 10 PURE WATER CU _ug/] PURE WATER Pb .Dg/l Figure 15. Pore-water and sediment concentration profiles of a) Cu, and b) Pb. Values are in units of pg/l and pg/g of dry sediment respectively. OX (oxidizable phase); MR (moderately reducible phase); TM (sum of WAS, EX, ER, MR and OX phases); and PW (pore water). The open pore water symbol at x=0 is the concentration in the bottom water. 53 most of the Cu and Pb in the hydromorphic fraction, although the relative importance of these two phases in sequestering the metals changes with sediment depth. In the I.S., the 0X phase sequesters the most Cu, while both the 0X and MR are of equal importance in sequestering Pb. In the sediments below the I.S., the MR phase becomes relatively more important in sequestering Cu, while MR and 0X phases remain nearly equal for Pb. A similar change in Cu partitioning between the I.S. and the sediments below the I.S. has been found in other areas of Lake Superior (Wilson et al. 1986). Concentrations of Cu and Pb in the REDOX layer are slightly higher than in the sediment above and below this layer. The pore water concentrations of Cu and Pb near the sediment-water interface are higher than in the lake water, Figure 15. This indicates the release of these metals from the sediment to the pore water as a result of their post- depositional remobilization. Both metal profiles are similar to the Mn pore water profile, (Figure 9). Pore water concentrations increase between 2.5 and 4 cm, decrease from 4 to 10 cm, increase between 10 and 14 cm, and increase or remain constant below 15 cm. However, concentration maxima are not always at the same depths. For example, the pore water maxima for Pb near the top of the core occurs at 2.5 cm, while the maxima for Cu is between 3.5 and 4.5cm and for Mn at 4.5 cm. This indicates that. different mechanisms affect the release of the metals from the sediment. 54 The similarity of the pore water profiles of Cu, Pb, and Mn in sections of the core and the slight enrichments of Cu and Pb in the MR phase (metals associated with Fe oxides) in the REDOX layer, indicate that the geochemical behavior of Cu and Pb is in part related to the Mn and Fe REDOX cycles. These results are similar to observations on trace metal behavior in marine hemipelagic sediments (Graybeal and Heath 1984). However, the lack of a direct correlation between the. Pb and Mn maxima in the first 5 cm and broad nature of the Cu maxima in the same zone, indicate that processes other than REDOX reactions are affecting the geochemical behavior of Pb and Cu. We suggest that these metals are released during the decomposition of organic matter. Such a process has been shown to be important in marine sediments (Gobeil et a1. 1987; Ridgeway and Price; 1987; Westerlund et a1. 1986; Fischer et a1. 1986; Klinkhammer 1980: Sawlan and Murray 1983)). At this site, organic carbon in the sediment decreases from approximately 3%, by weight in‘ the I.S. to 1.5% between 3 and 4.5 cm., Figure 4. Decreasing concentrations of organic carbon in sediment of Lake Superior with burial has been interpreted to be due to carbon loss from decomposition (Johnson et a1 1982). The most rapid decrease of organic carbon occurs between the I.S. and 2 cm into the sediment, which is also the interval for the most rapid decrease of Cu and Pb in the sediment. The increase in the pore water concentration of Cu and Pb, overlaps with their decrease in 55 the sediment. Copper appears to be lost from the 0X phase during burial while lead is lost from both the MR and 0X phases. These observations are consistent with a model for metal release to the pore waters during the decomposition of organic matter. 4.2.2 Flux model for copper and lead The pore water concentration gradients of Cu and Pb near the sediment-water interface suggest that these metals could be diffusing out of the sediments to the overlying lake water. The potential importance of this diffusive flux (Ed) on the cycling of Cu and Pb at the sediment-water interface was estimated from their pore water gradients. These fluxes were then compared to an estimated particle bound metal flux (Fs) or the rate of hydromorphic metal addition to the sediment-water interface from the water column. The diffusive flux (Fd) was estimated using Fick’s law; Pa = -¢*DB*d[Me]p/dx, where ¢ = average porosity of the sediment over Ax, DB = bulk sediment diffusion coefficient (cmz/yr), and d[Me]p/dx = measured pore water concentration gradient (pg/cm3)/cm. This use of Fick's law assumes that DB is constant