31. organic carbon, reduced sulfur, and iron · pdf fileprell, w. l., niitsuma, n., et al.,...

Prell, W. L., Niitsuma, N., et al., 1991Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 117

31. ORGANIC CARBON, REDUCED SULFUR, AND IRON IN MIOCENE TO HOLOCENEUPWELLING SEDIMENTS FROM THE OMAN AND BENGUELA UPWELLING SYSTEMS1

Kay C. Emeis,2 John W. Morse,3 and Linda L. Mays3

ABSTRACT

We examined sediments from Neogene and Quaternary sections of the Benguela and Oman upwelling systems(DSDP Site 532, ODP Sites 723 and 722) to determine environmental and geochemical factors which control and limitpyrite formation in organic-carbon-rich marine sediments.

Those samples from the upwelling sites, which contained low to moderate concentrations of total organic carbon(0.7%-3%), had C/S ratios typical of normal marine sediments, i.e., around 2.8. In these sediments, TOC availabilityprobably limited pyrite formation. Results that do not conform with accepted models were found for the sediments highin TOC (3^0-12.4%). The organic matter was of marine origin and contained considerable pyrolytic hydrocarbons, afact that we take as a sign of low degradation, yet significant concentrations of dissolved sulfate coexisted with it (> 5mmol/L in the case of Sites 532 and 723). Detrital iron was probably not limiting in either case, because the degree ofpyritization was always less than 0.65. Therefore, controls on sulfate reduction and pyrite formation in the organic mat-ter-rich sediments do not appear to conform simply to generally accepted diagenetic models. The data from these ther-mally immature, old, and organic-rich marine sediments imply that (1) the total reduced sulfur content of organic-richmarine upwelling sediments rarely exceeds an approximate boundary of 1.5% by weight, (2) the C/S ratio of these sedi-ments is not constant and usually much higher than the empirical values proposed for marine sediments. We concludethat sedimentary pyrite formation in upwelling sediments is limited by an as yet unknown factor, and that caution is ad-vised in using C/S ratios and C vs. S diagrams in paleoenvironmental reconstructions for organic-rich sediments.

INTRODUCTION

Recent investigations of the formation of pyrite have focusedon the role of organic matter decomposition via sulfate reduc-tion during shallow burial in marine sediments. Organic matterabundance and composition is perceived as one of three majorfactors that are generally considered to limit the amount of py-rite formed in sediments. The other two factors are the presenceof reactive detrital iron minerals, and availability of dissolvedsulfate (e.g., Berner, 1984, 1985). These parameters vary accord-ing to the prevailing depositional environment.

Several investigators (e.g., Sweeney, 1972; Goldhaber andKaplan, 1974; Berner, 1984; Leventhal, 1987) have demonstratedthat total organic carbon (TOC) and total reduced sulfur (TRS)correlate positively in sediments, and that the organic carbon toreduced sulfur ratio averages 2.8 (±0.8) in modern marine sedi-ments deposited beneath oxygenated seawater (e.g., Berner, 1982).Plots of total organic carbon (TOC) vs. total reduced sulfur(TRS) have been used to distinguish different depositional envi-ronments, i.e., normal marine (Berner and Raiswell, 1983 1984),euxinic (e.g., Leventhal, 1983), and freshwater (e.g., Davison etal., 1985). In euxinic environments, H2S present in the bottomwaters reacts with detrital iron. At low organic matter concen-trations this leads to low C/S ratios compared with normal ma-rine sediments. In freshwater and brackish environments, whereless sulfate is available for reduction, high C/S ratios are found(Berner, 1984, 1985; Berner and Raiswell, 1984; Davison et al.,1985). However, C/S ratios are, within a given environment, de-pendent upon the concentration and quality of TOC and rangesof C/S data are large. This led Raiswell and Berner (1985) to

1 Prell, W. L., Niitsuma, N., et al., 1991. Proc. ODP, Sci. Results, 117: Col-lege Station, TX (Ocean Drilling Program).

2 Geologisch-Paláontologisches Institut, Universitat Kiel, 2300 Kiel, FederalRepublic of Germany.

3 Department of Oceanography, Texas A&M University, College Station, Texas77843, U.S.A.

emphasize the utility of C vs. S plots in interpreting environ-ments.

One goal of this investigation is to examine the concept ofdetermining paleoenvironments based on carbon-sulfur system-atics in sediments from marine sections of Neogene to Holoceneage. Because the sediments of our sample sets are unweatheredand thermally immature, and because their facies are well estab-lished from chemical, lithological, and paleontological investi-gations (e.g., Hay, Sibuet, et al., 1984; Meyers et al., 1984;Prell, Niitsuma, et al., 1989), they constitute ideal targets to testthe traditional concepts of pyrite formation. A second objectivewas to test the method of whole-rock pyrolysis (Rock-Eval) as ameans to approximate organic matter character in the context ofsubstrate availability for bacterial fermentation and sulfate re-duction. Because of the central importance of organic matter tothe process of pyrite formation, we have gone beyond simplydetermining TOC concentration, and have used Rock-Eval py-rolysis to evaluate the origin of organic matter (Espitalié et al.,1985), and we propose that hydrocarbon yield during pyrolysismay be used to roughly determine its degree of degradation.This concept has been employed previously in order to elucidatethe effect of aerobic and anaerobic depositional conditions onorganic matter preservation (Pratt et al., 1986; Dean et al.,1986; Davis et al., 1988; Emeis and Morse, 1990; Emeis et al.,in press).

STUDY AREASThree sites drilled during Legs 75 and 117 of the Deep Sea

Drilling Project and Ocean Drilling Program have been chosento study the formation of authigenic iron sulfides during dia-genesis of sediments in the late Neogene and Quaternary. Theyare Site 532, situated offshore of southwest Africa in the Ben-guela upwelling area, and Sites 723 and 722 from the Arabiancontinental margin underneath the northwest Arabian Sea up-welling area (Fig. 1). The sediments studied range from late Mi-ocene to Pleistocene in age and encompass nearshore and hemi-pelagic marine deposits, which were deposited under aerobicand dysaerobic to anaerobic conditions in and beneath mid-wa-

517

K.-C. EMEIS, J. W. MORSE, L. L. MAYS

10° Ni

10° S

40° 30 10° E 20° 30°

56°E

Figure 1. A. Location map of DSDP Site 532 on Walvis Ridge underneath the Benguela upwelling area.B. Location of Sites 722 and 723 drilled during Leg 117.

