Holocene paleoceanography of the Bay of Biscay: Evidence for west-east linkages in the North Atlantic based on dinocyst data

Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Holocene paleoceanography of the Bay of Biscay: Evidence for west-east linkages in the North Atlantic based on dinocyst data Jena Zumaque a,⁎, Frédérique Eynaud b, Anne de Vernal a a b

GEOTOP, Université du Québec à Montréal, CP 8888, Montréal, Québec H3C 3P8, Canada Université de Bordeaux, EPOC, UMR 5805, F-33400 Talence, France

a r t i c l e

i n f o

Article history: Received 21 July 2016 Received in revised form 27 November 2016 Accepted 21 December 2016 Available online 23 December 2016 Keywords: Holocene Bay of Biscay Dinocysts Sea-surface conditions Climatic optimum Episodes of cooling

a b s t r a c t Paleoceanographical changes during the Holocene were reconstructed from the study of core MD95-2002 situated in the northern Bay of Biscay, which is marked by the direct influence of the northeastern return branch of the North Atlantic Drift. Palynological data, sea-surface condition estimates based on dinocyst assemblages and stable isotope measurements in planktic and benthic foraminifera reveal a strong influence of freshwater/meltwaters from both the proximal European sources and the more distal Laurentide Ice Sheet, which experienced delayed deglaciation. The data also indicate the setting of a climate optimum between 7 and 5.5 ka followed by a cooling trend, which is consistent with insolation changes and other regional records of climate changes. Superimposed on the long term trends, the reconstructions of sea-surface conditions evidence large amplitude changes at centennial to millennial time-scales, with seven episodes of cooling and low salinity since 11 ka that generally match episodes of dense sea-ice cover in the Labrador Sea. The west to east transfer of the seaice and/or meltwater signal across the North Atlantic evidenced from core MD95-2002 points to strong linkages between western and eastern North Atlantic, probably in relation to the North Atlantic Oscillation (NAO) mode of variability. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Oxygen isotope records from Greenland ice cores have highlighted high frequency climate variations during the last glacial period (Dansgaard et al., 1993) in contrast to the Holocene, which has long been considered relatively stable from a climatic point of view. Nevertheless, in the context of the current global warming, the scientific community has intensified the study of climate changes during the present interglacial in order to document the natural variability of climate under “warm” regime and to better assess on the actual impact of the anthropogenic forcing. In particular, special attention has been paid to the North Atlantic and adjacent lands as they represent key areas with regard to poleward heat transport through the North Atlantic Drift (NAD) as a major component of the Atlantic Meridional Overturning Circulation (AMOC) (cf. Carton and Hakkinen, 2011). Bond et al. (1997) were the first to highlight climate variability in the Holocene from marine records of the northern North Atlantic (cf. Andersen et al., 2004a). Since then, millennial-scale variability of climate parameters has been detected in Holocene records based on various tracers, including coccoliths, foraminifers, and dinocysts (e.g. ⁎ Corresponding author. E-mail address: [email protected] (J. Zumaque).

http://dx.doi.org/10.1016/j.palaeo.2016.12.031 0031-0182/© 2016 Elsevier B.V. All rights reserved.

Giraudeau et al., 2010; Bond et al., 2001; Hall et al., 2004; Solignac et al., 2006, 2008). However, some inconsistencies in the timing, periodicity and amplitude of the variations, as well as in the signature of early Holocene climatic optimum and long-term trend, have been noticed by several authors (e.g., Eynaud et al., 2004; Solignac et al., 2006; de Vernal and Hillaire-Marcel, 2006). The amplitude and the timing of the climatic optimum referred to as “Hypsithermal” differ from one site to another, but it seems well recorded and consistent along the main path of the NAD (e.g., de Vernal and Hillaire-Marcel, 2006). The heterogeneity in the Holocene records reflects complex dynamics of ocean and climate in the North Atlantic during the Holocene. In this context, the goal of the present study is to further document the climate variability of the present interglacial from the reconstruction of hydrographic conditions in the northern Bay of Biscay. For this purpose, we analysed core MD95-2002 retrieved from a site located under the direct influence of the northeastern return branch of the North Atlantic Drift (NAD) (Frew et al., 2000). The coring site is characterized by high sedimentation rates, especially during Termination I, during which they reach up to 50 cm/kyrs (Zaragosi et al., 2001; Auffret et al., 2002; Zaragosi et al., 2006). Here, we reconstructed seasurface conditions based on the analysis of dinocyst assemblages (e.g., Eynaud, 1999; Rochon et al., 1999; Penaud et al., 2009; de Vernal et al., 2013). Isotopic measurements of planktic foraminifera (Globigerina

404

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

bulloides and Globorotalia truncatulinoides) have been made to provide complementary information on the properties of the sub-surface water masses.

