Post-depositional migration of elements during diagenesis in brown clay and turbidite sequences in the North East Atlantic

00167037/84/53.00

+ .I0

Post-depositional migration of elements-during diagenesis in brown clay and turbidite sequences in the North East Atlantic S.COLLEY,J. THOMSON,T. R. S. WIL.S~Nand N. C. HIGGS Institute of Oceanographic Sciences, Brook Road, Wormley, Godalming, Surrey GUS SUB, U.K. (Received October IO, 1983; accepted in revised form March 7, 1984) Abstract-Twocoresfroma NE Atlanticpelagic clayareaof lowaccumulation rateeachcontain a single turbid&withdifferent age.thickness andcomposition. Diagenesis following introduction of exotic turbidite material into the pelagic clay sequence has resulted in distinctive colour changes The diagenetic process is thought to he driven by bacterially-mediated organic oxidation, and compositional differences in the turbidites and clays allow examination of the effects on metal concentrations of this process In one core, a long turbidite section emplaced 330,000 years ago is overlain by clay. Organic oxidation has apparently proceeded from the turbidite top downwards and has maintained a U concentration peak below the oxidation front marked by a colour change. The U source is believed to be U initially associated with the organic matter in the oxidised section of the turbidite. Vanadium and Cu behave similarly, but upward migration of the redox-sensitive metals Mn and Ni is also seen. In the second core an 8 cm turbidite section was emplaced about 170,000 years ago in a clay column. In this case organic consumption now appears complete, but evidence for diagenetic e5ects is found in a 16 cm compositional alteration‘halo’ of the underiyingpelagic clay. Fe(R) has been enriched and the hydrogenouscomponent of Mn, Co, Ni and Cu removed from the pelagic clay to form the halo. INTRODUCTION POST-DEPOSITIONAL mobilityof severalredox-sensitive elements (particularly Mn, but also Co, Ni, Cu and U) within the sediment column, as a result of changing Eh conditions with depth, has been discussed Previously for several, mainly hemipelagic, depositional environments (LYNN and BONATT~,1965; Lt et al., 1969; RONATTIet al., 1971; HARTMANNet al., 1976). Many of the cores described by these authors consist of a brown, oxidised upper layer, with underlying greygreen, reduced autochthonous sediment. Mineralogically the sediment is uniform, but there are major chemical differences between the two layers. Manganese is enriched in the upper layer, but other elements such as Ni, Co, P, La, Ba and MO have also been reported in higher concentrations in this oxidised zone. Conversely, Fe, C,, Cr, V, U and S have been reported as more concentrated in the lower reduced layer (VAN DER WEUDEN et al., 1970; B~NATTI et al., 197 1; KORNPROB~T eral., 1973; GARDNERet al., 1982). The dominant diagenetic process responsible for this sep aration is considered to be the lowering of Eh in the deeper parts of the sediment as a consequence of organic carbon oxidation. The process of diagenesis in unconsolidated marine sediments ideally requires study of both solid phase and interstitial water compositions. Results obtained from porewater analyses, many dire&d at the problem of growth of manganese nodules via metal supply from interstitial waters, have aided in the understanding of metal geochemistry in pelagic and hemipekgic sediments (FROELICH et al., 1979; KLINKHA~~MER, 1980;

two reasons: MnO, is thought to be the next favoured electron acceptor after O2 and NO; are consumed, or secondly Eh-pH conditions change so that Mn2+ is favoured over solid manganese oxides (BURDIGEand GIESKES,1983). In either case Mn(II) ions would then migrate along a concentration gradient to a level where oxidising conditions exist, which is usually in the up ward direction. Other elements such as Ni, and to a lesser extent Cu, tend to follow the dissolved Mn protiles (KLINKHAMMER,1980; KLINKHAMMER et al., 1982; SAWLANand MURRAY, 1983) suggesting that these metals are released during the reduction of the MnO, carrier phases. While many of the examples cited above are steady state, this paper is concerned with two cores from a North Fast Atlantic area dominated by pelagic sedimentation, but also subject to occasional turbidite incursions. Organic carbon concentrations are low in the pelagic brown clay sediment (~0.2%) but the interposition of turbidite units has introduced relatively organic-rich material. Subsequent organic carbon oxidation has affected the local redox conditions in the previously oxic pelagic sediment column. This paper is concerned with the effects of the introduction of one thick and one thin layer of exotic material into pelagic clays in terms of element mobilisation patterns reflected by the solid phase. METHODS Core locations, descriptions and sampling

One of thefew brownclay areasof the North EastAtlantic to thewestof theCapeVerdeAbyssal Plain. The KLINKHAMMER et uf., 1982; SAWLANand MURRAY, is situated anza has a rugged relief with basins below 6000 m and hills 1983). rising to 5000 m water depth. It is bounded in the east by When solid mangqrem oxide is buried below the the abyssal plain, to the south by an extension of the Cape oxic zone, Mn reduction may be favoured for one of Verde plateau and to the north by the Great Meteor &amount. 1223

TO the west the rugged termin extends, rising steadily, to the Mid-Atlantic Ridge. Two gravity cores from the deep basin taken on RRS Discoverycruises, are investigated here. Core 103 I 1 was colr lected using a staintess steel gravity corer with a $0 cm diameter, 2 m long barrel at 25”39.2%, 30”5?.6W in a water depth of 6 129 m. This core consists of an upper 60 cm layer of brown clay (Munsell Code 5YR 4/4), underlain by 157 cm of sediment with a grey hue. There is a sharp coiour change within this unit from yellow-grey (SY7/2) to olivegrey (SY5/2) material at 127 cm. Core 10400 was collected using a Kasteniot gravity corer with a 2 m X 15 cm square barrel at 25”42.4’N, 30”57.7W in a water depth of 6044 m. This core consists of brown clay (2SYR 4/4f interrupted by a green (1OY 514) layer at 4755 cm and an underlying red band (5YR 5/4) from 55 to 71 cm. Shipboard sampling of the cores was carried out on retrieval and the subsamples were returned to the laboratory where they were dried at 110°C and ground in agate. When larger sampIes from 103 I! were required for XRF analysis the core was resampled at the laboratory, where it had been stored at 4°C. No porewater analyses were made on either of these cores.

obtained from XKF. LECO and ISUIO~IC analysrs (or rhe Lwc, cores are tabulated as Tables 1-4. Ferrous iron was determined on samples from core 10400 by the method of WILSON(1955). The resuhs are given HI

