39. RARE EARTH, MAJOR, AND TRACE ELEMENT COMPOSITION …

Pisciotto, K. A., Ingle, J. C, Jr., von Breymann, M. T., Barron, J., et al, 1992Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 127/128, Part 1

39. RARE EARTH, MAJOR, AND TRACE ELEMENT COMPOSITIONOF LEG 127 SEDIMENTS1

Richard W. Murray,2 Marilyn R. Buchholtz ten Brink,3 Hans-J. Brumsack,4 David C. Gerlach,5 and G. Price Russ III5

ABSTRACT

The relative effects of paleoceanographic and paleogeographic variations, sediment lithology, and diagenetic processes onthe final preserved chemistry of Japan Sea sediments are evaluated by investigating the rare earth element (REE), major element,and trace element concentrations in 59 squeeze-cake whole-round and 27 physical-property sample residues from Sites 794, 795,and 797, cored during ODP Leg 127.

The most important variation in sedimentary chemical composition is the increase in SiO2 concentration through the Pliocenediatomaceous sequences, which dilutes most other major and trace element components by various degrees. This biogenic inputis largest at Site 794 (Yamato Basin), moderately developed at Site 797 (Yamato Basin), and of only minor importance at Site795 (Japan Basin), potentially reflecting basinal contrasts in productivity with the Yamato Basin recording greater biogenic inputthan the Japan Basin and with the easternmost sequence of Site 794 lying beneath the most productive waters. There are fewsystematic changes in solid-phase chemistry resulting from the opal-A/opal-CT or opal-CT/quartz silica phase transformations.Most major and trace element concentrations are controlled by the aluminosilicate fraction of the sediment, although the effectsof diagenetic silica phases and manganese carbonates are of localized importance.

REE total abundances (IREE) in the Japan Sea are strongly dependent upon the paleoceanographic position of a given sitewith respect to terrigenous and biogenic sources. REE concentrations at Site 794 overall correspond well to aluminosilicatechemical indices and are strongly diluted by SiO2 within the upper Miocene-Pliocene diatomaceous sequence. Eu/Eu* values atSite 794 reach a maximum through the diatomaceous interval as well, most likely suggesting an association of Eu/Eu* with thesiliceous component, or reflecting slight incorporation of a detrital feldspar phase. XREE at Site 795 also is affiliated stronglywith aluminosilicate phases and yet is diluted only slightly by siliceous input. At Site 797, Σ R E E is not as clearly associated withthe aluminosilicate fraction, is correlated moderately to siliceous input, and may be sporadically influenced by detrital heavyminerals originating from the nearby rifted continental fragment composing the Yamato Rise.

Ce/Ce* profiles at all three sites increase essentially monotonically with depth and record progressive diagenetic LREEfractionation. The observed Ce/Ce* increases are not responding to changes in the paleoceanographic oxygenation state of theoverlying water, as there is no independent evidence to suggest the proper Oceanographic conditions. Ce/Ce* correlates slightlybetter with depth than with age at the two Yamato Basin sites.

The downhole increase in Ce/Ce* at Sites 794 and 797 is a passive response to the diagenetic transfer of LREE (except Ce)from sediment to interstitial water. At Site 795, the overall lack of correlation between Ce/Ce* and L a / Y ^ suggests that otherprocesses mask the diagenetic behavior of all LREEs. First-order calculations of the Ce budget in Japan Sea waters and sedimentindicate that ~20% of the excess Ce adsorbed by settling particles is recycled within the water column and that an additional-38% is recycled at or near the seafloor. Thus, because the remaining excess Ce is only -10% of the total Ce, there is not a largesource of Ce to the deeply buried sediment, further suggesting that the downhole increase in Ce/Ce* is a passive response todiagenetic behavior of the other LREEs. The REE chemistry of Japan Sea sediment therefore predicts successive downholeaddition of LREEs to deeply buried interstitial waters.

INTRODUCTION

The chemical composition of sediments from a marginal oceanbasin records the geologic history of the basin and the source ter-rane(s) from a different perspective than the physical record providedby the sediments themselves. Subtle changes in the redox state of theoverlying water, for example, may be recognizable only in the chemi-cal distributions within the sediment column. Variations in the com-position of the sedimentary source and in the depositional position ofthe sediment with respect to metalliferous and terrigenous inputs alsoaffect the chemical composition of sediment. When a sedimentary

1 Pisciotto, K. A., Ingle, J. C, Jr., von Breymann, M. T., Barron, J., et al, 1992. Proc.ODP, Sci. Results, 127/128, Pt. 1: College Station, TX (Ocean Drilling Program).

2 Department of Geology and Geophysics, University of California at Berkeley,Berkeley, CA 94720, U.S.A. (Present address: Graduate School of Oceanography, Univer-sity of Rhode Island, Narragansett, Rl 02882, U.S.A.

3 Earth Sciences Department, Lawrence Livermore National Laboratory, Livermore,CA 94550, U.S.A.

4Geochemisches Institut, Goldschmidtstr. 1, D-3400, Gottingen, Federal Republicof Germany.

5 Nuclear Chemistry Division, Lawrence Livermore National Laboratory, Liver-more, CA 94550, U.S.A. (Gerlach, present address: Charles Evans and Associates, 301Chesapeake Drive, Redwood City, CA 94063, U.S.A.; Russ, present address: Chemistryand Material Sciences Department, Lawrence Livermore National Laboratory, Liver-more, CA 94550, U.S.A.

grain settles onto the basin floor, however, diagenetic processes beginto modify its chemistry, and the preserved signals of the sedimentarybasin's history may change with time. The degree of change willdepend on the susceptibility of the element(s) in question to mobilizeunder diagenetic conditions. In many cases the diagenetic reactionswithin the sediments provide important information regarding thebasin's geologic history.

The sedimentary sequences in the Japan Sea present a terrificopportunity to assess both the chemical sources to fine-grained sedi-ments and the post-depositional diagenetic processes that modify thesediment's primary chemical composition. The Japan Sea is a small,enclosed, marginal ocean basin in which the sedimentary sources anddepositional regimes are well characterized by lithostratigraphic se-quences recording 2000-5000 m of deposition (Fig. 1). Furtherresolution of the depositional environments is provided by the wealthof sedimentologic, biostratigraphic, and chemical data collected dur-ing Ocean Drilling Program (ODP) Legs 127 and 128. As such, thephysical stratigraphic framework is constructed within which chemi-cal variations may be placed. Moreover, Japan Sea sediments arecurrently undergoing many different stages of diagenesis, includingsulfate reduction, silica phase transformations, and carbonate disso-lution/recrystallization (Murray et al., this volume), and changes insediment chemistry may potentially be tied to these diagenetic proc-esses. In particular, the siliceous sequence provides an advantage

677

R. W. MURRAY ET AL.

45 N

40 =

140

Figure 1. Location map of the Japan Sea showing locations of Leg 127 Sites 794-797 and Deep Sea DrillingProject Leg 31 Sites 299-302. Bathymetry contoured in meters, a marks the approximate position of thestudy of REEs in settling particles near Site 794 (40.8°N, 138.7°E; Masuzawa and Koyama, 1989); * marksthe approximate position of the study of REEs in Japan Sea water (35.7°N, 133.2°E; Tanaka et al., 1990).Map modified from Tamaki, Pisciotto, Allan, et al., 1990.

when addressing large-scale diagenetic changes, as the diageneticstate of the sample can be tracked easily by the SiO2 phases as theyprogress through the successive opal-A/opal-CT/quartz diagenetictransitions.

This study attempts to resolve the lithologic and diagenetic pa-rameters that control the chemical composition of Japan Sea sedi-ments. Superimposed on both these effects are paleoceanographicchanges that have potentially modified the chemistry of Japan Seasediments through time. While paleoceanographic variations andpaleogeographic location each in turn influence sediment lithologyand diagenetic conditions through a given stratigraphic sequence, werestrict our definition of paleoceanographic variations to includethose large-scale changes in seawater composition that will substan-tially influence the rare earth element (REE) record in sedimentthroughout both Japan Sea basins. By tracking chemical behavioracross lithologic, diagenetic, and paleoceanographic boundaries, weaddress the relative importance of each potential influence under avariety of sedimentologic regimes.

Paleoceanographic Evolution of the Japan Sea

In order to place the chemical variations of the different lithologiesof sediments preserved in the Japan and Yamato basins within theproper stratigraphic context, the general paleoceanographic historyof the Japan Sea relevant to our chemical investigation is summarizedhere. Comprehensive presentations of the lithostratigraphic and pale-oceanographic changes are presented elsewhere in great detail(Tamaki, Pisciotto, Allan, et al., 1990; plus papers in this volume).

The preliminary paleoceanographic dates and sedimentation ratesdescribed herein and used throughout this paper are based on descrip-tions in Tamaki, Pisciotto, Allan, et al. (1990); more recent valuespresented in other papers in this volume supersede those used here.

Initial extension and rifting began in the Japan Sea in the early(?)Miocene to early middle Miocene and resulted in the gravity-flowdeposition of volcaniclastic sandstones and siltstones along the basinmargin. Accelerated deepening of the Yamato Basin during the earlyMiocene and of the Japan Basin during the early middle Miocene ischaracterized by fine-grained, diatomaceous and calcareous clay-richpelagic and hemipelagic muds. Bottom-water (500-1500 m) condi-tions were most likely oxic to dysaerobic throughout the entireMiocene, as indicated by foraminifer assemblages and bioturba-tion/lamination character. The three sites (794, 795, and 797) forwhich chemistry is reported here were either bathymetrically isolatedor too distal from terrigenous inputs during this time to receive coarseclastic deposits. Site 797 (Yamato Basin) may have been slightlydeeper than Sites 794 and 795. From the late Miocene to the Pliocene,an increase in diatom productivity resulted in hemipelagic diatoma-ceous and siliceous sedimentation. Bottom waters (1500-2500 m)were generally oxygenated and intrabasin circulation was well devel-oped, although communication between the Yamato and Japan basinsmay have been restricted during the middle late Miocene. After thesehighly productive conditions, bathyal hemipelagic muds were depos-ited from the Pliocene to the present. Sea-level fluctuations may havealso affected terrigenous inputs and resulted in periods of anoxic deepwater. Within the Quaternary section, preservation of dark-light cy-cles in silty clays may record fluctuations of terrigenous input, oxy-

678

RARE EARTH, MAJOR, AND TRACE ELEMENT COMPOSITION, LEG 127

genation level, type of organic matter, surface-water productivity, orsedimentation rate. These dark-light cycles are described extensivelyelsewhere in this volume.

Composite Lithostratigraphic Succession

The lithostratigraphic sequences that result from this paleoceano-graphic evolution are relatively similar at all three sites studied here(Fig. 2); details of the lithologic successions, terminology usage, andcriteria used to divide lithologic units are discussed elsewhere(Tamaki, Pisciotto, Allan, et al., 1990). Silica phase transformationshave a profound affect on sedimentary lithology at each site and arelocated on all chemical profiles. A composite downhole lithostrati-graphic sequence is described in the following, with key differencesamong sites highlighted. Note that the unit numbers do not correlateamong the sites (Fig. 2).

The uppermost -100 m below seafloor (mbsf) consists generallyof slightly to highly bioturbated gray clay, silty ashy clay, anddiatomaceous clay, which were deposited through the late Plioceneand Quaternary. Discrete ash beds (1-5 cm thick) are a volumetri-cally minor yet numerous component. Dark-light cycles tend to bemore notable in the uppermost tens of meters. Minor calcareouslayers (including diagenetic dolomite or ankerite, and manganesecarbonates) are found throughout. Diatom abundances decreasetoward the base of Unit I at Site 795 and increase toward the base ofUnit I at Site 797. Beneath this sequence are moderately to intenselybioturbated diatom and clayey diatom ooze. The age and duration of

deposition of this sequence is different from site to site (Fig. 2). Ashbeds (1-5 cm thick) are sparsely disseminated and very rare. Minordolomicrite and dolomitic nodules also occur, usually within concen-trations of foraminifers. At Site 794, the lowermost portion of thisdiatomaceous sequence (Subunit HB) consists of diatom clay similarto the purer diatom ooze except with less diatoms. Below the diatomooze is a middle to upper Miocene sequence consisting of moderatelyto intensely bioturbated diatomaceous clay and faintly laminatedsiliceous claystone, along with dolomite layers, pyrite, and ash inter-spersed throughout. Dolomite, ankerite, and silica cements are presentlocally in small nodules, stringers, lenses and laminae, as are Porcel-lanite and chert layers. The poor recovery through this section at Site795 may indicate a greater abundance of chert layers than was observedin the recovered samples. Thin carbonate layers are relatively commonat Site 797. Beneath the siliceous claystone at Site 794 are moderatelybioturbated claystone, silty claystone, and calcareous claystone, withan increasing diagenetic dolomite and ankerite component.