over the depth range of interest and neglects burial of pore water and compaction. The concentration gradients (d[Me]p/dx) are considered linear from the sediment-water 56 interface to the maxima in the upper portion of the pore water profiles, Figure 15. Diffusion coefficients (D°*) for 4°C are calculated from the data of Li and Gregory (1974) assuming a linear change of the diffusion coefficients at infinite dilution (Do) between 0° and 18°C. The solute metals are assumed to Ibe free ions. The bulk. sediment diffusion coefficient is calculated as DB = ¢2*Do* (Lerman 1975). Because (1) no visible evidence of bioturbation was detected in the core and (2) the effect of bioturbation on metal migration in Lake Superior is considered negligible (Johnson and Eisenreich 1979, Eisenreich personnel communication 1988), bioturbation is not considered in these calculations. The particle bound metal flux (F8) is estimated as the mass sedimentation rate (R) times the total hydromorphic metal concentration in the incoming sediment particles [Me]°. The mass sedimentation rates used are taken from Kemp et al (1978b) and also from the results of the Mn modeling in this study. The ratio of the diffusive and particle bound metal fluxes [(Fd/Fs)*100%] indicates the fraction of hydromorphic sediment bound Cu and Pb that is remobilized within the upper sediments (Callender and Bowser 1980). Because interfacial sediment (I.S.) is frequently not collected, errors may result in the calculation of diagenetic fluxes (Berner 1980). We evaluated the potential error in flux estimations by calculating two values for the bound metal flux, F One flux, F includes the metal 8’ S' 57 Table 5. Parameters used in the calculation of the fluxes of Cu and Pb. A porosity of 0.93 is assumed (see appendix, Table A). Concentrations are in pg/g, fluxes are in #9/0m2*Yr- Element R 2 [Me]o [Me]o F Fs (g/cm *yr) (I.S.) (0-1cm) (§.s.) (0-1) Cu 8*10-3 161 136 1.30 1.10 Cu 2.5410‘3 161 136 0.40 0.34 Cu O.68*10'3 161 136 0.11 0.09 Pb 3410"3 74.5 42.2 0.60 0.34 Pb 2.5410"3 74.5 42.2 0.19 0.11 Pb O.68*10'3 74.5 42.2 0.057 0.029 Element d[Me]/gx dx DB 2 Fd 2 (pg/cm ) (cm) (cm /Yr) (pg/cm *Yr) Cu 1.5410’5 3.5 109 0.15 Pb 1.7410‘3 2.5 146 0.23 R sediment flux rate [Me] concentration in I.S. or 0-1 cm interval as noted Fs, gs, sediment bound metal flux calculated using I.S. and 0-1 cm interval respectively. d[Me]p/dx concentration gradient of pore water metal over interval dx in top sediments DB diffusion coefficient of pore water metal adjusted for tortuosity - Fd estimated diffusive flux of pore water metals up to sediment-water interface 58 concentration in the I.S. as [Me]o while the other, F 5’! uses the concentration in the 0-1 cm. interval of the sediment core as [Me]o. The latter value would be typical of calculations made not including the I.S.. Table 5 lists the parameters used in the calculations, the estimated bound metal fluxes (F8 and Fs’) , and the diffusive fluxes of Cu and Pb. Estimated diffusive fluxes for Cu and Pb of are 0.15 and 0.23 pg/cm2*yr, respectively. The flux for Cu is comparable to fluxes of Cu in marine hemipelagic sediments. For example, Sawlan and Murray (1983) estimated from pore water profiles at 6 hemipelagic sites in the equatorial Pacific, that diffusive fluxes (Fd) of Cu from sediments ranged between 0.17 to 0.46 pg/cm2*yr. The data presented in table 5 illustrates that the calculation of the metal sedimentation rates (F are s) significantly affected by the value of the sedimentation rate. Furthermore, by not considering metal concentrations in the I.S., Fs is under estimated by 20% and 50% for Cu and Pb, respectively. The results in Table 6 show that significantly more Pb is remobilized than Cu and that when mass sedimentation rates are less than 25*10-3g/cm2*yr the amount of Pb remobilized becomes significantly greater than 100%. If 25*10-3 g/cm2*yr is representative of the mass sedimentation rate in the Caribou sub basin, then the amount of remobilized Cu and Pb is 38% and 121%, respectively. Thus the fraction of Cu and Pb remobilized in the hemipelagic 59 Table 6. Per cent of metal remobilized from surface sedimen s for various sedimentation rates at site 1388. R is in g/cm *yr. % remobilized = (Fd/Fs)*100% Sediment flux 8*10‘3 2.5410"3 0.68*10-3 rate (R) Element Cu 12% 38% 136% Pb 38% 121% 404% Fd = estimated diffusive flux from Table 4 Fs = estimated sediment bound metal flux, I.S. only, for various possible sediment flux rates discussed in text 60 sediments of Lake Superior