Table 5. RESULTS

Accumulation rafe.y and timinr! qf turbid& emplacemenr

.4nai_ysis XRF analyses for major and trace element contents were made with a Philips 14 10 X-ray spectrometer. Major element analyses were performed on glass dii manufactured from fusion mixtures according to the methods of NORRISHand HU-ITON(1969) and HARVEYet al. (1973) using a Cr-target tube. Overall ana@ticaI precision is 3% or better, except for MnO and P20J (5%). Trace elements, Na and Cl were determined on pressed powder pellets composed of 6-9 g of sediment with 0.5% wax binder. Na and Cl contents are precise to 10% and 5% respectively. A W-target tube was used for trace element analyses: precision is 2-4% for most elements except Co (6%) and Nb and Cu (6-9%). Carbonate and organic carbon contents were determined using LECO induction furnace and acid decomposition treatments. CaCOx values are measured with a precision of about 5% whereas the organic carbon data (obtained by difference) are precise to about 10%. AnaIyses for uranium and thorium a-emitting isotopes were performed by the method of THOMSON(1982). The data

Figure I shows the “ThxLm data plotted versus depth for the two cores. The mean accumulation rates shown are estimated by the conventional method (KU. 1976) from the decay of’X)T&,, with increasing depth in the sediment column. The zu)T~xms profile for core 10400 demonstrates that the pelagic sedimentation of brown clay is interrupted only in the green band (47-55 cm]. The lower contact of this band is sharp, showing little disturbance by bioturbation. Disturbance by organisms is evident at the upper contact, however, and brown mottfing, due to this process, is seen throughout the band. The green band is believed to be a turbidite. From the 230Tbx,.. profile in the brown sediment, the date of emplacement is estimated at 170,000 years before present. An important feature of the interpretation is that the red band below the turbidite, at 55-71 cm, is shown to be part of the brown clay section. The data in the red band contrast with consistent *q4,, values found in the turbidite the low off-trend 2q&, itself. In core 103 I I the z30)T~,Ccs,profile indicates that brown clay has been accumulating continuously since about 330,000 years before present, as estimated from the mean accumulation rate of 0.18 cm/IO3 years. MicmpataeontoiogicaI investigation confirms an age between 270.000 years (absence of Emiliuniu hu.xfe.vi) and 450,000 years ~P,~e~d~~erni~iania lacu~~~saoccasionally

present.

Depth tllteFfa1 cm

IO-13 20-23 31-33.5 52-55 ii 77-80 90 99 97-99.5 109 lO?-110 125 (F-127 127.5 130 136 F"" t56 15s15% 188 188.5-191

0.52

1,x 1.37 1.33 1.30 ND ND

0.95 0.92 0.94 0.92 0.90 0.3%

0.29 0.29 0.29 0.17 0.18 D.13

8.13 7.99 8.61 8.79 9.30 26.,5

98.09 98.15 98.21 98.05 95.83 97.79

2.29 2.39 2.39 2.35 3.17 2.21

24.09 24.30

0.40 0.46

ND ND

0.3! 0.39

0.13 0.13

26.70 76.44

98.06 97.95

1.94 2.05

7.15

24.36

ii.48

ND

0.3;

0.13

27.18

98.15

I.85

2.00

25.39

11.30

ND

0.35

a.13

29.49

98.25

1.75

0.03 0.04 0.03

2.15 2.01 t.99

25.74 26.54 26.87

0.27 0.33 0.32

ND NO ND

0.35 0.34 0.34

3.14 0.14 0.15

29.28 30.02 30.15

98.24 98.05 98.29

1.76 1.95 1.7,

3.22 3.24

0.04 0.03

1.94 1.88

21.02 26.97

0.29 0.31

NO no

0.33

0.15 O.ii

M.37 30.50

90.65 98.57

1.43 1.35

3.33

0.04

I.96

27.ll

J 1'1

ND

0.34

0.14

29.08

98.11

1.89

17.73 17.71 17.99 17.46 1;:;;

7.93 7.76 a.01 7.43 7.08 4.17

0.44 0.40 0.44 0.48 0.71 0.14

2.95 3.11 2.98 3.16 3.27 2.18

I.26 0.93 0.80 1.03 l.17 23.97

30.19 M.05

9.59 9.7%

4.38 4.06

0.05 0.12

2.16 2.21

29.75

9.58

4.03

3.12

27.95

9.10

3.51

0.03

27.72 26.43 26.37

9.07 9.01 8.88

3.43 3.19 3.19

y&

8.91 8.84

26.97

8.79

54.33 55.07 54.06 54.57 g.g

2.7: 2.60 2.76 2.84

2.81

AND DISCUSSXON

but Pliocene

discoasters

absent)

for

53.09 51.13 53.77 54.93 58.88 63.63 61.84 63.35 53.59 53.67 52.03 54.02 53.70 53.83 53.75 53.54 54.21 53.76

17.18 16.46 17.73 16.97 13.23 12.61 13.52 12.05 17.58 18.01 I?_88 18.22 17.90 17.79 18.21 18.49 18.21 18.19

siQ2Al103

8.03 8.14 8.20 8.17 8.11

a.11

7.54 7.41 7.92 7.28 5.17 4.90 5.23 4.75 8.75 8.74 7.91 8.17

Fe20w

0.41 0.39 0.39 0.28 0.19 0.18 0.12 0.17 0.05 0.06 0.28 0.37 0.47 0.47 0.49 0.46 0.50 0.48

MMl

2.94 2.87 2.93 3.10 2.73 2.83 2.95 2.85 2.93 2.86 2.74 2.88 2.86 2.76 2.91 2.84 2.86 2.73

MgO

2.39 0.88 0.74 1.07 1.96 2.31 2.10 2.32 0.58 0.55 0.58 0.55 0.58 0.56 0.64 0.65 0.72 0.64

CaO

2.86 2.89 2.97 2.83 2.29 2.51 2.55 2.46 2.93 2-91 2.70 2.88 2.93 2.96 2.96 3.01 3.06 2.99

KRO

1.29 2.02 f.88 1.33 2.05 2.43 2.30 2.05 1.49 1.35 I.34 1.53 1.49 1.41 1.48 1.31 1.47 1.51

NaRO

0.96 0.95 0.96 0.94 0.89 0.97 0.95 0.97 0.97 0.98 0.95 1.01 0.99 0.99 1.01 1.01 1.02 1.00

TiOR

0.20 0.17 0.16 0.19 0.22 0.20 0.?9 0.19 0.17 0.17 0.23 0.17 0.18 0.17 0.21 0.19 0.19 0.17

P205

9.84 12.37 8.97 9.23 7.18 7.48 7-68 8.30 8.94 8.84 8.05 8.56 8.72 8.72 8.59 8.33 8.33 8.23

LO1

Major element data, X wt (dry basis)

Major and trace element data for core 10400

* Cl + equivalent Ma, assuaing (wt X Na)/(wt X Cl) = 0.555

to-12 16-18 24-26 44-46 47-49 49-51 51-53 53-55 61-63 69-71 71-73 73-75 83-85 113-115 133-135 163-165 173-175 183-185

Depth interval w

Table 2.

