230Th ratio as a paleoproductivity proxy

EPSL ELSEVIER

Earth and Planetary

Science Letters 149 (1997) 85- 100

Enhanced scavenging of 231Pa relative to 230Th in the South Atlantic south of the Polar Front: Implications for the use of the 231Pa/ 230Thratio as a paleoproductivity proxy H.J. Walter

*,

M.M. Rutgers van der Loeff, H. Hoeltzen

Alfred Wegener Institute for Polar and Marine Research, Bremerhacen. Germany Received 25 November

1996; revised 24 March 1997: accepted

1 April 1997

Abstract The fractionation of 230Th and 23’Pa was investigated throughout the Atlantic sector of the Southern Ocean. Published scavenging models generally assume that the ‘3’Pa/ 230Th ratio of surface sediments is primarily determined by the mass flux of particles. This relationship holds north of the Polar Front, where low primary productivity coincides with ratios of unsupported *“Pa/ 230Th - X,(23’Pa/ 230Th) - in surface sediments below the production ratio of both radionuclides in the water column. However, we observed high zz’Pa/ 230Th ratios, conventionally interpreted as a high-productivity signal, in surface sediments south of the Polar Front, especially throughout the Weddell Sea, in contradiction with the low particle flux of this region. Measurements of both dissolved and particulate fractions of ‘3’Pa and *a’Th in the water column revealed a strong N-S decrease in the Th/Pa fractionation factor, from typical open ocean values around 10 north of the Polar Front to values between 1 and 2 south of 60”s. This observation clearly indicates that the high Ez’Pa/ ‘30Th ratios in surface sediments south of the Antarctic Circumpolar Current are produced by a N-S increase in the relative scavenging efficiency of 23’Pa relative to 230Th, most probably due to a change in the chemical composition of particulate matter, and not by a high mass flux. It is speculated that biogenic opal, suggested not to significantly fractionate 23’Pa and 13’Th, may explain the enhanced scavenging of 23’Pa to the south. This assumption is further supported by extremely high 23’Pa/ 230Th ratios up to 0.34 in material collected with sediment traps south of the Polar Front, where fluxes are primarily determined by biogenic opal. Based on these results we conclude that, in regions where the sedimenting flux is dominated by biogenic opal, the ‘“‘Pa/ 230Th ratio is not a reliable indicator for the mass flux of particles, thus limiting its use as a paleoproductivity proxy in the Southern Ocean. Kewords:

Southern Ocean: radionuclides;

scavenging;

fractionation;

1. Introduction Oceanic bioproductivity important factor controlling

deep ocean and atmosphere, which is termed as the ‘biological pump’ [l-3]. In the euphotic zone of the surface ocean, CO, is taken up by phytoplankton and converted into organic matter, which then may sink to the deep ocean. Attempts have been made to explain variations in the atmospheric concentration of CO, in the past by differences in the

between

is considered to be an the partitioning of CO,

* Corresponding author. Tel: f49 471 4831 5 18. Fax: 47 1 483 1 425. E-mail: [email protected] 0012-821X/97/$17.00 Copyright PII SO01 2-82 1X(97)00068-X

opal; paleoproductivity

+ 49

bioproductivity

0 1997 Elsevier Science B.V. All rights reserved

of the oceans

[3-51.

Recently,

it has

86

H.J. Walter et al. / Earth and Planetary Science Letters 149 (1997) 85-100

been argued that iron, as the limiting factor for today’s Southern Ocean productivity, was supplied in glacial periods by wind-blown dust 161. A resulting increase in bioproductivity could have contributed to the lower concentrations of CO, in the glacial atmosphere. Ideally, the productivity signal of the surface ocean should be stored in the underlying sediments hut high and variable remineralization rates of most of the biogenic detritus in the water column and in surface sediments leave behind changes in preservation efficiency rather than real changes in productivity [7-lo]. Hence, to reconstruct past changes in oceanic productivity we have to look for tracers (proxies) in oceanic sediments which have preserved their bioproductivity signal from the euphotic zone, independently from the preservation of biogenic material. In this paper we focus on the 231Pa/ 230Th ratio which has been proposed to be such a tracer [6,1 l171. 23’Pa (half-life 32500 yr) and z30Th (half-life 75200 yr> are natural radionuclides which are continuously produced in the water column by alpha decay of their dissolved progenitors 235U and 234U. As a result of its long oceanic residence time, the distribution of U in the ocean is very homogenous [18]. Consequently, 231Pa and 230Th are produced uniformly throughout the water column at a constant initial activity ratio of 0.093. In contrast to U, 231Pa and are scavenged to and 230Th are particle-reactive the sea floor within 50-200 yr and lo-40 yr, respectively [ 19-231. This small difference in particle reactivity of Pa and Th is responsible for a fractionation of the two radionuclides in the water column. Due to its short oceanic residence time over most of the ocean, the flux of 230Th to the sediment is assumed to equal its local rate of production in the water column, whereas the longer oceanic residence time for 231Pa allows this nuclide to be transported over basin-wide distances prior to being scavenged. In large areas of the world’s oceans it has been shown that the mass flux of particles is the primary factor controlling the scavenging of 231Pa, and thus determines the ratio at which 231Pa and 230Th are deposited in the sediment [ 151. This is well documented by 23’Pa/ 230Th ratios lower than their production ratio of 0.093 in areas of low particle flux (e.g. in the central gyres), where only a minor fraction of the 231Pa is scavenged to the underlying sediments, with

the remainder being transported laterally to regions of higher particle fluxes where it is removed more efficiently [ 19,201. Consequently, 23’Pa/ 230Th ratios in surface sediments in upwelling regions at ocean margins strongly exceed the production ratio. The enhanced scavenging at ocean margins is termed as ‘boundary scavenging’ (for recent reviews of this process see [ 11,24,25]). Based on this apparently well established relationship between, 23’ Pa/ 7-30Th ratio and mass flux 1171, the 231Pa/230Th ratio has been used to assess past changes in oceanic productivity of dated sediment cores [6,11,12,14,15]. In this model variations in the 23’Pa/ 230Th ratio through time were simply explained by changes in ocean productivity. The application of the 231Pa/ 230Th ratio as a paleoproductivity proxy requires a constant 230Th/ 231Pa fractionation factor (indicated by F), which is defined as: F=

(230JV 231 Pa)part ( 230Th/ 231 Pa)diss

(1)

However, this assumption does not seem to be valid throughout the oceans [22]. Rutgers van der Loeff and Berger [22] measured a southward decrease in the 230Th/ 231Pa fractionation factor on a N-S transect across the Antarctic circumpolar current in the South Atlantic, which was explained by a N-S increase in the scavenging efficiency of 231Pa relative to 230Th, which may be due to a change in the chemical composition of suspended particulate matter. Their findings, together with results of other authors [6,19,20,25-281, imply that, in certain oceanic areas, the chemical composition of particles has to be considered as an additional factor controlling the ratio at which 23’Pa and 230Th are scavenged from the water column. Furthermore, a recent study [23] has revealed that the hydrography may also control the scavenging of 23’Pa. Yu et al. found that boundary scavenging of 231Pa is not well pronounced in the Atlantic Ocean (in contrast to the Pacific Ocean [29]), where even in the upwelling region off Western Africa the 23’ Pa/ 230Th ratios in surface sediments do not exceed the production ratio. The lack of boundary scavenging of 231Pa in the Atlantic Ocean was ex-

H.J. Walter et al./ Earth and Planetary Science Letters 149 (1997) 85-100

plained by the short mean residence time of deep water (SO-100 yr [2]), a timescale similar to that of Pa scavenging and horizontal mixing, which does not allow high concentrations of dissolved 231Pa to build up in the central Atlantic Ocean, which are required for large-scale diffusive transport of 23’Pa to the ocean margins. Yu and co-workers [23] estimated that only half of the ‘3 ’ Pa produced in the Atlantic north of 50”s is deposited in the underlying sediments, with the remainder being advected to me south with North Atlantic Deep Water (NADW). They further suggested that this exported amount of dissolved ‘3’Pa is removed to the sediments after entering the Antarctic circumpolar current (ACC) south of 50”s. The enhanced scavenging of 231Pa in the Southern Ocean south of the Polar Front was speculated to result from the high particle flux within the productive Antarctic Circumpolar Current (ACC) [ 17,31,32]. If this assumption holds, we should expect a decrease in the 231Pa/ 230Th ratios south of the productive ACC, for which Yu et al. [23] found some support in the only three data points available to them. The objective of this study is to find out whether, in the South Atlantic, the ‘3’Pa/ 230Th ratio is still a reliable tracer for the mass flux of particles. To test whether earlier observations [22] of a latitudinal gradient in F, which are based on data from a single N-S transect, occur on a basin-wide scale, the water column was sampled on several transects throughout the South Atlantic. In addition, a large number of surface sediments were analyzed for 231Pa and ‘3nTh to find out whether the latitudinal gradient in F is also reflected in the sedimentary 231Pa/ ‘30Th record, which would have important consequences for the use of the a31Pa/ 230Th ratio as a paleoproductivity proxy. Furthermore, to investigate to what extent the breakdown of F in the Southern Ocean results from a N-S change in the chemical composition of particulate matter, 23’Pa/ 230Th ratios were also determined in sediment trap samples with a known chemical composition of sinking particles.

