' \.. .F..\.. . V» I‘.\ fit I Ik I I._tI.. .: .2 . I . I. It a?! is a 1. - I i .T .. I #rr},«a1.o\_ Jns_.MHccu uu.‘ 'I .4! II 0 “at. a! on c 903%” QoOquaN‘ov 0:... _ I...v|o.QoV¢Uot ”III... I... O .I. ‘Qr. (II- _ 7’13... «a‘OaI . . .....Iu I4 09' \ Q 01-11.! a ‘0 \C‘Ao...-.c:\ Q..\.¢.OIQ.OII .. 0.3. -.w.n. o. .¢IO qI .l- ‘10. I“ Init III-1. t LIP. \ .Od.“o"q to $31 5“; h‘l...‘ h‘.’0ill10l.'lfCo>Fe>Cr; (4) metal concentra- tions in the shelf sediments are dependent on grain-type; (5) Amazon Shelf Sediments remain oxic to a depth of 450 cm; (6) little reparti- tioning occurs with sediment depth. Dedicated to my parents for all their loving support and encouragement. ACKNOWLEDGEMENTS I would like to express my sincere thanks to the members of my committee Dr. David T. Long, Chairman, Dr. Grahame Larson, and Dr. Duncan Sibley for their criticism, support and encouragement in completing this project. I would like to thank North Carolina State University, particularly Dr. David DeMaster for supplying the cores for analysis. I would also like to thank my friends and colleagues with whose help and support have made this project endurable. This thesis was partially funded by a grant from Sigma Xi, the Research Society and a Research Scholarship from Marathon Oil Corporation. iv TABLE OF CONTENTS List of Tables . . . . . . . . . . . . . . . . . . . . List of Figures. . . . . . . . . . . . . . . . . . . IntrOdUCtion O O O O O O O O O O O O 0 I O O O O O 0 Significance and Research Goals. . . . . . . . . . . Nature of Metals in Continental Shelf Environments Sources of Metals to Shelf . . . . . . . . . Pathways and Sinks in Marine Systems . . . . Diagenesis of Metals in Marine Sediments . . St Udy Area 0 I O O O O O O O O O O O O O O O O O O O Meth0d0109y. O O O O O O O O O O O O O O O O O O O 0 Results and Discussion . . . . . . . . . . . . . . . . . . . . . Chemical Partitioning of Metals in Amazon Shelf Sediments. Distribution of Metals on the Shelf. . . . . . Diagenesis of Metals in Amazon Shelf Sediments source 0 O O O O O O I O O O O O O O O O O O 0 Cone] “51.0" 0 O O O O O O O O O O O O O O O O 0 O O 0 References 0 O O O O O O O O O O O O 0 O O O O O O O Appendices Appendix I Chemical Partitioning Data. . . . . Appendix II Graphite Control Settings For Perkin Elmer 560 Atomic Adsorption Spectrophotometer with HGA:2200 Graphite Furnace . . . . . . . . . . . . . . Appendix III B-Matrix of Heights from Q-mode and R-mode Factor Analysis. . . . . . . . . . . . . . . . . . . . Page 105 126 .130 LIST OF TABLES Table Page 1. Trace metal partitioning of Cr, Mn, Co, Cu, Fe, Ni, Zn in Amazon River suspended load and surficial shelf sediment as a function of sediment type. . . . . . . . . . . . . . . . . 41 2. Trace metal partitioning of Cr, Mn, Co, Cu, Fe, Ni, Zn in selected soils of the Amazon River Drainage Basin. . . . . . 43 vi Figure 1. 2. 12. 13. LIST OF FIGURES Study area and location of sampling stations . . . . . . . . Morphostructural map of Amazon Drainage Basin (Stallard, 1980). O O O O O O O O O O O O O O O O O O O O O O O O O O 0 Typical surface-water salinities ( 0/00) during wet season conditions (April-May, 1968) and dry season conditions (November, 1967) of the surface waters off the Amazon. Profiles in the upper right hand corners show vertical salinity gradients in profiles taken directly seaward of the Amazon mouth. (Milliman ££.él-. 1975). . . . . . . . . . Map of Amazon River, showing locations on mainstream tributaries where sediment loads were measured in 1977 the StUdy by Meade e_t fl. , 1979. O O O O O O O O O O 0 Distribution of grain-types at 0-2 depth on the Amazon continenta] Shelf. O O O O O O O I O O O O O O O O O O Areal distribution of mean grain size on the Amazon Continental Shelf (Nittrouer et al., 1983) . . . . . . Distribution of sediment accunulatgpa rates on the Amazon Continental Shelf based on Pb geochronology (Kuehl a _a—]O ’ 1982) O O O O O O I O O O O O I O O O 0 Clay mineral distribution along the Amazon Shelf from Nittrouer gt 31. (1983) and Gibbs (1977) . . . . . . . for Flow chart of selective chemical attacks . . . . . . . . . Nunerical representation of sediment type from Folk (1974) . Contour map of zinc in the detrital fraction of the surface (9-5 cm) sediment of the Amazon Continental Shelf. . . . . . Contour map of zinc in the hydromorphic fraction of the surface (0-5 cm) sediment of the Amazon Continental Shelf. . Contour map of total zinc in the surface (0-5 cm) sediment of the Amazon Continental Shelf. . . . . . . . . . . . . . . vii Page 15 17 19 21 24 26 29 Figure 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Plot of sediment type versus concentration of zinc total. Numbers refer to overlapped points . . . . . . . . . . . . . Plot of sediment type versus concentration of zinc in the detrital fraction. Numbers refer to overlapped points . . . Plot of sediment type versus concentration of zinc in the hydromorphic fraction. Numbers refer to overlapped points . Plot of sediment type versus concentration of zinc in the hydromorphic fraction plotted as ranked data . . . . . . . . Contour map of the ratio zinc in the detrital fraction divided by aluninun in the detrital fraction in the surface (0—5 cm) sediments of the Amazon Continental Shelf . . . . . Contour map of the ratio total zinc concentration divided by total aluminum concentration in the surface (0-5 cm) sediments of the Amazon Continental Shelf. . . . . . . . . . Contour map of the ratio zinc in the oxide fraction divided by iron and manganese in the oxide fraction in the surface (0-5 cm) sediments of the Amazon Continental Shelf . . . . . Contour map of the ratio barium in the oxide fraction divided by iron and manganese in the oxide fraction in the surface (9-5 cm) sediments of the Amazon Continental Shelf . . . . . Contour map of the ratio zinc in the hydromorphic fraction divided by zinc in the detrital fraction in the surface (0-5 cm) sediments of the Amazon Continental Shelf . . . . . Contour map of the ratio barium in the hydromorphic fraction divided by barium in the detrital fraction . . . . . . . . . Sediment type with depth for core 10 (Nittrouer gt al., 1983). O O O O O O O O O O O O O O O O O O O O O O O O 0 O 0 Concentration of chromium with depth in core 10. . . . . . . Concentration of iron with depth in core 10. . . . . . . . . Concentration of manganese with depth in core 10 . . . . . . Concentration of zinc with depth in core 10. . . . . . . . . Concentration of copper with depth in core 10. . . . . . . . Concentration of nickel with depth in core 10. . . . . . . . Concentration of bariun with depth in core 10. . . . . . . . viii Page 61 61 63 63 66 68 70 72 75 77 79 81 84 86 88 91 93 95 INTRODUCTION The geochemical cycle of any trace metal is defined in terms of sources, pathways, and sinks. In continental shelf environments conti- nental fresh water and marine waters mix. Upon mixing more than 95% of the terrigenous sediment contained in the freshwater settles out prior to 3 0/00 salinity (Milliman st 31., 1975), thereby making continental shelf sediments a possible major sink for trace metals in exogenic systems. Two major sources of trace metal influx into oceanic sediments exist; a river source and a hydrothermal source (Schutz and Turekian, 1965). Traditionally, continental runoff has been considered the major source of metals to seawater and marine sediments (Gibbs, 1965; Sholkovitz and Price, 1980). Recently, however, it has been suggested, and in part demonstrated, that hydrothermal activity associated with oceanic ridge systems could also be a significant source of metals to marine environments (Gordon and Corliss, 1979; Edmond 35 al., 1979; Hindom gt al., 1971; Brumsack, 1980; Dymond, 1981; Marchig gt al., 1982; Bischoff and Dickson, 1975). The significance of this source, though, has as yet not been assessed. The purpose of this research is to determine the controls on trace metal concentrations in continental shelf sediments as well as their source. The area for this study is the Amazon Continental Shelf which is a relatively contaminate free environment (Drever, 1982). The working hypothesis for this study is that patterns of trace metal chemical partitioning among various chemical phases in shelf sediments should reflect the controls on their chemical behavior and sources of metals to the sediment (Loring, 1976; Loring, 1982; Luoma and Bryan, 1981). SIGNIFICANCE AND RESEARCH GOALS The Amazon River Continental Shelf was selected as a study area because it is an unpolluted system whose river source drains through diverse environments. Hence, it offers an excellent Opportunity to study trace metal geochemical behaviors. It is also one of the major rivers of the world, supplying 18% of all river water to the ocean. Consequently, the study of the nature of trace metals in the associated shelf sediments should be particularly important in helping to understand the chemical processes occurring in shelf sediments. The goals of this research are to: (1) Determine the partitioning fate of selected trace metals upon interaction with shelf sediments, (2) Determine the fate of selected trace metals during shelf sediment diagenesis, (3) Define trace metal sources to shelf sediments. NATURE OF METALS IN MARINE AND CONTINENTAL SHELF ENVIRONMENTS This section discusses previous work on the geochemical cycle of elements in marine environments. The geochemical cycles are defined in terms of sources, pathways, and sinks of the elements. Discussion is focused on those aspects of the marine environment that can be related to the geochemical behavior of metals in continental shelf environ- ments. Trace metal diagenesis in marine sediments will also be discussed. Sources of Metals to Shelf: Sources of trace metals to marine and therefore continental shelf environments have been summarized by Schutz and Turekian (1965) as: (1) stream and river, (2) submarine volcanism, (3) submarine alteration or solution of non-volcanic material, (4) eolian, (5) glacial discharge, and (6) anthropogenic. Eolian and glacial discharge are considered minor sources of trace metal input into the continental shelf (Schultz and Turekian, 1965). Submarine alteration and submarine volcanism on the other hand are difficult to distinguish as unique sources to shelf sediments and are therefore classified together as a ridge source. In the case of the Amazon River system, the anthropogenic input is considered negligible (Drever, 1982). Ridge systems and river runoff can therefore be con- sidered the two major sources of trace metal input into the shelf sediments. Several marine environments have been studied (estuaries, deep ocean sediments, ridge systems, continental shelf and fjords) with regard to the delineation of separate trace metal sources. Factor analysis was used to define the relative importance of each source input (Imbrie and Van Andel, 1964). In this research two basic types of multivariant factor analysis were utilized, Q-mode and R-mode. Q-mode correlates two samples on the basis of the variables, whereas R-mode compares the relationships between the variables on the basis of the samples (Kerlinger, 1973; Nelson, 1981; Parks, 1970; Imbrie and Van Andel, 1964). Mathematically, Q and R-mode factor analysis treat each sample or variable as a vector and resolves it into a small number of component vectors (Imbrie and Van Andel, 1964). Frequently, the original or raw data matrix must be transformed, but this transformation should in no way affect the accuracy of the program because relative proportions of variables are preserved (Bopp, 1981; Loring, 1982). This transforma- tion of raw data allows every variable to have equal weight in the statistical program despite the ranges observed in variable's concentration. The results of Q-mode factor analysis are written in two matrixes, the FS-matrix and the B-matrix. The Fs-matrix is a table listing combinations of the variables that define the compositional end-members for each of the factors. The B-matrix is a matrix of the influence exerted by each end-member (i.e., matrix relates the intensity of correlation between the factors) on each sample (Bopp, 1981; Bopp and Briggs, 1981). In the scaled varimax factor listing (FS-matrix), if all variables were equally represented their varimax factor value would be 1.000. A strong positive value represents a positive correlation. A zero value represents no correlation and a strong negative value represents an inverse correlation with the matrix for that factor. A contour map of the matrix for each factor is generated and illustrates the distribution of factors in the study area (BOpp, 1981; Bopp and Biggs, 1981). Most marine environments that have been studied have shown several source inputs. For example, Bopp and Biggs (1981) have identified three sources of metals to Delaware Bay sediments (representative of (1) a terrigenous natural background level source; (2) an oceanic source; and (3) a river source). Holmes and Martin (1978) on the other hand have determined two sources of trace metal origin to the continen- tal shelf in the Gulf of Mexico (a natural background source and an anthropogenic source). In addition, Loring (1976, 1976b, 1979, 1982) has delineated two sources of trace metal input to the Saguenary Fjord; an industrial source and a natural source. Harding and Brown (1975) also investigated trace metal