Late Quaternary changes in biogenic opal fluxes in the Southern Indian Ocean

Available online at www.sciencedirect.com R

Marine Geology 202 (2003) 143^158 www.elsevier.com/locate/margeo

Late Quaternary changes in biogenic opal £uxes in the Southern Indian Ocean L. Dezileau a;
, J.L. Reyss b , F. Lemoine c b

a Programa Regional de Oceanogra¢a Fisica y Clima, Universidad de Concepcion, Concepcion, Chile Laboratoire des Sciences du Climat et de l’Environnement, Laboratoire mixte CEA^CNRS, Gif-sur-Yvette, 91198 Gif-sur-Yvette Cedex, France c Institut de Radioprotection et de Surete¤ Nucle¤aire, Fontenay aux Roses Cedex, France

Received 12 April 2002; received in revised form 15 July 2003; accepted 14 August 2003

Abstract Late Quaternary sedimentary and paleoenvironmental conditions in the southern Indian Ocean have been reconstructed from radioisotope and proxy element profiles (biogenic opal and organic carbon) measured on five sediment cores taken along a transect across the Indian sector of the Antarctic Circumpolar Current. Dissolutioncorrected opal rain rates were used to reconstruct past changes of opal productivity for this region. Records from these five cores indicate that opal productivity during glacial periods was lower than presently recorded south of the Antarctic Polar Front (APF), probably due to increased ice cover. North of APF, opal productivity was slightly greater during glacial periods than during the Holocene, probably in response to (1) the northward migration of the APF by approximately 5‡ latitude, (2) a northward transport of Si from the Antarctic Zone, and (3) an increase of Fe, necessary for opal-producing organisms, via upwelling and the erosion of the Kerguelen Plateau. We also invoke a decoupling between opal burial and organic carbon flux to the seabed to explain the variation in buried Si/C ratio between glacial and interglacial sediment. This decoupling is principally explained by better organic carbon preservation in the glacial sediments due to strong sediment focussing. An increase in glacial export paleoproductivity is not supported by the data, implying that bioproductivity variations in the Southern Indian Ocean are unlikely to have contributed to the glacial drawdown of atmospheric CO2 inferred from ice core data. : 2003 Elsevier B.V. All rights reserved. Keywords: Southern Ocean; Indian sector; biogenic opal; organic carbon; thorium; paleoproductivity

1. Introduction In order to understand biogenic silica produc* Corresponding author. Present address: Laboratoire de Dynamique de la Lithosphe're, Universite¤ de Montpellier 2, 34095 Montpellier cedex 05, France. E-mail address: [email protected] (L. Dezileau).

tion in the Southern Ocean, attention has focused on the accumulation history of opal in sediments (Mortlock et al., 1991; Charles et al., 1991; Bareille et al., 1991). However, there are some uncertainties, which have led to false conclusions when evaluating paleoproductivity from biogenic opal accumulation rates. First, preservation states of biogenic opal are expected to occur (Pichon et al., 1992). It is suspected that opal dissolution

0025-3227 / 03 / $ ^ see front matter : 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0025-3227(03)00283-4

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

144

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

occurs mostly at the sediment^water interface rather than deeper down core in the pore water (Broecker and Maier-Reimer, 1992). The extent of dissolution of biogenic silica depends on a number of factors including the residence time of biogenic silica in the di¡usion layer, the bulk accumulation rate, dilution by other components, and the assemblage composition because some diatom species are more prone to dissolution than others (Pichon et al., 1992). Second, opal accumulation rates in sediments cannot distinguish between changes in accumulation resulting from variations in productivity controlling vertical rain rate and those resulting from variations in sediment redistribution by currents. In the Southern Ocean, the latter uncertainty is especially signi¢cant because of the presence of relatively strong currents associated with the Antarctic Circumpolar Current (ACC) and Antarctic Bottom Water (FrancOois et al., 1993; Dezileau et al., 2000; Frank et al., 2000). Because sediment redistribution is common in the Southern Ocean, we use the 230 Th-normalization procedure to derive regional average pelagic rain rates preserved in the sediments from ¢ve cores taken along a transect across the ACC from 54‡S to 43‡S. We also use a transfer function to quantify silica dissolution in sediments developed by Pichon et al. (1992) in order to correct opal £uxes for dissolution. Using these cores, Bareille et al. (1991) attempted to reconstruct biogenic accumulation rates in order to estimate the opal paleoproductivity of this area. However, discrepancies between sediment accumulation rates and vertical rain rates require that the paleoproductivity reconstructions of Bareille et al. (1991) be reinterpreted. Past changes in productivity may be assessed, in principle, from the record of biogenic detritus buried in the seabed. Diatoms largely drive the export of organic carbon and nutrients to the deep sea in the Southern Ocean. Thus, opal rain rates are expected to re£ect the productivity of overlying surface waters. However, recent studies indicate that the carbon £ux/opal £ux ratio in glacial sediments of the Southern Ocean di¡ers from the corresponding ratio in interglacial sediments (Kumar et al., 1995; Anderson et al., 1998;

Bareille et al., 1998). Hence, in this paper we will also report a comparison between vertical opal rain rates and organic carbon rain rates in order to study the buried Si/C ratio. The ¢rst aim of this paper is to estimate and discuss the variations in glacial/Holocene opal productivity in the Southern Indian Ocean using variations of dissolution-corrected opal rain rates as a proxy. The second aim is to study the evolution of the buried Si/C ratio between glacial and Holocene periods.

2. Materials and methods 2.1. Oceanography and core material Today, the Southern Ocean can be divided into di¡erent zones de¢ned by a succession of three meridional frontal structures. The ¢rst front, located around 40‡S in the Indian sector and known as the Subtropical Front (STF), constitutes the northern border of the Southern Ocean. It is characterized by a narrowing of the isotherm contained between 12‡C and 14‡C (Deacon, 1966; Belkin and Gordon, 1996). The second front, near 45‡S in the Indian sector, is named the Subantarctic Front (SAF). This front is characterized by a pronounced minimum in the salinity of its subsurface water (Whitworth and Nowlin (1987)) and is di⁄cult to distinguish from the STF around the Kerguelen sector. The third front, near 52‡S in Indian sector, is the Antarctic Polar Front (APF). It consists of a narrowing isotherm contained between 3‡C and 5‡C during summer (Park et al., 1991; Belkin and Gordon, 1996). The APF is at the northern limit of wind-driven upwelling of deep water and is characterized by a pronounced gradient in dissolved silica (Van Bennekom et al., 1988). Besides a small southward increase in dissolved nitrate and phosphate, a drastic increase in dissolved silica gives way to enhanced opal productivity. The area delimited by the STF and SAF is usually called the ‘Subantarctic Zone’ (SAZ) and that delimited by the SAF and APF is known as the ‘Polar Frontal Zone’ (PFZ). The Antarctic Zone (AZ) is the entire region south of the APF (Fig. 1). To study the

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

145

Fig. 1. Map showing the location of the ¢ve cores sampled from three di¡erent zones within the Indian sector of the Southern Ocean. The average position of the Subtropical Front (STF), the Subantarctic Front (SAF) and the Antarctic Polar Front (APF) are indicated, after Belkin and Gordon (1996). Circumpolar zones include the Subantarctic Zone (SAZ), lying between the STF and the SAF, and the Polar Frontal Zone (PFZ), lying between the APF and the SAF. The AZ is the entire region south of the APF. Bottom relief and place names of the study area are also indicated: CP, Crozet Plateau; KP, Kerguelen Plateau; SEIR, South East Indian Ridge; SIB, South Indian Basin.

