the biogeochemistry of manganese and iron reduction in ... · the biogeochemistry of manganese and...

36
Earth-Science Reuiews, 35 (1993) 249-284 249 Elsevier Science Publishers B.V., Amsterdam The biogeochemistry of manganese and iron reduction in marine sediments David J. Burdige Department of Oceanography, Old Dominion University, Norfolk, VA 23529, USA (Received March 3, 1993; revised and accepted June 16, 1993) ABSTRACT Manganese and iron reduction in marine sediments are known to play important roles in the biogeochemical cycles of many elements, including carbon, sulfur, phosphorus and several trace elements. These reduction reactions affect these cycles on a variety of time scales, ranging from those as short as seasonal time scales (e.g., nutrient cycling in coastal ecosystems), to those as long as thousands to tens of thousands of years (e.g., glacial-interglacial transformations in deep sea sediments). In this review article I will briefly summarize the results of laboratory studies on the types of manganese and iron reduction that are known to occur in marine sediments, and then discuss the occurrence of these processes in different sedimentary environments. Particular efforts will be given to examining the rates and mechanisms of sedimentary manganese and iron reduction in relationship to other biogeochemical processes in sediments. INTRODUCTION chemical aspects of reductive dissolution re- actions (Stone and Morgan, 1987; Stumm Manganese and iron reduction in marine and Sulzberger, 1992); microbial manganese sediments are known to play important roles and iron reduction (Ehrlich, 1987; Lovley, in the biogeochemical cycles of many ele- 1987, 1991; Ghiorse, ]988; Nealson and My- ments, including carbon, sulfur, phosphorus ers, 1992); and manganese sediment geo- and several trace elements. For this reason chemistry (Glasby, 1984; Schimmield and considerable effort has been devoted to ex- Pedersen, 1990). Rather than simply dupli- amining the rates and mechanisms of these cating or up-dating these articles, in this re- reduction reactions in sediments. In this re- view article I will incorporate information view article I will begin by first briefly sum- from all of these different fields into a more marizing the results of laboratory studies on unified examination of the current knowl- the types of manganese and iron reduction edge on manganese and iron reduction in that are known to occur in marine sediments, sediments. This will then be followed by a discussion of Manganese and iron reduction also occurs these processes in different sedimentary en- in stratified marine water columns such as vironments. The attempt here will be to ex- the Saanich Inlet and the Black Sea (Emer- amine the rates and mechanisms of sedimen- son et al., 1982), in the low oxygen waters of tary manganese and iron reduction in rela- some open ocean environments (Landing and tionship to other biogeochemical processes. Bruland, 1987) and in the water columns of In recent years there have been several other many stratified lakes (Davison, 1985). A1- review articles that have examined topics re- though there are many similarities between lated to the subject of this review, including: sedimentary manganese and iron reduction 0012-8252/93/$06.00 © 1993 - Elsevier Science Publishers B.V. All rights reserved

Upload: hangoc

Post on 04-Jun-2018

220 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

Earth-Science Reuiews, 35 (1993) 249-284 249

Elsevier Science Publishers B.V., Amste rdam

The biogeochemistry of manganese and iron reduction in marine sediments

David J. Burdige Department of Oceanography, Old Dominion University, Norfolk, VA 23529, USA

(Received March 3, 1993; revised and accepted June 16, 1993)

ABSTRACT

Manganese and iron reduction in marine sediments are known to play important roles in the biogeochemical cycles of many elements, including carbon, sulfur, phosphorus and several trace elements. These reduction reactions affect these cycles on a variety of time scales, ranging from those as short as seasonal time scales (e.g., nutrient cycling in coastal ecosystems), to those as long as thousands to tens of thousands of years (e.g., glacial-interglacial transformations in deep sea sediments). In this review article I will briefly summarize the results of laboratory studies on the types of manganese and iron reduction that are known to occur in marine sediments, and then discuss the occurrence of these processes in different sedimentary environments. Particular efforts will be given to examining the rates and mechanisms of sedimentary manganese and iron reduction in relationship to other biogeochemical processes in sediments.

INTRODUCTION chemical aspects of reductive dissolution re- actions (Stone and Morgan, 1987; Stumm

Manganese and iron reduction in marine and Sulzberger, 1992); microbial manganese sediments are known to play important roles and iron reduction (Ehrlich, 1987; Lovley, in the biogeochemical cycles of many ele- 1987, 1991; Ghiorse, ]988; Nealson and My- ments, including carbon, sulfur, phosphorus ers, 1992); and manganese sediment geo- and several trace elements. For this reason chemistry (Glasby, 1984; Schimmield and considerable effort has been devoted to ex- Pedersen, 1990). Rather than simply dupli- amining the rates and mechanisms of these cating or up-dating these articles, in this re- reduction reactions in sediments. In this re- view article I will incorporate information view article I will begin by first briefly sum- from all of these different fields into a more marizing the results of laboratory studies on unified examination of the current knowl- the types of manganese and iron reduction edge on manganese and iron reduction in that are known to occur in marine sediments, sediments. This will then be followed by a discussion of Manganese and iron reduction also occurs these processes in different sedimentary en- in stratified marine water columns such as vironments. The at tempt here will be to ex- the Saanich Inlet and the Black Sea (Emer- amine the rates and mechanisms of sedimen- son et al., 1982), in the low oxygen waters of tary manganese and iron reduction in rela- some open ocean environments (Landing and tionship to other biogeochemical processes. Bruland, 1987) and in the water columns of In recent years there have been several other many stratified lakes (Davison, 1985). A1- review articles that have examined topics re- though there are many similarities between lated to the subject of this review, including: sedimentary manganese and iron reduction

0012-8252/93/$06.00 © 1993 - Elsevier Science Publishers B.V. All rights reserved

Page 2: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

250 ~>J. BURDIGE

and that which occurs in water column envi- nese and iron reduction. While there are ronments, this article will only discuss these natural settings in which manganese and iron processes in sediments, reduction occur in the presence in oxygen

(Sunda et al., 1983; Hong and Kester, 1986) it is likely that the vast majority of man-

THE CHEMISTRY AND MICROBIOLOGY OF MANGANESE AND IRON REDUCTION ganese and iron reduction in sediments oc-

curs under anoxic (or sub-oxic) conditions. While the absence of oxygen generally ap-

Definitions of iron and manganese reduction pears to be required for the occurrence of sedimentary manganese and iron reduction,

Under conditions of low Eh and pH, the the factors ultimately controlling these pro- reduced forms of manganese and iron (gen- cesses are more complex. Understanding the erally Mn 2+ and Fe 2+) are thermodynami- factors controlling sedimentary manganese cally favored over their oxidized forms (gen- and iron reduction often involves determin- erally Mn 4+ and Fe3+), which are generally ing: (1) the reductants involved in the reac- found in nature as solid oxides and oxyhy- tions; (2 ) the processes by which these reduc- droxides (Stumm and Morgan, 1981). Man- tants are produced; and (3) the direct or ganese reduction will therefore be defined as indirect role of biota in mediating either the the reduction of manganese oxides or oxyhy- reduction of the oxides or the production of droxides to Mn z+. Similarly, iron reduction the chemical reductants. In general, man- will be defined as the reduction of oxidized ganese and iron reduction in sediments may iron minerals to Fe 2+. be a strictly chemical (abiotic) process, or

Given the numerous problems in making may be mediated by biota (primarily bacte- and interpreting Eh measurements in natural ria). Furthermore, there appear to be several environments such as sediments (Stumm and different types of microbially-mediated metal Morgan, 1981; Lindberg and Runnells, 1984), oxide reduction. While many studies have the p r e s e n c e / a b s e n c e of oxygen is often indicated the ways that manganese and iron taken as the "mas te r" variable defining the reduction may occur in sediments, there have environmental conditions favoring manga- been far fewer studies that have definitively

at ODU. Dr. Burdige received his B.A. in Chem- istry (with Honors) from Swarthmore College and his Ph.D. in Oceanography from the Scripps Institution of Oceanography, University of Cali- fornia, San Diego. Prior to joining the faculty at Old Dominion University he was a Post-Doctoral Research Associate in the Marine Sciences Pro- gram at the University of North Carolina at Chapel Hill. In addition to his interests in man- ganese and iron redox cycling, Dr. Burdige is also

Dr. Burdige is an Associate Professor in the interested in carbon and nitrogen diagenesis in Department of Oceanography at Old Dominion sediments, cycling of dissolved organic matter in University. He also holds a joint appointment in the oceans and mathematical modeling of biogeo- the Department of Chemistry and Biochemistry chemical processes.

Page 3: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 251

demonstrated the exact mechanism(s) by Burns and Burns (1979, 1981) suggest that which these processes actually occur in a the predominant manganese oxides found in given sediment, sediments include: todorokite, (Ca,Na,K)

(Mg, Mn2+)MnsO12.xH20; vernadite (also The chemistry of oxidized and reduced man- known as 6-MNO2), M n O 2 . n H 2 0 . m ( R 2 0 , ganese and iron in sediments RO, R203), where R = Na,Ca,Co,Fe,Mn; and

birnessite, (Na,Ca,K)(Mg,Mn 2 +)Mn6014 In the simplest sense, manganese oxides 5H20. However the results of laboratory

(or oxyhdroxides) in sediments are often re- studies of manganese oxidation (Hem and ferred to as MnO x, although other cations Lind, 1983; Murray et al., 1985) also suggest are also incorporated into the crystal lattices that other manganese oxides (e.g., hausman- of these materials. MnO x is a non-stoichio- nite, Mn304, or feitknechtite, /3-MnOOH) metric formula since x is greater than one should exist in sediments. Burns and Burns and generally less than two. This therefore (1979, 1981) and Schimmield and Pedersen implies that the "average" oxidation state of (1990) further discuss the implications of manganese in these materials is between two these laboratory studies in terms of man- and four. Whether this is caused solely by ganese oxide mineralogy in sediments. mixtures of Mn 2+ and Mn 4÷ in the solid, or The end-product of manganese reduction also involves the occurrence of Mn 3÷ has is generally been thought to be Mn 2+, al- been the subject of some debate. Mn 3+ is though soluble Mn 3÷ may also be produced generally thought to be thermodynamically (Ehrlich, 1987). Evidence from the chemical unstable relative to disproportionation to literature suggests that Mn3+-organic com- Mn 2+ and Mn 4÷ (Morgan, 1967), although plexes should be stable or metastable under studies have shown that Mn 3+ is indeed a range of anoxic and oxic conditions (e.g., found in certain manganese oxides (Glau- see a recent review in Luther et al., 1993), singer et al., 1979; Murray et al., 1985). and the possible occurence of dissolved

The mineralogy of manganese oxides in Mn3+-complexes in anoxic waters of the sediments can also vary. These oxides are mid-Chesapeake Bay has recently been dis- generally thought to be amorphous materi- cussed (Luther et al., 1993). However, als, and are often found as coatings on inor- whether such Mn3+-complexes occur in sedi- ganic (e.g., clays) or biogenic (e.g., siliceous ment pore waters and if so, whether they are tests) sedimentary particles. They are also produced by oxidative or reductive processes usually found in close association with sedi- are both questions that will clearly require mentary iron oxides. Attempts to identify further study. The Mn 2+ produced during sedimentary manganese oxides is difficult, in sedimentary manganese reduction is gener- part because the small crystal size and lack ally found dissolved in pore waters as either of long-range ordering within the crystal a "free" ion or more likely an organically- structure often precludes analysis by conven- complexed ion (Elderfield, 1981). However tional X-ray diffraction techniques. Studies in anoxic sediments where manganese con- using techniques such as transmission and centrations and carbonate alkalinity reach scanning electron microscopy, electron dif- sufficiently high levels, some of this divalent fraction and X-ray photoelectron spec- manganese may precipitate out of the pore troscopy have, however, proven useful in fur- waters as a pure or mixed carbonate phase ther describing the crystal chemistry and (Holdren et al., 1975; Grill, 1978; Suess, 1979; mineralogy of sedimentary manganese and Aller, 1980; Pedersen and Price, 1982; iron oxides (Burns and Burns, 1979, 1981; Thomson et al., 1986; Middelburg et al., Glausinger et al., 1979; Murray, 1979; Mur- 1987). Schimmield and Pedersen (1990) fur- ray et al., 1985). ther discuss the factors controlling man-

Page 4: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

252 D,J. BURDIGE

ganese solubility under these conditions. Millward and Moore, 1982; Balistrieri and Iron oxide mineral phases that have been Murray, 1986). Iron and manganese redox

identified in sediments include goethite (c~- cycling in sediments then also leads to associ- FeOOH), ferrihydrite (5 Fe20 3. 9 H20), ated diagenetic cycles of these other ele- akageneite (/3-FeOOH), and lepidocrocite ments (e.g., Jenne, 1968; Turekian, 1977), (y-FeOOH) (Burns and Burns, 1977, 1981; since the reduction/dissolution and oxida- Murray, 1979). As with the identification of t ion/precipitation of these metal oxides also manganese oxides in sediments, similar prob- causes the cycling of these transition ele- lems also exist in the identification of sedi- ments and heavy metals between sedimen- mentary iron oxides. Another iron oxide rain- tary pore waters and solid phases (see Coastal eral found in both marine and freshwater and estuarine sediments for further details).

sediments is magnetite (Fe304), a magnetic Chemical (abiotic)manganese and iron reduc- phase containing both Fe 2+ and Fe 3+. In tion sediments, magnetite is often a detrital com- ponent derived from the weathering of conti- A wide range of organic and inorganic nental rocks or submarine (hydrothermal) compounds are able to chemically reduce vulcanism (Murray, 1979). However authi- manganese and iron oxides. These possible genic magnetite can also be produced both in reductants include: sulfide (Kessick and the presence and absence of oxygen (Blake- Thomson, 1974; Pyzik and Sommer, 1981; more, 1975; Bell et al., 1987; Karlin et al., Burdige and Nealson, 1986); nitrite (Bartlett, 1987; Bazylinski, 1990; Lovley, 1990), and 1981); Fe 2+ (for manganese oxides)(Postma, under some conditions anaerobic magnetite 1985; Myers and Nealson, 1988b; Burdige et production may be the result of microbial al., 1992); organic acids such as pyruvate and iron reduction (Lovley, 1990). At the same oxalate (Stone, 1987a) and certain aromatic time, magnetite can itself undergo reductive compounds (Stone, 1987b; LaKind and Stone, dissolution in the presence of sulfide (Can- 1989; Stone and Ulrich, 1989). The reduction field and Berner, 1987). of iron oxides by sulfide has important impli-

The more common end-product of iron cations in the formation of sedimentary pyrite reduction is, however, Fe z+ . In contrast to and plays an important role in the global Mn 2÷ though, ferrous iron is far less soluble sulfur cycle (e.g., Berner, 1982, 1984; Garrels and forms insoluble phases such as: siderite, and Lerman, 1984). However, this subject is FeCO3; pyrite, FeS2; iron monosulfides, FeS; beyond the scope of this review, and the and vivianite Fe3(PO4)2"8H20 (Emerson reader is therefore urged to examine the and Widmer, 1978; Martens et al., 1978; references cited here for further details on Suess, 1979; Berner, 1980, 1984; Postma, this topic. 1982). These differences in the solubility of While the compounds discussed above may reduced iron and manganese can, in some chemically reduce manganese and iron ox- circumstances, lead to significant differences ides, their occurrence in natural settings is in the diagenetic cycling of these elements often a result of biological activity. As such, across redox boundaries (Burdige and Neal- understanding the importance of these abi- son, 1986). otic reduction processes in sediments also

Finally, it should be noted that both man- involves determining the mechanisms by ganese and iron oxides are known to take up which these reductants are produced, and a variety of transition elements and heavy the role biota play in this production.

metals (e.g., As, Cu, Ni, Zn, Co, Mo and Pb), Microbial manganese and iron reduction either by adsorption to the oxide surface or by direct incorporation into the crystal struc- The processes by which microorganisms ture (Murray, 1975; Burns and Burns, 1981; couple the reduction of manganese or iron

Page 5: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 253

oxides to the oxidation of organic corn- otic manganese reduction coupled to sulfide pounds (or H e) are often referred to as dis- oxidation, in that elemental sulfur (S °) ap- similatory (or heterotrophic) iron or man- pears to be the predominant end-product of ganese reduction. There have been several the abiotic reaction (Burdige and Nealson, recent review articles on this subject (Ehrlich, 1986). This type of microbially-mediated 1987; Lovley, 1987, 1991; Ghiorse, 1988; manganese reduction is apparently catalyzed Nealson and Myers, 1992), and these pro- by chemolithotrophic bacteria (i.e., organ- cesses will therefore only be briefly summa- isms that use the energy of this reaction for rized here. autotrophic CO 2 fixation, or chemosynthe-