2. Environmental setting Core MD95-2002 (47°27′N; 08°32′W) was retrieved during the IMAGES 101 expedition of RV Marion Dufresne in May–July 1995 (Bassinot and Labeyrie, 1996; Auffret et al., 2002). It was collected at 2174 m water depth on the Meriadzec Terrace, in the northern part of the Bay of Biscay 600 m above the Biscay abyssal plain (Eynaud et al., 2007) (Fig. 1). The deep relief in the Bay of Biscay constitutes the seaward prolongation of the Berthois Spur that forms a morphological boundary along the Celtic and Armorican margins (Bourillet et al., 2006) that splits the sedimentary load from the shelf towards the Celtic Fan westward and the Armorican Fan eastward. These fans are two of the three major deep-sea fans of the French Atlantic Margin (Zaragosi et al., 2000). During low-stands of eustatic sea-level, they were mainly fed by discharges from the “Fleuve Manche paleoriver” that drained a large part of northwestern Europe (Eynaud et al., 2007). This important fluvial paleosystem, which extended from the southern North Sea to the Bay of Biscay, comprises the English Chanel, a portion of the continental “canyon-dominated” slope, and the two deep-sea turbidite systems mentioned above (i.e. Celtic and Armorican Fans) (cf. Toucanne et al., 2011). Presently, surface waters are under the direct influence of the warm NAD, which contribute to the North Atlantic gyre (e.g., Sutton and Allen, 1997). At the study site sea-surface temperature (SST) and salinity (SSS) are 11.7 ± 0.6 °C and 35.54 ± 0.05 and 17.5 ± 1.0 °C and 35.58 ± 0.10 in winter and summer respectively (World Ocean Atlas, 2001; Conkright et al., 2002). The Slope Current (SC) also carries warm and salty waters

of the Eastern North Atlantic Water (ENAW) that occupies the water column down to 800 m (e.g., Lazure et al., 2008). Below the SC, from about 800 to 1300 m, a branch of the warm but very salty (35.7) Mediterranean Overflow Water (MOW) overlies the Labrador Sea Waters (LSW), which is characterized by salinity ranging from 35 to 35.5 (Cossa et al., 2004). Diluted LSW salinity signal in the eastern part of the Bay of Biscay is episodically induced by a diapycnal mixing favored by the proximity of the continental slope (van Aken, 2000). 3. Material and methods 3.1. Stratigraphy and chronology of the core Core MD95-2002 consists of 30 m of hemipelagic clays. Detailed Xray of the core shows no lamination, nor evidence of turbidity currents or erosion within the Holocene section (Auffret et al., 2002). The core was the subject of many studies that established a robust chronology spanning the last 30,000 years (Zaragosi et al., 2001, 2006; Eynaud et al., 2007, 2012). The late Pleistocene chronostratigraphical framework is based on 20 AMS 14C dates from monospecific Neogloboquadrina pachyderma left coiled (Npl) or Globigerina bulloides (Gb) populations. The stratigraphy is also constrained from the δ18O record and planktonic foraminifer assemblages (Zaragosi et al., 2000, 2001). For the chronostratigraphy of the Holocene section, we have combined the 2 AMS 14C dates published by Zaragosi et al. (2006) and 4 additional AMS 14C dates obtained from Globigerina bulloides samples (Table 1). The age vs. depth relationship was established from the 14C ages using the CLAM software (Blaauw, 2010; http://chrono.qub.ac.uk/blaauw/ clam.html), which uses calibrations similar to that of CALIB 7.0.2. The calibration was made from Marine13 (cf. Reimer et al., 2013) taking into account a marine reservoir effect of 405 years. No additional

Fig. 1. Location of the study core MD95-2002 (large square: 47°27′N; 08°32′W; 2174 water depth) and other cores to which we refer in the text (small circles: LO09-14, Andersen et al., 2004a; RAPID-12-1K – Thornalley et al., 2009; MD95-2015 – Eynaud et al., 2004; HM03-133-25- Solignac et al., 2008; HU90-013-013P – Solignac et al., 2004; KS10b – Mojtahid et al., 2013; MD95-2043 – Fletcher et al., 2012). The main surface currents are indicated as follows: Irminger Current (IC), North Atlantic Drift (NAD), Norwegian Atlantic Current (NwAC), Slope Current (SC), East Greenland Current (EGC). Dashed black lines represent the Iceland Scotland Overflow Water (ISOW).

