12. INTERSTITIAL WATER STUDIES, LEG 15 - STUDY OF CO2 RELEASED FROM STOREDDEEP SEA SEDIMENTS1

Jaw-Long Tsou, Douglas Hammond, and Ross Horowitz, Lamont-Doherty GeologicalObservatory, Columbia University, Palisades, New York

ABSTRACT

Thirty-five samples of deep sea sediments were collected onDSDP Leg 15 and stored in sealed kettles for over a year. ThePCO2 of these kettles was monitored as a function of time andtemperature. When CO2 is removed for analysis, the decreaseobserved in kettle PCO2 is consistent with the assumption thatequilibration of water and calcite controls the alkalinity and pHof the interstitial water. An empirical relation between ΣCC>2 andPCO2 is selected which allows correction for CO2 removal andcomparison of PCO2 to other carbonate parameters (i.e., pH)measured on Leg 15. The resolution of the data is limited, but pHand PCO2 appear to be consistent within 0.1 pH units.

Calculation of the degree of calcite saturation is used tosuggest that pressure may significantly influence ionic equilibriabetween clay and pore waters, causing samples from Site 149 toappear understaurated with calcite.

Sediment samples from regions with strong [SC 4=] gradientsshow increases in PCO2 with time, indicating active sulfatereduction continuing in the kettles. Comparison of the CO2generation rate in the laboratory with that in situ at Sites 147 and148 suggests that the process may be limited by the availability ofsuitable organic substrates, which is a function of temperature.

INTRODUCTION

The carbonate chemistry of the interstitial fluids of deepsea sediments presents a challenging problem to geo-chemists. The system, defined here as including alkalinity,total CO2, pH, the partial pressure of CO2 (PCO2), Ca

+ +

(titrated Ca), total HCO3~, and total CO 3= , may be

completely defined by a knowledge of any two parameters,the equilibrium constants of the reactions which relatethem, and the degree of calcite saturation.

A number of processes rpay affect the system:1) Bacterial oxidation of organic matter to produce

CO 2.2) Solution and precipitation of carbonate minerals.3) Cation and hydroxyl exchange between clay minerals

and interstitial water which may be very important inbuffering pH.

4) Solution of conjugate bases such as NH3, PO 4= , FeO,

and reduction of weak bases to stronger bases (such asSO 4

= to S=).In an attempt to study these processes, and also to

evaluate the pH measurements made on shipboard, 35samples of sediment were collected in sealed kettles onDSDP Leg 15 and returned to Lamont-Doherty GeologicalObservatory where the PCO2 was monitored as a functionof time and temperature. Thirteen of these samples are

Lamont-Doherty Geological Observatory Contribution No. 1911.

from the Cariaco Trench (Site 147), twelve from the AvesRise (Site 148), and ten from the Venezuelan Basin (Site149). One core-storage kettle was broken in transportation.

These three sites are interesting to geochemists, becausethey cover a wide range of sedimentary and chemicalenvironments. The Cariaco Trench is a structural depressionon the continental shelf surrounded by water less than 200meters deep. Because it is separated from the deep water ofthe Caribbean Sea and receives a large amount of organicmaterial, its bottom water has become anoxic and has anunusually high temperature (17°C). Site 147 (882 m belowsea level) is a rather uniform calcareous clay with authigenicdolomite, calcite, and pyrite. A high sedimentation rate (50cm/103y) prevails, the material is rich in organic material(dry weight carbon 2%), and is consequently ananaerobic environment. The Venezuelan Basin is a deep(Site 149 is 3472 m below sea level), open basin with loworganic content (~0.1%), and low sedimentation rate (Av.0.8 cm/103y). Sulfate changes very little with depth in thecore, indicating limited biological activity. The sedimentchanges from a calcareous clay above 250 meters to adiatom ooze below. The Aves Rise site (1223 m deep) isalso a calcareous clay with an average sedimentation rate of3 cm/103y. Sulfate is reduced in the upper 100 meters ofthe sediment.

COLLECTION AND ANALYTICAL PROCEDURE

The shipboard handling of samples is described inHorowitz et al. (this volume). Briefly, a 10 to 15 cm sectionwas cut from the core and moved to a dry box flushed with

851

J-L. TSOU, D. E. HAMMOND, R. HOROWITZ

argon. In an effort to avoid oxidation, the sediment surfacewas scraped away and the remainder sealed in kettles forstorage. The kettles were approximately half filled withsediment. CO2 loss from the sediment during this processand during coring is difficult to estimate, but should besmall except for samples from the Cariaco Trench wherethe gas content was so high it caused difficulties in drilling.The gas pressure was sometimes sufficient to blow thesediment out from the pipe on the ship.