Dissimilatory manganese and iron reduc- sis), although the organisms responsible for tion appear to be strictly anaerobic pro- this process have not yet been identified. cesses, although many manganese and iron Furthermore, iron oxides are apparently not reducing organisms are facultative anaerobes used by these organisms in an analogous (i.e., in the presence of oxygen many use 0 2 reaction with sulfide (Aller and Rude, 1988). as an electron acceptor). The manganese and This process may be important in certain iron reducing bacteria that have been exam- sediments where physical or biological mix- ined in pure culture generally use only a ing of the sediments brings solid phase sul- limited number of relatively simple com- tides into contact with manganese oxides (AI- pounds as their energy source (e.g., hydro- ler and Rude, 1988; King, 1990; Canfield et gen, lactate, acetate or formate), although an al., 1993), and will be discussed below in iron reducing organism has also been ob- greater detail. served to oxidize a wide range of aromatic The types of microbial manganese and iron organic compounds (Lovley, 1991). Microbial reduction described above may be consid- manganese and iron reduction coupled to ered "direct" microbial reduction, in that the organic matter oxidation also appears to re- organisms appear to directly (and presum- quire direct contact of the bacteria with the ably enzymatically) catalyze the reduction oxide surface (Tugel et al., 1986; Arnold et processes. There are, however, other types of al., 1988; Ghiorse, 1988; Myers and Nealson, "indirect" microbial manganese and iron re- 1988a). Finally, lab studies have shown that duction in which the reaction occurs as a the rates of microbial manganese and iron result of perhaps localized acid production, reduction are a function of the mineralogy or or production of reduced compounds (e.g., specific surface area of the oxides undergo- sulfide, Fe 2+ and certain organic acids)which ing reductive dissolution (Lovley and Phillips, then react with the oxides. Under these con- 1987b, 1988; Burdige et al., 1992), with amor- ditions for example, sulfate reducing bacteria phous oxides generally showing greater reac- could then be considered indirect manganese tivity than more highly crystalline forms, and iron reducing bacteria. However, the results of Burdige et al. (1992) suggest that differences in the rates of micro- GENERAL ASPECTS OF SEDIMENTARY MAN- bial manganese reduction with different ox- GANESE AND IRON REDUCTION ides are also a function of the absolute rate of the process itself. Early studies of manganese redox cycling

Recent studies have also suggested the in sediments led to the observation that man- existence of another type of direct microbial ganese is reduced from the +4 to the +2 manganese reduction, in which microorgan- oxidation states when sedimentary conditions isms appear to be able to couple the reduc- change from "oxidizing" to "reducing" with tion of manganese oxides to the oxidation of depth (Murray and Irvine, 1895; Wangersky, sulfide to sulfate (Aller and Rude, 1988). 1962; Lynn and Bonatti, 1965; Li et al., 1969; This process is distinctly different from abi- Bonatti et al., 1971). The manganese oxides

Page 6: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

254 D.J. BURDIGE

involved in this process ultimately have ex- Pore Water Mn (pM) ternal (e.g., river or aeolian) sources, al-

0 lO 20 though internal cycling of the oxides as a result of these redox changes also plays a o ~ - ~- . . . . ~

role here (see below). The depth of the manganese "redox" ol id phase Mn

boundary in sediments is generally assumed 10 to occur where oxygen concentrations in the pore waters go to near-zero levels, based in ~" part on the examination of pore water man- o . . . . ganese and oxygen profiles (Shaw et al., 1990; ~ 2 0 " " . . . . . - ,

Reimers et al., 1992). Oxygen is therefore ~ 2 assumed to be the primary sedimentary oxi- ~ '\ :

dant of pore water manganese (Aller, 1990; ',, Reimers et al., 1992). The Fe2+ /Fe 3+ redox 3o ~ boundary generally occurs below the man- Pore water Mn ,'

ganese redox boundary in many sediments (see below), and also often results in a brown 40 ~------~-----~ - ~ -

to green color transition in non-sulfidic (e.g., 0 4 8 12 pelagic) sediments (Lyle, 1983). This color change is reversible, and is apparently associ- Solid Phase Mn (mg/g) ated with the oxidation and reduction of iron Fig. 1. Idealized steady-state pore water and solid in smectite clays, rather than the reductive phase manganese profiles resulting from the sedimen-

tary redox cycling of manganese. The curves shown dissolution of sedimentary iron oxides. Fi- here were determined with the model equations pre- nally, at least in pelagic sediments, pore wa- sented in Burdige and Gieskes (1983). As discussed in ter iron profiles suggest that c o m p o u n d s the text, the depth of manganese redox boundary is other than 0 2 (i.e., manganese oxides and found just below the depth of the peak in solid phase

nitrate) are likely more important oxidants of manganese (i.e., 13 cm). Fe 2+ during sedimentary iron redox cycling (see below), state conditions and in the absence of signifi-

With these redox changes in the sedi- cant bioturbation (Gratton et al., 1990; ments, the reduction of manganese and iron Rabouille and Gaillard, 1991) to a well-de- oxides leads to the production of their re- veloped peak in solid phase manganese just duced, and generally more soluble, forms above the redox boundary (see Fig. 1). This (see Fig. 1). As discussed in the previous redox boundary moves upwards at a rate section, the dissolved Mn 2+ and Fe 2+ pro- equal to the sedimentation rate, and there- duced by these reactions may precipitate out fore relative to the sediment-water interface of the sediment pore waters as sulfide, car- the redox boundary and this manganese peak bonate or phosphate phases. However, these remain at a fixed depth in the sediments. reduction reactions generally also lead to In addition to papers which have qualita- pore water gradients across the redox bound- tively described this redox cycling, several ary of the reduced forms of these metals, steady-state models have quantitatively ex- The resulting upward diffusion of these ions amined manganese reduction and redox cy- then leads to their re-oxidation and the for- cling in sediments (Michard, 1971; Holdren mation of new oxides. Subsequent burial and et al., 1975; Robbins and Callender, 1975; reduction of these authigenic oxides contin- Burdige and Gieskes, 1983; Aller, 1990; ues this redox cycle. For manganese, this Gratton et al., 1990; Rabouille and Gaillard, internal redox cycling leads, under steady- 1991). These models have examined this pro-

Page 7: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCI'ION IN MARINE SEDIMENTS 255

cess both in terms of the sediment man- and iron reduction in marine sediments, at ganese redox cycle discussed above, and in the end of the article a brief discussion is terms of the role of manganese reduction in also presented for comparative purposes on sediment organic matter remineralization the occurence of these processes in freshwa- (see below). Similar models for iron reduc- ter sediments. tion and sedimentary iron redox cycling have not been presented, although Bender and Heggie (1984) used a variation of the Bur- MANGANESE AND IRON REDUCTION IN

PELAGIC SEDIMENTS dige and Gieskes (1983) manganese model to estimate iron reduction rates in pelagic sedi- ments (see below). Broadly speaking, manganese and iron re-

The remainder of this article will discuss duction in pelagic sediments has been exam- manganese and iron reduction in pelagic, ined in the context of: (1) their role in sedi- continental margin, estuarine, coastal and mentary organic matter oxidation (e.g., Froe- salt marsh sediments. The division of the lich et al., 1979; Bender and Heggie, 1984); following sections is based on this classifica- (2) their role in the formation of manganese tion of sedimentary environments, and was nodules (e.g., Moore, 1981; Dymond et al., used primarily for organizational purposes. It 1984); (3) their role in the formation of mul- should be noted that in some cases these tiple, manganese rich layers in pelagic sedi- divisions are somewhat arbitrary, and that merits (e.g., Berger et al., 1983; Finney et al., there is often a continuum among these sedi- 1988) and (4) their occurrence during the mentary environments. Finally, although the diagenesis of deep-sea turbidites (e.g., Colley primary focus of this article is manganese et al., 1984; Wilson et al., 1985).

ELECTRON CONCENTRATION PROCESSES ACCEPTOR AG°(kJ/mol)

. ' ' ' O 2 Aerobic Respiration 02 -3100

' NO~,// Denitrification NO~ -3030

/ Manganese Mn(IV) -3090 to -2920 Reduction (e.g., MnO2)

SO 2-

uJ

+ Iron Reduction Fe(lll) -1410 to -1330 (e.g., FeOOH)

/ Sulfate Reduction SO42- -380

Fig. 2. Hypothetical pore water profiles predicted by the successive utilization of inorganic compounds as terminal electron acceptors in the decomposition of sedimentary organic matter (modified from Froelich et al., 1979). The ranges in AG ° values for manganese and iron reduction are based on the choice of minerals phases presumed to be undergoing reductivc dissolution.

Page 8: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

256 D.J. 13URDIGE

The role o f manganese and iron reduction in sidered to be less efficient than denitrifica- the oxidation o f sedimentary organic matter tion, although this is based on free energy

calculations which assume that the man- Studies in recent years have examined sed- ganese oxides undergoing reductive dissolu-

imentary manganese and iron reduction in a tion are well characterized, crystalline min- larger perspective than that discussed in the eral phases. However, the free energy of previous section, looking at these processes formation of manganese (or iron) oxides in the context of their role in the oxidation of varies not only among different oxide phases, sedimentary organic matter. In many sedi- but for a given phase also depends on the ments, there appears to be a well-developed available surface area, differences in the av- sequence of electron acceptors used in the erage Mn oxidation state of the phase and oxidation of sedimentary organic matter, with variability in the degree of crystallinity of the the "biogeochemical zonation" of these rein- phase (Murray, 1979; Hem et al., 1982). ineralization processes apparently controlled These factors should cause differences in the by the free energy yield for the oxidation of energetics and rates of microbial manganese sedimentary organic matter by each electron and iron reduction with different oxides, as acceptor (Claypool and Kaplan, 1974; Froe- has been observed in lab studies of these lich et al., 1979). Under steady-state condi- processes (Lovley and Phillips, 1987b; Bur- tions this leads to a series of well-developed dige et al., 1992). zones in the sediments where each process These observations then suggest that there predominates, with regions in the sediments could be a certain degree of "overlap" both where the reduction of manganese and iron between the energetics of these sub-oxic pro- oxides should be coupled to the oxidation of cesses, and their zonation in pelagic sedi- organic matter (see Fig. 2 ). The depth scales ments. This appears to be more important over which these remineralization processes for denitrification and manganese reduction, occur can vary among different sedimentary possibly because of similarities in the re- environments, and are controlled by factors ported AG ° values for these processes (see such as the organic carbon rain rate to the Fig. 2). Consistent with this suggestion, man- sediment surface, the bottom water oxygen ganese reduction has been observed to occur content, the sedimentation rate and the bio- before the complete reduction of pore water turbation rate (Froelich et al., 1979; Bender nitrate in several pelagic sediments (Klink- and Heggie, 1984; Emerson, 1985; Canfield, hammer, 1980; Lyle, 1983; Salwan and Mur- 1993). Deep-sea pore water profiles are of- ray, 1983; also see Fig. 3). These pore water ten consistent with this biogeochemical zona- data also sometimes suggest the possibility of tion model (Froelich et al., 1979; Klinkham- overlap between the zones of manganese and mer, 1980; Jahnke et al., 1982; Salwan and iron reduction, although differences in the Murray, 1983; Bender and Heggie, 1984; energetics of these processes do not strongly Murray and Kuivala, 1990). support this occurrence (see Fig. 2). Further-

As processes coupled to the oxidation of more, ferrous iron is known to chemically organic matter, microbial manganese and reduce both nitrate (Buresh and Moraghan, iron reduction are less efficient than aerobic 1976) and manganese oxides (see references (oxic) respiration and more efficient than in Chemical (abiotic) manganese and iron re- sulfate reduction (which is generally referred duction and Fig. 4), and these reactions may to as anoxic metabolism; see Fig. 2). Along contribute to the appearance of a more dis- with denitrification, manganese and iron re- tinct boundary between the zone of iron re- duction are referred to as "sub-oxic" rein- duction and the zones of the other sub-oxic ineralization processes (e.g., Froelich et al., processes. 1979). Manganese reduction is generally con- The apparent upper boundary of the iron

Page 9: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 257

Concen t ra t ion (gM) of these reac t ions b e t w e e n fe r rous iron and

0 2o 4o 6o 8o m a n g a n e s e oxides and ni t ra te , and for the 0 , , , , , , fact tha t the oxidant of the upward ly diffus-

/ e ing fe r rous iron is likely not 0 2 but r a t h e r is ~ e Nitrate e i the r m a n g a n e s e oxides or n i t ra te (Klink-

p , , ~ o h am m er , 1980; S~rensen, 1987; also see Fig. , ~ 4).

10 •

E \ CONCENTRATION . . . . . . . .

t-- ' ~ • / ' ( ) 2 J Corg +02 " ~ CO2+H20 ~ - /X ~ Manganese NO V Mn2++O 2 ~ MnO2

20 /~ A Corg + NO3 - ~ CO 2 + N 2 Corg + MnO 2 ~ CO 2 + Mn 2+

L~ ~ .r ~ -~Mn2 + MnO2 + Fe2+ - ) " Mn2++ FeOOH

• ~ ~ NO~ + Fe 2* ~ N 2 + FeOOH /~ Iron •

/ I A •

30 Corg + FeOOH ~ CO 2 + Fe 2+ Fig. 3. Pore water concentrations of nitrate (dots), ', manganese (filled triangles), and iron (open triangles), all versus depth, in a core collected at MANOP site M (re-drawn from data presented in Klinkhammer et al., 1980). This site is in the eastern equatorial Pacific, ;ONCENTRATION 20-30 km east of the spreading axis of the East Pacific ~,~ = Corg + 02 ~ C O 2 + H20 Rise, in a water depth of about 3100 m. ,-'b~" . . . . . . . . . j Mn2++ O z ~ MnO=

r educ t ion zone (as in fe r r ed f rom the g r e e n - ~ Fe =÷ *o2 --> F.oo. b rown color t rans i t ion in sed iments ) gener- ~ Cor~*NO~ --> CO2.N= ally occurs be low the m a n g a n e s e r educ t ion , / ~ \M C°rg *Mn02 ~ C02+Mn2÷

zone and at the d e p t h o f comp le t e n i t ra te ~ ' ~ ~ n2+

removal f r om the po re waters (Lyle, 1983). ~ \ ~ Because of reac t ions b e t w e e n fe r rous iron and m a n g a n e s e oxides and ni t ra te , any Fe 2 + c.,g. F.OOH ---> CO~ * F2 ÷

p r o d u c e d by i ron r educ t ion in the p r e sence o f m a n g a n e s e oxides or n i t r a te (i.e., in zones ', o f m a n g a n e s e r educ t ion a n d / o r deni t r i f ica- t ion) should be re -oxid ized back to i ron(III )

oxides, r a t he r t han accumula t e in the po re Fig. 4. A.(above) Pore water profiles in which the waters . This may then explain the c lea re r oxidation of iron in the pore waters is coupled to the sepa ra t ion o f the a p p a r e n t zone of i ron re- reduction of either manganese oxides or nitrate (see duc t ion f rom the zones of e i the r deni t r i f ica- equations to the right). Note that similar profiles are

generally seen in many pelagic sediments (e.g., see Fig. t ion or m a n g a n e s e reduc t ion , regard less of 3). B.(below) Hypothetical pore water profiles that any actual over lap b e t w e e n these sub-oxic would exist if 02 was the primary oxidant of pore processes [also see Myers and Nea l son water iron (see equations to the right). Note that here (1988b) and Canf ie ld (1993) for a f u r t he r it is assumed that iron oxidation is no t coupled to

either the reduction of manganese oxides or nitrate. discussion of this p h e n o m e n a ] . T h e lack of a Therefore iron must diffuse through the zones of man-

l inear po re wa t e r i ron prof i le t h rough the ganese reduction and denitrification to a depth in the zones of m a n g a n e s e r educ t ion and deni tr i f i - sediments where the upward iron flux reaches the cat ion provides f u r t h e r ev idence in suppor t downward oxygen flux.