is similar to amounts remobilized in hemipelagic sediments in the ocean. For example, Fischer et a1 (1986) estimated that 45% and 100% of the Cu and Pb, respectively, are recycled in hemipelagic sediments at site H of the MANOP program. Allowing for uncertainties in the mass sedimentation rate, remobilization of Cu and Pb appears to be significant at site 1388 in the Caribou sub basin of Lake Superior . The fate of the remobilized metals is unclear. They could be adsorbed by the I.S. sediments before escaping to the water column, adsorbed by the sediments settling to the sediment surface, or taken up by biota. Regardless of their fate, the amount of metal remobilization suggests that sediment concentration profiles of Cu and Pb at this site may not indicate either the timing or amounts of the anthopogenic input of metal to Lake Superior. CHAPTER 5 CONCLUSIONS 5.1 SUMMARY AND CONCLUSIONS Studies of sediments and pore waters at a site in the Caribou sub-basin, Lake Superior, were made to evaluate the effects of early diagenesis on trace metal behavior. Selective chemical extractions on the sediments were made to discern what hydromorphic phases are controlling metal behavior. Diagenetic modeling and calculation of fluxes across the sediment-water interface were used to quantify the extent of metal remobilization between the sediments and the pore waters and the overlying lake waters. Pb, Zn, Cu, Fe, and Mn concentrations were measured in the hydromorphic phases of sediments collected by gravity core and compared to their concentrations in surface sediments (I.S.) collected by manned submersible. The submersible allowed a representative sampling of the sediment-water interface. Mn, Cu, and Pb were measured in pore waters at the same site. The following conclusions can be made: 1. A transition zone of interfacial sediment 0.5 to 1.5 cm thick exists between the particles settling through the water column and the sediment column. This interfacial 61 62 material is referred to as the 'fluff’ because of its in situ appearance. It resembles the transition zone or interfacial sediments found in marine environments in having elevated concentrations of organic carbon and trace metals such as Cu and Pb. 2. The dominant phases sequestering the metals are the moderately reducible (metals associated with Fe and Mn oxides) and. the: oxidizable (organics and sulfides), but their importance is different for each metal. The moderately reducible phase sequesters the most Zn, the oxidizable phase the most Cu, and the oxidizable and the moderately reducible phases both sequester most of the Pb in subequal proportions. 3. The I.S. is highly enriched in Pb, Cu, and Zn compared to the lower sediments, but is not enriched in F6 and Mn. The relative importance of phases sequestering the metals changes from the I.S. to the sediments below the I.S. . This suggests that the metals are undergoing remobilizat ion during early diagenesis . Diagenetic remobilization of the metals is only evident when the individual hydromorphic phases are studied. Study of total metals or total hydromorphic metals in the sediments obscures such interpretations of metal behavior. 4. In most cases the concentrations of Pb and Cu in the various phases of the I.S. are an exponential extrapolation of their concentrations in the sediment below the I.S. . However, there are chemical differences, such as the Zn 63 concentrations in the I.S. and the sediment below, that suggest the I.S. may have some unique properties. 5. The general sediment concentration profiles of Pb, Cu, and Zn are similar to typical "anthropogenic" profiles in lakes. The Zn "anthropogenic" profile was not found in earlier studies. A comparison of the Pb and Cu profiles to an earlier study suggests that in the earlier study the I.S. was not sampled. 6. The concentration profiles of Fe and Mn reflect early diagenetic processes. Mn and Fe oxides in the REDOX layer have not accumulated significant excesses of Cu, Zn and Pb, even through such accumulations occurs in the surface sediments. This was interpreted to indicate that the decay of organic matter in the surface sediments supplies metal to the oxides near the sediment-water interface , but not to those at depth. 7. Manganese pore water and sediment concentration profiles indicate its diagenetic remobilization. The profiles are decoupled from their theoretical diagenetic steady-state relationship. Although recent changes in sedimentation rates could account for some of the decoupling, changes in the fluxes of 02 and Mn2+ must also have occurred. These changes must be in a direction to cause a deeper penetration of 02 into the sediments and a reduced rate of Mn2+ oxidation. 