changes in biological productivity for the Indian sector of the Antarctic Ocean, it is necessary to sample the various subsystems of the ocean. For this reason, we selected ¢ve cores that characterize each of the three major zones, the SAZ, the PFZ and the AZ. Five gravity cores from the southern £ank of the Southeast Indian Ridge (Fig. 1; Table 1) were collected between 3000 and 3500 m water depth beneath the eastward-£owing ACC. Core MD 88773 was previously studied by FrancOois et al. (1997). We also examined one core from the Kerguelen Plateau (MD 84552), which was collected at 1800 m water depth. All cores were recovered

Table 1 Location of the gravity cores Cores

Latitudes (‡S)

Longitudes (‡E)

Water depth (m)

MD MD MD MD MD MD

43.5 46.5 46.1 46 52.6 54.9

79.8 88.1 90.1 96.5 109.5 73.8

3205 3460 3420 3290 2460 1780

94-102 94-104 88-769 88-770 88-773 84-552

during the APSARA II (1984) and IV (1988) and PACIMA (1994) ocean drilling programs. The gravity cores were sampled and then split into separate slices with a resolution between 5 and 25 cm per sample depending on macroscopic changes in lithology and previously determined stratigraphy. The measurement of each sample represents the average value for the corresponding core section. 2.2. Stratigraphic framework and age model A combination of results derived from oxygen isotope measurements, siliceous microfossil bio£uctuation, and AMS 14 C ages was used to establish the stratigraphy of each core (Dezileau et al., 2000). 2.3. Analytical methods U and Th isotopes were measured by K spectrometry following chemical preparation as described in Dezileau et al. (2000). All data have been previously published in Dezileau et al. (2000). The biogenic opal data (Table 2) used in this study have been previously published in Ba-

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

146

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

Table 2 Data used in this study Cores MIS

102

770

104

769

552

773

1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3

Vertical rain rates

Opal rain rates

(g cm32 ka31 )

(g cm32 ka31 )

1.17 U 0.07 1.25 U 0.08 1.12 U 0.07 1.49 U 0.09 0.91 U 0.05 1.04 U 0.06 0.99 U 0.06 1.02 U 0.06 1.12 U 0.07 1.16 U 0.07 0.85 U 0.05 0.89 U 0.05 2.25 U 0.14 1.12 U 0.07 0.97 U 0.06 1.50 U 0.09 1.60 U 0.10 1.25 U 0.08

0.06 U 0.01 0.08 U 0.01 0.08 U 0.01 0.10 U 0.01 0.48 U 0.03 0.71 U 0.04 0.21 U 0.01 0.25 U 0.02 0.37 U 0.02 0.07 U 0.00 0.18 U 0.01 0.27 U 0.02 1.78 U 0.11 0.42 U 0.03 0.37 U 0.02 1.30 U 0.08 1.10 U 0.07 1.00 U 0.06

Dissolution

Organic C

(%)

Dissolution-corrected opal rain rates (g cm32 ka31 )

(%)

Vertical organic C rain rates (mg cm32 ka31 )

30 60 60 58 38 49 30 30 30 30 30 40 51 74 75 54 55 59

0.09 U 0.01 0.20 U 0.02 0.20 U 0.02 0.24 U 0.02 0.77 U 0.08 1.39 U 0.14 0.30 U 0.03 0.36 U 0.04 0.53 U 0.05 0.10 U 0.01 0.26 U 0.03 0.43 U 0.04 3.63 U 0.36 1.62 U 0.16 1.48 U 0.15 2.80 U 0.28 2.44 U 0.24 2.44 U 0.24

0.13 0.19 ^ 0.25 1.1 ^ ^ ^ ^ 0.27 0.67 ^ 0.37 0.57 ^ 0.66 0.49 ^

1.5 U 0.1 2.3 U 0.2 ^U^ 3.7 U 0.3 10.1 U 0.8 ^U^ ^U^ ^U^ ^U^ 3.1 U 0.2 6.1 U 0.5 ^U^ 8.3 U 0.7 6.4 U 0.6 ^U^ 9.9 U 0.8 7.8 U 0.6 ^U^

Vertical rain rates refer to 230 Th normalized total mass accumulation rates. Opal rain rates are corrected for sediment redistribution but not opal dissolution. Dissolution (%) is estimated using the diatom transfer function of Pichon et al. (1992). Dissolution-corrected opal rain rates are estimated by dividing opal rain rates by (1-% dissolution). The organic carbon data has been previously published in Bareille et al. (1998). Organic carbon rain rates are corrected for sediment redistribution. Vertical rain rates of core MD 88773 are from FrancOois et al. (1997).

reille et al. (1991, 1998) and were compiled in Dezileau et al. (2000). The opal content has been obtained by chemical dissolution following the procedure of Mortlock and Froelich (1989) with a general precision better than 2%. The organic carbon data (Table 2) used in this study are from Bareille et al. (1998). Organic carbon determinations were made after decarbonatation using a Leco carbon^sulfur analyzer.

3. Results 3.1. Past changes in sea surface temperatures In order to understand past variations in hydrology, Huston (1980) and Prell (1985) have developed a method allowing evaluation of past sea surface temperature (SST), based on counting foraminifera found in deep-sea sediments. Labeyrie et al. (1996), Salvignac (1998) and Lemoine (1998) have used this method on sediment cores MD

94102 (43‡S), MD 88769 (46‡S), MD 88770 (46‡S) and MD 84552 (55‡S) to reconstruct past SST variations for the central part of the Indian sector of the Southern Ocean (from 60‡E to 100‡E in longitude). The results of these reconstructions are shown in Fig. 2a,b. Each ¢gure shows the typical increase in SST from the Last Glacial Maximum (LGM) to the Holocene period. Moreover, if we compare the SST obtained from core MD 88769 and MD 88770 (Fig. 2a), located roughly at the same latitude (46‡S), we observe large di¡erences in SST for most part of marine isotope stage 3 (MIS 3), reaching more than 2.5‡C (around 5.5‡C for core MD 88769 and 3.0‡C for core MD 88770). These large SST di¡erences indicate that synchronous global temperature variations do not occur in the studied area. The di¡erences in SST imply the existence of a large meridional gradient in temperature and suggest the presence of a frontal pattern, very similar to modern day APF, between cores MD 88769 and MD 88770. During MIS 2,

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

Fig. 2. (a) SSTs calculated using foraminiferal transfer functions for the last 60 kyr for cores MD 94102, MD 88770 and MD 88769 (Salvignac, 1998; Labeyrie et al., 1996; Lemoine, 1998). (b) Summer SSTs reconstructed for the LGM for sediment cores MD 88770 and MD 84552.

the SST gradient between cores MD 88769 and MD 88770 was lower. This is probably due to the presence of the APF at the location of the two cores. During the Holocene period, we can identify a short warming episode from 12 to 9 ky BP from the SST reconstruction for core MD 84552 (Fig. 2b) that is not synchronous with the global warming following the LGM recorded by core MD 88770. Lemoine (1998) attributed this warming episode to a temporary southward incursion of the APF. In conclusion, these data support the idea that a thermal frontal structure,