Page 10: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

258 D.J. BURDIGE

Based on sediment concentrations, man- tween reducing conditions in pelagic sedi- ganese and iron oxides should have a large ments and the importance of sub-oxic remin- capacity for oxidizing sedimentary organic eralization, because the sites studied by Ben- matter (Reeburgh, 1983). However, evidence der and Heggie (1984) also had relatively to date suggests that manganese and iron shallow ( ~ 1-5 cm) surface oxic zones. reduction are not quantitatively important in Bender and Heggie (1984) also presented oxidizing organic matter in pelagic sedi- a series of "theoretical" flux calculations ments. Bender and Heggie (1984) used pore which place general constraints on the im- water data from five pelagic sediments to portance of aerobic respiration, denitrifica- estimate rates of sedimentary nitrate, man- tion and manganese and iron reduction in ganese and iron reduction, and then used the oxidation of organic matter in pelagic "Redfield"-type ratios to convert these rates sediments. These calculations agree with to rates of organic carbon oxidation by each their pore water results, again indicating the electron acceptor. Their results suggest that minor role that secondary oxidants should reactions involving all "secondary" oxidants play in organic matter remineralization in (nitrate, manganese and iron oxides and sul- pelagic sediments. Similar results were ob- fate) account for less than 10% of the total tained by Rabouille and Gaillard (1991), us- organic carbon oxidized in these sediments, ing a steady-state diagenetic model for or- with aerobic respiration accounting for the ganic matter oxidation that includes aerobic vast majority (> 90%) of the sedimentary respiration, denitrification and manganese organic matter remineralization. At the same reduction. Iron reduction was not included time, it should also be noted that the occur- in their model, although its omission does rence of reactions such as iron oxidation not significantly affect this conclusion. coupled to either manganese or nitrate re- While manganese and iron reduction do duction complicates the use of calculated not appear to be significant in the remineral- manganese, nitrate and iron reduction rates ization of organic matter in pelagic sedi- to estimate amounts of carbon oxidized by ments, they may be important in other as- each sub-oxic process. However, this does pects of organic carbon dynamics in these not significantly affect their conclusion that sediments. In almost all marine sediments it sub-oxic processes do not account for a sig- is observed that some fraction of the organic nificant fraction of the observed organic car- carbon deposited to the sediment-water in- bon remineralization in the sediments they terface appears to escape remineralization. examined. An understanding of the factors controlling

In contrast, Sayles and Curry (1988) pre- this burial (or preservation) of sedimentary sented pore water calculations which suggest organic matter is important for several rea- that denitrification and manganese reduction sons, in part because this process represents account for up to ~ 40% of the total carbon the primary long-term repository for organic oxidized in two deep-sea sediments with rel- carbon in the global carbon cycle (Emerson atively shallow ( ~ 2-3 cm) surface oxic zones, and Hedges, 1988). Bender and Heggie (1984) In three other less reducing sites, with much showed that the amounts of organic matter thicker ( > 20 cm) oxic surface zones, sub-oxic remineralized by secondary oxidants (prim- processes account for < 10% of the total arily nitrate, manganese oxides and sulfate) carbon remineralized, a value similar to that appear to be comparable to the amounts of observed by Bender and Heggie (1984). organic matter preserved in the sediments. However when the results of both of these Based on this observation, they suggested studies (Bender and Heggie, 1984; Sayles that factors affecting organic carbon oxida- and Curry, 1988) are compared, one sees tion by secondary oxidants may thus play an that there is not a simple relationship be- important role in controlling the preserva-

Page 11: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 259

tion of organic matter in pelagic sediments, model in Fig. 2, there have been no defini- A similar conclusion was also reached by tive studies to date that have demonstrated Rabouille and Gaillard (1991) in their model- the occurrence of microbial manganese and ing study, iron reduction coupled to the oxidation of

In contrast, Emerson and co-workers sedimentary organic matter where pore wa- (Emerson, 1985; Emerson et al., 1985; Emer- ter data suggests that it is occurring. In addi- son and Hedges, 1988) have argued that since tion, the discussion above further indicates the vast majority of organic matter in pelagic that the use of pore water profiles to deter- sediments is oxidized by O2, small changes in mine these biogeochemical zones is still quantities affecting the rate of aerobic respi- somewhat tentative, due in part to an incom- ration (e.g., the organic carbon rain rate to plete understanding of all the chemical and the sediment surface or the bioturbation rate) biological reactions affecting manganese and ultimately control the preservation of sedi- iron redox cycling in pelagic sediments (e.g., mentary organic matter. Secondary oxidants the importance of manganese or nitrate re- are therefore presumed to play a minor role duction coupled to ferrous iron oxidation in both the remineralization and preserva- versus organic matter oxidation). Studies in tion of organic matter in pelagic sediments, all of these areas are needed to resolve these Unfortunately, it is difficult at this time to uncertainties. evaluate these seemingly contradictory asser- tions, in part because of the above-discussed The role of manganese and iron reduction in uncertainties in using pore water profiles to the formation offerromanganese nodules infer rates of manganese and iron reduction coupled to organic matter oxidation. In addi- Manganese reduction in pelagic and tion, recent studies have also demonstrated hemi-pelagic sediments may be involved in the important role that other factors m a y the formation of ferromanganese nodules, play in controlling the preservation of sedi- since under some circumstances manganese mentary organic matter. These include both reduction in the sediments and the resulting the composition of the sedimentary organic upward diffusion of dissolved manganese matter itself (Westrich and Berner, 1984; across the sediment-water interface may be Middelburg, 1989; Burdige, 1991), as well as an important source of manganese to nod- the possible importance of oxygen (and pro- ules (Calvert and Price, 1977; Grill, 1978; cesses associated with oxic conditions in sedi- Moore et al., 1981; Dymond et al., 1984; ments) in affecting the remineralization of Glasby, 1984). These processes may also play certain types of organic compounds found in a role in the growth of freshwater manganese sediments (Henrichs and Reeburgh, 1987; nodules (see the discussion below and in Emerson and Hedges, 1988; Canfield, 1989b; Dean et al., 1981). Jahnke et al., 1989; Pedersen and Calvert, In a detailed study of the geochemistry of 1990; Henrichs, 1992; Lee, 1992). These ob- manganese nodules at several contrasting servations therefore suggest that the factors sites in the Pacific, Dymond et al. (1984) controlling organic matter preservation in referred to this process as the "sub-oxic dia- marine sediments are a function of both the genesis" nodule accretionary process. How- organic matter itself as well as the specific ever in the sites they studied, this process remineralization processes associated with its appears to be involved in the formation of oxidation. Future studies are clearly needed nodules only at a hemi-pelagic site which to resolve this complex problem, underlies productive waters in the eastern

Finally, it should be noted that while tropical Pacific (MANOP site H). Further- deep-sea pore water profiles are generally more, while nodule composition data suggest consistent with the biogeochemical zonation the occurrence of this sub-oxic diagenetic

Page 12: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

260 D.J. BURDIGE

mechanism in the formation of nodules at Fig. 5) and interfacial pore water manganese this site, pore water data are apparently in- profiles (Heggie et al., 1986; see Fig. 6) pro- consistent with the possibility. This is the vide additional evidence in support of the case because in these and other hemi-pelagic occurrence of manganese reduction in some sediments, the "major" zone of manganese "oxic" surficial sediments. These authors dis- reduction is generally not sufficiently close to cuss the possible role that this type of man- the sed iment -water interface (e.g., see Fig. ganese reduction may play in manganese 3), and any manganese produced at depth nodule formation, and Kalhorn and Emerson does not likely escape oxidation and precipi- (1984) used a simple box model to demon- tation in the upper, oxic sediments (see also strate that this rate of surficial sediment the discussion in M A N G A N E S E AND manganese reduction at MANOP site H is IRON R E D U C T I O N IN C O N T I N E N T A L ~ 5 times the manganese accumulation rate M A R G I N SEDIMENTS). To reconcile these in nodules of this region. These authors also apparent contradictory observations Dymond suggest several possible explanations for the et al. (1984) suggested that nodule accretion occurrence of this manganese reduction in as a result of sub-oxic diagenesis is not a oxic surficial sediments, including: (1) man- steady-state process, and that it occurs be- ganese reduction in anoxic microenviron- cause of episodic high productivity events ments within the sediments; (2) possible pro- that lead to transient "zones" of manganese duction of dissolved organic compounds ca- reduction in sediments near the sed iment - pable of reducing manganese oxides in the water interface, presence of oxygen and (3) differences in the

Sediment manganese oxidation state mea- forms and reactivity of manganese oxides surements (Kalhorn and Emerson, 1984; see found at the sediment -water interface rela-

Solid phase manganese (Mn s, mg/g) x in MnO x

0 20 40 60 80 1.4 1.6 1.8 2.0 0 , ,

4 \ A'x Mns o

E • o

c- 1 2

Q .

c~ 16

• o

20 pw o

o • BC6

,., BC12 o o BC12 24 , I ~ I

0 10 20 30

Pore water manganese (Mnpw, pM)

Fig. 5. Solid phase and pore water manganese concentrations (right), and the oxidation state of manganese in the solid phase (x in MnOx; left), all versus depth, in box cores collected at MANOP site H (re-drawn from data presented in Kalhorn and Emerson, 1984). Site H is in the eastern equatorial Pacific (water depth 2590 to 3420 m), about 900 km east of the East Pacific Rise in a sediment plain covered with manganese nodules. As discussed in the text, the minimum in the value of x at a depth of ~ 2 cm is taken as an indication of the occurrence of manganese reduction in the otherwise "oxie" surface sediments. Note the "primary" zone of manganese reduction below a depth of ~ 12-14 cm.

Page 13: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

B I O G E O C H E M I S T R Y O F M A N G A N E S E A N D I R O N R E D U C T I O N I N M A R I N E S E D I M E N T S 261

Ni t ra te (~tmol/kg) M a n g a n e s e

(nmol/kg) (pmol/kg) 20 40 60 0 40 80 120 10 20 30

0 ' 'o -~r-'°'~6~ , o 0 t - . . . .

I0 @ / • o o ~...~° /@" / o " • 0 •

/ / ko . ' ~ . • O - - 4 / / 4 -- i

• o To ~ / / • O •

/ / \ \ • 0 0 • E / / , \

v • 0 0 • 8 y 8 \ I

~_ • 0 •

e o o •

0 0 O

16 16 ; , Fig. 6. Fine scale pore water profiles of manganese and nitrate, versus depth, in replicate cores collected at M ANOP site M (re-drawn from data presented in Heggie et al., 1986). Note the scale break in the concentration axis of the manganese profile. The observed manganese production in the upper 1 cm of these sediments is in excess of that predicted from oxic organic matter remineralization and opaline silica dissolution, and is therefore assumed to come from the reduction of sedimentary manganese oxides (Heggie et al., 1986). As discussed in the text, this manganese reduction appears to differ from that which occurs in the "primary" zone of manganese reduction below ~ 5 cm. These profiles show the same general trends as those seen in previous studies at site M (see Klinkhammer et al., 1980, and Fig. 3), although better resolution near the sediment-water interface has been obtained here.

tive to those found deeper in the sediments remobilization of manganese in this surface (i.e., those which are produced during the layer, since manganese reduction is thermo- deeper diagenetic recycling of manganese in dynamically favored over iron reduction (see the "main" zone of manganese reduction). Fig. 2; note that this anoxic sediment cover While there is evidence in the literature con- might result from the rapid sedimentation of sistent with each of these possibilities, none phytoplankton debris after the "crash" of a of them has been clearly shown to be respon- bloom in the water column). These anoxic sible for the occurrence of manganese reduc- conditions then transform the nodule surface tion in apparently oxic surficial sediments, into one which is now enriched in iron, due

Related non-steady state phenomena may to this selective manganese removal. The also play a role in the formation of alternat- thickness of this "new" iron-rich layer, rela- ing manganese and iron rich layers common tive to the now deeper manganese-rich layer, in many marine and freshwater nodules depends on the time of nodule accretion (Sorem and Fewkes, 1977; Moore, 1981). In under oxic conditions versus the time of Oneida Lake, Moore et al. (1980) and Moore burial under reducing condition. Removal of (1981) have shown that manganese nodules this anoxic sediment cover (e.g., by a storm generally accrete a manganese-rich layer event) allows for the accretion of new man- when exposed to oxygenated lake bottom ganese rich material at the nodule surface, waters. However, temporary burial of nod- and the repetition of this process leads to the ules by organic-rich sediment apparently observed banding in these nodules. This me- leads to the development of anoxic condi- chanism may also play a role in the growth of tions at the nodule surface and the selective deep-sea nodules, although deep-sea nodules

Page 14: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

262 D.J. B U R D | G E

have not been examined in the context of of Burdige and Gieskes (1983)predicts that such processes, only one manganese peak should be ob-

served in sediments (e.g., see Fig. 1). There- The effects of Quaternary paleoceanographic fore, the existence of multiple manganese changes on manganese reduction in pelagic peaks has been attributed to the occurrence sediments of non-steady state manganese redox cycling.

In the simplest sense, these non-steady state A common observation in many pelagic conditions result from the net migration (rel-

and hemi-pelagic sediments is the occur- ative to the sediment-water in terface)of the fence of multiple solid phase manganese manganese redox boundary, since under peaks in the sediments (e.g., see Fig. 8). Such steady state conditions the redox boundary multiple manganese peaks are observed over moves upwards at a rate equal to the sedi- a variety of depth and time scales, in some mentation rate and therefore remains at a cases occurring as mm-scale laminae, rich in fixed depth in the sediments. Assuming that both manganese and iron (e.g., see the dis- the manganese redox boundary occurs at the cussion in Wilson et al., 1986). However, depth in the sediments where pore water multiple manganese rich layers of cm-scale oxygen goes to zero (see Fig. 2 and GEN- thickness are also found in the upper meter ERAL ASPECTS OF SEDIMENTARY to several meters of some pelagic and hemi- M A N G A N E S E AND I R O N R E D U C - pelagic sediments, over time scales that go TION), the depth of this boundary will in- back not only to the most recent glacial-in- crease with time (i.e., occur at a deeper terglacial transition (i.e., the Glacial-Holo- depth) if, for example, bottom water oxygen cene boundary) but also cover much of the concentrations increase, the flux of organic late Quaternary (Froelich et al., 1979; Gard- carbon to the sediments decreases, a n d / o r her et al., 1982; Berger et al., 1983; Price and the rate constant of sedimentary organic Froelich, 1987; Finney et al., 1988; Price, matter oxidation decreases (e.g., due to 1988; Dean et al., 1989; Mangini et al., 1990). changes in the reactivity of this material).

In the following discussion, the occurence Similarly, the depth of this redox boundary of these manganese peaks will be taken to be will decrease if these parameters change in indicators of either the present-day location the opposite direction. It should also be noted or the previous location(s) of the manganese here that changes in these parameters do not redox boundary (see the discussion below represent the only ways to cause shifts in the and in G E N E R A L ASPECTS OF SEDI- depth of the manganese redox boundary, and MENTARY MANGANESE AND IRON they are listed here simply as examples of REDUCTION), based on the assumption possible ways the depth of this boundary can that the predominant form of manganese in be shifted. these peaks is manganese oxides. However, if Evidence in support of such relationships some (or all) of this manganese actually ex- between the depth of the manganese redox ists as a carbonate (rather than an oxide) boundary in sediments and the parameters phase, these peaks would likely form some discussed above comes from several observa- depth (on the order of several tens of cm) tions. The first are experimental and mod- below the manganese redox boundary elling results (Emerson, 1985; Emerson et (Pedersen and Price, 1982; Burdige and al., 1985) that indicate the appropriate rela- Gieskes, 1983; Thomson et al., 1986; Schim- tionships between these parameters and the mield and Pedersen, 1990). This occurence depth in pelagic sediments at which pore would obviously greatly affect the interpreta- water oxygen concentrations go to zero. In tions discussed below, addition, other studies have observed that

Under steady-state conditions the model the depth of the present-day manganese re-

Page 15: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 263

dox boundary (as inferred from, e.g., man- ary into the zone of manganese reduction, ganese pore water profiles)is inversely corre- where it may then slowly dissolve and be lated with the flux of organic carbon to the reprecipitated at the new redox boundary sediments (Lynn and Bonatti, 1965; Swin- (see the discussion below for further details). banks and Shirayama, 1984; Finney et al., In contrast, an upshift of the redox bound- 1988). ary may lead to the entrapment of the relict

A deepening redox boundary leads to a manganese peak in the new zone of man- situation in which the downward progressing ganese reduction (see Fig. 7). This peak will redox front "leaves" behind the manganese then either be preserved in the sediments, or oxide peak at the old redox boundary. "New" undergo reductive dissolution and be re-pre- manganese oxides are precipitated at the cipitated at the new redox boundary. The current redox boundary from the upward dif- extent to which either of these occurs is fusion of the standing stock of pore water dependent on several factors which will be manganese in the deeper oxygen-depleted discussed below. Nevertheless, if the redox sediments. A distinct manganese peak will be upshift occurs rapidly (relative to rates of seen to grow in at the new redox boundary if manganese reduction and sediment accumu- this redox downshift is relatively rapid and lation) and the new redox boundary remains occurs a sufficient distance from the original stable for a sufficient time period, a distinct redox boundary (see Fig. 7). Furthermore, manganese peak will be seen to grow in at when the old manganese peak is buried down the new redox boundary (see Fig. 7). to the new redox boundary, it then becomes Since pore water manganese profiles re- involved in manganese redox cycling at this spond more rapidly to changes in sediment new boundary. However, the possibility also redox conditions than do solid phase man- exists that some, or all, of this "old" peak ganese profiles (Froelich et al., 1979; Peder- may be buried through the new redox bound- sen et al., 1986; Price, 1988; Schimmield and

Solid Phase Mn Conc. ~ In-growth of a Mn peak at the new redox boundary

~ n e of Mn Oxidation i ~ J ZoneofMnOxidation [ ZoneofMnOxidation - - ~)eod°Xdary

12- redox ........... Dissolution of the relict

G) S - - b ° u n d a r y ~ Mn ~ x ~, Relict Mn peak at t h e z o n e of Mn reduction t~, ~ former redox boundary* ..... " Mn peek at the former

redox boundary • "...(--a. . . . . . redox ~ . . . . . boundary

I Zone of Mn reduction • Zone of reduction In-growth of a Mn peek at the new redox boundary

The Distribution of Solid Phase Upward Shift of the Manganese Redox Downward Shift of the Manganese Redox Manganese Under Steady State Boundary Boundary

Conditions * Note that the relict Mn peek is unaffected by diegenetlc redox cycling until it is buried to the depth of the new redox boundary.

Fig. 7. The effects on solid phase manganese profiles due to changes in the depth of the redox boundary between the zones of manganese oxidation and reduction. The two figures to the right of the steady state profile represent manganese profiles that would be observed some time after the "instantaneous" (relative to the time scales of other sedimentary processes) upshift or downshift of the steady-state redox boundary. These changes in the depth of the manganese redox boundary are presented here in this fashion to more clearly illustrate the ways in which multiple manganese-rich layers can form in pelagic sediments.