8. The concentrations of Cu and Pb in the sediment decreases with depth. Their pore water profiles indicate a 64 potential diffusive flux of these metals from the upper sediments to the overlying lake water. These upward fluxes may be nearly as large as the downward flux of sediment bound Cu and Pb and suggest substantial remobilization of Cu and Pb is occurring in the sediment column. Metal released during decay of organic matter may be the principle source for these metals to the pore water. The amount of Cu and Pb remobilization during early diagenesis is large at this site and may prevent interpretation of their sediment concentration profiles as historical records of metal inputs to the lake. 9. The cycling of Mn, Cu and possibly Pb within the hemipelagic. sediments of' the Caribou sub basin in Lake Superior is analogous to that observed in hemipelagic marine sediments. The cycling is similar in terms of benthic fluxes, sediment and pore water profiles, amounts of metal remobilizations during early diagenesis, and source of the released metals. 65 5.2 RECOMMENDATIONS FOR FURTHER WORK Some very important questions about trace metal behavior in Large Lakes are left unanswered by this study. The spatial variability should be examined at a greater number’ of sites in the Great Lakes, and, possibly some similar marine settings. Pore water and solid phase data should be examined together in a similar fashion to this study. The general behavior of the trace metals could be atypical at this one site. More attention should be given to determining the flow of metals within the lake as a whole. This should involve capturing suspended and/or settling particles from the water column and examining them by methods similar to those used on the bottom sediments. This would allow better mass balance models to be developed than the simple ones used in this study. The diffusive fluxes of copper and lead out of the sediments appear to be important factors in the budget of these metals in the bottom sediments. In the whole lake budgets they might not be; but these facts are yet to be determined. These recommendations, and others that are not directly related to the topic of this thesis, are among the goals of the continuing N.O.A.A./N.U.R.P. Great Lakes of the World research ongoing here at M.S.U. I am fortunate to have the chance to address these unanswered questions as part of this ongoing program. APPENDICES 66 APPENDIX A. SEDIMENT AND PORE WATER DATA Table A. Porosity and organic carbon concentration at site 1388. sample depth porosity organic carbon (C!!!) (wt%) I.S. 0.0 N.M.l 2.84 6-1 0.5 0.96 2.43 6-2 1.5 0.91 2.18 6-3 2.5 0.92 1.53 6-4 3.5 0.91 1.50 6-5 4.5 0.89 1.40 6-6 5.5 N.M. 1.08 6-7 6.5 0.85 0.80 6-8 7.5 N.M. 0.96 6-9 9.0 0.83 1.13 6-10 10.0 0.79 1.48 6-11 11.0 0.79 1.41 6-12 12.0 0.82 1.63 6-13 13.5 0.78 1.40 6-14 15.5 0.84 1.51 6-15 18.0 0.80 1.54 6-16 21.0 0.83 1.67 6-17 24.0 0.73 1.30 6-18 27.0 0.77 1.43 6-19 30.5 0.75 1.31 (1) N.M.=not measured. Table 67 B. Iron in sediment fractions site Concentrations in pg/g of dry sediment. sample depth EX WAS ER MR OX (cm) I.S. 0.0 257.0 5.6 47 5660 1580 6-1 0.5 43.0 5.4 96 6240 1120 6-2 1.5 38.2 7.2 89 6460 1180 6-3 2.5 18.6 6.2 84 6210 1480 6-4 3.5 31.2 3.8 89 7600 1560 6-5 4.5 12.3 5.0 110 6010 1580 6-6 5.5 143.0 3.4 142 5820 1450 6-7 6.5 103.0 3.8 129 5890 1170 6-8 7.5 218.0 84.2 291 14900 2400 6-9 9.0 60.0 10.0 198 7950 2260 6-10 10.0 N.M 4.2 58 5450 1880 6-11 11.0 5.9 1.4 78 4060 955 6-12 12.0 8.8 3.4 115 5030 997 6-13 13.5 2.5 6.0 92 4860 1230 6-14 15.5 N.M 7.2 94 5460 1330 6-15 18.0 2.7 5.6 67 5100 987 6-16 21.0 25.0 11.0 189 5410 1670 6-17 24.0 25.7 2.6 55 4990 1250 6-18 27.0 4.4 1.6 96 4960 1530 6-19 30.5 5.8 0.0 82 4970 1520 (1) N.M.=not measured. 1388. 68 Table C. Manganese in sediment fractions site 1388. Concentrations in pg/g of dry sediment. sample depth EX WAS ER MR OX (cm) I.S. 0.0 5.76 45.1 354 320 N.M.1 6-1 0.5 46.8 24.0 380 295 N.M. 6-2 1.5 37.4 20.0 442 295 N.M. 6-3 2.5 29.2 22.0 475 295 54.0 6-4 3.5 29.8 21.0 527 405 64.4 6-5 4.5 25.8 24.0 579 345 78.5 6-6 5.5 5.60 7.3 1980 875 164 6-7 6.5 2.52 5.9 2380 1080 155 6-8 7.5 4.96 N.M. 13600 7520 649 6-9 9.0 4.08 6.7 1260 503 114 6-10 10.0 1.04 4.3 28 70 39.4 6-11 11.0 0.24 3.6 42 45 22.9 6-12 12.0 3.00 6.1 46 50 26.0 6-13 13.5 0.52 3.5 48 50 23.7 6-14 15.5 10.5 14.0 31 90 30.8 6-15 18.0 0.24 3.9 46 85 23.7 6-16 21.0 5.52 N.M. 62 85 33.4 6-17 24.0 0.52 3.3 44 110 32.2 6-18 27.0 0.60 6.7 68 95 33.1 6-19 30.5 1.12 6.1 65 105 37.9 (1) N.M.