sources in the sediments of the Pamlico River Estuary, North Carolina. They found that the sources include a background (natural weathering) input source, and an industrial (anthropogenic) input source. Bender st 21., (1971) studied sediments in the East Pacific Rise and found a seawater source and a ridge volcanism source. Edmond gt al., (1982) dealt with sediments in the East Pacific Rise and concluded that the hydrothermal input in the oceans are substantial compared with fluvial transport. Heath and Dymond (1981) identified 5 sources of input to Nazca plate sediments (a biogenic, detrital, hydrothermal, authigenic and solution residue). The detrital, hydrothermal and biogenic sources are the three dominate sources of trace metal input to the Plate sediments. The detrital cone ponent is the major input source for trace metals to the Plate sedi- ments near the continental coast of South America, whereas the hydro- thermal component is the major source of input to ocean sediments near the East Pacific Rise. The hydrothermal source is undetectable at 2500 km east of the East Pacific Rise (Heath and Dymond, 1981; Dymond, 1981). The previous research cited above has shown that (1) when an anthropogenic source is present it is dominate; (2) when a study area is located near a ridge system, the hydrothermal source is dominate; (3) when a study area is located near a continental coast, the detrital input is dominate and (4) if 2500 km from a ridge source, ridge input is undetectable. Therefore, the concentrations and distributions of trace metals in a continental shelf environment should be able to be interpretated in terms of a single source. Pathways and Sinks in Marine Systems: The pathways of metals in exogenic systems can be defined in terms of five mechanisms of transport (Kharkar gt al., 1968). They are: (1) in solution (free and complexed). (2) in the lattice of minerals of the suspended load, (3) associated with oxides of the suspended load, (4) adsorbed on the minerals of the suspended load, and (5) associated with solid organic material. Trace metals can be removed from active transport in the marine environment by: (1) mechanical deposition of larger particles, coagulation of the colloidal fraction, humus colloids uptake plankton microorganism extraction, precipitation as oxide coatings, adsorption in cation exchange sites, and (in anaerobic conditions) forming insoluble sulfides (Krausk0pf, 1956; Morozov, 1979; Harding and Brown 1975). AAAAAA \IO\ 01 43- um vvvvvv These seven removal mechanisms would be the major sinks of trace metals in marine and therefore continental shelf systems. The chemical nature of the pathways and sinks of metals are very similar as shown above and reflect whether or not the metal is in an active state of transport. The chemical nature of a metal in exogenic systems (independent of its state of transport) is defined then by the chemical state it is in: dissolved (free and complexed), adsorbed on or in exchange position in clays, precipitated as hydroxides, adsorbed on or co-precipitated with Fe and Mn oxides, associated with organic matter, precipitated as sulfides, in lattice sites of minerals. AAAAAaA \I Ch 0'1 «9 m N H VVVVUVV States of 2-6 are the states of the metal associated with the sediment. States 2-5 are collectively called the hydromorphic fraction or phase and state 6 is called the detrital or residual fraction. Metals in these fractions are normally differentiated by a series of selective chemical attacks on the sediment as discussed in the methods section. After the sediment has been chemically characterized, the metals are said to be partitioned among the different fractions of the sediment. The following is a summary of work that has been done in defining these fractions for metals in marine sediments. Loring (1975, 1976, 1976b, 1979, 1982) studied trace metals in the sediments of the Saguenary Fjord. He found that the highest concentrations are found in mud (fine-grained) sediments and that the lowest concentrations occur in the sandy (coarse) sediments (i.e. Co, Ni, Cr, Zn, Cu, and Pb con- centrations increase with decreasing grain size). The detrital phase -of Co, Ni, Cr, Zn, Cu, and Pb concentrations account for 71-98% of the total elemental concentrations in Fjord sediments. Holmes and Martin (1978) trace Cr, Cu, Fe, Mn, Ni, Zn, Ba and Pb migration in a continen- tal shelf environment in the northwest Gulf of Mexico. They concluded that the most important factor affecting physical and chemical process within the sedimentary environment are variation in sediment accumula- tion rates. Brannon st 31., (1979) researched Cu, Fe, Mn, and Zn in sediments from Mobil Bay, Alabama. They defined selective extraction phases investigated as adsorbed (ion exchangeable) on sediment materi- al, reducible (solubility and migration controlled by oxidation-reduc- tion reactions), bound in organic matter and residual. Iron in the reducible phase has a value of 68-79%. 0f secondary importance to Fe is the organic phase (12-19%). Combined, the reducible and organic phases represent 90% of the total Fe concentration in the sediments. Manganese has 15-35% of total metal concentration in the reducible phase, 35-44% in the organically bound phase and 8-12% in the residual phase. Together, the organically bound, residual and reducible phases account for 78-80% of total Mn concentrations in the sediments. Copper has 51-61% of total metal retained in the residual phase. The organi- cally bound phase is of secondary importance with 32-41% of total metal concentration. Combining the residual and organic phases, they repre- sent 92% of total Cu in the sediments. Zinc has 52-59% of total metal concentration held in the organic phase and 38-44% in the residual phase. United, they equal 95% of total Zn concentration to the Bay sediments (Brannon st 31., 1974). Harding and Brown (1975) studied Co, 10 Cr, Cu, Ni, Pb and Zn in the Pamlico River Estuary. They found that factors affecting trace metal uptake in sediments included an enrich- ment of trace metal concentrations in clays and organic matter. They stated that the surficial distribution of fine sediments appear to be due to the patterns of water circulation. Trace metal incorporation with clay sediments and organic matter in the estuary were dependent on temperature, pH, Eh and clay mineral type (the relative importance of each was undefined). Harding and Brown (1975) explain trace metal dis- persal in the estuary based on the relative mobility of the elements. Cr is the most immobile element studied. It quickly drops out of the water column upon entering the estuary and is not dispersed throughout the Bay sediments. The moderately immobile trace elements, Co, Cu, Ni and Pb (relatively compared to the other elements studied) remain in solution, are circulated from their industrial source and are dispersed in clays and organic matter throughout the estuary sediments. Cosma st 31, (1979) studied trace metals (Cr, Cu, Ni, Mn) in surface Sediments in the continental shelf area between Arenzano and Capo Noli in the Liqurian Sea. Cr was controlled by pollution inputs. Nickel concen- trations were accounted for by stream input of eroded basic rock outcrops. Capper concentrations are associated with the organic frac- tion. Manganese concentrations are related to sediment textures. Turekian and Imbrie (1966) researched deep sea sediments from the Atlantic Ocean. They found that the trace elements Ba, Co, Cu, Ni, Pb, Cr and Mn showed no correlation with clay mineralogy or depth of water. There was a strong correlation of metals Mn, Ni, Co with areas of low clay accumulation rates possibly due to a fine grained pelagic compo- nent. Morozov (1979) studied migration of Fe, Mn, Zn, Cu, Ni, Co, Cr, 11 and Pb around the mouths of several rivers. The general conclusions of this investigations were: Trace metal concentration in the suspended and dissolved forms decreased in the sequence river-estuary-sea-ocean; The portion of suspended Fe and Mn forms decrease from the river (97%) to ocean (7%); the mechanical deposition of the larger particles occurs near the mouth of a river; coagulation of the colloidal fraction and the humus colloids take up a large fraction of trace metals at the mouth of a river; and sea/fresh water zones show extensive plankton microorganism development which extract metals from the water column. Summarizing past research these conclusions can be drawn: (1) trace metal concentrations increase with decreasing grain size; (2) there is no correlation with clay mineralogy or depth of water from the Atlantic deep-sea sediment core samples; (3) an important factor affecting processes in a basin is the variation in sediment accumula- tion rates; and (4) the relatively immobile elements will drop out of the water column upon entering an estuary from a river source. Diagenesis of Metals in Marine Sediments: Chemical conditions of sediments can change with depth as a result of compaction and age (Bonatti gt al., 1971). Factors affecting the chemical conditions are: changes in Eh, sediment compositon, tempera- ture, bioturbation, and pH (Yen and Tang, 1977). Eh values normally decrease with depth in sediments (Addy st 31., 1976). The relatively mobile elements should be enriched in the oxidized sediment zone (Ni, Fe, Mn, Co) (Addy gt al., 1976; Holmes and Martin, 1978; Heath and Dymond, 1981); while less mobile elements like zinc and c0pper show little migration in the sediment column (Addy 33 al., 1976). Brooks st 323’ (1968), Bonatti et al., (1971) concluded that Fe and Cu are not 12 mobile elements and show no overall trend to increase with sediment depth, and Fe seems to show inconsistent behavior during diagenesis. Ba and Pb are also relatively immobile elements and according to Holmes and Martin (1978) show no overall trend with depth. Brannon ft 31., (1979) stated that no patterns of migration for the elements Fe, Mn, Cu and Zn could be conclusively drawn. Generally, if an environment is reducing, the elements with higher redox potentials should go into solution and if an environment is oxidizing they should precipitate. In this way the higher redox potential elements migrate up the sedi- ment column, creating an enriched metal concentration on the top centimeters of the ocean floor (Addy 23.21“: 1976). In summary, these investigations show that: (1) Ni, Co, Mn should be enriched in the upper centimeters of the sediment column; (2) Ba, Pb, Zn, Cu should not be enriched nor depleted in the upper centimeters of the sediment column; and (3) Conflicting Fe behavior exists. This summary shows that much research still needs to be done. No research involving source delineation in unpolluted continental shelf sediments has been reported. In this investigation, the chemical par- titioning data on the Amazon Continental Shelf coupled with factor analysis can aid in the resolution of trace metal sources to shelf sediments. Chemical partitioning data with depth in Amazon Shelf sedi- ments will help to eliminate conflicting diagenetic trace metal behav- ioral observations. Exploring trace metal behaviors encountered when fresh and oceanic waters merge will aid in the investigation of the controls or aid in determining the controls of geochemical cycling of trace metals. STUDY AREA The study area is located on the Continental Shelf at the mouth of the Amazon River (Figure 1) and extends from a latitude of 52°N to 46°E and a longitude of 2°S to 6°N. The Amazon River Basin lies entirely in the central equatorial area, extending from 5°N to 20°S longitude and from 50°H to 77°w latitude (Gibbs, 1967; Stallard, 1980). The Amazon River drains a diverse climatic area ranging from the artic Andean Mountains to the tropical rain forests. Temperatures vary from less than 15°C in the Andeas to 28°C in the tr0pical forests (Gibbs, 1967). The Amazon River Drainage Basin is the largest in the world, draining an area 6.3 x 106 km2 (Keller, 1962). The river drains diverse geological environments (Figure 2) comprising active orogenic zones, epiorogenic uplift zones, stable cratonic zones and active sedi- mentary basin zones (Gibbs, 1967; Stallard, 1980). The Amazon River has an average discharge of 1.7 x 105 m3/sec, supplying 18% of all water from rivers into the ocean (0ltman, 1968; Drever, 1982). Sedi- ment influx ranges from 8-9 x 108 tonnes/year; influx was measured at 0bidos a site marking 80% river flowage (Figure 4) (Holeman, 1968). The dissolved load is 2.9 x 108 tonnes/year (Meade gt al., 1979). The clay content at the mouth of the river consists of 33% illite, 31% kaolinite, 27% montmorillite, and 2% chlorite (Milliman et al., 1975). Precipitation in the Amazon River Basin ranges from less than 2000 mm/year to over 3500 mm/year (Stallard, 1980; Hoffman, 1968). The 13 14 Figure 1. Study area and location of sampling stations. Figure 2. “SEE ELSE fllflflllflfifliflfl El 16 Morphostructural Map of Amazon Drainage Basin (Stallard, 1980). Amazon Trough Subandean Trough Shields Western Cordillera (Cordillera Occidental) Intercordilleran Zone (includes the altiplano regions) Eastern Cordillera (Cordillera Oriental) Subandean Uplifts \ION U'I-wao—b o o o o o o 0 Symbol key: CONSTRUCTIVE STRUCTURAL RELIEF ELEMENTS Trend of folded young mountain ranges of the Andean system. Crest of horst mountains and monoclimes. Fault flexure. Lithological. Step, escarpment. DESTRUCTIVE NON-STRUCTURAL RELIEF ELEMENTS Old erosion. Surfaces of Mesozoic Tertiary age. Occurrence of inselbergs. Crest of residual relief. ACCUMULATIVE RELIEF ELEMENTS Quaternary fluvial. Alluvial or eolian deposits (in the Andes including glacio-fluvial and volcanic tectonic mudflow deposits). Occurrence of Late-Tertiary and Quaternary lacustrine or marine deposits. Pleistocene loess with ash-admixtures. Sand dunes. Pleistocene glacial deposits. Occurrence of volcanoes. Salt flat. Lake. Figure 3. 