Page 16: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

264 D.J. BURDIGE

Pedersen, 1990), pore waters quickly adjust boundary occurs relatively slowly compared to changes in the depth of the manganese to the rate of sediment accumulation a n d / o r redox boundary and begin precipitating a the in situ rates of manganese oxidation and new solid phase manganese peak at the new reduction, one should see a broad diffuse redox boundary. Therefore in sediments con- solid phase manganese peak whose one limit taining multiple solid phase manganese is the old redox boundary and whose other peaks, the manganese pore water profile may limit is the position of the current redox be useful in distinguishing which of the man- boundary. Again, pore water manganese pro- ganese peaks is the active diagenetic feature files should be useful in elucidating the posi- associated with the current redox boundary, tion of the current (active) redox boundary. Relict peaks observed above the active redox Examples of what may be relatively rapid boundary are indicative of a redox downshift, and slow redox upshifts can be seen in Fig. 8 while relict peaks found below the current for two cores from the equatorial Atlantic redox front likely result from a redox upshift, originally described by Froelich et al. (1979). However, the possibility also exists that relict In contrast, Pedersen et al. (1986) presented peaks found below the current redox bound- results from a core collected near the East ary may be features associated with a redox Pacific Rise in which there appears to have downshift (e.g., see the discussion above), been a ~ 10 cm redox downshift which they The interpretation of these peaks in the con- suggested occurred within the past 1000 yr. text of the redox history of a sediment and However model calculations presented by the controls on sedimentary redox conditions Price (1988) indicate that this redox down- will be discussed below in greater detail, shift may have occurred much more recently,

Finally, if the movement of the redox perhaps within the past 10-20 yr.

Pore water manganese (Mnpw I.tM)

0 20 40 60 10 20 0 ' r ' , ~ ' ' '

~ M n s

1 • ~ ' A

20 -

~ •

° \ ~ 4O (1) Mnpw ~ Mnpw

P 60 A

Core 5GC1 Core 14GC1 J I L I i , l J I i

10 20 30 0 2 4 6

Solid phase manganese (Mns,mg/g) Fig. 8. Pore water (open triangle) and solid phase (dots) manganese, both versus depth, in cores collected in the eastern equatorial Atlantic (re-drawn from data in Froelich et al., 1979). Based on the discussion in the text (also see Fig. 7) it appears as if the rate of redox upshift was more rapid in core 5GC1 than in core 14GC1. However other phenomena may also possibly explain the observed differences between these two cores.

Page 17: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 265

As discussed in the beginning of this sec- sponse to a relatively recent, short-term in- tion, multiple solid phase manganese peaks crease in ocean fertility that occurred within are observed in many pelagic and hemi- the last few thousand years. pelagic sediments (e.g., see Fig. 8). In dis- Evidence for such a recent redox upshift cussing the processes leading to the forma- can be seen, for example, in core 5GC1 in tion of such layers in Glacial-Holocene sedi- the eastern equatorial Atlantic (Fig. 8) where merits in the eastern equatorial Pacific, Ber- the manganese pore water profile indicates ger et al. (1983) suggested that the upper that the upper manganese peak is found at layer (at 7-10 cm) is the active redox bound- the current redox boundary. Price (1988) also ary while the deeper layer (near 30 cm and presented scaling calculations similar to those centered on the Glacial-Holocene bound- described in Froelich et al. (1979) which sug- ary) is found at the relict redox boundary, gest that the in-growth of such an upper They further proposed that this resulted from manganese peak after a redox upshift can a decrease in ocean fertility (primary produc- occur in less than 5000 yr. Finally, cores tion) which occurred during the transition showing such multiple manganese peaks oc- from Glacial to Holocene conditions (Peder- cur predominantly in sediments underlying sen, 1983). However, because such a de- eastern equatorial upwelling systems in the crease in primary productivity should have Atlantic and Pacific Oceans (Price, 1988; led to a decrease in carbon fluxes to the Price and Froelich, 1987), and these authors sediments, Price (1988) later recognized that have suggested that the formation of these such a decrease in ocean fertility at the be- multiple manganese peaks is related to rela- ginning of the Holocene should have shifted tively rapid, large-scale climate events such the active redox boundary to deeper depths as the E1 Nif io /Southern Oscillation (EN- in the sediments (e.g., see the redox down- SO). In support of this suggestion, they noted shift scenario in Fig. 7 and the discussion in that changes in sedimentary organic carbon Thomson et ai., 1984). The occurrence of the concentrations are generally associated with relict manganese peak below the current re- these multiple manganese layers, which they dox boundary as a result of a redox upshift interpreted as indicators of variation in the further indicated to Price (1988) that if the organic carbon rain rate to these sediments scenario of Berger et al. (1983) was correct, over time. then the present-day manganese profile could In an attempt to further quantify the time only have been obtained if this relict peak scales over which such multiple manganese had also been buried through the current peaks form in these sediments, Price (1988) redox boundary, developed a non-steady state manganese dia-

Price (1988) also offered an alternate ex- genetic model based on the steady-state planation for these observations. She sug- equations presented by Burdige and Gieskes gested that a drop in ocean fertility during (1983). In her model the depth of the man- the early part of the Holocene indeed estab- ganese redox boundary is externally specified lished a deeper redox boundary in these sed- and allowed to vary with time. She also as- iments, and resulted in the formation of the sumed (as was the formalism in Burdige and deeper manganese peak. The upper (or relict) Gieskes, 1983) that the rate of manganese pre-Holocene (i.e., Glacial) manganese peak reduction is first order with respect to the was then buried to the early Holocene redox amount of solid phase manganese present, boundary, where it was incorporated into the and therefore that relict manganese peaks manganese redox cycling occurring at this trapped in the manganese reduction zone boundary. She then proposed that the forma- undergo slow, continual reductive dissolution tion of the second, surficial manganese peak (e.g., see Fig. 7). While these assumptions (above the deep manganese peak) is a re- have certain disadvantages (see below), her

Page 18: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

266 D.J. BURDIGE

model results lead to several interesting sug- the sediments (see Fig. 10). One of the most gestions concerning the relationship between interesting points to come from these calcu- observed sediment manganese depth profiles lations is that the required increase in car- and paleoceanographic conditions during the bon fluxes to the sediments (which then leads Holocene. to such a redox upshift) need only have oc-

For a core in the eastern equatorial Pa- curred within the last ~ 1500-2000 yr either cific, calculations with the Price (1988)model as a single event (as in Fig. 10) or as an indicate that alternating upshifts and down- interrupted sequence (as in Fig. 9) to create shifts of the manganese redox boundary over the present-day sedimentary manganese dis- the past ,,, 2000 yr (caused perhaps by tributions. From a paleoceanographic stand- ENSO-type climate and primary productivity point, this suggests that the present-day variations) can lead to a manganese distribu- "normal" patterns of equatorial upwelling tion in the sediments that is consistent with and carbon fluxes to the sediments of these the observed data (see Fig. 9). Furthermore, regions are not typical of the entire Holocene, for core 5GC1 in the eastern equatorial At- and that the early Holocene ocean was likely lantic (see Fig. 8), her model calculations less productive and perhaps more reminis- also show that the upper manganese peak in cent of what is now seen during presumably these sediments can form within approxi- "atypical" ENSO conditions (Price, 1988). mately 2000 yr of a single redox upshift in Similar changes in the depth of the man-

Mns(mg/g) Mns(mg/g)

0 0 5 10 15 0 5 10 15 A~.~ I I ~ I I

I A 10

.... - A ~ A . . . . . . A • ~ ' J " ! • • E : ! -.,

2 0 - : • • " ' . . . • " " • t = O

0 • . . . . . t = O • ..........

30 / ......A. 80 yr 2200 yr

40 " • : ) ' - " ¢

Fig. 9. The results of Price's (1988) non-steady state manganese model applied to data from a core (RC 52) in the eastern equatorial Pacific (data from P.N. Froelich, pers. commun, to B. Price; note that other data from this core are reported in Murray et al., 1984). The initial conditions used for each model run (i.e., the t = 0 profiles) were determined as discussed in Price (1988) and are shown in each figure as a dotted line. The final profiles for the two model simulations are shown as solid lines, and the actual data are shown as triangles. A.(left) A model simulation in which the manganese redox boundary is alternately assigned to the depths of 16 cm and 27 cm using a time schedule for the occurrence of E N S O events derived with data from the Quelccayo ice cap in Peru. There a six "instanta- neous" redox shifts in this t ime schedule over the 1480 yr simulation, with the time between redox shifts ranging from 120 to 350 yr. Price (1988) assumed that ENSO-domina ted time periods (which are characterized by higher sea surface temperatures and reduced equatorial upwelling) are periods of lower carbon flux to the sediments. This then leads to more oxidizing conditions in the sediments, and therefore to a redox downshift. Conversely, ENSO-poor times are assumed to be dominated by a redox upshift in the sediments, due to increased upwelling and the resulting increase in carbon flux to the sediments. B.(right) A model simulation in which the manganese redox boundary is alternately assigned to the depths of 16 and 19 cm at 270 yr intervals over 2200 yr.

Page 19: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 267

ganese redox boundary apparently also oc- (see Fig. 11). In contrast to the Price (1988) curred several times earlier in the Quater- model, Finney et al. (1988) assumed that a nary. Gravity cores from hemi-pelagic sedi- relict manganese peak trapped in the man- ments in the eastern equatorial Pacific (MA- ganese reduction zone by a redox upshift can NOP site H) exhibit several peaks in both become "permanently" buried in the sedi- solid phase manganese and organic carbon ments, since they suggest that little of this concentrations over the past ~ 400,000 yr, peak is remobilized once it is buried below and Finney et al. (1988) have suggested that the redox boundary. The roughly 100,000 yr they are caused by periodic changes in redox periodicity seen in the depth of the man- conditions that are similar to those described ganese redox boundary at this site appears to above for the most recent glacial-interglacial be consistent with other paleoceanographic (i.e., the Glacial-Holocene) transition. To information on changes during the Quater- further examine these changes they devel- nary of quantities (e.g., surface water pro- oped a simple semi-quantitative model that ductivity) which likely affect the depth of this used the organic carbon and manganese pro- redox boundary (Finney et al., 1988; Lyle et files to hindcast the depth of the manganese al., 1988). redox boundary over the age of the cores The models of Price (1988) and Finney et

al. (1988) have provided important insights into possible relations between manganese

a%(mgtg) depth profiles in pelagic and hemi-pelagic 00 8 16 24 32 sediments and Quaternary paleoceanograph-

I ~ ' ' ' ic changes that affect sedimentary organic matter remineralization and burial. How- ever, as discussed above, there appear to be

10 questions about several key assumptions made in these models, in part because of the ways that both models specify the depth of

2o " .... the manganese redox boundary and the ki- ""'"-.... t--0 °E f ~ ""'" netics of manganese reduction. This problem

could be overcome by developing a more . 3 o rigorous time-dependent diagenetic model

" I t--aooo yr (e.g., an expansion of the steady state model of Rabouille and Galliard, 1991) which, for

40 example, would directly calculate the depth of the redox boundary with model-derived oxygen profiles (which would be determined

5o using parameters such as the bottom water oxygen concentration and the organic carbon flux to the sediments). Such an approach

6o would appear to be useful in further examin-

Fig. 10. The results of Price's (1988) non-steady state ing the conclusions of these studies. manganese model applied to data from core 5GC1 in the eastern equatorial Atlantic (see Fig. 8; data from The occurrence of manganese and iron reduc- Froelich et al., 1979). The initial condition used for this tion during the diagenesis of deep-sea tur- model run (i.e., the t = 0 profile) was determined as bidites discussed in Price (1988) and is shown as a dot ted line. The final profile (shown as a solid line), results from a 2000 yr simulation following an instantaneous redox The deposition of organic-rich turbidites upshift from 26 to 7 cm. o v e r m o r e organic-poor pelagic sediments is

Page 20: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

268 D.J. BURDIOE

an additional non-steady-state system in dizing organic carbon at a downward pro- which manganese and iron redox processes gressing redox front within the turbidite. This play important roles. Turbidites represent process has been described as one in which sediments t ransported by turbidity currents these oxidants "burn down" through the or- from the continental shelf down the conti- ganic rich sediment, and hence this redox nental slope and into the deep ocean basins front has been described as a burn-down (Open University, 1989). In general, tur- front (Wilson et al., 1985). In the context of bidites contain levels of organic carbon that the discussion in the third part of this section are elevated over those in most pelagic sedi- (and Fig. 7), these turbidite systems are simi- ments, and the thicknesses of turbidite se- lar to sediments in which there is a slow quences range from tens of centimeters to redox downshift. The current location of the several meters. Initial geochemical studies of burn-down front serves as a locus for sub- these systems focussed on the effects of the surface aerobic respiration, iron and man- deposition of a single turbidite in a pelagic ganese oxidation, and denitrification, as can sediment on organic mat ter remineralization be seen in pore water, solid phase and deni- and mobilization of redox-sensitive elements trification depth profiles (S0rensen et al., (Colley et al., 1984; Wilson et al., 1985, 1986; 1984; Wilson et al., 1985, 1986; Wallace et Wallace et al., 1988; Jahnke et al., 1989). al., 1988). The burn-down process also re- However, later studies have also examined suits in color change in the turbidite from the geochemical consequences of repeated green to brown (Lyle, 1983) and the remobi- turbidite deposition in pelagic sediments lization of redox sensitive elements such as (DeLange, 1986;Thomson et al., 1987; Buck- iron, manganese and uranium. The time ley and Cranston, 1988; Dean et al., 1989; scales over which this burn-down process oc- Thomson et al., 1989). curs are related to quantities such as the

The deposition of an organic-rich turbidite thickness and initial organic carbon content over organic-poor pelagic sediments leads to of the turbidite, the pelagic sedimentation a situation in which oxygen and nitrate dif- rate, and bottom water oxygen and nitrate fuse downwards through the turbidite, oxi- concentrations (Colley et al., 1984; Wilson et

O -

5 -

u 10

-!,- i - o.. w 15

20

2 5 -

o 2 0 0 4 0 0

A G E ( k y r ) -

Fig. 11. Variations in the depth of the manganese redox boundary at MANOP site H over the past 400,000 yr, as determined using the model presented by Finney et al. (1988). Reprinted with permission from Paleoceanography, vol. 3, p.181, copyright 1988 by the American Geophysical Union.

Page 21: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

B1OGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 269

al., 1985). However, model calculations sug- Hines et al., 1991; Reimers et al., 1992). gest that burn-down fronts moves downward However, as with pelagic sediments, the at rates comparable to pelagic sedimentation mechanisms of these processes in margin rates (Colley et al., 1984; Wilson et al., 1985, sediments have not been clearly defined. Es- 1986). timates of manganese and iron reduction

In terms of manganese and iron redox rates in margin sediments (and their impor- chemistry, it is generally assumed that the tance in sedimentary organic carbon oxida- turbidite represents a source of dissolved iron tion) have also been made by modeling both and manganese which is then oxidized at the pore water and solid phase sedimentary pro- downward progressing redox front. The ulti- files (Heggie et al., 1987b; Bender et al., mate source of this upwardly diffusing iron 1989; Aller, 1990; Reimers et al., 1992). While and manganese is generally assumed to be this modeling approach does not always yield managnese and iron reduction occurring in an accurate estimate of the true rates of the reduced turbidite sediments, although the metal oxide reduction in sediments (see the exact mechanisms by which these reductions discussion in previous section and in Can- occur (i.e., what is the reductant), and its field, 1993), the results of these calculations relationships to other geochemical processes suggest that in the absence of significant in the turbidite sediments have not been well physical or biological "mixing" (or particle described. Several recent studies have exam- transport) of sediments, manganese and iron ined these problems more directly, including reduction play minor roles in organic matter a discussion of the possible role of ferrous remineralization in continental margin sedi- iron reduction of manganese oxides in tur- ments. bidite sediments (Thomson et al., 1987, 1989; The reasons for this observation likely stem Buckley and Cranston, 1988). This reaction is from differences in the ways that organic of some interest here because it appears to matter oxidants are transported in sedi- lead to secondary color changes in the tur- ments. In the absence of sediment "mixing" bidite sediments (referred to as diagenetic processes (e.g., bioturbation), sedimentation "haloes") beyond the brown-green color is the dominant process by which manganese change seen at the burn-down front (see and iron oxides move downward in sedi- above). Understanding the formation of these ments, while diffusion in sediment pore wa- diagenetic haloes appears to be of some use ters is the dominant transport process of in interpreting the post-depositional history dissolved oxidants such as oxygen, nitrate of interbedded pelagic and turbidite sedi- and sulfate. As such, in most margin sedi- ments, particularly those in which turbidite ments the diffusion of these pore water oxi- emplacement occurs on a time scale more dants appears to be sufficiently fast relative rapid than that for the complete burn through to sediment organic carbon fluxes and sedi- (or oxidation) of the uppermost turbidite mentation of manganese and iron oxides. (Buckley and Cranston, 1988). This then leads to a situation in which metal

oxide reduction plays a minor role in overall MANGANESE AND IRON REDUCTION IN CON- organic carbon remineralization, with the TINENTAL MARGIN SEDIMENTS predominant processes being aerobic respi-

ration and denitrification (Bender et al., The occurrence of manganese and iron 1989; Reimers et al., 1992), and occasionally

reduction in continental margin sediments sulfate reduction (Christensen, 1989; Hines has been demonstrated with pore water pro- et al., 1991). It should also be noted that files (Aller et al., 1986; Heggie et al., 1987b; Bender and Heggie (1984) placed similar, Bender etal . , 1989;Aller, 1990;Jahnke, 1990; and more quantitative, constraints on the Murray and Kuivala, 1990; Shaw et al., 1990; relative roles of manganese and iron reduc-