2. Sampling and analytical methods The water column was sampled by in situ pumping to collect the dissolved and particulate fractions

?31Pa

and

230

81

Th. Samples were taken at 16 staof tions on several N-S transects across the Antarctic Circumpolar Current throughout the Atlantic sector of the Southern Ocean during Polarstern expeditions ANT X/5 (August-September 19921, ANT X/6 (October-November 1992) and ANT XI/4 (MarchMay 19941, and at four stations in the Southern Weddell Sea during ANT IX/3 (January-March 1991) (Fig. 1). Surface sediment samples were taken at each pumping station and from a large number of additional locations, including expedition ANT IX/2 (Fig. 1). Details on station locations are given in the respective cruise reports ([33-351, Kuhn. in prep.). At each pumping station 300-1400 1 of seawater were filtered in situ by several COSS-filtration systems through a 1 p,rn nuclepore filter and two MnO,-coated cartridges. A detailed description of the analytical procedures for the cartridges is given in [22]. Briefly, cartridges were spiked with appropriate amounts of “9Th and ‘33Pa as yield tracers and leached on board under recirculation with a solution containing 0.02 N hydroxylammonium chloride, 0.07 N HNO, and 0.01 N HF. We found that the leaching procedure was not quantitative for Th (only 40-70% was leached from the cartridges), which prevented an isotopic equilibrium between the 229Th spike and the Th isotopes in the sample being reached. Therefore natural 1-34Th, which is in radioactive equilibrium with its parent ‘38U below a depth of 100 m [36], was used as yield tracer for the calculation of dissolved ‘30Th. The leaching procedure for ‘3’ Pa was found to be quantitative. Collection efficiencies for ‘3’Pa and ‘j”Th, based on the activities of ‘3’ Pa and ‘34Th leached from the two cartridges in series [37], mostly ranged between 60% and 80% and 85% and 95%, respectively. Filters were dried (not weighed), spiked with the same yield tracers and digested in a microwave using a mixture of 10 ml HNO, (65%), 2 ml HF (40%). 1 ml H202 (30%) and 2 ml HClO, (70%). After digestion. Th and Pa were separated on anion exchange resin (AGl -X8, 100-200 mesh) and further treated using the method in [38]. Pa was purified by conventional TTA extraction [38], except for the filters from expedition ANT X1/4, where an electroplating method was used. Good results were obtained by the electroplating procedure using polypropylene rather than glass columns (Francois. pers. commun.) for the

88

H. J. Walter et al. /Earth

and PlaneraN Science Letters 149 (1997) 85-100

separation of Pa from Th to prevent the release of silica. The surface sediment samples were analyzed for 231Pa, Th and U isotopes according to the following procedure: 0.5 g of dry sediment was spiked with 229Th and 236 233Pa, U, and digested in a mixture of 10 ml HNO, (65%), 5 ml HF (40%) and 5 ml HClO, (70%). Th and U isotopes were separated, purified and electroplated [38]. For Pa, the electroplating method was applied. After electroplating onto silver discs, activities of 231Pa, Th and U isotopes were determined by alpha counting, and activities of 234Th and ‘33Pa by beta counting. Chemical yields for Pa were around 50% for the conventional TTA extraction method (except for ANT IX/3, which gave only lo-30%). Higher chemical efficiencies for Pa (6080%) were achieved by the new electroplating method. Chemical yields for Th and U were around 80% and 50-60%, respectively (except for a number of cartridges with Th efficiencies < 50% due to incompIete leaching, see above). The tracers 236U and 229Th were calibrated with certified standard solutions prepared from pure U,O,

and ThO, (PTB Braunschweig, uncertainty 2.5% (3cr)), respectively. For the calibration of the 233Pa tracer a definite amount of tracer solution was evaporated on a platinum disc and beta counted. Calibrations were checked by replicate analysis of the uranium-bearing sediment standards DL-la [39] and UREM-I 1 [40]. Blank contributions for 23’Pa and 230Th to the sediments were negl$$ble -but blanks contributed mostly around 2-5% to Pa m the first and 3-25% in the second cartridge in series. 230Th blanks generally contributed 2-8% (only one cartridge used). Contributions were highest for the suspended particulate matter, with blanks for 231Pa of 3-75% and l-30% for 230Th. Due to the low chemical efficiency for Pa of the cartridges (only lo-30%) during ANT 1X/3, blank contributions for 231Pa were higher, with 8-20% for the first and 14-70% for the second cartridge. Only data with blank contributions less than 50% were included in the results. The measured activities of 230Th and 231Pafor the surface sediments were corrected for detrital, U-supported 230Th and 231Pa, which is assumed to be in secular equilibrium with the detrital 234U and 235U.

-40

40

-50

50

-60

60

0 surface sediments El In-situ pumps

q in-situ pumps and surface sediments

A

sediment traps Fig. 1. Map showing sample locations

H.J. Walter et al./Earth

respectively. For samples with a 234U/ 232Th ratio less than 0.8, excess activities of 230Th (Zz’Th) and 13’Pa (‘,z ’ Pa) were calculated as follows: ;:‘Th

=?30~t.,,

f;“pa

=239Jam

_ 234~ _

(2)

0.046 .?34U

(3)

where m stands for measured activity and the factor 0.046 is the activity ratio of 235U/ ‘34U in detrital material. Sediments with 233U/ 232Th exceeding 0.8 were assumed to contain authigenic uranium and in those cases unsupported activities of 230Th and ‘3’ Pa were calculated using ‘32Th [19]: ;,“‘Th =“OTh,

_ R .?3?Thm

z,“Pa Ez3’Pam - 0.046.

R ,232Th

(4) (5)

where R is the ‘3RU/ 232Th ratio of average detrital material and is based on those samples with 134u, 232 Th less than 0.8. R has been found to vary with latitude, with values around 0.6 north of 60”s decreasing to the south to values between 0.3 and 0.4. These regional variations are included in the calculation of the unsupported activities. Ingrowth of

-60”

89

and Planetary Science Letters 149 (1997) 85-100

230Th and 23’Pa from authigenic U can be excluded because of the young age of the sediments. Although the calculation of Ea’Pa and fi” Th may overestimate or underestimate the supported activities for some of the samples, this does not significantly affect the conclusions drawn in this study because of the high ‘30Th/ “‘Th ratio in most of the samples.

3. Results An overview of :\3’Pa/ ‘30Th ratios of surface sediments throughout the Atlantic sector of the Southern Ocean is given in Table 1 and Fig. 2. The ratios are surprisingly well correlated with latitude (Fig. 3a). As expected from the low primary productivity 171, lowest ft’Pa/ ‘30Th ratios, far below the production ratio of 0.093, are observed in sediments north of the Polar Front. The southward increase above the production ratio within the Antarctic Circumpolar Current (ACC) persists throughout the Weddell Sea. where values are mostly clustered around 0.15. Results of dissolved (< 1 km) 0"

-30”

30”

1

-40

-40

-50

!

k

ACClWeddii-~Gyreboutid.

-60

-60” 0

<0.093

A

q

0.093 - 0.130 0.131 - 0.16

+

> 0.16

-30”

0”

30”

Fig. 2. Overview of excess 23’Pa/‘30Th ratios in surface sediments of the Atlantic sector of the Southern Ocean. The positions of the Subantarctic Front (SAF), the Polar Front (PF) and the ACC-Weddell Gyre boundary (AWB). according to Orsi et al. [41]. are also shown.

I

of ““.I.

““Th,

50”23.7’S

51”4l.O’S

PS 1779

PS 1780

I I .4’S

67 50.5’S

68”50.4’S

69’22.0’s

70007.9’S

PS 1976

PS 1978

PS 1979

PS 1981

76”24.8’S

PS 2055

PS 2072

4735

16”29.g’W

38”32.5’W

37”00.3’W

45”00.7’S

48”3O.I’S

50’22.3’s

51”31.9’S

52*39.4’s

PS 2262

PS 2269

PS 2271

PS 2273

30”33.l’W

31”22.3’W

33”14.7’W 3345

3642

478 I

5396

4886

5111

PS 2257

5341

I’ W

50’06.4’W

44.=31.0’S

5372

3609

4322

3320

2674

2123

4428

394

44’27

43O58 3’S

PS 2256

03’57.6’E

M”l4.9’E

06”15.8’E

OO”53.4’E

06”14.I’E

06”13.7’E

20”46.8’W

30”21.2’W

4526

4795

17”53.6’W

IJ”15.2’W

4919

4857

4625

20”50.7’W

24”og.a’w

PS 2254

expedition

68’17.4’s

60”33.9’S

PS 2052

X/J

66”48.3’S

PS 2049

ANT

68”5OS’S

69’05.2’s

PS 2040

I.4’S

71’05.7’s

69-O

I

PS 2039

PS 201

PS 1991

expedition

67”13.3’?,

PS 1974

IX /3

67”28.7’5

PS 1968

ANT

4860

27”07.5’W

66”37.6’S

PS 1966

31”06.6’W

4800

30”17.8’W

66-16.6’s

PS I964

4777

35”26.9’W

65413.2’S

PS 1961

4727 4736

37”54.8’W

65”40.3’S

65”24.7’S

37’44.6’W

4434 4684

42”30.2’W

45’48.2’W

5016

PS 1959

64*49.2’s

18”36.6’W

4258

15”16.4’W

PS 1957

64’24.4’S

PS 1955

1X / 2 expedition

55”