18 Typical Surface-water salinities (O/OO) during wet season conditions (April-May, 1968) and dry season conditions (November, 1967) of the surface waters off the Amazon. Profiles in the upper right hand corners show vertical salinity gradients in profiles taken directly seaward of the Amazon mouth. (Milliman gt_al., 1975). 20 Figure 4. Map of Amazon river, showing locations on mainstream tributaries where sediment loads were measured in 1977 for the study by Meade gt_al., 1979. 21 .i . , ' .7 ‘c ~~ ‘-‘-.‘i-‘ ' -"'- I" i '7 2"!» ' .. thwartdiw. - ' _, ;‘~ ’ . _ . . k ‘ ,\.; “one...“ .QJ H g... ..,—; "We... ‘ 1 i ‘. . Pacific Ocean i _ ’41.‘ j 830 Paulo “‘ dc Olivenci Sumo Antb‘nio da 19d 9 u ) ‘ ‘1' z 43a <9 + Itapet'la I, Mampum 4 0 $6,. 1 W Rio New 9 Manaus+ ~C.-- ”‘z- Rio Madeira mm + is)“. + 22 river has a low buffering capacity and pH values range from 6.5-7.5 throughout the year (Schmidt, 1972). The waters of the Amazon also appear thoroughly mixed vertically (Gibbs, 1967). The Amazon River flows onto the shelf and into the Guiana Current. The Guiana Current is an extension of the South Equatorial Current which flows in a northwesterly direction along the east coast of South America (Ryther gt al., 1967; Eisma st 31., 1971). At the mouth of the river, 95% of the terrigenous sediment from surface waters settle out prior to 3 0/00 salinity (Milliman gt al., 1975). Figure 3 shows sur- face water salinities for two seasons--the dny season, November, and the wet season, April-May. The width of the Shelf ranges from 100 to 300 km. The bathymetric chart (Figure 1) of the Continental Shelf reveals a relatively flat inner shelf to the 40 meter isobath. There is an abrupt change at this point and a steepening in slope appears until the 60 meter isobath. The change of slape represents the foreset of a subaqueous prograding delta (Kuehl gt al., 1982; Nittrouer 33 al., 1983). Surface grain size type distribution are shown in Figure 5. The upper two centimeters of shelf sediments are composed of (1) a mud (silt and clay) zone occupy- ing the inner shelf seaward and northwestward of the Amazon River mouth and (2) a sand zone dominating the rest of the outer shelf region. Three detrital sedimentary deposits are observed from the surface sedi- ment on the Amazon Continental Shelf (1) an outer shelf sand deposit; (2) an inner shelf mud deposit; and (3) mud interbedded with sand deposit (K“9h].EE.El°’ 1982; Nittrouer gt al., 1983). The areal dis- tribution of sediment grain size is in Figure 6. Outer shelf sands (median grain size 1.5-3.0 0) show moderate to good sorting. Inner Figure 5. 23 Distribution of grain types at 0-2 cm depth on the Amazon Continental Shelf (Nittrouer gt al., 1983). 24 Partially lndurated Substrate 111‘ v 0 2° 5° 'OOK'“ :5.- ....._ :’ |_§'-:i--.;..-ff T ISODIUH m RIGID?! -'nTI nxx r1 . 0° 00‘“ Ch.n0.I Mangd Island Figure 6. 25 Areal distribution of mean grain size on the Amazon Continental Shelf (Nittrouer gt al., 1983). 26 <4 sand .0 \ ‘ ’ 8 Si" 0 2 Maroio'lslond ‘2' e :f'Ofifi-‘v 27 shelf muds (median grain size 6-9 0) are poorly sorted (Nittrouer gt ‘31., 1983). Several models have been proposed to explain the sedimentary structures and grain size distribution on the Amazon Continental Shelf (Milliman et al., 1975; Gibbs, 1976; Nittrouer gt al., 1983). Milliman 'gtnal., (1975) hypothesizes that no modern muds are being deposited on the Amazon Shelf, except on a small nearshore zonal area. However, Milliman 3: al., (1975) states that two distinct periods of sedimenta- tion have occurred in Quaternary time. One occurring at high sea level and the other at low sea level (60-80 meters below present sea level). During high sea level periods, inner shelf sediment accumulation occurs only in a narrow nearshore belt of mud. The majority of muds, however, are deposited at low sea levels. In direct opposition to Milliman gt al,, (1975) theory are Pb-210 data on shelf sediments. The Pb-210 dates indicate that modern mud accumulation is presently occurring (Figure 7) (Kuehl 33 al., 1982; Gibbs, 1976). A second theory to explain the physical processes on the Amazon Continental Shelf was pr0posed by Gibbs (1976). Gibbs (1976) states that the Amazon River waters move offshore as a plume and is carried northwestward. Upon entry into the shelf, river particles settle out and are transported landward by bottom currents. A landward fining of particle size should result. Recent data, however, reveal an absence of reverse grading, contradicting the sedimentation model of Gibbs (1976) (Figure 8) (Nittrouer gt al., 1983; DeMaster, personal communi- cation 1982). A recent model proposed by Nittrouer st 21. (1983) suggests that modern sediment accumulation occurs as a result of a subaqueous delta 28 Figure 7. Distribution of sediment acguaulation rates on the Amazon Continental Shelf based on Pb geochronology. (Kuehl £131.,1982). > 2 cm/yr < 2 cm/yr I 90 I. ll I‘ll I ‘l’lll I I!!! If I ll ‘5 < 0.1 cnvyr 7 {Hi-I d NHH 1 i 1. ‘ 1 1 ’uu ‘ 1 l l mom» 60 20 0 m MOIOIS ISO”. I '0‘ Amazon RIVQI ." '. can. 30 .Axmmav manpw new Amwmfiv .Hm HM gmsogupwz Eon; mpmzm :o~cE< as» mcopm coppznwgumwv ngmcws AmPu .m 953... 31 .33: so); cog< Ea... 3:230 8w 8m 8.. Sn o8 P b r B 2.5....0829: " .2 03.5.39. " v. 2.5 u _ “ems-£U u U j :3: 336.-- 33: ..o .o 83.2.2 95. “ “ ‘- LIIIIO // you O fo— T8 ram wound 32 prograding over relict basal sands. Nittrouer 33 al., (1983) hypothe- sizes that a turbulent jet emanates from the river mouth and flows across the inner shelf. Their hypothesis is consistant with sediment accumulation rates based on Pb-210 data (Figure 7), sedimentary deposit data (Figure 5) and sediment grain distribution data (Figure 5). At low sediment accumulation rates, the effects of bioturbation on the disruption of sedimentary structures increases (i.e., original stratification is preserved in areas of highest sediment accumulation rates with minimal bioturbation). A strong correlation therefore exists between accumulation rate and sedimentary structure (Kuehl gt Eflr’ 1982). The model of Nittrouer st 21., (1983) is the most consis- tant with present data and is therefore the accepted theory for use in this research. METHODOLOGY A research cruise to the Amazon Continental Shelf took place in October 1979. Shelf sediments (Figure 1) were sampled using an N.E.L.- Reineck type box corer (20 cm x 30 cm cross-sectional area), in order to maintain core stratigraphic integrity. These cores and soil samples from Belem (Figure 1) were immediately subsampled and shipped to North Carolina State University Laboratories. At North Carolina State University a series of analysis including grain-size analysis and Pb-210 analysis for sediment accumulation rate data were performed (Nittrouer et al., 1983). The cores were again subsampled for trace metal analysis in the Michigan State University Geochemistry Laboratory. The vials contain- ing the Amazon sediments were placed in a 40°C oven for 96 hours, to completely dry the wet core samples. Each sample was dry sieved with a brass U.S. standard .212 mm sieve Opening (65 mesh) sieve. After each sample, the sieve was placed in a sonic cleaner filled with distilled water for washing. After 10 minutes in the sonic cleaner, the sieve was rinsed with double distilled water and placed on its side to dry to prevent atmospheric or dust contamination (Bopp, 1981). To insure no contamination of the brass sieve on sediment samples, two identical samples were processed; one passing through the sieve, the second unsieved. Differences in the two samples were not statistically 33 34 different, therefore the sieve offered no significant source of contam- ination to the sample. Each sample (5.000 grams dry weight of the less than .212 mm frac- tion) was weighed on a Mettler HL 52 scale and placed directly into pre—treated polyethylene centrifuge bottles. (Pre-treatment of the polyethylene centrifuge bottles included the following: (1) washing with soap and water, (2) rinsing with distilled water, (3) soaking in a l:3 HCL solution overnight, (4) rinsing in double distilled water three times and (5) thoroughly drying). A series of sequential chemical extractions were performed on each 5.000 gram sample (Figure 9). Extractions were performed in order to determine the partitioning of trace metals from four hypothesized fractions: Hypothesized sediment fraction Chemical response of sediment to attack (1) Clay Exchangeable (2) Fe, Mn oxide Hydromorphic Easily and moderately reduced (3) Organic Oxidized (4) Residual Detrital Resisent (Gephart, 1982). The procedure utilized is a combination of attacks performed by Tessier st 21" (1979); Gibbs (1977); Gephart (1982); and Gupta and Chen (1975) and is summarized below: (1) Exchangeable: 40 ml of 1 M MgClz, pH=7 were added to the 5.000 gram less than .212 mm dry sediment sample. The mixture was con- tinuously agitated at room temperature for one hour (Tessier gt al., 1979; Gibbs, 1977). 35 Figure 9. Flow chart of selective chemical attacks. DRY SEDIMENT I .212 mm SIEVED 5.000 gms 40 ml 11M MgClz @ pH=7 for 1 hour 1 CENTRIFUGE : Decant, leachate defined as Exchangeable Fraction 1 NASH RESIDUE (SEE TEXT) 100 mlI.O4 M NHZOH-HCL in 25% (v/v) HOAc @ 96 i 3°C for 6 hours l CENTRIFUGE : Decant, leachate defined as Oxic Fraction l NASH RESIDUE (SEE TEXT) l 15 ml .OZM HNO3 and 25 ml 30% H O @ pH=2 @ 85t 2°O Tor 2 hrs add 15 ml 30% H20 heat @ 85t 2°C for hrs cool add 25 ml 3.2 M NH4OAc in 20% (v/v) HNO3 + distilled water to 100 ml volume CENTRIFUGE : Decant, leachate defined as Organic Fraction NASH RESIDUE (SEE TEXT) DaY (at 40°C overnight) 'T—T'T'l .200 gms sample and 1.000 gms LiBO3 Fuse Dissolve in 5 ml HCL + 50 ml H O Dilute to 1 0 ml volume Defined as Fraction 36 (2) Fe-Mn Oxide: The residue from (1) was leached with 100 ml of 0.04 M NHZOH-HCL in 25% (v/v) HOAc. This extraction was performed at 96 i 3°C for six hours (Tessier 3131., 1979; Gephart, 1982). (3) Organic: The residue from (2) was extracted with 15 ml of 0.02 M HNO3 and 25 ml of 30% H202 pH=2 (adjusted with HNO3) heated to 85i2°C for two hours with occasional agitation (every 15—20 minute agitation). At the conclusion of two hours a second ali- quot of 15 ml 30% H202 was added. The sample was then returned to the water bath of 85i2°C for three additional hours, occasionally agitating. After three hours, the sample was cooled to room tempera- ture, at which time 25 ml of 3.2 M NH4OAc in 20% (v/v) HNO3 and double distilled water (diluting sample to 100 ml) were added. This mixture was stirred continuously at room temperature for 30 minutes (Gephart, 1982; Gupta and Chen, 1975). (4) Residual: 0.200 grams of dry residue from (3) were fused with 1.000 gram of LiBO3 at 1000°C in graphit crucibles for 15 minutes. The fused sample was removed from the furnace and immediately placed in a pre-prepared solution of 5 ml HCL and 50 ml double distilled water. Once dissolved, the sample was diluted to 100 ml with double distilled water. Duplicate fusions were run on each sample (Perkin Elmer Atomic Adsorption Instruction Manual, 1973; Gephart, 1982). Upon completion of each attack, except the residual fraction, the sample was centrifuged for 12 minutes at 15,000 RPM. The supernate was drawn off and stored in pre—treated 250 ml polyethylene storage bottles. The residue sample was then rinsed with double distilled water and centrifuged at 15,000 RPM for 12 minutes, prior to subsequent extraction phases. 37 The supernate from each selective chemical extraction phase and the residual LiBO3 fraction were analyzed by either a flame atomic adsorption 560 spectrOphotometer or by a graphite furnace atomic adsorption HGA-2200 spectrOphotometer. Atomic adsorption control set- tings for the elements analyzed by flame attachment in all matrixes were modified from conditions stated in the Perkin Elmer Atomic Adsorp- tion Instruction Manual (1978). The Instruction Manual does not, how- ever, offer optimized control settings for different matrixes when ana- lyzing by graphite furnace. Optimization for each element analyzed by the graphite furnace for each of the four matrixes had to be performed. Optimization for each element in each matrix analyzed was in accordance with guidelines set forth in the Perkin Elmer Atomic Adsorption Instruction Manual for furnace (1978). Optimized control settings for the graphite furnace found in this study are summarized in Appendix II. All sample sediments were also analyzed for total organic carbon content by a modified Halkley-Black titration method (Gaudette st 31., 1974). A .200 gram dried, less than .212 mm sieved sediment sample was placed in a 500 ml Erlenmeyer flask. Precisely 10 ml of 1 N K2Cr207 solution was added to each sample. The flask is then gently swirled to mix sediment and solution. Next, 20 ml of concen- trated H2504 were added to the flask and again gently swirled for one minute. The mixture in the flask was allowed to stand at room temperature for 30 minutes. A standardization blank was run with each batch of samples. After standing 30 minutes, the solution was diluted to 200 ml volume with double distilled water and 10 ml 85% H3PO4, .20 grams NaF and 15 drops of diphenylamine indicator were added to the flask. 