ANALYTICAL METHODOLOGY

Sample Preparation

Most samples in this study are splits of the whole-round squeeze-cake residues from the nylon- and Teflon-lined interstitial-watersqueezer described by Brumsack et al. (this volume). The sampleswere analyzed for their major element, trace element, and REEcomposition as described in the following. Not every squeeze cakewas prepared for geochemical analysis, but the continuity of the

Site 794Yamato Basin

Site 795Japan Basin

1 0 0 -

200-

E 300-

Q.CDQ 400-

500-

600

Unit ISilty clay

Unit IIDiatom

ooze(HB)

Diatomclay

Unit III

Siliceousclaystone

UnitVSilty calc.

clay

Unit ISilty clay

Unit IIDiatom

ooze

Unit IIIDiatomclaystone

Unit IV

Siliceousclaystone

Site 797Yamato Basin

Unit I

Silty clay

Unit IIDiatom siltclay ooze

Unit IIISilty clay

Unit IVClaystoneSil.Cly.Stn.

Unit VSiliceousclavstone

•2.0 Ma

5.1 Ma

11.3 Ma

Figure 2. Lithostratigraphic summary of Leg 127 sites analyzed for sedimentary chemistry. Note that the unit numbers do not correlate among thesites. Unit IV at Site 794 was not sampled in this study.

679

R. W. MURRAY ET AL.

Table 1. Major element data (wt%, volatile free) from Leg 127 squeeze-cake residues and physical-property samples.

Core, section,interval (cm)

127-794 A-

1H-3, 140-1453H-5, 140-1456H-4, 145-1509H-4, 145-150

12H-4, 145-15015H-4, 145-15022X-4, 145-15025X-4, 145-15029X-4, 145-15033X-2, 140-14536X-4, 145-150

127-794B-

12R-1, 145-15018R-4, 145-15024R-2, 140-145

127-795 A-

1H-4, 145-1502H-4, 145-1503H-4, 145-1504H-4, 145-1505H-4, 145-1506H-4, 145-1507H-4, 145-1508H-4, 145-1509H-4, 145-150

10H-1, 145-15012H-4, 140-14515H-4, 140-14518H-4, 140-14521X-4, 140-14525X-3, 140-14529X-2, 140-14530X-2, 145-15031X-4, 145-150

a34X-l, 100-102a34X-3, 100-102a34X-4, 100-10234X-5, 140-145

a34X-6, 100-102a34X-7, 39-40a35X-l, 100-10235X-2, 145-150

a35X-3, 100-102a35X-4, 100-102a35X-6, 100-102a37X-l, 99-10037X-1, 145-150

Depth(mbsf)

4.423.750.879.3

107.8136.3203.5232.5271.4309.2337.9

406.4468.5523.7

6.015.324.834.843.853.862.872.381.886.8

110.2138.7158.8187.5224.8262.4272.3285.1308.9311.9313.4315.3316.4317.3318.6320.6321.6323.1326.1337.9338.4

Subunit

IAIAIAIBIIAIIAIIAHBHBIIIAIIIA

IIIBIIIBV

IAIAIAIAIAIAIAIAIAIBIBIIIIIIII

IIIAIIIAIIIAIIIBIIIBIIIBIIIBIIIBIIIBIIIBIIIBIIIBIIIBIIIBIVAIVA

SiO2

62.064.362.766.970.279.280.871.665.367.366.5

79.974.364.3

64.164.564.662.366.166.965.863.364.565.467.770.971.477.371.474.673.375.273.972.475.574.075.374.774.876.377.073.470.280.576.9

A12O3

17.7

17.915.5

16.3

14.09.298.60

11.015.215.015.3

9.1611.112.6

17.217.2

15.1

17.015.815.3

17.0

16.916.417.2

14.613.2

13.310.813.1

11.811.911.311.1

12.2

11.712.312.1

11.611.011.310.312.1

13.59.1

10.8

CaO

1.01

1.710.71

0.71

0.730.45

0.48

1.110.68

0.630.67

2.17

0.432.42

1.27

1.081.08

1.201.501.15

1.01

1.060.86

0.931.16

0.831.060.72

0.89

1.09

0.930.980.95

1.000.980.88

0.84

1.10

0.810.720.86

0.960.98

0.620.66

MgO

3.25

2.993.41

2.99

2.792.04

2.20

4.144.74

4.134.52

1.984.764.88

3.073.15

3.373.01

2.352.85

3.093.07

3.15

3.052.97

2.882.512.02

2.83

2.35

2.482.122.22

2.58

2.612.39

2.801.84

2.30

2.172.012.44

2.70

1.682.02

Na2O

3.74

2.74

2.82

2.80

3.012.602.17

2.49

3.062.14

2.50

1.41

1.451.61

2.792.54

2.952.64

3.683.02

3.12

2.982.992.82

2.87

2.482.622.40

2.572.58

2.55

2.493.823.27

2.512.75

2.413.46

3.462.362.95

2.893.042.142.37

K2O

3.53

3.62

3.333.52

2.902.02

1.702.24

3.073.42

3.41

1.72

2.984.19

3.36

3.563.103.213.41

2.99

3.273.22

3.30

3.42

2.812.522.482.07

2.532.08

2.23

2.072.20

2.292.002.27

1.952.42

2.181.93

1.912.19

2.501.692.02

Fe 2O 3

7.87

5.84

10.4

5.86

5.393.77

3.42

5.09

6.796.305.71

3.18

4.336.62

7.03

6.92

8.589.626.37

7.005.71

7.85

7.856.09

7.006.455.84

4.115.874.84

5.795.125.13

5.364.03

4.75

3.904.25

4.77

4.584.295.26

6.153.734.51

MnO

0.15

0.100.24

0.180.370.17

0.21

1.83

0.490.390.72

0.08

0.132.75

0.39

0.15

0.510.210.180.12

0.24

0.730.17

0.110.24

0.09

0.100.09

0.16

0.09

0.100.09

0.09

0.09

0.080.08

0.080.08

0.08

0.080.070.08

0.080.050.07

TiO2

0.71

0.70

0.76

0.690.580.39

0.350.44

0.610.61

0.58

0.360.420.48

0.730.75

0.590.71

0.590.63

0.72

0.730.71

0.750.61

0.550.550.47

0.580.52

0.62

0.570.55

0.630.530.57

0.56

0.500.54

0.54

0.510.61

0.690.460.51

P 2 O 5

0.130.120.110.080.080.060.050.070.080.100.07

0.060.080.15

0.140.130.110.140.100.120.110.110.100.120.100.080.080.070.080.070.090.080.080.090.070.080.060.080.070.080.080.090.090.070.07

profiles was assured. Samples were split aboard JOIDES Resolutionby cutting with a knife down the longitudinal axis of the compressedcake. An additional suite of physical-property residue samples fromSites 795 and 797 were split aboard the ship and analyzed for majorand trace elements (no REE) as described in the following. On shore,after oven-drying at ~150°C for 3-5 hr, all samples were powderedby hand in an agate mortar and pestle.

Analytical Procedure and Uncertainty

Approximately 20 g of squeeze-cake residue was powdered andsplit for REE and major/trace element analysis. Major and traceelements were analyzed by X-ray fluorescence (XRF) at X-Ray AssayLaboratories, Don Mills, Ontario, Canada. REEs were analyzed byICP-mass spectrometry (ICP-MS) at Lawrence Livermore NationalLaboratory. Complete sample dissolution prior to introduction intothe ICP-MS was achieved by HNO3-HF-HC1O4 microwave digestion

in sealed Teflon beakers, as detailed elsewhere (Murray et al., 1991).Samples were prepared and run in random order (i.e., mixed sites anddepths) on different analytical runs, days, and ICP-MS instruments.Values of Ba determined from XRF were used to correct for ionizationsuppression of REEs (average -30%) during ICP-MS analysis; theaccuracy for REEs therefore is as good as the Ba XRF values(5%-<10%). Replicate analyses of solutions indicate that the preci-sion of the Ba ionization correction is ±4% for light REEs (LREEs;La to Nd) and ±8% for heavy REEs (HREEs; Er to Lu), even thoughthe % suppression is nonreproducible and may vary between runs ofthe same solution on different days by 10%-20%. As determined byreplicate analyses of USGS reference rocks MAG-1, SCo-1, andSGr-1 (Taylor and McLennan, 1985; Jarvis and Jarvis, 1985), thegeneral precision of the ICP-MS analysis is ~<5% for LREEs and5%-10% for HREEs (Murray et al., 1991). Gd is systematicallyhigher (18%) in the reference rocks, which is most likely due tointerferences on 157Gd from 138BaF+ or 140CeOH++ (Murray et al,

680


Table 1 (continued).


127-795B-

1R-1, 140-145a 9R-2,111-113"11R-3, 81-83a12R-3, 96-9813R-2, 140-145

"13R-3, 63-65a14R-3, 127-12919R-1, 140-145

127-796A-

2H-4, 145-1505H-5, 145-1509X-2, 145-150

11X-1, 145-150

127-797B-

1H-2, 145-1502H-4, 145-1503H-5, 140-1454H-4, 145-1505H-4, 145-1506H-4, 140-1457H-4, 145-1508H-4, 145-1509H-4, 140-145

10H-4, 145-15012H-4, 145-15015H-4, 145-15018H-4, 145-15021X-4, 140-14524X-5, 140-14527X-6, 140-14528X-1, 145-150

a30X-l , 99-10130X-1, 145-150

a31X-2, 98-100a31X-5, 96-98a32X-2, 98-10032X-3, 145-150

a32X-5, 99-100a33X-l , 100-10133X-2, 145-150

a33X-3, 100-10134X-2, 145-150

a 42X-1 ,47-49a 44X-1,29-31a 46X-1 ,40-42a47X-l , 107-109a48X-l , 100-102

Depth(mbsf)

366.6445.0464.0475.4483.9484.6495.0540.1

9.239.261.879.8

3.011.922.830.940.449.859.468.978.387.9

106.9135.4163.9191.4221.7252.3254.6273.5274.0284.7289.2294.4296.4298.9302.5304.5305.5314.2388.4407.6427.0437.3446.9

Subunit

IVAIVBIVBIVBIVBIVBIVBIVB

IAIAIBIB

IAIAIAIAIAIAIAIAIAIBIBIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

IVAIVAIVAIVAIVBIVBVVV

SiO2

67.388.172.573.470.972.263.479.0

65.362.167.167.6

62.660.565.057.662.565.165.067.464.966.666.964.159.875.175.369.367.163.268.669.164.265.065.865.263.864.163.662.982.988.484.575.381.9

A12O3

15.25.7

13.813.514.514.119.011.2

17.215.616.015.9

17.317.318.016.417.817.717.416.117.816.015.917.29.93

12.111.114.414.615.714.513.015.415.116.215.415.816.716.717.57.35.26.9

11.87.5

CaO

0.840.380.840.760.960.702.070.64

1.896.621.611.74

1.191.310.769.621.010.710.810.740.960.601.030.585.470.390.430.570.680.911.041.011.171.250.940.810.890.690.970.940.360.340.420.520.45

MgO

2.670.902.272.152.212.162.361.58

2.752.472.542.55

3.282.933.133.303.062.873.252.822.083.271.933.643.302.712.753.534.374.433.653.384.484.064.104.194.584.363.823.872.081.751.933.393.61

Na2O

2.381.171.961.821.971.952.311.33

3.002.642.802.97

2.652.742.372.203.002.742.512.593.712.473.882.681.702.052.052.512.544.232.404.023.473.872.653.173.092.723.052.481.501.241.561.741.43

K2O

2.971.152.562.642.732.711.812.11

2.872.952.902.68

3.683.433.783.293.713.453.663.174.143.364.023.372.052.392.192.622.742.962.592.283.002.632.862.933.053.293.393.931.561.061.542.621.72

Fe2O3

7.842.265.285.036.085.398.213.69

6.096.816.205.66

8.3010.85.856.477.816.416.336.125.486.674.987.214.284.565.276.036.817.386.206.396.826.856.447.167.557.047.376.833.781.582.733.942.96

MnO

0.080.030.080.080.070.070.060.03

0.090.110.170.10

0.090.110.090.330.350.280.230.300.150.180.880.43

12.90.200.280.260.350.290.250.210.550.260.240.230.310.320.240.810.110.060.080.120.13

TiO2

0.680.270.580.570.640.580.630.46

0.700.640.670.70

0.740.770.830.710.710.660.700.640.660.720.520.690.370.520.520.590.700.740.650.580.760.720.690.710.750.700.740.720.320.240.310.450.31

P 2 O 5

0.100.040.070.070.070.060.140.05

0.140.130.100.10

0.150.160.140.130.120.100.110.090.120.090.090.090.170.070.080.080.100.120.140.130.140.200.140.120.110.100.140.130.080.060.080.080.07

Note: Analyses by XRF at X-Ray Assay Laboratories, Don Mills, Ontario, Canada.a Physical-property residue sample.