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

405

Table 1 Radiocarbon ages and sedimentation rates of the Holocene section of the core MD95-2002. Depth (cm)

Material dated

Laboratory number

Corrected 14C ages in yr BP

Calendar Age (cal. yr BP)

Reference

0–1 34–35 72–73 104–105 140–141 240–241

G. bulloides G. bulloides G. bulloides G. bulloides G. bulloides N. pachyderma l.

LSCE – 99360 OS-98831 OS-98831 OS-98831 LSCE – 99361 LSCE-99362

1660 ± 70 3830 ± 25 6740 ± 35 8510 ± 40 9080 ± 90 10,790 ± 100

1199 (2σ = −160; +129) 3777 (2σ = −95; +87) 7268 (2σ = −96; +91) 9134 (2σ = −129; +125) 9823 (2σ = −268; +288) 12,248 (2σ = −171; +192)

Zaragosi et al., 2001 This paper This paper This paper Zaragosi et al., 2001 Zaragosi et al., 2001

correction was applied. The age model was constructed by interpolation between tie points using smooth spline relationship. Calculated sedimentation rates vary between 11.4 cm/kyrs in the middle-late Holocene and up to 50 cm/kyrs during the early Holocene (Table 1; Fig. 2).

3.2. Micropaleontological analyses In order to document paleoceanographical changes throughout the Holocene, micropaleontological analyses were performed down to 200 cm on 1 cm thick samples collected at intervals of 2 cm. This sampling aimed to obtain a higher time resolution than that of Eynaud (1999), which was done at 10 cm interval. The 79 samples were first wet sieved to separate fractions using 106 μm and 10 μm mesh sieves. The fraction N106 μm was used for foraminiferal picking whereas the 10–106 μm fraction was used for palynological preparations. Palynological results were combined with those of Eynaud (1999).

3.2.1. Dinoflagellate cysts (dinocysts) Palynological preparation followed the procedure described by de Vernal et al. (1999) and Eynaud (1999). The dinocyst determination was done with a Leica Microscope at a ×400 magnification. About 300 dinocysts were identified and counted in each sample. Species identification followed Rochon et al. (1999). Relative abundances were calculated relative to the total sum of dinocysts and absolute abundances (concentration cysts/cm3) were obtained from the marker grain method, which yields results with a reproducibility of about ±10% (de Vernal et al., 1996; Mertens et al., 2009). In addition to dinocysts, all other palynomorphs were counted. They notably include pollen, spores, Pediastrum algae, and reworked pre-Quaternary palynomorphs. Along the western European margin, both Pediastrum algae and the ratio Rd/ Md [reworked dinocysts/modern dinocysts] constitute robust proxies of terrigenous advection (e.g. Zaragosi et al., 2001; Eynaud et al., 2007; Penaud et al., 2009). Here we only present summary data as the detailed counts can be found in Zumaque (2014). 3.2.2. Quantitative reconstruction of past sea-surface conditions To reconstruct sea-surface parameters, the modern analogue technique (MAT) has been applied to the dinocyst assemblages (relative abundances) using a script developed by Guiot and Brewer (www. cerege.fr/IMG/pdf/formationR08.pdf) for the R software (“R” version 2.7.0, R Development Core Team, 2008). This approach consists in a statistical assessment of the distance between the fossil and modern assemblages. Hydrographic parameters were estimated from an average weighted inversely to the distance or dissimilarity of the 5 best analogues (e.g. Kucera et al., 2005; Guiot and de Vernal, 2007). The modern assemblages are compiled in a database developed from the analyses of surface sediment samples. The database used here contains 1492 modern samples from North Atlantic, Arctic and North Pacific oceans and their adjacent seas. Information on dinocyst assemblages and oceanic data are available on the GEOTOP http://www.geotop.ca/website (cf. also de Vernal et al., 2013). The reconstructed parameters include sea-surface temperature (SST) and sea-surface salinity (SSS) of the three warmest (summer) and the three coldest months (winter). Uncertainties of reconstructions were evaluated by testing the predictive ability of the approach based a subset of modern dinocyst data (e.g. Guiot and de Vernal, 2007). The difference between the observed and reconstructed values led to calculate the errors of prediction, which are ±1.2 °C and ±1.7 °C for winter and summer SST, and ±2.0 and ±2.3 for winter and summer SSS. The high values of SSS errors are mainly due to the sites located in low salinity domains such as the Arctic or river mouths which are characterized by high variability with regard to salinity and where the actual hydrographic measurements are rare and not always accurate. In core MD95-2002, the variations of salinity during the Holocene are mostly above 32. In this domain, the error of prediction is ± 0.47 and ± 0.41 for winter and summer SSSs respectively.