Page 22: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

270 D.J. BURDIGE

tion in the remineralization of organic mat- ter [see Mackin and Swider (1989) and Aller ter in pelagic sediments (see previous sec- (1993) for a further discussion of such phe- tion), nomena in coastal marine sediments]. Since

In contrast, studies of Amazon continental little of the Mn z+ produced in the sediments shelf and the Panama basin sediments (Aller escapes as a benthic flux, oxygen actually et al., 1986, 1991; Mackin et al., 1988; Aller, serves as the ultimate oxidant for organic 1990) have indicated ways that sediment mix- matter remineralization in these sediments. ing processes can lead to situations in which As a result, manganese oxides (and man- manganese and iron reduction play more sig- ganese redox cycling) therefore simply serve nificant roles in sedimentary organic matter to shuttle electrons between 0 2 and organic remineralization. In Panama Basin sedi- matter in the sediments (see Fig. 12), in ways ments, intense macrofaurlal activity leads to similar to those in which manganese redox extensive sediment mixing in the upper ~ 30 cycling in coastal sediments may couple sul- cm of sediment (Aller, 1990). Coupled with fide oxidation to 0 2 reduction (Burdige and high manganese inputs to these sediments, Nealson, 1986; Aller, 1993; see also Salt due to their close proximity to hydrothermal marsh sediments). sources of manganese on the East Pacific This coupling of bioturbation, manganese Rise, this leads to a situation in which man- redox cycling and organic matter remineral- ganese reduction appears to be responsible ization in Panama Basin sediments increases for all of the observed benthic carbon rein- the importance of anoxic decomposit ion in ineralization. At the same time, much of the the sediments while at the same time de- oxygen flux into these sediments is directly creasing the storage of reduced reaction utilized in the re-oxidation of reduced me- products, and hence "diagenetic evidence" tabolites such as Mn 2+ or N H 2, rather than of anoxic pathways of organic matter decom- in the oxidation of sedimentary organic mat- position (Aller, 1990). However, another way

A . H20 MnO2 Corg

02 Mn 2÷ CO2

B. H20 Mn02 HzS C02

O: Mn2+ SO~- Corg Fig. 12. Possible ways that manganese redox cycling can interact with the redox cycling of other elements in sediments (modified from Burdige and Nealson, 1986, and Aller, 1993). A. The coupling of manganese reduction to organic matter oxidation, and Mn z+ oxidation to O: reduction (e.g., as in Panama Basin sediments; Aller, 1990). B. The coupling of manganese reduction to sulfide oxidation (produced by sulfate reduction coupled to organic matter oxidation), and Mn 2+ oxidation to 0 2 reduction (e.g., as in Long Island Sound sediments; Aller, 1993).

Page 23: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

B I O G E O C H E M I S T R Y O F M A N G A N E S E A N D I R O N R E D U C T I O N IN M A R I N E S E D I M E N T S 271

that these interactions may affect sediment result of bacterial sulfate reduction (i.e., sul- processes involves any potential effects asso- fide production) followed by sulfide oxida- ciated with the preservation of sedimentary tion using iron and manganese oxides. How- organic matter under anoxic conditions. AI- ever, the extent to which bacterial sulfate though in a net sense oxygen appears to be reduction can occur in the presence of "mi- the ultimate oxidant of organic matter in crobially-reducible" iron oxides is uncertain. these sediments, the specific mechanism by Lovly and Phillips (1987a) suggested that the which organic matter is remineralized is ap- presence of these oxides should inhibit bac- parently anoxic decomposition via man- terial sulfate reduction, although Sorensen ganese reduction (Fig. 12). Therefore, if there (1982) demonstrated the co-occurrence of are factors that allow (or favor) the preserva- bacterial sulfate reduction and iron reduc- tion of organic matter (or certain types of tion in coastal marine sediments. These lat- organic compounds) under anoxic conditions, ter results suggest that some of the observed then such preservation may be possible in iron reduction in Amazon shelf sediments these sediments, in spite of the net coupling could be related to a coupling of bacterial of organic matter oxidation to 0 2 reduction, sulfate reduction, sulfide oxidation and iron

In the sediments of the Amazon continen- reduction. However, net sulfate reduction tal shelf, several factors lead to a situation in appears to be of minimal importance in these which iron reduction (and to a lesser extent sediments, based on sulfate pore water gradi- manganese reduction) appears to be the pre- ents and solid phase sulfur concentrations dominant process by which sedimentary or- (Aller et al., 1986). Therefore, if iron and ganic matter is remineralized (Aller et al., manganese reduction in these sediments are 1986, 1991; Mackin et al., 1988). These fac- coupled to sulfate reduction and sulfide oxi- tors include: (1) relatively low reactive or- dation, then the complete oxidation of sul- ganic carbon concentrations; (2) relatively fide and other reduced sulfur intermediates high solid phase iron concentrations and (3) (e.g., S °, FeS, FeS 2) back to sulfate would intense physical mixing or reworking of the have to occur under "almost complete anoxic sediments. 2mpb profiles and non steady-state (no free O2)" conditions (Aller et al., 1986). modeling of pore water profiles suggest that While such sulfide oxidation reactions have the upper 50 to ~ 150 cm of these sediments been shown to occur with manganese oxides are mixed on roughly seasonal time scales, (Aller and Rude, 1988), similar reactions have and this mixing apparently "recharges" sedi- not yet been demonstrated under surficial mentary oxidants when reduced metabolic sediment conditions with iron oxides (e.g., end-products (e.g., Fe z+ or Mn 2+) are ex- see the discussion in Aller and Rude, 1988). posed to oxygenated shelf bottom waters (A1- Resolving these issues clearly has important ler et al., 1986; Mackin et al., 1988). The implications on how iron and manganese re- timing of these processes and the chemical duction are coupled to the biogeochemical characteristics of the sediments therefore ap- cycling of carbon and sulfur not only in Ama- pear to "poise" these sediments at a point zon shelf sediments, but in other sediments where iron reduction dominates organic mat- subject to similar types of physical mixing. ter remineralization. Finally, manganese and iron reduction in

At the same time, the exact mechanisms continental margin sediments may also play a by which iron and manganese reduction are role in affecting the oceanic water column coupled to organic matter oxidation in these distributions of these elements. In most sediments are still unclear. Aller et al. (1986) pelagic and many continental margin sedi- suggested that metal oxide reduction could merits, the depth scales of the biogeochemi- be directly coupled to sedimentary organic cal zones in Fig. 2 are such that very little (if matter oxidation, or could be an indirect any) of the manganese and iron produced by

Page 24: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

272 ~.J. BURDIGE

Sul fa te (raM) ~ C 0 2 (mM) M a n g a n e s e (gM) Iron (~M)

4 8 12 0 4 8 12 100 200 100 200 300

0 0 0 ~ l l k

4 4 4 4 ~-- -

st. N ~ :J: 3•8•92 ,- B S 8 ? '

K / ~, 8 , /,

'~ 12 : 12 ,,,~,, ,2 / 12

16 16 16 16

8 16 24 0 8 16 24 200 400 400 800 1200

z ~

4 4 - je

st. M ,,- 8 8 ~ 8

3•7•92 ~" : ° + cl 12 12 12 I 12

3 •

16 16 16 16

8 16 24 0 8 16 24 0 100 200 100 200 300 o ¶ o . . . . . o f l , o I . . . . .

4 4 I 4 . ~ 4 , t~

st. S o

12 12 1'2 ~ 12

16 16 16 16

Fig. 13. Pore water profiles of: dissolved sulfate, ZGO 2 (total dissolved inorganic carbon), manganese and iron, all versus depth, at three sites in the Chesapeake Bay (Burdige, Homstcad and Lustwcrk, unpub, data). Site N is in the northern Bay and has relatively low bottom water salinities (~ 3-10 psu). These pore water profiles suggest a spatial separation between surficial zones of manganese and iron reduction and a deeper zone of sulfate reduction (below ~ 4 cm). Site M is in the mcsoha[Jnc portion of the Bay (bottom water salinitcs of ~ 10-20 psu) where low oxygen or ano×ic conditions generally occur in the summer months (Smith ctal., 1992). The sediments at this site arc highly reducing and su]fidic. These observations, along with measured rates of sedimentary sulfate reduction (Marvin and Capone, 1992) suggest that much of the manganese and iron reduction in these sediments may ]ikc|y occur by sulfide oxidation. Site S is in the southern Bay (bottom water salinitics of ~ 20-30 psu) and has integrated annual rates of sulfate reduction that arc half those of the mid-Bay region (Marvin and Capone, ]992). However these sediments arc also inhabited by organisms that actively bioirrigatc and mix the sediments. Evidence for this bioirrigation and sediment mixing is based on both visual observations of macrofauna in the sediments, and on the lack of significant sulfate and ~CO 2 pore water gradients in spite of measured rates of sedimentary sulfate reduction and ~GO 2 production (Marvin and Gaponc, ]992; Burdigc, Homstcad and Lustwerk, unpubl, data). Manganese and iron pore water profiles indicate the occurrence of metal oxide reduction in site S sediments, although the mechanisms by which these processes occur are unknown at the present time.

Page 25: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 273

metal oxide reduction ultimately escapes the (Johnson et al., 1992). However, these au- sediments. In general this occurs because the thors also suggest that the sedimentary source reduced forms of these metals either precipi- of this dissolved manganese is not the reduc- tate at depth as insoluble phases (e.g., man- tive dissolution of manganese oxides, but ganese carbonates or iron sulfides), or are rather is manganese remobilized near the oxidized in the sediments upon upward diffu- sediment-water interface during the oxic sion. These solubility controls notwithstand- (aerobic) remineralization of sedimentary or- ing, the occurrence of an almost complete ganic matter and calcium carbonate dissolu- internal redox cycling of these metals is con- tion. Furthermore, measured benthic man- trolled by a balance between the thickness of ganese fluxes are largest from sediments in the "oxidized" surficial zone, the downward the shallowest (< 100 m) water depths, and fluxes of possible oxidants (e.g., O 2 diffusion), decrease with water depth as one moves to the upward fluxes of dissolved manganese sediments in water depths where the oxygen and iron and the kinetics of manganese and minimum zone in the water column impinges iron oxidation, on the California continental shelf (600-1000

The possibility does exists, however, that m). Such a distribution of benthic fluxes vs. some fraction of the dissolved manganese water depth also appears to be inconsistent and iron produced at depth during reductive with previous suggestions for how margin dissolution may escape the sediments as a sediments may act as sources of manganese benthic flux. Using manganese as an exam- to the open ocean (Martin and Knauer, 1984; pie, this could occur in a situation in which Heggie et al., 1987a; Landing and Bruland, the manganese redox boundary in the sedi- 1987). Clearly, further studies are required to ments is sufficiently close to the sediment- determine if the observations of Johnson et water interface as a result of, e.g., large al. (1992) are representative, in general, of organic carbon fluxes to the sediment or l o w other margin sediments. Such studies should bottom water oxygen concentrations. The also examine the possibility that the pro- "residence time" of the upwardly diffusing cesses responsible for this Mn 2+ production manganese in the surficial oxic sediments near the sediment-water interface are simi- could then be sufficiently short, relative to lar to the types of "oxic" manganese reduc- the time scales (or rates) of manganese oxi- tion in pelagic sediments described by Kai- dation, and lead to a "leakage" of dissolved horn and Emerson (1984) and Heggie et al. manganese through the oxic sediments to the (1986) (see The role of manganese and iron overlying waters. Several authors have sug- reduction in the oxidation of sedimentary or- gested that this situation may indeed occur ganic matter, and Figs. 5 and 6). in some continental margin sediments, and that the dissolved metals which leave the MANGANESE AND IRON REDUCTION IN sediments as a benthic flux are then trans- COASTAL, ESTUARINE, AND SALT MARSH ported by cross-shelf mixing processes to the SEDIMENTS open ocean (Martin and Knauer, 1984; Heg- gie et al., 1987a; Landing and Bruland, 1987). Coastal and estuarine sediments These authors have further suggested that this process may play an important role in Pore water profiles from coastal and estu- affecting open ocean water column distribu- arine sediments indicate the occurrence of tions of manganese and perhaps iron. manganese and iron reduction in these sedi-

Recent in situ benthic flux measurements ments (Holdren et al., 1975; Aller, 1980, 1993; in California continental margin sediments Elderfield et al., 1981; Hines et al., 1984; have indeed demonstrated the occurrence of Sundby and Silverberg, 1985; Sundby et al., manganese fluxes from these sediments 1986; S~rensen and J~rgensen, 1987; Can-

Page 26: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

274 D.J. BURDIGE

field, 1989a; Canfield et al., 1993 and others; portance of manganese reduction via ferrous also see Fig. 13). Furthermore, direct rate iron oxidation (Postma, 1985; Myers and measurements verify the occurrence of these Nealson, 1988b; Burdige et al., 1992). Fi- processes (Aller, 1980, 1993; S0rensen, 1982; nally, given the nature and intensity of mac- Lovley and Phillips, 1987b; Aller and Mackin, rofaunal activity in many of these sediments 1989; Canfield, 1989a; Canfield et al., 1993; (e.g., bioturbation), oxic and anoxic processes Lustwerk and Burdige, 1993). However, be- often appear to "co-exist" in surficial sedi- cause sulfate reduction is generally the pre- ments, since the activity of these organisms dominant pathway by which organic matter is creates "a mosaic of biogeochemical mi- oxidized in these sediments (J0rgensen, 1982; croenvironments, rather than (the) vertically Martens and Klump, 1984; Henrichs and stratified distribution" of remineralization Reeburgh, 1987; Mackin and Swider, 1989), processes that is seen in Fig. 2 (Aller, 1982). manganese and iron reduction are thought to These phenomena may then lead to a situa- play minor roles here in the oxidation of tion in which solid phase sulfides are brought sedimentary organic carbon. At the same into contact with manganese oxides, and un- time though, the rates of all of these pro- der these circumstances manganese reduc- cesses (e.g., sulfate, manganese and iron re- tion may also be coupled to the oxidation of duction as well as overall organic carbon these sulfides by chemolithotrophic bacteria remineralization) have generally not been si- (Aller and Rude, 1988; King, 1990; Canfield multaneously determined in any given sedi- et al., 1993). ment, precluding a definitive examination of As an example of how some of these pro- the relative roles of these processes in over- cesses may occur, a recent study of mana- all carbon remineralization. A recent study ganese cycling in Long Island Sound sedi- by Canfield et al. (1993) is an exception, and ments has shown that the vast majority of the in this study they observed that in certain manganese reduction occuring in these sedi- Danish coastal sediments manganese and merits appears to result from sulfide oxida- iron reduction can be significant processes tion, and that most of the Mn 2+ produced by (and in one place the only important process) this process is internally re-oxidized by the by which organic matter is remineralized, oxygen flux into the sediments (Aller, 1993).