=not measured. Table 69 D. Copper in sediment fractions site Concentrations in pg/g of dry sediment. sample depth EX WAS ER MR OX (cm) I.S. 0.0 1.4 4.8 2.3 37.9 115.0 6-1 0.5 0.8 4.6 6.6 45.9 77.7 6-2 1.5 0.5 0.4 1.8 20.5 37.6 6-3 2.5 <0.1 0.6 0.6 10.8 17.2 6-4 3.5 <0.1 0.8 1.0 11.5 17.6 6-5 4.5 N.M.1 1.2 1.9 14.6 16.6 6-6 5.5 <0.1 2.4 2.9 17.0 15.5 6-7 6.5 1.7 3.6 2.6 16.9 13.0 6-8 7.5 0.2 1.4 2.7 20.2 8.6 6-9 9.0 1.0 3.0 1.0 14.7 13.4 6-10 10.0 N.M. 4.2 1.4 14.1 21.0 6-11 11.0 <0.1 1.4 2.3 14.5 15.8 6-12 12.0 0.2 <0.1 2.3 14.8 18.3 6-13 13.5 0.3 <0.1 2.3 15.0 15.8 6-14 15.5 N.M. 1.4 2.9 15.3 20.6 6-15 18.0 0.3 2.4 3.4 13.3 14.9 6-16 21.0 N.M. 4.4 3.3 14.1 22.7 6-17 24.0 0.3 1.0 2.2 12.9 18.5 6-18 27.0 0.5 1.6 4.3 12.1 21.4 6-19 30.5 0.2 0.8 3.0 11.8 23.7 (1) N.M.=not measured. 1388. Table 70 E. Lead in sediment fractions site Concentrations in pg/g of dry sediment. sample depth EX WAS ER MR OX (cm) I.S. 0.0 6.4 7.0 1.3 31.1 28.7 6-1 0.5 4.0 3.0 0.8 18.9 15.5 6-2 1.5 <0.5 1.0 <0.3 8.2 7.4 6-3 2.5 0.8 <0.4 <0.3 3.5 3.0 6-4 3.5 <0.5 <0.4 <0.3 2.1 1.9 6-5 4.5 N.M. <0.4 <0.3 1.9 2.3 6-6 5.5 <0.5 <0.4 <0.3 2.3 2.3 6-7 6.5 <0.5 <0.4 <0.3 2.7 1.9 6-8 7.5 <0.5 <0.4 <0.3 <0.2 <0.2 6-9 9.0 <0.5 <0.4 <0.3 1.9 1.5 6-10 10.0 N.M. <0.4 <0.3 1.8 2.7 6-11 11.0 <0.5 <0.4 <0.3 2.6 1.5 6-12 12.0 <0.5 <0.4 <0.3 2.7 1.5 6-13 13.5 <0.5 1.0 <0.3 2.1 2.1 6-14 15.5 N.M. <0.4 <0.3 2.6 1.9 6-15 18.0 <0.5 <0.4 <0.3 2.4 3.0 6-16 21.0 N.M. <0.4 <0.3 2.1 2.7 6-17 24.0 <0.5 2.0 <0.3 1.4 2.1 6-18 27.0 <0.5 1.0 <0.3 1.4 2.3 6-19 30.5 <0.5 <0.4 <0.3 1.5 2.3 (1) N.M. = not measured. 1388. 71 Table F. Zinc in sediment fractions site 1388. Concentrations in pg/g of dry sediment. sample depth EX WAS ER MR OX (cm) I.S. 0.0 1.9 28.0 14.7 77.3 22.0 6-1 0.5 6.1 5.7 9.6 52.0 12.1 6-2 1.5 3.0 2.6 5.9 38.7 8.6 6—3 2.5 1.3 1.0 3.0 25.0 6.9 6-4 3.5 1.3 1.0 2.8 23.3 11.2 6-5 4.5 N.M.l 1.0 2.9 21.7 10.8 6-6 5.5 0.5 0.7 5.0 21.5 10.6 6-7 6.5 0.6 0.6 4.2 19.8 9.2 6-8 7.5 0.9 N.M. 8.4 22.7 8.8 6-9 9.0 1.2 0.3 1.3 28.5 21.4 6-10 10.0 N.M. 0.3 0.9 24.1 17.1 6-11 11.0 0.3 0.5 2.0 22.7 12.1 6-12 12.0 0.5 0.3 1.6 23.1 11.9 6-13 13.5 0.6 0.4 1.4 22.9 12.7 6-14 15.5 N.M 0.5 0.9 23.1 14.5 6-15 18.0 0.8 0.8 1.0 21.1 11.5 6-16 21.0 N.M. N.M. 1.8 24.2 15.1 6-17 24.0 0.3 0.5 0.5 19.7 13.3 6-18 27.0 0.2 1.1 1.3 19.1 14.4 6-19 30.5 0.5 0.6 1.1 19.7 15.3 (1) N.M.=not measured. 72 Table G. Pore water metals site 1388. sample depth Mn Cu Pb (cm) m9/1 #9/1 #9/1 B.w.1 0.002 2.8 0.81 6-1 0.5 0.010 3.3 0.87 6-2 1.5 0.007 4.5 1.32 6-3 2.5 0.018 4.6 5.40 6-4 3.5 0.023 8.7 1.62 6-5 4.5 0.059 8.0 N.M.2 6-6 5.5 N.M. N.M. N.M. 6-7 6.5 0.016 4.0 0.97 6-8 7.5 0.008 3.6 0.89 6-9 9.0 0.005 6.6 1.50 6-10 10.0 0.017 17.6 1.39 6-11 11.0 0.101 19.9 1.61 6-12 12.0 0.153 10.7 1.50 6-13 13.5 0.107 18.8 2.86 6-14 15.5 0.171 8.2 0.74 6-15 18.0 0.293 15.1 1.57 6-16 21.0 0.229 9.6 1.44 6-17 24.0 0.314 18.3 1.44 6-18 27.0 0.526 12.3 N.M. 6-19 30.5 0.582 14.3 2.50 (1) B.W.=bottom water. (2) N.M.=not measured. 73 APPENDIX E. STEADY STATE EXPERIMENTS. To determine if the selective chemical attack scheme of Gephart (1982) was suitable for the sediments examined in this study, "steady state experiments" were performed. The purpose of these experiments was not to make a detailed investigation of selective chemical extractions, but only to confirm that the experimental conditions were optimal. This was done by subjecting 1 gram samples of sediments from Lake Superior to the extractions. The sediments used in the experiments were from core 2, taken in 1986 on the same cruise as site 1388 and treated in exactly the same manner. The sediments were all similar to those found at site 1388 in being largely clay. The concentration of selected elements was monitored as a function of time during the course of the leaching experiments. Small aliquots of leachate were removed during the experimental run and analyzed for selected metals by A.A.S. The leaching runs were done sequentially in the same order and under the same conditions as given in Table 2. Only a limited number of analyses were possible on each aliquot because of the limited volume available. Very small 74 Table H. Steady state experiments. EXCHANGABLE FRACTION TIME Fe BBL3 Mn 2-5 hr. #9/9 #9/9 0 0 0 0.6 22 21 1.1 12 20 1.5 15 25 2.0 29 17 3.1 - 27 4.1 31 32 WEAKLY ACID SOLUBLE FRACTION TIME Mn BBL3 hr. #9/9 0 0 0.6 88 1.6 85 4.1 99 6.6 102 7.9 108 9.2 128 EASILY REDUCIBLE FRACTION TIME Fe BBL3 Fe 2-5 Mn 2-5 hr. #9/9 #9/9 #9/9 0 0 0 0 0.27 42 95 212 0.55 . 