Fig. 2. Sedimentation rates (red line) and age vs. depth relationship (black line) in the upper 250 cm of core MD95-2002 (cf. Table 1 for 14C ages). Grey lines represent the interval of confidence of the age model. Blue and red dots represent minimum and maximum calibrated dates for each 14C age. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.2.3. Stable isotope Oxygen and carbon isotope analyses have been performed on two planktonic foraminifer species and one benthic taxon. Specimens of the epipelagic taxon Globigerina bulloides (G. bulloides) have been retrieved from the 150–250 μm fraction and those of the mesopelagic

406

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

taxon Globorotalia truncatulinoides (G. truncatulinoides) and the benthic taxon Uvigerina peregrina (U. peregrina) have been retrieved from the N250 μm fraction. The benthic shells have been roasted at 250 °C in a vacuum for 2 h. Samples have been then analysed using a Micromass IsoprimeTM mass spectrometer coupled to a MultiCarbTM preparation system at GEOTOP-UQAM. Isotopic ratios are expressed in ‰ versus VPDB (Vienna Pee Dee Belemnitella), defined with regard to NBS-19 (Coplen, 1988). The analytical uncertainty (1σ) of the laboratory is of ± 0.05‰ for both carbon and oxygen isotope compositions. Due to methodological differences and a lack of intercalibration, we did not used previous data obtained on G. bulloides and published in Auffret et al. (2002). 4. Results and interpretations 4.1. Dinocysts assemblages Dinocysts are abundant throughout the sequence, with concentrations of the order of 103–104 cysts/cm3 (Fig. 3). Prior to 5.6 ka, mean values of 10,000 dinocysts/cm3 are recorded, with maximum concentration ranging up to 31,000 dinocysts/cm3, which indicates cyst fluxes of up to 744 cysts.cm−2.yr−1 and suggests very high productivity (Fig. 3). After 5.6 ka, the concentrations decrease to mean values of 4000 dinocysts/cm3 indicating fluxes of the order 48 cysts/cm2·yr. In addition to dinocysts, abundant reworked pre-Quaternary palynomorphs are recorded, especially prior to 9.5 ka, which indicates important terrigenous input related to the deglaciation of the western European margins (e.g. Zaragosi et al., 2001; Eynaud et al., 2007; Penaud et al., 2009). Thirty dinocyst taxa were identified (cf. Eynaud, 1999; Zumaque, 2014). Among those, eight taxa dominate the assemblages: Spiniferites