PS 1954

ANT

PS 1782

3549

14”04.5’W

2556

48013.9’S

PS 1777

Il~O2.2’W

50”57.l’S

PS 1775 2516

55’27.5’s

PS 1772

07”30. I’W

3276

04”27.6’E

52”35S’S

PS 1768 4136

3749

04”5 I .8’E

5 l”49.9’5

PS 1765

Ol”lO.O’E

3717

05”45.3’E

5OOO9 2’S

PS 1759

PS 1755 4263

45”37.2’S

PS 1752

07”06. I%

44”29.3’S

PS 1751

47047.3’S

Cm)

O-l

O-I

O-I

O-I

O-l

O-I

O-l

O-l

O-I.2

O-1

O-5

O-I

o-5

O-I

o-5

0.5-2

0.5-2

0.5-Z

0.5-Z

0.5-2

o-o.5

o-o.5

0.5-2

0.5-2

o-o.5

o-o.5

o-o.5

0.5-2

o-5

o-5

o-5

O-l.5

o-5

l-2

l-2

O-5

o-5

l-2

o-5

o-2

(cm)

depth

(;;‘Pa,. Depth

“‘Pa

Water

excess

4507

“‘Pa,

4770

Longitude

10°28.3’E

“lTh.

09”36.5’E

expedition

Latitude

(dpm/g)

ANT VIII/3

Station

Activities

Table exces

“‘Th

(;:OTh)

* 0.06

f 0.03

+ 0.03

zt 0.03

f 0.05

* 0.08

f 0.06

f 0.04

+ 0.03

* 0.04

* 0.04

* 0.05

zt 0.04

k 0 06

* 0.08 * 0.07

* 0.04

f 0.05 f 0.06

0.78

0.69

k 0.03

It 0.03

0 99 * 0.08

0.98

I 05 * 0 04

0.98

I .Ol + 0.04

I .03 * 0.08

1.06 * 0.05

0.99

0.83

I.10

1.15 * 0.09

I .oo * 0.08

I I I iz 0.06

0.80

I .20 * 0.09

1.18 * 0.06

1.50fO.lI

I .23 f 0.06

2 lz*oo.lo

I .28 f 0.04

* 0.09

* 0.1 I

f 0.05

* 0.17

I S5 -f 0.06 2.40 e 0.09

2.40

27.93

I .22 * 0.05 I .69 * 0.06 I .61 + 0.06

I .28 * 0.05 I .76 f 0.06 I .67 + 0.06

2.13 * 0.08 2.51 fO.11 2.29 * 0.09 2.49 * 0. I I 2.59 + 0.12

2.2 I * o.og 2.57 * 0.1 I 2.34 It 0.09 2.54 f 0. I 1 2.68 it 0.12

14.91 * 0.51 15.50 + 0.49 f 0.59

0.85

1.13 kO.05 0.71

0.67

f 004

+ 0.04

0 85 * 0.04 0.75 * 0.04

0.18 6.78 zt 0.19

8.04*

0.75 * 0.04

* 0.04

I 72 + 0.1 I 1.12 f 0.05

1 74 * 0. I I 18.80 f 0.37

I .68 * 0.07

10.34 * 0.23

14.29 * 0.41

I .49 It 0.08 0.95 * 0.05

* 0.03 1.19iO.06

0.88 I .20 * 0.06

f 0.03

I .06 + 0.09

13.50 f 0.38

I .06 f 0.09

IO 74 * 0.32

0.89

I 97 * 0.10 5.16iO.17

2.05 * 0.10 5.21 + 0 91

13.1 I * 0.35 32.46

f 0.17

4.37 & 0.16

+ 0.16 f 0.76

4.48 26.20

* 0.03

I .09 * 0.05 0.73

l.16iO.05 8 * 0.03

9.80 i 0 32

0.88 f 0.04

I .92 * 0.08

8.30 f 0.260.7

* 0.04

* 0.08 0.97

2.00 7.57 * 0.21

13.43 * 0.39

I .42 zt 0.06

19.54 zt 0.61

18.65

* 0.0 I

I .64 + 0.08

1.67 * 0.08

12.5lkO34

0.06

I .53 * 0.06

1.58 + 0.06

10.36 * 0.36

0.09 * 0.0 I

I .68 + 0.08

15.69 f 0.41

I .84 + 0.09

I .91 * 0.09 1.74 * 0.08

13.65 + 0.54

+I 0.30

15.38 + 0.58

12.61

I3 40 zt 0.38

0.71 * 0.03

* 0.03

0.76

9.72 * 0.34

1.18~0.05

6.85 f 0.22

1.19 c 0.05

I .49 f 0.06

10.34 zt 0.28 7.84 It 0.25

I .05 * 0.04

I .06 f 0.04

I I .72 f 0.26 1.49 + 0.06

0.71 * 0.04

* 0.04

0.71

7.27 + 0.23

* 0.50

f 0.09

0.97 * 0.05

* 0.05

0.97

8.65 f 0.30 1.55 f 0.06

I .07 * 0.04

1.07 * 0.04

10.96 f 0.28

1.28 * 0.05

* 0.05

I.28

14.57 f 0.42

I .33 + 0.08

I .5 I c 0.08

2.68 * 0.13

36lztO.13

4.30

2.37 * 0. I 1

3.44 * 0 I5

3.78 * 0.13

3.40 + 0.14

I.18

3.35 * 0.18

2.78 * 0.16

2.92

3.57 * 0.18

366ztO.18

3.42 f 0.13

3.18 * 0.15

3.61 f 0 22

3.63 k 0.19

I.33 I.54 * 0.06

3.48 * 0.12

3.26 f 0.13

3.91 * 0.17

3.39*0.13

0.30 f 0.03

0.75

0.56

0.22 * 0.03

0.81

I .06 f 0.07

0.27

0.26

8.82 + 0.22

0.43

2.29 * 0.09

* 0.09

2.29

+ 0.78

30.09

I .92 * 0.07

fA’P,

2.38 * 0.10

0.79 f 0.06

samples *“Pa

I .94 * 0.07

sediment

42.39 * 0.10

* 0.55

I .02 * 0.05

f 0.9

=OTh

for all surface

28.16

2’oTh

33.19

1.19*0.10

*“Th

and t:‘Pa/

I .33 * 0.06

I .27 i 0.04

I70*0.13

I.49

0.86

0.92

n.d.

n.d.

0.5 I * 0.03

0.74fO.10

0.75

0.60

0.62

0.58

0.47 * 0.03

0.68 f 0.04

2%”

+ 0.96

* 0.29

f 0.55

* 0.45

* 0.38

f 0.27

* 0.49

* 0.80

* 0.59

+ 0.34

* 0.36 i- 0 56

* 0.31

* 0.64

f 0.42

* 0.55

f 0.50

* 0.46

+ 0.80

f 0.39

+ 0.31

5 0.44 f 0.28 6.33 + 0.24

761+0.23

9.73

17.82

13.24 + 0.40

12.52

9.73

3 I .43 * 0.98

12.05 k 0.37

25.21

7.47 * 0.26

8.70 f 0.36

6.43 rt 0.26

12.43

I .07 * 0.09

18.73 + 0.65

17.45

14.51

14.00

13.68

10.39 * 0.35

9.09 * 0.33

12.10 + 0.54

14.05

Il.28

12.12

8.02 * 0.37

5.36 + 0.19

7.66 * 0.30

9.89

11.39*032

7.14

27.42

10.23

8.49

8.67

14.31

29.52

27.69

32.51

:;‘Th

k 0.004

* 0.004

* 0.003

* 0.004

230Th

f 0.004

* 0.007

f 0.010

* 0.008

* 0.008

* 0.008

* 0.005

f 0.006

* 0.009

* 0.01

* 0.009

f 0.009

* 0.009

* 0.008

* 0.008

zt 0.008

+ 0.006

+ 0.006

f 0.007

* 0.005

* 0.008

* 0.005

* 0.007

* 0.010

0.119

* 0.007

0.1 I2 f 0.006

0.116

0.098

0.090

0.085

0.092

0.166

0.170

0. I78 * 0.009

0.105

0. I34 * 0.008

0.151

0.161

0.080

0.143

0.146

0.161

0.184

0. I62 jz 0.008

0 I61 f 0.010

0.174

0. I44 * 0.009

0.136~0008

0. I48 * 0.007

0.145

0.160

0.142

0.155

0.151

0.093

0.100

0.088

0. I52 f 0.009

0.1 I4 + 0.008

0. I23 f 0.006

0.089

0.078

0.070

0.074

;‘,’ Pa/

1

O-I

1916

11-S

5322 5324 464 I SOS6 52MY SO83 1261

OS”01 dE

U7”46 &

34-3 I .BE

4E

24’58

36”3J

lX”1S 4E

38CJ9 XE

6?“07 4s

62”57.8S

66’00. i S

63-I I 6s

60”‘22.5S

57‘35.h

4Y”30 js

PS 2578

PS 2579

PS 2589

PS 26,oo

PS 2602

PS 2604

PS2611

8E

o-0.5

5349

02’1 I.iE

O-5

(1-S

0-S

O-S

O-S

O-S

0-v

5012

03”I 2 4w

61’13 iS

o-5

59g26.tk

5193

PS 2577

.31”.35 ZE

PS 2575

IA

expedition

4Pl

X1/4

PS 2562

ANT

O-S

I

409

“6”0”.2’W

I’S

4x-30

PS 2376

O-5

234 I

06”0”.2’W

o-5

3660

06”Oo S’W

53”59 8’S

PS 2372

o-5

o-s

57”03.3’S

PS 2371

4059 so39

58-29.2’S

5

U-O s

U-5

n-s

O-5

OS”59 9’W

OS”5Y.6’W

55-5 I. I’S

PS 2369

PS 2370

3756

05943.O’W

46”92.3’S

3715

PS 2368

2060

W

06”OO. I’ W

PS 2367

06aO0.0’