98.70 97.54 98.42 98.15 94.79 lUO.05 99.43 99.46 97.98 98.14 94.69 98.36 97.93 97.69 98.39 98.03 98.44 97.45

Total

3.42 ND 0.83 I.75 3.50 4.25 4.92 ND 0.50 ND 0.33 0.42 0.01 ND 0.11 0.03 0.09 0.10 0.05 ND 0.15 ND 0.12 0.10

It: 2.13 1.25 Nn: 2.16 NO ND 1.96 ND 2.22 0.50 0.08

1.96 2.55 2.34 2.85 2.38 2.36 2.19 2.53 2.16 2.05 2.13 I.%

NaCl* CaCO, C-v

114 112 120 116 92 95 90 96 120 120 116 119 12.3 125 120 125 123 123

Rb

193 141 143 145 151 171 165 154 129 133 134 138 148 148 141 146 148 144

Sr

39 37 34 40 36 41 41 38 38 41 39 38 40 40 47 44 46 39

Y

227 215 214 229 324 378 393 324 211 216 215 223 226 224 223 217 224 213

Zr

28 27 28 27 24 25 26 24 26 29 29 32 31 28 30 28 30 26

Mb

117 123 121 121 77 74 68 64 74 7U 105 115 120 115 117 120 123 136

Ni

60 54 48 45 28 28 23 24 20 21 106 89 62 64 66 60 61 57

Co

169 160 167 155 115 109 106 114 161 167 165 168 169 170 174 172 171 165

V

Trace element data. ppa

89 86 90 99 79 78 76 85 91 91 89 92 91 89 90 90 89 88

Cr

146 141 155 120 77 74 72 85 105 130 138 147 137 123 131 142 145 183

Cu

116 115 116 113 82 83 79 84 116 115 112 116 120 117 119 117 120 148

Zn

S. Colley et

I”6 :dble 3.

Uramum

and thorium

~sotopl~

datb

'0, iore

103!'

ai.

together

with

CaCO?

and

:

u=i

dfl3iyse-

__ ‘HU

Depth cm

0 i0 20 30

_.

i”9 2:

80 90 99 109 116 120 125 127.5 130 132 134 136 148 156 164-165 176 188 200

u

Th

ppm

ppm

2.66iO.12 2.52m.08 2.74m.10 2.45m.09 2.54m.09 2.56m.07 2.36m.07 2.lOm.07 1.47m.05 1.574.05 1.46m.05 1.58rn.05 1.50m;05 2.00m.09 1.86M.06 7.4Om.20 12.2m.3 12.6M.3 ti.2m.4 8.1m.4 4.5m.2 5.1m.2 3.7m.2 3.28m.01 3.35rn.06 3.21m.11

17.om.5 12.9m.1 16.8x1.9 14.5m.1 12.9m.1 13.1m.4 12.4m.5 11.5m.4 6.2iO.2 8.2m.4 7.8m.3 6.3m.3 6.2m.4

6.4m.3 5.1m.4 6.1m.4 5.9m.3 5.1m.3 5.1m.2 5.3m.3 5.5m.3 4.2M.2 5.2m.3 5.3m.3 4.0m.2

3.5m.2, 0.99m.04 0.51m.02 0.54m.02 0.65M.03

6.4m.2

y&L;;

6.im.3 6.4M.3 6.8m.2 6.3m.3 6.6iil.2 6.5M.3

1.00i0.05 0.95m.04 0.98rn.04 0.92m.04 0.95m.04 0.91m.03 0.88M.03 0.83m.03 0.99m.04

12.7m.2 7.9m.3 9.7m.2 6.3m.l 2.59m.06 3.43m.07 2.41m.07

0.94m.03

1.39m.06 1.6Om.06 2.01m.08 2.44m.14 3.7om.13 3.99m.11 4.iom.10 3.98M.13 3.76rn.11

1.1om.04 l.lOM.04 1.27rn.06 i.32m.07

4.lm.3 3.4m.2

6.85fl.2 6.6iD.2 7.3m.2 6.2rn.2 6.9M.2 7.2m.3

"'Th z-Y-Th Activity ratio

;'30 u Activity ratio

Th u

i .5om.o5 l.iim.02 l.lim.03 1.05m.02 i.03m.03 1.01m.0‘4 1.06m.04 1.06M.03 i.03m.05 1.02m.03 1.06m.02 1.05m.04

t :27m:07 i.e4*0.07 1.92m.10 1.96m.06 2.03m.11

1.69m.04 1.39m.05

3.43m.12 2.87m.07 2.32m.09 2.39m.09 2.13m.05 2.12m.08

2.09m.05 2.23m.oc

: 3LTh

Z3dU

LLTh excess dpmfg

dpWg

dw/g

1.99m.09 1.80m.06 2.OOm.07 1.68M.06 1.8o*o.o7 1.73m.05 1.56m.05 1.30m.05 1.09m.04 l.llM.03 1.2Om.04 1.3010.04 l.42m.05 1.97ti.08 2.08m.06 6.10M.10 lO.lM.3 9.9m.3 fl.7m.3 6.lKI.3 3.5m.l 4.0m.i 2.83rn.13 2.49m.08 2.66m.05 2.53m.08

52.4t1.3 24.7t1.3 39.6i2.0 22.3m.6 8.1rn.3 10.9m.3 7.2m.2 4.7m.1 2.09m.07 2.77m.37 3.03m.12 3.09m.11 3.70m.17 6.1Om.10 6.4Om.20 7.3Om.20 6.OOm.10 $.3Om.20 6.OOm.20 4.5om.10 3.4im.13 3.74m.14 3.5im.09 3.25m.12 3.34m.07 3.53m.t;

--~..-.--.

___-____..__-

C,,.$

_-_.~_.___

50.4t1.3 22.9il.3 37.6i2.0 20.6rn.6 6.3Om.30 9.2oio.30 5.60m.20 3.40*0.10 l.OOm.08 1.66m.37 1.83m.13 1.79m.12 2.28M.18 4.10m.10 4.30m.20 1.20rn.M

0.68m.15 0.76rn.14 0.68rn.09 l.OOf0.14

,.a::'.

