Anomalously low alkenone temperatures caused by lateral particle and sediment transport in the Malvinas Current region, western Argentine Basin

Deep-Sea Research I 47 (2000) 2369}2393

Anomalously low alkenone temperatures caused by lateral particle and sediment transport in the Malvinas Current region, western Argentine Basin Albert Benthien*, Peter J. MuK ller Universita( t Bremen, FB Geowissenschaften, Postfach 330440, 28334 Bremen, Germany Received 4 August 1999; accepted 6 December 1999

Abstract We analysed the alkenone unsaturation ratio (;)Y ) in 87 surface sediment samples from the  western South Atlantic (53N}503S) in order to evaluate its applicability as a paleotemperature tool for this part of the ocean. The measured ;)Y ratios were converted into temperature using  the global core-top calibration of MuK ller et al. (1998) and compared with annual mean atlas sea-surface temperatures (SSTs) of overlying surface waters. The results reveal a close correspondence ((1.53C) between atlas and alkenone temperatures for the Western Tropical Atlantic and the Brazil Current region north of 323S, but deviating low alkenone temperatures by !23 to !63C are found in the regions of the Brazil}Malvinas Con#uence (35}393S) and the Malvinas Current (41}483S). From the oceanographic evidence these low ;)Y values  cannot be explained by preferential alkenone production below the mixed layer or during the cold season. Higher nutrient availability and algal growth rates are also unlikely causes. Instead, our results imply that lateral displacement of suspended particles and sediments, caused by strong surface and bottom currents, benthic storms, and downslope processes is responsible for the deviating ;)Y temperatures. In this way, particles and sediments carrying  a cold water ;)Y signal of coastal or southern origin are transported northward and o!shore  into areas with warmer surface waters. In the northern Argentine Basin the depth between displaced and una!ected sediments appears to coincide with the boundary between the northward #owing Lower Circumpolar Deep Water (LCDW) and the southward #owing North Atlantic Deep Water (NADW) at about 4000 m.  2000 Elsevier Science Ltd. All rights reserved. Keywords: Alkenones; ;)Y ; Sea-surface temperature (SST); Surface sediments; South Atlantic Ocean;  Brazil}Malvinas Con#uence

* Corresponding author: Tel.: #49-421-218-7108; fax: #49-421-218-3116. E-mail address: [email protected] (A. Benthien). 0967-0637/00/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 7 - 0 6 3 7 ( 0 0 ) 0 0 0 3 0 - 3

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2370

1. Introduction Over the past decade there has been growing interest in the alkenone method for reconstructing past sea-surface temperatures (SSTs). This method is based on the discovery that certain Haptophyte algae, especially the coccolithophores Emiliania huxleyi and species of the genus Gephyrocapsa, synthesize long-chain (C }C ),   unsaturated ketones (alkenones) in di!erent proportions depending on the growth temperature of the algae (Marlowe, 1984; Brassell et al., 1986; Prahl and Wakeham, 1987; Volkman et al., 1995; Conte et al., 1998). Brassell et al. (1986) introduced the temperature-dependent alkenone unsaturation index ;) , which, in a simpli"ed form,  uses the di- and triunsaturated C alkenones only (Prahl and Wakeham, 1987):  ;)Y "[C ]/[C #C ].     The tetraunsaturated C alkenone, included in the original de"nition of the index,  is mostly omitted in calibration studies as it is not generally present in marine particulate matter and sediments. For more thorough descriptions of the history, problems and improvements of the alkenone paleotemperature method the reader is referred to the reviews of Brassell (1993), Conte et al. (1998) and MuK ller et al. (1998). The use of ;)Y as paleotemperature proxy requires a "rm calibration of the index.  Most authors have relied on the E. huxleyi culture equations of Prahl and Wakeham (1987) and Prahl et al. (1988), whose general validity for paleoceanographic applications has been con"rmed by many sediment-based studies (Sikes et al., 1991; RosellMeleH et al., 1995; Pelejero and Grimalt, 1997; Sonzogni et al., 1997a, b; MuK ller et al., 1997, 1998; Herbert et al., 1998). In a previous calibration study, MuK ller et al. (1998) presented sediment-based ;)Y -temperature relationships for the eastern South Atlantic and the global ocean  between 603S and 603N. The aim of the present study was to establish if these calibrations are also valid for the western South Atlantic. For this purpose, we analysed 87 surface sediments between 53N and 503S (Fig. 1) covering a range in sea-surface temperature from 7 to 283C. We found a good agreement between ;)Y and estab lished calibrations in the tropical}subtropical western South Atlantic north of 323S but lower values than predicted in the western and northern Argentine Basin. Potential factors that may have caused these deviations are discussed in this paper.

2. Oceanographic setting The study area encompasses the western South Atlantic Ocean from 53N to 503S and 603W to 203W (Fig. 1). The upper-level circulation in this area is dominated by the Brazil and Malvinas (Falkland) Currents (Peterson and Stramma, 1991; Peterson, 1992). Near 103S, the South Equatorial Current bifurcates northward into the North Brazil Current and southward into the Brazil Current. Most of the water carried by the South Equatorial Current enters the North Brazil Current (Peterson and Stramma, 1991). The weaker Brazil Current #ows southward along the continental

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2371

Fig. 1. Bathymetric map of the western South Atlantic indicating the locations of the surface sediments analysed in this study and the general surface-water circulation patterns (modi"ed after Peterson and Stramma, 1991; Peterson, 1992). The core locations were grouped in accordance with the biogeochemical provinces of Longhurst (1995): Western Tropical Atlantic (WTRA), Brazil Current region (BRAZ), South Atlantic Tropical Gyre (SATL), Brazil}Malvinas Con#uence (BMC, considered here as a separate province), Malvinas (Falkland) Current region (MALV). Numbers indicate the core transects 1}10.

margin of South America and strengthens under the in#uence of a recirculation cell (Reid et al., 1977; Gordon and Greengrove, 1986). Near 383S, the Brazil Current encounters the Malvinas Current, which transports cold subantarctic water northward along the Argentine continental shelf. The Malvinas Current originates as a branch of the Antarctic Circumpolar Current east of the Drake Passage and turns sharply northward over the Malvinas (Falkland) Plateau (Peterson, 1992). It is

2372

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

assumed to be a strong current with signi"cant bottom #ow, but actually very few velocity estimates exist in the literature (e.g. Peterson, 1992; Garzoli, 1993; Peterson et al., 1996). At the junction both currents separate from the continental margin and #ow seaward in a southeastward direction towards 433S. The strong thermohaline frontal region formed after the separation is referred to as the Brazil}Malvinas Con#uence and has been the objective of several hydrographic investigations (e.g. Olson et al., 1988; Gordon, 1989; Garzoli, 1993; Provost et al., 1995). The Con#uence is a major frontal encounter of subtropical and subantarctic waters with strong horizontal surface-water temperature gradients, often reaching 13C km\ (Olson et al., 1988). The latitude at which the currents separate from the coast varies seasonally with a northward penetration of the Malvinas Current during austral winter and a southward shift of the Brazil Current during austral summer. Associated with the Con#uence is a complex array of large cyclonic and anticyclonic eddies (Legeckis and Gordon, 1982) also extending in the deep layers along the continental slope (Maamaatuaiahutapu et al., 1992, 1998). This intense eddy activity is assumed to cause short period, strong current events, so-called benthic storms (Gardner and Sullivan, 1981; Hollister and McCave, 1984; Flood and Shor, 1988). The Brazil}Malvinas Con#uence region not only marks a rapid transition between upper-level current regimes but also is at the crossroads of several major water masses of the world ocean (Provost et al., 1995). Seven water masses have been identi"ed in the Con#uence area from the vertical distributions of temperature, salinity, dissolved oxygen, and nutrients (Reid et al., 1977; Con#uence Principal Investigators, 1990). Brie#y, the Thermocline Water (TW) of subtropical origin and the Subantarctic Surface Water (SASW) carried by the Brazil and the Malvinas Currents, respectively, meet in approximately the upper 800 m of water depth. Between 500 and 1500 m the northward #owing Antarctic Intermediate Water (AAIW) is detected. The North Atlantic Deep Water (NADW), at about 2500 m depth, #ows southward along the continental slope before it turns eastward near 383S (Maamaatuaiahutapu et al., 1992). It separates the oncoming Circumpolar Deep Water (CDW) into an upper branch (UCDW) and a lower branch (LCDW), the latter often referred to generically as Antarctic Bottom Water (AABW). Beneath the lower CDW, the Wedell Sea Deep Water (WSDW) #ows northward (Georgi, 1981; Peterson, 1992).