1991), although visual inspection of shale-normalized plots of datafrom this study do not show a systematic Gd enrichment. Final LREEconcentrations agree well (±4%) with previous studies of other JapanSea sediments determined by instrumental neutron activation analysis(Masuzawa and Koyama, 1989); HREE data are difficult to compareas a result of the non-shalelike Tb, Yb, and Lu ratios presented in thedata table of Masuzawa and Koyama (1989), as well as the lack ofHREE agreement between their table and shale-normalized REE plot.

REE Terminology

REE abundances in marine sediments commonly are normalizedto average shale values to minimize the effect of the odd-evenvariability in elemental abundances. Normalization values used here(Table 1) are an average of North American, European, and Russian

shale composites used by previous workers; note that Tb values fromthis composite are probably too high, causing an apparent Tb deple-tion in the shale-normalized REE plot (Murray et al., 1991). Tm andLu values may also be high (R. L. Korotev, pers. comm., 1991).Nevertheless, we use these values because they are widely used bymarine chemists (see discussion in Sholkovitz, 1988) and are there-fore of most use for comparing results between studies of oceanwaters and marine sediment. Moreover, use of the North AmericanShale Composite (NASC) for ICP-MS REE data is handicappedbecause not all HREEs in NASC are analyzed by a common technique(Gromet et al., 1984). Shale values used here for normalizationpurposes do not necessarily represent the average REE terrigenousinput pattern (Sholkovitz, 1988, 1990; Condie, 1991). We calculatethe Ce anomaly using the ratio Ce/Ce* = (Cesample/Ceshale)/Ce*, whereCe* is the predicted value obtained by linear interpolation between

681

R. W. MURRAY ET AL.

shale-normalized La and Pr values. The Eu anomaly (Eu/Eu*) iscalculated similarly, using values of shale-normalized Sm and Gd.Ce/Ce* and/or Eu/Eu* ~l does not imply flatness across any portionof the shale-normalized REE spectrum, including the Ce and Euregions. Precision of Ce/Ce* and Eu/Eu* is estimated to be -5%, withaccuracy estimated at ~<10% for Ce/Ce* and higher for Eu/Eu*because of potential interferences on Gd (<15%, see previous discus-sion). Contrasts in LREE vs. HREE behavior are quantified by theratio Lan/Ybn = (Lasample/Lashale)/(Ybsample/Ybshale).

RESULTS AND DISCUSSION

Major element (volatile free), REE, and trace element data for allLeg 127 sites are reported in Tables 1 and 2, and the respectiveconcentration profiles are also presented for Sites 794, 795, and 797(Figs. 3 through 5). Concentration data from Site 796 are presentedin Tables 1 and 2 but are not discussed, as only four samples wereanalyzed. Correlation coefficient values (r) between various elementsat each site are listed in Tables 3 through 5; except where noted, listedvalues are for correlations through the entire sequence and may masklocal associations not visible at discrete intervals. Throughout thediscussion, we often reference data to Al because it is primarily froma terrigenous source and can therefore be used to monitor varyingterrestrial inputs and because Al is relatively insoluble and immobileduring diagenesis (e.g., Hower et al., 1976; Boles and Franks, 1979).Results are presented here on a site-by-site basis in order to synthesizethe different styles of chemical behavior responding to the lithologicand diagenetic influences through a common stratigraphy. Site 794 isdiscussed in the greatest detail, as many of the aspects of chemicalbehavior at Site 794 are observed at Sites 795 and 797 and need notbe repeated. The site presentations are followed by discussion of theREE compositions of Japan Sea sediments and their implications forthe sedimentary history of the Japan Sea.

The downhole profiles of the raw concentration data allow assessmentof the broad associations of particular elements to different lithologies.For example, as will be shown in the following, SiO2 concentrations arecommonly associated with diatomaceous sequences. The Al-normalizedprofiles show how an elemenfs concentration changes with respect toterrigenous input. If a given Al-normalized profile records elementalbehavior that cannot be related to the inferred terrigenous or siliceousinputs, then the element must be responding to a third parameter, forexample, phosphate. In all profiles, significant deviations (±15%) fromthe average, especially if part of a larger trend, may be interpretable inthe context of stratigraphic heterogeneity controlled by local parameters.In some cases, comparison between different profiles on a point-by-pointbasis elucidates the causes behind the perturbations. In other cases,however, sharp and isolated increases or decreases in concentration arenot associated between elements at the same site, which limits successfulinterpretation.

Site 794

Major Elements

The most important variation in major element concentrations atSite 794 is the varying SiO2 concentration, which dilutes concentra-tions of the other major elements to different degrees (Fig. 3A andTable 3). This is seen clearly, for example, through the diatom oozeand diatomaceous clay of Unit II, where SiO2 concentrations reach amaximum of 80.8 wt% in Subunit IIA, before decreasing through thesiliceous clays of Subunit HB. SiO2 concentrations remain essentiallyconstant at -66 wt% through the opal-A/opal-CT diagenetic transi-tion. The elevated SiO2 values below the opal-CT/quartz transitionwithin Subunit IIIB may reflect the increasing amount of silicacements, stringers, and lenses in Subunit IIIB (Tamaki, Pisciotto,Allan, et al., 1990). Within the claystones of Unit V, SiO2 concentra-tions return to average shale values.

In addition to the expected Si/Al increase, the Na/Al ratio alsoincreases through Unit II (Fig. 3B). This profile may reflect a strongaffiliation of Na with the biogenic opal or an affiliated clay phase,with Na being adsorbed from either interstitial waters or seawater, assuggested by Donnelly and Merrill (1977), or perhaps may record aninherited chemical signature from detrital inputs (e.g., Plagioclasefeldspar) that are not uniform through time. We will discuss these andother potential causes further, in the context of the Eu/Eu* data.

Other Al-normalized major element profiles show downhole vari-ation at Site 794 (Fig. 3B). K/Al, Fe/Al, Mn/Al, Ca/Al (excepting onepoint), Mg/Al, and P/Al increase from Unit III to Unit V, the sameinterval through which Si/Al decreases, implying response to theconcentration of the siliceous component. Alternatively, these in-creases may reflect the diagenetic formation of carbonate/phosphatewithin and below Unit III. Fe/Al also shows a relative maximumwithin the diatom clay of Subunit HB. The overall slight Mg/Alincrease from the seafloor through the diatomaceous ooze of SubunitIIA is consistent with an additional Mg component potentially record-ing adsorption onto biogenic opal (e.g., Donnelly and Merrill, 1977).

Rare Earth Elements

The REEs are affected by increases and decreases in the biogenicSiO2 component at Site 794, as shown by r = -0.93 between ZREEand SiO2 (Table 6) and visually apparent from the pronounced ZREEminima through the siliceous Subunit IIA and EREE increases fromUnit III to Unit V (Fig. 4 and Table 2). IREE/A1 values are essentiallyconstant through Unit II, indicating an association of REEs withaluminosilicate phases. This is also suggested by high r values be-tween IREE and A12O3, TiO2, K2O, Fe2O3, and P2O5, which tend tobe present in shalelike relative abundances (e.g., Gromet et al., 1984).As such, all may be responding to the SiO2 dilution.

Ce/Ce* values are all positive and increase consistently fromvalues of ~ 1.08 for shallow samples to a maximum of Ce/Ce* = 1.61within the siliceous claystone of Subunit IIIA (Fig. 4). The near-sur-face values are consistent with the Ce/Ce* value -1.07 reported byMasuzawa and Koyama (1989) from surficial sediment 50-100 kmnorth of Site 794 (Fig. 1). There is no change in trend through thediatomaceous interval of Unit II, nor is there a change in value acrossthe opal-A/opal-CT diagenetic transition. Ce/Ce* decreases sharplyto Ce/Ce* = 1.15, however, across the opal-CT/quartz transitionbetween Subunits IIIA and IIIB. As discussed previously, SiO2 con-centrations tend to increase through this interval as well, whichsuggests that the Ce/Ce* decrease is a diagenetic change, with thenewly precipitating biogenic quartz phases recording a lower Ce/Ce*value than the primary sediment. The relatively high r value betweenCe/Ce* and MnO (0.81; Table 6) suggests that the excess Ce isassociated with Mn.

Eu/Eu* varies only slightly through the sedimentary column, averag-ing 1.03 ± 0.05, except for a large and systematic increase within thediatom ooze of Unit II, where Eu/Eu* peaks at 1.28 (Fig. 4). Thecorrelation coefficient of r=0.69 between Eu/Eu* and SiO2 suggeststhat Eu/Eu* is affiliated with the biogenic siliceous component, whichis controlling the negative r values between Eu/Eu* and the remainingelements (Table 6). There is essentially no change in Eu/Eu* acrosseither the opal-A/opal-CT or opal-CT/quartz diagenetic boundary.The single high Eu/Eu* point in Unit III corresponds with largedecreases in CaO and Sr and with an increase in Lan/Ybn; no othermajor or trace elements deviate significantly in that sample from theoverall trend.

Shale-normalized Eu anomalies have been reported only frommarine waters affected by aeolian input (Eu/Eu* < 1) or hydrothermalinput (Eu/Eu* > 1; Elderfield, 1988). Neither of these potential inputscan be responsible for the observed positive Eu anomalies throughthe diatomaceous Unit II sequence as (1) the Eu anomaly peak at Site794 is positive, not negative, and (2) there clearly is no independent

682


evidence of hydrothermal input into the sequences (e.g., metalliferoussediment or Ce/Ce* « 1, and Fe2O3 averages a low 5.75 ± 1.90).Eu/Eu* is not likely to respond to diagenetic influences within thesediment column (see discussion in McLennan, 1989). The relation-ship between Eu/Eu* and SiO2 implies that the siliceous componentis carrying a slightly positive Eu anomaly, although this would requireEu fractionation during construction of diatom tests or during adsorp-tion from seawater and/or incorporation of a Eu-enriched phase intoor onto diatom tests. A slight relative increase in the primary detritalfeldspar component could also cause an inherited positive Eu/Eu*anomaly in the bulk sediment, even if the feldspars have been diage-netically altered, due to the high Eu/Eu* of Plagioclase and alkalifeldspars (e.g., Eu/Eu* ~7^K); Dymek and Gromet, 1984; Taylor andMcLennan, 1985). The coincidence of the peaks in Eu/Eu* and Na/Almay also be consistent with the presence of a detrital feldspar com-ponent in this interval, although the small feldspar component impliedby the slight bulk Eu/Eu* increase (from ~l to -1.3) relative tofeldspar Eu/Eu* (-7-40, as noted previously) may not supply suffi-cient Na to generate the Na/Al increase. Although the fact that the Euenrichment occurs only at Site 794 (where the biogenic input is mostpronounced) argues for a siliceous influence, we cannot rigorouslyresolve these potential mechanisms given the current data base, andwe hope that additional information about this interval will be gainedfrom mineralogical studies at this site (e.g., Fagel et al., 1991).