I _. __. ND 0.30 0.24 ND 0.13 NO 0.28 0.14 0.16 0.11 0.06 0.07 ND ND 0.54 0.51 0.69 ND ND 0.41 0.79 0.67 0.79 0.49 1.03 0.55

no ! .08 0.67 ND 2.92 ND 2.17 15.7 41.4 43.3 44.2 47.3 ND 45Y 46:s 48.3 NO ND 41.6 48.3 47.4 47.9 41.5 41.5 4!.7

_-.. .-._

ND = Not determined

the whole carbonate section (60-2 17 cm) underlying the brown clay (P. P. E. WEAVER, pers. commun.). This carbonate section is also interpreted as a turbidite, emplaced around 330,000 years before present. In this core bioturbation with resultant sediment mixing ap pears to have a%cted the uppermost 20 cm of the pelagic clay, and burrow mottles are also found near the upper boundary of the turbidite. Table 6 lists the average compositions for the pelagic sections in cores 103 11 and 10400, together with values for average shale and for average Pacific clays. It can be seen that the pelagic clay sections for both cores have similar contents of most elements. Diagenetic processes in core 10311 No palaeontological or sedimentological differences were found between the upper (60-127 cm) yellow-

Table 4.

Uranium and thonum

isotopic

data for core 234"

Depth cm

u

Th

PPm

ppm

G

10 25 45 50 54 60 68 71-73 a5 110 135 160 181

2.72kO.10 2.75tO.06 2.75t0.10 2.77kO.10 3.13io.10 2.62iO.07 2.59tD.07 2.68t0.09 2.71t0.09 2.79iD.07 2.85*O.C8 2.74m.10 2.87iO.09

17.6tC.9 16.9Ml.8 15.2f0.7 10.7io.3 10.6tD.5 17.9f0.6 17.1f0.9 13.3io.5 16.6t0.5 19.4t1.0 18.3k0.7 17.3t0.5 18.2tO.7

6.5i0.4 6.liO.3 5.5io.3 3.9io.2 3.4iO.2 6.8eO.3 6.6tO.4 5.oto.3 6.1i0.2 6.9t0.4 6.4f0.3 6.3iO.3 fi.3m.3

--____-___

238 U Activity Ratio 1.18iO.05 1.01io.03 0.91io.04 :.93io.o4 O.88io.03 0.92io.03 0.97to.03 0.92kG.03 0.92*L'.o4 U.86ko.03

grey and lower ( 127-2 17 cm) olive-grey sections of the turbidite in core 103 Il. Similarly, there is no major compositional distinction between the two sections when the data are expressed on a carbonate-free basis. (This is necessary because some carbonate dissolution has occurred in the upper section (Fig. 2)). The major differences between the turbidite and the pelagic clay sediment in this core are the high CaC03 content, low hydrogenous element values (but see discussion below), and relatively high organic content of the former sediment type. Figure 2 shows the concentration profiles of C,, U, the 234U/238U activity ratio, and CaCOj for the core. One marked feature is the lower organic carbon content of the upper section of the turbidite relative to the olive-grey section. LYLE(1983) has demonstrated that such green colours are produced by the reduction of a fraction of the iron in clay structures to Fe(H). 104OC _. 230Th 232 Th Activity Ratio

‘m: 3.i9*0:10 2.11+0.05 i.74tO.06 3.02tD.06 2.52io.08 2.46~1.06

234" dpW9

2.39iO.08 2.07t0.05 1.88*0.07 1.92fC.07 2.06iO.07 1.8OiC.05 1.87rn.05 1.84fG.06 1.a7t0.06 1.81iO.05 1.8OiO.05 1.82iO.07 1.78*0.06

230Th dwf9

54.5i2.4 3c.9t1.2 12.1tc.5 5.5iO.l 4.5t0.2 13.1m.4 10.5io.5 7.9ic.3 6.4i0.2 4.2t0.2 2.95t0.12 2.16tlJ.08 2.06S.lG

Z30Thexcess dpnfg

52.1k2.4 28.8t1.2 10.3io.5 3.58f0.17 2.42to.19 11.3io.4 8.6i0.5 6.1i0.3 4.5t0.2 2.39*0.20 1.15io.13 0.34+0.11 0.28i0.12

Diagenesis in pelagic clays Table

5.

Values

of FeT.

lG4QO.

FeT

and Fe(l1)

Fe(II)

and v

is calculated from

for core eT from XRF results

a wet chemical

analysis

for FeO.

Depth

interval cm

FeT

Fe(II)

I

%

Fe(IIh100 FeT

%’

10-12 16-18

5.27 5.18

0.15 0.2c

2.85 3.86

24-26 44-46 47-49 49-51 51-53 53-55

5.54 5.G9 3.62 3.43 3.66 3.32

C.16 0.20 0.31 0.44 0.54 0.43

2.89 3.93 10.22 12.83 14.75 12.95

61-63 69-71

6.12 6.11

0.53 0.44

8.66 7.20

71-73 73-75

5.53 5.11

G.24 c.14

4.34 2.45

83-85 113-115

5.62 5.67

o”05

0090

133-135 163-165 173-175 183-185

5.69 5.14 5.71 5.67

i, D

: 0 0

;

In this case the colour change, denoting

the Fe(III)/ Fe(R) redox boundary, is just below the step in the organic carbon profile. We believe that the organic material originally present in the upper part of the turbidite has been destroyed by oxidation. The rate of this process would be controlled by the rate of diffusion of oxygen downward from the sea bed, and thus would become slower as the oxidation front became more deeply buried. Such oxidation of organic matter is believed to be bacteriallymediated and results from reduction of the available oxidant producing the greatest free energy change (FROELICH et al., 1979; EMERSONet al., 1980). After oxygen itself has been consumed, the predicted sequence for reduction is NOI and MnOr, FezOr and then S04. Since no porewater data are available, the point reached in this sequence is unknown, but MnOr does appear to have been utilised. Subsequently the overall process is referred to simply as oxidation, although an Eh gradient must exist in the column. While it is not possible to model the relevant processes in detail, a semiquantitative treatment is possible. A numerical model has been produced in which the linear slope of the oxygen diffusion profile is used to calculate the flux of oxygen, and hence the carbon oxidised, during each timestep. The new, deeper position of the organic-rich surface is calculated from this information. This in turn determines the new slope of the oxygen profile for the succeeding timestep. The accumulation of new sediment is also included. Table 7 summarises the input data appropriate to station 103 11. Curves of the depth of the organic-rich layer against time are shown in Fig 3 for various values of surlicial sediment oxygen consumption. The stoichiometry of the reaction assumed is: 13802 + C,06H2630, ,,,N,6P + 18HCO; 124COr + 16NO; + HPO$- + 140H20.