SO”59 Y’S

49”OO.O’S

PS 2366

3117

W

06”00.3’

55”00.5’s

O-S

“-5

PS 2365

4040

56-04.3’S

PS 2364

4156

48”““. I’S

PS 2363

06”5”.6’W

O-S

2688

06”““. I’W

O-5

3194

8’ W

05”59.9’W

I’S

53’00

06-02

55”““. I’S

PS 2362

expedition

58’58 6’W

57”59. I’W

56”56.6’W

48”59.6’W

PS 2361

ANT X/6

.53-36.4’S

O-I

PS 2353

o- I

4372

55”15.5’S

PS 2342

4539

56”01.9’S

PS 2339

n- I

4027

53”59.3’W

57-09. I’S

PS 2336

0-I

4SSS

52~oo.o’w

57055.1’S

PS 2334

I)- I

11)3x

59”02.4’S

PS 2331

O-I

5224

44”5”.8’W

6OOO6O’S

O-I

PS 2320

0-I

39”42 3’W

PS 2312

I666

35’34 6’W

S9”03.5’S

59’49 6’S

PS 2307

X / 5 expedition

PS 2299

ANT

57”3”.6’S

PS 228X

I’S

59’44

57”-15,1’S

PS 2283

I

?SS3

,I.

1 n-1 0 ! 0-I

a-

56.49

7’S

S4”lX I’S

PS 22x0

expedition

F’S 2276

ANT X/S 0 5s i

0 0.1

087_tOO7

f 0.06

* 0.02

2 0.03

t 0 02

i 0 02

+ 0 05

i 0 04

l6k

0.07 C 0 05

* 0 US

k 0 II

0 80 + 0 04

0 57 + 0.04

0 48 i 0.03

0.54

* 0.03 f 0.03

* 0.03

IR48ytO.76

176

f 0.47

0.41

045

XL f U 14

i

+ 0.44

ls778+U?L

II

I7 s1+

22.71

2

f NY7

1Y 3s * 0.56

1I3

+ 0.19

37 75 f

I441

“.I I

f 0 09

I1

1 73 5 n.00

I .xx + U 08

2.66

3 40 f 0

3 03 * 0.09

2.51 * 0.10

3.92 zt 0.12

2 74 ?; 0 08

3 202

2.73 i 0.08

I 20i 004

0 57 + 0 03

f 0.05

12

I 72 i 0 “6

I xx + 0 OR

2.65 i 0 09

3.36 i 0 I I

3 03 A 0.09

2 43 f U IO

387_C”

2 70 * 0.08

3 16fUli

2 68 k II U8

I I7 i 0.04

0 56 + 0.03

0 OS

I 34 i

615+012

7.87 * 0 25

* 0.25

I.35

I .?? + 0.04

0.56

0.84

I .63 i 0.08

-

f 0.03 r 0 03

U 57 * 0 03

0.85

0.55

I .?6 -t 0.04

* 0.04

I.27

f 0.04

I .65 + 0.08

1661

0 43 i 0 0s

+ 0.04

0.7 I * 0.03

0.85

I .32 f 0.05

0 16iO.OI

* 0.03

I 38 + 0.06 0.53

i 045 * 0.14

? 0.23 + 0 22

f”3l

k 0.28 + 0 I4

t

I 22

f 0.51 i 0.5 I

+ “..a

IS 59 1 0.36

I I 65 t 0 43

I7 II

2 I .84 r 0.5 I

20.07

14.90

2’) 55 k I “7

1XSY+Ohl

176S*O88

36.78

13 x3 * O.-I4

6.00

7.61 k 0 29

Il.63

7.66 k 0.19

IO 05 i 0 28

10.00 t 0 33

3.66 -t 0.17

4.94

781

IO.31

4 93 * 0.20

5.57 * 0.2 I

I .72 * 0.08

4.02

12.31

I4 27 f 0.45

f 0.07

I.76

3?2i_OI2 Y x2 f 0 39

I.01

f 0 03

A 0.04

* 0.05

i-o.01

k O.“?

+ 0.20

6.29 * 0.28

4.07

8 99 f 0.38

7SlfOIX

3.65 t 0 1’)

57XtOl6

7 2” + 0 3 I

6 3 I * 0.17

0.88 zt 0 04

* 0.02

I 03 * 0.04

0.73

0.85

I.33

021

0.57

I .4? i 0.06

I .RO + 0.07

0 93 * 0.04

0.34

0 79 * 0.04

f 0.03

I .28 + 0.05 0.77

I 24 * 0.04

I

J 26 c 0. I

+ 0.04

0.38 + 0.02

0.83

0 78 f 0 “3

I .3? * 0.05

l3*005 Il2+0.“.5

f 0.04

I

0.86 0 sY f 0.02

+ 0 04

0 59 f 0.02

0.86

0 75 L 0 03 0 Y3 i 0 “7

I I .99

* 0.24

0 76 r II 03 n 93 t 0 03

8.09 f 0.16

IO.41

10.32 f U 29

3.92 + 0.14

so3+0.19

30.55

J??~UIl

? 0.21

8.13 * 0.28

IO.65

5 05 + 0.16

5.68 * 0 17

2.68 f 0.06

2.76 + 0 I4

2.46 + 0.08

? II

17htnin

0.88

022~001

0.38 * 0.03

0.69 + 0.03

0.58 i 0 03

0.46 + 0.03

0 47 + 0.04

0 39 f 0.03

0.13*002

0.46 * 0 03

0.5 I * 0.02

0.18 f 0.02

0.16 f 0.02

I..51 + 004

II

f 0.05 f 0

13.20 + 0.38

* 0.09 4.79

IS I6 + 0.38

f 0.32

+ 0.08

+ 0.09

IO.81

7.28 + 0.22

I 62 I.hR I 71 I 25

0 66 I 0 “4 ious

+ 0 3 I

4 06 f 0.09

2.29 f 0 08

061

9.85 446fo.16

I .28 * 0.04

I .40 zt 0.08

0.60

0 87 f 0.05

I 011 * 0OS “d nd.

0.76

“83

I) 97 i 0 04

0.59

“36i003

0 63 & 0.03

0 36 c 0.04

0.43

0.36 -.k 0 02

0.58 i 0.03

0.40

0.51

n.d.

I36+021

0.35 i_ 0.04

0 48 k 0.02

I.24

0.71 * 0.04

0.89 i 0 04

0.89 i 0 05

0 99 i 0.05

0 85 + 0 05

0.99 + 0 06

0 73 f 0 01

I

800r0.15

* U 01

0 03

n 72 i

I.42

” n1

3 86 + n.

0 49 f

6 “4 i I) I

3 Is

’ 43 5 0 2’

6 ho * 0.22

nqttno2

0 01

0 12 + 0 04

:

0.4s + 0 04 0 7’

03htO”~

0 h,, _t 0 0:

III 0.013 A 0.010

,h

i 0.003

I i (IOfJh

+ 0.00’) oi~nioo”S

0 ihl

0IS4+0007

0 I SJ k U.006

o I47 f 0.006

0. I63 i O.OOY

0 I3

0 “‘IS i 0 004

0 174 r 0010

0.073

0.094 + 0.005 0 085 i 0 004

* 0 008 + 0.009

+ 0.003

i

2 Y) s

5

g m a

m “. 2 2

.s

z 0.140 0.176

0.055

0 I s9 + 0.007

s D g -u F

\

R

s

2

;

3

0.147 i 0.010 0 “84 * 0 004

0 069 + 0.003 0.130 f 0008 0.256 f 0.015

0.238 0.172

0. I32 * 0.009 0.094 * 0.008

0.123 + 0006 0.1 I2 i 0.006

0 090 * 0.005

0. I OS + 0.008

0. I25 i- 0.007

0 I88 * 0.012

0. I42 * 0.008

0 14’) ? 0 008

U.lb2+“oII

IJ I49 f 0.008

” I?o* 0007 0 I 30 + 0 007

03”57.6’E

37”00.3’W

33”OS.l’W 23”57,3’W

60”33.9’S

ANT X / 5 expedition PS 2262 48”30. I’S

50”24.O’S 54”38. I’S

59”44. I’S

60”13.4’S

56”38.O’S

PS 2072

PS 2269 PS 2216

PS 2283

PS 2323

PS 2337

54”55.I’W

44”52,7’W

23”16.5’W

06”15.O’E

61”58.1’S

5106

4598

4766

4744 4450

5396

5370

3900

3360 3264

Cm)

Water depth

and particulate

PS 2054

Longitude

of dissolved

26”07.O’W OO”53.4’E

Latitude

(dpm/m’)

ANT IX / 3 expedition PS 1999 ?3”37.3’S PS 2049 69”05.2’S

Station

Table 2 Activities

2500 4500 500 500 2000 3000 4000 500 2000 3000 4000 500 1500 2500 3500 3500

2100 85 275 295 2335 160 3900

f f f * f f *

0.035 0.0492 0.052 0.027 0.025 0.035 0.035

0.660 * 0.035 1.035 * 0.044 0.365 f 0.024 0.483 k 0.022 1.013 f 0.034 I.032 * 0.033 0.940 * 0.024 0.730 + 0.030 I.265 + 0.046 1.132 + 0.039 1.351 + 0.043 0.999 * 0.041 I .3786 f 0.068 1.154 f 0.043 1.298 e 0.043 0.951 f 0.030

1.237 0.516 0.935 0.895 1.443 0.821 1.771

““Th,,,,.