518

ORGANIC CARBON, REDUCED SULFUR, IRON IN UPWELLING SEDIMENTS

ter oxygen-minimum zones (Hay, Sibuet, et al., 1984; Prell,Niitsuma, et al., 1989). Similarity in facies and dominant inputby calcareous nannofossils distinguishes these sites from thosestudied by Emeis and Morse (1990) on the Peru margin, whereprimary production is dominated by diatoms.

Site 532, DSDP Leg 75, is located on the eastern part of theWalvis Ridge on the Southwest African continental margin. Thedominant lithology is hemipelagic calcareous and siliceous bio-genic ooze that accumulated at rates between 25 and 50 m/m.y.and that shows pronounced dark-light cyclicity on a scale ofmeters. From smear slide analyses, Hay, Sibuet, et al. (1984) es-timate that individual biogenic components range from 80% to90% in Pleistocene and Pliocene sediments. Clay-sized clasticmaterial accounts for 10%-20% in the Pleistocene section, andincreases downhole to 50% in the late Miocene. In the biogenicfraction, nannofossils range from 20% to 50% and diatoms arepresent in quantities of 10%-40% in late Pliocene strata. Theopal-rich late Pliocene sediment is also characterized by a peakin TOC concentrations of up to 8.65%. The increase in diatomsand TOC in the late Pliocene reflects an increase in primary pro-ductivity and a strengthening of the Benguela upwelling system(Meyers et al., 1983, 1984).

Site 723 (18°03.079'N and 57°36.561 'E) was drilled in water808 m deep and recovered uppermost Pliocene to Holocene(?)foraminifer-bearing nannofossil oozes to calcareous clayey silts(Fig. 2). The sedimentation rate is high throughout the section(approximately 200 m/m.y.) and records variability in eolianand biogenic input during the depositional period in great detail(Prell, Niitsuma, et al., 1989). Abundance of TOC and the oc-currence of laminations in intervals of latest Pliocene to earlyPleistocene suggest that bottom-water oxygen contents were lowerand precluded bioturbation; the resulting laminated sequencescoincide with strata that were cemented by dolomite during dia-genesis, and in which biogenic silica was preserved.

Site 722 is located on Owen Ridge (16°37.312'N and59°47.755'E) in water 2030 m deep. The oldest sediments hereare early Miocene silty clay stones, but the dominant lithology isnannofossil ooze to diatomaceous nannofossil ooze with lighterand darker colored alternations on a scale of meters. Coincidentwith dark colors are TOC concentrations > 2% by weight, anda decrease in CaCO3. We studied organic matter character andsulfide content across one of the cyclic light-dark intervals inCore 722A-16X (146-147 mbsf) and in a few sediments fromother depths, hoping to find a clue to the depositional condi-tions leading to the light-dark cyclicity and their chemicalfingerprint.

EXPERIMENTAL METHODSSamples of Holes 723A and 722A were freeze dried and

ground immediately after core splitting on board ship; those ofHole 532B were immediately frozen upon core retrieval, andfreeze dried after shipment. The nature of the organic matterand organic carbon concentration was measured by Rock-Evalpyrolysis (Espitalié et al., 1985). The parameters determined byRock-Eval pyrolysis that are used here are TOC, the hydrogenindex (HI, in mg hydrocarbons/g TOC), and the oxygen index(OI, in mg CO2/g TOC) as determined after standardizationwith an appropriate standard (IFP 55000).

By using Rock-Eval pyrolysis results of immature marinesediments, we attempt to address the question of organic mattercomposition, i.e., its degree of refraction and its origin, whichare crucial parameters that exert control on the extent of micro-bial sulfate reduction of sedimentary organic matter. Whole-rock pyrolysis is a standard tool of hydrocarbon exploration todetermine the origin and thermal maturity of insoluble organicmatter (kerogen) in hydrocarbon source rocks. The method isbased on the assumption that organic matter of aquatic sources

o —

50 —

100 —

150 —

Z> 200 -–α=

z

250 —

300 —

350 —

400 —

1 >

ü 8o *><j | a.

1H •2H •3H 1

T3J5H 16H 17H 18H •9H 1I O X PTTxT"™1 2x113X•1 4XP15X•16Xl_rTxr^•isxl19X•20X•2 1 X •22X•"237r"124X•25X•26XP27X•28Xp29X•30xN

HZffH32X•33X•T T T P1M | •36X BJ T X " ^ ^38x139xp4OX•41 X •I7x"^™143x144X| 145XJ46X1

c

1

Hole

α>

<

α>α>uoo

X

α>cα>oo

</)

ia.

ene

oo'Id

α>

lat

723A

Lithology

Fades I

i i Illl

i |l rl 1 1 1 1 1

;; in 11 IIi , •M1 1 1 1 1 1

! t { ' ' " " '

1 ' ''ITΠVI

i i f1!1! 1 1 1 1

; N>l 1l l•|l|•|,;i|i|l|l|l|l|i|

I I I I I I I

1 ' I l l l

i i )]I|I

i > •i * i t i i i ' i i t i * *

. , iiii 1111 iii

'. IYITIYΠI i hi hi i' in ii in

1 : VlTl ITI'' l l H I I I I I

' 'l '. ••jll II II

U1111111111111

11,1,1.1,1,1,: I

: ; i i i I I I I I,1,1,1,1,1,1,1,j TITI'!1!1

1 1 1 1 1 II- J| 1 1 1 1 1 1

I i 1 I 11 | 1 11 | 1 i

| »i, l , l<•ii|i,i|i,i|i|i

"I I'I lYi I1 1 1 1 1 1 1 1 1 • 1 1 1

11,1,1,1,1,1,1,ii 1111111111111JU.1,1,1,1,1,TITIYIYVI 1 1 1 1 1 1

' ' T I T 1 1

< il|l,lI ii 11 ii,i ii 111

< IY IY IYΠ*i 1111111111111FITiiYiYiYi1

,|li111 ill li!| 1> l l l l l l l'll,llllllllllla>i•i•LI 1111111111 i 11CI!I!