3. Material and methods 3.1. Surface sediment samples The sediments analysed in this study (Fig. 1) were recovered with box and multicorers during several R/V Meteor cruises in the western South Atlantic (Bleil et al., 1993, 1994; Wefer et al., 1994; Segl et al., 1994). The water depths at the core locations ranged from 200 to 5820 m (Table 1). Surface sediments of box cores (0}2 cm) were taken on board with plastic syringes and stored at a temperature of about 43C. Cores obtained by multicorers were sectioned into 1 cm slices and stored deep frozen at !183C. In the home laboratory, all samples were freeze-dried and ground in an agate

Core type

Sample depth (cm)

!8.17 !8.20 !8.56 !8.57 !8.53 !8.74 !8.92

!20.82 !23.98 !23.99 !20.96

Brazil Continental Margin, 8.53S (transect 1) 2201-1 MUC 0}1 BRAZ 2202-5 MUC 0}1 BRAZ 2206-1 MUC 0}1 BRAZ 2205-4 MUC 0}1 BRAZ 2204-1 MUC 0}1 BRAZ 2207-2 MUC 0}1 BRAZ 2208-1 MUC 0}1 BRAZ

Brazil Continental Margin, 21}243S (transect 2) 2125-2 MUC 0}1 BRAZ 2102-1 MUC 0}1 BRAZ 2101-1 GKG 3}4 BRAZ 2124-1 MUC 0}1 BRAZ

Latitude

2.31 2.29 4.24 5.14 0.01 0.00 !3.68 2.21 !1.27 !4.03

Province

WTRA WTRA WTRA WTRA WTRA WTRA WTRA WTRA WTRA WTRA

Western Tropical Atlantic, 53N}43S 1503-2 GKG 0}2 1504-1 GKG 0}2 1515-2 GKG 0}2 1514-4 GKG 0}2 2215-8 MUC 0}1 2216-2 MUC 0}1 1501-1 GKG 0}2 1506-1 GKG 0}2 2213-1 MUC 0}1 2212-1 MUC 0}1

Core (GeoB)

!39.86 !41.20 !41.21 !39.56

!34.46 !34.26 !34.48 !34.35 !34.02 !34.14 !33.70

!30.65 !31.29 !43.67 !46.58 !23.49 !23.10 !32.01 !35.18 !24.15 !25.62

Longitude

1542 1805 1814 2003

793 1148 1382 1797 2085 2590 3975

2298 2980 3125 3511 3711 3926 4258 4267 4314 5584

0.896 0.878 0.859 0.914

0.929 0.938 0.939 0.960 0.941 0.950 0.932

0.946 0.925 0.945 0.942 0.946 0.950 0.929 0.935 0.949 0.931

Water ;)Y  depth (m)

25.8 25.3 24.7 26.4

26.8 27.1 27.1 27.8 27.2 27.5 26.9

27.3 26.7 27.3 27.2 27.3 27.5 26.8 27.0 27.4 26.9

¹ *) (3C)

25.3 24.5 24.5 25.2

27.5 27.5 27.6 27.6 27.5 27.5 27.4

27.4 27.4 27.5 27.5 26.7 26.7 27.2 27.5 26.5 26.7

SST + (3C)

!0.03

*dO ()

0.5 !0.07 0.8 0.2 1.2 !0.02 (continued on next page)

!0.7 !0.4 !0.5 0.2 !0.3 0.0 !0.5

!0.1 !0.7 !0.2 !0.3 0.6 0.8 !0.4 !0.5 0.9 0.2

*¹ (3C)

Table 1 Station data and ;)Y results for surface sediments from the western South Atlantic, sorted according to the water depth in each transect. Also, given is the  temperature obtained from ;)Y (¹ ) using the global core-top calibration (;)Y "0.033¹#0.044) of MuK ller et al. (1998) and the annual mean sea-surface  *)  temperature (SST ) as provided by the NOAA Climate Server (http://ferret.wrc.noaa.gov/fbin/climate}server) (Levitus and Boyer, 1994). *¹ indicates the + di!erence between alkenone and atlas temperatures (¹ !SST ) and *dO is the di!erence between predicted and measured benthic isotopic values (see *) + Section 5.4 for calculations) A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393 2373

!20.62 !21.27 !21.73 !22.09 !23.04 !25.51

Santos Plateau, 29}343S 2829-3 MUC 2827-2 MUC 2828-1 MUC 2830-1 MUC

Rio Grande 2817-3 2818-1 2819-2 2820-1 2821-2 2022-3

0}2 0}2 0}2 0}2

SATL SATL SATL SATL !30.87 !31.48 !31.48 !29.02

!30.92 !30.87 !30.85 !30.82 !30.45 !34.44

BRAZ BRAZ BRAZ BRAZ BRAZ BRAZ

Latitude

Rise, 313S, and Mid Oceanic Ridge MUC 0}2 SATL MUC 0}2 SATL MUC 0}2 SATL MUC 0}2 SATL MUC 0}2 SATL MUC 0}2 SATL

0}1 0}1 0}1 0}1 0}1 0}1

Province

!26.74 !27.10 !27.18 !27.29 !27.49 !27.91 !28.65 !29.11 !29.14

MUC MUC MUC MUC MUC MUC

2130-1 2126-1 2119-2 2118-1 2117-4 2116-2

Sample depth (cm)

Brazil Continental Margin, 27}293S (transect 3) 2105-3 MUC 0}1 BRAZ 2106-1 MUC 0}1 BRAZ 2107-5 MUC 0}1 BRAZ 2104-1 MUC 0}1 BRAZ 2108-1 MUC 0}1 BRAZ 2109-3 MUC 0}1 BRAZ 2110-1 MUC 0}1 BRAZ 2111-2 MUC 0}1 BRAZ 2112-1 MUC 0}1 BRAZ

Core type

Core (GeoB)

Table 1. (continued)

!43.43 !40.73 !40.72 !44.00

!38.07 !38.17 !38.34 !38.44 !38.82 !20.91

!46.74 !46.50 !46.46 !46.38 !46.23 !45.87 !45.52 !45.22 !43.38

!37.10 !38.93 !38.55 !38.02 !36.65 !36.00

Longitude

3523 3702 3750 3815

2943 3110 3435 3606 3936 4016

202 502 1052 1505 1991 2513 3003 3498 4009

2113 2537 2958 3482 4047 4164

0.789 0.808 0.754 0.805

0.731 0.767 0.759 0.760 0.750 0.648

0.863 0.859 0.845 0.857 0.860 0.853 0.860 0.806 0.802

0.912 0.924 0.893 0.913 0.872 0.834

Water ;)Y  depth (m)

22.6 23.2 21.5 23.1

20.8 21.9 21.7 21.7 21.4 18.3

24.8 24.7 24.3 24.6 24.7 24.5 24.7 23.1 23.0

26.3 26.7 25.7 26.3 25.1 23.9

¹ *) (3C)

22.1 21.9 21.9 22.9

22.2 22.2 22.2 22.3 22.5 19.5

23.5 23.4 23.4 23.4 23.4 23.2 22.9 22.8 22.9

25.4 25.1 25.1 25.1 25.1 24.6

SST + (3C)

0.5 1.3 !0.4 0.2

!1.4 !0.3 !0.5 !0.6 !1.1 !1.2

1.3 1.3 0.9 1.2 1.3 1.3 1.8 0.3 0.1

0.9 1.6 0.6 1.2 0.0 !0.7

*¹ (3C)

!0.09

!0.10 !0.08 0.04

!0.27 !0.09 !0.02

0.03 !0.12 !0.32 0.07

*dO ()

2374 A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

!37.04 !37.21 !37.21 !37.41 !37.41 !37.54 !37.54 !37.61 !37.61 !37.83 !37.83 !38.47 !38.47

!37.81 !38.51

Argentina Continental Margin, 38}393S (transect 6) 2701-4 MUC 0}1 BMC 2703-7 MUC 0}1 BMC

!35.54 !35.54 !35.60 !35.60 !35.75 !35.75 !35.98 !35.98 !36.33 !36.33 !37.01 !37.01

!31.90 !32.50 !33.50

(transect 5) BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC

SATL SATL SATL

Uruguay Continental Margin, 37}383S 2801-2 MUC 1}2 2802-2 MUC 0}1 2802-2 MUC 1}2 2803-1 MUC 0}1 2803-1 MUC 1}2 2804-2 MUC 0}1 2804-2 MUC 1}2 2805-1 MUC 0}1 2805-1 MUC 1}2 2806-6 MUC 0}1 2806-6 MUC 1}2 2807-1 MUC 0}1 2807-1 MUC 1}2