147

similar to the modern day APF, has persisted over the last 60 kyr with a northward shift of the APF of about 5‡ latitude during the glacial period compared with the Holocene. These results are consistent with those from other studies of the Atlantic and Indian Oceans (Charles et al., 1991; Mortlock et al., 1991; Bareille et al., 1998). Some workers argue that the fronts could not have migrated signi¢cantly because their position is constrained by interaction of the £ow of the ACC with bottom topography throughout much of Southern Ocean (Moore et al., 2000; Anderson et al., 2002). This argument is unlikely to have in£uence in our case, because most cores are located away from large topographic features (with the exception of the core MD 84552), and are located above the South Indian Basin. In the area represented by these cores, a northward shift in the mean location of the APF during the last glacial period is considered possible. For this reason, throughout this paper, reference to the location of the APF implies that it is 5‡ latitude north during glacial periods and at its present-day position during the Holocene. This has an implication on the position of the Southern Westerlies because ocean circulation and the position of the APF are directly related to the mean position of these winds. Our results suggest a northward shift of the Southern Westerlies during MIS 2/3 compared with the Holocene. The direction of Late Pleistocene shifts of the westerly wind belt ^ either equatorward (e.g. Heusser, 1989) or poleward (Markgraf, 1989) ^ has been debated. The Southern Andean region is ideal for monitoring late Quaternary paleoclimate at mid-latitudes in the Southern Hemisphere, because it is one of the few land areas crossed by the Southern Westerlies. Therefore, paleocontinental and paleoceanographic records along a meridional transect can be used to track the ice-age history of the Southern Westerly stormtracks. Recent results based on paleoclimatic modelling (Hulton et al., 1994), glacier £uctuations (Clapperton et al., 1995), palynological paleoclimatic studies (Moreno, 1997; Moreno et al., 1999) and marine sediments (Lamy et al., 2000, 2001, 2002) suggest an equatorward shift of the Southern Westerlies during the LGM in

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

148

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

Chile. These results which are not consistent with the majority of climate models (Wyroll et al., 2000; Valdes, 2000) are consistent with our data. 3.2. Reconstruction of opal rain rates corrected for dissolution 3.2.1. Reconstruction of opal rain rates: 230 Th method The 230 Th-normalization procedure used here is described in detail by Bacon (1984), Suman and Bacon (1989) and FrancOois et al. (1990). Brie£y, 230 Th is produced at a rate of 2.63 1035 dpm cm33 kyr31 in the oceanic water column from the radioactive decay of dissolved 234 U in seawater. Because of the long (4 105 a) residence time of U in the ocean, its concentration in seawater shows no signi¢cant geographical variations (Ku et al., 1977). Since Th is strongly adsorbed by marine particulate matter, its residence time in the water column is short ( 6 40 a), and the rate of 230 Th delivery to the sediments is practically equal to the production rate, i.e. none of the 230 Th is lost by decay in the water column. Measurements of 230 Th with deep-sea sediment traps (Anderson et al., 1983; Yu et al., 2001) and a study of the global distribution of 230 Th £ux to ocean sediments constrained by GCM modelling (Henderson et al., 1999) have con¢rmed that the production and the vertical £ux of 230 Th carried by settling particles are nearly in balance. Exceptions have been found in areas where advection of water masses with short residence times occur, such as the case shown for the Atlantic sector of the Southern Ocean (Rutgers van der Loe¡ and Berger, 1993; Walter et al., 1997). Thorium-230 normalized £ux calculations are based on the assumption that the £ux of 230 Th to the sea£oor is constant and equal to the production rate (Bacon, 1984; Suman and Bacon, 1989; FrancOois et al., 1990). Excess 230 Th activity in settling particulates is then inversely related to total mass £ux. Excess 230 Th activities (decay- and ingrowth-corrected) in sediments can thus be used as reference against which the vertical £ux of other sedimentary components can be calculated : L :Z:f i F i ¼ 230

Thex 

ð1Þ

where Fi is the rain rate of the sedimentary component i; L is the constant production rate of 230 Th in the water column (2.63U105 dpm cm33 kyr31 ); Z is the water depth; fi is the wt% of sedimentary component i; 230 Thex ‡ is the activity of decay- and ingrowth-corrected excess 230 Th in sediment. This approach has been successfully applied where lateral re-sedimentation is important (Suman and Bacon, 1989; FrancOois et al., 1993, 1997; Kumar et al., 1995; Anderson et al., 1998, 2002; Frank et al., 2000; Dezileau et al., 2000; Chase et al., 2003). Down-core records of opal rain rate (corrected for sediment redistribution but not opal dissolution) are shown in Fig. 3. Application of Eq. 1 to 230 Thex ‡ data reveals relatively low opal rain rates for the four northern cores with values below 0.15 g cm32 kyr31 throughout the Holocene (Fig. 3). In cores MD 88769, MD 94104 and MD 88770 opal rain rates are modestly higher during MIS 2 and 3 than during the Holocene (0.2^1 g cm32 kyr31 ). In core MD 84552, which is located on the Kerguelen Plateau, opal rain rates during the Holocene are represented by values of up to 3 g cm32 kyr31 while lower values are evident during glacial periods. For core MD 88773 (not shown on Fig. 3) relatively high opal rain rates (around 1 g/cm2 kyr) are recorded throughout the investigated period of time (FrancOois et al., 1997). 3.2.2. Comparison between the opal accumulation rate and the opal rain rate In Fig. 4, a comparison of the average opal accumulation and the opal rain rate for the Holocene, MIS 2 and MIS 3 along the transect of this study is presented to illustrate the importance of correcting for sediment redistribution when reconstructing particle £uxes in the past. Without correction for sediment redistribution, the pattern of opal accumulation suggests that an area of high particle £uxes remained at a position south of the APF during the past 40 kyr (Fig. 4). In a previous study (Dezileau et al., 2000), we investigated the e¡ects of circulation of water masses on deep marine sedimentation. We showed that lateral transport of biogenous and detrital material contributes 50^90% of sediment input

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

149

tion for sediment redistribution (Fig. 4b), except that opal rain rates are much lower than opal accumulation rates for the South Indian Basin. The variation of Holocene opal rain rates with latitude is consistent with the known present-day trends in biological productivity (Pondaven et al., 2000a,b), i.e. relatively high rain rates found south of the APF and minimum values occurring at the SAF (Fig. 4b). We notice that opal rain rates in the PFZ for the last glacial period were slightly higher than during the Holocene. The high opal rain rates during MIS 2 and 3 in the Antarctic zone were modestly lower than those found during the Holocene in MD 88773 (FrancOois et al., 1997), and signi¢cantly lower than those of the Holocene recorded in the core from the Kerguelen Plateau (MD 84552). After correction for sediment redistribution, the opal rain rate variations can be explained by changes in two factors: biogenic opal productivity and/or dissolution. Pichon et al. (1992) studied and attempted to quantify opal dissolution in cores located in the Southern Indian Ocean. In our study the importance of the two factors is assessed.