1227

This composition of the organic material corresponds to the Redlield ratios normally assumed for deepocean org+ic material (JAHNKE et al., 1982). It is assumed that consumption of oxygen occurs only at the sediment water interface and at the upper surface of the organic-rich layer (the “oxidation front”). Hence, the oxygen diffusion profile between these sites is linear. The respiration at the sediment-water interface causes an offset to the diffision profile, so that the latter, when extrapolated upward, intersects the sediment-water interface at an oxygen value less than the bottom-water value. At the depth of the oxidation front, oxygen is assumed to fall to zero concentration. Hence, the linear oxygen gradient is defined by the known bottom water value for dissolved oxygen, the instantaneous depth of the oxidation front, and the assumed offset. The magnitude of this oxygen profile offset has been estimated by inspection of oxygen profile data from oxic sediments in the literature (WILSON, 1978; MURRAY and GRUNDMANIS, 1980, GRUNDMANIS and MURRAY, 1982): the effect of various estimates is shown in Fig. 3. Lack of such data at the present sites constitutes a major uncertainty in this application of the model. A further uncertainty is the possible upward diffusion of reduced chemical species from deeper within the sediment. The quality of the available data at this site does not at present justify the further elaboration of this rather crude model. However, it serves to illustrate the important point that the downward diffusive flux of oxygen to the buried reaction front is adequate to oxidise the organic carbon of the turbidite layer to the 65 cm depths observed at station 10311. It is, in fact, likely to underestimate this oxidation depth, since diagenesis attributable to electron-acceptors other than oxygen is not included. The most significant such acceptor is nitrate. Simple assumptions suggest a maximum nitrate flux to the front equivalent to about 30% of the oxidative diagenesis. It is thus possible that anaerobic diagenesis makes a significant contribution to the total diagenetic process. A notable feature of Fig. 2 is the peak in U concentration just beneath the colour change boundary at 127 cm. This is interpreted as a progressive postdepositional relocation of U in the turbidite as a result of the downward advance of the oxidation front. The initial U content of the turbidite is high, 3.3 ppm in the linear unaffected section at 175-200 cm. Assuming an initial [U] of 3.3 ppm, around 65% (68 &cm*) of the U deficit ( IO6 &XI*) of the upper oxidised section of the turbidite is now present in the peak. The U mobilised from the oxidised section appears to have been originally associated with the organic matter of the turbidite, since the U levels now present in the oxidised section are 2.7 ppm on a carbonate-free basis, similar to the values in pelagic clay (Tables 3 and 4). An association of authigenic U with C, in marine sediments is well-known (MANGINI and D~MINIK, 1979).

Gem

Marl bier

Marl

FE.

with organic

Red clay

day

clay

hwn

I. Core descriptions and ‘-IL,

contenl

rate

,-_ _._-- _

_ __.___._”___-“_ ..,“.__.__.

for the two cores studied. Sedimem

cm / lO?years

data-vprsusdepth

-0-29

Sedimentation

/

accumulation

-

-

f

-@18cm

Sedimentatian

*

f z

IO311

< 1.0

rates are also shown

230Th /*%

I

*

-II/_

IO3 years

rote

Diagenesis Table 6.

1229

in pelagic clays

Average major and trace element compositions for the different sediment types present in cores 10311 and 10400 compared with average shale (Wedepohl, 1971) and average Pacific pelagic clay (Turekian and Wedepohl, 1979). 1961; Bischoff et al _~.

. _..

10311 Element

Pelagic clay a:O-60cm b,c: 5,4

10400 Pelagic clay O-47cm 4.4

10400 Pelagic clay 71-185cm

Avera e shale B

Average Pacific pelagic claye

10400 Red clay 55-71cm 2.2

10400 Green clay 47-55cm 434

10311 Carbonatef

10311 Carbonatef

6C-127cm 5.4

127-ZlZcm 6.3

SiO2

54.30

53.20

53.60

58.90

54.90

53.60

61.90

52.60

51.10

A1203

17.60

17.10

18.10

16.70

16.60

17.80

12.90

17.00

17.10

Fe203T

7.64

7.54

8.10

6.91

7.70

8.75

5.00

7.14

6.26

MnO

0.49

0.37

0.40

0.09

0.56

0.06

0.17

0.16

0.07

MgO

3.09

2.96

2.82

2.60

3.40

2.90

2.84

3.79

3.80

cao

1.04

1.27

0.62

2.20

0.70

0.57

2.17

-

K20

2.77

2.89

2.94

3.60

2.70

2.92

2.45

0.76

0.60

TiO2

0.93

0.95

1.00

0.78

0.78

0.98

0.95

0.66

0.65

'2'5

0.24

0.18

0.19

0.16

0.25

0.17

0.20

0.23

0.28

Rb

120

116

122

140

110

120

93

116

122

Sr

149

156

143

300

710

131

16C

Y Zr

38

38

42

41

150

40

39

36

34

216

221

221

160

150

214

355

186

197

Nb

30

28

29

18

14

28

25

17

14

Ni

123

121

119

68

210

72

71

74

75

co Y

58

52

71

19

113

21

26

23

24

162

163

169

130

117

164

111

164

151 123

Cr

93

91

90

90

64

91

80

120

CU

133

141

143

45

230

118

77

87

91

Zn

118

115

121

95

165

116

82

121

119

a b c d

Interval studied Number of samples averaged for Mjor elements Number of samples averaged for trace elements From Wedepohl (1971) From Bischoff, et (1979). Rb, Zr and Nb from Turekian and Uedepohl (1961) f Expressed on a carbonate-free basis

e

The greater part of the uranium mobilised during oxidation has moved progressively downwards, BONA-IT et al. (1971) similarly found that uranium was enriched at depth in an organic-rich, hemipelagic environment in the Eastern Pacific. It is probable that the remaining 35% of U lost from the upper section of the turbidite has diffused upwards, and possibly has even been lost from the sediment to seawater. To a first approximation the radioactive decay sequence nuclides *?J, 2uU and *qh have similar activity profiles (Fig. 4a for WJ and *qh). The *qh maximum, however, occurs at a slightly shallower depth than those for the U isotopes, as would be expected if the U isotope peaks are still moving downwards slowly and increasing in size. There is also clear evidence of 234U movement relative to 238U, as some of the highest 2uU/238U activity ratios yet reported in marine sediments occur on the upper side of the U concentration peak (Fig. 4b). This isotope differentiation is well documented, and is believed to be due to a preferential migration of 2wU in the (VI) oxidation state following in situ production from 238U in the (IV) oxidation state. Only 30% ofthe *YJ so produced is susceptible to this effect (Ku, 1965; KOLODNYand KAPLAN,1970; MANGINIand DOMINIK, 1979). Up ward migration of *W has been dominant in this core