—u> —α>- •σo c(0 n>

U.

c a

432 .3 Total depth

Legend

Parallel laminae (notscale): Facies Ml

L imestone/dolomite:Facies I I

Calcareous ooze

Calcareouε chalk

| Ciay/claystone

Figure 2. Lithologic summary and stratigraphy of Site 723.

519


is richer in hydrogen (i.e., in lipids) than terrestrial organic mat-ter, which contains relatively more oxygen (i.e., in carbohy-drates and lignin).

Most immature sediments of marine origin contain organicmatter in the range of 0.2% to 0.5% (Hunt, 1972) that has un-dergone severe microbial degradation. As a consequence, thehumic substance of marine sediments is a highly condensedpolymer of complex structure. The condensation to macromo-lecular compounds severely hampers complete characterizationof the proto-kerogen, and commonly monitored labile com-pounds of TOC (e.g., amino compounds, carbohydrates, hy-drocarbons, chlorins) rarely contribute more than 10% of theentire TOC. This proto-kerogen, however, still contains morehydrogen-bearing and heterocompounds (N,S,O-bearing mole-cules) when compared to thermally mature kerogen and terres-trial organic matter, and the abundance of pyrolyzable com-pounds (of which hydrocarbons are a sizable portion; Emeis etal., in press) may be a good, if crude, approximation of the sub-strate quality for bacterial fermentation and sulfate reduction.The abundance of pyrolytic hydrocarbons normalized to TOC,i.e., the hydrogen index, may thus be useful to distinguish highdegrees of biodegradation from moderate to low degradation(based on the ratio of HI to organic carbon, i.e., the ratio of la-bile lipid-rich material to TOC). We are aware that a stringentand traditional use of pyrolysis parameters is problematic in thecontext of most immature sediments, because errors at low or-ganic matter concentrations lead to erratic and exaggerated num-bers, and the general perception is that HI and OI are onlymeaningful at TOC > 1 %. Such low values are scarce in case ofthe upwelling sediments studied here, and rarely do the TOCvalues drop below a threshold of 1% TOC, below which mineralmatrix effects may interfere with the hydrocarbon yield duringpyrolysis (Katz, 1983). The upwelling sediments are thereforeuniquely suited for our purpose.

To verify TOC measurements with the Rock-Eval instrument,TOC in samples from all sites was determined by subtractingcarbonate-C from total-C obtained from combustion in a CHNElemental Analyzer (Perkin Elmer Model 240). The methodused on samples from Site 532 also yields the concentration oftotal nitrogen in the sample. We found that the TOC determina-tions with both instruments agree well in the case of the samplesets studied here. All TOC concentrations given in the tables arethose derived by the difference method. Carbonate determina-tions were performed on a Coulometrics CO2 Analyzer afteracidification with 2 N Hcl.

Total inorganic reduced sulfur (TRS) was measured on allsamples and determined by the Cr(II) reduction technique ofZhabina and Volkov (1978), as modified by Canfield et al.(1986). The method to measure acid-volatile sulfides (AVS; Wes-trich, 1983) uses cold 6 N HC1 and stannous chloride. Extractedsulfide concentrations were determined by titration with Pb-C1O4, using a silver-sulfide reference electrode. All AVS concen-trations were below our detection limit of 4 mmol/g, and wehave taken the total inorganic reduced sulfur fraction to be rep-resentative of TRS. Iron was measured after extraction withcold 1 N HC1 for 12 hr and analyzed by flame atomic absorp-tion spectrometry to give a measure of reactive iron abundancesand to calculate the degree of pyritization (Berner, 1970). Thedegree of pyritization (DOP) is given as

DOP (molar ratio) = PyriteFe/(PyriteFe + ReactiveFe),

where PyriteFe is well approximated as 0.5 TRS, based on themolar ratio of iron and sulfur in pyrite. In the past various acidconcentrations have been used to extract reactive iron (e.g.,Berner, 1970, used hot 12 N HC1, while Canfield, 1989, used asequential extraction that ended with a sodium dithionite solu-

tion of pH = 4.8). Consequently, our calculated degrees of py-ritization may be different from those derived from other ironextraction techniques; it was pointed out that our "reactive"iron may actually include iron from iron-bearing silicates.

Average C/S ratios were calculated from the ratio of averageTOC and average TRS values. Other uncertainties are reportedas standard deviations.

RESULTS

Site 532The highly productive Benguela upwelling system, where sedi-

mentation rates range from 25 to 62 m/m.y., resulted in consid-erable burial of TOC at Site 532 beneath oxygenated bottomwaters since the Miocene (Hay, Sibuet, et al., 1984). Data fromthis site are presented in Table 1. Organic carbon concentrationsare highly variable in interlayered light/dark hemipelagic oozesand range from 0.70% to 8.65%, with an average of 3.97%.Maximum TOC values occur in the late Pliocene at about 100mbsf, an increase that is accompanied by an increase in biogenicsilica and has been attributed to an increase in primary produc-tivity and a higher degree of preservation. Below this peak,TOC concentrations are considerably lower.

Fairly uniform organic C/N ratios at Site 532 average 12.9and range from 8.6 to 15.9 (see also Hay, Sibuet, et al., 1984).Analysis of macerals and lipids (Rullkötter et al., 1984; Meyersand Dunham, 1984) corroborate the finding of pyrolysis thatthe bulk of the organic matter in the sediments recovered at Site532 is of marine origin, even though contributions from terres-trial sources were apparent in all methods. Concentrations ofTRS range from 0.24 to 1.44 wt%. The relation between TOCand TRS is illustrated in Figure 3. The upper 100 m of the sec-tion has an average C/S value of 5.83 (± 2.18, n = 15), whichdecreases to 3.33 (± 1.08, n = 17) below 125 m. These ratiosare at the high end of values for normal marine environments.