Fig. 3. Opal rain rates (corrected for sediment redistribution but not opal dissolution) as a function of age for the ¢ve cores.

at the base of the South East Indian Ridge (SEIR) with lateral input increasing during glacial periods. This suggests that, although most of the material found on the sea£oor comes initially from surface water production, the sedimentation in the South Indian Basin is mainly controlled by redistribution by deep ocean currents. The pattern of opal rain rates is not very di¡erent after correc-

3.2.3. Reconstruction of dissolution-corrected opal rain rate Bareille (1991) and Pichon et al. (1992) have established a transfer function to quantify the dissolution of diatom silica in Southern Ocean sediments. The relationship between the amount of silica dissolution and changes in diatom species distribution was built by controlled progressive dissolution of biogenic silica in recent sediment samples from box-core tops, each representative of modern diatom species assemblages. In each experiment, the amount of dissolved silica was measured. Factor analysis (Q-mode) identi¢ed four factors accounting for 83% of the variance. The ¢rst three factors are controlled by surface water properties (mostly temperature). The fourth factor (F4) is the only one correlated with loss of silica in the reference samples (R = 0.9). Pichon et al. (1992) then de¢ned an empirical relationship between the F4 value and the percentage of dissolved silica. Because dissolution experiments start with surface sediment that has already suf-

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

150

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

Fig. 4. Time slices of the (a) opal accumulation rates (not corrected for sediment redistribution and dissolution), and (b) opal rain rates from 43‡S to 55‡S. The triangle on the X-axis marks the approximate position of the APF for each period. MD 88773 data are from FrancOois et al. (1997).

fered dissolution of opal, it is postulated that the % dissolution obtained by using the empirical relationship determined by Pichon et al. (1992) is a minimum. However, by comparing opal £uxes in sediment trap (Pondaven et al., 2000a,b) to 230 Thnormalized accumulation rates of opal in the POOZ Dezileau et al. (2003) have shown that the percent of opal dissolution is underestimated by only 14%. The intensity of dissolution of Bareille (1991) and Pichon et al. (1992) was estimated for each sediment level in cores MD 88770 (PFZ), MD 88773 (AZ) and MD 84552 (AZ) and for each marine isotopic stages in cores MD 94102 (SZ), MD 88769 (PFZ) and MD 94104 (PFZ). Both the corrected and uncorrected opal rain rates used in

this study are published in Dezileau et al. (2003) but their signi¢cance was not discussed. 3.2.3.1. Subantarctic Zone (North of 45‡S). For core MD 94102, low opal rain rates ( 6 0.1 g/cm2 kyr) are recorded throughout the investigated period of time (Fig. 5). Estimating opal dissolution by applying the diatom transfer function of Pichon et al. (1992) reveals a di¡erence in opal dissolution from 60% in the last glacial to only 30% in the Holocene. After correcting for dissolution, the dissolution-corrected opal rain rates in the SAZ remain at values lower than 0.2 g/cm2 kyr (Fig. 5). Thus, the low opal £uxes observed during the investigated period of time are apparently due to low productivity of biological opal.

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

151

Fig. 6. Dissolution-corrected opal rain rates for ¢ves cores from the Indian sector of the Southern Ocean, collapsed onto a meridional transect for illustration. Holocene values are presented as circles; MIS 2 values as squares. The white (black) triangle on the X-axis marks the approximate position of the APF during the Holocene (MIS 2). MD 88773 data are from FrancOois et al. (1997).

glacial opal rain rates (0.2^1 g/cm2 kyr) are up to a factor of 3 higher than the Holocene ones (0.07^ 0.2 g/cm2 kyr). Opal dissolution estimates did not reveal a signi¢cant di¡erence between glacial (30^ 45%) and Holocene values (30^58%). The higher dissolution-corrected opal rain rates (up to 1.7 g/ cm2 kyr in core MD 88770) observed during glacial periods indicate a quite signi¢cant increase of production of siliceous organisms compared to the Holocene.

Fig. 5. Dissolution-corrected opal rain rates (corrected for sediment redistribution and dissolution, circles) and opal rain rates (squares) as a function of age for the ¢ve cores.

3.2.3.2. Polar Frontal Zone. For cores MD 94104, MD 88769 and MD 88770, low opal rain rates are recorded throughout the Holocene and are modestly higher during the glacial stages. The

3.2.3.3. Antarctic Zone. For core MD 84552, located on the Kergulen Plateau, high opal rain rates, between 0.7 and 2.6 g/cm2 kyr, were determined for the Holocene and low values (0.4 g/cm2 kyr) for the glacial period (Fig. 5). Estimates of opal dissolution from a diatom transfer function reveal an increase in opal dissolution from 51% in the Holocene to 74% during MIS 2, and 75% during MIS 3. Dissolution-corrected opal rain rates are still generally higher during the Holocene (between 0.9 and 3.7 g/cm2 kyr) than during MIS 2 (0.8^1.3 g/cm2 kyr). Low dissolution-corrected opal £uxes recorded during glacial stages re£ect a decrease in the production of siliceous organisms. For core MD 88773 (not shown on Fig. 5) relatively high opal rain rates (around 1 g/cm2 kyr) are recorded throughout the inves-

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

152

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

Fig. 7. Comparison between dissolution-corrected opal rain rates and organic carbon rain rates during glacial and Holocene periods. Holocene values are presented as circles; MIS 2 values as squares. The white (black) triangle on the X-axis marks the approximate position of the APF during the Holocene (MIS 2). MD 88773 data are from FrancOois et al. (1997).

tigated period of time (FrancOois et al., 1997). Estimates of opal dissolution (Bareille, 1991) did not reveal any clear di¡erence between glacial and Holocene, which remained static at a value around 55% throughout the investigated period of time. Thus, the relatively high values of dissolution-corrected opal rain rates (2 g/cm2 kyr) observed during both the Holocene and MIS 2 and 3 indicate a constant high productivity of opal in this region over time.

4. Discussion 4.1. Paleoceanographic implications In the following we discuss changes in dissolu-

tion-corrected opal rain rates to the sea£oor in the Indian sector of the Southern Ocean from MIS 2 to the Holocene. Glacial conditions in this region can be seen more clearly (Fig. 6) by comparing a latitudinal transect of dissolutioncorrected opal rain rate for the Holocene with the same transect for MIS 2. At sites well to the south of the APF, represented by cores at V53 and 55‡S in Fig. 6, dissolution-corrected opal rain rates during MIS 2 were lower than during the Holocene. These results are consistent with those from other studies in the Atlantic and Paci¢c Ocean (FrancOois et al., 1997; Frank et al., 2000; Anderson et al., 2002; Chase et al., 2003) and suggest that diatom productivity during MIS 2 was lower than during the Holocene at latitudes south of the APF. Lower diatom productivity during MIS 2 south of APF may have been driven either by a reduced supply of Si to surface waters via upwelling, by less e⁄cient use of upwelled Si by diatoms, or by an increase in sea-ice cover. Considering di¡erent hypotheses, recent studies have shown that the rate of upwelling to Southern Ocean surface waters during the LGM was not lower than that existing today (see Anderson et al., 2002). Moreover, silicon stable isotope measurements suggest that the percentage utilization of silicic acid by diatoms in the Southern Ocean during the last glacial period was diminished relative to the Holocene (De La Rocha et al., 1998). Consequently, increased coverage of this area by sea-ice during glacials could explain the concurrent drop in both opal productivity and silicic acid use during MIS 2. This is consistent with an estimate from diatom-based reconstructions that annual sea-ice cover persisted for 4 months (X. Crosta, pers. commun., 2000) during MIS 2 in the southernmost core MD 84552. However, Gersonde and Zielinsky (2000) established that a longer period of ice cover is needed to explain our low rates of opal sedimentation. There are discrepancies between the approaches to reconstruct sea-ice extent used by Crosta et al. (1998) and the one proposed by Gersonde and Zielinsky (2000). Chase et al. (2003) have presented a relationship between the average number of ice free days per year and the 230 Thnormalized opal burial during the Holocene. If we