mirabilis/hyperacanthus, Operculodinium centrocarpum, Bitectatodinium tepikiense, Brigantedinium spp., Nematosphaeropsis labyrinthus, Selenopemphix nephroides, Spiniferites ramosus and Impagidinium aculeatum (Fig. 3). The relative abundance of taxa allowed us to distinguish two main ecostratigraphic zones. Zone 1 covers the lowest part of the record up to 9.9 ka. It is characterized by the dominance of O. centrocarpum, B. tepikiense, and Brigantedinium spp., with significant occurrences of Pentapharsodinium dalei. This assemblage is typical of temperate continental margins characterized by low surface salinity and stratified upper water masses such as the Gulf of St. Lawrence (cf. Rochon et al., 1999). There is a progressive decrease in the above-mentioned taxa relative to S. mirabilis in the upper part of zone 1, which suggest a warming in surface waters (cf. Rochon et al., 1999). Zone 2, from 9.9 to the top of the sequence is characterized is characterized by the co-dominance of S. mirabilis and O. centrocarpum and a high number of accompanying species. The overall assemblages suggest temperate conditions, but variations in the occurrence of accompanying taxa reflect changes in the properties of surface waters. From 9.9 to 3.5 ka, the most common secondary taxa include B. tepikiense, Brigantedinium spp., N. labyrinthus and S. ramosus in addition to Selenompemphix nephroides, which records its maximum percentages. The significant occurrence of Protoperidiniales taxa such as S. nephroides indicates neritic conditions and suggests relatively low salinity (cf. Rochon et al., 1999). From 3.5 to 2.1 ka, maximum occurrence of the thermophilic and oligotrophic taxa Impagidinium aculeatum (e.g. Radi and de Vernal, 2008) with minimum occurrence of Protoperidiniales suggest warm and saline conditions. Finally, after 2.1 ka, the recurrence of P. dalei and Protoperidiniales may relate to a slight cooling and lower salinity.

Fig. 3. Diagram of dinocyst assemblages vs. age in core MD95-2002. The relative abundance of the main taxa is expressed in percentages. The concentration of dinocysts and Pediastrum are expressed in nb per cm3. The dashed line delimitates the ecostratigraphic zones discussed in the text.

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

4.2. Reconstruction of sea-surface conditions The reconstruction of sea-surface conditions from the application of MAT to dinocyst assemblages provide reliable results inasmuch as the assemblages are characterized by high concentrations, statistically reliable counts (N300 specimens per sample) and close analogues in the reference database. All of the analogues are from the North Atlantic and adjacent seas. Results are reported in Fig. 4. The sequence is marked by large amplitude variations of each parameter. The reconstructed SSSs vary from 31 and 36. They are lower than the modern mean except around 4 and 2.5 ka. Seasonal SSTs also show strong variations, from 0 to 12 °C and from 14 to 24 °C in winter and summer respectively. The winter to summer contrasts of temperature are larger than modern for most of the study interval. This together with the generally lower than present salinity suggest a strong neritic influence and/or important dilution with freshwater during most of the Holocene in addition to relatively strong stratification in the upper water layer. The record of SSS and winter SST shows a general trend of increasing salinity during the early-mid Holocene with particularly low values prior to 10.1 ka and optimum values reached between 4.3 and 2 ka. In contrast, summer SSTs varied around modern values through the Holocene with optimum conditions, significantly warmer than at present, being recorded between 7.5 and 5.2 ka before a general cooling trend. 4.3. Stable isotope composition of foraminifera The oxygen isotope composition of G. bulloides and G. truncatulinoides ranges between 0.29‰ and 1.84‰, and 1.2‰ and 2.28‰ respectively

407

whereas δ18O values of U. peregrina vary from 2.93‰ to 3.69‰ (Fig. 5). G. bulloides is an epipelagic taxon usually living in the 100 upper meters of the column water, and more specifically at about 50–70 m which corresponds to the bottom of the mixed layer (Farmer et al., 2008). The deep dwelling mesopelagic foraminifer G. truncatulinoides lives between 200 and 600 m in the water column and prefers colder temperatures although it is often characteristic of subtropical gyres (Schiebel and Hemleben, 2005). Hence, the difference of ~ 1‰ in δ values of G. bulloides and G. truncatulinoides is related to their respective habitat and would imply a ΔT ~ 4 °C (Frew et al., 2000) and/or different δ18Owater –values between surface and sub-surface waters. The G. bulloides δ18O curve shows some variations, with peaks of high values N1‰ recorded prior to 10.4 ka and at about 7.9 ka, and a slight decrease until about 6.5 ka, which might indicate an increase of temperature and/or a decrease of salinity, both implying a density decrease. G. truncatulinoides presents a δ18O record more noisy than that of G. bulloides with variations from 0.5 to 1‰ vs. VPDB. Although the δ18O record of the two species differs, similar broad trends are recorded showing a decrease in the lower part of the core switching to an increasing trend after about 5.5 ka. The U. peregrina δ18O record, also highly variable, exhibits two peaks of values N3.5‰ at about 4.7 and 3.8 ka. The carbon isotope composition of G. bulloides and G. truncatulinoides ranges between −1.42‰ and −0.4‰, and 0.28‰ and 1.44‰, respectively. Such a difference no doubt reflects distinct depth preferences of the two taxa, as G. bulloides is epipalegic and G. truncatulinoides meso-bathypelagic (e.g., Kucera et al., 2005), which might explain different conditions of calcite precipitation (e.g., Ravelo and Hillaire-Marcel, 2007). Indeed, G. bulloides occupies the maxima