Lan/Ybn values, which indicate the relative behavior of LREEs toHREEs, remain in the range -0.97 to -1.34 through Units I and II(Fig. 4), only slightly less than the estimated average terrigenous inputof Lan/Yb,, of -1.3 suggested by Sholkovitz (1990) and generallyagreeing with the value of ~ 1 suggested by Condie (1991). Within thediatomaceous ooze of Subunit IIA, La/Ybn values reach a relativeminimum of 0.97 and increase through the diatomaceous clay ofSubunit HB to La/rb,, = 1.34. Below Unit II, Lan/Ybn values showmore scatter, yet on average are lower in the siliceous claystones ofUnit III and the silty calcareous clay of Unit V. The low La,/Ybn valueof 0.69 is coincident with high CaO and Ca/Al. The potential for acausal link between the changes in Lan/Ybn and Ce/Ce* below theopal-A/opal-CT transition at Site 794 will be addressed later.

Trace Elements

Cr, Rb, Y, Zr, and Nb concentrations all have similar downholeprofiles that decrease through Subunit IIA (Fig. 5A); concentrationsalso correlate negatively with the SiO2 concentration (Table 3) indi-cating that these elements are not associated with the biogenic sedi-mentary component. Ba is enriched slightly through Subunit IIA,although Sr values are within the same range as Sr values in the siltyclays and siliceous claystones above and below the interval.

The essentially constant Al-normalized profiles of Rb, Zr, and Nb(Fig. 5B) further suggest that these elements are affiliated predomi-nantly with the aluminosilicate phase. Most r values of the traceelements are greatest with A12O3 and TiO2 (Table 3). Of these ele-ments, however, Y/Al is enriched in the clay-rich diatomaceousSubunit HB, and Nb/Al seems slightly elevated through the siliceousclaystone of Unit III. Cr/Al may be slightly higher through thediatomaceous interval of Unit II and also through portions of thesiliceous claystones of Unit III, implying that a portion of the Cr iscarried by the siliceous component. Sr/Al and Ba/Al clearly areenriched through the diatomaceous intervals, confirming their bio-genic associations.

Site 795

Major Elements

The diatom-rich sequences at Site 795 (Units II and III), althoughrelatively high in SiO2, do not show the well-developed SiO2 maxi-mum that is present at Site 794 (Fig. 3A). Instead, SiO2 concentrationsgradually increase from near-surface values of -64 wt% to a maxi-

mum of-77 wt% at the opal-A/opal-CT transition (-325 mbsf). Thesemaximum concentrations are slightly lower than the maximum SiO2

observed at Site 794 (-81 wt%). The SiO2 scatter through Unit IV isaffected by sampling resolution and reflects the increasing abundanceof diagenetic Porcellanite and chert sequences, which commonlycontain zones of elevated SiO2 concentrations. The remaining majorelements show absolute concentrations that generally follow Al andare diluted by the SiO2 component (Fig. 3A and Table 4).

Si/Al also shows a only slight increase through the ooze (Fig. 3B),consistent with the minimal SiO2 increase described previously. Be-low the opal-A/opal-CT transition, Ti/Al, Na/Al, Fe/Al, and Mg/Alall decrease to minimum values at depth (Fig. 3B), implying somediagenetic behavior, although the specific relationships are unclear.The Na/Al profile does not show the pronounced maximum throughthe siliceous interval that was recorded at Site 794 (Unit II at bothsites). Al-normalized Ca, P, and, most noticeably, Fe profiles eachshow an increase from the diatom ooze of Unit II to a relativemaximum through the diatom claystones of Unit III (Fig. 3B).

Rare Earth Elements

Σ R E E concentrations at Site 795 decrease from -190 to -160 ppmthrough Unit I, decrease further to -115 ppm in the diatom ooze ofUnit II, and decrease only slightly to the base of Unit III, where ZREEis -100 ppm (Fig. 4). ZREE appears higher in the siliceous claystonesof Unit IV, although only a few samples were analyzed. The highΣ R E E and ZREE/Al at 43.7 mbsf are coincident with an anomalouslylow Eu/Eu* as well as high Na2O and Zr concentrations. Interelementcorrelations (Table 6) show that the REEs are only slightly diluted bythe siliceous biogenic component and are closely affiliated with thealuminosilicate component (A12O3, r=0.92; K2O, r = 0.94). The otherhigh r values (e.g., P2O5) also reflect the aluminosilicate input, as theAl-normalized major element ratios are similar to that of averageshale, as at Site 794.

The Ce/Ce* profile increases from values ranging between 0.98 and1.11 in Unit I to higher values (-1.30) at depth in Unit IV (Fig. 4). Thereis no change in Ce/Ce* across either the opal-A/opal-CT or opal-CT/quartz diagenetic transition. Correlation coefficients, however, sug-gest a weak and negative association with virtually all components exceptSiO2, most likely reflecting a SiO2 dilution effect (Table 6). Note,however, that unlike at Site 794 Ce/Ce* does not correlate with MnO.Most Eu/Eu* values range between 0.90 and 1.08 and are essentiallyuniform (within error; Fig. 4). The anomalously low Eu/Eu* in Unit Icorresponds with high ZREE, as described previously. Unlike at Site 794,there is no Eu/Eu* maximum within the diatom ooze of Unit II (nor isthere a concomitant Na or Na/Al increase).

The Lan/Ybn profile (Fig. 4) can be divided into three sections: (1)increasing from -1 to -1.25 through Unit I claystones, (2) decreasingto a broad relative minimum of-0.84 through the diatom ooze of UnitII and diatomaceous claystone of Unit III, and (3) increasing again to1.26 through Unit IV siliceous claystones. Despite the continuity ofthe profile, Lan/Ybn correlates only moderately (and negatively) withSiO2 and Eu/Eu* (Table 6). The low La/rb,, (= 0.49) at the base ofUnit I is coincident with high Eu/Eu* and low Ce/Ce*, as well as highMn/Al and Ca/Al, and as such probably reflects REE fractionationwithin a manganese carbonate.

Trace Elements

Some trace element profiles at Site 795 (Fig. 5A) are generallysmooth through the entire sequence (e.g., Sr, Zr), although othersshow scatter around a constant trend (e.g., Nb). Chromium concen-trations double within Unit III to values between 120 and 160 ppm.This is the same interval through which Ca/Al, P/Al, and Fe/Alprofiles increased as well (see the preceding). Sr/Al and Ba/Al, whichat Site 794 showed a clear siliceous influence, record only an ex-tremely slight increase through Unit II (Fig. 5B). Correlation coeffi-

683

Table 2. Rare earth element and trace element data (ppm) from Leg 127 squeeze-cake residues and physical-property samples.


dShale

127-794A-

1H-3, 140-1453H-5, 140-1456H-4, 145-1509H-4, 145-150

12H-4, 145-15015H-4, 145-15022X-4, 145-15025X-4, 145-15029X-4, 145-15033X-2, 140-14536X-4, 145-150

127-794B-

12R-1, 145-15018R-4, 145-15024R-2, 140-145

127-795A-

1H-4, 145-1502H-4, 145-1503H-4, 145-1504H-4, 145-1505H-4, 145-1506H-4, 145-1507H-4, 145-1508H-4, 145-1509H-4, 145-150

10H-1, 145-15012H-4, 140-14515H-4, 140-14518H-4, 140-14521X-4, 140-14525X-3, 140-14529X-2, 140-14530X-2, 145-15031X-4, 145-150

b34X-l, 100-102b34X-3, 100-102b34X-4, 100-10234X-5, 140-145

b34X-6, 100-102b34X-7, 3 9 ^ 0b35X-l, 100-10235X-2, 145-150

b35X-3, 100-102b35X-4, 100-102b35X-6, 100-102b37X-1,99-10037X-1, 145-150

La

41

33.733.338.934.825.322.413.620.629.630.428.6

16.123.540.0

34.940.035.532.243.731.733.130.032.533.626.021.321.618.423.519.017.916.7n.a.n.a.n a19.0n.a.n.a.n.a.17.2n.a.n.a.n a.n.a.15.9

Ce

83

81.677.486.881.966.056.645.263.493.690.1

102.6

40.659.794.8

82.784.481.474.795.674.7

74.866.373.272.856.449.357.143.958.247.247.9

45.0n.a.n.a.n.a.46.8n.a.n.a.n.a.47.4n.a.n.a.n.a.n.a.43.4

Pr

10.1

9.489.58

10.39.006.935.994.076.088.338.628.50

4.586.51

10.2

9.4511.29.358.39

11.38.588.677.818.558.868.295.926.324.796.395.245.274.95n.a.n.a.n.a.5.42n.a.n.a.n.a.5.07n.a.n.a.n.a.n.a.4.68

Nd

38

32.229.436.131.323.520.614.120.928.529.329.1

15.421.637.3

31.638.232.829.337.030.229.727.229.4

29.628.220.221.017.522.018.017.116.7n.a.n.a.n.a.17.7n.a.n.a.n.a.17.2n.a.n.a.n.a.n.a.15.8

Sm

7.5

6.375.657.016.294.523.95

Eu

.61

.32

.27

.62

.39

.04

.052.90 0.824.04 0.955.715.906.37

.27

.27

.56

3.21 0.704.397.84

6.548.62 :6.736.137.486.456.425.646.266.018.413.20 (3.36 (3.88 (5.053.723.523.46n.a.n.a.n a3.86n.a.n an.a.3.54n an.a.n an.a.3.20

.08

.81

.46..05.45.37.27.45.40.26.33.31

1.06).69).67).94.11

191).87).86n.a.n.a.n a187n.a.n.a.n.a.J.85n a.n.a.n an.a.J.78

Gd

6.35

5.195.166.155.344.433.492.613.555.164.876.37

2.843.707.39

6.427.116.245.977.586.576.425.625.845.777.283.443.103.614.793.663.653.64n.a.n.a.n a3.88n.a.n.a.n.a.3.68n.a.n.a.n an.a.3.25

Tb

1.23

0.880.750.810.800.610.540.420.530.830.780.88

0.430.620.96

0.881.080.910.801.070.880.810.780.750.781.180.380.470.490.640.540.480.48n.a.n.a.n a0.54n.a.n.a.n.a.0.50n.a.n.a.n an.a.0.42

Dy

5.49

4.864.125.254.763.603.292.263.194.584.205.07

2.473.506.55

5.346.895.264.835.995.184.794.584.484.686.742.983.283.333.803.172.902.94n.a.n.a.n a3.42n.a.n.a.n.a.3.03n.an.a.n an.a.2.60

Ho

1.34 :

0.84 ;0.67 ;0.76 ;o.9i ;0.520.540.330.590.69 ;0.64 2

Er

.75

..29

..36

..43.52.77.53.15.71.17.12

0.92 2.90

0.46 .500.62 2.041.19 3.56

0.93 ;1.08 :1.04 ;0.92 ;1.23 :i.04 :0.97 :o.9i :0.84 :0.84 ;

.89

.55

.86..79.38

..68

..52

..64

..38>.54

1.32 4.440.500.590.610.75 :0.580.590.62n.a.n.a.n.a.0.78n.a.n.a.n.a.0.63n.a.n.a.n an.a.0.54

.68

.77

.75>.19.84.64.72

n.a.n.a.n a.96

n.a.n an.a..83

n an.a.n an.a..54

Tm

0.63

0.340.300.480.290.240.280.160.230.340.250.38

0.210.270.56

0.370.540.380.340.470.370.320.380.350.360.930.190.180.250.330.260.250.24n.a.n.a.n a0.33n.a.n an.a.0.26n an.a.n an.a.

0.23

Yb

3.51

2.432.172.482.231.651.611.201.812.341.942.52

2.011.853.67

2.793.662.442.553.262.522.332.392.242.284.541.141.431.57

2.151.841.681.70n.a.n.a.n a1.94n.a.n an.a.1.83n an.a.n an.a.1.56

Lu

0.61

0.340.300.290.290.24

0.230.140.230.310.280.42

0.210.190.62

0.370.650.410.380.490.360.360.380.310.330.760.160.180.24

0.330.280.260.27n.a.n.a.n a0.30n.a.n.a.n.a.0.28n an.a.n an.a.