0}2 0}2 0}2

(transect 4) BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC BMC

MUC MUC MUC

Uruguay Continental Margin, 35}373S 2813-1 MUC 0}1 2813-1 MUC 1}2 2812-3 MUC 0}1 2812-3 MUC 1}2 2811-1 MUC 0}1 2811-1 MUC 1}2 2810-2 MUC 0}1 2810-2 MUC 1}2 2809-2 MUC 0}1 2809-2 MUC 1}2 2808-3 MUC 0}1 2808-3 MUC 1}2

2826-1 2825-3 2824-1

!55.02 !54.20

!54.17 !53.98 !53.98 !53.70 !53.70 !53.54 !53.54 !53.44 !53.44 !53.14 !53.14 !52.32 !52.32

!52.55 !52.55 !52.39 !52.39 !52.27 !52.27 !51.98 !51.98 !51.52 !51.52 !50.63 !50.63

!40.97 !41.43 !42.50

587 1187

491 1007 1007 1162 1162 1836 1836 2743 2743 3542 3542 4481 4481

588 588 1041 1041 1789 1789 2909 2909 3539 3539 4541 4541

3949 4352 4512

0.340 0.366

0.398 0.394 0.402 0.416 0.415 0.401 0.402 0.417 0.422 0.441 0.432 0.446 0.446

0.490 0.507 0.469 0.482 0.458 0.464 0.486 0.482 0.475 0.492 0.495 0.492

0.627 0.613 0.568

9.0 9.8

10.7 10.6 10.8 11.3 11.2 10.8 10.8 11.3 11.5 12.0 11.8 12.2 12.2

13.5 14.0 12.9 13.3 12.5 12.7 13.4 13.3 13.1 13.6 13.7 13.6

17.7 17.2 15.9

14.9 14.9

15.7 15.7 15.7 15.8 15.8 15.9 15.9 15.9 15.9 16.0 16.0 16.2 16.2

17.6 17.6 17.6 17.6 17.6 17.6 17.7 17.7 17.8 17.8 17.9 17.9

21.6 21.3 20.6

!0.30

!0.17

!0.31

!0.19

!0.30

!5.9 !5.1 (continued on next page)

!5.0 !5.1 !4.9 !4.5 !4.6 !5.1 !5.1 !4.6 !4.4 !4.0 !4.2 !4.0 !4.0

!4.1 !3.6 !4.7 !4.3 !5.1 !4.9 !4.3 !4.4 !4.7 !4.2 !4.2 !4.3

!3.9 !4.1 !4.7

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393 2375

!47.44 !47.44 !47.33 !47.33 !47.33 !47.31

Argentina Continental Margin, 473S (transect 9) 2719-1 GKG 0}2 MALV 2719-3 MUC 0}1 MALV 2722-2 MUC 0}1 MALV 2722-2 MUC 1}2 MALV 2722-1 GKG 0}2 MALV 2718-1 MUC 0}1 MALV

!38.51 !38.51 !39.29 !39.29 !39.24 !39.24

!43.67 !43.67 !43.87 !43.87 !43.91 !43.91 !43.99

BMC BMC BMC BMC BMC BMC

Latitude

Argentina Continental Margin, 43}443S (transect 8) 2712-1 MUC 0}1 MALV 2712-1 MUC 1}2 MALV 2714-5 MUC 0}1 MALV 2714-5 MUC 1}2 MALV 2715-1 MUC 0}1 MALV 2715-1 MUC 1}2 MALV 2729-1 MUC 1}2 MALV

1}2 0}2 0}1 1}2 0}1 1}2

Province

!41.42 !41.42 !41.95 !41.95 !42.37 !42.37

MUC GKG MUC MUC MUC MUC

2703-7 2703-6 2734-2 2734-2 2705-7 2705-7

Sample depth (cm)

Argentina Continental Margin, 41}423S (transect 7) 2708-5 MUC 0}1 BMC 2708-5 MUC 1}2 BMC 2707-4 MUC 0}1 BMC 2707-4 MUC 1}2 BMC 2706-6 MUC 0}1 BMC 2706-6 MUC 1}2 BMC

Core type

Core (GeoB)

Table 1. (continued)

!60.09 !60.09 !58.62 !58.62 !58.62 !58.97

!59.33 !59.33 !58.00 !58.00 !57.66 !57.66 !56.97

!57.30 !57.30 !56.32 !56.32 !55.54 !55.54

!54.20 !54.20 !54.34 !54.34 !53.36 !53.36

Longitude

684 684 2383 2383 2384 2991

1230 1230 2361 2361 3277 3277 4643

396 396 3167 3167 4740 4740

1187 1192 2295 2295 4501 4501

0.301 0.299 0.219 0.226 0.263 0.235

0.266 0.266 0.314 0.281 0.290 0.293 0.383

0.280 0.302 0.293 0.303 0.374 0.384

0.373 0.396 0.324 0.325 0.421 0.421

Water ;)Y  depth (m)

7.8 7.7 5.3 5.5 6.6 5.8

6.7 6.7 8.2 7.2 7.5 7.5 10.3

7.2 7.8 7.5 7.8 10.0 10.3

10.0 10.7 8.5 8.5 11.4 11.4

¹ *) (3C)

9.0 9.0 9.0 9.0 9.0 9.0

10.9 10.9 10.8 10.8 10.9 10.9 11.0

12.0 12.0 12.0 12.0 12.2 12.2

14.9 14.9 14.4 14.4 15.1 15.1

SST + (3C)

!1.2 !1.3 !3.7 !3.5 !2.4 !3.2

!4.2 !4.2 !2.6 !3.6 !3.4 !3.4 !0.7

!4.8 !4.2 !4.5 !4.2 !2.2 !1.9

!4.9 !4.2 !5.9 !5.9 !3.7 !3.7

*¹ (3C)

!0.18

!0.25

*dO ()

2376 A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

!47.31 !47.16 !47.16

!37.60 !44.21 !44.21 !44.48 !44.48

MALV MALV MALV

Northern and Central Argentine Basin 2814-3 MUC 1}2 SATL 2731-1 MUC 0}1 MALV 2731-1 MUC 1}2 MALV 2730-1 MUC 0}1 MALV 2730-1 MUC 1}2 MALV

1}2 0}1 1}2

!48.91 !48.01 !48.01 !47.96 !47.96 !47.96

MUC MUC MUC

Argentina Continental Margin, 483S (transect 10) 2723-2 MUC 0}1 MALV 2727-1 MUC 0}1 MALV 2727-1 MUC 1}2 MALV 2724-6 GKG 0}2 MALV 2724-7 MUC 0}1 MALV 2724-7 MUC 1}2 MALV

2718-1 2717-8 2717-8

!39.07 !51.42 !51.42 !53.25 !53.25

!57.88 !56.54 !56.54 !56.17 !56.18 !56.18

!58.97 !56.49 !56.49

4949 5691 5691 5817 5817

569 2819 2819 4788 4799 4799

2991 4495 4495

0.570 0.428 0.438 0.411 0.424

0.294 0.232 0.231 0.312 0.277 0.299

0.249 0.324 0.331

15.9 11.6 11.9 11.1 11.5

7.6 5.7 5.7 8.1 7.1 7.7

6.2 8.5 8.7

17.6 13.1 13.1 12.3 12.3

7.9 8.5 8.5 8.5 8.5 8.5

9.0 9.2 9.2

!1.7 !1.5 !1.2 !1.2 !0.8

!0.3 !2.8 !2.8 !0.4 !1.4 !0.8

!2.8 !0.7 !0.5

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393 2377

2378

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

mortar. For alkenone analysis, we used sediments from the upper 2 cm assuming that they are of late Holocene age (see below). 3.2. Alkenone analysis Alkenones were extracted from 1 to 10 g aliquots of freeze-dried and homogenised sediment. We used UP 200H ultrasonic disrupter probes (200 W, amplitude 0.5, pulse 0.5) and successively less polar mixtures of methanol and methylene chloride (CH OH, CH OH : CH Cl 1 : 1, CH Cl ), each for 3 min. The three extracts were       combined, desalted with deionized water, dried over Na SO and concentrated under   N . The extracts of cores GeoB 2701}2734 and GeoB 2801}2830 were additionally  puri"ed by passing them over a silica gel cartridge (Varian Bond Elut; 1 cm/100 mg) and then saponi"ed to remove possibly interfering fatty acid methyl esters, which may a!ect the relatively small C peak in sediments from cold-water regions. For the  saponi"cation, we added 0.3 ml of 0.1 M KOH in CH OH : H O (90 : 10) to the   extract, which was then heated at 803C in a capped vial for 2 h. After cooling, the alkenone-containing fraction was obtained by partitioning into hexane, evaporated, and "nally taken up in 25 ll of a 1 : 1 CH OH : CH Cl mixture.    The extracts were analysed by capillary gas chromatography using a HP 5890A gas chromatograph equipped with a HP Ultra 1 fused silica column (50 m;0.32 mm; 0.52 lm), split/splitless injection (1 : 10 split modus), and helium as carrier gas. The oven temperature was programmed from 50 to 1503C at 303C min\, 150 to 2303C at 83C min\, and 230 to 3203C at 63C min\ with a 45 min hold at 3203C. Analytical precision was better than 0.02 ;)Y units ($0.63C) based on replicate extractions and  injections of selected samples and laboratory internal reference sediments. At most sites in the Argentine Basin, we analysed the two uppermost sediment slices (0}1 and 1}2 cm, Table 1). The values agreed within $0.022;)Y units (in most cases within  $0.01 units) and were averaged for the purpose of this study.