as the high 2uU/238U ratios are in the oxidised section immediately above the U peak (Fig. 4b). Further interpretation of these isotope activity ratios is complicated by the presence of U in different phases and oxidation states and is outside the scope of this paper. The mechanism invoked above to explain the uranium peak has similarities to that suggested for the formation of the roll-front uranium deposits found at the interface between oxidised and unoxidised sandstone (MAYNARD, 1983), although the level of concentration reached here is much less and diffusive rather than advective transport is involved. Of the elements measured in this core only V and Cu show signs of similar behaviour (Fig. 5). These two metals are also found in association with U in the Colorado Plateau deposits which may be genetically similar to roll-front deposits (see MAYNARD, 1983, p. 17 1). BONATT1 et al. (197 1) similarly found enrichments of U, V and Cr at depth in Pacific hemipelagic sediments. No evidence for such behaviour by Cr is found in this core, and we conclude that it was not present in association with the original turbidite organic matter, unlike U, V and Cu. Other trace metals appear to have been mobilised, but in these cases movement has been predominantly upwards so that the progressive intensification seen

1230

S. Colley et

Organic carbon (%)

Uranium (ppm)

i( 0.0

0.2 OJ I I

0.6 0 8 I I

ai. 135J /‘% (activity ratio)

I.0 ,

i

i

I

i i i

\

t

.-

.

.

.

.

!

\ . \

FIG. 2. Descriptive core log and depth for con 103 I I.

profiles for C,,

uranium, the *"U/2'8U activity ratio and CaCOJ versus

on the downward migration of U is absent. Manganese has been concentrated in a peak immediately above the turbidite/clay boundary at 57 cm (Fig. 5). In terms of the model, this would have occurred on resumption

Table 7.

A summary of the input data used to construct a numerical rode1 Appropriate ta station 10311

Psr&wers

used in nmdel (station 10311)

= 0.18 cm/iO'yr Pclrglc acc\nulation rate Organrc cwbon originally in turbidite = o.BZdry wt = 0.22 m nales Bottom sea-water oxygen content = 0.85 x 10-S cm2 set-1 Free solution dWfurlon coefficient (01 porosity of sedlmnt Carswed untfon)(B) - 0.70 Ocpth of uniform surficial reaction layer assured * 2 cm Equations: T0rtwxity e

=

J-$

where F IS formation factor

F

-

V3

Ullman and Aller (1982)

Effective diffusion coefflciant (0,) Oxygen flux in porewater =

00

=

OBm2

=

W2

4

Rarults: Penetration of the oxidation front rntn the turbidite (d) for various ass& values of activity at the sediinzntwater Interface and trns since crnplacmnt activity(lmles c.-2 WC-')

d(cm) 2x104 yr

105 yr

2x105 y'

22:: : ;;:::

17.6 15.3

33.5 39.4

43.2 M.8

::;z : ;$

21.1 19.4

45.0 so.2

64.4 58.0

of pelagic sedimentation after deposition of the turbidite. At that time the depth of the pelagic clay cap would be at a minimum and diffusion of oxygenated seawater into the turbidite less hindered, with the result that the rate of downwards oxidation would be at a maximum. On utilisation of MnOr as an oxidant, Mn(II) would diffuse in solution and be oxidised to the solid phase again at higher levels in the sediment column (FROELICHetd., 1979; BURDIGEand GIESKES, 1983). The Mn peak so produced would then remain unaSxted by subsequent oxidation as it would always be located in a zone of oxidised porewater, although the locus of manganese immobilisation would move downwards in time as oxidation proceeded. It can be seen from Fig. 5 that the oxidised turbidite section does indeed have higher Mn levels than the lower section. A similar behaviour is observed for Ni (Fig. 5). EIONATTI et al. (197 I) also found a predominant upward movement of Mn, Ni and also Co in a hemipelagic environment, where precipitation occurred in an upper oxidised zone. Diagenetic processes in core 10400 The major compositional contrasts between the short green section (47-55 cm) and the over- and underlying pelagic clay in this core are its higher contents

1231

Diagenesis in pelagic Clays

FIG. 3. The thickness of the oxidised layer plotted against time elapsed since the turbidite event at station 1031 I (calculated using the model parameters given in Table 7). The upper diagonal line represents the increasing thickness of pelagic sedimentation. The lower (diagenetic) curves are drawn for various values of suficial oxygen demand in units of IO-l3 moles cmm2se& (assumed constant over the upper 2 cm). As surficial demand rises, less oxygen is available for reaction at the surface of the deep organic-rich layer (close hatching) so that the oxidised layer deepens less rapidly.

of SiOz , CaO, and Zr, and lower contents of Rb, AlrOs,

V, Cr and Th (Table 6). This is interpreted as being due to higher heavy mineral (SQ and Zr) and calcium carbonate (Ca and Sr) contents in the green section, with the remaining elements lower because of dilution. These ditliinces reinforce the 2% interpretation that this unit is a turbidite with a different genesis from the remainder of the pelagic clay section. In contrast, Table 6 shows that the longer red section (55-7 1 cm) in the core is similar in composition to the pelagic clay in terms of SiO2, A&03, CaO, Zr, Cr, Rb, Ti02, V and Th contents, and confirms the ‘?&.,_ interpretation that this section is modiEed pelagic clay. The only compositional differences observed between the red band and the underlying brown clay are its higher Fe content (increased by 0.65% with all Fe expressed as Fe20s) and its lower contents of Mn, Co, Ni and possibly Cu (Fig. 6 and Table 6). The red band is therefore interpreted as a diagenetic ‘halo’ caused by the overlying turbidite. The compositional distinction of the red band indicates that the alteration involves ion transport of iron into the halo and removal from it of redox-sen-

sitive Mn and its congeners. A redox replacement reaction of the type:

Md+ + 2Fez -

Mn$+ + 2F<’

(I)

was therefore considered. Inventory calculations for the red band show that there is an excess of 54 mg/ cm2 Fe whereas 46 mg/cm2 Mn has been removed. There is therefore a shortfall of Fe for a reaction such as (l), but more importantly oxidation state analyses (Table 5) demonstrate that an amount of Fe(R) is present in the red band sufficient to account for all the iron excess. Iron therefore appears to have been concentrated as an Fe(R) mineral (such as siderite). The action of another unidentiEed reductant must therefore be invoked to explain the observed manganese mobilisation. Because high C, values are not observed anywhere in this core, organic oxidation is probably now inactive or proceeding at a slow rate on the refractory organic fraction (MULLER and SUEJB, 1979; MULLER and MANGINI, 1980). The most reducing site in this se& iment column would have been in the turbidite, because it initially contained the C, to fuel the diagenetic

S Colley 234

0

12

3

U. 4

FIG. 4a. Profiles of 2)4U and 2qh turbidite at station 103 1 I.