*3”Th and B’ Pa, dissolved Depth (m)

1f 0.006

0.140 0.076 0.152 0.197

f + + f

0.025 0.0 I9 0.012 0.022

0.109 * 0.010 0.218 f 0.021 0.195 * 0.020

0.05 1f 0.009 0.136 f 0.015 0.386 k 0.020

0.078 f 0.012 0.233 * 0.021

0.072 * 0.005

0.062 f 0.005 _

0.07

* *“Th,,,

and particulate

“‘Th

0.299 0.318 0.219 0.227 0.276 0.353 0.327 0.272 0.335 0.310 0.3 I2 0.306 0.383 0.349 0.295 0.309

0.386 0.154 0.228 0.242 0.461 0.231 0.421

0.048 0.034 0.034 0.046 0.076 0.041 0.043

, and

f 0.020 f 0.018 + 0.018 k 0.014 & 0.015 * 0.035 f 0.020 f 0.022 + 0.017 f 0.014 f 0.024 f 0.022 k 0.022 f 0.017 f 0.015 * 0.020

k * f f f k f

23’ Pa&s,.

*“Pa/

“‘Pa)

0.005 0.008 0.008 0.019 0.023 0.017 0.017 0.025 0.017 0.009 0.014 0.014

+ f f f

0.003 0.002 0.003 0.002

f 0.004 f 0.004 f 0.003

* 0.002 f 0.004 f 0.005

& 0.002 + 0.002

0.023 f 0.006 0.021 f 0.006 0.014 f 0.003

* 23’Pap,,,

(““Th/

f f f f f f f

0.46 1f 0.312 f 0.599 + 0.465 + 0.272 f 0.344 + 0.348 + 0.364 + 0.261 f 0.272 f 0.226 f 0.276 f 0.289 f 0.301 * 0.226 k 0.325 f

0.313 0.298 0.243 0.297 0.320 0.281 0.238

“‘Pa/

fractionation

0.039 0.022 0.064 0.036 0.017 0.03 I 0.023 0.033 0.016 0.015 0.019 0.023 0.023 0.018 0.013 0.023

0.040 0.072 0.038 0.058 0.053 0.051 0.025

0.119 0.121 0.090 0.070

f f + +

0.030 0.039 0.0 18 0.014

0.152 k 0.029 0.076 f 0.014 0.130 f 0.019

0.147 f 0.042 0.139 f 0.026 0.061 f 0.01 I

0.070 f 0.029 0.032 k 0.01 I

0.187 f 0.043

0.339 f 0.095

0.331 + 0.082

231Pa, xlTh,,”

factor (F) for the water column ““Th,,,,

samples

2.42 2.48 2.51 4.65

k f f f

0.55 0.82 0.53 0.99

1.71 f 0.34 3.60 + 0.70 1.74 f 0.29

1.85 f 0.54 2.48 f 0.52 5.75 f I.1 I

6.62 f 2.77 9.73 * 3.40

1.27 f 0.32

0.94 f 0.3 1

0.94 f 0.26

F factor

56”09. I’S

57”44.7’S

55”5l.l’S W29.2’S

PS 2359

PS 2360

PS 2369 PS 2370

5324

5084

5120

4290

34”3 1.6E

38”35.4E

38”JY .6E

63”llhS

57%.9S

49°30.h

PS 2600

PS 2604

PS2611

Errors are given as 1 (r of propagated uncertainty -= count rate not sufficiently above background.

PS 2579

07’46.48

43”l l.iS

PS 2562

62’57.h

3980 4980

5220

OS”59.6’W 05”59.9’W

3480

1887

lS”25.5’W

06”28.6’W

4910

23”lX.S’W

3 l”35.iE

ANT IX/4

expedition 56”59.6’S

ANT X/6 PS 2358

-

3000

of counting

2820 3790

300 3500 4620

2000 3500

1900 3500

2700 4220

4000 3000 4000 2000 3000 2500 2500 4Ooo

f 0.019 + 0.027 It 0.023 + 0.032 + 0.033 & 0.038 + 0.026 k 0.050 f 0.055

statistics

k f k * + + k * f

0.006 0.028 0.006 0.009 0.004 0.007 0.006 0.014 0.020

0.101 + 0.008 0.193 f 0.017

0.090 * 0.009 0.204 k 0.014

0.054 f 0.005 0.101 f 0.007

0.049 f 0.005 0.079 + 0.006

0.134 * 0.005 0.193 f 0.007

0.132 0.393 0.079 0.148 0.049 0.099 0.062 0.119 0.286

and blank

0.754 * 0.03 1 0.993 + 0.043

0.491 f 0.026 1.073 + 0.043 1.374 f 0.067

1.245 f 0.052 1.613 f 0.064

1.473 f 0.099 1.737 f 0.092

0.593 * 0.024 0.84 1+ 0.034

0.732 I. I25 1.002 0.881 0.981 1.074 0.923 1.259 I.433

-

& 0.016 * 0.020 f 0.019 * 0.015 * 0.022 + 0.017 f 0.017 + 0.021 f 0.019

0.373 * 0.029 0.437 + 0.034

0.164 f 0.015 0.366 + 0.022 0.419 f 0.021

0.385 f 0.022 0.418 f 0.020

0.431 f 0.022 0.5 10 f 0.024

0.3 1 1 + 0.043 0.369 + 0.022

0.310 0.335 0.350 0.337 0.288 0.361 0.291 0.394 0.447

f + + + f + + f f

0.001 0.003 0.00 I 0.002 0.001 0.002 0.001 0.003 0.004

0.004 * 0.002 0.017 * 0.002

0.022 * 0.003 0.029 f 0.004

0.013 * 0.002 0.021 f 0.002

0.011 f 0.002 0.015 + 0.002

0.005 f 0.002 0.008 f 0.003

0.011 0.038 0 009 0.016 0.007 0.012 0.005 0.016 0.036 * & f k f f f * f

0.023 0.017 0.021 0.022 0.024 0.019 0.020 0.021 0.017

0.49s _+ 0.044 0.441 * 0.039

0.346 + 0.036 0.341 f 0.025 0.305 * 0.022

0.309 f 0.022 0.259 + 0.016

0.293 k 0.025 0.293 f 0.021

0.524 f 0.075 0.439 _+ 0.031

0.474 0.298 0.350 0.383 0.293 0.336 0.315 0.313 0.312

+ + f * f f f + f

0.012 0.009 0.019 0.012 0.022 0.018 0.023 0.027 0.016

0.075 + 0.017 0.090 + 0.014

0.243 + 0.040 0.144 f 0.021

0.243 k 0.039 0.202 + 0.027

0.225 f 0.043 0.193 f 0.033

0.038 f 0.0171 0.044 * 0.0141

0.086 0.098 0.108 O.IlO 0.146 0.1 16 0.080 0.132 0.126

+ * f f * + f * f

0.73 0.34 0.59 0.43 0.35 0.49 1.14 0.51 0.34

6.62 f 1.58 4.88 f 0.87

1.40 + 0.25 2.12 f 0.15

1.27 k 0.23 1.28 f 0.19

1.30 + 0.27 1.52 f 0.28

3.75 + 2.06 0.02 f 3.20

4.93 3.04 3.24 3.49 2.02 2.88 3.93 2.38 2.47

94

H.J. Walter et al. /Earth

0.31 o.251

(a)

and Planetary Science Letters 149 (1997) 85-100

I

t

Polar,Front ~

23’ Pa/ 230Th are shown in Table 2 and Fig. 3b. The data designate a rough N-S decrease across the ACC, with lowest values occurring in the central Weddell Sea (Fig. 3b), where dissolved 230Th accumulates relative to 23’Pa, due to upwelling and ingrowth [221. 23’Pa/ 230Th ratios on susp ended particles (Table 2 and Fig. 3c) show the same N-S trend as the surface sediments (Fig. 3a) with lowest values far below the production ratio in the northernmost stations, typical for open ocean environments [19211, increasing strongly across the ACC to the south.

4. Discussion 01

1 75

I

40

45

50

55 60 latitude south

65

70

4.1. Fractionation Ocean

of 23’Pa and ‘-‘OTh in the Southern

(b)

01 40

45

50

55 60 latitude south

65

70

75

@I

T

PolarFront

0.4.