The kettles were refrigerated (T = ~ 4°C) from the timeof collection until the first set of measurements was carriedout, about 5 months later. While still refrigerated, analiquot of gas (about 1/6 of the total) was taken from eachkettle for infrared (IR) analysis and the kettle refilled withnitrogen to about 1 atmosphere total pressure. To deter-mine the CO2 content, each gas sample was diluted 10 to50 times with N2 and compared with laboratory standards.Results are listed in Table 1.

TABLE 1Partial Pressure of CO2 at 4°C and 22°C

KettleNumber

Depth(m) DSDP No.

15-147 Cariaco Trench (10°42.65

123456789

10111213

4.225.2

14204.5

566484

104128148177

15-148 Aves Rise

141516171819202122232425

610263563709090

107107161210

147B-1-3147A-1-2147B-1-4147B-2-2147B-2-6147B-1-4147B-6-2147B-7-4147B-9-4147B-11-3147C-2-2147C44147C-7-4

SampleInterval

(cm)

4'

pco2a

(mb)

'C

NPCO/I

(mmol/1)

PCO/L

(mb)

'N, 65° 10.46'W) Depth below sea surface:

93-11060-9695-11038-5338-532-15

96-14077-97

111-12669-84

0 4 10-34

119-129

9.92(0.284)6.454.47

11.17.959.33

12.731.116.312.312.27.65

1.3_

0.870.451.20.771.41.62.82.51.61.70.92

(13°24'N, 63°45'W) Depth below sea surface:

148-14148-2-1148-3-31484-3148-7-3148-8-3148-10-3148-10-3148-124148-124148-18-2148-234

15-149 Venezuelan Basin (15° 06

262728293031323334

73101136157

323

230315369

149-9-5149-12-5149-164149-18-3149-2-2149-4-3149-26-2149-35414941-5

135-150135-15070-85

120-135135-15045-60

75-9090-10590-105

120-135

.25'N, 69C

135-150105-120

0-15105-120135-150135-150105-120105-120105-120

2.041.932.782.474.376.272.85_

2.882.742.002.16

'21.95'W)

1.060.4781.952.001.271.185.651.132.28

0.230.360.370.290.280.500.19_

0.210.190.140.12

20.4—

13.27.64

20.716.413.025.950.722.621.016.015.3

1223 m.

5.844.881.116.81

10.910.78.099.056.666.755.185.80

NPCO2d

z(mmol/1)

882 m.

2.4—

1.61.62.21.43.14.57.36.45.24.35.7

0.671.11.31.01.11.41.90.821.11.00.600.50

Depth below sea surface: 3472 m.

0.0730.0310.150.160.120.110.490.0870.20

3.331.522.451.511.802.918.325.681.59

0.320.130.280.560.190.311.010.470.17

22° C

Pco2b

(mb)

18.7—

13.5(4.30)19.316.512.916.646.220.418.911.912.5

5.874.597.975.98

(15.0)11.0

8.649.226.387.004.536.36

2.941.382.140.8581.372.497.255.090.731

PCH4b

(mb)

n.d.-

n.d.n.d.

8.46n.d.

0.9241.706.701.200.3

-0.10.838

n.d.-

n.d.n.d.

-n.d.n.d.n.d.

1.170.593.364.99

n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.n.d.

Pco2a

(mb)

19.7-

13.7——

17.7—--——

11.8-

5.805.157.787.13

10.1--

9.476.70---

3.311.522.260.943—---—

aGas chromatograph data.bInfrared data.cTotal CO2 lost from 1 liter of pore water to gas phase in bottles before the first measurement at 4°C. (Calculated from

estimated volume of the sample sediments and measured water content of nearby sediments. DSDP data).

"Total CO2 lost from 1 liter of pore water before the first measurement at 22°C.n - d Not detectable.

852

STUDY OF CO2 RELEASED FROM STORED DEEP-SEA SEDIMENTS

When this set of measurements was completed, thekettles were removed from cold storage and allowed to sitat room temperature (T = 23°±1.5°C). The extremesensitivity of PCO2 to temperature (see Table 1 wherePCO2 at 22°C is a factor of 2 greater than PCO2 at 4°C)indicated that a better control of temperature was needed.The kettles were moved to a thermostated laboratory andplaced in styrofoam boxes where the temperature wasnearly constant (22°±0.5°C) for 100 days. A set of PCO2measurements was then made (Table 1). One hundred dayslater, a second set of PCO2 measurements was made at22°C. Some of this second set of measurements were madeusing IR analysis and all using a gas chromatograph (GC)which required only 0.5 cc of gas (about 1% of the gas ineach kettle). At this time, the partial pressure of methane(PCH4) was also determined in each kettle. All these resultsare listed in Table 1.