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

use this relationship, we estimate that annual seaice cover persisted for 6 months. During these 4^6 months, snow covered sea-ice reduced the availability of light in the surface waters at 55‡S (Fig. 6), which was probably the most signi¢cant factor limiting opal productivity in this area. At sites north of the actual position of the APF, represented by cores at 46 and 46.5‡S in Fig. 6, dissolution-corrected opal rain rates during the MIS 2 (between 0.3 and 0.8 g/cm2 kyr) were slightly higher than during the Holocene (between 0.1 and 0.3 g/cm2 kyr). These results are consistent with those from other studies of the Paci¢c Ocean (Anderson et al., 2002; Chase et al., 2003) and suggest that diatom productivity would have been only modestly greater during MIS 2 compared to the Holocene. There is a clear di¡erence in the Atlantic sector where opal rain rates were as high north of the actual position of the APF during the MIS 2 as the maximum rates observed south of the APF during the Holocene ( s 2 g/cm2 kyr; Frank et al., 2000). Higher diatom productivity during MIS 2 north of the actual position of the APF in the Indian sector may have been driven either by an increase supply of Si to surface waters and/or by an increase in Fe. This is discussed below. SSTs, at V+3‡C, calculated using foraminiferal transfer functions (Lemoine, 1998), suggest that cores at V46‡S (particularly core MD 88770) were close to the APF (Figs. 2 and 6). Thus, the increase of opal productivity during MIS 2 compared to the Holocene is probably a consequence of an increase supply of Si to the surface waters in this area due to a northward migration of the APF by approximately 5‡ of latitude. In this case, the APF during MIS 2 stayed as today both as a physical water mass boundary and an ecological boundary. This is in contrast to the interpretations of Moore et al. (2000), Anderson et al. (2002) and Chase et al. (2003) who suggest that the two likely became uncoupled at the LGM. During MIS 2, extensive sea-ice cover limited diatom productivity and silicic uptake south of the APF (Anderson et al., 2002; Chase et al., 2003). Upwelled Si that went unused south of the APF because of ice cover was transported

153

northward, permitting enhanced opal productivity in the region. With Si but not Fe being supplied to waters north of the actual position of the APF during MIS 2, primary producers in this region are limited. Diatoms in the Indian sector have probably received some Fe from the erosion of the basaltic hence Fe-rich Kerguelen Plateau by strong currents associated to the ACC (Dezileau et al., 2000), from upwelled Fe that went unused south of the APF because of ice cover, and also may be from Patagonian dust. As opal productivity was clearly more important in the Atlantic sector than in the Indian sector north of the actual position of the APF during MIS 2, we suggest that Fe £uxes in the Atlantic sector were greater than in the Indian sector. This is consistent with the decreasing Atlantic^Paci¢c gradient dust £ux estimated by Mahowald et al. (1999) and the interbasin gradient of lithogenic £uxes estimated by Chase et al. (2003). Bopp et al. (2003) have assessed the impact of dust deposition rates on marine biota and atmospheric CO2 using a state-of-the-art ocean biogeochemistry model. The model results suggest that the glacial increase in Patagonian dust deposition was most likely insigni¢cant for increasing Fe supply to the Indian sector of the Southern Ocean. Thus, most likely Fe was supplied via upwelling and the erosion of the Kerguelen Plateau by strong currents associated to the ACC. Opal productivity was lower during MIS 2 than during MIS 3 (Fig. 5) although Fe was probably more important during MIS 2 than during MIS 3. This result can be explained by an unfavorable light-mixing regime during MIS 2. Pondaven et al. (2000b) found that present-day high wind conditions prevailing during winter (average 14 m s31 ) would keep algal biomass at low levels, whereas decreased wind speed in summer (7 m s31 ) would allow moderate strati¢cation (mixed layer depth 50^100 m), leading to a moderate increase in biomass and major nutrient utilization. An intensi¢cation of the Westerlies is supported by the increased wind strengths during MIS 2 inferred from the concentrations of dust and sea salt extracted from the Vostok ice core (Petit et al., 1999). An intensi¢cation of the ACC probably due to an increase in westerly wind strengths may

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

154

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

also be deduced from marine sediment studies (Dezileau et al., 2000). Thus, the mixed layer (wind-dependent) was probably deeper during MIS 2 than during MIS 3 inducing an high-iron low biomass case (Blain et al., 2001) during MIS 2. This process could also explain the modest increase and not a high increase of opal productivity during MIS 2 relative to the Holocene. At site MD 94102 (43‡S), located in the northern part of the PFZ, i.e. close to the Subantarctic Front (Fig. 6), dissolution-corrected opal rain rates during MIS 2 were similar to those recorded during the Holocene. Low dissolution-corrected opal rain rates (0.2 g/cm2 kyr) recorded in this region are presumably due to unfavorable SST conditions and low concentration of dissolved silica. Burckle et al. (1987) argued that the northern limit of high productivity by diatoms is controlled by temperature, which e¡ects the metabolic processes of diatoms. Burckle et al. (1987) noticed a general decrease in diatom abundance in the surface waters of the Southern Ocean when SSTs are higher than 8‡C. Fiala and Oriol (1990) found that Antarctic diatoms could not survive in temperatures above 6^9‡C. Reconstructed SSTs (Salvignac, 1998) for core MD 94102 are higher than 8‡C throughout the investigated period of time. These SST values may have prevented Antarctic diatoms from extending north of 43‡S in the Indian sector. The other important factor that needs to be mentioned is the concentration of dissolved silica. At present, the northern part of the PFZ is characterized by a low concentration of dissolved silica. Reconstructed SSTs (Salvignac, 1998) for core MD 94102 clearly show that the APF does not reach this latitude (Fig. 2a) throughout the investigated period of time. Thus, we suggest that the diatom abundance has been limited by a low supply of Si to surface waters via upwelling in this region. The low productivity of siliceous organisms in this northernmost core may therefore be explained by SSTs which have been always above 8‡C and low supply of Si to surface waters via upwelling. Low opal £uxes recorded throughout MIS 2 for this northernmost core con¢rm that the siliceous belt, which has probably been associated with the APF, did not reach this latitude. This is a