4. Results Table 1 lists the ;)Y results for the surface sediments from the western South  Atlantic together with the station data. We consider SST values from 0 m depth as representative for the mixed surface layer in accordance with previous core-top calibration studies (e.g. Rosell-MeleH et al., 1995; Sonzogni et al., 1997a, b; Herbert et al., 1998; MuK ller et al., 1998). To simplify the evaluation of the results, we grouped the core sites from north to south in accordance with the biogeochemical provinces de"ned by Longhurst (1995). These provinces are the Western Tropical Atlantic (WTRA), the coastal Brazil Current (BRAZ), the South Atlantic Tropical Gyre (SATL), and the Malvinas Current region (MALV). The area of the Brazil-Malvinas Con#uence (BMC) is considered here as an additional province (Fig. 1). Fig. 2 shows the relationship between ;)Y and annual mean SST for the surface  sediments from the western South Atlantic in comparison to the E. huxleyi culture equation of Prahl et al. (1988) and to core-top calibrations for the eastern South

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2379

Fig. 2. Relationship between ;)Y and annual mean SST (Levitus and Boyer, 1994) for surface sediments  from the western South Atlantic (see also Table 1). Also, shown are the E. huxleyi culture calibration of Prahl et al. (1988) and the core-top calibrations for the eastern South Atlantic and the global ocean (MuK ller et al., 1998). The shaded area indicates the standard error of estimate for the global calibration ($0.05;)Y  units or $1.53C). ;)Y values deviating by more than this value from the global calibration line are  considered anomalous in the present study.

Atlantic and the global ocean between 603N and 603S (MuK ller et al., 1998). Due to the wide range of latitudes covered, annual mean SST ranges from 7.9 to 27.63C and ;)Y from 0.22 to 0.96. The highest ;)Y values (0.79}0.96) were recorded in the warm   waters (22}283C) of the Western Tropical Atlantic and the Brazil Current region. They are in good accordance with the calibrations shown in Fig. 2. Further to the southeast, the ;)Y decreased to values between 0.57 and 0.81 in the South Atlantic  Tropical Gyre. Whereas most of the values from this province agree within the uncertainty of the global calibration, three samples from the southern rim of the Santos Plateau (GeoB 2824}2826) showed signi"cantly lower alkenone temperatures compared to the Levitus data (by 3.9}4.73C, Table 1, Fig. 3).

2380

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

Fig. 3. Bathymetric map illustrating the di!erences (*¹, Table 1) between alkenone-derived temperatures using the global core-top calibration of MuK ller et al. (1998) and annual mean atlas SST at 0 m (Levitus and Boyer, 1994). Isolines were created with the Kriging method (Davis, 1986). See caption of Fig. 1 for the province names.

In the region of the Brazil}Malvinas Con#uence (BMC) between 35 and 423S, all sediment samples had lower ;)Y values than expected from the calibrations in Fig. 2.  The ;)Y values ranged from 0.29 to 0.50 suggesting temperatures between 7.5 and  13.83C. These values are lower, by 2}63C, than atlas SST at the same sites (transects + 4}7, Table 1). The lowest ;)Y values (0.22}0.38) were obtained in the domain of the Malvinas  Current (43}493S). A closer look at the data from transects 8}10 (Table 1) reveals that only ;)Y temperatures from the upper (569 and 684 m) and lower (4495}4799 m) 

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2381

continental margin were in good agreement (*¹(1.33C) with the observed surfacewater temperatures. All sediments from intermediate water depths (&1200}3300 m) showed alkenone temperatures deviating by !2.4 to !4.23C.

5. Discussion Several biological and sedimentological factors may explain the low alkenone temperatures in the western Argentine Basin, i.e. the water depth and season of alkenone production or erosion and lateral sediment transport by currents. In the following evaluation, we use oceanographic criteria to discuss which process most likely caused the anomalous ;)Y values in this region.  5.1. Subsurface alkenone production Coccolithophores live in the euphotic zone where they use sunlight for photosynthesis. The extent of this zone is a direct result of the amount of suspended particles in the water and de"ned as that depth where the irradiance is reduced to 1% of its value at the surface (Morel and Berthon, 1989). It varies from about 20 m in eutrophic water to about 120 m in oligotrophic water (Longhurst, 1993). A major question regarding the sedimentary ;)Y signal is at which water depth this temperature signal was  produced. Production of alkenones below the mixed layer could thus result in lower ;)Y temperatures than those observed at the surface. Prahl et al. (1993) described  such a situation in the northeast Paci"c Ocean at 423N. The ;)Y signal recorded by  their sediment traps was not in accordance with the mixed layer source and was believed to derive from a subsurface chlorophyll maximum (SCM) within the thermocline. In the Malvinas Current region, however, it is unlikely that signi"cant alkenone production occurs below the surface mixed layer. Fig. 4 displays the general hydrographic features of the upper 100 m of the water column for selected sites in the Brazil}Malvinas Con#uence (transect 5) and the Malvinas Current region (transect 8). The "gure combines typical temperature and nitrate pro"les (Levitus and Boyer, 1994) with the depth of the photic zone (light grey) according to Longhurst (1995) and that of the chlorophyll a maximum (dark grey) as reported by Gayoso (1995) and Carreto et al. (1995). The alkenone temperatures (thick vertical lines) are derived from the corresponding surface sediments. According to Longhurst (1995), this province is characterised by a shallow photic depth between 30 and 40 m, which is persistent throughout the year. The mixed layer extends below this depth for most of the year (March}November), reaching its maximal thickness (200 m) in austral winter (July}September). Only during summer (December}February) is the top of the thermocline illuminated, but the production in the subsurface chlorophyll maximum remains low just as in the other seasons (Longhurst, 1995). This general characterisation of the province is in agreement with the few available studies concerning the vertical distribution of coccolithophores in the Malvinas Current region. Gayoso (1995) examined samples that had been collected during November}December 1989

2382

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

Fig. 4. Depth pro"les for seasonal water temperatures and average nitrate concentrations (Levitus and Boyer, 1994) for typical sites in the Brazil}Malvinas Con#uence and the Malvinas Current region. Also indicated is the ;)Y temperature based on the global calibration of MuK ller et al. (1998), the photic zone  (light grey) after Longhurst (1995), and the depth of the chlorophyll a maximum (dark grey) as reported by Gayoso (1995) and Carreto et al. (1995).

along three transects o! RmH o de la Plata. She found maximum cell concentrations of E. huxleyi at 25 m depth, although coccospheres of this species were present down to a depth of 75 m, and highest concentrations of chlorophyll a at about 30 m. Carreto et al. (1995) also reported highest chlorophyll a concentrations at about 30 m during

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2383

spring and summer in the Argentinean Sea south of the RmH o de la Plata, whereas the (less signi"cant) winter maximum occurred close to the surface. From this evidence, we assume that the ;)Y signal was essentially produced in the  upper 40 m of the water column. The surface water characteristics summarised in Fig. 4 clearly show that the anomalously low ;)Y temperatures cannot be explained by  subsurface production within the thermocline unless one assumes that alkenones are exclusively produced during the winter season. However, this is not supported by the seasonal production characteristics in this region (see below). Moreover, the fact that biased and una!ected ;)Y signals occur under the same hydrographic conditions  (Fig. 4c and d) points to another mechanism as well. The high nutrient concentrations in the surface waters of the Malvinas Current region (e.g. 3}7 lM at the locations shown in Fig. 4) also argue against a predominance of alkenone production below the mixed layer. As the surface waters are not depleted in nutrients, there is no reason for alkenone producers to sink into the nutricline (e.g. Knappertsbusch, 1993). It should be noted that the hydrographic situation is di!erent at the Paci"c site studied by Prahl et al. (1993), where nitrate is reduced to low levels ((0.3 lM) in the mixed layer and subsurface production appears to be more signi"cant. In summary, we suggest that the anomalous ;)Y ratios of the surface sediments  from the western Argentine Basin cannot be explained by subsurface alkenone production in the thermocline. The temperature signal was probably produced in the mixed layer and modi"ed by another process.