94 105 303 0.85 146 104 329 1.17 - 187 395 1.43 217 176 407 1.87 304 267 490 75 TABLE H. (cont.) MODERATELY REDUCIBLE FRACTION TIME Fe BBL3 Fe 2-15 Fe 2-5 Mn 2-5 Zn 2-5 hr. #9/9 #9/9 #9/9 #9/9 #9/9 0 0 0 0 0 O 0.5 4990 3140 2530 176 6.3 1.6 9330 5380 5830 272 17.4 4.1 14740 7340 7710 298 21.4 6.1 14710 8340 9090 330 32.7 7.4 15970 8370 9000 323 - 8.4 - - - 376 30.0 76 aliquots,(0.1 ml.), were removed, so as not to change the solid to solution ratio significantly. The first four extractions only were monitored. The results are summarized in Table H. The criterion used to judge how well the extractions were tuned was to look for a distinct plateau or leveling off in metal released during the progress of the extraction. If one was found that corresponded to one found by Gephart (1982) the conditions and the reaction time was judged to be suitable. The steady state experiments indicated that the times recommended by Gephart (1982) were also suitable for sediments from Lake Superior. The exchangeable (EX) fraction was monitored for Fe in one sample and for Mn in another. Metal release for the two samples was mostly complete after 1 hour, the recommended time, Figure A. The results were rather ’noisy'from these runs, limited readsorption or reequilibration of the sediments with the near neutral pH solution is possibly the cause of this scatter. In the weakly acid soluble fraction, (WAS), Figure B., such low metal concentrations were released that only Mn in one sample could be monitored. Most of the manganese was released within one hour of leaching. The recommended time was 5 hours for this extraction. Although the recommended time was longer than necessary, no reason to alter the sequence was seen. The concentration of released manganese remained stable after the first hour. 77 EXCHANCEABLE FRACTION ;:;;e SPL. BBL3 n SPL. 2-5 40 F} T T T T T F— §3D r— // ._ C3 / / LJJ / \ / m 20 r— ~+ \ / - 3; + D ‘L l U 1 1 1 1 1" O 1 2 3 4 5 TIME (hours) Figure A. Steady state experiment on exchangeable (EX) fraction. Arrow is time recommended by Gephart (1982). WEAKLY ACID SOLUBLE FRACTION Mn SPL. BBL3 60 D 1 l 1 1" O 2 4 6 8 IO TIME (hours) Figure B. Steady state experiment on weakly acid soluble (WAS) fraction. Arrow is time recommended by Gephart (1982). PPM DRY SEDIMENT 0.) CD p— 78 Figure C. shows the results for the easily reducible (ER) fraction for' Mn in one sample and for Fe in two samples. There is much more rapid release of Mn in the first half hour than afterward but a satisfactory steady state is not reached. After the initial burst of Mn and Fe release Fe and. Mn are evolved at a nearly constant rate for the duration, (1.87 hrs.), of the experiment. It appears that more than one mineral is attacked by this reagent. This accords well with Gephart (1982) findings. The reaction time recommended by Gephart (1982) was retained in this study. It would seem to be the best possible compromise between leaching out the interpreted first, most easily reducible, mineral or group of minerals completely and minimally affecting the more slowly reduced mineral(s). It is interpreted that the most easily reduced mineral(s) are the. hydrous IMn-oxides this extraction is targeted for, and that the more slowly reduced mineral(s) are Fe-oxides. Figure D. shows the results for the moderately reducible fraction of sediment (MR). The recommended reaction time for this extraction ‘was 6 hours. A good ’plateau’ in leachate concentrations was found at this time for Fe, Mn, and Zn in selected samples. No reason to alter the recommended reaction time was inferred from these results. PPM DRY SEDIMENT 79 EASILY REDUCIBLE FRACTION «4 Fe SPL. BBL3 Fe SPL. 2-5 XMn SPL. 2—5 SDD —' ‘ ' ‘ ' ,X. 400-— XX _ BDD~ XX ._ 2002 X / ._ 1DD — /+..—- ——)e _ O 0.4 0.8 1.2 1.6 2 TIME (hours) Figure C. Steady state experiment on easily reducible (ER) fraction. Arrow is time recommended by Gephart (1982). 8O MODERATELY REDUCIBLE €~§e ggt. e . 15 J ._ TE 12r~ —. 2: C3 LLJ U) 8 H _ >_ 92 g: 4 - ._ El 4 f D 1 1 1" O 4 6 IO TIME (hours) MODERATELY REDUCIBLE B*Mn SPL. FRACTION + Zn SPL- 400 ~’ ‘ ‘ L.40 p.