very similar situation to the one found by Asmus et al. (1999) in the eastern Atlantic sector of the Southern Ocean. 4.2. Variations in biogenic silica and organic carbon rain rates Here we discuss changes in dissolution-corrected opal rain rates and organic carbon rain rates to the sea£oor in the Indian sector of the Southern Ocean from MIS 2 to the Holocene. Our aim is to determine whether dissolution-corrected opal rain rates represent a reliable proxy of export paleoproductivity or whether there was a glacial decoupling of organic carbon and biogenic opal accumulation (Kumar et al., 1995; Anderson et al., 1998; Bareille et al., 1998). During the Holocene period, maximum biogenic opal and organic carbon deposition has occurred in the AZ of the Southeast Indian sector (Fig. 7). During glacial times, the belt of biogenic opal and organic carbon rain rates shifted northwards. A closer comparison of last glacial and Holocene patterns suggests that there are signi¢cant di¡erences between the two periods. First, glacial biogenic opal rain rates (1.1 g/cm2 kyr) never reach values as high as those observed in the Holocene accumulation belt (1.8 g/cm2 kyr). During glacial times the average dissolution-corrected opal rain rates is equal to 0.94 g/cm2 kyr (n = 5) corresponding to a 20% reduction compared to the Holocene (1.20 g/cm2 kyr). The glacial increase of biogenic silica rain rates in the PFZ^SAZ does not compensate glacial Antarctic decrease south of the present position of the PFZ. Secondly, during glacial times the average organic carbon rain rate is equal to 6.5 mgC/cm2 kyr (n = 5) corresponding to a 25% increase compared to the Holocene (5.3 g/cm2 kyr). Although the spatial coverage of cores for which opal and organic carbon burial rates have been reconstructed remains too limited to attempt a quantitative budget for opal and carbon burial during MIS 2, if our results are representative of the entire Southeastern Indian Basin, then we can suggest that this region is characterized by a reduction in opal rain rates that is not recorded by organic carbon rain rates. One way to reconcile these ob-

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

servations is to invoke a decoupling between opal rain rates and organic carbon £uxes to the sea bed. Kumar et al. (1995), Anderson et al. (1998) and Bareille et al. (1998) have reached a similar conclusion in the Atlantic and Indian sectors of the Southern Ocean. What is the origin of this decoupling ? Either organic carbon was delivered to glacial sediments of the PFZ with reduced amounts of opal, or there is a di¡erential change in the degradation/preservation of the biogenic components, or both. These two scenarios are discussed further below. Firstly, the scenario involving the delivery of organic carbon to glacial sediments of the PFZ together with reduced amounts of opal. There are two possible hypotheses to explain this decoupling. An ecological shift of dominant species assemblages away from diatoms to phytoplankton without silicic shells (coccolithophorids; Anderson et al., 2002) could contribute to the observed decoupling between carbon £ux and opal accumulation in the Indian sector. Moreover, diatom assemblages have lower Si/C ratios during MIS 2 north of the actual position of the APF. Recently, Hutchins and Bruland (1998) have shown that Fe plays a crucial role in the control of the Si/C ratio, and during times of increased Fe inputs diatom assemblages exhibit lower Si/C ratios (production of less silici¢ed organisms). We suggest that the increased Fe fertilization of surface waters during MIS 2 (Martin, 1990; Kumar et al., 1995) may have resulted in decreased Si/C ratios north of the present-day position of the APF. Secondly, the scenario of change in the preservation of organic matter. Decoupling between opal and organic carbon accumulation can be explained by more e⁄cient organic carbon preservation in the PFZ during glacial times. Dezileau et al. (2000) have clearly shown an increase in sediment focusing in the PFZ during glacial times. Consequently, we suggest that the preservation of organic carbon during glacial times has been controlled by intense sediment focusing in the PFZ. Even if these e¡ects on the organic carbon preservation are not quanti¢able, we suggest that this e¡ect may explain a major part of the strong

155

decoupling between opal rain rates and organic carbon £uxes to the sea bed. In conclusion, the origin of the decoupling between C and Si during glacial times in the PFZ can be explained as the result of a combination of factors including: a change in phytoplankton communities, the in£uence of Fe on the Si/C ratio in diatoms and, most likely, a better preservation of organic components in the sediment. During glacial times, in the Indian sector of the Southern Ocean the average biogenic opal rain rate showed a 20% reduction and the average organic carbon rain rate showed a 25% increase compared to the Holocene. In this area, the increase of the average organic carbon rain rate does not necessarily mean an increase in export production of 25%. This result may be due to a better preservation of organic carbon in the sediment. In this case, if we consider only the organic carbon £ux we overestimate export production. On the other hand, the decrease of the average opal rain rate does not necessarily mean a decrease in export production of 20%. Indeed, it is possible that there was a change in the composition of phytoplankton communities or a change in the Si/C ratio during glacial times. In this case if we consider only the opal £ux, we underestimate export production. Finally, in the Indian sector of the Southern Ocean, we conclude that the range in past variability of export production is between 320 and +25% during glacial times compared to the Holocene. This relatively small di¡erence between the two estimates of export productivity suggests there has little change between glacial times and the Holocene in the Indian sector. The export paleoproductivity of the Atlantic sector of the Southern Ocean during MIS 2 has also been found to be similar to today (Frank et al., 2000) while in the Paci¢c sector paleoproductivity was lower than today (Chase et al., 2003). An increase in export paleoproductivity during glacial periods is not supported by the data, implying that bioproductivity variations within the Southern Ocean are unlikely to have contributed to the glacial drawdown of atmospheric CO2 inferred from ice core data.

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

156

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

productivity was similar to Holocene values in the Indian sector of the Southern Ocean.

5. Conclusions Radioisotope and proxy element pro¢les were measured on ¢ve sediment cores forming a transect across the Southern Indian Ocean of the ACC. Records of opal and organic carbon £uxes to the sea£oor were obtained by normalization to 230 Th. During glacial periods, we observe that dissolution-corrected opal rain rates are slightly greater than during the Holocene in the PFZ. This is probably a consequence of the northward migration of the frontal system by approximately 5‡ latitude, a northward transport of Si from the AZ and an increase in Fe via upwelling and the erosion of the Kerguelen Plateau by strong currents associated to the ACC. In the Antarctic zone, glacial biogenic opal rain rates were similar to the Holocene except for the Kerguelen Plateau where rates have decreased. We suggest that the availability of light in surface waters above the Kerguelen Plateau was reduced during the last glacial period by snow covered sea-ice, which would have signi¢cantly limited opal productivity in the area. The results of our study favor an interpretation that overall opal productivity for the Indian sector of the Southern Ocean is slightly lower during glacial periods. As far as export paleoproductivity is concerned, dissolution-corrected opal rain rates were lower during the glacial period by up to 20% while organic carbon rain rates were higher by up to 25%. One way to reconcile these apparently contradictory observations is to invoke a decoupling between opal burial and organic carbon £ux to the sea bed during the last glacial period. This decoupling can be explained by a number of factors, including a change in the composition of phytoplankton communities (other than diatoms), a change in the Si/C ratio in siliceous planktonic organisms due to increased Fe input in the Southern Ocean, and, most importantly, by a better preservation of organic components in the sediment. The results of our study indicate that the evolution of export production ranges between 320 and +25% during glacial times compared to the Holocene in the Indian sector of the Southern Ocean. This small range means that glacial export