Acid-soluble Fe concentrations range from 0.38 to 0.92 wt%.The DOP exhibits a rather narrow range of 0.45 to 0.63. Theseintermediate DOP values indicate that Fe was probably not lim-iting for pyrite formation in these sediments. Also, the sulfateconcentrations are high enough (4.9-31 mmol/L) to indicatethat it was not a limiting factor for pyrite formation (Gieskes etal., 1984; Berner, 1985). Organic matter character from pyroly-sis in Hole 532B clusters in an area that is characteristic of im-mature marine organic matter in a plot of the HI vs. OI (Fig. 4).As can be seen in Figure 5, a loose correlation exists between thetwo parameters HI and TOC for samples from Hole 532B. Thispositive correlation implies that the labile, hydrogen-rich com-ponents are more abundant at high TOC preservation. Becausethe relationship should be constant at uniform composition ofthe organic matter (the HI is normalized to organic carbonabundance), we interpret this correlation as an indication ofpreferential preservation of hydrogen-bearing organic materialat high TOC burial (Pratt et al., 1986).

Sites 723 and 722TOC concentrations in sediments from Site 723 average 3.17

wt%, and show a peak in upper Pliocene to lower Pleistocenelaminated and biosiliceous oozes (Prell, Niitsuma, et al., 1989;see Table 2). In contrast to samples from Site 532, no correla-tion is apparent in a plot of TOC vs. HI. The average HI is 351in the sample set from Site 723 (Fig. 5). TRS concentrations ex-hibit a wide scatter relative to TOC (Fig. 3). The average TRSconcentration in sediments from Site 723 is 0.47 wt%. Thisvalue is about half that observed for sediments from Site 532.

A possible explanation for the lower TRS values at Site 723 isthat they may arise from higher calcium carbonate content ofthe sediments. Although the TOC content of the sediments does

520


Table 1. Results of analyses for Site 532. n.d. = not determined.

Core, section,interval (cm)

Depth(mbsf)

CaCO3

(wt°7o)TOC

(wt%) C/NTRS

(wt%) C/SFe(sol)(wt%)

DOP(molar) HI OI

2H-2, 75-824H-2, 29-395H-3, 12-2213H-1, 28-4113H-2, 56-6517H-2, 10-1617H-2, 54-6822H-1, 30-4022H-2, 24-3822H-2, 52-5622H-2, 56-6022H-2, 72-7622H-2, 80-8422H-2, 84-8822H-2, 88-9227H-3, 70-8334H-2, 25-3034H-2, 34-4134H-2, 48-5434H-2, 18-12643H-2, 00-10843H-3, 06-1243H-3, 19-2643H-3, 26-3247H-2, 43-5547H-2, 55-6747H-2, 16-13056H-1, 64-7056H-1, 85-9156H-1, 97-10556H-1, 19-12666H-3, 44-49

MeanStandard deviationMinimumMaximum

5.714.019.752.153.971.071.491.793.193.493.593.693.793.793.8

116.7143.6143.7143.8144.5179.7180.3180.4180.5194.0194.2194.8229.0229.3229.4229.6267.1

28.5014.0033.1018.7012.705.70

16.0010.3017.3012.1012.607.20

n.d.12.9011.5016.3017.3018.9016.4024.6016.9014.0020.7015.9026.8024.5041.6022.5020.1022.6028.4024.80

18.877.705.70

41.60

2.115.091.675.624.465.122.935.523.907.437.848.656.436.018.451.942.593.963.163.111.994.343.993.601.612.193.851.623.072.680.701.43

3.972.140.708.65

10.6 0.4212.7 0.7612.9 0.2911.7 0.9413.5 0.8513.1 .4412.3 0.7815.9 (10.8 (14.114.215.214.714.614.013.5 (13.3 (

).94).63.20.14.26.20.02.13

).75).56

13.2 0.9712.1 0.9013.2 0.7413.1 0.7614.613.214.2 (11.3 <12.7 (11.811.4 (12.1

.12L07).94).72).851.02).84.02

11.9 0.798.6 0.24

11.5 ().50

n.d. 0.87n.d. 0.28

8.6 0.2415.9 .44

5.06.75.86.05.23.63.85.96.26.26.96.95.45.97.52.64.64.13.54.22.63.93.73.82.22.63.81.93.03.42.92.9

4.6n.d.1.97.5

0.45n.d.n.d.n.d.n.d.0.92n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.38n.d.n.d.n.d.n.d.n.d.0.63n.d.0.760.47n.d.n.d.0.47n.d.0.50n.d.0.50

0.560.170.380.92

0.45n.d.n.d.n.d.n.d.0.58n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.63n.d.n.d.n.d.n.d.n.d.0.61n.d.0.520.47n.d.n.d.0.61n.d.0.58n.d.0.47

0.550.070.450.63

338411375264n.d.501316300258314340344423371347278283182169172229340274294242240194199256210161258

28781

161501

231147244

89n.d145164

8196888588928870

14914788

124128157110106114186144102166131127288191

1345270

288

CO

cc

2.0

1.5

1.0

0.5

-

o

D

•

D

\-µ

Site 722

Site 723

Hole 532B

/

© rQQo π

. . J . . . .

/

• βD π π

CP

D α C

a o

Normal marine (C/S = 2

7/

•> * .

D

o a πα n

. i . • • • i • • • • i •

8)

•

D

. . i

• •

0 1 2 3 4 5 6 7 8 9

TOC (wt%)

Figure 3. Organic carbon to reduced sulfur weight ratios in samples from the Oman margin andBenguela upwelling. Note the lack of correlation for samples from Site 723.

521


600

500

400

×α>•oc

c 300α>σ>oi_

•o>.