5.2. Seasonality in alkenone production A second factor that may bias ;)Y ratios is a high seasonality in alkenone  production. This factor is not important in the tropical part of the western Atlantic, where the seasonalities both in primary production and mixed layer temperature are very weak (Levitus and Boyer, 1994; Longhurst, 1995). The seasonal SST variation, for example, is about 13C in this region and thus close to the uncertainty of the alkenone paleotemperature method (Fig. 2). Hence, even if the seasonality in production were high, this would have no measurable e!ect on the resulting ;)Y value.  In the domains of the Brazil Current and the South Atlantic Tropical Gyre, primary production is generally enhanced in austral spring and summer and lowest in winter (Longhurst, 1995; Antoine et al., 1996). Local upwelling o! Cabo Frio (21}233S) also shows a seasonal maximum in spring and summer and a minimum in winter (Campos et al., 1996). However, this region is characterised by upwelling episodes of a few days duration occurring frequently throughout the year (Longhurst, 1995), which will not produce a clear seasonal signal. The ;)Y results for transect 2 from this upwelling area  and for the other sites in the subtropical western Atlantic north of 323S (Table 1) are consistent with these general productivity patterns re#ecting annual mean SST rather than a seasonal signal. In the Con#uence and Malvinas Current region, seasonal e!ects on ;)Y could be  more important as the SST contrast between summer and winter is higher (on average about 63C; Levitus and Boyer, 1994). The primary production rate is again enhanced

2384

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

during spring and summer and lowest in winter as in the subtropical regions (PodestaH and Esaias, 1988; Longhurst, 1995; Antoine et al., 1996). Satellite imagery also reveals highest frequencies of classi"ed coccolithophore blooms (presumably E. huxleyi) during austral spring and summer (October}February) and reduced activities during the rest of the year (Brown and Yoder, 1994; Brown and PodestaH , 1997). The blooms are observed along the shelf break and slope from o! the RmH o de la Plata to south of the Malvinas Islands, occurring episodically throughout the year with maxima in austral summer. According to these production characteristics it is unlikely that the biased low ;)Y values in the Malvinas Current region were caused by preferential alkenone  production during the cold season. Even if alkenone production were exclusively con"ned to the more productive spring period, the average temperature value recorded by ;)Y would be numerically similar to annual mean SST. We thus assume  that the ;)Y values in the Malvinas Current region should re#ect a temperature that  is close to the yearly average just as in the tropical}subtropical provinces of the western South Atlantic. This assumption is consistent with the results of other studies, which also indicate that the seasonality in production has no signi"cant e!ect on the ;)Y record of sediments in tropical to temperate oceanic regions (Sonzogni et al.,  1997a, b; Herbert et al., 1998; MuK ller et al., 1998).

5.3. Nutrient availability and growth rate It has been suggested, mainly on the basis of culturing experiments with E. huxleyi and G. oceanica, that alkenone-derived temperatures may be a!ected by biological factors such as genetic di!erences between species (Volkman et al., 1995; Sawada et al., 1996; Conte et al., 1998) and variations in nutrient availability and growth rates (Popp et al., 1998; Epstein et al., 1998). In the Malvinas Current region, the nutrient levels in surface waters are generally higher than in the less productive Brazil Current and the tropical western Atlantic (e.g. Levitus and Boyer, 1994). Therefore, the anomalously low ;)Y values in the western Argentine Basin could be due to the higher nutrient  availability and enhanced growth rates as indicated by some culture experiments (Epstein et al., 1998; Conte et al., 1998). However, this cannot explain the observation that biased and una!ected ;)Y values occur very close together and underneath  the same production system as observed, for example, at sites GeoB 2712 and GeoB 2729 (Fig. 4c and d). Other examples of such situations are transects 9 and 10 in the southwestern Argentine Basin and samples from the southern Santos Plateau, where abrupt ;)Y transitions from biased to normal values occur at water depths between  3949 and 3815 m (Table 1). Furthermore, if nutrient availability and growth rate had a signi"cant e!ect on sedimentary ;)Y values, other coastal upwelling systems, such  as o! California and Namibia, should also show deviating cold ;)Y values, which is  not the case (Herbert et al., 1998; MuK ller et al., 1998; Kirst et al., 1999). The above considerations also suggest that di!erences between species and in the physiological status of alkenone-synthesizing populations as reported by Conte et al. (1998) are unlikely explanations for the anomalous ;)Y ratios in the Argentine Basin. 

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2385

5.4. Erosion of Holocene sediments Another possible explanation for the low ;)Y ratios in the Argentine Basin could  be that the examined surface sediments were of glacial rather than of Holocene age. We cannot completely rule out this possibility, although the shipboard acoustical systems PARASOUND and HYDROSWEEP were employed during all Meteor cruises to "nd suitable coring locations with good sound penetration and an undisturbed sediment cover. This was particularly important in the western Argentine Basin, where erosional features at the continental margin alternate with intervals of almost undisturbed sedimentation (Segl et al., 1994; Bleil et al., 1994). For most of our tropical}subtropical core locations, a Holocene age of the surface sediments is veri"ed by the occurrence of the planktonic foraminifer Globorotalia menardii (Bleil et al., 1993; Wefer et al., 1994) and by oxygen isotope stratigraphy (e.g., RuK hlemann et al., 1996; DuK rkoop et al., 1997; Wol! et al., 1999; B. Donner, S. Mulitza, S. Niebler, unpublished data). It is particularly important to note that G. menardii was also present, albeit in low amounts, in core-tops from stations GeoB 2811}2813 and GeoB 2803}2804 at the Uruguay continental margin (Bleil et al., 1994), where ;)Y appears to be strongly biased (transects 4 and 5, Table 1). Unfortunately, G.  menardii could not be used to verify the Holocene age of more southern sediments in the Argentine Basin as this is beyond the biogeographical limit of this warm-water species (Berger et al., 1985). To further assess the age of the surface sediments, we compared dO values of the benthic foraminifer Cibicides wuellerstorx (d ) with equilibrium dO  values for calcite (d ) at 21 stations '1000 m water depth between 83S and    473S. d was corrected by #0.64 to compensate for the vital e!ect (Duplessy et  al., 1984). Foraminiferal isotopic data and direct measurements of temperature and dO of the bottom water (d ) at 13 of these stations were kindly provided by M.   PaK tzold. In addition, we used GEOSECS d values as integrated for the western   South Atlantic by Paul et al. (1999) and calculated d values applying the modern   relationship between d and salinity (S) in the deep sea (d     (SMOW)"1.529S!53.18, Zahn and Mix, 1991) and salinities of the World Ocean Atlas (Levitus and Boyer, 1994). As this relationship probably is not valid for the Antarctic Intermediate Water, it was applied only at water depths '1500 m. The equilibrium fractionation of d was "nally obtained from the paleo   temperature equation of Erez and Luz (1983), as solved for d "d #75.33!(5675!(17!¹)/0.03) .      The di!erences between predicted and measured benthic isotopic values (*d"d !d ) should be close to zero for Holocene sediments and in the order     of !1.5 for glacial sediments (e.g. Bickert and Wefer, 1996). Our calculations yield *d values between #0.13 and !0.39 (mean$p"!0.13$0.12) for the 21 locations where benthic isotopic values were available (Table 1). Although the deviations between predicted and measured dO values are slightly higher, on average, in the Argentine Basin (*d"!0.24$0.06, n"7) than in the Brazil Basin (*d"0.08$0.11, n"14), these results are consistent with a Holocene age of the sediments.

2386

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

5.5. Lateral particle and sediment displacement The above considerations imply that the low ;)Y values observed in the western  Argentine Basin cannot be convincingly explained by biological processes in surface waters or glacial-age surface sediments. We therefore favour another mechanism and suggest that the discrepancies between alkenone and sea-surface temperatures were caused by lateral displacement of suspended matter and sediments. Surface waters in the region of the Brazil}Malvinas Con#uence are characterised by strong horizontal temperature gradients that may reach 13C km\ (Olson et al., 1988). Gordon (1989), for example, described a situation for October 1984 when the surface-water temperature increased at the frontal zone from about 83 to 143C within 50 km. During this month, the frontal zone was located between 373S and 433S extending from the upper continental slope southward to the continental rise ('5000 m). However, this frontal region is characterised by a high seasonal and annual (spatial) variability and complex eddy patterns (e.g. Legeckis and Gordon, 1982; Con#uence Principal Investigators, 1990) and thus only weakly represented by the averaged SST data of the World Ocean Atlas (Fig. 5). The possible lateral advection of particles during sinking through the water column can be estimated by assuming average sinking rates and current velocities. A realistic range for the sinking rate of bulk particles intercepted by sediment traps is 50}200 m day\ (Siegel and Deuser, 1997). Using these rates and current velocities between 20 and 40 cm s\ for the Malvinas Current and the underlying Antarctic Intermediate Water (Peterson et al., 1996), we estimate a possible northeastward displacement of

Fig. 5. Mean spring sea-surface temperatures (October}December, Levitus and Boyer, 1994) re#ecting the northward #owing Malvinas Current and the southward #owing Brazil Current in the region of the Argentine Basin. Heavy lines indicate our core transects 4}10 (see Table 1).