The radius of the core sediments is about 4 cm. If wetake 2 × 10~5 cm2/sec as the diffusion constant for CO2, amean transportation distance for CO2 in the sediments in100 days is calculated to be 14 cm. Tortuosity and porosityare not taken into account. However, cracks on thesediments are sufficient to cover their effects. Therefore thePCO2 we measured is considered to be the PCO2 of thecarbonate system of pore waters. To establish the reliabilityof PCO2, another comparison of pH measurements ismade. Several kettles were opened and punch-in pH (PIpH)was determined with an Orion pH meter (model 801 digitalpH/mV meter) and two electrodes (a Beckman No. 40471glass electrode and an Orion 90-02 double junction refer-ence electrode). They are listed in Table 2 along with twosets of pH calculated from the most recent PCO2 measure-ments: one set was derived from PCO2 and Gieskes' (thisvolume) titration alkalinity, and one from PCO2 andcalcium (see Gieskes; Presley et al.; Sayles et al.; Hammond,this volume), assuming calcite saturation.

A third of the discrepancy between measured andcalculated pH can be attributed to pH measurement error(±0.04 pH unit). The remainder is probably due to changesin Ca++ and alkalinity which have occurred during storage(a discrepancy of 0.08 pH units could be explained by a20% change during storage in ALK, [Ca++], or a 20% errorin the calcite saturation assumption). The reasonableagreement between measured pH and that calculated fromPCO2 and alkalinity indicates that the observed PCO2 isclose to being in equilibrium with the interstitial water after100 days. The even closer agreement between measured pH(±0.06 pH unit, Table 2) and that calculated from PCO2and [Ca++] is strong evidence that the water is inequilibrium with calcite and that [Ca++] is less susceptibleto change during storage than is alkalinity.

Derivation of the equilibrium constants used for thiscalculation is discussed in the following section.

EQUILIBRIUM CONSTANTS

The chemical equilibria of carbonate system in sea watercan be shown as follows (Li, 1967):

PC02)gas

[C0,=]3 J aq

aq

TABLE 2Comparison of Laboratory pH and PÇQ Measurements

KettleNo.

919202122232830

Pco2a

(mb)

46.211.08.649.476.707.002.141.37

ALKb

(meq/1)

29.148.086.906.905.625.622.322.44

Cac

(mmol/1)

5.615.125.025.025.595.59

21.9910.03

p H d

7.357.387.427.387.457.437.527.73

P He

6.997.317.367.347.397.387.357.62

P Hf

7.117.257.307.397.507.417.357.53

aMeasured at 22°C.Gieskes, this volume.

cSee Hammond, this volume."Calculated from alkalinity + PCO2eCalculated from calcium + Pço

J-L. TSOU, D. E. HAMMOND, R. HOROWITZ

K* f H C O 3 - L

K s p .*

CafCO

Kf C a + + ff Ca CO

sp

(3)

(4)

where

: fraction of free ions in total concentration.

* : denote pore water quantities (others are seawater quantities).

L : Lyman's apparent constants for sea water(Lyman, 1957).

K : apparent solubility product of calcite (Maclntyre,1965).

a : solubility of CO2 in sea water (Li and Tsui,1971).

Thus, using the major element data determined on squeezedwater, equilibrium constants can be computed for eachsample.

EFFECT OF TIME AND CO2REMOVALON PCO2

After 100 days, the gas and water should be inequilibrium. When the system is perturbed by removingCO2 from the gas phase for analysis, a new equilibriumPCO2 will be attained. This PCO2 offers a clue about theinteractions occurring between the water and the solidphases over this 100-day time span. Several mechanismsmay operate when CO2 is removed for analysis:

1) Solid phases and water do not exchange:Σ C O 2 = initial — gas removedALK = initial

2) Solid carbonate (calcite) equilibrates with the ECO2dissolved in the water:

Σ C O 2 = initial — gas removed — calcite ppt.ALK = initial — 2 × calcite ppt.[Ca++1 = initial - calcite ppt.[Cθ3=]=Ks*p/[Ca++]

3) Clay minerals exchange ions with the solution keep-ing the pH constant:

Σ C O 2 = initial — gas removedpH = initial

4) Clay minerals fix pH and water equilibrates withcalcite:

Σ C O 2 = initial — gas removed + calcite dissolved[Ca++] = initial + calcite dissolved[CO3