200

100

Type 1 1

/

. /

/

1D

D

π

•

a

aα [

* J

Φ

•D

*

Gα

αα

o ° # α

D °•

*

^ ^Type III

π

D

αα

»

•

G

•

O

D

•

O

Site 722

Site 723

Hole 532B

°

100 200 300 400

Oxygen index (01)

Figure 4. Plot of HI vs. OI for samples studied here. All samples can be considered to contain predomi-nantly marine organic matter rich in hydrogen-bearing molecules. A shift to lower HI and higher OI, i.e.,to the evolution path of kerogen type III (terrestrial material), may be effected by advanced bacterial degra-dation of kerogen and proto-kerogen due to recycling and reworking of sediment and organic matter.

not correlate with CaCO3 (Fig. 6A), TRS exhibits a significantnegative correlation with CaCO3. When TRS and TOC are cal-culated on a CaCO3-free basis, we recognize a considerableoverlap between values from Sites 532 and 723 (Fig. 6B), andthe mean TRS concentrations, calculated on a CaCO3-free ba-sis, is almost identical (Site 532 = 1.07 wt%, Site 723 = 1.04wt%). This is an interesting similarity, because it is the samevalue as is found in sediments from the upwelling system off-shore Peru (Emeis and Morse, 1990). At present, we are uncer-tain whether the similarities are coincidental, or whether theyare expressions of more fundamental and as yet unrecognizedprocesses controlling TRS formation in organic-rich marinesediments.

While the mean TRS concentrations, calculated on a CaCO3-free basis, are almost identical at these sites, mean correctedTOC concentrations are considerably different (4.9 wt% at Site532, 7.0 wt% at Site 723), as are mean DOP values (0.55 at Site532, 0.42 at Site 723). The differences in mean TOC concentra-tions are reflected in the differences in mean C/S ratios of 4.6 atSite 532 and 6.8 at Site 723. Both values are considerably higherthan the typical value of 2.8 for "normal marine" sediments.

A limited number of samples measured from Site 722 makeinterpretations of results for ridge sediments more tenuous. Av-erage TOC concentrations in the hemipelagic sediments fromSite 722 are 1.2 wt%, and a mean of 0.29 wtVβ TRS results in anaverage C/S wt ratio of 4.1 (Table 3). The samples studied fromthis site are low in TOC, and their hydrogen and oxygen indices

show that organic matter is in part degraded and oxidized ma-rine material with a considerable proportion of terrestrial or-ganic matter. The small number of samples and the lack of irondata for this site make it difficult to judge the effects of in-creased organic-matter burial in dark members of the light/dark cyclicity on reduced sulfur concentrations. TRS and TOCappear to be correlated positively, but values never exceed 0.5wt Vo TRS in the organic-matter-rich dark intervals that have thehighest hydrogen indexes.

DISCUSSION

This study was undertaken to investigate the organic carbon-authigenic sulfide system in Neogene and Quaternary sedimentswhich have not been exposed to major secondary processes,such as exposure to weathering or meteoric ground waters. TheBenguela and Oman margin upwelling sites were chosen be-cause the environment was marine, with sediments which havemajor variations in TOC content primarily associated with theintensity of upwelling. It was consequently possible to investi-gate the influence of TOC concentration on the sedimentarysulfide system under relatively constant environmental condi-tions. Our investigation differs from previous work in that itattempts to address the problem of pyrite formation in sedimen-tary environments, where the diagenetic zone of sulfate reduc-tion is vastly expanded as compared to more frequently investi-gated nearshore sediments. Even though the amount of sulfatereduced and sulfide formed in the sediment below the immedi-

522


600

500

I 400

x

•o

α>

o•σ>.

300

200

100

CPπ D

° ° n n• ü

O DUO

Ch

- o-O

O Site 722

• Hole 532B

π Site 723

0 1 2 3 4 5 6 7 8 9 1 0

TOC (wt%)

Figure 5. Diagram of TOC vs. HI for samples studied here. Note the lack of correlation for Site 723.

ate sediment/water interface decreases exponentially with depth,and much of the sulfide thus formed may be reoxidized in mi-crobially mediated reactions, we can test the generally acceptedmodel for pyrite formation in an environment of slow turnoverrates.

The three primary factors which are generally considered tolimit formation of sulfides in sediments are dissolved sulfate, re-active iron minerals, and metabolizable organic matter (e.g.,Berner, 1970). In samples from both Sites 532 and 723 sulfateconcentrations in pore waters are not depleted to the pointwhere sulfate was limiting, and the moderate DOP (< 0.65) ofiron implies that reactive iron may not be limiting either. (Itshould be noted that we used a less harsh method for determin-ing the DOP than has traditionally been used. The harshermethods would probably have given even lower DOP values.)Consequently, it would be reasonable to expect that metaboliz-able organic matter was the limiting component for sulfide min-eral formation in these sediments, as it is in many normal ma-rine sediments (Berner, 1984).

However, there are generally high concentrations of organicmatter present in all samples of the unconsolidated and ther-mally immature sedimentary section from both Sites 532 and723. Pyrolysis results of Site 723 suggest that the organic matteris not degraded to the point that it could limit sulfate reduction(Emeis et al., this volume). Westrich and Berner (1984) havedemonstrated that it is not simply the quantity, but also thequality of organic matter which determines the pool of metabo-lizable TOC in sediments. In most samples from Sites 532 and723, the hydrogen-rich labile constituents were abundant; Daviset al. (1988) found that an HI of 150 limited sulfate reduction in

black shales of Cretaceous age. It appears that a sufficient poolof metabolizable organic matter is present in our set of samplesfor sulfate reduction to proceed.

These results present a dilemma for understanding what con-trols sedimentary sulfide production in these TOC rich marinesediments, because none of the factors traditionally consideredlimiting, e.g., labile organic matter, sulfate, and iron, are at lowenough concentrations to limit pyrite formation. Sulfate is un-ambiguously not limiting. The fact that we often observe muchhigher (>0.8) DOP values in other sediments is suggestive thatthe low DOP's observed here are not grossly misleading in indi-cating that reactive Fe is probably not limiting.