38 The solution in the flask was then back titrated with 0.5 N fer- rous ammonium sulfate solution. The color progressed from an Opaque green-brown to green upon the addition of the ferrous ammonium sulfate solution. The color continuously shifted upon further titration to a bluish-black-grey; at this point the addition of a few dr0ps of ferrous solution shifted the color to a brilliant green end point (a one-drop end point) (Gaudette gt al., 1974). The results of the analysis were calculated by the following equation: Organic Carbon = 10(1-T/S)((.34)(.OO3)(100/.20)) T= sample titrated; S = standardization blank titration; .003 = meq weight of carbon; .34 = normality of K2Cr407; N = weight of sediment sample; 10 = volume of KZCr407 (Gaudette et al., 1974). The total organic carbon (TOC) content is presented in Appendix I. The data from the chemical partitioning attacks were reduced by factor analysis, both R and Q-mode. Factor analysis can aid in reveal- ing simple patterns from complex data sets. Data from the chemical partitioning attacks fulfills the three requirements of factor analysis utilization: (1) conveniently coded -- values in ppm, (2) large number of sample volume -- 84 samples, chemically frac- tionated into four phases, analyzed for 10 elements, and (3) prior knowledge of relationships is inadequate (Kerlinger, 1973; Imbrie and Van Andel, I964). The data were factored by factor analysis in original (raw) form and after transformation of the data by natural log. The transformation of the data was done in order to see if any improvements on correlation could be made since geochemical data frequently does not have a normal distribution. RESULTS AND DISCUSSION Results of the chemical partitioning studies for the metals are presented in Appendix I. These results will be discussed in three parts; general nature of partitioning of metals in the sediments, dia- genesis of metals in the sediments in terms of metal partitioning, and source of trace metals to the sediments as interpreted from R and Q-mode factor analysis of the partitioning data. Chemical Partitioning of Metals in Amazon Shelf Sediments Two aspects of the chemical partitioning of the metals in the sed- iments are discussed: (1) a general summary of the nature of the par- titioning in the sediments and (2) the nature of the partitioning over the shelf. In Table 1 the partitioning results from Appendix I are summarized as a function of sediment type. The sediments in the study area were classified by Nittrouer st 31., (1983) as to type using the classification scheme of Folk (1974). Sediment types are defined as a function of where the sediment plots on a ternary diagram with respect to the three end-member compositions of clay-sand-slit. Numerical representation of sediment types found by Nittrouer st 31., (1983) is shown below in Figure 10: Figure 10. Numerical representation of sediment type from Folk (1974). (MY Type 1 = sand Type 4 = sand-silt-clay Type 2 = sand-silt Type 5 = clayey-silt Type 3 = silt-sand Type 6 = silty-clay 40 Four chemically defined sediment fractions are shown on Table 1 (exchangeable, Fe-Mn oxide, organic matter, and detrital). The values are the average of the percent of the metal in a particular fraction. Since these values are averages of percents, the totals of a metal in a sediment type are not necessarily 100%. Table 1 also present a summary of the work of Gibbs (1977) on the partitioning of selected metals in the suspended load of the Amazon River. Gibbs samples were taken off Macapa' Brazil and therefore would not be influenced by marine water (Gibbs, 1977). Although Gibbs sam- ples were taken approximately five years before the present study sam- ples were taken, little has happened climatically or environmentally that would change the nature of the partitioning of metals in the sus- pended load. It is suggested then that the results of Gibbs (1977) in general reflect the nature of metal partitioning in the suspended sedi- ment today. Gibbs data can therefore be used to study gross dif- ferences in metal partitioning in the sediments between the freshwater environment of the Amazon River and the marine continental shelf envi- ronment. Table 2 is a summary of partitioning of metals in the soils. Soil samples were taken at Belem. A summary of the behavior of each metal follows. SuSpended load sediments refers to Gibbs (1977), shelf sediments refer to surficial (O—Scm) Continental Shelf sediments. Chromium: Most of the Cr (usually greater than 90%) is associated in the residual fractions of the suspended load, shelf sediments, and soils and suggests that most of the Cr addition to the oceans is in the detrital phase of sediments. Of the hydromorphic phases in the shelf sediment, Fe-Mn oxides are the most important in sequestering Cr; 41 00. 0. N8. NN N.e_ 00.0N n.0N 8_.00 88.08 _.o. 00.8N 0.0N N8.8N .0.0N oN. 0N 88¢ 88. 8 08. 8 8N.0 N8. N .8.8 0N.n .0.8 00.0 80.0. 0.N_ No.0 8_.0 08. 0 000 0e.8 0N.0_ 00.8 ¢¢.N_ N._n 00.8N 50.0 88.8 .N.N_ o... 88.N_ 0_.n_ 80.8 oxo 58. 08. 888. 80. 800. 8Ne. 088. N... 08.. oN.N N.N 0N._ 0N.8 x8 >008 8 >008 8 >008 8 8 >008 8 >008 8 8 8 N. 0. N . n N .0 0..- N »058 0 8 0 n N _ 0N... (I. N 0.. N no 08 N .0.80 80.. NN.80 88.N 00.80 N8._8 00.N 8N. 00 N0.N 00.00 .8.00 00.05 880 N0. N0.N N00. .8.N 0.. 0... 80.. 80. NN. _ .Ne. 00. 00.. N8.0. 050 00.N 00.0 00.. «0.8 80.N No.0 00.0 No.N N0.8 8_.N 08.N 0N.0 No.0N oxo 80. «N. 08.. 0N0. 0.0 0.0 0.0 840. 8N0. «on. N_. 00. .N.8 xm >008 8 >008 8 >008 u u >008 8 >008 u 8 8 N. 0. N _ n N .0 u--- N »as8 0 8 0 n N _ 8N; I..- N 0,. N 00 00.5 .8.0N N_.0 .8.0N 80.8 .o.oe N0.08 00.N_ 00. 08 N._N 00.88 00.88 00.88 880 08.. NN.0 N0._ No.8 0N. .N.8 0N.08 on. 08. _ n... 08.. 80.8 N0. 8 050 08.8 00.88 N.N. 80.08 .N.0 .N.08 00.8e 00.8 88.8N N.0_ 0N.8n N_.Ne 8..08 oxo o.N_ 0_.8. 8.8. N8.0_ N8._ No.0 N_.__ 8._N 0N.8N .0.0 00.0 Ne.8. 80. X8 >008 8 >008 u >008 8 8 >008 u >008 a u 8 N. 0. N _ n N .0 .0.. 8 »as8 0 8 0 n N _ «NJ I... N 00 N 5: .N._ 00. 80 00.4 .N.88 8.. 00.80 80.N8 80.. 88.80 N... N0.80 88.58 0..00 88¢ 08. N0. _ N8._ 88.. 00. 8... 80.. N0. 00. .... .8.. 88.. 80.8 memo 8N. 08.8 N8.8 N8.0 88. N8.8 NN.8 No._ 00.8 0.0. 8N.N. 80.0 00.8 80x0 00. NNo. 8.. 000. 8.. 800. .o. No. N_. NN. NN. «N. 00.8 0x8 >08 8 >08 8 >08 8 8 >08 8 S>08 u 8 08 N. 0. N . n N .0 u nu »:5m 0 m V m N — 0N3? QIIII ‘ N0 000*L3m 000*L3m 00.3.2.5 000*L3m 000+Lzm wonkLam oonwLam —: .Om 5.058 5.058 5.058 5.058 5.058 ..058 ahmum .855.0 50 oa>+ +coe_vom No :o_+uc:N o no +coe_uom N Lo>_m co~oe< :. :N .z .00 .00 .oo .5: — m4moc 000m00+m . A+coogoa :— mo:_o>v co_+uog+ c. _o+oe +cooLoa 00 80.0508 2 0o 0 000>©o mo+nv crocxc: Lo mc_88_zm :o_+0000 .000.8omN co_+omg0 0.00m000 :o_+0000 00.xo 02-008 co.+oog+ 00000800 >_omoo_\0_00mcmcoxm0 80.0608 00 gossazn .0N0.. c0.08.0.8803 x 00 00 00800 00>0 0502.008 42 80000008 000. 0000 808 00 >0008 .NN0.. 800.0m 0N.8 0..NN 00.8 00..N 80.0 88.08 8N.N0 0N.0 00.NN 0.0. 0N.00 08.8N -- 800 0N0. N0.0 0N0. 00.8 8NN. 00.0 NN.0 00. N0.N 0N.N N8.0 00.8 -- 000 0..8 00.N. 08.8 N0.NN 0N.0 80..8 .0.0N NOHN 0N.0N 00.0 N0.NN 8...N -- 0x0 00.. 00.. 880. 000. 0N.. 08.. 800. N0 8 N0.. 00.N 80.. 0N.. -- x0 >008 8 >008 8 >008 8 8 >008 8 >008 8 8 -- N. 0. N . 8 N .0 -- 8 F028 0 8 0 8 N . 0N.0 -- 8 00 N 8 0N 88.N 8..N0 NN.8 N8.00 008. NN.80 88..0 00.N N8.00 00.N 00.00 0..N0 00.08 800 .N.. .0.0 .0.0 0N.0 .N.. 8N.8 NN.0 0N.0 .N.0 00.. 0N.. ...8 8N.8. 000 .0.N 80.N 80.. N0.0 N0.N 00.0. 0N.8. NN.N 8N.0. 80.. 08.0 00.N 00.80 0x0 808. 80. 00N. 0.8. N.0. 80.. .N0. 800. NN. 8NN. 88. 00.. 0..8 x0 >008 8 >008 8 >008 8 8 >008 8 >008 8 8 8 N. 0. N . 8 N .0 -- 8 0058 0 8 0 8 N . 0N.0 -- 8 00 N 8 .z NN.8 8..N0 .N.0 0N.N0 80.. 08.N0 8N.0N N8.8 00.80 8.0. 08.8N N0.N0 N0.00 N800 N08. NN. 080. 8.. 00. 00. 8.. 08. 08. N0.. N.. NN. N.0 0000 00.8 88.N. .N.0 N8.N. .0.. N8.N. 8..0N N0.N 00.8. 0..0. 0..0N 00.N. 80.00 80x0 000. 80. 8N0. 000. 000. 80. 800. .N.. N8.. .. .0.. 80. 0.0 0x0 >008 8 >008 8 >008 8 8 >08 8 0.>08 8 8 08 N. 0. N . 8 N .0 -- 8 m»0s8 0 8 0 8 N . 0N.0 0-- 8 N0 N 8 ooo0gam 000003m oon803m mom0gam oon8gam oum0gam ouo0gam v__0m 0.058 0.058 0.058 0.058 0.058 0.058 0bwum .800.0 00 7—0150. — Bank 43 08.00 No.80 08.00 N0.00 00.00 08.00 N8.00 .0.00 00.N0 80.00 .8.00 880 .0N. 0N.8 8.. N08. 00. N0. 0 0 N0. 0.. 8N. 000 .z 0 0 0 0 0 0 0 0 0 0 0 0x0 .2 08. N... 08. 08N. 00. 08. N0. 000. 80.. 88. 8N. x0 N8.00 N0.00 N0.00 8N.N0 N0.80 00.00 08..N .N.08 80.0N NN.N0 8N.00 800 N00. 00.. 0.. 88. 0 0N. 08. N0. N8. 08. 00. 000 0. N0.8. 008.8. 00.N. 88... N0.8 N0.0N N..0N 0..00 88.0N 00.0. 0..0. 0x0 00 .00. 88N. 00N. N08. 00. 00.0 .0.. N.0 0N.N 8..N 8..8 x0 0N.N0 N..00 0N.N0 00.0N 00.00 08.88 .0..8 .8.8. N0.8N N8.0. 08.08 800 00 00.0 NN.0 80... N0.0N .0.0. 80.8. .0.08 .N.00 ...80 .N.80 00.08 000 80.N N..0 N0.8 00.. 0..0N 00.0N 0N.8N 80.8 88.0 8N.0 80.8 0x0 00 8N.. 0 00.N 00.. N8.0. 00.0 80.8 00.0. 00.0 08.8 .N.N x0 0.00 00.00 00.00 00. 00. 00. 00 00.00 .N.00 00 00. 800 0 80. 000. 0 0 0 0 0 0 0 0 000 00 0N. ... N0. 0 0 0 0 0 0 0 0 0x0 00 0 0 0 0 0 0 0 N0. 0N. 0 0 x0 80.00 0N.00 0N.00 00.00 .0.N0 80.00 N0.N0 .0.80 80..0 N0.80 0N.00 800 80.. 00. .N. 08. .0.N 00.. N8.. N..N 88. 0.. 00.N 000 0: 0.. 080. 80. 80. 0 0 0 0 0 0 0 0x0 0: N0.8 0..0 00.N No.0. 0 0 0 0 00.0 00.0 .N.0 x0 N..00 00.N0 00.00 00.00 00..0 00.80 00.N0 80..0 0N.80 00.N0 8..00 0800 .0.0 00.0 N0.N 88.0 00.0 00.0 .8.0 00.0 0N.8 80.0 00.8 8000 00 08.0 0N.N 00.N 80. 0N. 00. 08. N0. 00. 08. 0N. N0x0 00 00. N.. .N. NN. 88. 08. 00. 00. .0. N0. N.. .x0 -ILPI- IIRII Ilkli Ilkll. IIhII. Ilkll IIhII Ilull. IILrI. .IIRII -III- . . . . ..8 .8N .8. .... 8.N 0.0 .N 0 0 0 00 -8N -8. -.... 8.N 0.0 -N 0-0 08500.00: ..08 N00oo ..08 c.80m 0m0:_0go 00>_m coN02< 00+ 00 m__om 00+00_08 c. :N ..z .00 8:0 .oo 0:: .00 +0 m:_co_+.+000 .0805 00008 N MAm_0moo.\0_n00mc0zoxw. 44 00.N0 0N.0N 00.08 N8. 8N.N0 .8... 00..8 .0.00 NN.88 0800 88.. NN.N 8N.8 N0.0 N8.N 00.8 00.8 NN.0 .0.8 8000 NN.0 80.0 00.NN 80.00 00..N 00.00 N8.0. 00... 0N.0N N0xo 0N 80.0 N8.N. N0.0N N0.08 8N.8. 8..08 N0.00 N0.0. .0.08 .x0 88 . . ..8 .8N .8. .... 8.N 0.0 .N 0 00 -8N -8. -.... 8.N 0.0 -N 0-0 08:00.00: ..08 N0000 ..08 ..0.+000. .N 0.00N 45 organic matter is second with the exchangeable fraction playing a relatively minor role. This nature of Cr partitioning in sediments is consistent with past work (Gephart, 1982). The relative partitioning of Cr amount in the sediment fraction does not appear to be strongly dependent on grain-type in the shelf sediments. There is, however, significantly more Cr in the hydromor- phic fraction of the sediments composed of sand (Type 1) than the other sediment types, with the Fe-Mn oxide fraction becoming very dominate in the sequestering of Cr. The nature of Cr partitioning in the sand sediment is consistant with the higher amount of hydromorphic Fe in this sediment type with respect to the other sediments (Table 1). In general, there is slightly less metal in the hydromorphic frac- tion of the shelf sediments compared to the suspended load sediments. Organic matter is more important in sequestering Cr than Fe-Mn oxides in the suspended load. The hydromorphic fraction of the soil better reflects the partitioning in the suspended load rather than the shelf sediments, in that the organic fraction is dominate. The exchangeable and Fe-Mn oxide fractions have very minor roles in the sequestering of Cr in the soil samples. The nature of Cr partitioning in the Amazon River environment sug- gests that Cr undergoes a repartitioning from river to shelf (e.g., organic fraction to Fe-Mn oxide fraction) and that the suspended sedi- ment could be a source for dissolved Cr to seawater (e.g., less Cr in hydromorphic fraction of shelf sediments). The direction of reparti- tioning (organic matter to Fe-Mn oxide) suggests that organic matter could be a source of the dissolved Cr within the sediment. 46 Manganese: The Fe-Mn oxide fraction and residual fraction, combined account for the majority of Mn concentration in both suspended load and shelf sediments (approximately 80-95%). The amount of Mn in the Fe-Mn oxide fraction, the dominate fraction, decreases from suspended load to marine environment, whereas Mn in the detrital fraction in both sus- pended load and shelf remains similar (33%). The Mn organic fraction concentrations are of minor importance (5%) and are similar for shelf and suSpended load sediments. The exchangeable fraction accounts for less than 1% total Mn in the suspended load sediments but increases in the shelf sediments to 15%. In general Mn can be quite mobile in this system since most of the Mn occurs in the hydromorphic phase. Mn partitioning in the oxide, organic and detrital fractions of shelf sediments, appear to be influenced by grain-type. There is more Mn hydromorphics in clay sediments (Type 6) than in the other sedi- ments. Within the hydromorphic fraction, the oxide fraction is domi- nate and increases from sand (Type 1) to clay (Type 6) sediments, 35% to 55% respectively. The organic fraction behaves similar to the oxide fraction (e.g., Type 1 at 1.8% increases to Type 6 at 55%). This Mn behavior is contrary to Cr hydromorphic behavior which decreases from sands (Type 1) to clays (Type 6) and Cr detrital fraction behavior which increases from sand (Type 1) to clay (Type 6) sediments. The Mn in soils have a different distribution pattern among the four fractions than either the suspended load or shelf sediments. 