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2387

Fig. 6. Bathymetric map illustrating the general bottom water #ow patterns in the western Argentine Basin (modi"ed after Flood and Shor, 1988), together with the *¹ isolines from Fig. 3.

particles by 100}1000 km in the upper 1500 m of the water column. Hence, particles carrying the cold water signal of the Malvinas Current could easily be transported into regions with much warmer surface waters (see Fig. 5). A second mechanism may be the redistribution of sediments by bottom currents. The circulation pattern at the western margin of the Argentine Basin (Fig. 6) is characterised by strong boundary and bottom currents (Peterson, 1992; Peterson et al., 1996). These vigorous currents and their interaction with the sea #oor are responsible for a pronounced benthic nepheloid layer (Eittreim et al., 1976; Biscaye and Eittreim, 1977). In addition, the high kinetic energy related to the Brazil}Malvinas Con#uence (Dickson, 1983; Flood and Shor, 1988) may induce short-lived, intense bottom current events (so-called benthic storms), which can play an important role in reworking and redistributing bottom sediments (Gardner and Sullivan, 1981; Hollister and McCave, 1984). Resuspension of sediments and downslope transport is mirrored in particle-size distributions at the Argentine continental margin (Ledbetter, 1986; HoK ppner and Henrich, 1999) as well as in high-resolution seismic records, which reveal extensive areas of mass #ow east of the RmH o de la Plata (Klaus and Ledbetter, 1988). This is the area where, due to the Con#uence, highest eddy kinetic energy is expected and where our data show the highest temperature di!erences between estimated and atlas temperatures (Figs. 3 and 6).

2388

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

The suspended sediments appear to be transported both downslope and northward with the boundary and bottom currents of Antarctic Intermediate Water (AAIW) and Circumpolar Deep Water (CDW) into areas with warmer surface waters. Our coretop ;)Y data give no evidence for a southward transport of suspended sediments with  the North Atlantic Deep Water (NADW) between 2000 and 3000 m (Fig. 3). This suggests, that the bottom velocities of this water mass are signi"cantly lower compared to the AAIW and CDW, and the net transport of sediments thus is northeastward. Additionally, Maamaatuaiahutapu et al. (1992) showed that the NADW turns eastward at approximately 383S, being less important in the regions south of the Brazil}Malvinas Con#uence. It is likely that the processes believed to bias ;)Y in the  western Argentine Basin also will a!ect other paleotemperature proxies in a similar way. It is well known, for example, that diatoms (Burckle and Stanton, 1975; Jones and Johnson, 1984) and even foraminifera (Boltovskoy et al., 1996) are displaced northward over long distances in the western South Atlantic. Another interesting feature of our core-top ;)Y results is the sudden change from  biased to normal values between station GeoB 2826 (3949 m) and station GeoB 2828 (3750 m) at the southern Santos Plateau (Table 1) southwest of the Vema Channel (Figs. 1 and 3). The Vema Channel connects the Argentine Basin with the Brazil Basin (Le Pichon et al., 1971) and is an important passage for the northward #ow of Lower Circumpolar Deep Water (LCDW) below about 4000 m, while the overlying NADW #ows southward (Hogg et al., 1982, 1996). The NADW}LCDW boundary coincides with the depth horizon below which sediments show anomalous alkenone temperatures. The three cores (GeoB 2824, GeoB 2825, and GeoB 2826) exhibiting large alkenone temperature anomalies of 3.9}4.73C compared to atlas SST values (Table 1) were recovered from water depths between 4512 and 3949 m at the southern mouth of the Vema Channel. The other seven core-top samples analysed from this region (GeoB 2817}2821 and GeoB 2827}2828) derived from shallower water depths (2943}3936 m) and reveal no signi"cantly biased alkenone temperatures. The average di!erence between atlas and alkenone SST is !0.43C for these stations ranging from !1.4 to #1.33C (Table 1). The close accordance between the LCDW}NADW boundary and the water depth where ;)Y turns from biased to una!ected values strongly  suggests that the LCDW transports suspended sediments into the northern Argentine Basin, thus biasing the sedimentary ;)Y record at water depths greater than about  4000 m.

6. Summary and conclusions We have compared alkenone ;)Y ratios of surface sediments from the western  South Atlantic to modern atlas temperatures of overlying surface waters (Levitus and Boyer, 1994) and to established culture (Prahl et al., 1988) and core-top calibrations (MuK ller et al., 1998). Our results reveal signi"cant di!erences between the tropical} subtropical western South Atlantic and the Malvinas Current region in the western Argentine Basin that must be considered in future paleotemperature reconstructions:

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2389

(1) Sediments from the western tropical Atlantic, the Brazil Current region, and the South Atlantic Tropical Gyre north of 323S show ;)Y values in agreement with  annual mean atlas SST and the above-mentioned calibrations. Adding these values together with recently published data from the California margin (Herbert et al., 1998) to the global core-top data set compiled by MuK ller et al. (1998) yields an updated equation based on 456 core locations (;)Y "0.033¹#0.047, r"0.960) that is  virtually identical with the previous one (;)Y "0.033¹#0.044, n"370, r"0.958).  This shows that the calibration basis for ;)Y as paleotemperature tool becomes  increasingly stable. (2) By contrast, sediments from the region of the Brazil}Malvinas Con#uence (35}393S) and the Argentine continental slope between 413 and 483S generally show lower alkenone temperatures by 3}63C compared to annual mean atlas SST. Only the ;)Y values of sediments from the upper continental slope at 47}483S and from the  deep Argentine Basin below about 4500 m were found to be in accordance with the surface-water data. The oceanographic characteristics of the Malvinas Current region (shallow photic depth, high nutrient concentrations in the mixed layer, perennial coccolithophore production with spring}summer maximum, minor subsurface production) suggest that the primary (unbiased) ;)Y signal in this region should re#ect the annual mean  temperature of the mixed layer, just as is true in the eastern South Atlantic (MuK ller et al., 1998) and in many other tropical to temperate oceanic regions (Sikes et al., 1991, 1997; Rosell-MeleH et al., 1995; Pelejero and Grimalt, 1997; Sonzogni et al., 1997a, b; Herbert et al., 1998). The low alkenone temperatures derived in the western Argentine Basin can thus not be explained by production below the mixed layer or during the cold season. Di!erences in growth rate and nutrient availability are also unlikely causes, because this cannot explain the simultaneous occurrence of biased and una!ected ;)Y ratios below the same production system. Instead, our results imply  that suspended particles and sediments are displaced northward and downslope by the strong surface and bottom currents prevailing in the western Argentine Basin. In this way, particles and sediments carrying a cold water ;)Y signal of coastal or  southern origin are deposited in o!shore and northern areas with warmer surface waters. Our results further indicate that the northern limit beyond which sedimentary ;)Y values remain una!ected by the described processes must occur somewhere  between 303S and 363S (the limits of our transects 3 and 4) at the continental margins o! Uruguay or Brazil. In the northern Argentine Basin, the water depth above which ;)Y changes from biased ('3940 m) to una!ected values ((3820 m) coincides with  the boundary between the northward #owing LCDW and the southward #owing NADW at about 4000 m. Acknowledgements We thank the o$cers, crew and scientists aboard R/V Meteor for their help during core recovery and sampling, and Dietmar Grotheer and Ralph Kreutz for assistance