Even if reactive Fe does limit pyrite production, the coexist-ence of the sulfate with high concentrations of only moderatelydegraded organic matter for millions of years remains unex-plained. Work performed in our laboratory on ODP sedimentsfrom beneath the Peru upwelling systems is yielding similar re-sults (Emeis and Morse, 1990). For some reason bacterial sul-fate reduction fails to proceed as far as would be expected in or-ganic-rich, hemipelagic marine sediments from beneath highlyproductive upwelling systems. This may be caused by some notyet understood aspect of bacterial ecology, and is certainly wor-thy of further investigation.

Raiswell and Berner (1987) have recently examined many ofthe factors which influence C/S ratios in the post-depositionalenvironments encountered in ancient sediments. They empha-sized the need to use a large number of samples in drawing con-clusions about depositional environments. It was also pointedout that the influences of many of the factors encountered inthis study, such as subsurface supply of sulfate from evaporitic

523


Table 2. Results of analyses for Site 723.


1-1, 43_441-1, 113-1141-2, 43-441-3, 113-1141-4, 43-441-4, 144-1451-5, 43-441-5, 113-1142-1, 43-442-1, 113-1142-2, 43-442-3, 43-442-3, 113-1142-4, 43-442-5, 43-442-6, 43-442-7, 43-443-1, 43-443-2, 43-443-2, 127-1283-2, 138-1393-3, 43-443-3, 113-1143-4, 43-443-4, 144-1453-5, 43-443-6, 43-444-3, 43_446-4, 144-1457-2, 149-1507_2, 43-448-1, 43-448-4, 140-1509-2, 43-449-4, 144-14510-6, 80-8211-1, 149-15013-4, 119-12014-2, 149-15014-3, 70-7115-6, 0-117-2, 0-119-4, 64-6520-4, 30-3121-6, 45-4622-3, 29-3023-2, 0-124-3, 0-125-5, 43-4426-4, 24-2527-2, 43-4428-2, 48-4929-6, 119-12030-1, 0-131-2, 144-14532-5, 119-12032-6, 43-4433-2, 78-7933-2, 88-8933-2, 118-11933-2, 138-13933-2, 148-14933-3, 20-2133-3, 28-2933-3, 38-3933-3, 48-4933-3, 8-933-4, 40-4136-6, 0-1

Mean

Depth(mbsf)

0.41.11.94.14.95.96.47.18.28.99.7

11.210.412.714.315.817.317.919.420.320.420.921.622.423.423.925.430.552.358.056.966.271.877.381.393.495.9

119.8126.8127.5140.9154.3177.2186.7199.4204.4212.3223.4236.5244.5251.3261.1211A278.3290.8304.7305.4315.3315.4315.7315.9316.0316.2316.4316.4316.5316.1312.0333.6

Standard deviationMinimumMaximum

CaCO3

(wt%)

53.760.163.955.351.650.243.143.1n.d.57.6n.d.44.043.355.851.252.751.262.762.151.030.754.669.260.256.454.850.850.363.960.554.163.756.260.757.857.354.663.365.364.961.163.551.761.958.856.263.144.554.950.750.728.929.469.068.460.861.6n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.54.635.7

54.99.2

28.969.2

TOC(wt<Vo)

5.804.802.513.501.491.801.312.861.562.932.733.565.052.251.682.274.773.593.384.586.833.203.952.180.711.231.661.012.044.501.823.121.093.852.422.271.093.493.292.972.512.061.762.652.112.923.044.514.605.192.216.114.254.904.573.063.714.622.953.933.253.594.073.172.913.323.904.662.91

3.171.310.716.83

TRS(wt%)

0.300.190.200.200.430.400.360.280.410.380.530.600.320.450.340.410.360.270.280.370.660.300.210.270.490.410.630.470.510.400.380.190.530.360.290.290.700.430.410.290.520.530.520.540.500.510.520.720.470.590.610.700.400.160.300.580.380.710.590.530.730.690.640.740.670.680.810.491.11

0.470.180.161.11

c/s19.125.712.417.33.54.53.6

10.13.87.85.16.0

16.05.05.05.6

13.213.311.912.410.310.518.68.21.53.02.62.24.0

11.34.8

16.72.1

10.88.47.81.58.28.1

10.14.83.93.44.94.25.85.96.39.78.83.68.8

10.531.415.15.39.86.55.07.44.55.26.44.34.34.94.89.42.6

6.8n.d.

1.531.4

Fe(sol)(wt%)

n.d.0.27n.d.0.48n.d.n.d.n.d.0.50n.d.0.50n.d.n.d.0.45n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.33n.d.n.d.n.d.n.d.0.37n.d.n.d.0.410.31n.d.0.37n.d.0.29n.d.n.d.n.d.0.52n.d.n.d.n.d.n.d.0.490.39n.d.n.d.0.540.590.430.54n.d.n.d.0.29n.d.0.34n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.45n.d.

0.420.090.270.59

DOPmolar

n.d.0.37n.d.0.27n.d.n.d.n.d.0.33n.d.0.40n.d.n.d.0.38n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.36n.d.n.d.n.d.n.d.0.52n.d.n.d.0.450.35n.d.0.45n.d.0.46n.d.n.d.n.d.0.33n.d.n.d.n.d.n.d.0.470.53n.d.n.d.0.430.460.550.53n.d.n.d.0.48n.d.0.49n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.0.49n.d.

0.430.080.270.55

HI

402n.d.370n.d.241231327n.d.354n.d.364345n.d.321476285242299309343487415n.d.n.d.220259185n.d.399113279397251381452n.d.254437416417422379266395402392395n.d.444n.d.n.d.n.d.435433n.d.402n.d.397375393375369374353360388388n.d.399

35577

113487

OI

169n.d.189n.d.159163204n.d.189n.d.154116n.d.15114014185

181133133109157n.d.n.d.269162104n.d.164n.d.150116179104123n.d.16611010281

11811913611410080

115n.d.

75n.d.n.d.n.d.

8271

n.d.88

n.d.76

n.d.104n.d.11091

11511299

113n.d.I l l

1293971

269

n.d. = not determined.