0f major importance in Mn sequestering in soils is the detrital fraction (at greater than 89%). The hydromorphic fraction is of minor impor- tance with the oxide fraction being the most insignificant. This is a drastic change from shelf and suspended load sediments which had the 47 oxide fraction being of primary importance. In general, the data from Table 1 indicates that Mn undergoes repartitioning within the hydromor- phic fraction from a river to a marine environment (e.g., oxide frac- tion to exchangeable fraction). 222315: ‘The detrital fraction is the fraction accounting for most of Co in shelf sediments and soils (90% and 99%, respectively). The domi- nance of the detrital fraction over the hydromorphic fractions has been found previously (Loring, 1979). The detrital fraction is also domi- nate (44%) versus the exchangeable, organic or oxide fractions. In the suspended load sediments, the hydromorphic fraction is the dominate fraction for Co concentration. Within the hydromorphic fraction for the suspended load sediments, the Fe-Mn oxide fraction dominates (28%). of secondary importance is the organic fraction (19%), and of least importance is the exchangeable fraction (8%). The partitioning of Co in shelf sediments and soils are similar and differ from suspended load sediments. Shelf sediments and soils have 90-99% of Co sequestered in the detrital fraction with the hydro- morphic fraction playing a minor role. However, in the suspended load (River) sediments the hydromorphic fraction dominates (e.g., 56%). There are slight grain-type variations observed in Co. The hydro- morphic fraction increases from sand (Type 1) to clay (Type 6) sedi- ment. Co partitioning among various grain-types in shelf sediments is similar to Mn partitioning (e.g., the hydromorphic fraction increases from Type 1 to Type 6). Generally, the nature of Co partitioning in this system suggests that Co goes through repartitioning from river to shelf (e.g., hydro- morphic fraction to detrital fraction) and that the change in relative 48 importance of the hydromorphic fraction (56% to 6%) could be a source for dissolved cobalt to seawater. £22233; Approximately 79% of the Cu is associated in the detrital fraction (suspended, shelf, and soil horizons). Therefore, most of the Cu in sediments brought into the ocean is in the detrital phase. The hydromorphic fraction (suspended and shelf) comprise the remaining 21% of Cu sequestered in sediments. Within the hydromorphic fraction, the oxide fraction is dominate (8-13%), followed by the organic fraction (6%) and the exchangeable fraction (2-5%). Comparing the organic and detrital fractions for suSpended loads and shelf sediments, similar percent concentration values exist (6% organic fraction and 79% detri- tal fraction). However, the exchangeable fraction draps from 5% to 2% and the oxide fraction increases from 8% to 13% in suspended versus shelf sediments. There is little significant change among the different grain-types (Types 1-6) in the detrital or hydromorphic fractions, although, there is a slight influence of grain-type on the exchangeable fraction. The exchangeable fraction decreases from Type 1 (2%) to Type 6 (.3%). The nature of Cu partitioning between a soil core (sample from Belem) and soil horizons (A-D) are different. Soil horizons have a similar partitioning trend for Cu in shelf and suspended load sedi- ments. The soil core has a residual fraction of Cu relative percent value of 23-55%. The hydromorphic fraction has a relative Cu concen- tration value of 21% for suspended load sediments, shelf sediments and soil horizons and increases to 45-77% for soil core. In general, there is little repartitioning between the hydromor- phic and detrital fractions from river to shelf environments. But, 49 within the hydromorphic fraction repartitioning occurs. Repartitioning of Cu from river to shelf environments is via the exchangeable fraction to the oxide fraction. This suggests that the oxide fraction is play- ing a greater role in sequestering Cu and that the exchangeable frac- tion may release Cu to seawater (i.e., be a source of Cu to seawater). .1392; The major phases of Fe in the suspended sediments are the detri- tal fraction (46%) and oxide fraction (46%). There is a dramatic increase (46%-82%) in Fe detrital fraction and subsequent decrease in hydromorphic fraction from suspended load to shelf sediments. The dom— inate Fe fraction within the hydromorphic phase of the shelf sediments is the oxide fraction (17%). The organic and exchangeable phase are of minor significance. Shelf sediments are slightly grain-type dependent. The detrital fraction in Type 1 is at a lower relative percent value than Type 6 (73% to 82%, respectively). There is a corresponding decrease in hydromorphic fraction from Type 1 to Type 6. Soils (core and horizon) and shelf sediments have similar parti- tioning data. The detrital fraction dominates Fe partitioning with the oxide fraction of secondary importance. The organic fraction and exchangeable fraction are of minor importance. Repartitioning does occur in Fe sediment concentrations between river and shelf sediments. The hydromorphic fraction decreases Fe con- centration values while the residual fraction increases from river to shelf environments. This suggests that the suspended load can be a source of Fe to seawater. M15531. In the suspended load sediments the relative partitioning of Ni is dominated by the oxide fraction (45%) and the detrital fraction 50 (38%). 0f minor importance are the organic fraction (13%) and the exchangeable fraction at 3%. The shelf sediments have different parti- tioning of Ni. Most of the Ni in the shelf sediments are in the detri- tal fraction (87%). Partitioning of Ni in the hydromorphic fraction of the shelf sediments is oxide fraction (7%), organic fraction (5%), exchangeable fraction 1%. Shelf sediments appear slightly dependent on grain-type. The residual fraction has a slightly higher concentration percentage value in Type 1 grain-type (94%) as opposed to Type 6 (87%). Corresponding to this change, the hydromorphic fraction increases from Type 1 (6%) to Type 6 (13%). Soils, both horizons and core samples, behave in a similar parti- tioning pattern as shelf sediments but not as suspended load sediments behave. Soils appear to have one dominate fraction (detrital 95—99%) for Ni. The hydromorphic fraction is minor and within this fraction, the oxide fraction is at undetectable levels. The nature of Ni partitioning from suspended load to shelf envi- ronment suggests Ni repartitioning. The oxide fraction in the sus- pended load sediment repartitions to the detrital fraction in shelf sediments. This repartitioning of hydromorphic fraction (especially the oxide fraction) can be a potential source for Ni to seawater from suspended load sediments. 21355 Gibbs (1977) did not analyze for Zinc in suSpended load sedi- ments and therefore there will be no comparison between suspended load and shelf sediments for zinc. However, partitioning data for zinc is available for soil samples, shelf sediment and shelf sediment changes with grain-type (Table 1). The residual fraction (73% for shelf 51 sediments) is the dominate fraction of Zn in shelf sediments. The oxide fraction is of secondary importance at 21% with both organic (5%) and exchangeable (1%) fractions being of minor importance. Grain-type dependency is observed in the residual as well as hydromorphic fractions. There is an increase in detrital Zn for Type 1 (64%) to Type 6 (77%). Corresponding to this increase, is a relative decrease in the hydromorphic fraction (Type 1 at 36% to Type 6 at 23%). All three individual phases (exchangeable, oxide, organic) within the hydromorphic fraction decrease from Type 1 (sand) to Type 6 (clay). The soil horizons display a similar behavior as shelf sediments. The detrital phase is dominate (76-78%) and the hydromorphic is of lesser importance. However, soil core samples differ. They display two dominate fractions, the detrital and the oxide. Insmmmw: (1) Only Cr and Cu have the majority of their concentration in the residual fraction for all three environments (suspended load, shelf, soil). (2) Two fractions dominate the partitioning of Mn, Co, Ni, Fe in suspended load sediments (almost equal fractions): oxide and detrital. (3) In general, most metals (90%) are in a combination of residual and oxide fractions. (4) The hydromorphic fraction is dominate in controlling the behavior of Mn in both suspended and shelf environments. (5) All metals except for Mn and Cu show a decrease in the organic fraction from river to shelf environments. (6) All metals except for Mn and Cu show a decrease in the per- centage of the metals in the hydromorphic fraction from river to shelf environments. (7) In general, the relative partitioning trends as a function of grain-type are preserved for all metals. 52 (8) The hydromorphic fraction for Mn, Co, Cu, Ni all have grain- type trends which increase element concentrations from Type 1 and Type 6, Cr and Zn have a reverse trend. (9) The relative partitioning within the hydromorphic fraction is highly variable between soil and shelf environment, except for Cr, the relative partitioning of the metals between the soils and suspended material are not similar. (10) Cr, Co, Cu, Fe, Zn, Ni are partitioned similarly in shelf sediments (detrital 73-94%; oxide 4-27%; organic .2-6%; exchangeable 0-2%). (11) Relative partitioning of Mn, Cr, Cu in the sediment percent is the same in river and shelf environments. (12) Differences were observed in the relative partitioning of Fe, Co, and Ni between the sediments at the river and those of the shelf. The hydromorphic fraction (especially the oxide fraction) is much smaller in the shelf sediments. The data can be interpretated to suggest that: (1) Significant repartitioning occurs between water and shelf environments. (2) Upon mixing with seawater the sediment could be a source of metals to seawater, the order of importance of this source decreases in the order Ni>Co>Fe>Cr. (3) Decay (oxidation) of organic matter could be the source of metal to seawater within the sediment. Distribution of Metals on the Shelf: The behavior of selected trace metals in this system were studied by the construction of contour maps of elemental concentrations in the surface sediment (Figures 11-13). Distribution maps were created for Zn and Ba. Zinc behavior was studied as a representative of a metal with both significant hydromorphic as well as residual partitioning. 53 Figure 11. Contour map of zinc in the detrital fraction of the surface (9-5 cm) sediment of the Amazon Continental Shelf. 54 DETRI'I'AI. ZINC 55 Figure 12. Contour map of zinc in the hydromorphic fraction of the surface (8-5 cm) sediment of the Amazon Continental Shelf. 56 HYDROMORPHIC ZINC Marcie; Island 57 Figure 13. Contour map of total zinc in the surface (0-5 cm) sediment of the Amazon Continental Shelf. 58 ZINC TOTAL Moroio' Island 59 Ba was chosen for study because it is an alkaline element and should behave differently than the transition metals studied. Concentrations of zinc in the continental shelf sediments are shown in Figures 11, 12, and 13 representing absolute values of Zn in the detrital fraction, hydromorphic fraction and total sediment, respectively. The concentration patterns are similar in each figure and appear to reflect the distribution of sediment types on the shelf (Figure 5). The concentration of Zn decreases from core location 2 (Type 6) to core locations 3,4,5,6 (Type 1). Core locations 30 and 33 (Type 1) have corresponding lower concentrations than core locations 31 and 32 (both of Type 5). Core locations 35, 29 and 28 all of similar type (Type 6) display similar concentrations. Cores 47-57, again reflect grain-type composition; they display similar grain-type compositions (Type 5) and therefore similar concentration values. The dependency on grain-type for metal concentrations in the shelf sediments can be further demonstrated by plotting elemental concentra- tion in the sediment versus grain-type. Figures 14, 15, and 16 are such plots for Zn-total, Zn-residual and Zn-hydromorphic, reSpectively. These Figures show distinct clusters of Zn concentrations depending on grain-type, with the biggest difference between Type 1 and Types 5 and 6 (sand and silt-clay, respectively). If the concentrations are plotted as ranked data, rather than as absolute data, the differences are made even clearer. For example, Figure 17 is a plot of sediment type versus Zn hydromorphic using ranked concentrations. One can clearly see that the concentration distribution of the Zn in the shelf sediment can be explained by the distribution of the various sediment Figure 14. Figure 15. 6O Plot of sediment type versus concentration of zinc total. Numbers refer to overlapped points. Plot of sediment type versus concentration of zinc in the detrital fraction. Numbers refer to overlapped points. SEDINENT TYPE I 6.0+ 4.0+ SEDIMENT 1w: 2’ I I II 2 II I 0.0+ 61 I I 2:: III3 I I II II I3 aI I 2 x I I III 2x I I I x t f Jr 4 4. fi‘ 0. 