It also seems highly likely that mechanisms Little Mn 2+ escapes these sediments, much of manganese and iron reduction other than of the sediment oxygen uptake is accounted microbial reduction coupled to organic mat- for by manganese oxidation, and overall, ter oxidation may be important in coastal manganese redox cycling appears to couple and estuarine sediments. In general, pore sedimentary sulfide oxidation to O 2 reduc- water and rate profiles indicate that iron and tion (see Fig. 12; see also Burdige and Neal- manganese reduction is localized in a region son, 1986). As discussed above, bioturbation quite close (i.e., within the upper few cm) to plays a major role in driving these processes, the sediment-water interface (e.g., see Fig. in part by transporting manganese oxides 13). Given typical dissolved sulfide and sul- and organic carbon into deeper, reduced fate reduction rate profiles in such sedi- (sulfidic) zones (Aller, 1993). Future studies ments, much of the manganese and iron re- in other estuarine and coastal sediments are duction could then be mediated by abiotic needed to continue to resolve the role(s) of chemical reactions involving dissolved sulfide manganese (and iron) reduction in all of the (Pyzik and Sommer, 1981; Burdige, 1983; redox cycles discussed above. Burdige and Nealson, 1986). Manganese and Because of the association of numerous iron pore water profiles also show a signifi- trace elements (e.g., Mo, Cu, Zn, Co, As and cant "overlap" in the zones of iron and man- the rare earth elements) with manganese and ganese reduction, suggesting the possible im- iron oxides (see The chemistry of oxidized and

Page 27: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

B I O G E ( ) C H E M I S T R Y ( )F M A N G A N E S E A N D I R O N R E D U C T I O N IN M A R I N E S E D I M E N T S 275

reduced manganese and iron in sediments), treme cases the presence of these surficial sedimentary manganese and iron redox cy- iron oxides may also actually lead to a net cling may play important roles in the cycling uptake of phosphate by estuarine sediments of these elements in nearshore environments (Boynton and Kemp, 1985; Garber, 1987; (Hines et al., 1984; Sundby et al., 1986; Burdige, 1989). This iron oxide "trapping" Westerlund et al., 1986; Sholkovitz and El- (or storage) of sedimentary phosphorus is derfield, 1988). Given the close proximity of not, however, always a permanent phenom- the zones of manganese and iron reduction ena. The development of seasonal anoxia or to the sediment-water interface, these pro- low oxygen conditions in the bottom waters cesses may then lead to benthic fluxes of of estuarine and coastal environments such these trace elements along with manganese as, e.g., the mid-Chesapeake Bay (Smith et and iron benthic fluxes (see also the previous al., 1992) can cause the reduction of surficial section). Such benthic fluxes may affect the sediment iron oxides and the release of phos- estuarine geochemistry and riverine fluxes of phate stored "on" these oxides. This then these elements (see references cited above leads temporarily to elevated phosphate and Hunt and Kelly, 1988). Similarly, many fluxes from such sediments, above those pre- of these elements are found at elevated lev- dicted by pore water profiles and molecular els in coastal waters due to man's activities, diffusion (Cornwell and Boynton, in prep.). and these concentrations may have deleteri- It should also be noted that similar inter- ous (or "toxic") effects on coastal ecosys- actions between iron oxides and phosphate terns. The cycling (and benthic fluxes) of have been shown to occur in pelagic and these elements in conjunction with sedimen- hemi-pelagic sediments (Berner, 1973; Froe- tary manganese and iron reduction may lich et al., 1977; Morse and Cook, 1978; therefore affect the extent to which coastal Filipek and Owen, 1981), and that phosphate sediments represent "permanent" reposi to- adsorbed to these oxides may account for ries for these elements. ~ 10% of the phosphorus ultimately buried

Sedimentary iron redox cycling may also in marine sediments (Froelich et al., 1982). play an important role in the dynamics of As a result of these interactions, iron reduc- phosphorus cycling in estuarine and coastal tion in pelagic and hemi-pelagic sediments ecosystems. It is well documented that a sig- also leads to "excess" phosphate production nificant fraction of the nutrients required for in sediment pore waters above that predicted primary production in these ecosystems are simply from phosphate liberation as a result supplied by the remineralization of sedimen- of organic mater oxidation (e.g., Froelich et tary organic matter and their resulting ben- al., 1979). However, most of the phosphate thic fluxes (Nixon, 1981; Klump and Martens, produced in this fashion is likely retained 1983). However, iron oxides strongly adsorb within the sediments, given the depths at dissolved phosphate and remove it from sedi- which iron reduction occurs in most pelagic ment pore waters (e.g., Krom and Berner, and hemi-pelagic sediments (i.e., the major- 1980), and the existence of such oxides near ity of the upwardly diffusing phosphate pro- the sediment-water interface can act as an duced during iron reduction is likely scav- effective trap for the upwardly diffusing enged by iron oxides found in the upper, phosphate produced by these remineraliza- oxidized portions of the sediments). Iron re- tion reactions (Krom and Berner, 1981; Cal- duction in these sediments therefore likely lender and Hammond, 1982; Sundby et al., has minimal impacts on the marine phospho- 1986, 1992; Klump and Martens, 1987). In rus cycle. part, this leads to an increased sediment

Salt marsh sediments residence time for recycled phosphorus rela- tive to recycled nitrogen or carbon (e.g., Interest in iron reduction in salt marsh Klump and Martens, 1987), and in some ex- sediments stems in part from a recognition of

Page 28: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

276 D.J. BURDIGE

the importance of sulfur redox cycling in the possibility of a more dynamic manganese re- overall biogeochemistry of salt marshes, and dox cycle in salt marsh sediments than that the apparent strong relationship between iron predicted by total solid phase concentrations and sulfur redox cycling in these environ- alone. The conceptual model presented by ments (Lord and Church, 1983; Giblin and Burdige and Nealson (1986) also indicated Howarth, 1984; Howes et al., 1984; Cutter ways that manganese redox cycling could be and Velinsky, 1988; Hines et al., 1989; Luther as important as iron redox cycling in redox et al., 1991, 1992 and others). However, while processes, in general, in salt marsh sedi- iron redox cycling in salt marshes is generally ments. All of these observations suggest that assumed to be closely linked to sulfur redox more detailed work on manganese redox cy- cycling, Luther et al. (1992) recently pre- cling in salt marsh sediments is needed to sented a conceptual model which suggested better resolve the role (and importance) of that organic ligands as well as reduced sulfur manganese reduction in these sediments. compounds act as reductants of iron (III) minerals in salt marsh sediments. In their MANGANESE AND IRON REDUCTION IN LAKE model iron reduction is presumed to be an SEDIMENTS abiotic chemical reaction, although they also suggested that bacteria a n d / o r salt marsh Since many lakes undergo seasonal strati- plants likely produce the organic compounds fication, much of the manganese and iron that act as the reductants. Recent results redox cycling in lakes occurs in the water also provide general support for this sugges- column, across the redox boundary which tion that the production of dissolved organic generally separates the epilimnion and the matter by salt marsh plants may play impor- hypolimnion (e.g., Davison, 1985). However, tant roles in salt marsh biogeochemical pro- in lakes that do not undergo such seasonal cesses. Based on the results of their studies, stratification, manganese and iron reduction Hines et al. (1989, 1993) have suggested that often occurs in the sediments (Robbins and the exudation of dissolved organic carbon Callender, 1975; Dean et al., 1981; Davison, (e.g., acetate) from the roots of salt marsh 1985; Sholkovitz, 1985; Cornwell, 1986; plants may account for a significant portion Cornwell and Kipphut, 1992; Aguillar and of the observed sulfate reduction in salt Nealson, 1993). As in other sedimentary sys- marshes during the active growing season of tems, the mechanisms of manganese and iron the plants, reduction in lake sediments are poorly char-

In contrast to our knowledge about iron acterized. While many of the mechanisms of redox cycling in salt marsh sediments, much manganese and iron reduction observed in less is known of manganese redox cycling. In marine sediments likely also occur in lake most salt marsh sediments total, solid phase sediments, the lack of significant sulfide pro- manganese concentrations are generally duction in most lacustrine systems (due to much lower (up to two orders of magnitude) low sulfate concentrations) limits the impor- than those of total iron (Lord, 1980; Nixon, tance of sulfide-mediated reduction. This 1980; Luther et al., 1991), and this observa- then suggests that much of the metal oxide tion led Luther et al. (1991) to conclude that reduction in lake sediments could be coupled manganese redox cycling does not play an to organic matter oxidation, although future important role in iron and sulfur redox cy- studies are clearly needed to verify this as- cling in these environments. However pore sumption. Manganese and iron reduction in water iron and manganese concentrations in lake sediments also appears to play impor- salt marsh sediments are often more compa- tant roles in the formation of lacustrine man- rable in magnitude (Giblin and Howarth, ganese nodules (e.g., Dean et al., 1981; 1984; Luther et al., 1991), suggesting the Moore, 1981). Furthermore, as in coastal

Page 29: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDU( ' I ' ION IN MARINE SEDIMENTS 277

sediments, iron redox cycling in lake sedi- thank Joris Gieskes for initially stimulating ments exerts a strong influence on the phos- my interest in manganese redox cycling in phorus dynamics of lake ecosystems (e.g., sediments, and Ken Nealson for his support Emerson and Widmer, 1978; Cornwell, 1987). and encouragement over the years. I also

thank the many people who sent me preprints

CONCLUSIONS and reprints of their work on manganese and iron reduction in the early stages of my work

Manganese and iron reduction in marine on this review article. Rigel Lustwerk, Greg Cutter, Juli Homstead and two anonymous

sediments play important roles in the biogeo- reviewers read earlier versions of this manu- chemical cycles of many elements. These re- script, and I thank them for their useful duction reactions affect these cycles on a suggestions. Financial support during the variety of time scales, ranging from those as short as seasonal time scales (e.g., nutrient preparation of this manuscript was provided

by the National Science Foundation, the Pc- cycling in coastal ecosystems), to those as troleum Research Fund of the American long as thousands to tens of thousands of

Chemical Society and the Toxics Research years (e.g., glacial-interglacial transforma-

Program of the Chesapeake Bay Environ- tions). In this review I have attempted to

mental Effects Committee. summarize the results of studies of these processes, indicate the relationships of man- ganese and iron reduction to other biogeo- chemical processes in sediments, and discuss REFERENCES what I believe are important areas of future research. In general, previous studies o f Aguillar, C. and Nealson, K.H., 1993. Manganese re- manganese and iron reduction in sediments duction in Oneida Lake, N.Y.: Estimates of spatial have taken two approaches. The first has and temporal manganese flux. Can. J. Fish. Aquat.

involved studies aimed at describing the Sci., in press. Aller, R.C., 1980. Diagenetic processes near the sedi-

mechanisms of these processes, while the ment-water interface of Long Island Sound, II. Fe second has involved studies examining the and Mn. Adv. Geophys., 22: 351-415. geochemistry of these processes. Future work Aller, R.C., 1982. The effects of macrobenthos on should emphasize the "inter-weaving" of chemical properties of marine sediments and overly-

these two approaches, thereby providing bet- ing water. In: P.L. McCall and M.J.S. Tevesz (Edi- tors), Animal-Sediment Relations, The Biogenic

ter information not only on the mechanisms Alteration of Sediments. Plenum Press, New York, of, and the controls on, sedimentary man- pp. 53-102. ganese and iron reduction, but also on how Aller, R.C., 1990. Bioturbation and manganese cycling these reduction reactions are coupled to at the sediment-water interface. Philos. Trans. R.

Soc. London, A 331: 51-68. other sedimentary biogeochemical processes. Such information can then be used in devel- Aller, R.C., 1993. The sedimentary Mn cycle in Long

Island Sound: its role as intermediate oxidant and oping m o r e a c c u r a t e and predictive models the influence of bioturbation, 0 2, and C,~rg flux on for systems in which sedimentary manganese diagenetic reaction balances. J. Mar. Res., in press. and iron reduction o c c u r . Aller, R.C. and Rude, P.D., 1988. Complete oxidation

of solid phase sulfides by manganese and bacteria in anoxic marine sediments. Geochim. Cosmochim.

ACKNOWLEDGEMENTS Acta, 52: 751-765. Aller, R.C. and Mackin, J.E., 1989. Open-incubation,

Many of the ideas presented here resulted diffusion method for measuring solute reaction rates in sediments. J. Mar. Res., 47: 411-440.

from discussions over the years with n u m e r - Aller, R.C., Mackin, J.E. and Knox Jr, R.T., 1986. ous friends and colleagues, and to all of Diagenesis of Fe and S in Amazon inner shelf them I offer many thanks. In particular, I muds: apparent dominance of Fe reduction and

Page 30: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

278 D.J. BURDIGE

implications for the genesis of ironstones. Cont. Boynton, W.R. and Kemp, W.M., 1985. Nutrient re- Shelf Res., 6: 263-289. generation and oxygen consumption by sediments

Aller, R.C., Aller, J.Y., Blair, N.E., Mackin, J.E., Rude, along an estuarine salinity gradient. Mar. Ecol. Prog. P.D., Stupakoff, I., Patchineelam, S., Boehme, S.E. Ser., 23: 45-55. and Knoppers, B., 1991. Biogeochemical processes Buckley, D.E. and Cranston, R.E., 1988. Early diagen- in Amazon shelf sediments. Oceanography, 4: 27-32. esis in deep sea turbidites: the imprint of paleo-

Arnold, R.G., DiChristina, T.J. and Hoffmann, M.R., oxidation zones. Geochim. Cosmochim. Acta, 52: 1988. Reductive dissolution of Fe(III) oxides by 2925-2939. Pseudomonas sp, 20(I. Biotechnol. Bioeng., 32: Burdige, D,J., 1983. The Biogeochemistry of Man- 1081-1096. ganese Redox Reactions: Rates and Mechanisms.

Balistrieri, L.S. and Murray, J.W., 1986. The surface Ph.D. Diss., Scripps Inst. Oceanol., Univ. Calif., San chemistry of sediments from the Panama Basin: The Diego, 251 pp. influence of Mn oxides on metal adsorption. Burdige, D.J., 1989. 1988 Sediment Monitoring Pro- Geochim. Cosmochim. Acta, 50: 235-2243. gram for the Southern Chesapeake Bay. O D U R F

Bartlett, R.J., 1981. Nonmicrobial nitrite-to-nitrate Tech. Rep., 89-6. transformations in soils. Soil Sci. Am. J., 45: 1054- Burdige, D.J., 1991. The kinetics of organic matter 1058, mineralization in anoxic marine sediments. J. Mar.

Bazylinski, D.A.~ 1990. Anaerobic production of Res., 49: 727-761. single-domain magnetite by the marine, magneto- Burdige, D.J. and Gieskes, J.M., 1983. A porewater / tactic bacterium, strain MV-1. In: R.B. Frankel and solid phase diagenetic model for manganese in ma- R.P. Blakemore (Editors), Iron Biominerals. Plenum rine sediments. Am. J. Sci., 283: 29-47. Press, New York, pp. 69-77. Burdige, D.J. and Nealson, K.H., 1986. Chemical and

Bell, P.E., Mills, A.L. and Herman, J.S., 1987. Biogeo- microbiological studies of sulfide-mediated man- chemical conditions favoring magnetite formation ganese reduction. Geomicrobiol. J., 4: 361-387. during anaerobic iron reduction. Appl. Environ. Mi- Burdige, D.J., Dhakar, S.P. and Nealson, K.H., 1992. crobiol., 53: 2610-2616. The role of manganese oxide mineralogy on micro-

Bender, M.L. and Heggie, D.T., 1984. Fate of organic bial and chemical manganese reduction. Geomicro- carbon reaching the deep sea floor: a status report, biol. J., 10: 27-48. Geochim. Cosmochim. Acta, 48: 977-986. Buresh, R.J. and Moraghan, J.T., 1976. Chemical re-

Bender, M., Jahnke, R., Weiss, R., Martin, W., Heggie, duction of nitrate by ferrous iron. J. Environ. Qual., D.T., Orchardo, J. and Sowers, T., 1989. Organic 5: 320-325. carbon oxidation and benthic nitrogen and silica Burns, R.G. and Burns, V.M., 1977. Mineralogy. In: dynamics in San Clemcnte Basin, a continental bor- G.P. Glasby (Editor), Marine Manganese Deposits. derland site. Geochim. Cosmochim. Acta, 53: 685- ch. 7. Elsevier, Amsterdam, pp. 185-248. 697. Burns, R.G. and Burns, V.M., 1979. Manganese miner-

Berger, W.H., Finkel, R.C., Killingley, J.S. and als. In: R.G. Burns (Editor), Marine Minerals. MSA Marchig, V., 1983. Glacial-Holocene transition in Short Course Notes, 6: 1-46. deep-sea sediments: manganese-spike in the east- Burns, R.G. and V.M. Burns, 1981. Authigenic oxides. equatorial Pacific. Nature, 303: 231-233. In: C, Emiliani (Editor), The Sea, 7. The Oceanic

Berner, R.A., 1973. Phosphate removal from sea-water Lithosphere. Wiley-Interscience, New York, pp. by absorption on volcanogenic ferric oxides. Earth 875-914. Planet. Sci. Lett., 18: 77-86. Callender, E. and Hammond, D.E., 1982. Nutrient

Berner, R.A., 1980. Early Diagenesis, A Theoretical exchange across the sediment-water interface in Approach, Princeton University Press, 241 pp. the Potomac river estuary. Estuarine Coastal Shelf

Berner, R.A., 1982. Burial of organic carbon and pyrite Sci. 15: 395-413. sulfur in the modern ocean: its geochemical and Calvert, S.E. and Price, N.B., 1977. Geochemical varia- environmental significance. Am. J. Sci., 282: 451- tion in ferromanganese nodules and associated sedi- 473. ments from the Pacific Ocean. Mar. Chem., 5: 43-

Berner, R.A., 1984. Sedimentary pyrite formation: an 74. update. Geochim. Cosmochim. Acta, 48: 605-615. Canfield, D.E., 1989a. Reactive iron in marine sedi-

Blakemore, R.P., 1975. Magnetotactic bacteria. Sci- ments. Geochim. Cosmochim. Acta, 53: 619-632. ence, 190: 377-379. Canfield, D.E., 1989h. Sulfate reduction and oxic respi-

Bonatti, E., Fisher, D.E., Joensuu, O. and Rydell, H.S., ration in marine sediments: implications for organic 1971. Postdepositional mobility of some transition carbon preservation in euxinic environments. Deep- elements, phosphorus, uranium and thorium in deep Sea Res., 36: 121-138. sea sediments, Geochim. Cosmochim. Acta, 35: Canfield, D.E., 1993. Organic matter oxidation in ma- 189-201. rine sediments. In: R. Wollast, L. Chou and F.

Page 31: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 279

Mackenzie (Editors), Interactions of C, N, P and S control of mineralogical and chemical composition Biogeochemical Cycles. NATO-ARW. Springer, by multiple accretionary processes. Geochim. Cos- Berlin, pp. 333-363. mochim. Acta, 48: 1913-1928.

Canfield, D.E. and Berner, R.A., 1987. Dissolution and Ehrlich, H.L., 1987. Manganese oxide reduction as a pyritization of magnetite in anoxic marine sedi- form of anaerobic respiration. Geomicrobiol. J., 5: ments. Geochim. Cosmochim. Acta, 51: 645-659. 423-431.