1

ti Production

’

1

50

55 60 latitude south

65

70

ratio-

As we have shown above, Zz’Pa/ 230Th ratios in surface sediments of the South Atlantic are characterized by a strong latitudinal gradient, with high !jf’ Pa/ 230Th ratios far above the production ratio south of the Polar Front, extending to the Antarctic continent. According to usual scavenging theory, throughout the oceans the depositional flux of ‘30Th is assumed to almost equal its local rate of production in the water column with little net lateral advection [19,20,23,42,43]. If this assumption also holds for the Southern Ocean, the high za’Pa/ ‘j”Th ratios south of the Polar Front suggest that, in addition to the local production, a considerable portion of the ‘31Pa deposited in sediments south of Polar Front must have been supplied by lateral transport. The average zz’ Pa/ 230Th ratio of about 0.15 in surface sediments south of Polar Front would suggest a lateral contribution of ‘“‘Pa to the sediment inven-

Fig. 3. (a) Plot of excess ‘I’ Pa/ “‘Th ratios in surface sediments of the South Atlantic versus latitude. (b) Dissolved f < 1 mm) and (c) particulate 231Pa/ 230Th ratios throughout the Atlantic sector of the Southern Ocean. Only samples in the depth interval from 1500 m below the surface and 500 m above the sea floor are shown. Error bars represent 1 v uncertainty propagated from counting statistics and blank.

H.J. Walter et al./Earth

Table 3 Opal contents and ;‘;‘IPa/ j,?‘Th ratios of sinking particles, Trap

PF l-4 B01+2 WS3

Latitude

50”09.O’S 54”20.2’S 64”54.1’S

Longitude

OY46.4’E 03”20.2’W 02”33.8’W

and F factors (errors as in Table 2)

Mass flux (g m-’ aa’)

Opal (S)

a:,’ Pa/,?:‘Th

Cm) 700 450 360

38.3 53.1 33.7

40 54 70

0.322 + 0.066 0.344 f 0.034 0.249 & 0.016

Depth

95

and Planetary Science Letters 149 (IY97J 85-100

* fz3’Pa/ ““Th),,,, 0.599 i 0.064 0.465 f 0.036 0.280 * 0.058

F factor

1.86 f 0.33 1.35+0.17

I. I2 f 0.24

Data on mass flux and opal content of PF l-4 and WS 3 are from [46], and of BO 1 + 2 from Wefer and Fischer (unpublished) ’ interpolated from the latitudinal gradient of dissolved ‘j’Pa/ “‘Th ratios (Fig. 5).

tory of almost 40%. This model was applied to calculate the ‘3’ Pa mass balance for the Southern Ocean [23]. Assuming a mean zz’Pa/ 230Th rain ratio of 0.17, Yu et al. estimated that the depositional flux of 23’Pa throughout the entire Southern Ocean balances in situ production plus “‘Pa import from the Atlantic Ocean, suggesting this region to be an important sink for 13’Pa. Based on the fact that scavenging of 23’Pa in excess of its production in the water column is well known from high productivity regions at ocean margins (boundary scavenging), previous authors have hypothesized that this process might also occur in the Southern Ocean [23,44]. They suggested that the enhanced scavenging of 23’Pa in the Southern Ocean south of the Polar Front primarily results from the high particle flux within the Antarctic Circumpolar Current (ACC). If this assumption holds, then we should expect a corresponding decrease in the 2:’ Pa/ ‘30Th ratio in surface sediments south of the ACC (e.g. in the Weddell Sea), where particle fluxes are thought to be much lower than further north [45,46]. However, indistinguishable iz ’ Pa/ 230Th ratios in surface sediments south of the ACC from those underlying the ACC are in conflict with this model. The same conclusion can be drawn from the water column distributions of 23’Pa and 230Th. While in an earlier study [44] the high zz’Pa/ 230Th ratios in rapidly accumulating sediments underlying the ACC were hypothesized to result from quantitative stripping of the dissolved concentrations of 23’Pa and 230Th from the water column, because of the high particle flux, our data confirm recent results [22] and clearly indicate that there is no significant depletion of either radionuclide in the water column within the

ACC relative to the less productive regions to the north and south. Furthermore, the observed N-S decrease in dissolved ‘3’Pa/ 230Th ratios across the ACC is in conflict with what should be expected from published scavenging models [ 19-2 1,471. In the scenario described in Yu et al. [23]; namely, the southward flow of NADW into the ACC, the dissolved 23’Pa/ 230Th ratio of this water mass should increase as it enters the area of enhanced particle flux because of the preferential removal of 13’Th over 231Pa. Although total mass fluxes measured in sediment traps within the ACC, located at 50”s (PF-1) and 55”s (BO 1 + 2) with 38.3 g rn-’ y-’ [46] and 53.1 g m-’ Y-’ (Fischer and Wefer, unpubl. data), respectively (Table 3) exceed most values observed for the typical open ocean (7-45 g rn-’ y - ‘, [48]), they are much lower than values occurring in upwelling regions at ocean margins (often > 100 g m-’ y-l, [17,48]. At this mass flux in the ACC, it is not likely that 23’Pa is scavenged significantly above its rate of production in the water column [ 171. Consequently, the enhanced scavenging of 23’Pa within the ACC does not seem to result from an enhanced mass flux as postulated previously of particles, [6,11,12,14,15,23], but is more likely explained by a N-S change in the Th/Pa fractionation factor (F). This is clearly illustrated in Fig. 4, where F is plotted against latitude. North of 48”s high values of F, around 10, typical for the open ocean [19-211, indicate a strong preference for the adsorption of Th on particles, while to the south F decreases gradually to relatively constant values between 1 and 2 south of 60”s. In theory, the strong N-S decrease of F may result either from a southward decrease in the scavenging efficiency of ‘30Th, or from a respective

96

H.J. Walter et al. /Earth

.

16’

and Planetary Science Letters 149 (1997) 85-100

I

Polar Front

14.

;I 2 gi O40

45

50

55 latitude

60

f 6.5

70

75

south

Fig. 4. Latitudinal gradient of the Th/Pa fractionation in the South Atlantic. Errors as in Fig. 3.

factor (F)

increase in the scavenging efficiency of 23’ Pa. However, as long as we do not have data on the activities of both radionuclides on a weight mass basis, and consequently on their distribution coefficients, or on their inventories in the sediment, this question cannot be answered. In line with general scavenging theory [19,20,42,47], in the following we will therefore assume that the scavenging efficiency of 230Th does not change with latitude. 4.2. Effect of the composition

of particles

on F

The strong latitudinal gradient of F in the South Atlantic south of the Polar Front can only be explained by a N-S change in the chemical composition of particulate matter, thus confirming recent speculations [22]. Particles with a clear preference for the adsorption of 230Th relative to 231Pa are only observed north of 48”S, while to the south the abundance of a particulate phase with a high scavenging efficiency for 231Pa increases, which dominates scavenging south of 60”s. Values of F close to 1 show that this particulate hase does not substantially fracP tionate 230Th and 3’Pa. MnO, is thought not to fractionate 230Th and 23’Pa (e.g. [49]), and scavenging by MnO,-coated particles has been suggested to explain values of F around 1, found at ocean margins (e.g. in the Panama and Guatemala basins, [20]). Further evidence for unpreferential scavenging of 23’Pa and 230Th comes from hydrothermal plumes,

rich in Fe and Mn, emanating near mid-ocean ridges [27]. Although MnO,-coated cartridges do fractionate 230Th and 231Pa (based on the slightly higher collection efficiencies for 230Th than for 231Pa), this effect seems to be small and a N-S increase in the availability of MnO,-coated particles in the water column could explain the N-S trend of F. Since we have no data on the Mn and Fe contents of suspended matter south of the Polar Front these possibilities cannot be further assessed. However, we are not aware of sources like hydrothermal activity that would cause a N-S increase in suspended Mn and Fe oxyhydroxides. Biogenic opal is also suggested to have a high affinity for adsorption of 231Pa: this is hypothesized on the basis of field observations [6,11,22,23,28,31 I and laboratory experiments, carried out by Anderson et al. [49], who determined an F factor of 1.1, indicating that opal does not significantly fractionate 230Th and 231Pa. Available data on the composition of settling particles in the South Atlantic, based on sediment trap studies [45,46] show a strong N-S increase in the content of biogenic opal, contributing up to 70% of the total annual flux (Table 3). This N-S increase in the content of biogenic opal on sedimenting particles agrees well with the N-S breakdown of F, suggesting this phase to be a possible carrier for 231Pa to the sediment, although data on the opal contents of suspended particulate are not available. To investigate this question further we have determined 231Pa/ 230Th ratios in sinking particles with high contents of biogenic opal (Table 3). In Fig. 5 these data (which will be discussed in more detail elsewhere, manuscript in prep.) are shown together with the dissolved 231Pa/ 230Th ratios measured at similar water depths (85-500 m, Table 2). Assuming that the sinking particles are representative of the suspended particles, and that they are in equilibrium with respect to exchange of dissolved 23’Pa and ‘30Th, an estimation of F can be made. Although at the trap sites the dissolved concentrations of 23’Pa and 230Th were not measured, it is obvious from Fig. 5 that the dissolved 23’Pa/ 230Th ratios are primarily determined by the geographic latitude, with high values of 0.6 at 50”s decreasing to relatively constant values of 0.25-0.30 south of 60”s. This well established relationship can be used to assess the

H.J. Walter et al. /Earth and Planetary Science Letters 149 C1997) 85-100

vary by a factor of 14, opal contents of particles are indistinguishable [46]. Furthermore, in the Western Weddell Sea (WS-I), where the mass flux has been shown to be extremely low [45], opal contents of particles are highest, reaching almost 80%.