30. 60. 90. I20. 150. PPM ZINC TOTAL I I. I ZII2II I II I I 23 2II I I I I 2 I 3 I 0 ; 4. 4 4. #8 5. 30. 55. BO. 105. I30. PPM ZINC DETIITAL Figure 16. Figure 17. 62 Plot of sediment type versus concentration of zinc in the hydromorphic fraction. Numbers refer to overlapped points. Plot of sediment type versus concentration of zinc in the hydromorphic fraction plotted as ranked data. SEDIMENT TYPE SEDIMENT TYPE 63 I 2I I4 2I " 2 I332: I 2 I I I I 2 I“ 2 II I -O 7.0 N.0 2l-O 28-0 35.0 PPM ZINC HYDROMOIIPHIC I II I IIIII II I IIIIII I Illlx I I I I I I I II II IIII I .0 0 0' : 4. a 0- lo- 20. 30. 40. SO- ZINC HMOMORPHIC RANKED 64 type over the shelf. The dependency of metal concentration on grain— type is supported by the works of Gupta and Chen (1975), Loring (1979). On the Amazon Continental Shelf, grain-type appears to control metal concentration. These results suggest that physical processes (mechanical deposi- tion as a function of sediment type and shelf hydrodynamics) are mainly reSponsible for the distribution of metals on the shelf. In an attempt to eliminate the effect of grain-type on metal con— centrations when interpretating partitioning data, ratioed concentra- tions were studied. By using ratios, other possible controls (biogenic and geochemical) on metal distributions in shelf sediments can be deduced (Long and Gephart, 1982). Ratios of metal concentration to Al were used to study metal dis- tribution independent of biogenic effects (Sholkovitz and Price, 1980). Figures 18 and 19 are concentration maps of Zn residual/Al residual and Zn hydromorphic/Al total, respectively. Both maps show a strong depen- dency on grain-type and suggests that biogenic controls on the distri- bution of Zn in the shelf sediment is minor. Zn, Ba/Fe + Mn oxide ratios were also utilized. The Fe + Mn oxide ratios were chosen because of the dependency of the oxide fraction on grain-type and by utilizing the ratio, normalization with respect to grain—type dependency was sought (Tessier st 21:, 1982). Figure 20 is a concentration map of Zn oxide/Fe + Mn oxide. The Zn oxide ratio has a uniform distribution across the shelf. Figure 21 is of Ba oxide/Fe + Mn oxide. The Ba oxide ratio has a high concentration (dump) at the mouth of the river and values decrease away from the coast. Figure 21 could be related to salinity, as the river waters mix with the oceans a 65 Figure 18. Contour map of the ratio zinc in the detrital fraction divided by aluminum in the detrital fraction in the surface (8-5 cm) sediments of the Amazon Continental Shelf. 66 z- DETRITAL/Al 02mm 67 Figure 19. Contour map of the ratio hydromorphic zinc concentration divided by total aluminum concentration in the surface (0-5 cm) sediments of the Amazon Continental Shelf. 68 Zn HYDROMORPHIC/AI TOTAL Moroio' Island . ' 69 Figure 20. Contour map of the ratio zinc in the oxide fraction divided by iron and manganese in the oxide fraction in the surface (8-5 cm) sediments of the Amazon Continental Shelf. 70 Zn ()0 D E/Fo -M- 0 XIDE Mar ' ' 010 lslmd . 71 Figure 21. Contour map of the ratio barium in the oxide fraction divided by iron and manganese in the oxide fraction in the surface (0-5 cm) sediments of the Amazon Continental Shelf. 72 Ba OXIDE/Fe - Mn OXIDE __) z 1.23:0 - ' 020 so mums-.“u’a' atom Int 20 «vol Moroio’ Island 73 drop Miterrigenous sediment load occurs before 3 0/00 salinity (Nfilliman.gtqal., 1975) (Figure 3). Both maps (Figures 20 and 21) no longer show the strict grain-type dependency for the distribution of the metal in the sediment, each map having a slightly different distribution pattern. Figures 22 and 23 show Zn hydromorphic/Zn residual and Ba hydro- morphic/Ba residual, respectively. Both Figures (22 and 23) display an increase in hydromorphic/detrital ratios as one moves in a direction across the shelf, away from the coast. Also observed in Figure 22 is the high ratio Zn values associated with sediments of Types 1 and 2 (sands). These Figures suggest that either the hydromorphic fraction is a source for metals to seawater resulting in a decrease in hydromor- phic fraction or that metal is repartitioned from the hydromorphic fraction to the detrital resulting in an increase in residual fraction to account for observed ratio trends. Diagenesis of Metals in Amazon Shelf Sediments: Sediment samples were taken as a function of depth at locations 10 and 42 (Figure 1). Core 42 and core 10 were found to have similar trends and therefore only core 10 will be discussed. Metal concentra- tions versus depth profiles are in Figures 24-31 for core 10. Figure 24 shows sediment type for samples with depth. Figure 25 is a plot of Cr concentration with depth and shows that the organic and oxide phases are of minor importance in controlling Cr concentration and have a con- stant value throughout the core. Chromium concentrations in the detri- ‘tal phase, however, show several fluctuations. There is a decrease at 50 cm (silt composition), an increase at 220 cm (clay composition), a 74 Figure 22. Contour map of the ratio zinc in the hydromorphic fraction divided by zinc in the detrital fraction in the surface (w-S cm) sediments of the Amazon Continental Shelf. 75 Zn HYDROMORPHlC/Zn DETRITAI. 76 Figure 23. Contour map of the ratio barium in the hydromorphic fraction divided by barium in the detrital fraction. 77 Ba HYDROMORPHlC/Ba DETRITAL Figure 24. 78 Sediment type with depth for core 10 (Nittrouer gt_gl., 1983). O 79 7 HS; .0.. ‘ w-Si"‘C|ay 100- . be 4 M’SW'C'W CLAY ‘30‘ 4 W‘SI"’C'OY 1 Sand no 6 -‘ 250— 0 8W - 5 s wsm 4 l ”0‘ ‘ 5"“ smo sm ‘ said—Silt-{loy ’50- 4 w-“f'clay ‘00- C be O (by 450‘ SEDIMENT TYPE CORE )0 80 cowpmgucmucoo Lu Page» i Hop . :o_pomc$ puu_cumu u mum . co_uomgc muwxo . oxo < :owuumg$ uwcmmgo . omo —- “hex Fonsxm .oH mgoo c_ zuamu zpw; sswaocgu mo cowumcucmocou .mN mg=m_u 81 53.2083U 0— ECU .9 0.3 am. 3. a... o... ow 82 decrease at 310 cm (sand), and an increase at 350 cm (silt composi- tion). This suggests for Cr that the detrital fraction is grain-type dependent. Iron concentration versus depth (Figure 26) shows relatively con- stant concentration values for Fe in the oxide and detrital phases. However, a slight decrease of Fe in these fractions is observed at 360 cm where a compositional change from sand-silt—clay to silty-clay occurs. The exchangeable, and organic phases show a fluctuation in Fe concentration at 100 cm; the exchangeable phase increases while the organic phase decreases. At 200 cm there is a decrease in Fe concen- tration and at 350 cm, another decrease in both exchangeable and organ- ic phases. These changes in concentration reflect changes in composi- tion. (Type 4 has a higher concentration value than Type 6). This suggests that for Fe, grain-type plays a significant part in metal concentration. Mn concentration with depth (Figure 27) shows that the amount of Mn in the exchangeable, residual and organic fractions is fairly con- stant with depth. Manganese in the oxide fraction (the major phase of Mn in the sediment) changes with depth as a function of sediment type. The highest concentration values are in the sand-silty-clay composi- tion, the lower concentrations appear in the sand and silty-clay compo- sitions. Thus, for Mn, the concentration changes with depth and nature of its partitioning suggests that Mn concentration is grain-type depen- dent and that the total Mn concentration pattern reflects the Mn oxide pattern. Figure 28 is a plot of Zn concentration versus depth. The Zn plot shows that concentrations in the organic and oxide phases are constant 83 :o_umcucmucou mm Pmuou . Heb . cowuomg» pmuwgumu u mum . cowuomcm mv_xo . oxo 4 cowuomcw u_:wmgo . umo -- :o_uumcm w_nmmmcm;oxo . xm o "max ponsxm .oH «Lou =_ suaou :pwz :og_ mo cowumgucwucou .om mg=m_u 84 208. o. m¢OU ...Oh OX0 00 Xm (8 85 =o_»mcpcmu:ou :2 page» u FOP l co_uomgw quwgumv . mum . :owuomgw wvwxo u oxo 4 cowuomc$ uwcmmco u omo -— cowuomc$ mpnmmmcmsoxm . xm o "aux Fonsam .oH mcoo cw gamma saw: mmwcmmcme yo covumcuzmocou .NN mesmwu 86 mmm2<02<<¢ o— w¢OU AX THU? &. .r ..r a. a. .r a. 8:623: it or L: 0.. m. ééééééé g s 87 Figure 28. Concentration of zinc with depth in core 10. Symbol key: II ORG - organic fraction A OXD - oxide fraction I RES - detrital fraction - TOT - total Zn concentration i DEPTH IN CM 2 9 330- 88 nm x "MIMIC 40 BO .0 70 .0 DO 100 110 120 130 1‘0 4 1 4 1 J _1 1 1 n I I ’J ” a a I f" L’ \\~ \~‘ \~ ‘~ ‘s W I -—'J A--" ”’ ’ I”’ ” “‘ \ “‘ ~~ ~g \ \ \ \\ ‘5 -~‘\ CORE IO ZINC 89 with depth except at 330 cm. There is a decrease of Zn in the organic phase and an increase of Zn in the oxide phase. There are two fluctua- tions of Zn in the residual phase with depth: at 150 cm and at 310 cm. Both depths show a decrease in detrital concentration of Zn. At 150 cm and 310 cm depths the sediment composition changes to the sand (Type 1). Other compositional changes (sand-silty—clay and silty-clays) appear to have no discernable effect on concentrations of Zn in the sediment. Figure 29 is a plot of Cu concentration with depth. Copper in the oxide fraction appears constant with depth and is relatively small. The organic fraction also is constant with depth except for an increase at 300 cm. The organic fraction increases drastically due to encoun- tering a grain-type compositional change at this point. Copper in the detrital fraction is of dominate control on its concentration in these sediments and fluctuates throughout the sediment core. The Cu detrital fraction appear to be controlled by the sediment type (higher concen- tration values are associated with clay type sediment and lower concen- tration values are associated with sand type composition). Figures 30 and 31, nickel and barium concentrations with depth, respectively, show similar grain-type dependencies. The organic and oxide fraction are relatively unimportant and remain constant with depth. Little repartitioning occurs with depth. The detrital fraction has two major low level concentration points at 150 cm and 300 cm for both barium and nickel profiles. These low level concentration points correspond to a sand grain-type (Type 1). High concentration levels correspond to a clay grain-type (Type 5 and Type 6). Nickel and barium Figure 29. 90 Concentration of copper with depth in core 10. Symbol key: II ORG - organic fraction A OXD - oxide fraction 0 RES - detrital fraction - TOT - total Cu concentration 4__.._..—.__—__4 - 91 92 cowumcucmocoo Pz quop . he» . cowpumcw Pwuwgpmu . mum . cowuumcw muwxo . oxo 4 558.; 0.23.8 .. $5 I "xmx ponsxm .oH wcoo cw :uamu cpwz meuwc mo cowumgpcmucou .om 925.: 93 amxgz O— mcou ...... be (we. O. o.v an of 3x22 in W3 NI Hun 94 cowpmcucmocou mm pope» . ho» . cowgumcy FouPLumu . mum . "aux FonEAm .oH mcou cw guamu ;u_z Ezwcmn we co_umcacmucou .8 95m: 55.3 9 EU 95 on. 00¢ 00. 0 96 concentrations with depth are, therefore, grain-type dependent and show little repartitioning. In general, there exists a relationship between sediment type and concentration of the detrital fraction. The hydromorphic phase is more independent of sediment type with depth than they are at the surface (excluding Mn oxide fraction). The relative importance of the hydro- morphic phases in sequestering metal does not indicate major reparti- tioning of the metals with depth within the hydromorphic fraction. These observations can be interpreted to indicate that to a depth of 450 cm on the shelf, the metal concentrations in the sediment are dependent on sediment type. However, the relatively high amount of Mn in the Fe-Mn oxide fraction results in the profile of total Mn concen- tration with depth to be independent of sediment type as the Mn in the oxide fraction gets redistributed. Core 10 sediments are also considered oxic throughout the core length. The lack of decreasing metal concentration in the oxide phase and no observable oxidized sediment zone enrichment for Ni, Fe, Mn, and Co which is expected when reducing conditions are encountered [Addy st 31,, 1976; Holmes and Martin, 1978; Heath and Dymond, 1981]. These observations lead to the conclusion that core 10 does not encounter a reducing environment and is therefore considered oxic throughout the core. segues: Partitioning data for the surface (0-5 cm) shelf sediments were arranged in a Q-mode and R-mode factor analysis array. The data was factored in raw form and in transformed form (i.e., natural log matrix) (Loring, 1982). The B-matrix for factor analysis of raw R-mode, 97 transformed R-mode, transformed R-mode PA1, raw Q-mode, and transformed O-mode PA1 appear in Appendix III. Q-mode will be discussed below. The results of R-mode factor analysis will not be discussed, however the data appears in Appendix III. PA1 (principal factoring iteration) was employed because one factor dominated. PA1 may extract other vari- ables which cannot be determined by other factor methods (Statistical Package for the Social Sciences, 1975). The B-matrix is the influence of each end member on the total metals found in each sediment sample (Bopp and Biggs, 1981). Q-mode factor analysis, both raw and transformed PA1 forms, show high communality (i.e., squared multiple correlation) of about .9 and has one factor accounting for greater than 90% factor influence on the shelf. In conclusion, only one factor appears to be controlling the source of trace metals to the shelf sediments. This factor is interpreted to indicate that the Amazon River is the dominant source. Other factors accounting for an anthropogenic source or a biogenic source are undetectable. A hydrothermal source is also undetectable, which is in support with past work (Dymond and Heath, 1982). CONCLUSIONS This research examined the chemical partitioning of Fe, Mn, Zn, Cu, Pb, Ni, Cr, Co, Ba and Al among four theoretical phases within continental shelf sediments and soils from Belem. The conclusions of this study are: (1) The Amazon River is dominate over all other sources of trace (2) (3) (4) (5) (6) metals to shelf sediments. Significant repartitioning of metals in sediment occurs when sediments move from a river to an ocean environment. The suspended load sediment of the river can be a potential source of metals to seawater, when the sediment enter seawater. The order of importance for this source decreases in the sequence Ni>Co>Fe>Cr. Metal concentrations and distribution in and on the shelf are mainly dependent on the grain-type of the sediment. Amazon shelf sediments do not encounter reducing conditions to a depth of 450 cm. Little repartitioning of metals in the sediment occurs with depth. 