Acknowledgements Piston cores were recovered during the ASPARA II (1984), ASPARA IV (1988) and PACIMA (1994) cruises. We are most grateful to the o⁄cers and crew of R/V Marion-Dufresne. These cruises were funded by the Territoires et Terres Australes et Antarctiques FrancOaises. We thank Martin Frank and Roger FrancOois for their constructive comments which signi¢cantly improved the manuscript. L.D. received ¢nancial support from ECOS-SUD. References Anderson, R.F., Bacon, M.P., Brewer, P.G., 1983. Removal of 230 Th and 231 Pa from the open ocean. Earth Planet. Sci. Lett. 62, 7^23. Anderson, R.F., Kumar, N., Mortlock, M.R.A., Froelich, P.N., Kubik, P., Dittrich-Hannen, B., Sutter, M., 1998. Late-Quaternary changes in productivity of the Southern Ocean. J. Mar. Syst. 17, 497^514. Anderson, R.F., Chase, Z., Fleisher, M.Q., Sachs, J., 2002. The Southern Ocean’s biological pump during the Last Glacial Maximum. Deep-Sea Res. II 49, 1909^1938. Asmus, T., Frank, M., Koschmieder, C., Frank, N., Gersonde, R., Kuhn, G., Mangini, A., 1999. Variations of biogenic particle £ux in the southern Atlantic sections of the Subantarctic Frontduring the Late Quaternary: Evidence from sedimentary 231Paex and 230Thex. Mar. Geol. 159, 59^74. Bacon, M.P., 1984. Glacial to Interglacial changes in carbonate and clay sedimentation in the Atlantic ocean estimated from 230 Th measurements. Isot. Geosci. 2, 97^111. Bareille, G., 1991. Flux se¤dimentaires: Pale¤oproductivite¤ pale¤ocirculation de l’oce¤an Austral au cours des 150,000 dernie'res anne¤es. The'se, Univ. de Bordeaux I, 259 pp. Bareille, G., Labracherie, M., Labeyrie, L.D., Pichon, J.J., Turon, J.L., 1991. Biogenic silica accumulation rate during the Holocene in the Southeastern Indian Ocean. Mar. Chem. 35, 537^551. Bareille, G., Labracherie, M., Bertrand, P., Labeyrie, L., Lavaux, G., Dignan, M., 1998. Glacial^Interglacial changes in the accumulation rates of major biogenic components in Southern Indian Ocean sediments. J. Mar. Syst. 17, 527^539. Belkin, I.M., Gordon, A.L., 1996. Southern Ocean fronts from the Greenwich meridian to Tasmania. J. Geophys. Res. 101, 3675^3696. Blain, S., Tre¤guer, P., Belviso, S., Bucciarelli, E., Denis, M., Desabre, S., Fiala, M., Martin-Je¤ze¤quel, V., LeFe'vre, J.,

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

L. Dezileau et al. / Marine Geology 202 (2003) 143^158 Mayzaud, P., Marty, J.-C., Razouls, S., 2001. A biogeochemical study of the island mass e¡ect in the context of the iron hypothesis: Kerguelen Islands, Southern Ocean. Deep-Sea Res. I 48, 163^187. Bopp, L., Kohfeld, K.E., Le Que¤re¤, C., Aumont, O., 2003. Dust impact on marine biota and atmospheric CO2 during glacial periods. Paleoceanography 18, 1046. Broecker, W.S., Maier-Reimer, E., 1992. The in£uence of air and sea exchange on the carbon isotope distribution in the sea. Global Biochem. Cy. 6, 315^320. Burckle, L.H., Jacobs, S.S., McLaughlin, R.B., 1987. Late austral spring diatom distribution between New Zealand and the Ross Ice Shelf, Antarctica: Hydrographic and sediment correlations. Micropaleontology 33, 74^81. Charles, C.D., Froelich, P.N., Zibello, M.A., Mortlock, R.A., 1991. Biogenic opal in southern ocean sediments over the last 450,000 years: Implications from the surface water chemistry and circulation. Paleoceanography 6, 697^728. Chase, Z., Anderson, R.F.A., Fleisher, M.Q.F., Kubik, P.W., 2003. Accumulation of biogenic and lithogenic material in the Paci¢c sector of the Southern Ocean during the past 40,000 years. Deep-Sea Res. II 50, 799^832. Crosta, X., Pichon, J.J., Burckle, L.H., 1998. Application of Modern Analog Technique to marine Antarctic diatoms: Reconstruction of maximum sea-ice extent at the Last Glacial Maximum. Paleoceanography 13, 284^297. Clapperton, C.M., Sugden, D.E., Kaufman, D.S., McCulloch, R.D., 1995. The last glaciation in Central Magellan Strait, southernmost Chile. Quat. Res. 44, 133^148. Deacon, G.E.R., The subtropical convergence. In: Fairbridge, R. (Ed.), The Encyclopedia of Oceanography. New York, pp. 884^885. De La Rocha, C.L., Brzezinski, M.A., DeNiro, M.J., Shemesh, A., 1998. Silicon-isotope composition of diatoms as an indicator of past oceanic change. Nature 395, 680^683. Dezileau, L., Bareille, G., Reyss, J.L., Lemoine, F., 2000. Evidence for strong sediment redistribution by bottom currents along the Southeast Indian Ridge. Deep-Sea Res. I 47, 1899^1936. Dezileau, L., Bareille, G., Reyss, J.L., 2003. 231 Paex ‡/230 Thex ‡ ratio as proxy for past change in opal £uxes in the Southern Indian Ocean. Mar. Chem. 81, 105^117. Fiala, M., Oriol, L., 1990. Light-temperature interactions on the growth of Antarctic diatoms. Polar Biol. 10, 629^636. FrancOois, R., Bacon, M.P., Suman, D.O., 1990. Thorium 230 pro¢ling in deep-sea sediments: High-resolution records of £ux and dissolution of carbonate in the equatorial Atlantic during the last 24,000 years. Paleoceanography 5, 761^787. FrancOois, R., Bacon, M.P., Altabet, M.A., Labeyrie, L.D., 1993. Glacial/interglacial changes in sediment rain rate in the SW Indian sector of Subantarctic waters as recorded by 230 Th, 231 Pa, U, and N15 N. Paleoceanography 8, 611^629. FrancOois, R., Altabet, M.A., Yu, E.-F., Sigman, D.M., Frank, M., Bacon, M.P., Bohrmann, G., Bareille, G., Labeyrie, L.D., 1997. Contribution of southern ocean surface-water strati¢cation to low atmospheric CO2 concentrations during the last glacial period. Nature 389, 929^935.