0.6. 0.5.

e 5: 0.4. 2 0.3, z N 0.2.

91

4.3. The “.3’Pa / 230Th ratio as a paleoproductiuity I

“,‘1 45

50

proxy 55

60

65

70

Latitude South Fig. 5. Plot of excess “’ Pa/ 230Th ratios of sediment traps (filled squares) and dissolved ( < 1 mm) “‘Pa/ “‘Th ratios (open triangles) measured in the depth interval between 85 and 500 m, versus latitude. The stippled line indicates the N-S trend of dissolved “‘Th ratios. Errors as in Fig. 3. “‘Pa/

dissolved 23’Pa/ 230Th ratios at the respective trap sites. Thus, at PF l-4 and BO 1 + 2, F values were calculated based on the dissolved 23’Pa/ 230Th ratios available at the same latitude, whereas for WS 3 an average value of 0.28 + 0.056 for the region south of 60”s was taken. F values show a slight N-S decrease, from 1.86 + 0.43 at PF l-4, to 1.35 + 0.17 at BO 1 + 2, to 1.12 f 0.24 at WS 3, coinciding with a respective increase in the content of biogenic opal of sinking particles from 40% to 70% (Table 31, thus strongly supporting the hy othesis of unpreferential P scavenging of 23’ Pa and 2 ‘Th by biogenic opal in the surface ocean. Based on these lines of evidence, we suggest the high ‘3’Pa/ 230Th ratios in surface sediments south of the Polar Front throughout the South Atlantic result from enhanced scavenging of “‘Pa by biogenic opal, rather than from a high mass flux of particles. Note that there is a general ocean-wide trend for the relative abundance of diatoms in the phytoplankton to increase as primary productivity increases [50,51], implying that high opal contents of particulate matter could be interpreted as a signal of high productivity. If so, in the Southern Ocean the ‘31Pa/ 230Th record of the surface sediments could still be a reliable tracer for export production. However, as can be seen from sediment trap data [46] in this region, there is no such relationship. At Maud Rise (WS-3 and WS-41, where annual mass fluxes

Our study in the South Atlantic has unequivocally revealed that, south of the Polar Front, the simple relationship between ‘3 ’ Pa/ ‘30Th ratio and particle flux is not valid. In particular, the high ‘3’ Pa/ ‘30Th ratios measured in surface sediments throughout the Weddell Sea cannot be interpreted as a signal of high particle flux because of the low particle flux of this region relative to the ACC, based on sediment trap data 145,461. Although, in the Western Weddell Sea, minimum particle fluxes of 8 g m-’ aa’, estimated from the flux of opal out of the sediment of 100 mm01 m-’ a-l [52], and the average opal content of sinking particles of 70-80% [45,46], are much higher than the sediment trap results (0.37 g m-’ a-‘), values are still much lower than in the ACC [52]. Thus, the high 23’Pa/ 230Th ratios observed are more likely explained by an increased adsorption of 1-31Pa to biogenic opal, which contributes up to 80% of the total annual flux [45]. Consequently, in regions where the sedimenting flux is dominated by biogenic opal, the ‘3’Pa/ 130Th ratio does not seem to be a reliable indicator for the mass flux of particles, although a definite conclusion in terms of a relationship between opal content and fractionation factor cannot yet be given. The strong affinity of ‘3 ’ Pa to opal has important implications for the use of the 231Pa/ ‘30Th ratio as a paleoproductivity proxy in dated sediment cores, because changes with time in the opal content of the sedimenting flux could have influenced the scavenging efficiency of 23’Pa [6,12]. We suggest that predictions about particle fluxes from the 23’Pa/ ‘30Th ratio can only be inferred if high fluxes of biogenic opal to the sediment can be excluded. However, reliable data about the opal content of sedimenting particles can only be expected in regions of high bulk accumulation rates, where the preservation efficiency of opal is high. In contrast, in regions of very low bulk accumulation rates, biogenic opal can dis-

H.J. Walter et al. /Earth and Planetary Science Letters 149 (1997) 85-100

98

solve almost quantitatively at the sediment surface. This effect is well known throughout the Weddell Sea, where sinking particles are primarily composed of biogenic opal (up to 80% of the total flux [45]) while contents of opal in surface sediments are extremely low ( < 1% [53]). Although in this region most of the biogenic opal production is already recycled in the water column [30], the flux of dissolved silica out of the sediments (100 mmol mm2 a- * in the Central Weddell Sea [52]), which can be regarded as a minimum estimate of the Si flux to the seafloor, is still 12% of the production rate of biogenie opal in the surface water (800 mm01 me2 a- ’ [30]). With a bulk accumulation rate of 1 g cm-’ ka-’ and an opal content of less than 1% of the sediments, the burial rate of biogenic opal is estimated to be in the order of 1 mmol m* a-‘, which illustrates that only 1% of the biogenic opal reaching the sea floor is buried. In contrast to the Weddell Sea, the flux and preservation of biogenic opal are much higher within the opal belt underlying the ACC, with burial rates of more than 60% of the rimary production [52]. Based on the constant ! 31Pa/ 230Th ratios in surface sediments throughout the South Atlantic south of the Polar Front, with opal contents varying from less than 1 to almost 80%, we suggest the primary opal-influenced 231Pa/ 230Th signal to be preserved in the sediment, even if recycling of opal in the surface sediment is almost quantitative. Consequently, we suggest that the 231Pa/ 230Th ratio cannot be applied as a reliable paleoproductivity proxy in regions where the flux to the sediment has been dominated by opal. Variations in the 23’Pa/ 230Th ratios through time, normally interpreted in terms of changes in particle flux, could solely result from changes in the content of biogenic opal on sinking particles. Like every other proxy, the 231Pa/ 230Th ratio is affected by secondary factors, our results from the South Atlantic show, once more, that reliable reconstructions of paleoproductivity can only be made if several independent proxies are used together. Acknowledgements We lar-stem.

thank the captains and crew of RV PoWithout their help the comprehensive data

set of this study would never have been possible. Andreas Michel is thanked for his assistance on expedition ANT 1X/3. We thank Michael Schliiter for deploying the in situ pumps on expedition ANT X/5 and Gerhard Fischer for supplying the sediment trap material. We are very grateful to Mike Bacon and Bob Anderson for their constructive reviews. This is AWI contribution number 1243 and contribution number 147 of SFB 261. lUC1

References 111R. Eppley, B.J. Peterson, Particulate organic matter flux and planctonic new production in the deep ocean, Nature 282 ( 1979) 677-680. 121W.S. Broecker, T.-H. Peng, Tracers in the Sea, LamontDoherty Geol. Observ., Columbia Univ., 1982, 690 pp. COz change: The iron hy[31J.H. Martin, Glacial-interglacial pothesis, Paleoceanography 5 (1990) l-13. [41J.L. Sarmiento, J.R. Toggweiler, A new model for the role of the oceans in determining atmospheric pCO,, Nature 308 (1984) 621-624. the ocean carbon system variation [51R.S. Keir, Reconstructing during the last 150,000 years according to the Antarctic nutrient hypothesis, Paleoceanography 5 (1990) 253-276. [61N. Kumar, R.F. Anderson, R.A. Mortlock, P.N. Froelich, P. Kubik, B. Dittrich-Hannen, M. Suter, Increased biological productivity and export production in the glacial Southern Ocean, Nature 378 (1995) 675-680. [71W.H. Berger, K. Fischer, C. Lai, G. Wu, Ocean productivity and organic carbon flux, part 1: Overview and maps of primary production and export production, SIO Ref. Ser. 87-30, Scripps Inst. Oceanogr., 1987, 67 pp. [81 A.J. Van-Bennekom, G.W. Berger, S.J. van der Gaast, R.T.P. De Vries, Primary productivity and the silica cycle in the Southern Ocean (Atlantic Sector), Palaeogeogr. Palaeoclimatol. Palaeoecol. 67 (1988) 19-30. 191 G.M. Bareille, L. Labracherie, L.D. Labeyrie, J.J. Pichon, J.L. Turon, Biogenic silica accumulation rate during the Holocene in the southeastern Indian Ocean, Mar. Chem. 35 (1991) 537-552. [lOI CC. Niimberg, BariumfluB und Sedimentation im shdlichen Siidatlantik, Hinweise auf Produktivitltslnderungen im QuartZr, Ph.D. Thesis, Univ. Kiel, 1995. [Ill Y. Lao, R.F. Anderson, W.S. Broecker, Boundary scavenging and deep-sea sediment dating: constraints from excess 2soTh and “‘Pa, Paleoceanography 7 (1992) 783-798. [121 Y. Lao, R.F. Anderson, W.S. Broccker, S.E. Trumbore, H.J. Hofmann, W. Wolfi, Transport and burial rates of “Be and 23’ Pa in the Pacific Ocean during the Holocene period. Earth Planet. Sci. Lett. 113 (1992) 173-189. [I 31 Y. Lao, R.F. Anderson, W.S. Broecker, H.J. Hofmann, W. Wolfi, Particulate fluxes of 230Th, 23’Pa and ‘“Be in the