98 REFERENCES 90 Figure 29. Concentration of copper with depth in core 10. Symbol key: II ORG - organic fraction A OXD - oxide fraction 0 RES - detrital fraction - TOT - total Cu concentration 91 CORE l0 COPPER 92 :owumcucmucoo wz —muou u wow u :owuumgw quwgumu . mum o :owuomcw muwxo . :xo 4 :o_uomc$ owcmmco . wmo - $8. Page? .oH wcou cw cpamu saw: meuwc mo cowpmcacmocou .om «Lsmwu 93 any-v.2 o— m¢OU ...... 5. In“. O. a... cc .3822 2.: o.» W3 NI Hun 94 cowumcpcmucoo mm _muou . ho» . cowpumcw Fouwcumu . mum . ”aux Fonsam .oH mgoo cw sumac guy: Eswcmn Co cowumcucmucoo .Hm mesmwu 95 $5.3 o— meow out 009 Oop 96 concentrations with depth are, therefore, grain-type dependent and show little repartitioning. In general, there exists a relationship between sediment type and concentration of the detrital fraction. The hydromorphic phase is more independent of sediment type with depth than they are at the surface (excluding Mn oxide fraction). The relative importance of the hydro- morphic phases in sequestering metal does not indicate major reparti- tioning of the metals with depth within the hydromorphic fraction. These observations can be interpreted to indicate that to a depth of 450 cm on the shelf, the metal concentrations in the sediment are dependent on sediment type. However, the relatively high amount of Mn in the Fe-Mn oxide fraction results in the profile of total Mn concen- tration with depth to be independent of sediment type as the Mn in the oxide fraction gets redistributed. Core 10 sediments are also considered oxic throughout the core length. The lack of decreasing metal concentration in the oxide phase and no observable oxidized sediment zone enrichment for Ni, Fe, Mn, and Co which is expected when reducing conditions are encountered [Addy gt al,, 1976; Holmes and Martin, 1978; Heath and Dymond, 1981]. These observations lead to the conclusion that core 10 does not encounter a reducing environment and is therefore considered oxic throughout the core. game: Partitioning data for the surface (0-5 cm) shelf sediments were arranged in a Q-mode and R-mode factor analysis array. The data was factored in raw form and in transformed form (i.e., natural log matrix) (Loring, 1982). The B-matrix for factor analysis of raw R-mode, 97 transformed R-mode, transformed R-mode PA1, raw Q-mode, and transformed Q-mode PA1 appear in Appendix III. O-mode will be discussed below. The results of R-mode factor analysis will not be discussed, however the data appears in Appendix III. PA1 (principal factoring iteration) was employed because one factor dominated. PA1 may extract other vari- ables which cannot be determined by other factor methods (Statistical Package for the Social Sciences, 1975). The B-matrix is the influence of each end member on the total metals found in each sediment sample (BOpp and Biggs, 1981). Q-mode factor analysis, both raw and transformed PA1 forms, show high communality (i.e., squared multiple correlation) of about .9 and has one factor accounting for greater than 90% factor influence on the shelf. In conclusion, only one factor appears to be controlling the source of trace metals to the shelf sediments. This factor is interpreted to indicate that the Amazon River is the dominant source. Other factors accounting for an anthropogenic source or a biogenic source are undetectable. A hydrothermal source is also undetectable, which is in support with past work (Dymond and Heath, 1982). CONCLUSIONS This research examined the chemical partitioning of Fe, Mn, Zn, Cu, Pb, Ni, Cr, Co, Ba and Al among four theoretical phases within continental shelf sediments and soils from Belém. The conclusions of this study are: (1) The Amazon River is dominate over all other sources of trace metals to shelf sediments. (2) Significant repartitioning of metals in sediment occurs when sediments move from a river to an ocean environment. (3) The suspended load sediment of the river can be a potential source of metals to seawater, when the sediment enter seawater. The order of importance for this source decreases in the sequence Ni>Co>Fe>Cr. (4) Metal concentrations and distribution in and on the shelf are mainly dependent on the grain-type of the sediment. (5) Amazon shelf sediments do not encounter reducing conditions to a depth of 450 cm. (6) Little repartitioning of metals in the sediment occurs with depth. 98 REFERENCES REFERENCES Addy S. K., Presley B.J. and Ewing M. (1976) Distribution of manganese, iron and other trace elements in a core from the Northwest Atlantic. J. Sed. Pet. 46, (4), 813-818. Bender M., Broecker H., Gornitz V., Middel U., Kay R., Sun 8.5. and Biscaye P. (1971) Geochemistry of 3 cores from the East Pacific Rise. Earth and Planet. Sci. Lett. 12, 425-433. Bischoff J.L. and Dickson F.w. (1975) Seawater-basalt interaction at 200°C and 500 bars: Implications for origin of seafloor heavy-metal deposits and regulation of seawater chemistry. Earth and Planet. Sci. Lett. 25, 385-397. Bonatti E., Fisher D.E., Joensuu O. and Rydell H.S. (1971) Post depositional mobility of some transition elements, phosphorus, uranium, and thorium in deep sea sediments. Geochim. Cosmochim. Acta 35, 189-201. Bopp F. 111 and Biggs R.B. (1981) Metals in estuarine sediments: Factor analysis and its environmental significance. Science 214, 441-443. Bopp F. III (1981) Ph.D. Thesis: University of Delaware, 1-179. Brannon J.M., Rose J.R., Engler R.M. and Smith I. (1977) The distribution of heavy metals in sediment fractions from Mobil Bay, Alabama. Chemistry of Marine Sediments, (ed. T.F. Yen), 125-149. Brooks R.R., Presley B.J. and Kaplan I.P. (1968) Trace elements in the interstitial waters of marine sediments. Geochim. Cosmochim. Acta 32, 397-414. Brumsack H.J. (1980) Geochemistry of cretaceous black shales from the Atlantic Ocean. Chem. Geol. 31, 1-25. Corliss J.B., Dymond J., Gordon L.I., Edmond J.M., Von Herzen R.P., Ballard R.D., Green K., Hilliams D., Bainbridge A., Crane K., and Van Andel T.H. (1979) Submarine thermal springs on the Galapagos rift. Science 203, (4385), 1073-1083. 99 100 Cosma B., Drago M., Tucci S., Piccazzo M. and Scarponi G. (1979) Heavy metals in Ligurian Sea sediments: Distribution of Cr, Cu, Ni and Mn in superficial sediments. Mar. Chem. 8, 125-142. DeMaster D. (1982) Personal communication. North Carolina State University. Demina L.L., Gordeyev V.V. and Formina L.S. (1978) Forms of occurrence of Fe, Mn, Zn, and Cu in river water and suspended matter, and their variation in the zone of mixing of river and seawater. (As exemplified by Rivers of the Black Sea, Sea of Azov, and Caspian Sea Basins). Geochem. Intl. 15, (4), 146-163. Drever J.I. (1982) The Geochemistry of National Haters, Prentice-Hall, Inc., 163-165. Dymond J. (1981) Geochemistry of nazca plate surface sediments: An evaluation of hydrothermal, biogenic, detrital, and hydrogenous sources. Geol. Soc. Amer. 154, 133-173. Dymond J., Corliss J.B., Cobler R., Muratii C.M., Chou C. and Conard R. Composition and origin of sediments recovered by deep drilling of sediment mounds, Galapagos Spreading Center. 377-385. Edmond J.M. (1981) Hydrothermal activity at mid-ocean ridge axes. Nature 290, (5802) 87-88. Edmond J.M., Craig H., Gordon L.I. and Holland H.D. (1979) Chemistry of hydrothermal waters at 21°N on the East Pacific Rise. E05 60, (46), 864. Edmond J.M., Measures C., McDuff R.E., Chan L.H., Collier R., Grant 8., Gordon L.I. and Corliss J.B. (1979) Ridge crest hydrothermal activity and the balances of the major and minor elements in the ocean: The Galapagos data. Earth Planet. Sci. Lett. 46, 1-18. Edmond J.M., Von Damm K.L, McDuff R.E. and Measures 0.1. (1982) Chemistry of hot springs on the East Pacific Rise and their effluent disposal. Nature 297, 187-191. Eisma D. and Van Der Mavel H.N. (1971) Marine muds along the Guyana Coast and their origin from the Amazon Basin. Contrib. Mineralogy and Petrology 31, 321-334. Folk R.L. (1974) Petrology of sedimentary rocks. Hemphill Publ. Co. 182. Gaudette H.E., Flight, N.R., Toner, L. and Floger, D.N. (1974) An inexpensive titration method for the determination of organic carbon in recent sediments. J. Sed. Pet. 44, (1), 249-253. 101 Gephart C.J. (1982) Relative importance of iron-oxide, manganese-oxide, and organic material on the adsorption of chromium in natural water sediment systems. MS Thesis, Michigan State University. Gibbs R.J. (1965) The Geochemistry of the Amazon River Basin. Ph.D. Thesis, University of California, San Diego. Gibbs R.J. (1967) The geochemistry of the Amazon River system: Part I the factors that control the salinity and the composition and concentration of the suspended solids. Geol. Soc. Amer. Bull. 78, 1203-1232. Gibbs R.J. (1976) Amazon River sediment transport in the Atlantic Ocean. Geology 4, 45-48. Gibbs R.J. (1977) Transport phases of transition metals in the Amazon and Yukon Rivers. Geol. Soc. Amer. Bull. 88, 829-843. Grupta S.K. and Chen K.Y. (1975) Partitioning of trace metals in selective chemical fractions on nearshore sediments. Environ. Letters 10, (2), 129-158. Harding S.C. and Brown H.S. (1976) Distribution of selected trace elements in sediments of Pamlico River, Estuary, North Carolina. Environ. Geol. 1, (3), 181-190. Heath G.R. and Dymond J. (1981) Metalliferous-sediment deposition in time and space: East Pacific Rise and Bauer Basin Northern Nazca Plate. Nazca Plate: Crustal Formation and Andean Convergence. 175-197 0 Hoffmann J.A.J. (1975) Climatic Atlas of South America HMO: Hungary. Holeman J.N. (1968) The sediment yield of major rivers of the world. Water Resources Research 4, 737-747. Holmes C.N. and Martin E.A. (1978) Migration of anthropogenically induced trace metals (barium and lead) in a continental shelf environment. Amer. Chem. Society 672-676. Imbrie J. and Van Andel T.H. (1964) Factor analysis of heavy-mineral data. Geol. Soc. Amer. Bull. 75, 1131-1156. Kerlinger F.N. (1973) Foundations of behavioral research. Holt, Rinehart and Winston, Inc., New York. 582-692. Kharkar D.P., Turekian K.L. and Bertine K.K. (1968) Stream supply of dissolved silver, molybdenum, antimony, selenium, chromium, cobalt, rubidium and cesium to the oceans. Geochim. Cosmochim. Acta 32, 285-298. 102 Kitano Y., Sakata M. and Matsumoto E. (1980) Partitioning of heavy metals into mineral and organic fractions in a sediment core from Tokyo Bay. Geochim. Cosnochim. Acta 44, 1279-1285. Krauskopf K.B. (1956) Factors controlling the concentration of 13 rare metals in seawater. Geochim. Cosmochim. Acta 9, 1-328. Kuehl S.A., Nittrouer C.A. and DeMaster D.J. (1982) Modern sediment accunulation and strata formation on the Amazon Continental Shelf. Marine Geology 49, 279-300. Long D.T. and Gephart C.J. (1982) Relative importance of iron-oxide, manganese-oxide, and organic material on the adsorption of chromium in natural water sediment systems. Proceeding from October Geological Society of America National Conference. Loring D.H. (1976) The distribution and partition of zinc, copper, and lead in the sediments of the Saguenary Fjord. Can. J. Earth Sci. 13, 960-971. Loring D.H. (1976b) Distribution and partition of cobalt, nickel, chromium, and vanadium in the sediments of the Saguenary Fjord. Can. J. Earth Sci. 13, 1706-1718. Loring D.H. (1979) Geochemistry of cobalt, nickel, chromium and vanadiun in the sediments of the estuary and Open Gulf of St. Loring D.H. (1982) Geochemical factors controlling the accunulation and dispersal of heavy metals in the Bay of Fundy Sediments. Can. J. Earth Sci. 19, (5), 930-944. Luoma S.M. and Bryan 8.N. (1981) A statistical assessment of the fonn of trace metals in oxidized estuarine sediments emloying chemical extractants. The Science of the Total Environment 17, 165-196. Marchig V. and Gundlach H. (1982) Iron-rich metalliferous sediments on the East Pacific Rise: Prototype of undifferentiated metalliferous sediments on divergent plate boundaries. Earth Planet. Sci. Lett. 58, 361-382. Martin J.M. and Meybeck M. (1979) Elemental mass-balance of material carried by major world rivers. Marine Chem. 7, 173-206. Meade R.H., Curtis N., DoVale C.M., Edmond J.M., Nordin C.F. Jr. and Rodrigues F.M. Costa. (1979) Sediment loads in the Amazon River. Nature 278, 161-163. Milliman J.D., Summerhayes C.P. and Barretto H.T. (1975) Oceanography and suspended matters off the Amazon River February - March 1973. Jo SEdo PEtFO]. 