Canfield, D.E., Thamdrup, B. and Hansen, J.W., 1993. Elderfield, H., 1981. Metal-organic associations in in- The anaerobic degradation of organic matter in terstitial waters of Narragansett Bay sediments. Am. Danish coastal sediments: Fe reduction, Mn reduc- J. Sci., 281: 1184-1196. tion, and sulfate reduction. Geochim. Cosmochim. Elderfield, H., McCafrey, R.J., Luedtke, N., Bender, Acta, 57: 3867-3883. M. and Truesdale, V.W., 1981. Chemical diagenesis

Christensen, J.P., 1989. Sulfate reduction and carbon in Narragansett Bay sediments. Am. J. Sci., 281: oxidation rates in continental shelf sediments, an 1021-1055. examination of offshelf carbon transport. Cont. Shelf Emerson, S., 1985. Organic carbon preservation in Res., 9: 223-246. marine sediments. In: E.T. Sundquist and W.S.

Claypool, G.E. and Kaplan, I.R., 1974. The origin and Broecker (Editors), The Carbon Cycle and Atmo- distribution of methane in marine sediments. In: spheric CO~: Natural Variations Archaen to Pre- I.R. Kaplan (Editor), Natural Gases in Marine Sedi- sent. AGU Geophys. Monogr., 32: 78-87. merits. Plenum Press, New York, pp. 99-139. Emerson, S, and Widmer, G., 1978. Early diagenesis in

Colley, S., Thomson, J., Wilson, T.R.S. and Higgs, anaerobic lake sediments II. Thermodyamic and N.C., 1984. Post-depositional migration of elements kinetic factors controlling thc formation of iron during diagenesis in brown clay and turbite se- phosphate. Geochim. Cosmochim. Acta, 42: 1307- quences in the north east Atlantic. Geochim. Cos- 1316. mochim. Acta, 48: 1223-1235. Emerson, S, and Hedges, J.l., 1988. Processes control-

Cornwell, J.C., 1986. Diagenetic trace-metal profiles in ling the organic carbon content of open ocean sedi- Arctic lake sediments. Environ. Sci, Technol., 20: ments. Paleoccanography, 3: 621-634. 299-302. Emerson, S., Kalhorn, S., Jacobs, L., Tebo, B.M., Neal-

Cornwell, J.C., 1987. Phosphorus cycling in arctic lake son, K.H. and Rosson, R.A., 1982. Environmental sediments: adsorption and authigenic minerals, oxidation rate of manganese(II): bacterial catalysis. Arch. Hydrobiol., 109: 161-179. Geochim. Cosmochim. Acta, 46: 1073-1079.

Cornwell, J.C. and Kipphut, G.W., 1992. Biogeochem- Emerson, S., Fischer, K., Reimers, C. and Heggie, D., istry of manganese- and iron-rich sediments in Too- 1985. Organic carbon dynamics and preservation in Ilk Lake. Alaska. Hydrobiol., 240: 45-60. deep-sea sediments. Deep-Sea Res., 32: 1-21.

Cutter, G.A. and Velinsky, D.J., 1988. Temporal varia- Filipek, L.H. and Owen, R.M., 1981. Diagenetic con- tions in sedimentary sulfur in a Delaware salt marsh, trois on phosphorus in outer continental shelf sedi- Mar. Chem., 23: 311-327. ments from the Gulf of Mexico. Chem. Geol., 33:

Davison, W., 1985. Conceptual models for transport at 181-204. a redox boundary. In: W. Stumm (Editor), Chemical Finney, B.P., Lyle, M.W. and Heath, G.R., 1988. Sedi- Processes in Lakes. Wiley, Chichester, pp. 31-54. mentation at MANOP site H (eastern Equatorial

Dean, W.E., Moore, W.S. and Nealson, K.H., 1981. Pacific) over the past 400,000 years: climatically Manganese cycles and the origin of manganese nod- induced redox variations and their effects on transi- ules, Oneida Lake, New York, USA. Chem. Geol., tion metal cycling. Paleoceanography, 3: 169-189. 34: 53-64. Froelich, P.N., Bender, M.L. and Heath, G.R., 1977.

Dean, W.E., Gardner, J.V. and Hemphill-Haley, E., Phosphorus accumulation rates in metalliferous sed- 1989. Changes in redox conditions in deep-sea sedi- iments on the East Pacific Rise. Earth Planet. Sci. ments of the subarctic north Pacific Ocean: Possible Lett., 34: 351-359. evidence for the presence of north Pacific deep Froelich, P.N., Klinkhammer, G.P., Bender, M.L., water. Paleoceanography, 4: 639-653. Luedtke, N.A., Heath, G.R., Cullen, D., Dauphin,

DeLange, G.J., 1986. Early diagenetic reactions in in- P., Hammond, D., Hartman, B. and Maynard, V., terbedded pelagic and turbiditic sediments in the 1979. Early oxidation of organic matter in pelagic nares abyssal plain (western north Atlantic): conse- sediments of the eastern equatorial Atlantic: sub- quences for the composition of sediment and inter- oxic diagenesis. Geochim. Cosmochim. Acta, 43: stitial water. Geochim. Cosmochim. Acta, 50: 2543- 1075-1090. 2561. Froelich, P.N., Bender, M.L., Luedtke, N.A., Heath,

Dymond, J., Lyle, M., Finney, B., Piper, D.Z., Murphy, G.R. and DeVries, T.A., 1982. The marine phos- K., Conrad, R, an'd Pisias, N,, 1984. Ferroman- phorus cycle. Am. J. Sci., 282: 474-511. ganese nodules from MANOP sites H, S, and R - - Garber, J., 1987. Benthic-pelagic coupling in the

Page 32: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

280 D.J. BURDIGE

Chesapeake Bay. In: M.P. Lynch and E.C. Krome Henrichs, S.M. and Reeburgh, W.S., 1987. Anaerobic (Editors), Perspectives on the Chesapeake Bay: Ad- mineralization of marine sediment organic matter: vances in Estuarine Sciences. US EPA Chesapeake rates and the role of anaerobic processes in the Bay Progam CBP/TRS 16/87, pp. 11-34. oceanic carbon economy. Geomicrobiol. J., 5: 191-

Gardner, J.V., Dean, W.E., Klise, D.H. and Baldauf, 237. J.G., 1982. A climate-related oxidizing event in Hines, M.E., Lyons, Wm,B., Armstrong, P.A., Orem, deep-sea sediments from the Bering Sea. Quat. W.H., Spencer, M.J. and Gaudette, H.E., 1984. Res., 107: 91-107. Seasonal metal remobilization in the sediments of

Garrels, R.M. and Lerman, A., 1984. Coupling of the Great Bay, New Hampshire. Mar. Chem., 15: 173- sedimentary sulfur and carbon cycles--an improved 187. model. Am. J. Sci., 284: 989-1007. Hines, M.E., Knollmeyer, S.L. and Tugel, J.B., 1989.

Ghiorse, W.C., 1988. Microbial reduction of iron and Sulfate reduction and other sedimentary biogeo- manganese. In: A.J.B. Zehnder (Editor), Biology of chemistry in a northern New England salt marsh. Anaerobic Microorganisms. Wiley, Chichester, pp. Limnol. Oceanogr., 34: 578-590. 305-331. Hines, M.E., Bazylinski, D.A., Tugel, J.B. and Lyons,

Giblin, A.E. and Howarth, R.W., 1984. Porewater evi- W.B., 1991. Anaerobic microbial biogeochemistry in dence for a dynamic sedimentary iron cycle in salt sediments from two basins in the Gulf of Maine: marshes. Limnol. Oceanogr., 29: 47-63. evidence for iron and manganese reduction. Estuar-

Glasby, G.P., 1984. Manganese in the marine environ- ine Coastal Shelf Sci., 33: 313-324. ment. Oceanogr. Mar. Biol. Annu. Rev., 22: 169- Hines, M.E., Banta, G.T., Giblin, A.E., Hobbie, J.E. 194. and Tugel, J.B., 1993. Acetate concentrations and

Glausinger, W.S., Horowitz, H.S. and Longo, J.M., oxidation in salt marsh sediments. Limnol. Oceano- 1979. Preparation and magnetic properties of man- gr., in press. ganite. J. Solid State Chem., 29: 117-120. Holdren, G.R., Jr., Bricker III, O.P. and Matisoff, G.,

Gratton, Y., Edenborn, H.E., Silverberg, N. and Sund- 1975. A model for the control of dissolved man- by, B., 1990. A mathematical model for manganese ganese in the interstitial water of Chesapeake Bay. diagenesis in bioturbated sediments. Am. J, Sci., In: T.M. Church (Editor), Marine Chemistry in the 290: 246-262. Coastal Environment. ACS Sym. Ser., 18, pp. 364-

Grill, E.V., 1978. The effects of sediment-water ex- 381. change on manganese deposition and nodule growth Hong, H. and Kester, D.R., 1986. Redox state of iron in Jervis Inlet, British Columbia. Geochim. Cos- in offshore waters of Peru. Limnol. Oceanogr., 31: mochim. Acta, 42: 485-494. 512-524.

Heggie, D., Kahn, D. and Fischer, K., 1986. Trace Howes, B.L., Dacey, J.H. and King, G.M., 1984. Car- metals in metalliferous sediments, MANOP Site M: bon flow through oxygen and sulfate reduction interfacial pore water profiles. Earth Planet. Sci. pathways in salt marsh sediments. Limnol. Oceano- Lett., 80: 106-116. gr., 29: 1037-1051.

Hcggie, D., Klinkhammer, G. and Cullen, D., 1987a. Hunt, C.D. and Kelly, J.R., 1988. Manganese cycling in Manganese and copper fluxes from continental mar- coastal regions: response to eutrophication. Estuar- gin sediments. Geochim. Cosmochim. Acta, 51: ine Costal Shelf Sci., 26: 527-558. 1059-1070. Jahnke, R.A., 1990. Early diagenesis and recycling of

Heggie, D., Maris, C., Hudson, A., Dymond, J., Beach, biogenic debris at the seafloor, Santa Monica Basin, R. and Cullen, J., 1987b. Organic carbon oxidation California. J. Mar. Res., 48: 413-436. and preservation in NW Atlantic continental margin Jahnke, R., Heggie, D., Emerson, S. and Grundmanis, sediments. In: P.P.E. Weaver and J. Thomson (Edi- V., 1982. Pore waters of the central Pacific Ocean: tors), Geology and Geochemistry of Abyssal Plains. nutrient results. Earth Planet. Sci. Lett., 61: 233- Blackwell, Oxford, pp. 215-236. 256.

Hem, J.D. and Lind, C.J., 1983. Nonequilibrium rood- Jahnke, R.A., Emerson, S.R., Reimers, C.E., Schuffert, els for predicting forms of precipitated manganese J., Ruttenberg, K. and Archer, D., 1989. Benthic oxides. Geochim. Cosmochim. Acta, 47: 2037. recycling of biogenic debris in the eastern tropical

Hem, J.D., Robertson, C.E. and Fournier, R.B., 1982. Atlantic Ocean. Geochim. Cosmochim. Acta, 53: Stability of /3MnOOH and manganese oxide depo- 2947-2960. sition from springwater. Water Resour. Res., 18: Jenne, E.A., 1968. Controls on Mn, Fe, Co, Ni, Cu and 563-570. Zn concentrations in soils and water: the significant

Henrichs, S.M., 1992. Early diagenesis of organic mat- role of hydrous manganese and iron oxides. In: tcr in marine sediments: progress and perplexity. Trace Inorganics in Water. ACS Adv. Chem. Set., Mar. Chem., 39: 119-149. 73.

Page 33: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

B I O G E O C H E M I S T R Y O F M A N G A N E S E A N D I R O N R E D U C T I O N IN M A R I N E S E D I M E N T S 281

Johnson, K.S., Berelson, W.M., Coale, K.H., Coley, Lindberg, R.D. and Runnells, D.D., 1984. Ground T.L., Elrod, V.A., Farley, W.R., Iams, H.D., Kil- water redox reactions: an anlysis of equilibrium gore, T.E. and Nowicki, J.L., 1992. Manganese flux state applied to Eh measurements and geochemical from continental margin sediments in a transect modeling. Science, 31: 925-927. through the oxygen minimum. Science, 257: 1242- Lord, C.J. III., 1980. The chemistry and cycling of iron, 1245. maganese and sulfur in salt marsh sediments. PhD

J0rgensen, B.B., 1982. Mineralization of organic matter Diss., Univ. Delaware, 177 pp. in the sea b e d - - t h e role of sulphate reduction. Lord, C.J. III and Church, T.M., 1983. The geochem- Nature, 296: 643-645. istry of salt marshes: sedimentary diffusion, sulfate

Kalhorn, S. and S. Emerson, 1984. The oxidation state reduction and pyritization. Geochim. Cosmochim. of manganese in surface sediments of the deep sea. Acta, 47: 1381-1391. Geochim. Cosmochim. Acta, 48: 897-902. Lovley, D.R., 1987. Organic matter remineralization

Karlin, R., Lyle, M. and Heath, G.R., 1987. Authigenic with the reduction of ferric iron: a review. Geomi- magnetite formation in suboxic sediments. Nature, crobiol. J., 5: 375-399. 326: 490-493. Lovley, D.R., 1990. Magnetite formation during micro-

Kessick, M.A. and Thomson, B.M., 1974. Reactions bial dissimilatory iron reduction. In: R.B. Frankel between manganese dioxide and aqueous sulfide, and R.P. Blakemore (Editors), Iron Biominerals. Environ. Lett., 7: 163-173. Plenum Press, New York, pp. 151-166.

King, G.M., 1990. Effects of added manganic and ferric Lovley, D.R., 1991. Dissimilatory Fe(III) and Mn(IV) oxides on sulfate reduction and sulfide oxidation in reduction. Microbiol. Rev., 55: 259-287. intertidal sediments. FEMS Microbiol. Ecol., 73: Lovley, D.R. and Phillips, E.J.P., 1986. Organic matter 131-138. rcmineralization with reduction of ferric iron in

Klinkhammer, G.P., 1980. Early diagenesis in sedi- anaerobic sediments. Appl. Environ. Microbiol., 51: ments from the eastern equatorial Pacific, II. Pore 683-689. water metal results. Earth Planet. Sci. Lett., 49: Lovley, D.R. and Phillips, E.J.P., 1987a. Competitive 81-101. mechanisms for inhibition of sulfate reduction and

Klump, J.V. and Martens, C.S., 1983. Benthic nitrogen methane production in the zone of ferric iron re- regeneration. In: E.J. Carpenter and D.G. Capone duction in sediments. Appl. Environ. Microbiol., 53: (Editors), Nitrogen in the Marine Environment. 2636-2641. Academic Press, San Francisco, pp. 411-457. Lovley, D.R. and Phillips, E.J.P., 1987b. Rapid assay

Klump, J.V. and Martens, C.S., 1987. Biogeochemical for microbially available ferric iron in aquatic sedi- cycling in an organic-rich marine basin--5. Sedi- ments. Appl. Environ. Microbiol., 53: 1536-1540. mentary nitrogen and phosphorus budgets based Lovley, D.R. and Phillips, E.J.P., 1988. Novel mode of upon kinetic models, mass balances, and the stoi- microbial energy metabolism: organic carbon oxida- chiometry of nutrient regeneration. Geochim. Cos- tion coupled to dissimilatory reduction of iron and mochim. Acta, 51: 1161-1173. manganese. Appl. Environ. Microbiol., 54: 1472-

Krom, M.D. and Berner, R.A., 1980. Adsorption of 1480. phosphate in anoxic marine sediments. Limnol. Lustwerk, R.L. and Burdige, D.J., 1993. Iron and man- Oceanogr., 25: 797-806. ganese reduction in Chesapeake Bay sediments.

Krom, M.D. and Berner, R.A., 1981. The diagenesis of Trans. Am. Geophys. Union (EOS), 74:326 (abstr.). phosphorus in a nearshore marine sediment. Luther, G.W., III, Ferdelmen, T.G., Kostka, J.E., Geochim. Cosmochim. Acta, 45: 207-216. Tsamakis, E.J. and Church, T.M., 1991. Temporal

LaKind, J.S. and Stone, A.T., 1989. Reductive dissolu- and spatial variability of reduced sulfur species and tion of goethite by phenolic reductants. Geochim. porewater parameters in salt marsh sediments. Bio- Cosmochim. Acta, 53: 961-971. geochemistry, 14: 57-88.

Landing, W.M. and Bruland, K.W., 1987. The contrast- Luther, G.W., Kostka, J.E., Church, T.M., Sulzberger, ing biogeochemistry of manganese and iron in the B. and Stumm, W., 1992. Seasonal iron cycling in Pacific Ocean. Geochim. Cosmochim. Acta, 51: 29- the salt marsh sedimentary environment: the impor- 43. tance of ligand complexes with Fe(II) and Fe(III) in

Lee, C., 1992. Controls on organic carbon preservation: the dissolution of Fe(III) minerals and pyrite, re- The use of stratified water bodies to compare in- spectively. Mar. Chem., 40: 81-103. trinsic rates of decompositin in oxic and anoxic Luther, G.W. III, Nuzzio, D.B. and Wu, J., 1993. systems. Geochim. Cosmochim. Acta, 56: 3223- Speciation of manganese in Chesapeake Bay waters 3335. by voltammetric methods. Anal. Chim. Acta, in

Li, Y.H., Bischoff, J. and Mathieu, G., 1969. The press. migration of manganese in Artic Basin sediment. Lyle, M., 1983. The brown-green color transition in Earth Planet. Sci. Lett., 7: 265.