=]=K*p/[Ca++]pH = initial

In Figure 1, the percentage of PCO2 change from thefirst set of 22°C measurements to the second set is plottedagainst the percentage of CO2 removed (NPCO2/ALK X100, assuming titration alkalinity Σ C O 2 ) . Also plottedare curves representing cases 1, 2, 3, and 4. Cases 1 to 4 arequite similar for samples with low alkalinity and high

calcium, or those with high alkalinity and low calcium. Ifthere were no CO2 released from sediments to the gas phasein the kettle, after an aliquot of gas was removed for IR,analysis PCO2 would be reduced by 15 to 22% (see A, B inFigure 1). However, almost all the points distribute along aline traversing cases 1, 2, 3, and 4. The deviation of all thepoints are caused by the following factors:

1) Molecular diffusion of CO 2.2) Reactions between sediments and pore water.3) Bacterial activities.If diffusion controls the PCO2, all the points should lie

on the line representing case 1. However, almost all of themtraverse it toward reactions lines (cases 2, 3, 4) and evencross case 4 to the region where PCO2 increases.

The group of samples (solid symbols) showing anincrease in PCO2 may be attributed to CO2 production bysulfate-reducing bacteria and will be discussed later. Manyof the remainder (open symbols) fall between cases 1 and 2,indicating that some exchange takes place between solidand aqueous phases. Although it is possible that the silicatephases may contribute to this exchange, the change inPCO2 can be explained by assuming that calcite precipita-tion is the mechanism controlling pH and alkalinity in thesystem over this time span.

Magnesium-rich coatings (Weyl, 1967) and organic coat-ings (Chave, 1965; Suess, 1970) have been shown to inhibitequilibration between calcite and sea water, and Cooke(1971) has shown that the composition of the fluid maycontrol the chemistry of the surface rather than vice versa.Apparently some of these processes prevent the attainmentof complete equilibration.

Although suggesting short-term pH and alkalinity con-trol by carbonate minerals, these observations do noteliminate the possiblity that clay minerals exert stronginfluences, i.e., ion exchange and pH buffering, over thelonger time span available for diagenesis.

CONSISTENCY OF LABORATORY PCO2AND SHIPBOARD pH MEASUREMENTS

ON INTERSTITIAL WATER

One purpose of this study was to test the consistency ofpH and PCO2 measurements. To do this, a relation must befound between PCO2 and the CO2 removed for analysis.Since stochiometric reactions (such as calcite equilibration)cannot be written, the best approach is to extrapolate thedashed line on Figure 1 to the amount of CO2 removedbefore the first analysis. This should be comparable to theconditions under which warm squeezed (WS) samples werecollected. The slope of the line is —5, i.e., for every percentof initial Σ C O 2 removed, there is 5 percent decrease inPCO2. The same correction can be made for PCO2measurements at 4°C, enabling them to be compared withcold squeezed (CS) samples. These extrapolations areplotted in Figure 5.

Figure 2 schematically illustrates the calculation modelsused to compute pH. In one series (model A), theextrapolated PCO2, the titration alkalinity, and the calciumdata were used to compute Σ C O 2 and the degree of calcitesaturation at the storage temperature (T§) (saturationresults are discussed in the following section). This Σ C O 2

854

STUDY OF C O 2 RELEASED FROM STORED DEEP-SEA SEDIMENTS

- 6 - 4

NPCO2/ALK 1%)

2 0 2 8

15

10

0

CMO

oα.CM

Ou0-

-10

-15

-20

-25

- 3 0

extrapolation\ \ choice

SulfateReducingRegion

o Cariaco Trench

A Aves Risea Venezuela Basin

i 2

29

C 0 2 ADDED

TO WATER

COfe REMOVED-*

FROM WATER

Figure 1. PCO2 change between two successive measurements at 22°C (APCO2/PCO2) vs. the ratio of CO2 removed foranalysis to total alkalinity (NPCO2/ALK). (Solid symbols are in or near regions of strong sulfate gradients; dashed lineis chosen to relate PCO 2 and ΣCO2; s e e text for explanations of model curves.)

855

J-L. TSOU, D. E. HAMMOND, R. HOROWITZ

PCO2 (Ts)

measuredNPCO2/ALK

(Ts)

COz~ removal correction I

TITRATION

ALK (Ts)

Extrapoluted

Pco (Ts)

C a + +

(Ts)

USE CA**(Ts)

CALCULATEDEGREE ofCALCITE

SATURATIONtTs

Assume calpite soturotion

CALCULATE

ΣCOZ Aik(Ts)

CONSERVE ICOjALK

MODEL A'PH

AT Tm

Lj MODEL A" '

Pco,AT 2 2 * |

EQUILIBRATE WITH CALCITE

, Ij MODEL A** J

I P P A I

MODEL A'J>H

AT Tm

ICON;