0.24

ΣREE

204

18217219918214012289

128183181196

91129216

18720918717122017317315616817015711112110113110610499

n.a.n.a.n a107n.a.n.a.n.a.103n.a.n.a.n an.a.94

Ce/Ce*

1.00

1.121.061.071.141.221.201.481.381.461.361.61

1.151.181.15

1.110.981.091.111.051.111.081.061.081.030.931.081.191.151.161.161.201.21n.a.n.a.n.a.1.13n.a.n.a.n.a.1.24n.a.n.a.n an.a.1.23

Eu/Eu*

1.00

0.981.011.061.030.991.211.281.081.001.021.05

1.001.151.02

0.971.120.960.970.720.950.930.960.94


La,/Ybn

1.00

1.191.311.341.341.311.190.970.97

1.081.340.97

0.691.090.93

1.070.941.251.081.151.081.221.071.24


Cr

959716592817261102928978

677772

899582977388988788103928990819592148160142153n.a.98n.a.1241399013515616112789

Rb

132132139132117966686116135122

759982

12813910012110396119102121

1388990759910597999410798

n.a.96n.a.869769911101068175

Sr

160188121184144144163146147116212

155128157

173155148176173171179153140185145121123116131138127109121141n.a.136n.a.160115109118135136108112

Y

19242724

141325311927

201926

153827

3820

11281719

1117241718

20n.a.29

n.a.18161913

37

Zr

152146140162105996979109126116

688095

14716812516124312011411214515811411895751189186867591n.a.112n.a.9178996210411277137

Nb

29172116191213

17

30

231015

2921253126191216262721

24

12132518

n.a.

n.a.

181518272417

Ba

975826121027501670260035802000192014503810

mo27803220

810740121013106831450881745117011709389061070924

9928919058249511010n.a.847n.a.10608978098719631090784823

MU

RR

AL

.

Table 2 (continued).

Core, section,

interval (cm) Ce Sm Eu Gd Tb Dy Ho Er Tm Yb I R E E Ct/Ce* Eu/Eu* LanA rbn Cr Rb Sr Ba

127-795B-

1R-1,b9R-2,

b HR-3,b12R-3,

13R-2,b13R-3,b14R-3,

19R-1,

127-796A-

140-145111-11381-8396-98140-14563-65127-129140-145

22.1n.a.n.a.n a24.9n.a.n.a.17.8

58.7n.a.n.a.n a74.5n.a.n.a.48.7

6.44n.a.n.a.n a7.42n.a.n.a.4.87

21.4n.a.n.a.n a24.8n.a.n.a.16.2

4.40n.a.n.a.n a4.97n.a.n.a.2.51

1.05n.a.n.a.n a1.16n.a.n.a.0.46

4.56n.a.n.a.n a4.79n.a.n.a.2.51

0.60n.a.n.a.n a0.63n.a.n.a.0.29

3.74n.a.n.a.n a3.67n.a.n.a.2.17

0.76n.a.n.a.n a0.74n.a.n.a.0.43

2.26n.a.n.a.n.a.2.11n.a.n.a.1.23

0.33n.a.n.a.n.a0.30n.a.n.a.0.16

2.24n.a.n.a.n.a.2.05n.a.n.a.1.21

0.34n.a.n.a.n.a.0.30n.a.n.a.0.17

129n.a.n.a.n.a.152n.a.n.a.99

1.20n.a.n.a.n.a.1.34n.a.n.a.1.28

1.00n.a.n.a.n.a.1.02n.a.n.a.0.79

0.84n.a.n.a.n.a.1.04n.a.n.a.1.26

95100918986855877

10855

11311512768106

1265512512613512823194

2711

19293018

13049111114123118128114

30

12

16231813

9457178198088008011290687

2H-4, 145-1505H-5, 145-1509X-2, 145-15011X1, 145-150

127-797B-

1H-2, 145-1502H-4, 145-1503H-5, 140-1454H-4, 145-1505H-4, 145-1506H-4, 140-1457H-4, 145-1508H-4, 145-1509H-4, 140-14510H-4, 145-15012H-4, 145-15015H-4, 145-15018H-4, 145-15021X-4, 140-14524X-5, 140-14527X-6, 140-14528X-1, 145-150

b30X-1,99-10130X-1, 145-150

b31X-2, 98-100

b31X-5, 96-98"32X-2, 98-10032X-3, 145-150

b32X-5, 99-100"33X-1, 100-10133X-2, 145-150

b33X-3, 100-10134X-2, 145-150

b42X-1,47^9

b44X-1,29-31b46X-1,40-42

b47X-l, 107-109

b48X-l, 100-102

24.642.625.823.9

37.831.844.233.746.035.736.535.296.232.181.036.620.328.923.831.729.4n.a.27.8n.a.n.a.n.a.32.1n.a.n.a.39.8n.a.33.4n.a.n.a.n.a.n.a.n.a.

56.696.055.556.6

85.270.110170.210682.582.081.417887.016092.751.383.279.990.9101n.a.92.3n.a.n.a.n.a.103.8n.a.n.a.120.6n.a.124.2n.a.n.a.n.a.n.a.n.a.

6.7312.47.746.81


23.349.227.723.1


5.2814.78.134.02


1.413.731.861.03


5.1114.37.743.35


0.732.361.170.53


4.3113.06.273.47


0.792.501.370.70


2.398.904.111.90


0.351.790.780.34


2.379.044.211.59


0.381.860.880.26


134272153128

189158227166235185188185385185334203116172160190198n.a.187n.a.n.a.n.a.211n.a.n.a.248n.a.231n.a.n.a.n.a.n.a.n.a.

1.081.020.961.08


1.161.101.001.21


0.890.400.521.29


82858897

114100108839693104906910263105588711489171167105144167171941561519615196116921167360

109100111116

157140146120130132148118171128139145711101071021361371051161281041121301401231381268167699465

231247129154

155156161347139180154135180207138179181154184182133175233210160211191198190228237193137108127130102

25323217

3213413635273229651650

10

29213013302226382943293736401213163611

126120128119

188166195133186172121130465136349132471331001331111051191031161101281071151341121435737627958

18

1715

1519151428211425107

971211182322

2214

39

112329

23151412

20

303025301410640

769835778576802172014508851080278017103190254030603960316023803910390033202510335030203350306031603260239031501320151019601850

Note: Analyses of non-REE trace elements by XRF at X-Ray Assay Laboratories, Don Mills, Ontario, Canada. Missing values < 10-ppm detection limit, n.a. = not analyzed for.aAverage shale values used for REE normalization (Haskin and Haskin, 1966; Sholkovitz, 1988).bPhysical-property residue sample.

σ

R. W. MURRAY ET AL.

Site 794, Yamato Basin

Unit IIDtm. siltycly. gpzβ

Unil IIISilty^lay

Unit IVCly. stn.Sil.cjy.stn

UnilVSil.cly.stn


o

100

200

300

400

500

600

0

100

200

300

400

500

600

Unil ISj[ty_clay

Unit IIDiatom

ooze

(dial, clay)

Unit III

Siliceousclaystonβ

Fe^O.

. 1 1 1 2 2 3

MnO CaO MgO

Site 795, Japan Basin

Unit I

Silty clay

' "Urfii (T "

Diatom

Unit III

Siliceousclaystonβ

0

100

200

300

400

500

MnO CaO


2

MgO

Unit I

Silty clay

Unit IIDtm. siltycty.ooze

Unit III ~S£ty_clay_

Unit IVCly. stn.Sil.cjy.stn.

UnilVSil.cly.stn

2 4 6 8 10

MnO

0 . 5 1.0 1.5

CaO

2 3

MgO

0.05 0.10 0.15

686

Figure 3. A. Downhole profiles of major element concentrations (wt%, volatile free) at Sites 794, 795,and 797. The siliceous component within the upper -300 m is defined clearest at Site 794. Note thesporadic high concentrations of SiO2 in Unit IV, Site 795. Isolated peaks of CaO, MgO, and P 2O 5 aremost likely due to discrete manganese carbonates and stringers (cf. Matsumoto et al., this volume). Seetext for expanded lithologic descriptions. A, CT, and Q refer to opal-A, opal-CT, and quartz, respectively.Note the change in depth scale for Site 797. B. Al-normalized major element concentration profiles atSites 794,795, and 797. Symbols as in Figure 3A. The siliceous interval at Site 794 is clearly delineatedby the Si/Al peak within Unit II. Note also the large Na/Al increase in Unit II at Site 794 and the greatlyenriched Si/Al values in Unit V at Site 797. Note scale changes between sites.


o

100

^ 2 0 0a| • 300

S" 400

500

600

Site 794, Yamato BasinUnit I

Silty_clay_

Unit IIDiatom

ooze

(dial day)

Unit III. Siliceous

claystone

UnitV

5 6 7 8 0.045 0.050 0.055 0.2 0.3 0.3 0.4 0.5Si/Al Ti/Al Na/AI K/Al


0.5 0.6 0.7 0.8Fe/AI

.01 .1 1 0.0250.050 0.075 0.2 0.4Mn/AI Ca/AI Mg/AI

0.006 0.008 0.010P/Al

Figure 3 (continued).

687

R. W. MURRAY ET AL.

o

100

200

3 0 0

400

500

600


yUnit II

Diatomooze

(dial clay)

Unit IIISiliceousdaystone

UnitV;». calc. dy.

100 200

ΣREE15 20 25 30 35

ΣREE/AI(×W4)

1.2 1.4 1.6

Ce/Cβ'

1.0 1.1 1.2

Eu/Eu'

0.8 1.0 1.2

La /Yb


16 18 20 22 24 26ΣREE/AI(X10<)

150200250300350 20 30 40

ΣREE ΣREE/AI (× w4)

1.2 1.4 1.6

Ce/Ce'

0.8 0.9 1.0

Eu/Eu'

1 2

La_/Yb

Figure 4. Downhole profiles of REE data at Sites 794, 795, and 797. Symbols as in Figure 3A. All Ce/Ce*values increase essentially monotonically with depth, with the exception of depths >400 mbsf at Site 794 (seetext). Note the large Eu/Eu* increase at Site 794 through Unit II; this accompanies the Na/Al increase shownin Figure 3B. EREE enrichments in Unit I at Site 797 are due to the heavy-mineral component. Note the changein depth scale for Site 797. Error bars are of representative maximum error.

cient values suggest a moderate affiliation of Rb with aluminosilicatesand of Sr with carbonates as well as aluminosilicates (Table 4).

Site 797

Major Elements

The SiO2 profile at Site 797 is intermediate in nature to the SiO2

profiles at Site 794 and Site 795, in that, like at Site 794, SiO2

records some relative increase in concentration through the diatomooze of Unit II (Fig. 3A). As at Site 795, however, the concentra-tion increase at Site 797 is not clearly delineated. SiO2 concentra-tions through Unit II reach a relative maximum only of -75 wt%(dry). The lowest value within the diatomaceous interval (SiO2 =59.8 wt% at 163.9 mbsf) reflects dilution by Mn, Ca, and P, mostlikely caused by a manganese carbonate nodule or stringer in theparticular sample (Fig. 3A). There is no discrete change in SiO2

concentration across the opal-A/opal-CT diagenetic boundary.The high concentrations near the base of Unit IV and into Unit Vreflect the increasing proportion of chert and Porcellanite at depth.Other major element concentrations respond to the SiO2 profile tovarying degrees (Fig. 3B and Table 5). With the exception of the163.9 mbsf data point (see the preceding), all of the major elementsare diluted through the diatomaceous sediments of Unit II andappear to behave relatively similarly downhole. MgO increaseswith SiO2 at depth, perhaps reflecting an association with thediagenetic SiO2 phases.

The Si/Al profile at Site 797 delineates the siliceous intervalswithin Unit II as well as the pronounced diagenetic Si enrichment at

depth (Fig. 3B). Through Unit II, this Si/Al profile is similar to thatat Site 794, although Si/Al ratios peak at ~6 at Site 797 and at ~8 atSite 794. Most other Al-normalized major element profiles do notpresent significant lithologic or diagenetic correlations. K/Al, how-ever, seems to increase just below the opal-A/opal-CT transition andthrough Units IV and V, implying an affiliation with a diageneticallyforming phase. Na/Al does not show the maximum within the diato-maceous interval that was observed at Site 794, although it doesincrease in the silty clay of Unit III (Fig. 3B). Ca/Al may also be higherin Unit III.