?h, 5

dpm/g 6

8

7

91011

within the calcareous

process. No sulphide was detected in the solid phase of either the green or red bands by the methylene blue method following acidification by 5M HCI (GIJSTAFSSON, 1960). We conclude that reduction in this thin turbidite was sufficient to utilise iron oxides, but insufficient

to utilize SO:-, and that this was the source

234

eI

a!

for the Fe(l1) enrichment now observed 111the red band. Of Mn and the trace elements associated with manganese oxides (Ni and Co) only Co gives any indication of subsequent immobilisation. Refixation of Co appears in a peak immediately below the lower boundary of the red band (Fig. 6). which contains excess Co amounting to 40% of that removed. No corresponding Mn and Ni relocalisations are observed, suggesting that these elements have been lost by upward diffusion. This would require an anoxic overlying column. The most likely period for such a condition would have been shortly after emplacement of the turbidite before further pelagic sediment accumulated. Despite the indications that oxidation of the turbidite organic matter was achieved quickly (loss of Mn and Ni) and is complete (low C,,), there is another mdication which suggests that fully oxic conditions have not yet been established. The evidence for this IS the existence of traces of Fe(II) at all depths to 75 cm. While the Fe(I1) contents on the green and red bands are highest and similar, the Fe(II)/Fe(total) values in the green band are higher (Fig. 6). As discussed previously, LYLE (1983) interprets such green colours as indicative of reducing conditions. The red band sediment can be turned green experimentally by treatment with sodium dithionite as described by Lyle. so its natural Fe(H) content is evidently insufficient to develop green colouration. The red colour of this halo must be a reflection of its decreased Mn content rather than its additional Fe(H)

l-4

xl” activity ratio

1.3

.....

g“8o t’M ..._...t,...... -“i’t

90

0.9-

o+i 1 0

t 1-m

-I-“=“”

I

-t;-1321M*XlMuM

+------

OLIVE

GREY

MARL

___

t

I 2

1 4

, 6

I 8

I 10

12

14

U Pm FIG. 4b. Plot of the *‘4U/2’sU activity ratio versus U content for the turbidite section in core 10311. The numbers by the points represent sample depths (cm). High values of the activity ratio are seen in samples above the uranium peak maximum (yellow-grey marl) indicating upward migration of *“U.

1233

Diagencsis in pelagic clays Ni ppm

MnO%

50

loo

150

I

i i

.

FIG. 5. Solid phase concentration-versu.vdepthprofil@for MnO, Ni, Cu and V (expressed on a carbonate_ free basis) in core 10311.

It is noteworthy on this interpretation that only the authigenic component of Mn, Co and Ni in the red band appears to have been remobilised, presumably by reduction of grain coatings. The residual levels of these elements now observed (Fig. 6 and Table 6) are

co ppm

0.1 0.2 0.3 04 0.5

t

:

.

I

.I /

50

150

100

2

Fal 0,

x

4

8

6

I

:’ .

I

.. /*I f -7

Ni ppm

MnO % 100

50

similar to those of shale, suggesting that the sediment matrix remains una&cted. For these elements, the diagenetic alteration causing the halo has therefore been analogous to the operation desired of a good selective leaching agent.

10 1

:’ \ .

./

I

-W-\

7 ‘-7.

-7

.

.

.

I.

I

i

i

I

i

i I i FIG. 6. Solid phase concentration-vemudepth profiles for Co, MnO, Ni, Fe201 and Fe(II)/Fer X 100% for core 10400.

CONCLUSIONS

( 197I) Post-deposctional mobiht~ of some trans1tiun ei~

The discovery of two turbidites of different character in pelagic clay sections has allowed an investigation of metal mobility in response to organic diagenesis. interpretations

of the diagenetic

alterations

have been

based on the general&d scheme which has emerged for deep-sea organic oxidation. These interpretations must be seen as conditional because no pore water information is available. In the case ofthe long turbidite (Core 103 11) an active process has been identified, but in the case of the short turbidite (Core 10400) organic combustion appears complete pretation is phenomenological. In Core 10311:

and the inter-

i) The initial organic carbon content of the turbidite has been oxidised progressively from the top downwards. A simple numerical model based on oxygen diffusion from bottom water indicates that this is the likely primary mechanism. ii) Uranium (and to a lesser extent Cu and V) are released from organic matter association in the upper oxidised section of the turbidite, and immobiiised at depth. As the oxidation front proceeds the U peak is progressively moved downwards and augmented. Around 65% of the U mobilised is now located in the peak, which is probably a marker of a particular Eh/ DH condition. iii) Some preferential migration of 2341Jupwards from the U peak is seen.

ments, phosphorus, uranium and thorium in deep va se& iments. G~wJz~rn.~f~~~~~r~~z it4a 35, 189-B 1 BURDIGED. J. and CIESKES J. M. t 1983)A pore w~ter/sohd phase diagenetic model for manganese in marine sediments. .dmer. J .%i 283, 29-47 EMERSONS.. JAI~NICER.. B~NIXH M. L.. FROEI.I(T~ P. h.. KLINKHAMMERG. I’.. BOWSERC. and SETLO(‘IIC;. -, ( 19X0\ .-

Early diagnesis in sediments from the eastern equatonal Pacific. 1.Pore water nutrient and carbonate results Earth Plonef Sci. Leff 49, 58-80.

FROELICH P. N.. KLIYKHAMMER G. P.. BENO~R XI I... LUEDKE N. A.. HEATH G. R.. CULLEN D.. DA~JPHINP.,

HAMMONDD.. HARI‘MAAB.. and MAYNARDV, (1979) Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Grc&im. C’osmochim .4cfa 43. 1OTC-1090 GARDNER J. V., DEAN W. E.. KL.ISE D. H. and

BAWAU;