{ MODEL B'7], PCO2 ,

, AT 22" |T — -

EQUILIBRATE

XI MODEL B** J

CONSERVE TCθ2 ALK

MODEL B'pH

AT Tm

1WITH CALCITE

I

AT 22*

MODEL B *pH

AT Tm

Figure 2. Flowchart of calculation for comparing extrapo-lated PCO2 data with other pore water data (Ts = 4° or22°C: temperature of squeeze and storage; Tm =temperature of which pore water pH was measured;dashed boxes are calculations done only on 4°C PCO2).

and the alkalinity were then used to calculate pH at the pHmeasurement temperature (T M , model A'). To estimate theeffect on pH which could be caused by calcite precipitationbetween water collection and pH measurement, model A*was used, which assumes that calcite precipitates until itreaches saturation. Models A" and A** were used tocalculate PCO2 at 22°C on the basis of the 4°C PCO2measurements. They have the same constraints as A' andA* and the results are compared to the 22° extrapolatedPCO2 in Figures 5a-c.

The second approach (model B) was to take [Ca++] andPCO2 and, assuming calcite saturation, calculate Σ C O 2 , andalkalinity. These two quantities were used, as in model A,to obtain model B' (pH calculated from Σ C O 2 and ALK atT M ) and model B* (pH of calcite-equilibrated solution).Models B" and B** were used for calculation of PCO2 at22°C from the 4°C measurements (Figures 5a-c).

The results of models A', A*, B', B*, and the pHmeasurements are plotted in Figures 3a-c for 4°C samplesand in Figures 4a-c for 22°C samples. In Site 148 and theupper 50 meters of Site 147, model pH is consistently 0.2pH units lower than the observed values. This may beattributed to CO2 production in SO 4

= reducing regions andwill be discussed later. The remaining regions show generalagreement (about 0.1 pH units) between the cluster of fourmodel pH values and the observed value. Unfortunately,there is insufficient resolution to choose one of the fourmodels as superior, and all that can be established is arather crude agreement of about 0.1 pH units betweencalculated pH and measured pH. The models are interestingin illustrating the effect calcite precipitation betweencollection and pH measurement can have on pH, usually adecrease of about 0.1 pH unit.

The extrapolated PCO2 (at 22°C and 4°C) is plotted inFigure 5a-c for comparison to that predicted from the 4°CPCO2 with models A", A**, B", B**. Again, the sulfate-

reducing regions show higher observed PCO2 than themodels predict, suggesting the continuance of CO2 produc-tion. The other regions show crude agreement (about 25%)between observed and model PCO2, but again the data lacksufficient resolution to choose a superior model.

It has been suggested (Hammond, this volume) that anincrease in temperature, which causes uptake of divalentcations and release of univalent cations (see Table 3), mayalso cause release of H+ ion (or uptake of OH~) andsolution of carbonate. This effect would make the observedPCO2 only about 10% greater than the model PCO2, andcannot be resolved.

DEGREE OF CALCITE SATURATION

The problem of calcite saturation has been discussedextensively in the literature and, as pointed out earlier,calcite and water are rarely in equilibrium. Keeping this inmind, the degree of saturation (DS) of calcite is defined as:

DS = [CO,

DS was calculated using the extrapolated PCO2, [Ca+ +],

and alkalinity (from Gieskes; Presley et al., Sayles et al.,and Hammond, this volume). Results are listed in Table 4and show considerable variation with temperature and fromhole to hole. The high values ( 1.8) in the Cariaco Trenchare probably due to CO2 loss which has not been correctedfor during sediment collection, and consequently theextrapolated PCO2 is too low. The 4°C results on samplesfrom the Aves Rise are reasonable but the 22°C resultsappear undersaturated. As mentioned earlier, many of thesesamples appeared to have active sulfate reducers which hadbeen at work longer and at a higher CO2 production rate(as will be shown later) before the 22° C measurements thanbefore the 4°C measurements. Since the PCO2 wasobserved to increase with time despite CO2 removal foranalysis, the extrapolated PCO2 was probably much toohigh.

Sediments from the Venezuelan Basin appear to beconsiderably undersaturated by nearly the same degreecalculated by Hammond (this volume) with pH andalkalinity measurements. The cause of this discrepancy isnot obvious, but the answer probably lies in the effects ofpressure on ionic equilibria between water and silicatephases. The DS calculated by Hammond for Sites 147, 148,and 149 shows smaller values for progressively deeper sites.If a decrease in pressure favors a rapid ion exchangereaction between clays and water, which might be writtenas

N a + ) clayC a H2° N a + + H C a + + clay

and perhaps be driven by a net increase in volume for thesystem, the alkalinity of the water would appear smallerand the pH more acid than in situ conditions. If the kineticsof calcite solution are not sufficiently rapid for equilibriumto be re-attained before squeezing takes place, the waterwill appear undersaturated. It is interesting to note that onesample (149-16-4-WS) was squeezed overnight and had analkalinity 25 percent greater than the corresponding CS

856

6.7pH(C.S.)