Rare Earth Elements

Σ R E E increases slightly through the silty clays of Unit I (Fig. 4).The two high IREE (385 ppm at 78.3 mbsf; 334 ppm at 106.9 mbsf)each correspond with low Eu/Eu* and high La,/Ybn, as well as withhigh Na2O, K2O, Y, Zr, and Nb, indicating a local heavy-mineralcomponent specific to these samples (although not zircon, as La/Y^would be low), which does not occur at either Site 794 or Site 795.Although Σ R E E may be slightly diluted through the uppermostdiatomaceous sequences of Unit II, the lowest Σ R E E at 163.9 mbsfreflects the influence of manganese carbonate, as mentioned pre-viously. Σ R E E increases continually from this relative minimum at163.9 mbsf to maximum values of -240 ppm just below the opal-A/opal-CT diagenetic transition. Σ R E E is not responding to SiO2

dilution (r -0), and r < 0.45 between Σ R E E and A12O3 is the lowestimplied aluminosilicate affiliation of all three sites (Table 6). These rvalues are markedly different from what is observed at either Site 794

688



100 200 300

Zr40 80

Nb1 2 3

Ba/1000

0

100

f 200

•E 300

g• 400Q

500

600


Silty clay' Unl IfDiatom

(dial clay)

' "Unit iii "Siliceous

daystonβ

10 14 17 20

Cr/AI14 16 18

Rb/AI20 30

Sr/Al Y/Al14 16 18 20

Zr/AI2 3 •

Nb/AI200 400 600

Ba/AI


10 20 30

Cr/AI14 18 22

Rb/AI20 30

Sr/Al2 3 4 5

Y/Al6 10 20 30 40

Zr/AI2 4 6 8

Nb/AI10 250 500 750

Ba/AI

Figure 5. A. Downhole profiles of trace element concentrations (ppm) at Sites 794, 795, and 797.Symbols as in Figure 3A. Note the change in depth scale for Site 797. B. Al-normalized trace elementconcentration (× I04) profiles at Sites 794, 795, and 797. Symbols as in Figure 3A. Note the change indepth scale for Site 797.

689

R. W. MURRAY ET AL.

Table 3. Interelement comparison using correlation coefficient (r) values, Site 794.

SiO2 A12O3 CaO MgO Na 2O K2O Fe 2 O 3 MnO a TiO2 P 2 O 5 Cr Rb Sr Y Zr Nb Ba

1.00

1.001.00

0.48 0.83 0.68 0.57 1.000.69 1.000.93 1.00

1.000.52 0.53 1.00

0.93 0.52 1.000.45 1.00

-0.46 1.00

Note: -r indicates negative correlation. -0.45 < r < 0.45 not shown (-95% confidence interval).a Determined without two nonrepresentative values from 232.5 and 523.7 mbsf (see text and Fig. 3).

SiO2

A12O3

CaOMgONa2OK2OFe 2 O 3

MnOTiO2

P 2 O 5

CrRbSrYZrNbBa

1.00-0.91

-0.55-0.52-0.90-0.87

-0.88-0.78-0.58-0.76

-0.70-0.77

1.00

0.650.800.72

0.950.610.500.90

0.550.910.49

1.00

0.60

0.58

0.62

1.00

0.52

0.69

0.63

0.67

1.000.69

0.730.83

0.66

0.560.69

1.00

0.840.680.840.71

0.570.68


SiO2 A12O3 CaO MgO Na 2O K2O Fe 2 O 3 MnO TiO2 P 2 O 5 Cr Rb Sr Y Zr Nb Ba

SiO2

A12O3

CaOMgONa 2OK2OFe 2 O 3

MnOTiO2

P 2 O 5

CrRbSrYZrNbBa

1.00-0.96-0.69-0.85

-0.86-0.92-0.57-0.89-0.87

-0.61-0.83

-0.69

-0.47

1.000.690.77

0.830.850.490.880.82

0.720.86

0.71

1.00

0.62

0.490.71

0.86

0.48

1.00

0.760.770.570.840.71

0.490.62

0.51

0.46

1.00

0.46

1.000.730.560.810.69

0.730.590.460.75

1.000.560.770.83

0.69

0.590.520.56

1.000.450.53

1.000.77

0.730.71

0.57

1.00

0.620.83

0.560.450.53

1.00

-0.47

1.000.620.480.46

1.001.00

0.58 1.001.00

0.54 1.00

Note: -r indicates negative correlation. -0.45 < r < 0.45 not shown (-95% confidence interval). EREE VS. accumulation rate (g/cm /k.y.) = -0.50.


SiO2 A12O, CaO MgO Na 2 O K2O Fe 2 O 3 MnO TiO2 P 2 O 5 Cr Rb Sr Y Zr Nb Ba

SiO ,Al-,0,CaOMgONa.O

K,0Fe\θ,MnO"TiO,

P2°5CrRbSrYZrNbBa

1.00-0.87-0.77-0.52-0.59-0.78-0.81

-0.83-0.73

-0.71-0.54

-0.64

1.000.67

0.640.940.82

0.930.66

0.890.470.590.86

1.00

0.740.600.74

0.700.89

0.59

1.00

0.53

0.58

0.56

0.49

1.000.570.59

0.640.530.480.660.460.45

0.56

1.000.680.450.800.57

0.89

0.690.86

1.00

0.880.73

0.730.48

0.73

1.001.000.76

0.840.540.450.80

1.00

0.61

0.59

1.00

0.570.45

1.00

0.620.79

1.00

0.62

1.000.52 1.00

1.001.00

Note: -r indicates negative correlation. -0.45 < r < 0.45 not shown (-95% confidence interval). Determined without nonrepresentative values for MnO (163.9 mbsf),CaO (30.9, 163.9 mbsf), P 2 O 5 (163.9, 294.4 mbsf), Zr-Nb (78.3, 163.9 mbsf), and Sr (30.9 mbsf) (see text and appropriate figures).

690


Table 6. Correlation coefficients (r) for REE data from Leg 127 squeeze-cake residues.

Site 794 Site 795 Site 797

SiO2

A12O3

CaOMgONa2OK2OFe 2O 3

MnOTiO2p

2 o 5

CrRbSrYZrNbBa

ΣREECe/Ce*Eu/Eu*'Lan/Yh

ΣREE

-0.930.79

0.61

0.930.800.480.760.75

0.470.68

0.680.71

1.00

-0.57

Ce/Ce*

0.81

-0.52

0.56

1.00

Eu/Eu*

0.69-0.69-0.46

-0.59-0.48

-0.62-0.54

-0.54

-0.66-0.49-0.57

0.57

1.00

Lan/Ybn

-0.490.63

-0.48

0.580.460.51

0.72

0.510.81

0.76

1.00

ΣREE

-0.460.920.680.830.530.940.730.520.810.87

0.700.82

0.71

1.00-0.62

Ce/Ce*

0.53-0.62-0.52-0.73-0.69-0.65-0.48

-0.49-0.66

-0.63

1.00

Eu/Eu*

1.00-0.69

Lan/Ybn

-0.68

1.00

Σ R E E Ce/Ce* Eu/Eu*

0.810.55

0.70

0.62

-0.47

-0.51

-0.47

-0.50

-0.730.650.62

0.75

1.00

-0.460.46

0.75

-0.62

0.62

-0.510.50

-0.570.450.77

-0.60

1.001.00

1.00 0.53 -0.73 1.00

Note: - r indicates negative correlation. -0.45 < r < 0.45 not shown (-95% confidence interval). All data from Site 797 determined withoutnonrepresentative values for MnO (163.9 mbsf), CaO (30.9, 163.9 mbsf), P 2 O 5 (163.9, 294.4 mbsf), ^REE-Zr-Nb (78.3, 106.9 mbsf), Eu/Eu*(78.3 mbsf) Lan/Ybn (78.3, 106.9, 163.9 mbsf) and Sr (30.9 mbsf) (see text and appropriate figures).

a Determined for Site 794 without nonrepresentative values from 232.5 and 523.7 mbsf (see text).b Determined for Site 795 without nonrepresentative values from 110.2 mbsf (see text).

or Site 795, and suggest an influence on EREE of an additionalsedimentary component other than the siliceous or aluminosilicatefraction. Note also that these r values are calculated while excludingthe anomalously high ZREE relating to high Y, Zr, and Nb near thebase of Unit I. In general, ZREE is correlated rather weakly at Site797 compared with the other two sites; the only r > 0.8 is with Na2O(r = 0.81), with other associations more weakly implied betweenMnO, K2O, and Y (Table 6).

Ce/Ce* values average 1.09 ± 0.07 through the silty clay of Unit I,before increasing essentially monotonically to a maximum of 1.67 withinUnit IV (Fig. 4). Eu/Eu* values average 0.90 ± 0.03 through the upper49.8 m of Subunit IA, before increasing to slightly greater values (~ 1.01± 0.05) downhole to Subunit IVA. The low value (Eu/Eu* = 0.77) at 78.3mbsf occurs in one of the samples with a potential heavy-mineraldominance (see the preceding). La/Yi),, decreases consistently downholefrom near-surface values of -1.4 to values within Unit IV ranging from-0.9 to 1. The relatively high La,/Ybn = 2.53 (78.3 mbsf) and = 2.04(106.9 mbsf) within Unit I is coincident with high EREE, Na/Al, Y/Al,and Zr/Al, as well as with low Eu/Eu*, as discussed previously. The lowLa/Y~bn = 0.57 at 163.9 mbsf corresponds with low ZREE as well as withhigh MnO, CaO, and P2O5 and suggests REE fractionation into a man-ganese carbonate or phosphatic nodule or stringer.

Trace Elements

Trace element profiles at Site 797 in general show little variationthrough the silty clays of Unit I and the diatomaceous sequence ofUnit II (Fig. 5 A). The Y, Zr, and Nb peaks within Unit I are due to theheavy-mineral input discussed previously. Cr, Rb, and perhaps Srincrease into Unit III; as at Site 795, Cr is enriched greatly, yet showsno strong interelement correlation (Table 5). All elements, except forNb, decrease in concentration below the opal-A/opal-CT transitionthrough Unit IV. Al-normalized trace element profiles (Fig. 5B)indicate that most trace elements are affiliated with aluminosilicatephases, although some profiles (e.g., Sr/Al) increase at depth.

REEs IN JAPAN SEA SEDIMENTS

REE total concentrations and relative fractionations are poten-tially responding to the combined influences of (1) paleoceanographicand paleogeographic variation, (2) sediment lithology, and (3) diage-netic modification. The relative importance of each of these influ-ences in controlling the REE composition of Japan Sea sediment ateach site is synthesized in this section. Overall, the REEs are associ-ated predominantly with the detrital fraction, which is diluted bybiogenic input that is driven by productivity in the overlying watercolumn. Due to the fact that the Japan Sea is an enclosed marginalbasin dominated by terrigenous and biogenic sources, REE scaveng-ing by metalliferous components is unimportant, although the processis a major influence elsewhere in marine sediment (e.g., Ruhlin andOwen, 1986; Murray et al., 1991).