2390

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

in the gas chromatographic analyses. We also thank Martina PaK tzold for providing dO data and Nils Andersen, Thorsten Bickert, Gaute Lavik, Jean Lynch-Stieglitz, Ralph Schneider, and Laurence Vidal for discussions and comments. Critical reviews by F.G. Prahl and two anonymous reviewers improved the manuscript and are greatly appreciated. This research was funded by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 261 at Bremen University, Contribution No. 298) and the European Community (contract ENV4-CT97-0564). References Antoine, D., AndreH , J.-M., Morel, A., 1996. Oceanic primary production. 2. Estimation at global scale from satellite (coastal zone color scanner) chlorophyll. Global Biogeochemical Cycles 10 (1), 57}69. Berger, W.H., Killingley, J.S., Metzler, C.V., Vincent, E., 1985. Two-step deglaciation: C-dated highresolution dO records from the tropical Atlantic Ocean. Quaternary Research 23, 258}271. Bickert, T., Wefer, G., 1996. Late Quaternary deep water circulation in the South Atlantic: reconstruction from carbonate dissolution and benthic stable isotopes. In: Wefer, G., Berger, W.H., Siedler, G., Webb, D.J. (Eds.), The South Atlantic: Present and Past Circulation. Springer, Berlin, pp. 599}620. Biscaye, P.E., Eittreim, S.L., 1977. Suspended particulate loads and transports in the nepheloid layer of the abyssal Atlantic Ocean. Marine Geology 23, 155}172. Bleil, U. and cruise participants, 1993. Report and Preliminary Results of METEOR Cruise M 23/2, Rio de Janeiro}Recife, 27.02.}19.03. 1993. Berichte, Fachbereich Geowissenschaften, Vol. 43, UniversitaK t Bremen, Bremen, 133pp. Bleil, U. and cruise participants, 1994. Report and Preliminary Results of METEOR Cruise M 29/2, Montevideo}Rio de Janeiro, 15.07.}08.08. 1994. Berichte, Fachbereich Geowissenschaften, Vol. 59, UniversitaK t Bremen, Bremen, 153pp. Boltovskoy, E., Boltovskoy, D., Correa, N., Brandini, F., 1996. Planktic foraminifera from the southwestern Atlantic (30}603S): species-speci"c patterns in the upper 50 m. Marine Micropaleontology 28, 53}72. Brassell, S.C., 1993. Applications of biomarkers for delineating marine paleoclimatic #uctuations during the Pleistocene. In: Engel, M.H., Macko, S.A. (Eds.), Organic Geochemistry. Plenum Press, New York, pp. 699}738. Brassell, S.C., Eglinton, G., Marlowe, I.T., P#aumann, U., Sarnthein, M., 1986. Molecular stratigraphy: a new tool for climatic assessment. Nature 320, 129}133. Brown, C.W., Yoder, J.A., 1994. Coccolithophorid blooms in the global ocean. Journal of Geophysical Research 99 (C4), 7467}7482. Brown, C.W., PodestaH , G.P., 1997. Remote sensing of coccolithophore blooms in the western South Atlantic Ocean. Remote Sensing of Environment 60, 83}91. Burckle, L.H., Stanton, D., 1975. Distribution of displaced Antarctic diatoms in the Argentine Basin. Third Symposium on Recent and Fossil Marine Diatoms, Lamont-Doherty Geological Observatory Contribution, Vol. 2235, Kiel, Germany, pp. 283}292. Campos, E.J.D., Ikeda, Y., Castro, B.M., Gaeta, S.A., Lorenzzetti, J.A., Stevenson, M.R., 1996. Experiment studies circulation in the western South Atlantic. EOS 77 (27), 253, 259. Carreto, J.I., Lutz, V.A., Carignan, M.O., Cucchi Colleoni, A.D., De Marco, S.G., 1995. Hydrography and chlorophyll a in a transect from the coast to the shelf-break in the Argentinian Sea. Continental Shelf Research 15 (2/3), 315}336. Con#uence Principal Investigators, 1990. Con#uence 1988}1990: an intensive study of the southwestern Atlantic. EOS 71(41), 1131}1137. Conte, M.H., Thompson, A., Lesley, D., Harris, R.P., 1998. Genetic and physiological in#uences on the alkenone/alkenoate versus growth temperature relationship in Emiliania huxleyi and Gephyrocapsa oceanica. Geochimica et Cosmochimica Acta 62 (1), 51}68. Davis, J.C. (Ed.), 1986. Statistics and Data Analysis in Geology. Wiley, New York, 646pp. Dickson, R.R., 1983. Global summaries and intercomparisons: #ow statistics from long-term current meter moorings. In: Robinson, A.R. (Ed.), Eddies in Marine Science. Springer, Heidelberg, pp. 278}353.

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2391

Duplessy, J.-C., Shackleton, N.J., Matthews, R.K., Prell, W., Ruddiman, W.F., Caralp, M., Hendy, C.H., 1984. C record of benthic foraminifera in the last interglacial ocean: implications for the carbon cycle and the global deep water circulation. Quaternary Research 21, 225}243. DuK rkoop, A., Hale, W., Mulitza, S., PaK tzold, J., Wefer, G., 1997. Late Quaternary variations of sea surface salinity and temperature in the western tropical Atlantic: evidence from dO of Globigerinoides sacculifer. Paleoceanography 12 (6), 764}772. Eittreim, S., Thorndike, E.M., Sullivan, L., 1976. Turbidity distribution in the Atlantic Ocean. Deep-Sea Research 23, 1115}1127. Epstein, B.L., D'Hondt, S., Quinn, J.G., Zhang, J., Hargraves, P.E., 1998. An e!ect of dissolved nutrient concentrations on alkenone-based temperature estimates. Paleoceanography 13 (2), 122}126. Erez, J., Luz, B., 1983. Experimental paleotemperature equation for planktonic foraminifera. Geochimica et Cosmochimica Acta 47, 1025}1031. Flood, R.D., Shor, A.N., 1988. Mud waves in the Argentine Basin and their relationship to regional bottom circulation patterns. Deep-Sea Research 35 (6), 943}971. Gardner, W.D., Sullivan, L.G., 1981. Benthic storms: temporal variability in a deep-ocean nepheloid layer. Science 213, 329}331. Garzoli, S.L., 1993. Geostrophic velocity and transport variability in the Brazil}Malvinas Con#uence. Deep-Sea Research I 40 (7), 1379}1403. Gayoso, A.M., 1995. Bloom of Emiliania huxleyi (Prymnesiophyceae) in the western South Atlantic Ocean. Journal of Plankton Research 17 (8), 1717}1722. Georgi, D.T., 1981. Circulation of bottom waters in the southwestern South Atlantic. Deep-Sea Research 28 (9), 959}979. Gordon, A.L., 1989. Brazil}Malvinas Con#uence * 1984. Deep-Sea Research 36 (3), 359}384. Gordon, A.L., Greengrove, C.L., 1986. Geostrophic circulation of the Brazil}Falkland Con#uence. DeepSea Research 33 (5), 573}585. Herbert, T.D., Schu!ert, J.D., Thomas, D., Lange, C., Weinheimer, A., Peleo-Alampay, A., Herguera, J.-C., 1998. Depth and seasonality of alkenone production along the California margin inferred from a core top transect. Paleoceanography 13 (3), 263}271. Hogg, N., Biscaye, P.E., Gardner, W., Schmitz Jr., W.J., 1982. On the transport and modi"cation of Antarctic Bottom Water in the Vema Channel. Journal of Marine Research 40 (Suppl.), 231}263. Hogg, N.G., Owens, W.B., Siedler, G., Zenk, W., 1996. Circulation in the deep Brazil Basin. In: Wefer, G., Berger, W.H., Siedler, G., Webb, D.J. (Eds.), The South Atlantic: Present and Past Circulation. Springer, Berlin, pp. 249}260. Hollister, C.D., McCave, I.N., 1984. Sedimentation under deep-sea storms. Nature 309, 220}225. HoK ppner, R., Henrich, R., 1999. Kornsortierungsprozesse am Argentinischen Kontinentalhang anhand von Siltkorn-Analysen. Zentralblatt fuK r Geologie und PalaK ontologie 7}9, 897}905. Jones, G.A., Johnson, D.A., 1984. Displaced diatoms in Vema Channel sediments: Late Pleistocene/ Holocene #uctuations in AABW #ow. Marine Geology 58, 165}186. Kirst, G.J., Schneider, R.R., MuK ller, P.J., von Storch, I., Wefer, G., 1999. Late Quaternary temperature variability in the Benguela Current system derived from alkenones. Quaternary Research 52, 92}103. Klaus, A., Ledbetter, M.T., 1988. Deep-sea sedimentary processes in the Argentine Basin revealed by high-resolution seismic records (3.5 kHz echograms). Deep-Sea Research 35 (6), 899}917. Knappertsbusch, M., 1993. Geographic distribution of living and Holocene coccolithophores in the Mediterranean Sea. Marine Micropaleontology 21, 219}247. Le Pichon, X., Ewing, M., Truchan, M., 1971. Sediment transport and distributions in the Argentine Basin. 2. Antarctic bottom current passage into the Brazil Basin. In: Ahrens, L.H., Press, F., Runcorn, S.K., Urey, H.C. (Eds.), Physics and Chemistry of the Earth, Vol. 8. Pergamon Press, Oxford, pp. 29}48. Ledbetter, M.T., 1986. Bottom-current pathways in the Argentine Basin revealed by mean silt particle size. Nature 321, 423}425. Legeckis, R., Gordon, A.L., 1982. Satellite observations of the Brazil and Falkland currents*1975 to 1976 and 1978. Deep-Sea Research 29 (3A), 375}401. Levitus, S., Boyer, T., 1994. World Ocean Atlas. 4: temperatures, NOAA Atlas NESDIS 4. U.S. Government Printing O$ce.