7.1 7.3 7.5 7.7

50

100

150 -

200

Dep\h

o×

_ A Δ × O

A Δ<

A ΔX

6.8

X O

lα) Cαriαco TrenchI i I . I

100

^ 200

3 0 0

4 0 0

pH(C.S.)

7.0 7.2 7.4 7.6 7.8 6.6 6.8

A QΔ ×A

00

00

6.7pH (W.S.)

7.1 7.3

50

E- 100SZ

α.

s

150

2 0 0

-6.20

O X

(α) Cαriαco Trench

I I I I I I I

6.8 7.0pH(W.S.)

7.2 7.4 7,6 7.8 6.6

100 -

ε 200 -

300 -

4 0 0

I 1o

0

90m |

o

AΔ0 A Δ

JD Δác

OA Δ»

OA C*

A Δ

X PH

Δ B'

Δ A" J

• A*

(b)Aves

1 • '

• Ix 1

X

i\

>

>

measured

Models

Rise

• I

\

(W.S.

1

1

(7.92)

X

)

100

E- 2 0 0

a.Q

3 0 0

4 0 0

pH IW.S.)7.0 7.2

> A

7.4 7.6

A Δ O( c ) Venezuela Basin

Figures 4a-c. Comparison of the measured warm squeeze pH with those calculated with different models (see Figure 2).

2 0

PC02 (mb)40 60 80 100

5 0

100

150

2 0 0

PC0 2 (mb)8 12 16 2 0

+ * x

(α) Cαriαco Trench

100

T; 2 0 0

3 0 0

4 0 0

+ Extrαpoloted PCOa ( 4 β C )x Extrapolated PCO2 (22°C)

o Model A"

• Model A * *Model B"

* Model B * *

(b) Aves RiseI

P C 0 2 (mb)

8 12 16 2 0

100

E-£200

300

400

Qß A

(ci Venezuela Basin

Figures 5a-c. Comparing the extrapolated PCO2 at 22°C with those calculated from different models using the extrapolated PCO2 at 4°C. (See dashed boxes inFigure 2.)

J-L. TSOU, D. E. HAMMOND, R. HOROWITZ

TABLE 3Effects of Squeezing Temperature on Some Major Elements

Location

1. Cariaco Trench e

2. Aves Risee

3. Venezuelan Basine

4. San Pedro Basina

Cold

466.1462.2464.3

Warm

470.4467.8470.4

Δ(W-C)

+4.3+5.6+6.1

(W-C)/C

+0.93%+1.2 %+1.3 %

Cold

7.987.576.73

Warm

9.419.237.78

(W-C)

+1.43+1.66+1.05

(W-C)/C

+17.9%+21.9%+15.6%+13.3%

Cac Mgc

5.607.95

21.2

5.587.56

19.3

-0.02-0.39-1.9

-0.4%-4.9%-9.1%-4.9%

39.046.144.3

36.742.942.8

-2.3-3.2-1.5

Cl< ALKC

-5.9%-6.9%-3.4%-2.5%

543.1549.4553.9

_

543.1555.1555.0

_

±0+1+1_

.7

.1

±0+0.3%+0.2%+1.4%

20.144.372.57 d

_

20.444.502.5 7 d

_

+0.30+0.13±0.00_

+1.5%+3.0%±0.0%

_

aBischoff et al., 1970; T c = 5°C; T w = 22.5°C.

°Sayles et al., this volume.cGieskes, this volume.d I n upper 16 samples (0 - 230 m), average warm squeeze datum (2.54) is higher than cold datum (2.49) by

1.9%. In lower 8 samples (259 - 379 m), average warm squeeze datum (2.64) is lower than cold datum (2.74)by 3.2%. For the whole core, the average change is zero.

Concentration in mmol/1. T c = 4°C; T w = 22°C.

sample, which was squeezed for only 30 minutes. Almostall other samples had CS and WS alkalinities within a fewpercent of each other. The longer time available for calciteequilibration in this WS sample after a rapid clay reactioncould be the source of this anomaly. Thus, the DScalculations from PCO2 portray a picture similar to thosefrom pH. Their divergence from the equilibrium value of1.0 is probably due to the unmeasured amounts of CO2 lostduring sampling and changes in pore water alkalinity duringstorage.