157

Frank, M., Gersonde, R., Rutgers Van der Loe¡, M., Bohrmann, G., Nurnberg, C.C., Kubik, P.W., Suter, M., Mangini, A., 2000. Similar glacial and interglacial export bioproductivity in the Atlantic sector of the Southern ocean: Multiproxy evidence and implications for glacial atmospheric CO2 . Paleoceanography 15, 642^658. Gersonde, R., Zielinsky, U., 2000. The reconstruction of late Quaternary Antarctic sea-ice distribution 9 the use of diatoms as a proxy for sea-ice. Paleogeogr. Palaeoclimatol. Palaeoecol. 162, 263^286. Henderson, G.M., Heinze, C., Anderson, R.F., Winguth, A.M.E., 1999. Global distribution of the 230 Th £ux to ocean sediments constrained by GCM modelling. Deep-Sea Res. I 46, 1861^1893. Heusser, C.J., 1989. Southern westerlies during the Last Glacial Maximum. Quat. Res. 31, 423^425. Hulton, N., Sugden, D.E., Payne, A., Clapperton, C.M., 1994. Glacier modelling and the climate of Patagonia during the Last Glacial Maximum. Quat. Res. 42, 1^19. Huston, W.H., 1980. The Aghulas Current during the late Pleistocene Analysis of modern faunal analogs. Science 207, 64^66. Hutchins, D.A., Bruland, K.W., 1998. Iron-limited diatom growth and Si:N uptake ratios in a coastal upwelling regime. Nature 393, 561^564. Ku, T.L., Knauss, K.G., Mathieu, G.G., 1977. Uranium in open ocean: Concentration and isotope composition. Deep-Sea Res. 24, 1005^1017. Kumar, N., Anderson, R.F., Mortlock, R.A., Froelich, P.N., Kubik, P., Dittrich-hannen, B., Sutter, M., 1995. Increased biological productivity and export production in the glacial Southern Ocean. Nature 378, 675^680. Labeyrie, L., Labracherie, M., Gorfti, N., Pichon, J.J., Vautravers, M., Arnold, M., Duplessy, J.-C., Paterne, M., Michel, E., Duprat, J., Caralp, M., Turon, J.L., 1996. Hydrographic changes of the Southern Ocean (southeast Indian sector) over the last 230 kyr. Paleoceanography 11, 57^76. Lamy, F., Klump, J., Hebbeln, D., Wefer, G., 2000. Late Quaternary rapid climate change in northern Chile. Terra Nova 12, 8^13. Lamy, F., Hebbeln, D., Ro«hl, U., Wefer, G., 2001. Holocene rainfall variability in southern Chile: A marine record of latitudinal shifts of the Southern Westerlies. Earth Planet. Sci. Lett. 185, 369^382. Lamy, F., Ru«hlemann, C., Hebbeln, D., Wefer, G., 2002. High- and low-latitude climate control on the position of the southern Peru^Chile Current during the Holocene, Paleoceanography, 17(2). Lemoine, F., 1998. Changements de l’hydrologie de surface (tempe¤rature et salinite¤) de l’oce¤an Austral en relation avec les variations de la circulation thermohaline au cours des deux derniers cycles climatiques. Ph.D. Thesis, Universite¤ Paris VI, 169 pp. Mahowald, N., Kohfeld, K., Hansson, M., Harrison, S.P., Prentice, I.C., Schulz, M., Rodhe, H., 1999. Dust sources and deposition during the last glacial maximum and current climate: A comparison of model results with paleodata from

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart

158

L. Dezileau et al. / Marine Geology 202 (2003) 143^158

ice cores and marine sediments. J. Geophys. Res. Atmos. 104, 15895^15916. Markgraf, V., 1989. Reply to C.J. Heusser’s ‘Southern Westerlies during the Last Glacial Maximum’. Quat. Res. 31, 426^432. Martin, J., 1990. Glacial^interglacial CO2 change: The iron hypothesis. Palaeoceanography 5, 1^13. Moore, J.K., Abbott, M.R., Richman, J.G., Nelson, D.M., 2000. The Southern Ocean at the last glacial maximum: A strong sink for atmospheric carbon dioxide. Global Biogeochem. Cy. 14, 455^475. Moreno, P.I., 1997. Vegetation and climate near Lago Llaquihue in the Chilean lake district between 20200 and 9500 14 C yr BP. J. Quat. Sci. 12, 485^500. Moreno, P.I., Lowell, T.V., Jacobson, G.L., Denton, G.H., 1999. Abrupt vegetation and climate changes during the last glacial maximum and last termination in the Chilen lake district: A case study from canal de la Puntilla (41 S). Geogr. Ann. 81, 285^311. Mortlock, R.A., Froelich, P.N., 1989. A simple method for the rapid determination of biogenic opal in pelagic marine sediments. Deep-Sea Res. I 36, 1415^1426. Mortlock, R.A., Charles, C.D., Froelich, P.N., Zibello, M.A., Saltzman, J., Hays, J.D., Burckle, L.H., 1991. Evidence for lower productivity in the Antarctic Ocean during the last glaciation. Nature 351, 220^223. Park, Y.H., Gamberoni, L., Charriaud, E., 1991. Frontal structure and transport of the Antarctic Circumpolar Current in the south Indian Ocean sector, 40-803E. Mar. Chem. 35, 45^62. Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola, J.M., Basile, I., Bender, M., Chappellaz, J., Davis, M., Delaygue, G., Delmotte, M., Kotlyakov, V.M., Legrand, M., Lipenkov, V.Y., Lorius, C., Pepin, L., Ritz, C., Saltzman, E., Stievenard, M., 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399, 429^436. Pichon, J.J., Bareille, G., Labracherie, M., Labeyrie, L.D., Baudrimont, A., Turon, J.L., 1992. Quanti¢cation of the biogenic silica dissolution in southern ocean sediments. Quat. Res. 37, 361^378. Pondaven, P., Ragueneau, O., Tre¤guer, P., Hauvespre, A.,

Dezileau, L., 2000a. Resolving the opal paradox in the Southern Ocean. Nature 405, 168^172. Pondaven, P., Ruiz-Pino, D., Fravalo, C., Tre¤guer, P., Jeandel, C., 2000b. Interannual variability of Si and N cycles at the times-series station KERFIX between 1990 and 1995 ^ a 1-D modelling study. Deep-Sea Res. I 47, 223^257. Prell, W.L., 1985. The Stability of the Low Latitude Sea Surface Temperatures: An Evaluation of the CLIMAP Reconstruction with Emphasis on the Positive SST Anomalies. Departement of Energy, Washington, DC. Rutgers van der Loe¡, M.M., Berger, G.W., 1993. Scavenging of 230 Th and 231 Pa near the Antarctic Polar Front in the South Atlantic. Deep-Sea Res. 40, 339^357. Salvignac, M.E., 1998. Reconstitution des tempe¤ratures de surface dans l’Oce¤an Austral Indien. The'se, Universite¤ de Bordeaux I, 214 pp. Suman, D.O., Bacon, M.P., 1989. Variations in Holocene sedimentation in the North American basin determined from 230 Th measurements. Deep-Sea Res. 36, 869^878. Valdes, P., 2000. South American paleoclimate model simulations: How reliable are the models? J. Quat. Sci. 15, 357^ 368. Van Bennekom, A.J., Berger, G.W., Van der Gaast, S.J., De Vries, R.T.P., 1988. Primary productivity and the silica cycle in the Southern Ocean (Atlantic Sector). Palaeogeogr. Palaeoclimatol. Palaeoecol. 67, 19^30. Walter, H.J., Rutgers van der Loe¡, M.M., Hoeltzen, H., 1997. Enhanced scavenging of 231 Pa relative to 230 Th in the South Atlantic south of the Polar Front: Implications for the use of the 231 Pa/230 Th ratio as a paleoproductivity proxy. Earth Planet. Sci. Lett. 149, 85^100. Whitworth, T., III, Nowlin, W.D., Jr., 1987. Water masses and currents of the Southern Ocean at the Greenwich Meridian. J. Geophys. Res. 92, 6462^6476. Wyroll, K.H., Dong, B.W., Valdes, P., 2000. On the position of southern Hemisphere westerlies at the Last Glacial Maximum: An outline of AGCM simulation results and evolution of their implications. Quat. Sci. Rev. 19, 881^898. Yu, E.F., Francois, R., Bacon, M.P., Fleer, A.P., 2001. Fluxes of 230 Th and 231 Pa to the deep sea: Implications for the interpretation of excess 230 Th and 231 Pa/230 Th pro¢les in sediments. Earth Planet. Sci. Lett. 191, 219^230.

MARGO 3398 5-11-03 Cyaan Magenta Geel Zwart