2392

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

Longhurst, A., 1993. Seasonal cooling and blooming in tropical oceans. Deep-Sea Research I 40 (11/12), 2145}2165. Longhurst, A., 1995. Seasonal cycles of pelagic production and consumption. Progress in Oceanography 36, 77}167. Maamaatuaiahutapu, K., Garc7 on, V.C., Provost, C., Boulahdid, M., Osiro!, A.P., 1992. Brazil}Malvinas Con#uence: water mass composition. Journal of Geophysical Research 97 (C6), 9493}9505. Maamaatuaiahutapu, K., Garc7 on, V., Provost, C., Mercier, H., 1998. Transports of the Brazil and Malvinas Currents at their Con#uence. Journal of Marine Research 56, 417}438. Marlowe, I.T., 1984. Lipids as paleoclimatic indicators. Ph.D. Thesis, University of Bristol, Bristol, 273pp. Morel, A., Berthon, J.-F., 1989. Surface pigments, algal biomass pro"les, and potential production of the euphotic layer: relationships reinvestigated in view of remote-sensing applications. Limnology and Oceanography 34 (8), 1545}1562. MuK ller, P.J., Cepek, M., Ruhland, G., Schneider, R.R., 1997. Alkenone and coccolithophorid species changes in late Quaternary sediments from the Walvis Ridge: implications for the alkenone paleotemperature method. Palaeogeography, Palaeoclimatology, Palaeoecology 135, 71}96. MuK ller, P.J., Kirst, G., Ruhland, G., von Storch, I., Rosell-MeleH , A., 1998. Calibration of the alkenone paleotemperature index ;)Y based on core-tops from the eastern South Atlantic and the global ocean  (603N}603S). Geochimica et Cosmochimica Acta 62 (10), 1757}1772. Olson, D.B., PodestaH , G.P., Evans, R.H., Brown, O.B., 1988. Temporal variations in the separation of Brazil and Malvinas Currents. Deep-Sea Research 35 (12), 1971}1990. Paul, A., Mulitza, S., PaK tzold, J., Wol!, T., 1999. Simulation of oxygen isotopes in a global ocean model. In: Fischer, G., Wefer, G. (Eds.), Use of Proxies in Paleoceanography * Examples from the South Atlantic. Springer, Berlin, pp. 655}686. Pelejero, C., Grimalt, J.O., 1997. The correlation between the ;) index and sea surface temperatures in the  warm boundary: The South China Sea. Geochimica et Cosmochimica Acta 61 (22), 4789}4797. Peterson, R.G., 1992. The boundary currents in the western Argentine Basin. Deep-Sea Research 39 (3/4), 623}644. Peterson, R.G., Stramma, L., 1991. Upper-level circulation in the South Atlantic Ocean. Progress in Oceanography 26, 1}73. Peterson, R.G., Johnson, C.S., Krauss, W., Davis, R.E., 1996. Lagrangian measurements in the Malvinas Current. In: Wefer, G., Berger, W.H., Siedler, G., Webb, D.J. (Eds.), The South Atlantic: Present and Past Circulation. Springer, Berlin, pp. 239}247. PodestaH , G.P., Esaias, W.E., 1988. Satellite-derived phytoplankton pigment concentrations along the shelf break o! Argentina, 1979}1980. EOS 69, 1144. Popp, B.D., Kenig, F., Wakeham, S.G., Laws, E.A., Bidigare, R.R., 1998. Does growth rate a!ect ketone unsaturation and intracellular carbon isotopic variability in Emiliania huxleyi? Paleoceanography 13 (1), 35}41. Prahl, F.G., Wakeham, S.G., 1987. Calibration of unsaturation patterns in long-chain ketone compositions for paleotemperature assessment. Nature 330, 367}369. Prahl, F.G., Muehlhausen, L.A., Zahnle, D.L., 1988. Further evaluation of long-chain alkenones as indicators of paleoceanographic conditions. Geochimica et Cosmochimica Acta 52, 2303}2310. Prahl, F.G., Collier, R.B., Dymond, J., Lyle, M., Sparrow, M.A., 1993. A biomarker perspective on prymnesiophyte productivity in the northeast Paci"c Ocean. Deep-Sea Research I 40 (10), 2061}2076. Provost, C., Gana, S., Garc7 on, V., Maamaatuaiahutapu, K., England, M., 1995. Hydrographic conditions in the Brazil}Malvinas con#uence during austral summer 1990. Journal of Geophysical Research 100 (C6), 10655}10678. Reid, J.L., Nowlin Jr., W.D., Patzert, W.C., 1977. On the characteristics and circulation of the Southwestern Atlantic Ocean. Journal of Physical Oceanography 7, 62}91. Rosell-MeleH , A., Eglinton, G., P#aumann, U., Sarnthein, M., 1995. Atlantic core-top calibration of the ;) index as a the sea-surface palaeotemperature indicator. Geochimica et Cosmochimica Acta 59 (15),  3099}3107.

A. Benthien, P.J. Mu( ller / Deep-Sea Research I 47 (2000) 2369}2393

2393

RuK hlemann, C., Frank, M., Hale, W., Mangini, A., Mulitza, S., MuK ller, P.J., Wefer, G., 1996. Late Quaternary productivity changes in the western equatorial Atlantic: Evidence from Th-normalized carbonate and organic carbon accumulation rates. Marine Geology 135, 127}152. Sawada, K., Handa, N., Shiraiwa, Y., Danbara, A., Montani, S., 1996. Long-chain alkenones and alkyl alkenoates in the coastal and pelagic sediments of the northwest North Paci"c, with special reference to the reconstruction of Emiliania huxleyi and Gephyrocapsa oceanica ratios. Organic Geochemistry 24 (8/9), 751}764. Segl, M. and cruise participants, 1994. Report and Preliminary Results of METEOR Cruise M 29/1, Buenos Aires * Montevideo, 17.06.}13.07. 1994. Berichte, Fachbereich Geowissenschaften, Vol. 58, UniversitaK t Bremen, Bremen, 94pp. Siegel, D.A., Deuser, W.G., 1997. Trajectories of sinking particles in the Sargasso Sea: modeling of statistical funnels above deep-ocean sediment traps. Deep-Sea Research I 44 (9}10), 1519}1541. Sikes, E.L., Farrington, J.W., Keigwin, L.D., 1991. Use of the alkenone unsaturation ratio ;) to determine  past sea surface temperatures: core-top SST calibrations and methodology considerations. Earth and Planetary Science Letters 104, 36}47. Sikes, E.L., Volkman, J.K., Robertson, L.G., Pichon, J.-J., 1997. Alkenones and alkenes in surface waters and sediments of the Southern Ocean: implications for paleotemperature estimation in polar regions. Geochimica et Cosmochimica Acta 61 (7), 1495}1505. Sonzogni, C., Bard, E., Rostek, F., Lafont, R., Rosell-MeleH , A., Eglinton, G., 1997a. Core-top calibration of the alkenone index vs. sea surface temperature in the Indian Ocean. Deep-Sea Research II 44 (6}7), 1445}1460. Sonzogni, C., Bard, E., Rostek, F., Dollfus, D., Rosell-MeleH , A., Eglinton, G., 1997b. Temperature and salinity e!ects on alkenone ratios measured in surface sediments from the Indian Ocean. Quaternary Research 47, 344}355. Volkman, J.K., Barrett, S.M., Blackburn, S.I., Sikes, E.L., 1995. Alkenones in Gephyrocapsa oceanica: implications for studies of paleoclimate. Geochimica et Cosmochimica Acta 59 (3), 513}520. Wefer, G. and cruise participants, 1994. Report and Preliminary Results of METEOR Cruise M 23/3, Recife * Las Palmas, 21.03.}12.04. 1993. Berichte, Fachbereich Geowissenschaften, Vol. 44, UniversitaK t Bremen, Bremen, 71pp. Wol!, T., Mulitza, S., RuK hlemann, C., Wefer, G., 1999. Response of the tropical Atlantic thermocline to late Quaternary trade wind changes. Paleoceanography 14 (3), 374}383. Zahn, R., Mix, A.C., 1991. Benthic foraminiferal dO in the ocean's temperature}salinity}density "eld: constraints on ice age thermohaline circulation. Paleoceanography 6 (1), 1}20.