45, 189-2060 Morozov N.P. (1979) Trace-element migration - form relationships for rivers, estuaries, seas, and oceans. Geochem. Intl. 16, (4), 161-164. 103 Nelsen T.A. (1981) The application of Q-mode factor analysis to su5pended particulate matter studies: Examples from the New York Bright Apex. Marine Geology 39, 15-31. Nittrouer C.A., DeMaster D.J. and Kowsmann R.O. (1981) The deltaic nature of Amazon shelf sedimentation as revealed by high resolution seismic reflection studies. E05 62, (17), 303. Nittrouer C.A., Sharara M.T. and DeMaster D.J. (1983) Variations of sediment texture on the Amazon Continental Shelf. J. Sed. Pet. 53, (1), 179-191. Ottman R.E. (1968) Reconnaissance investigations of the discharge and water quality of the Amazon River. U.S.G.S. Circ. 552, 16. Parks J.M. (1970) Fortran IV Program for Q-Mode cluster analysis on distance function with printed dendrogram. Computer Contribution 46. State Geological Survey, The University of Kansas. Perkin Elmer Atomic Absorption Spectroscopy Instruction Book (1973). Ryther J.J., Menzel D.H., Corwin N. (1967) Influence of the Amazon River outflow on the ecology of the western Trapical Atlantic 1. hydrography and nutrient chemistry. J. Mar. Res. 25, 69-83. Schmidt C.F. (1972) Amounts of suspended solids and dissolved substances in the middle reaches of the Amazon over the course of one year. Amazoniana 3, (11) 208-223. Schultz D.J. and Turekian K.K. (1965) The investigation of the geographical and vertical distribution of several trace elements in seawater using neutron activation analysis. Geochim. Cosmochim. Acta 29, 259-313. SIioJkovitz E.R. (1976) Flocculation of dissolved organic and inorganic matter during the mixing of river water and seawater. Geochim. Cosmochim. Acta 40, 831-845. Strolkovitz E.R. and Price N.B. (1980) The major element chemistry of suspended matter in the Amazon Estuary. Geochim. Cosmochim. Acta 44, 163-171. Stallard R.F. (1980) Major element geochemistry of the Amazon River system. PhD. Thesis: Hoods Hole Oceanographic Institution and hflassachusetts Inst. of Technology, 1-366. Statistical package for the social sciences (1975) (eds. Nie H.H., Hull (:.H., Jenkins J.G., Steinbrenner K. and Bent D.H.) McGraw-Hill Book Co. 468-508. TeSsier A., Campbell P.G.C. and Bi sson M. (1979) Sequential extraction larocedure for the speculation of particulate trace metals. Anal. Chem. 51, (7). 104 Tessier A., Campbell P.G.C. and Bisson M. (1982) Particulate trace metal speciation in stream sediments and relationships by grain size: Implications for geochemical exploration. J. Chem. Exp. 16, 77-104. Turekian K.K. and Imbric J. (1966) The distribution of trace elements in deep-sea sediments of the Atlantic Ocean. Earth Planet. Sci. Lett. 1, 161-168. Halkley A. and Black I.A. (1934) An examination of the Degthareff method for determining soil organic matter and a proposed modification of the chronic acid titration method. Soil Sci. 27, 29-38. Nindom H.L., Beck K.C. and Smith R. (1971) Transport of trace metals to the Atlantic Ocean by Three Southeastern Rivers. Southeastern Geology 12, (3), 169-181. Yen T.F. and Tang J.I.S. (1977) Chemical aspects of marine sediments. In Chemistry of Marine Sediments (ed. T.F. Yen), 1-38. APPENDICES APPENDIX I CHEMICAL PARTITIONING DATA 99 80 105 APPENDIX 1 Sample Location Processed prior to receiving sample (1 = yes; 2 = no) Bioturbated (1 = yes; 2 = no) Sample sediment type (Type 1-6 see triangular diagram) (Nittrouer 3131., 1983). Sediment accumulate rate based on Pb-210 data (Nittrouer et al., 1983). -_'-—' Depth in cm (low) Depth in cm (high) Total organic carbon Exchangable fraction in ppm Fe-Mn oxide fraction in ppm Organic fraction in ppm Residual fraction in ppm Total metal concentration in ppm (Addition of I - L values) Missing data Missing data Soils taken at Belem 81-85 = Soil horizons All samples sieved in less than .212 mm sieve prior to analysis. BARIUM 106 OOOmOO 00000000 000 00000000 OOOOOOOOOOOOOOOOOOOOOOO O OomN—wgoOONh@0880?"88800000000800mOOOOmOOmOOOOmOOOOOO*8h8 vawmvwmmvmmOw mwmvwmmNmmommemOvmmmmbmbvmmmmbehOmmOwawm thmOvammvwmvmmthOmmthmhmmmNb NmmNmmmemw®VNONmmvhmOmON mmmvrmvam mmmmmwNObvmOmNNmeVVIOOmhmeN wNmmemwmvvhmbm—Omv mww vonOmmmowathmVVOvvwmomehmmmOm thmwvvvthmwvmmwmm 00mmOOOO OOOOOOOOO OOOOOOOOOOO 000000 00 OOOOOOOOOOO §OOth0008mOOOOOOmOBOOOOOOOOOOOSOOmmoogggm0888mOOOCOOOOOO mOmwNwNthObmvvNvOmOmOmhmmwmmwmrFmmemmvmmvaanmwmowvwm mmNmonmethmummvmmm—mv—mmmmmwwNmmvmmNmObmmNomrmowmvmmNm NGOVFGVOMFNGQNO'QFOflWFO'OFOPflommV@ObmommwwfimmOVMMFhwwoovm mwv vamOmmmow—OhmOvvavvomofithmmmmmmwommNvawwavmmvmm hmOmmOONOmmFONNOOmOOh OmNOmOmOFthNOOVOOOmomNmFOmFMOOmMOOO wNvmvawVwNOVwmmvmmvm MVwmnwmvamO®NmmwmmOOmmm 0mm «mmvaN'w mmmN—mw—NQVNmmwmbmmmmvamNvam'mOVvawhmohPmNmmmohmmbmowr N IDVVVONNNGVO '-NNVIDVVQQLDIDDVVVNNNCOh'0LDFVVQV’NVOIDVV"NN 00> OOOOO Omb 0000000005 OONOOm me mmmwm *hv Cachmvono rvamm 0vNFmwammnmnvmvmmmmhwmmv«arev h 00 NNQI‘VGVVOQFNGWQFQQQOSNO’I‘I‘IDCDVDIDO"QIDI‘ONDC’H‘I‘NNCDIDI‘CDCD N10" Pv- NN.-Inv- Nmev—q-NNNB00000000000)OCC-v-PNv-F‘Pv-v-NPFV'IDNLDFIDNID""”IDID0000050N "QFQID MOON!) “(9(9va Qv-[xv-fsv-(p wwthNNmmmvvv “'mme COCOOO0000OOOOO0000OO00OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOmmmmmmmmmmmdmmoOOOOOOOOOOOOOOOOOOOOOOO0000mmvwwO WOmvavaN ON vaOum wwwwNNNm mmvvv vwmmvm NNNNNNNNdddOIdmO) 0505050501030501NNNNOMONMNGGG*P"P*""PNGGONO§ 0503 mmmmmmmmmmmmm O) ADD"?“"IDCDIDV0V?"0010'-VVIDID‘DIDIDIDIDIDIDID‘D‘DIDM"NIDIDVNNVIDIDIDIDVV0 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I v-u-C-v-v-Pv-v-v-v-v-w-Q-v-Pv-Pv-v-Pv-Pv-v-w-PwNNNNNv-C:NNNNNNNNNNNNNNN "*N"PNNNNNNNNNNNNNNNNN**PNFFP'?NF'PNNNFNNNNF*P*NNNNNNNNN NmvmmhmooooooOOOOOOOOOO'vamwhmmommhmmOvavmwhmmNNNNNNNNm ”wwwww wwwvaNNNNNmmnmmmmmmmvvvvvvvvv 107 FVVCDNQVCOONGOI‘F'NN men 0mm.Nvm bvv. 0Nm.N 000.0 000 0mm 050.vmm Ovm w mhw.N 000.0 va F th 00m.va hmv. 00N.N 000.0 0mm. N.0 00m.vpm Own. VNv.N 000.0 Ovm 00N 000.0m. 00v. OmN.N 000.0 Ohm N mm 0m>.mm 000.0 OhN. 000.0 00*. m. me.m. 0N0. va. 000.0 00. .ON OmN.m— bvo N.0. 000.0 Omw. .Ov www.mm hvo. bv0.w 000.0 0m0 mm mN®.mm 00v. mvn.p 000.0 00m. m 000.«@ hwo. 0.... 000.0 0.5 m Omh.0m th. mmN.p 000.0 00m. mm 0m0.vhm mv0.v 000.0 000.0 000.0 mm. 00m.waw mmN.v Om—.v 000.0 00.. m0. 00m.wm0« 0mv.v Ohm.o 000.0 00N. F—v OOm.mm0P Omp.m 00N.m 000.0 OmN mm. 00m.hvmp 000.m hwm.w 000.0 0®N 00. 00w.mv0P mmh.v 0mm.m 000.0 OVN. Om. 00m.mme mhv.m Ohm.m 000.0 0m— mm. 00m.mmmp 00¢.v 00N.m 000.0 0m. hm 00m.nwm mhv.v O—m.m 000.0 00N. vww 00¢.vmwp 0mm.w 0mm.h 000.0 OmO. me, OOm.vav Own.m me.m 000.0 OmN. mp. 00..0v—w mwv.v mm¢.h 000.0 Omp. mew 000.mwvw mow.m 00m.m 000.0 OhN. vvr 00v.wmv. mvv.¢ ONm.m 000.0 OmN. mm OmN.vmm 00>.v 0.0.m 000.0 000.0 I 4 x a _ I A.u.+coov z:_mm{0N0vooovva00055w 00v v00?00NN0NNP005v00N00w0v0v000005v0000 ' ' “4.4.: ‘..'..' '..'..’..'..'..' ‘ '..'..'..'..' ‘..'..'..’..‘..'..'..'.." ' '..‘ ' ' '..'..'..'..' .42.}. .. «V0005vN00555005050050000050P0N00000050000005000V* 00500 wNv500000Nmmm50v0wv0wvw00v005NmeNNN05—v000Nme 0w 0NN0N 0000v5vv00¢5500N059000*000N000500N0w505w00000v55 00000N0v N v-v-v-Nv- N v-v-N '- v-va-v-v-Q v-v- NmmmNNv-QQv-v-Nv-v-Nv-v-v-v-QQva-m 000000000000 00 000 00000 "0" 0 0 8800 008880000000000 0000808000000000 00 00000000 0000 0 00 00 MODNOSOOOO v000 0 000800800 00 000000Nm V 00 000000000 0 0 0 00000 00 00 0 00 0 :aPtNNN0003VQ :w00v0 000000000000000000000000000000000000000000000000000000000 00000000000000000000000000000000000000000000000000000—v00 v00 “0 00NO0N0 v00 000P0 www—NNN000QVQ wv 0 0 0 36 3 4 5 Lu NNNNNNNNdO’QO’O’O’O’O’éO; 0'; 00005000000300 8gNNNNNWONONGDMFP“'PFP'DMC‘DOC") Q IDID‘P'V'PIDCDCDVOVV'WD‘DID'VVID‘DCOIDIDIDIDIDIDID0000*NIDID'NN'MIDMIDVV0 99.99 0 I I I I I I O I I I I I I I I I I I I I m N'N'P"NNNNNNNNNNNNNNNNN'F"N""F"VNV'V'"NNN'NNNNV'V‘P"NNNNNNNNN < N0v005000000000000000000N0€005000005000?N0v08 gmNNNNNNNN0 rs FF"""""""""NNNNNNNONNNN P) NVVVVVVVVV '0 113 OrnmnmvN0000000 m0v50000NN0N0v0 wwNN5PNw500—000 0000000000vv0vv 005mmm5mw000 000vv00vv09v oowmwNv00V5000N ééééééédéééééééddddddddd' wév'éé'édbdéé04000000000 x H 0 0 0 00000000 0 O 0 0000 0 0 0 0000000 A.0.+couv 54(000 050$ 00 CIDOIDCLDOIDO"""NNNVF"IDCOF030505050 "V'OICDO’O" FPNNO’O‘O‘O’O” I I I I I k 0000000000000000N¢5v00000 NNNNNNNNNN""FNNNNNNNNNNNNNN *“NNMNVVID "W'NNI‘OOVV 0000000000000000?0'00000000 0000000000000000090?0000000 (DID MANGANESE 114 0000000000000000000000000000 00000000000000 0'550 00005w0 0—0 N0000000Q000N0 5v000000w005w0005N00000v0N5000va000FNN00 N50N 0—N00000P00000000000N0—N000—0000000NN—00V50N00N 00000000v0v0w000 50w5v0w0N0v00000vw0005500N5wwN00wN0 v00v00005500v0000v0v0050550v0v00N050N055Nv00000v000 '- v-v- 9O 93 347 183 OW) 730 800 000550050005505500505555550505005050050000000000000050505 0Nw00v#0500N00N00000v0v000v0P00v0v000v05000005v000000v0N0 0*05N00v0VN55v00v50v05500000v000vV000Nv0000VNv0000000v00N we - 50v00v00v0N00 000QNw—000v5v00000v0vN00Vw0v0505500vm00 0 N00N¢0N00000 ONGONN'VMDNOMQ'P *0 ww vav P00000r 00 O50500008800000000000000080080080000088008008000000800000 50000050 505505005505000 50 50 55000 00 00 055005 05500 N00N5000000v000vavN0050NNv0005N0vN50NNQN90NN00Q000VN000? 0500000N500V00000Nv0000'00—w00N0000005N8N0050wv 00*w0000w 00vava00N0N0v005000N0N0N0000NN000N0N¢ N5Pv00v00N5000P-P 0 N000v000vvN0N0000NNN000N000wwwwv NPwN0v0wNv0000Fv00 0005505005005000 050055 80000000000 000005000000 000050000 000vv00w0—000000 0N0000 050050N000 000000500000 50000w080 0000N00N00N0V50550*?000w000Nw0N550v 5000000000000N0005v 0 N0NwNNv5000vv0000~0Fv50P00000wP0V0000N05000900N000050NN50 w *0000v0009000N000w0005awe000N¢wgws vvav0000 wv v-v-u- .- O-v- v- gv-v-Nv-v-Nv- v-Ov-v: N8 v-v- 0 %0 000000000008808 00008 08 00 0 08 0 08 W8 0008 000000000 0 0 0 00000 00 00 0 00 0 000000000000000000000000000000000000000000000000000000000 N“0*PPNNNF0000000000000"*NP***'NF?*mNmFmNm?'P'*mg02020$N ”V Pv-v- wrrvNNN000vvv '“Wme 000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000OOOOOOOOOOOOOOOOOOOQQQ?P00 v00v0*v00N0 0N vowowwm wwvaN 000vvv '“mme NNNNNNNNGO‘O’O’U’OO’O’ 0050301050')O‘NNNNOMONONMGC‘O'PPP""*"GOOGMQO‘DO§CD 000000000000000 05050505 050'! UND""""ID(D(DV‘DVQ'-(D‘DIDV'VVW‘DCDIDIDIDIDIDIDIDCD‘D‘DN'PNIDII'D'NN‘PIDIDIDIDVVtD 0 q-v-v-v-v-v-v-v-v-v-q-v-v-v-v-v-v-v-v-v-v-c-Pv-O-v-v-NNNNNV-v-NNNNNNNNNNNNNNNQQQmQQQCV. 0010 000 NvNPw—NNNNNNNNNNNNNNNNvaFNPPvP*N*VPNNN*NNNN'*"NNNNNNNNN N0v00500000000000000000wN0v005000005000wN0v00500NNNNNNNN0 wwvwwwrrwwwvwwwwvwaw—wwNNNNNN0000000000¢¢vvvvvvv 115 005.00 NN5.00 000. 0v0. 00v.. 000. 0.00 0.00 00V.N0 005.0v 0N0. 0N0. 00N.N 0v... 0.00 0.00 .00.0v 000.0v 000. 0.0. 00v.. 000 0.00 0.00 000.Nv 005.00 00.. 0N0. 000.0 0V0. 0.00 0.00 000.N0 000.5v 00.. 5.0. 00N.0 050.N 0.00 0.00 555.0v 00N.0v 00N.. 000.0 5N0.N 00.. 0..0 0.0N 500.00 000.00 500. 000.0 050.N 00.. 0.0N 0.0. 00..00 000.N0 00v. 000.0 0V0. 00.. 0.0. .... 0.v.0N 000.0N 050. 000.0 500. 000. .... 0.5 00N.vv 000.0v 00.. .00. 0v0.0 000. 0.5 v.v 000.00 v00.50 #00. N00. 005.. 0.5. v.v 0.N 00N.00 500.v0 000. 000.0 0.0.0 000. 0.N 0.0 000.005 050.v0. 000.00 000 N0v 0.0.N. 000.0 0.N 0.0 000.000. 050 V5. 5v0.00 05N..N5 550.0. 00.. 0.N 0.0 000.000 000 00. 000.0v 050.000 0.0.50N CON. 0.. 0.0 005.000 000.00. 000 50 005 .N0 00v.00. 00N. 0.. 0.0 05N.050 000..0. 00N 00 00..00v 000.0v. 00N. 0.. 0.0 05N.000 005.00. 500.05 000.N00 00N.00 OVN. 0.00 0.0? 00..000 050 55. 5N0.00 050.000 500.00 00.. 0.0V 0.0V 0...000 000 N0. 0Nv.00 000.0.0 500.0 00.. 0.0? 0.00 000.N.5 000 00N 000.00 00v.000 000.00 00N. 0.00 0.00 0Q5.N00 000 00N 0N¢.55 00¢.0N0 000. .Eso mucmmgm> eemmu. mmmou. mHooo. mmomw. nmwow. ommmm. vomqo. mammw. wmnmm. mmofim. mNNmN. mmumm. cameo. ommex. wmavm. Namem. ~¢¢wm. onoo. Hmoam. macaw. auwpwczssoo quwcumu orgacosoguxg .wHPLHmn o_sacoeocnxs Fmp_cumv owsngosocu»; Pmpwcumu uwcacoaogca; —muwcumu owzagoEOLuz; quwgumu uwgncosocux; .mpwcumu u_;acoeocuxc —muwcumu owsncoaogu»; qu_cpmc owzacosocux; pmuwcuwc u_;acoeoguxg F< —< Lo so an an *2 wz :N :N c: :z oo 00 mm mu :0 no mm mm a_am_cm> 134 a.m~ a.m mamm~.n mmmaH.n makma. ammaa.u msaau. mNNmH.u mammfi. ma~¢a.: aaaaa.u mamma. amaafi. mHaaa.n mNNHa. amamm. mmama. aNmHa. mvmam. ameaa. mamaa.n HmmNH.u mxama. m Louumm a.mm ~.a camaH.u aHN¢H.n mmmma.u momma. mamaa.n «mama. mamHa.u aem-.: HNHNa.u maNmH. momma. ammam. amxma.u enmma. mmmHH.u mNHmN. mavmm. mvkaa.n comma. mmama.u aHmNa.u v Louomm mpcwwuwwmwoa mcoum Lagoon a.Hm m.~ muaaa. mmmaa.u Hmmvm.n mvmaa.u amaaa. memmm.n meefia.u maaaH.u Nuama.u mamma.u aHmHH. mmNNH. mmmam. ammvm. mam~a.u mvmaa. maaNH.u momma. mmaav. Hmeaa. mamma.n m couomm ~.vm m.a mHmNH. mamma. memmH. Numma. cumma. mumma. maamH. mamma. ammma. Hmmva. nevma. amama. Nfieafl. cumma. muamfi. mmama. momma. cammm. HmHaa. mmmaa. momma. N copomm ¢.¢¢ ¢.¢¢ Nmama. amNaH. mmaaH. Hmmma. muuaa. cmema.u mNHmH. NmmmH. asnma. gamma. aeama. aaaaa.n mmeafi. meaea. Haawfi. msmea. Nnmma. aHmaa.n mmmma.: ammma. aNmNa.u H Louumm mwmx_mc< coaomm wcoznm 3mm wocmwcm> .E=a wuccwgm> Hmuam. Nmmmm. mammm. cammm. nefimm. mammm. mmmfia. mmmam. HaNam. NaNHa. mHNmm. mmvfim. memca. mmamm. mmmma. ammea. mflmmm. aHmHN. camam. aeaam. Hafias. auw_mczesoa —muwcumv u_;acoeocnxs pmuwcumu uwsngoeocnx; quwgumu uwzacoeocu»; Pmuwcumu owzaLoEogua; qu_gumu uwgacoeocux; Pmuwcamu owsqcosocu»; Pmpwgpmu uwgqcosocu»; ~opwcumu u_;agoeocvas quwcumu opsacoeogum; Pmuwgumu u_:qgoeogux; p< _< La La an an wz *2 :N :N :2 :2 co co mm mm :a so mm mm :oocmo owcmaco _mpop «Eat: RIES H ”7111117111111 )Lllflllliujllfllfllllfll I)