Accordingly, paleogeographic position with respect to detrital andproductivity inputs is the most important controlling influence onΣ R E E concentrations within Japan Sea sediments (Fig. 6). Ce/Ce*does not respond to these inputs, as Ce/Ce* does not correlate withSiO2 or A12O3 to any significant degree, nor does the Ce/Ce* profilesystematically change through diatomaceous intervals. ZREE at bothSites 794 and 795 is controlled by terrigenous input from the nearbyJapan Arc; the relatively high correlations of ZREE with A12O3 andK2O suggest that most REEs are associated with clay minerals. Majorelement ratios (e.g., P2O5/A12O3) are also similar to that of averageshale. At Site 794, a pronounced Eu/Eu* increase through the diato-maceous Unit II may indicate an Eu-rich siliceous input or, alterna-tively, an additional feldspar contribution. In contrast to Sites 794 and795, Σ R E E at Site 797 correlates poorly with A12O3 and the othermajor elements, and may be controlled locally and sporadically by adetrital heavy-mineral association. Such heavy minerals most likelywould have originated from the rifted continental fragment compos-ing the Yamato Rise, located nearby to the west (Fig. 1). Althoughthere clearly is detrital input from the Yamato Rise beginning in thelate Pliocene (Fig. 2), earlier inputs may be masked by diagenetic

691

R. W. MURRAY ET AL.

WCoastal Japan Sea water

Ce/Ce* = 0.74

Siliceous Productivity

Figure 6. Schematic diagram summarizing processes that control EREE and Ce/Ce* at Sites 794, 795, and 797. Sites are positioned according to theirpaleogeographic position relative to the detrital inputs from the Japan Arc and Yamato Rise as well as suggesting the relative significance of biogenic inputfrom overlying water; sites are not positioned according to their actual location (Fig. 1). Note the change in scale between depth in seawater and sub-bottomdepth. Coastal Japan Sea water data from Tanaka et al. (1990; Fig. 1), settling particle data from Masuzawa and Koyama (1989; Fig. 1), sediment datafrom this work (Site 794). Site 794 is heavily influenced by terrigenous input from the Japan Arc as well as by the biogenic component from the highlyproductive overlying water. ZREE concentrations at Site 797 suggest input from the Yamato Rise. At Sites 795 and 797, the dashed line of productivityindicates the secondary role of siliceous input in controlling the EREE concentrations in the sediment. All three sites possess similar, diageneticallyinfluenced, downhole Ce/Ce* profiles; see the text for discussion of diagenetic mechanisms affecting the final preserved Ce/Ce* of Japan Sea sediment.

alteration. There is little correlation between ZREE and accumulationrate at any site (see discussion in Ruhlin and Owen, 1986; Murray etal., 1991), most likely because the accumulation rates are relativelysimilar and high throughout the Japan Sea (20-77 m/m.y.; Tamaki,Pisciotto, Allan, et al., 1990).

The detrital inputs (from the Japan Arc for Sites 794 and 795,from the Yamato Rise for Site 797) are diluted to varying degrees bythe biogenic input from the overlying seawater. ZREE values are mostinfluenced by biogenic SiO2 at Site 794, consistent with the locationof Site 794 in the eastern portion of the Japan Sea, where upwellingand biogenic productivity were perhaps greatest. ZREE values atSites 795 and 797 also are diluted by the biogenic input, although notto the same extreme as ZREE at Site 794. The two Yamato Basin sitesreflect greater siliceous input than Site 795 in the Japan Basin.

REE abundances and relative fractionations within Japan Sea sedi-ment do not record any basinwide paleoceanographic change. The con-sistent downhole increase in Ce/Ce* at each site preserves diagenetic,not paleoceanographic, processes. Below -400 mbsf at Site 794, Ce/Ce*seems to decrease due to effects of SiO2 diagenesis, as discussedpreviously. Although previous workers have related Ce/Ce* values insediments to the redox state of the overlying waters (e.g., Liu et al.,1988; see discussion of German and Elderfield, 1990a), there is noindependent geologic, stratigraphic, paleontologic, or chemical data tosuggest the existence of a monotonic and progressive oxygenation of

the Japan Sea which would be required to produce the Ce/Ce* profile(Tamaki, Pisciotto, Allan, et al., 1990). In fact, as described earlier,bottom-water conditions ranged from oxic to dysaerobic throughoutthe Miocene, were in general well oxygenated during the late Mioceneto Pliocene, and have recently included periods of anoxic deep water.None of these large-scale fluctuations in the oxygenation of bottomwaters are recorded by Ce/Ce* at any Leg 127 site. Indeed, becauseCe reduction can occur under suboxic (not completely anoxic) con-ditions (e.g., German and Elderfield, 1990b), the independently docu-mented variations of the Japan Sea dissolved oxygen should result inmany oscillations of Ce/Ce* values, were Ce/Ce* in sediment re-sponding faithfully to variations in the overlying water's oxygenationstate. Furthermore, a comparison of downhole Ce/Ce* profiles be-tween Sites 794 and 797 (Fig. 7), which should record the samepaleoceanographic signature because they are located within thesame basin (Fig. 1), indicates that Ce/Ce* correlates slightly betterwith depth than with age.

The consistent downhole increase of Ce/Ce* instead is explained bestby diagenetic fractionations. Masuzawa and Koyama (1989) measuredlarge positive Ce anomalies (-1.56 ± 0.34) for settling particles ofundifferentiated composition in the Japan Sea and concluded that Ce ispreferentially removed from Japan Sea waters, in association with Mn.This Ce is most likely Ce(IV), because Ce(IU) behaves similarly to theother trivalent REEs. Based on the less extreme positive Ce anomalies

692


1.0Ce/Ce*1.2 1.4 1.6 1.0

Ce/Ce*1.2 1.4 1.6

100

200

300

Q 400

500

600

• Site 794O Site 797

r=0.91

10

15

• Site 794O Site 797

r = 0.89

Figure 7. Profiles of Ce/Ce* at Sites 794 and 797 plotted downhole vs. depth and age. Note the overlapping plots of the two sites andthe slightly higher correlation coefficient when plotted vs. depth; better correlations would be expected vs. age, not depth, were Ce/Ce*responding to paleoceanographic variations in the overlying water.

from the surface sediment (Ce/Ce* = 1.07) than in the settlingparticles, Masuzawa and Koyama (1989) further suggested that theCe-bearing Mn(IV) oxide phases are liberated in conjunction with thedecomposition of organic material at or near the seafloor. By moni-toring both the bulk flux and Al flux near the seafloor they were ableto rule out significant resuspension of bottom sediment in theirparticulate samples (Masuzawa and Koyama, 1989). We note furtherthat much of the Ce liberation must be occurring while particles arerelatively high in the water column, as shown by the substantialdecrease in particulate Ce/Ce* below 1870-m water depth (Fig. 6).The fact that Ce is cycling actively in the water column suggests thatCe may also be relatively labile during diagenetic reactions withinthe sediment. Indeed, by definition, the increasing Ce/Ce* profile throughat least the upper 350 m of each site indicates that with continued burialCe concentrations are increasing in the sediment, relative to concentra-tions of the neighboring trivalent LREEs. Whether this relative Ceincrease is due to transfer of Ce from interstitial waters to sediment or toremoval of other LREEs from sediment remains to be seen; examinationof the Lan/Ybn ratio may help differentiate between the two processes.The following two paragraphs address this topic.

La,/Ybn values are mostly within the range of postulated averageterrigenous input (-1.0-1.3; Sholkovitz, 1990; Condie, 1991), ex-cluding the obvious nonrepresentative values discussed previously.Other subtle yet systematic deviations from this average are apparentat all three sites and, with the notable exception of Unit IV at Site 795,Lan/Ybn tends to increase with depth. These trends potentially reflectone or more of the following testable hypotheses: (1) variation in REEfractionation during adsorption from seawater, (2) changes in REEinput associated with changes in sediment sources to the basin, or (3)diagenetic remobilization and exchange with interstitial water. Mech-anism (1) is improbable, because REE fractionation from seawaterwould preferentially remove LREEs from seawater and add them tosediment, thereby increasing—not decreasing—sedimentary Lan/Ybn

values. Mechanism (2) may explain some of the variation, althoughits impact is unlikely to be large, given the relative uniformity ofmixed terrestrial sediments like these of similar age and tectonic

environment (e.g., Taylor and McLennan, 1985) and that there areonly slight differences in post middle Miocene clay mineralogythroughout these sites, as discussed previously (Fagel et al., 1991).La/Ybn values may, however, tend to decrease through the diatoma-ceous sequences, which may reflect primary compositional oradsorptive differences between biogenic and terrigenous particles.Mechanism (3) would require either the successive loss of LREEsfrom sediment to interstitial water, which seemingly contradicts theCe/Ce* increase discussed previously, or the transfer of HREEs fromsediment to interstitial water, which also is improbable, given therelatively restricted behavior of HREEs in other diagenetic and ma-rine environments (e.g., Fleet, 1984; Sholkovitz et al., 1989). Ifdiagenetic remobilization of LREEs actually is occurring, however,transfer of LREEs from sediment to interstitial water must be occur-ring to LREEs with the exception of Ce, as Ce/Ce* increases withdepth. This would require Ce, behaving anomalously due to its higheroxidation state, to be retained on the sediment while the remainingLREEs are transferred to interstitial waters.

Although Ce/Ce* does not correlate well with Lan/Yb,, for allpoints at a given site (with the exception of Site 797; Table 5), therenevertheless may be a genetic relationship between the loss of LREEs(except for Ce) from sediment and development of a positive Ceanomaly. It is necessary to evaluate whether the hypothesized LREEtransfer from sediment to the interstitial water could cause or accen-tuate the downhole Ce/Ce* increase observed at all sites. Consideringfirst Site 797 because the La/V^b,, and Ce/Ce* profiles are nearlylinear through the entire sequence and it is the only site with asuggested correlation between L a / Y ^ and Ce/Ce* (Table 6), theconsistent decrease from near-surface Lan/Ybn values of -1.40 toLan/Ybn ~ 0.95 at depth requires a 32% loss of La, with respect to Yb.Ce* would also decrease by 32%, resulting in an overall -50%increase in the calculated Ce/Ce*. At Site 797, then, if all Ce isretained on sediment, loss of LREEs with depth would predict aCe/Ce* change from -1.1 to -1.7. This predicted value agrees wellwith the observed profile at Site 797, where Ce/Ce* reaches maxi-mum values of -1.67 at depth. The data therefore are consistent with

693

R. W. MURRAY ET AL.

sympathetic generation of a positive Ce anomaly due to LREEtransfer from sediment to interstitial water, with Ce(IV) retained bythe sediment. At Site 794 (0-300 mbsf, over which Ce/Ce* increasessmoothly), the predicted Ce/Ce* due to LREE loss is -1.4, againagreeing well with the observed Ce/Ce* of -1.4. At Site 795 (usingLan/Ybn through Unit II and Unit III because Ce/Ce* is relativelyconstant within Unit I), the calculations predict an -70% increase inCe/Ce*, more than enough to account for the observed Ce/Ce* changeof -20% through the same interval. The increasing La,/Ybn throughUnit I, however, is not paralleled by a Ce/Ce* change, nor is theincreasing Laj/Yb,, through Unit IV, indicating that at Site 795 theremay not be a causal link between LREE behavior and downholedevelopment of the positive Ce anomaly, and that other processesseem to be masking LREE behavior. Considering all sites, then, atSites 794 and 797 (both in the Yamato Basin) there may be a geneticrelationship between LREE loss and generation of a positive Ceanomaly, whereas at Site 795 (Japan Basin) the exact relationshipbetween LREE behavior and Ce/Ce* remains unclear. This mayreflect fundamental contrasts between sedimentation in the two basins(Fig. 1), with the Yamato Basin sediments in general recording agreater influence by biogenic input. Overall, the evidence stronglysuggests that the consistent downhole increase in Ce/Ce* may be apassive response to LREE loss and need not involve significantremobilization of Ce. The contrast between the LREE transfer fromsediment to interstitial water in the Japan Sea and the previousobservations documenting LREE uptake by sediment from interstitialwater in estuaries (e.g., Sholkovitz et al., 1989) probably reflects themajor differences between the deep Japan Sea environment and theshallow estuarine systems; these contrasts include, but are not limitedto, sedimentation rate, age of burial and duration of diagenetic reac-tions recorded by sediment/interstitial water, depth of profile, andorganic matter content. Note also that the diagenetic processes docu-mented in the recent and shallowly buried estuaries (Sholkovitz et al.,1989) are not necessarily incompatible with the fractionations docu-mented in the Japan Sea deeply buried and deep-water sediment. Forexample, transfer of LREE to sediment may be occurring in theuppermost 40 cm of the Japan Sea sediment, as observed in theestuaries; we would not, however, be able to observe these shallowfractionations as our data profiles begin at 3 mbsf.