2C(org.) + SO 4= + 2H2O 2HCO3~ + H2S

H2S + FeO FeS + H2O

Ca+ + + 2HCO3 ~ CaCO3 + H 2O + CO2

so one unit of CO2 (i.e., NPCO2) is produced per unit ofSO4

= reduced. From Figure 6, the average production rateis 3 X lO~1 mmol/1 per 100 days or 3 × 10~6 mol/1 day.This can be compared with the in situ rate by fitting asteady state model to the observed [SO4

=] profile.Choosing

CO2GENERATION DURING STORAGE

In Figure 1, one group of samples showed PCO2increases with time. Virtually all of these came from, or justbelow, regions which showed strong gradients in [SO4

=],implying that viable communities of SO4

= reducers (i.e.,CO2 producers) are present in these sediments. Thedistance of each point from the dashed line in Figure 1 onthe NPCO2/ALK axis should approximate the amount ofCO2 generated over the 100 day period. Plotting thisagainst [SO4~] (Figure 6) shows little correlation, althoughan unexplained correlation exists between NPCO2/ALKand [SO4

=] (not shown). Apparently, in these sediments,the rate at which CO2 is produced is independent of[SO4

=] and depth and may, instead, be limited by the rateat which suitable organic substrates are produced. Althoughthe stoichiometry of the reaction is uncertain, it might bepostulated as

= R

Cz = [SO4=] at depth z

CQ = [SO4=] at depth 0

z=0

860

STUDY OF CO2 RELEASED FROM STORED DEEP-SEA SEDIMENTS

Table 4DEGREE OF SATURATION (D.S.) CALCULATED FROM Ca+\ ALK, AND EXTRAPOLATED P c o 2

KettleNo.

Depth(m)

15-147 Cariaco Trench16345789

10111213

4.24.55.2

1420566484

104128148177

15-148 Aves Rise141516171819202122232425

610263563709090

107107161210

Ca+ +

C.S.

3.61.51.759.14.23.55.236.155.456.156.406.30

9.227.387.377.585.795.574.724.726.026.028.97

10.1

15-149 Venezuelan Basin303126272829323334

32373

101136157230315369

15.417.122.822.510.511.9528.530.630.95

(mmol/l)a

W.S.

3.561.981.729.254.093.585.235.615.856.005.816.15

8.097.157.107.335.385.125.025.025.595.597.129.72

10.0311.6414.7616.3321.9921.7227.9230.6630.37

ALK (meq/l)a

C.S.

17.717.615.2

8.1813.820.024.627.830.032.931.423.7

4.003.505.235.027.617.666.72_

5.275.275.723.05

2.292.441.791.291.812.202.902.542.27

W.S.

17.7217.6015.17

7.9014.2210.0624.6429.1434.2332.8930.8023.74

Median

4.163.335.374.767.578.086.906.905.625.623.993.03

Median

2.442.722.061.292.322.292.972.422.22

Median

4°C

1.530.851.031.700.930.423.451.883.807.156.786.91

1.8

0.880.490.840.951.050.631.10—

0.830.900.920.640.90

0.540.770.590.840.410.620.362.330.81

0.59

D.S.

22° C

1.531.230.931.551.110.613.362.207.828.54

10.15.46

1.8

0.580.270.540.530.770.810.620.800.680.700.650.440.63

0.960.790.460.541.271.420.490.692.68

0.79

Gieskes; Presley et al.; Sayles et al.; Hammond, this volume.

R = sulfate utilization rate

D = diffusion coefficient of sulfate

Using D = 9.8 XIO~6 cm2/sec (Y.-H. Li, personal com-munication) and fitting this equation to the [SO4~]profiles in 147 and 148 (Figure 7) yields R l 4 7 = 1 X 1CT

7

mol/1 day and R 1 4 8 = 2 X 10-10 m o l / l day. Although thecorrelation may be coincidental, the increase in R withtemperature (in situ T147 17°C; T148 6-7°C) isstriking and may be a crucial factor in controlling sulfatereduction rate.

In Figure 7, PCH4 in the kettles is also plotted. Methaneis absent from regions with significant sulfate, as expectedon microbiological grounds (Desulfovibrio inhibits Methan-obacterium, B. Mechalas, personal communication). PCH4and [Sθ4=] show an inverse correlation in regions of low

sulfate, suggesting that the sulfate is probably a contami-nant in these samples and has destroyed some methaneduring storage.

CONCLUSIONS

Monitoring PCO2 in kettles of stored deep-sea sedimentsuggests several ideas:

1) pH measurements on interstitial water are probablyaccurate to 0.1 pH units. Insufficient resolution exists inthis PCO2 study to be any more precise, or to accuratelydetermine the degree of saturation of calcite.

2) The pU and alkalinity of interstitial water areprobably controlled by the interaction between bacterialmetabolism products such as CO2 and CH4 and carbonateminerals (at a constant temperature and pressure) over atime span of months to a few years.