— 2: L'J . E. 300 __ -3D :3 u: U) 92 z: E 100 I— —10 c 23 / O I J l {—4 D O 4 6 10 TIME (hours) Figure D. BBL3 2-5 2-15 2 2- l U'IUT Zn PPM DRY SEDIMENT Steady state experiments on moderately reducible (MR) fraction. Arrow is time recommended by Gephart (1982). BIBLIOGRAPHY 81 BIBLIOGRAPHY Allen, R.S. 1986. The Role of Particulate Matter in the Fate of Contaminants in Aquatic Ecosystems. Scientific Series #142, Environment Canada, Inland Waters Directorate National Water Research Institute Canada Centre for Inland Waters, Burlington, Ontario. Balistrieri, L.S. and Murray, J.W. 1984. Marine scavenging: trace metal adsorption by interfacial sediment from MANOP Site H. Geochimica et Cosmochimica Acta. 48:921- 929. Balistrieri, L.S. and Murray, J.W. 1986. The surface chemistry of sediments from the Panama Basin: The influence of Mn oxides on metal adsorption. Geochimica et Cosmochimica Acta. 50:2235-2243. Berner, R.A. 1980. Early Diagenesis. Princeton University Press. Burdidge, DJ. and Gieskes, J .M. 1983. A pore water/solid phase diagenetic model for manganese in marine sediments. American Journal of Science. 283:29-47. Callender, E., and Bowser, C.J. 1980. Manganese and copper geochemistry of interstitial fluids from manganese nodule-rich pelagic sediments of the northeastern equatorial Pacific Ocean. American Journal of Science. 280:1063-1096. Carignan, R. and Nriagu, J.O. 1985. Trace metal deposition and mobility in the sediments of two lakes near Sudbury, Ontario. Geochimica et Cosmochimica Acta. 49:1753-1764. Cornwell, J.C. 1986. Diagenetic trace-metal profiles in Art ic lake sediments . Environmental Science and Technology. 20: 290-302. Davison, W. 1982. Transport of iron and manganese in relation to the shape of their concentration-depth profiles. Hydrobiologica. 92: 463-471. Drever, J.I. 1988. The Geochemistry of Natural Waters. New Jersey: Prentice-Hall. Edgington, D.N. and Robbins, J.A. 1976. Records of lead deposition in Lake Michigan sediments since 1800. Environmental Science and Technology. 10:266-274. 82 Filipek, L.H. and Owen, R.M. 1979. Geochemical associations and grain size partitioning of heavy metals in lacustrine sediments. Chemical Geology. 26:105-117. Fischer, K., Dymond, J., and Lyle, M. and Soutar, A., and Rau, S. 1986 The benthic cycle of copper: Evidence from sediment trap experiments in the eastern tropical North Pacific Ocean. Geochimica et Cosmochimica Acta. 50: 1535-1543. ‘ Forstner, U. 1982. Accumulative phases for heavy metals in limnic sediments. in Developments in Hydrobiology 9: SedimentlFreshwgtgr Ipperaction, ed. P.G. Sly, pp. 269- 284. The Hague: Junk Publishers. Froelich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke, N.A., Heath, G.R., Cullen, D., Dauphin, P., Hammond, D., Hartman, B., and Maynard, V. 1979. Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Geochimica et Coshmochimica Acta. 43:1075-1090. Gaudette, H.E., Flight, W.R., Toner, W.R., and Folger, D.W. 1974. An inexpensive ‘titration. ‘method. for the determination of organic carbon in recent sediments. Journal of Sedimentary Petrology. 44:249-253. Gephart, C.J. 1982. Belgtiye Importance of Iron-Oxide, Manganese-Oxide and Organic Material on the Adsorption of Chromium in Natural Water Sediment Systems. M.S. Thesis, Michigan State University. Gobeil, C., Silverberg, N., Sundby, B., and Cossa, D. 1987. Cadmium diagenesis in Laurentian Trough sediments. Geochimica et Cosmochimica Acta. 51: 589-596. Graybeal, A.L., and. Heath, G.R. 1984. Remobilization of transition metals in surficial pelagic sediments from the eastern Pacific. Geochimica et Cosmochimica Acta. 48: 965-975. Gupta, S.K. and. Chen, K.Y. 1975. Partitioning' of trace metals in selective chemical fractions on near shore sediments. Environmental Letters. 10:129-158. Heggie, D., Kahn, D., and Fischer, K. 1986. Trace metals in metalliferous sediments, MANOP Site M: interfacial pore water profiles. Earth and Planetary Science Letters. 80:106-116. 83 Jenne, E.A. 1968. Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils and water: the significant roles of hydrous Mn and Fe oxides. American Chemical Society Advances in chemistry Series. 73:337-383. Johnson, T.C., and Eisenreich, S.J. 1979. Silica in Lake Superior: mass balance considerations and a model for dynamic response to eutrophication. Geochimica et Cosmochimica Acta. 43: 77-91. Johnson, T.C., Evans, J.E., and Eisenreich, S.J. 1982. Total organic carbon in Lake Superior sediments: Comparisons with hemipelagic and pelagic marine environments. Limnology and Oceanography. 27:481-491. Kemp, A.I.W., Williams, J.D.H., Thomas, R.I., and Gregory, M.I. 1978a. Impact of man’s activities on the chemical composition of the sediments of Lakes Superior and Huron. Water, Air, and Soil Pollution. 10: 381-401. Kemp, A.L.W., Dell, C.I., and Harper, N.S. 1978b. Sedimentation rates and a sediment budget for Lake Superior. Journal of Great Lakes Research. 