Page 34: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

282 D.J. BURDIGE

marine sediments: A marker of the Fe(lI1)-Fe(II) Moore, W.S., Ku, T.-L., MacDougall, J.D., Burns, redox boundary. Limnol. Oceanogr., 28: 1026-1033. V.M., Burns, R., Dymond, J., Lyle, M.W. and Piper,

Lyle, M., Murray, D.W., Finney, B.P., Dymond, J., D.Z., 1981. Fluxes of metals to a manganese nod- Robbins, J.M. and Brooksforce, K., 1988. The record ule: radiochemical, chemical, structural, and miner- of late Pleistocene biogenic sedimentation in the alogical studies. Earth Planet. Sci. Lett., 52: 151- eastern tropical Pacific Ocean. Paleoceanography, 171. 3: 39-59. Morgan, J.J., 1967. Chemical equilibria and kinetic

Lynn, D.C. and Bonatti, E., 1965. Mobility of man- properties of manganese in natural waters. In: S.D. ganese in diagenesis of deep sea sediments. Mar, Faust and J.V. Hunter (Editors), Principles and Geol., 3: 457. Applications of Water Chemistry. Wiley, Chich-

Mackin, J.E. and Swider, K.T., 1989. Organic matter ester, pp. 561-626. decomposition pathways and oxygen consumption in Morse, J.W. and Cook, N., 1978. The distribution and coastal marine sediments. J. Mar. Res., 47: 681-716. form of phosphorus in North Atlantic Ocean deep-

Mackin, J.E., Aller, R.C. and Ullman, W.J., 1988. The sea and continental slope sediments. Limnol. Ocea- effects of iron reduction and nonsteady-state diage- nogr., 23: 825-830. nesis on iodine, ammonium, and boron distributions Murray, J. and lrvine, R., 1895. On the manganese in sediments from the Amazon continental shelf, oxide and manganese nodules in marine deposits. Cont. Shelf Res., 8: 363-386. Trans. R. Soc. Edinburgh, 37: 721.

Mangini, A., Eisenhauer, A. and Walter, P., 1 9 9 0 . Murray, J.W., 1975. The interactions of metal ions at Response of manganese in the ocean to the climatic the manganese dioxide solution interface. Geochim. cycles in the Quaternary. Paleoceanography, 5: Cosmochim Acta, 39: 505-519. 811-821. Murray, J.W., 1979. Iron Oxides. In: R.G. Burns (Edi-

Martens, C.S. and Klump, J.V., 1984. Biogeochemical tor), Marine Minerals. MSA Short Course Notes, 6: cycling in an organic-rich marine basin--4. An or- 47-98. ganic carbon budget for sediments dominated by Murray, J.W. and Kuivala, K.M., 1990. Organic matter sulfate reduction and methanogenesis. Geochim. diagenesis in the northeast Pacific: transition from Cosmochim. Acta, 48: 1987-2004. aerobic red clay to suboxic hemipelagic sediments.

Martens, C.S., Berner, R.A. and Rosenfeld, J.K., 1978. Deep-Sea Res., 37: 59-80. Interstitial water chemistry of anoxic Long Island Murray, J.W., Balistrieri, L.S. and Paul, B., 1984. The Sound sediments. 2. Nutrient regeneration and oxidation state of manganese in marine sediments phosphate removal. Limnol. Oceanogr. 23: 605-617. and ferromanganese nodules. Geochim. Cos-

Martin, J.H. and Knauer, G.A., 1984. VERTEX: man- mochim. Acta, 48: 1237-1247. ganese transport through oxygen minima. Earth Murray, J.W., Dillard, J.G., Giovanoli, R., Moers, H. Planet. Sci. Lett., 67: 35-47. and Stumm, W., 1985. Oxidation of Mn(lI): Initial

Marvin, M.C. and Capone, D.G., 1992. Benthic micro- mineralogy, oxidation and ageing. Geochim. Cos- bial metabolism along the Chesapeake estuarine mochim. Acta, 49: 463-470. transect, abstr, presented at the 1992 ASLO Aquatic Myers, C.R. and Nealson, K.H., 1988a. Bacterial man- Sciences meeting, Santa Fe, NM. ganese reduction and growth with manganese oxide

Michard, G., 1971. Theoretical model for manganese as the sole electron acceptor. Science, 240: 1319- distribution in calcareouse sediment cores. J. Geo- 1321. phys. Res.. 76: 2179-2186. Myers, C.R. and Nealson, K.H., 1988b. Microbial re-

Middelburg, J.J., 1989. A simple rate model for organic duction of manganese oxides: Interactions with iron matter decomposition in marine sediments. Geo- and sulfur. Geochim. Cosmochim. Acta, 52: 2727- chim. Cosmochim. Acta, 53: 1577-1581. 2732.

Middelburg, J.J., deLange, G.J. and van der Weijden, Nealson, K.H. and Myers, C.R., 1992. Microbial reduc- C.H., 1987. Manganese solubility control in marine tion of manganese and iron: new approaches to pore waters. Geochim. Cosmochim. Aeta, 51: 759. carbon cycling. Appl. Environ. Microbiol., 58: 439-

Millward, G.E. and Moore, R.M., 1982. The adsorp- 443. tion of Cu, Mn and Zn by iron oxyhdroxide in Nixon, S.W., 1980. Between coastal marshes and coastal model estuarine solutions. Water Res., 16: 981-985. waters - -a review of twenty years of speculation and

Moore, W.S., 1981. I ron-manganesebanding in Oneida research on the role of salt marshes in estuarine Lake ferromanganese nodules. Nature, 292: 233- productivity and water chemsitry. In: P. Hamilton 235. and K.B. MacDonald (Editors), Estuarine and Wet-

Moore, W.S., Dean, W.E., Krishnaswami, S. and Bo- land Processes. Plenum, New York, pp. 437-525. role, D.V., 1980. Growth rates of manganese nod- Nixon, S.W., 1981. Remineralization and nutrient cy- ules in Oneida Lake, New York. Earth Planet. Sci. cling in coastal marine ecosystems. In: B.J. Neilson Lett., 46: 191-20/).

Page 35: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

BIOGEOCHEMISTRY OF MANGANESE AND IRON REDUCTION IN MARINE SEDIMENTS 283

and L.E. Cronin (Editors), Estuaries and Nutrients. mobilization in the interstitial waters of red clay and Humana Press, Clifton, N.J. pp. 111-138. hemi-pelagic sediments. Earth Planet. Sci. Lett., 64:

Open University Course Team, 1989. Ocean Chemistry 213-230. and Deep-Sea Sediments. Pergamon, Oxford, 134 Sayles, F.L. and Curry, W.B., 1988. ~13C, TCO 2 and pp. the metabolism of organic carbon in deep sea sedi-

Pedersen, T.F., 1983. Increased productivity in the ments. Geochim. Cosmochim. Acta, 52: 2963-2978. eastern equatorial Pacific during the last glacial Schimmield, G.B. and Pedersen, T.F., 1990. The geo- maximum (19,000 to 14,000 yr B.P.). Geology, 11: chemistry of reactive trace metals and halogens in 16-19. hemipelagic continental margin sediments. Rev.

Pedersen, T.F. and Price, N.B., 1982. The geochem- Aquatic Sci., 3: 255-279. istry of manganese carbonate in Panama Basin sedi- Shaw, T.J., Gieskes, J.M. and Jahnke, R.A., 1990. ments. Geochim. Cosmochim. Acta, 46: 49. Early diagenesis in differing depositional environ-

Pedersen, T.F. and Calvert, S.E., 1990. Anoxia vs. ments: The response of transition metals in pore productivity: What controls the formation of or- waters. Geochim. Cosmochim. Acta, 54: 1233-1246. ganic-carbon-rich sediments and sedimentary rocks? Sholkovitz, E.R., 1985. Redox-related geochemistry in AAPG Bull., 74: 454-466. lakes: alkali metals, alkaline-earth elements and

Pedersen, T.F., Vogel, J.S. and Southon, J.R., 1986. 137Cs. In: W. Stumm (Editor), Chemical Processes Copper and manganese in hemipelagic sediments at in Lakes. Wiley, Chichester, pp. 119-142. 21°N, East Pacific Rise: Diagenetic contrasts. Sholkovitz, E.R. and Elderfield, H., 1988. Cycling of Geochim. Cosmochim. Acta, 50: 2019-2031. dissolved rare earth elements in Chesapeake Bay.

Postma, D., 1982. Pyrite and siderite formation in Global Biogeochem. Cycles, 2: 157-176. brackish and freshwater swamp sediments. Am. J. Smith, D.E., Leffler, M. and Mackiernan, G., 1992. Sci., 282: 1151-1183. Oxygen Dynamics in the Chesapeake Bay, A Syn-

Postma, D., 1985. Concentration of Mn and separation thesis of Recent Research. Maryland Sea Grant from Fe in sediments - - I. Kinetics and stochiome- College, College Park, MD, 234 pp. try of the reaction between birnessite and dissolved Sorem, R.K. and Fewkes, R.H., 1977. Internal charac- Fe(II) at 10 ° C. Geochim. Cosmochim. Acta, 49: teristics. In: G.P. Glasby (Editor), Marine Man- 1023-1033. ganese Deposits. Ch. 6. Elsevier, Amsterdam, pp.

Price, B.A., 1988. Equatorial Pacific Sediments: A 147-184. Chemical Approach to Ocean History. Ph.D. Diss., S0rensen, J., 1982. Reduction of ferric iron in anaero- Scripps Inst. Oceanol., UCSD, 364 pp. bic, marine sediment and interaction with reduction

Price, B.A. and Froelich, P.N., 1987. Late Holocene of nitrate and sulfate. Appl. Environ. Microbiol., 43: redox changes in eastern equatorial Pacific and 319-324. Atlantic sediments: A time scale for recent in- Scrensen, J., 1987. Nitrate reduction in marine sedi- creases in equatorial upwelling. Trans. Am. Geo- ment: pathways and interactions with iron and sul- phys. Union (EOS), 68:1330 (abstr.). fur cycling. Geomicrobiol. J., 5: 401-421.

Pyzik, A.J. and Sommer, S.E., 1981. Sedimentary iron S0rensen, J. and J0rgensen, B.B., 1987. Early diagene- monosulfides: kinetics and mechanism of formation, sis in sediments from Danish coastal waters: micro- Geochim. Cosmochim. Acta, 45: 687-698. bial activity and M n - F e - S geochemistry. Geochim.

Rabouille, C. and Gaillard, J.-F., 1991. Towards the Cosmochim. Acta, 51: 1583-1590. EDGE: Early diagenetic global explanation. A S0rensen, J., Hydes, D.J. and Wilson, T.R.S., 1984. model depicting the early diagenesis of organic mat- Denitrification in a deep-sea sediment core from ter, O 2, NO 3, Mn and PO 4. Geochim. Cosmochim. the eastern equatorial Atlantic. Limnol. Oceanogr., Acta, 55: 2511-2525. 29: 653-657.

Reeburgh, W.S., 1983. Rates of biogeochemical pro- Stone, A.T., 1987a. Microbial metabolites and the re- cesses in anoxic sediments. Annu. Rev. Earth Planet. ductive dissolution of manganese oxides: Oxalate Sci., 11: 269-298. and pyruvate. Geochim. Cosmochim. Acta, 51: 919-

Reimers, C.E., R.A. Jahnke and D.C. McCorkle, 1992. 925. Carbon fluxes and burial rates over the continental Stone, A.T., 1987b. Reductive dissolution of man- slope and rise off central California with implica- ganese ( I I I / IV) oxides by substituted phenols. Envi- tions for the global carbon cycle. Global Bio- ron. Sci. Tech., 21: 979-988. geochem. Cycles, 6: 199-224. Stone, A.T. and Morgan, J.J., 1987. Reductive dissolu-

Robbins, J.A. and Callender, E., 1975. Diagenesis of tion of metal oxides. In: W. Stumm (Editor), Aquatic manganese in Lake Michigan sediments. Am. J. Surface Chemistry. Ch. 9. Wiley, Chichester, pp. Sci., 275: 512-533. 221-254.

Salwan, J.J. and Murray, J.W., 1983. Trace metal re- Stone, A.T. and Ulrich, H.J., 1989. Kinetics and reac-

Page 36: The biogeochemistry of manganese and iron reduction in ... · The biogeochemistry of manganese and iron reduction ... been the subject of some debate. ... generally thought to be

284 D.J. BURDIGE

tion stoichiometry in the reductive dissolution of son, T.R.S. and S0rensen, J., 1987. Geochemical manganese (IV) dioxide and Co(Ill) oxide by hydro- oxidation fronts in NE Atlantic distal turbidites and quinone. J. Colloid Interface Sci., 132: 509-522. their effects in the sedimentary record. In: P.P.

Stumm, W. and Morgan, J.J., 1981. Aquatic Chemistry. Weaver and J. Thomson (Editors), Geology and 2 ed. Wiley-Interscience New York, 780 pp. Geochemistry of Abyssal Plains. Geological Special

Stumm, W. and Sulzberger, B., 1992. The cycling of Publication, 31: 167-177. iron in natural environments: Considerations based Thomson, J., Higgs, N.C. and Colley, S., 1989. A geo- on laboratory studies of heterogeneous redox pro- chemical investigation of reduction haloes devel- cesses. Geochim. Cosmochim. Acta, 56: 3233-3257. oped under turbidies in brown clay. Mar. Geol., 89:

Suess, E., 1979. Mineral phases formed in anoxic sedi- 315-330. ments by microbial decomposition of organic mat- Tugel, J.B., Hines, M.E. and Jones, G.E., 1986. Micro- ter. Geochim. Cosmochim. Acta, 43: 339. bial iron reduction by enrichment cultures isolated

Sunda, W.G., Huntsman, S.A. and Harvey, G.R., 1983. from estuarine sediments. Appl. Environ. Micro- Photoreduction of manganese oxides in seawater biol., 52: 1167-1172. and its geochemical and biological implications. Na- Turekian, K.K., 1977. The fate of metals in the oceans. ture, 301: 234-236. Geochim. Cosmochim. Acta, 41:1139-1144.

Sundby, B. and Silverberg, N., 1985. Manganese fluxes Wallace, H.E., Thomson, J., Wilson, T.R.S., Weaver, in the benthic boundary layer. Limnol. Oceanogr., P.P.E., Higgs, N.C. and Hydes, D.J., 1988. Active 30: 372-381. diagenetic formation of metal-rich layers in N.E.

Sundby, B., Anderson, L.G., Hall, P.O.J., Iverfeldt, A., Atlantic sediments. Geochim. Cosmochim. Acta, 52: van der Loeff, M,M.R. and Westerlund, S.F.G., 1557-1569. 1986. The effect of oxygen on the release of cobalt, Wangersky, P.J., 1962. Sedimentation in three carbon- manganese, iron and phosphate at the sediment- ate cores. J. Geol., 70: 364. water interface. Geochim. Cosmochim. Acta, 50: Westerlund, S.F.G., Anderson, L.G., Hall, P.O.J., lver- 1281-1288. feldt, A., Rutgers van der Loeff, M.M. and Sundby,

Sundby, B., Gobeil, C., Silverberg, N. and Mucci, A., B., 1986. Benthic fluxes of cadmium, copper, nickel, 1992. The phosphorus cycle in coastal marine sedi- zinc, and lead in the coastal environment. Geochim. ments. Limnol. Oceanogr., 37: 1129-1145. Cosmochim. Acta, 50: 1289-1296.

Swinbanks, D.D. and Shirayama, Y., 1984. Burrow Westrich, J.T. and Berner, R.A., 1984. The role of stratigraphy in relation to manganese diagenesis in sedimentary organic matter in bacterial sulfate re- modern deep-sea sediments. Dee-Sea Rcs., 31: duction: The G model tested. Limnol. Oceanogr., 1197-1223. 29: 236-249.

Thomson, J., Wilson, T.R.S., Culkin, F. and Hydes, Wilson, T.R.S., Thomson, J., Colley, S., Hydes, D.J., D.J., 1984. Non-steady state diagenetic record in Higgs, N.C. and SCrensen, J., 1985. Early organic eastern equatorial Atlantic sediments. Earth Planet. diagenesis: The significance of progressive subsur- Sci. Lett., 71: 23-30. face oxidation fronts in pelagic sediments. Geochim.

Thomson, J., Higgs, N.C., Jarvis, I., Hydes, D.J., Col- Cosmochim. Acta, 49: 811-822. ley, S. and Wilson, T.R.S., 1986. The behaviour of Wilson, T.R.S., Thomson, J., Hydes, D.J., Colley, S., manganese in Atlantic carbonate sediments. Geo- Culkin, F. and Scrensen, J., 1986. Oxidation fronts chim. Cosmochim. Acta, 50: 1807-1818. in pelagic sediments: Diagenetic formation of

Thomson, J., Colley, S., Higgs, N.C., Hydes, D.J., Wil- metal-rich layers. Science, 232: 972-975.