861

J-L. TSOU, D. E. HAMMOND, R. HOROWITZ

0.7 -

(Vj(VI

S0.5I-

•SO.3 |

Sm

sOo

0.1

1

X6

(X6)

×25

1

XI9v20

CVI

lO

(VI

(VI

(VI

X

X

1 1

Time interval =110 ± 4days

average rate

^ X17 —* M

1 i

-

XI5

10

[so4=]T15

mmol/l

20

Figure 6. CO 2 generated at 22° C in HOdays vs. interstitial water [SO 4' ]. (Presley et al; Sayleset al.)

3) Pressure reduction from in situ to atmospheric maycause clay minerals to exchange cations and hydroxyl ions,significantly decreasing alkalinity and pH, and creatingundersaturation for calcite, which cannot dissolve rapidlyenough to compensate for this before interstitial watersamples are collected. This problem must be investigatedfurther before alkalinity and pH measurements can berelied on to discuss carbonate saturation.

4) The rate of CO2 production from sulfate reductionappears to increase with temperature. No correlation of ratewith depth of [SO4H was observed, suggesting that at aconstant temperature the process may be limited by theavailability of suitable organic substrates.

ACKNOWLEDGMENTS

The authors thank Dr. W. S. Broecker for suggesting thisproject, assisting in sample collection, and reviewing themanuscript. Dr. J. Simpson also reviewed the manuscriptand, with Dr. T. Takahashi, made many helpful suggestions.Drs. Sayles, Gieskes, and Presley kindly made their dataavailable prior to publication.

We also wish to express our thanks to the staff ofGlomar Challenger and the following persons: Tsung-HungPeng for guidance in using his IR and pH-meter, R. Pichulofor comments on the paper, G. Mathieu for designing thedilution system, and Marylou Zickl and Diann Warner fordoing the typing and drafting.

Financial support from National Science Foundationgrant NSF C-482 to collect the data and Atomic Energy

Commission grant AT( 11-1)2185 to interpret the results isgratefully acknowledged.

REFERENCES

Bischoff, J. L., Greer, R. E. and Luistro, A. O., 1970.Composition of interstitial water of marine sediments.Temperature of squeezing effect. Science. 167, 1245.

Chave, K. E., 1965. Carbonates: Association with organicmatter in surface sea water. Science. 148, 1723.

Cooke, R. C, 1971. The lysocline and calcium carbonatecompensation depth in the sea. Ph.D. Thesis, DalhousieUniversity.

Garrels, R. M. and Thompson, M. E., 1962. A chemicalmodel for sea water at 25°C and one atmosphere totalpressure. Am. J. Sci. 260, 57.

Li, Y.-H., 1967. The degree of saturation of CaCC«3 in theoceans. Ph.D. Thesis. Columbia University, New York.176 p.

Li, Y.-H. and Tsui, T.-F., 1971. The solubility of CO2 inwater and sea water. J. Geophys. Res. 76, 4203.

Lyman, J., 1957. Buffer mechanism of sea water. Ph.D.Thesis. Univ. California, Los Angeles.

Maclntyre, W. G., 1965. The temperature variation of thesolubility product of calcium carbonate in sea water.Fisheries Res. Board Canada Manuscript Rept. Ser. No.200.153 p.

Suess, E. 1970. Interaction of organic compounds withcalcium carbonate — I. Association phenomena andgeochemical implications. Geochim. Cosmochim. Acta.34,157.

Weyl, P. K., 1967. The solution behavior of carbonatematerials in sea water. Stud. Trop. Oceanogr. Miami. 5,178.

862

STUDY OF C O 2 RELEASED FROM STORED DEEP-SEA SEDIMENTS

5 0

100

150

200 i-

SO (mmol / I)

15 25

Δ—

A

- Δ

— Δ.ith

c

-

-

-

Δ

A*7 \' \

) A

A

A

i i

v\

-"—'

-__A

/

A

\\

&

C, = C0 - 7

---

_ * /

^ A

——

\

\•

\

\

\

\

i f

—

\

j 1-

+ 5×IO~'Z2

Cz s C

—

»

i i i

s

° •

Δ

A

O

•

1 —

/**

-

sea_ water

-O

_

0.33Z+IXI0"3Z2.

[so4•]

[ S 0 4 s ]PCH4

PCH4

1 1

•

147

148

147 "

148 -

-

-

-

-

-

-

4 6 8KETTLE PCH4 (mb)

Figure 7. [SO4 ] (Presley et al; Sayles et al.) and partialpressure of CH4 (in storage kettles} vs. depth. (Solidlines are the equations fitted to the sulfate profile.)

863