524


A 10

8 -

* 6

5

4 -

2 -

*•

••,

-•

••<•••••

D

α

D

α

••

»

D

D

D

D

D

•

o

D

αB o

αD

Site 532

Site 723

π π

D Q

K3 D

π BaD

tP D

B

20 40 60

CaCO3 (wt%)

2.0

1.5 -

i 1.0 -

0.5 -

0.0

•

•

* • *

• Site 532

a Site 723

D

e „ ‰,

en C Π • Q D Q

D Q O g

i80 20 40 60

CaCO3 (wt%)

80

2.0

1.5 -

1.0 -

0.5 -

0.0

»•**

i• / .

• D

• Site 532

a Site 723

/ O D D

• m.

D* if' π *

; GD Π

/ * ° α α

i1

ft

*>

<•

10 15 2 0

Corrected TOC (wt%)

Figure 6. A. Relationships between CaCO3 and TOC for samples studied from Sites 723 and 532. B. Relationships between CaCO3 and TRS for sam-ples studied from Sites 723 and 532. C. Carbonate-corrected TOC and TRS plot for Sites 532 and 723 shows considerable overlap of the two sites.C/S ratios do not conform to the "normal marine" ratio of 2.8.

minerals, H2S mobility, and maturation of organic matter, cansignificantly influence C/S ratios. Raiswell and Berner (1987)emphasized the utility of using vitrinite reflectance to estimatepost-depositional loss of organic carbon from sediments, whichin turn results in altered C/S ratios. Their results, along withthose of this study, point to both the need for and potential util-ity of a more sophisticated approach for characterization ofsedimentary organic carbon, when it is to be used with reducedsulfur concentrations.

The hydrogen index and organic carbon concentration arehigh enough in the organic-rich sediments at Site 532 to supportsulfate reduction. These observations, in conjunction with thefact that sulfate was present in interstitial waters (Gieskes et al.,1984; Prell, Niitsuma, et al., 1989) and that DOP is rarelyhigher than 0.5, suggest that only a fraction of organic matter,whether it be labile or not, is utilized during bacterial sulfate re-duction in the diagenetic environment of marine upwelling sedi-ments. Pyrite production in these sediments does not proceed to

525


Table 3. Results of analyses for Site 722.


1-4, 144-1453-4, 114-1156-4, 114-1159-4, 119-12012-4, 119-12015-5, 145-15016-2, 29-3016-2, 20-2116-2, 25-2616-2, 34-3516-2, 38-4016-2, 43-4416-2, 48-4916-2, 56-5716-2, 60-6116-2, 65-6716-2, 69-70

Mean

Depth(mbsf)

5.9425.0453.8482.48

111.59142.35146.39146.30146.35146.44146.49146.53146.58146.66146.70146.75146.79

Standard deviationMinimumMaximum

CaCO3

(wt%)

63.158.169.956.378.387.181.686.884.179.071.271.167.161.257.156.156.6

69.711.356.187.1

TOC(wt%)

0.350.831.132.100.410.210.580.020.441.031.681.941.831.981.901.882.06

1.200.760.022.10

TRS(wt%)

0.070.370.330.240.200.080.080.090.130.250.420.220.400.430.550.550.53

0.290.170.070.55

c/s

5.02.23.48.82.12.57.00.23.54.24.08.74.64.63.53.43.9

4.1n.d.

0.28.8

HI

223234182277154105188n.d.157226302258289267281249292

23058

105302

OI

n.d586441303865n.d579n.d661376325280289301300289292

421182280865

n.d. = not determined.

depletion of one of the components of the system, and sulfatereduction and H2S production appear to be limited by an as yetunknown process. The nearly identical TRS concentrations (ona carbonate-free basis) at Sites 532 and 723 may be indicative ofsuch a limiting factor.

CONCLUSIONS

An examination of the organic carbon-authigenic sulfidesystem in Neogene and Quaternary sediments from the Ben-guela and Oman upwelling systems indicates that C/S relation-ships in organic-rich sediments do not conform with those en-countered in "normal marine" sediments. The primary diffi-culty appears to be in our understanding of what limits bacterialdecomposition of the organic matter via sulfate reduction inthis type of sedimentary environment.

Our results also further demonstrate that care must be usedin applying carbon-sulfur relations to interpretation of paleoen-vironments. In upwelling sediments with high organic carboncontents (> 3 wtΦò), the C/S ratios are not, at present, applica-ble as paleoenvironmental indicators. Primary complicating fac-tors appear to be early diagenetic competition of organic speciesand possibly of metals for reduced sulfur ions, as well as subtledifferences in the "palatability" of organic matter. Prevalenceof oxic or anoxic conditions at the sediment/water interfaceduring deposition appears to have no effect on the abundanceof reduced sulfur.

ACKNOWLEDGMENTS

K.-C. E. gratefully acknowledges financial support by US-SAC and technical support by ODP during the course of thisstudy.

REFERENCESBerner, R. A., 1970. Sedimentary pyrite formation. Am. J. Sci., 268:1-

23., 1982. Burial of organic carbon and pyrite sulfur in the mod-

ern ocean: its geochemical and environmental significance. Am. J.Sci., 282:451-473.

1984. Sedimentary pyrite formation: an update. Geochim.Cosmochim. Acta, 48:605-615.

1985. Sulfate reduction, organic matter decomposition, andpyrite formation. Phil. Trans. R. Soc. London, 48:25-38.

Berner, R. A., and Raiswell, R., 1983. Burial of organic carbon and py-rite sulfur ion sediments over Phanerozoic time: a new theory. Geo-chim. Cosmochim. Acta, 47:855-862.

, 1984. C/S method for distinguishing freshwater from marinesedimentary rocks. Geology, 12:365-368.

[Author: please add reference for Canfield (1989) cited just before "Re-sults" section.]

Canfield, D. E., Raiswell, R., Westrich, J. T., Reeves, C. M., andBerner, R. A., 1986. The use of chromium reduction in the analysisof reduced inorganic sulfur in sediments and shale. Chem. Geol.,54:149-155.