H.J. Walter et al. /Earth and Planetarv Science Letters 149 f 1997) 85-100 northeastern Pacific, Geochim. Cosmochim. Acta 57 (1993) 205-217. [14] R. Francois, M.P. Bacon, M. Altabet, L.D. Labeyrie, Glacial/Interglacial changes in sediment rain rate in the SW Indian sector of subantarctic waters as recorded by ““Th, ‘“Pa, U and S15N, Paleoceanography 8 (1993) 611-629. [15] N. Kumar, R. Gwiazda, R.F. Anderson, P.N. Froelich. ” ’ Pa/ 230Th ratios in sediments as a proxy for past changes [I61

[17]

[18]

[ 191

[20]

[21]

in Southern Ocean productivity, Nature 362 (1993) 45-48. N. Kumar. Trace metals and natural radionuclides as tracers of ocean productivity, Ph.D. Thesis, Columbia Univ., New York, 1994. E.F. Yu, Variations in the particulate flux of 2’oTh and ‘“‘Pa and paleoceanographic applications of the ‘3’ Pa/ 230Th ratio. Ph.D. Thesis. Woods Hole Oceanogr. Inst., MA, 1994. J.H. Chen, R.L. Edwards. G.J. Wasserburg, “‘U, lJ41J and ‘32Th in seawater. Earth Planet. Sci. Len. 80 (1986) 241-251. R.F. Anderson. M.P. Bacon, P.G. Brewer, Removal of 23”Th and ‘“Pa from the open ocean, Earth Planet. Sci. Lett. 62 (1983) 7-23. R.F. Anderson, M.P. Bacon, P.G. Brewer, Removal of “‘Th and ‘“Pa at ocean margins. Earth Planet. Sci. Lett. 66 ( 1983) 73-90. ?31Pa and 230 Y. Nozaki, T. Nakanishi, Th profiles in the open ocean water column, Deep-Sea Res. 32 (1985) 12091220.

[22] M.M. Rutgers Van der Loeff, G.W. Berger, Scavenging of ‘?‘Th and 23’Pa near the Antarctic Polar Front in the South Atlantic, Deep-Sea Res. 40 (1993) 339-357. [23] E.F. Yu. R. Francois. M.P. Bacon, Similar rates of modern and last-glacial ocean thermohaline circulation inferred from radiochemical data. Nature 379 (1996) 689-694. [24] M.P. Bacon, Tracers of chemical scavenging in the ocean: boundary effects and large scale chemical fractionation, Phi10s. Trans. R. Sot. London Ser. A 320 (1988) 187-200. [25] R.F. Anderson. M.Q. Fleisher. P.E. Biscaye, N. Kumar, B. Dittrich. P. Kubik, M. Suter. Anomalous boundary scaveng?30Th* 23 I Pa* ing in the Middle Atlantic Bight: evidence from “‘Be and “‘Pb, Deep-Sea Res. II 41 (1994) 537-561. 1261 R.F. Anderson, Y. Lao, W.S. Broecker, S.E. Trumore, H.J. Hofmann. W. Wolfi, Boundary scavenging in the Pacific Ocean: a comparison of “Be and ‘“‘Pa. Earth Planet. Sci. Lett. 96 (1990) 287-304. [27] G.B. Shimmield. N.B. Price, The scavenging of U. of ““Th and “‘Pa during pulsed hydrothermal activity at 20”s East Pacific Rise, Geochim. Cosmochim. Acta 52 (1988) 669-677. [28] K. Taguchi. K. Harada, S. Tsunogai, Particulate removal of ‘“‘Th and “‘Pa in the biological productive northern North Pacific. Earth Planet. Sci. Lett. 93 (1989) 223-232. [29] H.-S. Yang. Y. Nozaki. H. Sakai, A. Masuda. The distribution of “‘Th and “‘Pa in the deep-sea surface sediments of the Pacific Ocean, Geochim. Cosmochim. Acta 50 (1986) 81-89. [30] A. Leynaert, D.M. Nelson, B. Queguiner, P. Treguer. The silica cycle in the Antarctic Ocean: is the Weddeil Sea 81ypical’?. Mar. Ecol. Progr. Ser. 96 (1993) 1-15.

99

[31] D.J. DeMaster, The marine budgets of silica and 32-Si. Ph.D. Thesis, Yale Univ., New Haven, 1979. [32] E. Boyle, Deep water distillation, Nature 379 (1996) 679680. [33] U. Bathmann, M. Schulz-Baldes, E. Fahrbach, V. Smetacek. H.-W. Hubberten, The expeditions Antarktis IX/ l-4 of RV in 1990/91, Ber. Polarforsch. IO0 (1992) I“Polarstern” 403. [34] R. Gersonde, The expedition Antarktis X/5 of RV “Polarstern” in 1992, Ber. Polarforsch. 131 (19931 1-167. [35] U. Bathmann, V. Smetacek. H. de Baar, E. Fahrbach, G. Krause, The expeditions Antarktis X/6-8 of RV “Polarstern” in 1992-93, Ber. Polarforsch. 135 (1994) l-236. [36] M.M. Rutgers Van der Loeff. J. Friedrich. U.V. Bathmann. Carbon export during the Spring Bloom a.t the Antarctic Polar Front, determined with the natural tracer “‘Th, Deep Sea Res. II 44 (1997) 457-478. [37] D.R. Mann, L.D. Surprenant, S.A. Casso. In situ chemisorption of transuranics from seawater. Nucl. Instrum. Methods Phys. Res. 223 (1984) 235-238. [38] R.F. Anderson. A.P. Fleer, Determination of natural actinides and plutonium in marine particulate material. Anal. Chem. 54 (1982) 1142-l 147. [39] H.F. Steger, W.S. Bowman. DL-la: A certified uraniumthorium reference ore. CANMET Rep. 8%IOE, CANMET. Energy. Mines Resour. Can.. 1980, 15 pp. [40] R.G. Hansen. E.J. Ring, The preparation and certification of a uranium reference material. Count. Mineral Technol. Rep. M84, Randburg, 1983, 8 pp. [41] A.H. Orsi. Th. Whitworth III. W.D. Nowlin Jr., On the meridional extent and fronts of the Antarctic Circumpolar Current, Deep-Sea Res. 42 (5) (1995) 641-673. [42] M.P. Bacon. C.-A. Huh. A.P. Fleer, W.G. Deuser, Seasonality in the flux of natural radionuclides and plutonium in the deep Sargasso Sea, Deep-Sea Res. 32 (1985) 273-286. 1431 R. Francois. M.P. Bacon, D.O. Suman, Th-730 profiling in deep-sea sediments: high resolution records of flux and dissolution of carbonate in the equatorial Atlantic during the last 24M)o years, Paleoceanography 5 (1990) 761-787 [44] D.J. DeMaster. The supply and accumulation of silica in the marine environment. Geochim. Cosmochim. Acta 4.5 (I98 I) 1715-1732. [45] G. Fischer, D. Fuetterer. R. Gersonde, S. Honjo. D. Osterman, G. Wefer. Seasonal variability of particle flux in the Weddell Sea and its relation to ice cover, Nature 335 (1988) 426-428. [46] G. Wefer. G. Fischer, Annual primary production and export flux in the Southern Ocean from sediment trap data, Mar. Chem. 35 (1991) 597-614. [47] M.P. Bacon, R.F. Anderson, Distribution of thorium isotopes between dissolved and particulate forms in the deep sea, J. Geophys. Res. 87 (1982) 2045-2056. [48] G. Wefer, Particle flux in the ocean: effects of episodic production, in: W.H Berger. V. Smetacek. G. Wefer (Eds.). Productivity of the Ocean: Past and Present. Wiley, New York. 1989. pp. 139-154.

100

H.J. Walter et al./Earth

and Planetary Science Letters 149 (1997) 85-100

[49] H.L. Anderson, R. Francois, S.B. Moran, Experimental evidence for differential adsorption of Th and Pa on different solid phases in seawater, EOS 73 (43s) (1992) 270. 1501 J. Dymond, M. Lyle, Flux comparisons between sediments and sediment traps in the eastern tropical Pacific: Implications for atmospheric CO, variations during the Pleistocene, Limnol. Oceanogr. 30 (1985) 699-712. [51] D.M. Nelson, P. Treguer, M. Brzezinski, A. Leynaert, B. Queguiner, Production and dissolution of biogenic silica in the ocean: Revised global estimates, comparison with re-

gional data and relationship to biogenic sedimentation, Global Biochem. Cycles 9 (1995) 359-372. [52] M. Schlllter, M.M. Rutgers van der Loeff, 0. Holby, G. Kuhn, Silica cycle in surface sediments of the South Atlantic. Deep-Sea Res., submitted. [53] M. Schluter, Early diagenesis of orcanic carbon and opal in sediments of the southern and western Weddell Sea. Geochemical analyses and modelling, Ber. Polarforsch. 73 (19901 156 pp.