4:276-287. Klinkhammer, G.P 1980. Early diagenesis in sediment from the eastern equatorial Pacific, II. pore water metal results. Earth and Planetary Science Letters. 49:81- 101. Klinkhammer, G., Heggie, D.T., and Graham, D.W. 1982. Metal diagenesis in oxic marine sediments. Earth and Planetary Science Letters. 61:211-219. Leckie, J.O., Benjamin, M.M., Hayes, K., Kaufman, G., and Altman S. 1980. Adsorption/coprecipitation of trace elements from water with iron oxyhydroxide. Electric Power Research Institute Research Project 910-1. LeLand, H.V.S., Shukla, S., and Shimp, N.L. 1973. Factors affecting distribution of lead and other trace elements in sediments of southern Lake Michigan. in Trace Metals and Metal-Organic Interactions in Natural Waters. ed. P. Singer, pp 89-138. Ann Arbor Science, Ann Arbor, MI. Lerman, A. 1975. Migrational processes and chemical reactions in interstitial waters. in The Sea, vol. 6, pp. 695-738. E.D. Goldberg, I.N McCave, J.J O'Brien, and J.H. Steele eds. New York: Wiley. Li, Y.H. and Gregory, S. 1974. Diffusion of ions in seawater and in deep-sea sediments. Geochimica et Cosmochimica Acta. 38:703-714. 84 Long, D.T. and Gephart, C.A. 1982. Competition analysis of Cr adsorption on Fe-oxides, Mn-oxides and organic matter in natural systems. Fall annual meeting, Geological Society of America Abst. with Programs. 14:549. Lyons, W.B., Hines, M.E., Gaudette, H.E. 1984. Major and minor element pore-water geochemistry of modern marine sabkhas. in Microbial Mats; Sprppgpalipes ed. P. Liss, pp. 411-423. New York. Malo, B. 1977. Partial extraction of metals from aquatic sediments. Environmental Science and Technology. 11:227-282. McKee, J.D., Wilson, T.P., Long, D.T., and Owen, R.M. 1986. Geochemical processes at the sediment/water interface in profundal lakes: experiments with a manned submersible. Fall annual meeting, Geological Society of America, Abst. with programs. 18:469. Murray; J.W., Balistrieri, L.S., and. Paul, B. 1984. The oxidation state of manganese in Marine sediments and ferromanganese nodules. Geochimica et Cosmochimica Acta. 48:1237-1247. Pedersen, T.F., Vogel, J.S., and Southon, J.R. 1986. Copper and manganese in hemipelagic sediments at 21°N, East Pacific Rise: diagenetic contrasts. Geochimica et Cosmochimica Acta. 50:2019-2031. Olivarez, A.M., Owen, R.M., and Long, D.T. 1988. Geochemistry of rare earth elements in benthic layer particulates and sediments of Lake Superior. Journal of Great Lakes Research (in press). Rea, D.K., Owen, R.M., and Meyer, P.A. 1981. Sedimentary processes in the Great Lakes. Reviews of Geophysics and Space Physics. 19: 635-648 85 Ridgeway, I.M., and Price, N.B. 1987. Geochemical associations and post-depositional mobility of heavy metals in coastal sediments: Loch Etive, Scotland. Marine Chemistry. 21: 229-248 Robbins, J.A. and Callender, E. 1975. Diagenesis of manganese in Lake Michigan sediments. American Journal of Science. 275:512-533. Robbins, J.A. and Gusting, J. 1976. A squeezer for efficient extraction of pore water from small volumes of anoxic sediment. Limnology and Oceanography. 21:905-909. Sawlan, J.J. and Murray, J.W. 1983. Trace metal remobilization in the interstitial waters of red clay and hemipelagic marine sediments. Earth and Planetary Science Letters. 64:213-230. Sigg, L. 1985. Metal transfer mechanisms in lakes:a the role of settling particles. in Chepical Processes in Lakes ed. W. Stumm, pp. 282-310. John Wiley, New York. Soutar, A. and Johnson, 3. 1981. Sampling the sediment-water interface--evidence for an organic-rich surface layer (Abst.). EOS. 62:905. Stumm, W. and Baccini, P. 1978. Man-made chemical perturbations of lakes. in kes' e ist eolo Phys; cs . ed . A. Lerman , pp . 91-126 . Springer-Verlag , New York. Tessier, A., Campbell, P.G.C., and Bisson, M. 1979. Sequential extraction procedune for the speciation of particulate trace metals. Analytical Chemistry. 51:844- 851. Thomas, R.L., and Dell, C.I. 1978. Sediments of Lake Superior. Journal of Great Lakes Research. 4:264-275. Westerlund, S.F.G., Anderson, L.G., Hall, P.O.J., Iverfeldt, A., van der Loeff, M.M.R., and Sundby, B. 1986. Benthic fluxes of cadmium, copper, nickel, zinc and lead in the coastal environment. Geochimica et Cosmochimica Acta. 50: 1289-1296. Wilson, T.F., Long, D.T., and Owen, R.M. 1986. The water/sediment interface via manned submersible. (Abst.). International Association of Great Lakes Research. 29th Conference: 52. a q'} "IIIIIIIIIIITIIIT