We further test the hypothesis of LREE (except Ce) transfer fromsediment to interstitial water and assess the budget of Ce in the JapanSea by combining the water column Ce/Ce* profile of Masuzawa andKoyama (1989), located 50-100 km north of Site 794 (Fig. 1), andthe downhole Ce/Ce* profile to -300 mbsf at Site 794 (Fig. 6). Inorder to track the behavior of Ce with respect to its neighboringLREEs and average shale, in addition to Ce/Ce* we use the calculatedCeexcess = Ceex = Cesample - [Lasample × (Ceshale/Lashale)], which facilitatescomparison of percent concentration change in Ce with respect to La,in units of ppm, and is also a more precise measurement than Ce/Ce*.A large exchange of Ce, with respect to the other LREEs, occurswithin the water column, as nearly 20% of the Ceex composing thelarge positive Ce anomaly on settling particles is recycled duringdescent through the water column (Ceex decreases from 16.3 to -13ppm, with Ce/Ce* decreasing from -2.10 to -1.28; Masuzawa andKoyama, 1989). Coastal waters from the southeastern Japan Seaaverage Ce/Ce* -0.74 (Tanaka et al, 1990); although these coastalwaters are distant from the Site 794 region (Fig. 1), this relative Cedepletion in water balances the smallest observed enrichment of -1.28in the settling particles (based on an initial Ce/Ce* -1.00). Therefore,in addition to the 20% Ceex recycled in the water column, there maybe a further 13-ppm Ceex that survives water column recycling and istransported to the seafloor environment. This 13 ppm represents themaximum amount of Ce, contributed by the particles, that is recycledat or near the seafloor (Fig. 6; Masuzawa and Koyama, 1989). Thesurficial sediment Ceex of -8 ppm (Ce/Ce* -1.07) indicates that 38%of this maximum Ceex is released to seawater at the seafloor. Thus,because of the great Ce exchange occurring at the sediment/water

interface, there is not a significant source of Ceex to the deeply buriedsediment. For example, if at Site 794 the entire Ceex of 8 ppmremaining in the surface sediment (-10% Cetotal) was released toshallow pore waters and subsequently transferred back to the sedi-ment upon deeper burial, it would generate a positive Ce anomaly ofonly Ce/Ce* -1.1, far less than the observed Ce/Ce* of -1.5. There-fore, the increases in Ce/Ce* with depth are not due to remobilizedCe from the original input(s). While it is difficult to refine furtherpotential diagenetic mechanisms without deeply buried REE intersti-tial-water data, the data sets and the Ce budget strongly suggest thatthe downhole increase in sediment Ce/Ce* is a passive response tothe behavior of the other LREEs and predict the successive downholeaddition of LREEs to deeply buried interstitial waters.

SUMMARY AND CONCLUSIONS

The most important variation in sedimentary chemical compositionthroughout the Japan Sea is the increase in SiO2 concentration throughPliocene diatomaceous sequences, which dilutes most of the other majorand trace element components by various degrees. This biogenic input islargest at Site 794 (Yamato Basin), moderately developed at Site 797(Yamato Basin), and of only minor importance at Site 795 (Japan Basin)and potentially reflects basinal contrasts in productivity, with the YamatoBasin in general recording greater biogenic input than the Japan Basinand the easternmost sequence of Site 794 lying beneath the most produc-tive waters of all three sites. There are few large or systematic changes insolid-phase chemistry resulting from the opal-A/opal-CT or opal-CT/quartz silica phase transformations. With the exception of certainelements on a site-specific basis, most major and trace element concen-trations are controlled by the aluminosilicate fraction of the sediment.The increasing amount of diagenetic phases with depth commonlyresults in greater concentrations of SiO2, CaO, and MnO, due to theirrespective affiliation(s) with diagenetic silica phases and manganesecarbonate nodules and stringers.

Σ R E E concentrations in the Japan Sea are strongly dependentupon the paleoceanographic position of a given site with respectto terrigenous and biogenic sources. ZREE values at Site 794 arestrongly diluted by the biogenic component through Unit II and UnitIII, and in general correspond well to aluminosilicate chemical indi-ces. Eu/Eu* values at Site 794 reach a maximum through the diato-maceous sequence as well, which most likely reflects a relative Euenrichment of the siliceous component or may suggest the slightincorporation of a detrital feldspar phase. ZREE at Site 795 also isstrongly affiliated with aluminosilicate phases yet is diluted onlyslightly by siliceous input. At Site 797, also in the Yamato Basin,Σ R E E does not correspond well with aluminosilicate indices, iscorrelated moderately to siliceous input, and appears sporadically andlocally influenced by detrital heavy minerals from the nearby riftedcontinental fragment composing the Yamato Rise.

Ce/Ce* profiles at all three sites increase essentially monotoni-cally with depth (except below 400 mbsf at Site 794) and recordprogressive diagenetic LREE relative fractionation. The observedCe/Ce* increases are not responding to changes in the paleoceano-graphic oxygenation state of the overlying water, as there are noindependent geologic, stratigraphic, paleontologic, or additionalchemical data to suggest the existence of a monotonic and progressiveoxygenation of the Japan Sea. Ce/Ce* correlates better with depththan with age at the two Yamato Basin sites.

Preliminary calculations suggest that at Sites 794 and 797 thetransfer of LREEs (with the exception of Ce) from sediment tointerstitial water may account for all of the observed relative Ceincrease. Therefore, the observed downhole increase in Ce/Ce* maybe a passive response to diagenetic behavior of the non-Ce LREEs.At Site 795, the overall lack of correlation between Ce/Ce* andLan/Ybn suggests that other process are occurring which mask thediagenetic behavior of all LREEs. First-order calculations of therelative Ce budget suggest that -20% of the Ce e x c e s s adsorbed by

694


settling particles is recycled within the water column and that anadditional -38% is recycled at or near the sediment/water interface(data from Masuzawa and Koyama, 1989). Thus, because the remain-ing 45% Ceex translates to 8-ppm Ceex and equals only -10% Cetotal,there is not a significant source of Ce to the deeply buried sediment,further suggesting that the downhole increase in Ce/Ce* is a passiveresponse to behavior of the other LREEs. The REE chemistry of theJapan Sea sediment therefore predicts the successive downhole addi-tion of LREEs to interstitial waters with deep burial.

ACKNOWLEDGMENTS

We thank Dr. E. H. De Carlo and an anonymous reviewer for theircomments and Dr. D. Nobes (U. of Waterloo, Canada) for providingthe physical-property residue samples. We thank the most helpful andprofessional technical staff of the ODP, and particularly the Leg 127shipboard personnel of Mary Ann Cusimano and Luis Alvarez (chem-istry technicians), Matt Mefferd (assistant lab officer), Barry Weber(electronic technician), and Don Sims (marine technician, X-ray lab).Larry Bernstein and Burney Hamlin assisted in keeping the shipboardchemistry lab running smoothly. We also thank the Leg 127 co-chiefscientists (Drs. K. Pisciotto and K. Tamaki) and ODP staff scientist(Dr. J. A. Allan) and curatorial representative (Peggy Myre Murray)for accommodating last-minute sample requests and for providingessential shipboard support. Additionally, we thank G. Hunt and J.Bazan at LLNL for assistance with the ICP-MS, as well as G.Anderson and G. Phelps for the sample preparation at University ofCalifornia at Berkeley. This work was partially supported by theInstitute for Geophysics and Planetary Physics at Lawrence Liver-more National Laboratory Grant # 90-41 and performed under theauspices of the U.S. Department of Energy by Lawrence LivermoreNational Laboratory under Contract W-7405-Eng-48. The work wasperformed while R. W. Murray held a shore-based JOI/USSAC OceanDrilling Fellowship, and funded additionally by a JOI/USSAC CruiseScience Support grant to R. W. Murray.

This manuscript contains material that is excerpted for a shortermanuscript, emphasizing REE behavior, published by Geochimica etCosmochimica Acta. Both manuscripts were submitted simultane-ously in accordance with ODP policy regarding dual submittal of ODPresults, and with full knowledge of the executive editor of Geo-chimica. We thank R. L. Cullers, C. R. German, R. L. Korotev, andR. A. Schmitt for their reviews of the Geochimica paper, many ofwhich were incorporated in this longer ODP contribution.

REFERENCES

Boles, J. R., and Franks, S. G., 1979. Clay diagenesis in Wilcox Sandstonesof southwest Texas: implications of smectite diagenesis on sandstonecementation. J. Sediment. Petrol, 49:55-70.

Condie, K. C , 1991. Another look at rare earth elements in shales. Geochim.Cosmochim. Acta, 55:2527-2531.

Donnelly, T. W., and Merrill, L., 1977. The scavenging of magnesium and otherchemical species by biogenic opal in deep-sea sediments. Chem. Geol.,19:167-186.

Dymek, R. E, and Gromet, L. P., 1984. Nature and origin of orthopyroxenemegacrysts from the St.-Urbain anorthosite massif, Quebec. Can. Min-eral., 22:297-326.

Elderfield, H., 1988. The oceanic chemistry of the rare-earth elements. Philos.Trans. R. Soc. London A, 325:105-126.

Fagel, N., Andre, L., Chamley, H., Debrabant, P., and Jolivet, L., 1991. Claysedimentation in a back-arc environment: the Japan Sea, ODP Leg 127.Terra, 3:283. (Abstract)

Fleet, A. J., 1984. Aqueous and sedimentary geochemistry of the rare earthelements. In Henderson, P. (Ed.), Rare Earth Element Geochemistry: NewYork (Elsevier), 343-373.

German, C. R., and Elderfield, H., 1990a. Application of the Ce anomaly as apaleoredox indicator: the ground rules. Paleoceanography, 5:823-833.

, 1990b. Rare earth elements in the NW Indian Ocean. Geochim.Cosmochim. Acta, 54:1929-1940.

Gromet, L. P., Dymek, R. F., Haskin, L. A., and Korotev, R. L., 1984. The"North American Shale Composite": its compilation, major and traceelement characteristics. Geochim. Cosmochim. Acta, 48:2469-2482.

Haskin, M. A., and Haskin, L. A., 1966. Rare earths in European shales, aredetermination. Science, 154:507-509.

Hower, J., Eslinger, E. V., Hower, M. E., and Perry, E. A., 1976. Mechanismof burial metamorphism of argillaceous sediment: 1. Mineralogical andchemical evidence. Geol. Soc. Am. Bull., 87:725-737.

Jarvis, I., and Jarvis, K. E., 1985. Rare-earth element geochemistry of standardsediments: a study using inductively coupled plama spectrometry. Chem.Geol., 53:335-344.

Liu, Y.-G., Mian, M.R.U., and Schmitt, R. A., 1988. Cerium: a chemical tracerfor paleo-oceanic redox conditions. Geochim. Cosmochim. Acta,52:1361-1371.

Masuzawa, T, and Koyama, M., 1989. Settling particles with positive Ceanomalies from the Japan Sea. Geophys. Res. Lett., 16:503-506.

McLennan, S. M., 1989. Rare earth elements in sedimentary rocks: influenceof provenance and sedimentary processes. In Lipin, B. R., and McKay, G.A. (Eds.), Geochemistry and Mineralogy of the Rare Earth Elements. Rev.Mineral., 21:169-200.

Murray, R. W., Buchholtz ten Brink, M. R., Gerlach, D. C, Russ, G. P., Ill,and Jones, D. L., 1991. Rare earth, major and trace elements in chert fromthe Franciscan Complex and Monterey Group, California: assessing REEsources to fine-grained marine sediments. Geochim. Cosmochim. Acta,55:1875-1895.

Ruhlin, D. E., and Owen, R. M., 1986. The rare earth element geochemistryof hydrothermal sediments from the East Pacific Rise: examinationof a seawater scavenging mechanism. Geochim. Cosmochim. Acta,50:393-400.

Sholkovitz, E. R., 1988. Rare earth elements in the sediments of the NorthAtlantic Ocean, Amazon Delta, and East China Sea: reinterpretation ofterrigenous input patterns to the oceans. Am. J. Science, 288:236-281.

, 1990. Rare earth elements in marine sediments and geochemicalstandards. Chem. Geol, 88:333-347.

Sholkovitz, E. R., Piepgras, D. J., and Jacobsen, S. B., 1989. The pore-waterchemistry of rare earth elements in Buzzards Bay sediments. Geochim.Cosmochim. Acta, 53:2847-2856.

Tamaki, K., Pisciotto, K., Allan, J., et al., 1990. Proc. ODP, Init. Repts., 127:College Station, TX (Ocean Drilling Program).

Tanaka, M., Shimuzu, H., and Masuda, A., 1990. Features of the heavyrare-earth elements in seawater. Geochem. J., 24:39-46.

Taylor, S. R., and McLennan, S. M., 1985. The Continental Crust: Its Com-position and Evolution: Oxford (Blackwell Scientific).

Date of initial receipt: 18 March 1991Date of acceptance: 29 August 1991Ms 127/128B-176

695

Documents

39. RARE EARTH, MAJOR, AND TRACE ELEMENT COMPOSITION …