Davis, H. R., Byers, C. W., and Dean, W. E., 1988. Pyrite formation inthe lower Cretaceous Mowry Shale: Effect of organic matter typeand reactive iron content. Am. J. Sci., 288:873-890.

Davison, W., Lishman, J. P., and Hilton, J., 1985. Formation of pyritein freshwater sediments: implications for C/S ratios. Geochim. Cos-mochim. Acta., 49:1615-1620.

Dean, W. E., Arthur, M. A., and Claypool, G. E., 1986. Depletion of13C in Cretaceous marine organic matter: Source, diagenesis, or en-vironmental signal? Mar. Geol., 70:119-157.

Emeis, K.-C, and Morse, J. W., 1990. Organic carbon, reduced sulfur,and iron relationships in sediments of the Peru margin, ODP Sites680 and 688. In Suess, E., von Huene, R., et al., Proc. ODP, Sci.Results, 112. College Station, TX (Ocean Drilling Program).

Emeis, K.-C, Whelan, J. K., and Tarafa, M., in press. Sedimentary andgeochemical expressions of oxic and anoxic conditions on the Perushelf. In Tyson, R., and Pearson, T. (Eds.) Continental Shelf An-oxia. Geol. Soc. London Spec. Publ.

Espitalié, J., Deroo, G., and Marquis, F , 1985. La pyrolyse Rock-Evalet ses applications. Rev. Inst. Fr. Pet., 40:563-579.

Gieskes J. M., Johnston, K., Boehm, W., and Nohara, M., 1984. Inter-stitial water studies, DSDP Leg 75. In Hay, W. W., Sibuet, J.-C, etal., Init. Repts., DSDP, 75. Washington (U. S. Govt. Printing Of-fice), 959-963.

Goldhaber, M. B., and Kaplan, I. R., 1974. The sulfur cycle. In Gold-berg, E. D. (Ed.) The Sea, Vol. 5, New York (John Wiley), 569-655.

Hay, W. W., Sibuet, J.-C, et al., 1984. Init. Repts. DSDP 75, Washing-ton (U. S. Govt. Printing Office).

Hunt, J. M., 1972. Distribution of carbon in crust of Earth. AAPGBull., 56: 2273-2277.

Katz, B. J., 1983. Limitations of 'Rock-Eval' pyrolysis for typing or-ganic matter. Organic Geochem., 4:195-199.

Leventhal, J. S., 1983. An interpretation of carbon and sulfur relation-ships in Black Sea sediments as indicators of environments of depo-sition. Geochim. Cosmochim. Ada., 47:133-137.

, 1987. Carbon and sulfur relationships in Devonian shalesfrom the Appalachian Basin as an indicator of environment of de-position. Am. J. Sci., 287;33-49.

Meyers, P. A., Brassell, S. C , Hue, A. Y., Barren, E. J., Boyce, R. E.,Dean, W. E., Hay, W. W., Keating, B. H., McNulty, C. L., Noharra,M., Schallreuter, R. E., Sibuet, J., Steinmetz, J. C , Stow, D., andStradner, H., 1983. Organic geochemistry of sediments recovered byDSDP/IPOD Leg 75 under the Benguela Current. In Thiede, J., andSuess, E., (Eds.), Coastal Upwelling: Its Sediment Record, Part B,New York (Plenum Press), 453-466.

Meyers, P. A., Brassell, S.C., and Hue, A.Y., 1984. Geochemistry oforganic carbon in South Atlantic Sediments from DSDP Leg 75. InHay, W.W., Sibuet, J.-C, et al., Init. Repts. DSDP 75/2, Washing-ton (Govt. Printing Office), 967-981.

Meyers, P. A., and Dunham, K. W., 1984. Geolipid comparison of bio-genic sediments from Deep Sea Drilling Project Holes 530B, AngolaBasin, and 532, Walvis Ridge. In Hay, W. W, Sibuet, J.-C, et al.,Init. Repts. DSDP 75/2, Washington (Govt. Printing Office), 1089-1095.

Pratt, L. M., Claypool, G. E., and King, J. D., 1986. Geochemical im-print of depositional conditions on organic matter in laminated-bio-turbated interbeds from fine-grained marine sequences. Mar. Geol.,70:67-84.

Prell, W. D., Niitsuma, N., et al., 1989. Proc. ODP, Init. Repts., 117:College Station, TX (Ocean Drilling Program).

Raiswell, R., and Berner, R. A., 1985. Pyrite formation in euxinic andsemi-euxinic sediments. Am. J. Sci., 285:710-724.

, 1987. Organic carbon losses during burial and thermal matu-ration of normal marine shales. Geology., 15:853-856.

526


Rullkötter, J., Mukhopadhyay, P. K., and Welte, D. H., 1984. Geo- Zhabina, N. N., and Volkov, 1.1., 1978. A method of determination ofchemistry and petrography of organic matter in sediments from various sulfur compounds in sea sediments and rocks. In Krumbein,Hole 53OA, Angola Basin, and Hole 532, Walvis Ridge, Deep Sea W. E., (Ed.), Environmental Biogeochemistry and Geomicrobiology,Drilling Project. In Hay, W. W., Sibuet, J.-C, et al., Init. Repts. Vol. 3: Methods, Metals and Assessment: Ann Arbor (Ann ArborDSDP 75/2, Washington (U.S. Govt. Printing Office), 1069-1087. Sci. Publ.): 735-746.

Sweeney, R. E., 1972. Pyritization during diagenesis in marine sedi-ments, [Ph. D. dissert.]. Univ. California, Los Angeles.

Westrich, J. T., 1983. The consequences and controls of bacterial sulfatereduction in marine sediments [Ph.D. dissert.]. Yale Univ., New Ha-ven.

Westrich, J. T., and Berner, R. A., 1984. The role of sedimentary or- Date of initial receipt: 28 June 1989ganic matter in bacterial sulfate reduction: The G model tested. Date of acceptance: 19 February 1990Limnol. Oceanogr., 29:236-249. Ms 117B-155

527

31. organic carbon, reduced sulfur, and iron · pdf fileprell, w. l., niitsuma, n., et al.,...

Documents