J. G. (1982) A climate-related oxidizing event in deepsea sediment from the Bering Sea. @car. Res. 18, 95 -107. GRUNDMANISV. and MtJRRAY J. W. (1982) Aerobic res-

piration in pelagic marine sedimenti. Gmchim. C’mmtxhwn Actn 46, IIOI-1120. Gt~s-F~Fssor*i L. f 1960) Determination of ultramicro amounts of sulphate as methylene blue: I. The colour reaction. .4naimt London 4, 1’7-235. HARTMANN M.. MULL.% P. J.. Suwj E. and \-4x DER WEIJDEN C. H. ( 1976) Chemistry of Late Quaternary sed-

iments and their interstitial waters from the N.W. African

continental margin. “A~c?c~r ” I;‘orsch-Erge~nfssu C-24, I II h7 HARVEYP. K., fAYI_oR R. D.. HENURVR. D. and BA~XXOFI F. (1973) An accurate fusion method for the anaiysis of rocks and chemically related materials by x-ray fluorescence spectrometry. X-Ruy Specfrttm 2, 33-34. JAHNKE R. A., EMERSONS. R. and MIJRRAY J. u’. (1982) A model of oxwen reduction. denitrification, and owmc

matter miner&ation

In Core 10400: i) A red diagenetic halo, 16 cm in iength, has been developed from brown clay immediately below a turbidite 8 cm thick. Iron as Fe(II) has been added to the halo, while Mn, Ni, Co and Cu (?) have been removed. The remaining levels of Mn and congeners in this band are now similar to shale values. Only Co shows signs of subsequent immobili~tion. ii) As a consequence of the presence of the turbidite, Fe(H) is still present at all depths to 75 cm, although only the turbidite has a sufficiently high Fe(II)/Fe(total) ratio to appear green. authors would like to thank Dr. P. P. E Weaver and Dr. F. Culkin of the Institute of Oceanmphic Sciences for useful discussions. They are also grateful to Dr. P. N. Froelich and Dr. M. Lyle for constructive reviews of an earlier version of this manuscript. This research has been carried out under contract for the Department of the Environment, as part of its radioactive waste management research programme. The results will be used in the fo~ulation of Government policy, but at this stage they do not necessarily represent Government policy.

Acknowledgmenu-The

REFERENCES BISCHOFFJ. L., HEATH G. R. and LEINEN M. (1979) Geo-

chemistry ofdeepsea sediments from the Pacific Manganese Nodule Province: DOMES sites A, B and C, In Marine Geology and Ocea~ograph31offhe Pacific manganese Nodtrle Province (eds. J. L. BlSCHOff and D. Z. PIPER). p. 397-436. Plenum Press. BONATTIE., FISHER D. E.. JOENSII~I0. and RVDELL H. S.

in marine sediments. L&wl. Oceunc~pr 27, 6 10-623. KLINKHAMMERG. P. ( 1980) Early diagenesis in sediments from the Eastern F~uatorial Pacific, Ii. Pore water metal results. Eurfh Planet. Sci. Leff. 49, 81-101.

KL]NKHAMMER G. P.. HECKLED. T. and GRAHAMD. W. ( 1982) Metal diagenesis in 0x1~ marine sediments. Enrfh Planer Sci. Lxff. 61, 2 I I-2 IO. KOLODNYY. and KAPLAN 1. R. (1970) Uranium wtopes in sea-floor phosphorites. Grochtm. Coswochirn. :tcfu 34, 3-24. KORNPROB~TC.. BOULEGUEJ. and MK”HARDG. (1973) Remobilisation du manganese dans les sediments superiicieis

du bassin angolais. C.R. ,kud. S’ci. Paris 276, Scrie D. 145-148. Ku T-L. (1965) An evaluation of the ‘34U/23sUmethod as a tool for dating pelagic sediments. J. Geophy,t Res. 70.3457.3474. Ku T-L. (t976) The uranium-ales methods of age determination. Ann. Rev. Eurfh Planer. Sci. 4, 347-379. I..I Y. H.. BXHOFF J. and MA~HIEU G. (1969)The migration of manganese in the Arctic basin sediments. Eur?h PlaneI. Sci. Left. 7, 265-270. LYLE M. (1983) The brown-green color transition in marine sediments: A marker of the Fe (II&Fe (II) redox boundary. Limnol.

Oceanogr. 28, 1026- 1033.

LYNN D. C. and BONATTIE. (1965) Mobility of man~ne~ in diagenesis of deep sediments. h&r. Geol. 3, 457-474. MANGIN~A. and DOM~NIKJ. (1979) Late Quaternary sapropel on the Mediterranean Ridge: U-budget and evidence for

low sedimentation rates. SGdiment. &o/. 23, 113-125. MAYNARD J. B. (1983) Geochemistry qf Sedimenfurv Ore

Lkpusifs. 305 pp. Springer-Verlag. M~JLL.ER P. J. and SUESSE. (1979) ~~u~ivity, ~dimen~ti~n rate and sedimentary organic matter in the oceans---l. Organic carbon preservation. Lkq-.% Res. 26, 1347-1362. MUI I ER P. J. and MRN(;INI A. ( 1980) Orgamc carbon de-

Diagenesis in pelagic clays composition rates in sediments of the Pacific manganese nodule belt dated by 2r”Th and 23’Pa. Earth Planer. Sci. Left. 51, 94-114. MURRAY J. W. and GRUNDMANIS V. (1980) Oxygen consumption in pelagic marine sediments. Science 209, 15271530. NORRISH K. and HUTTON, J. T. (1969) An accurate X-ray spectrographic method for the analysis of a wide range of geological samples. Geochim. Cosmochim. Acta 33, 431453. SAWLAN J. J. and MURRAY J. W. (1983) Trace metal remobilization in the interstitial waters of red clay and hemipelagic marine sediments. Earth Planer. Sci. Lett. 64,2 13230. THOMSONJ. (1982) A total dissolution method for determination of the a-emitting isotopes of uranium and thorium in deepsea sediments. Anal. Chim. Acfa 142,259-268. TUREKIANK. K. and WEDEPOHLK. H. (1961) Distribution

1235

of the elements in some major units of the earth’s crust. Bull. Geol. Sot. Amer. 72, 175-192. ULLMAN W. J. and ALLER R. C. (1980) Dissolved iodine flux-from estuarine sediments and implications for the enrichment of iodine at the sediment water interface. Geochim. Cosmochim. Acta 44, 1 I77- 1184. WEDEP~HL K. H. (1971) Environmental influences on the chemical composition of shales and clays. In Physics and Chemistry of rhe Earth (eds. L. H. AHRENS, F. PRESS, S. K. RUNCORNand H. C. UREY), Vol. 8, p. 307-333. Pergamon. VANDER WEUDEN,C. H.. SCHUILINGR. D. and DAS H. A. ( 1970) Some geochemical characteristics of sediments from the North Atlantic Ocean. Mar. Geol. 9, 81-99. WILSONA. D. (1955) A new method for the determination of ferrous iron in rocks and minerals. Bull. Geol. Surv. G.B. 9, 56-58. H.M.S.O. WILSONT. R. S. ( 1978) Evidence for denitrification in aerobic pelagic sediments. Nafure 274, 354-356.