Fig. 4. Reconstruction of sea-surface conditions in core MD95-2002 based on the modern analogue technique applied to the dinocyst assemblages. Red and blue colors are used for summer and winter conditions respectively. Bold curves represent smoothed record (three-points running mean). The vertical bands correspond to the modern temperature and salinity (mean and one standard deviation) as compiled within a radius of 30 nautical miles around the study site from the World Ocean Atlas (2001; Conkright et al., 2002). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

408

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

Fig. 5. Isotopic composition (δ18O and δ13C) of planktonic and benthic foraminifera in core MD95-2002 as expressed in ‰ vs. VPDB.

Fig. 6. Summarized information from MD95-2002 about the late glacial to interglacial transition and Laurentide ice (cf. Dyke et al., 2003). Blue bars represent episodes of cooling. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

productivity zone where oxidation of organic matter generates an isotope depletion whereas G. truncatulinoides grows up under the pycnocline. A difference of ~ 1‰ between G. bulloides δ13C-values and those of mesopelagic taxa such as Neogloboquadrina pachyderma levogyre (Npl) is frequently observed in North Atlantic (e.g., de Vernal and Hillaire-Marcel, 2006). Following Duplessy et al. (1984), a correction of +0.9‰ has been applied to the δ13C-values of U. peregrina that vary between 0.65‰ and 1.38‰. A similar carbon isotope composition of G. truncatulinoides and U. peregrina indicates a ventilation of sub-surface water masses down to the core site depth (2174 m), which is compatible with δ13C profiles of dissolved inorganic carbon (DIC) in the middle latitudes of the North Atlantic (cf. Kroopnick, 1985). The δ13C curve of G. bulloides shows values that discretely rise throughout Holocene, suggesting an increase of the productivity (Berger, 1978). A slight increase is also visible in the δ13C of both G. truncatulinoides and U. peregrina between 8 and 3 ka, which may reflect an increase in the ventilation of deep water masses during Holocene.

5. Discussion 5.1. Transition to full interglacial conditions: The proximal and distal record The multiproxy dataset compiled in Fig. 6 highlights two distinct phases in the transition to full interglacial conditions at the location of MD99-2002 core in the Bay of Biscay. The first phase, from the base of our record until about 9.9 ka, took place just after the Younger Dryas, when harsh conditions prevailed in North Atlantic (Broecker et al., 1988). It is marked by very low and variable SSS (Figs. 4 and 6), which suggests dilution with freshwater. This interval is also characterized by particularly high sedimentation rates reaching about 50 cm/kyrs, high concentrations of the freshwater algae Pediastrum and high Rd/Md ratios (Figs. 3 and 6). Hence, a proximal source of freshwater inputs from the adjacent European continent is likely as also seen earlier in the core during the LGM/Heinrich event 1 transition (e.g. Eynaud et al., 2012). Solomina et al. (2015) noted large retreat of Central European glaciers after the Younger Dryas and until 10.5 ka, thus leading to high meltwater discharges. Ivy-Ochs (2015) also pointed out that the glacial retreat was punctuated by cold phases marked by Alpine glaciers readvances (see also Ivy-Ochs et al., 2006). The maxima observed in the G. bulloides δ18O-curve (Figs. 5 and 6) before 10.6 ka might be the expression of these cooling pulses. Interestingly, the SSTs also show high variation although they are not perfectly synchronised with the δ18O-records. Furthermore, whereas the winter SST values remain low throughout the interval, the summer SSTs reach values higher than the modern ones (Figs. 4 and 6). Such strong seasonal contrast of temperature concomitant with low SSS likely relates to low thermal inertia in the mixed layer due to strong stratification. These conditions were favorable for high summer SST in response to the high summer insolation (Berger, 1978), which also led to warming of adjacent lands (cf. Naughton et al., 2007). The second phase took place after 9.9 ka. Sedimentation rates decreased sharply until they reach their modern values of about 12 cm/kyrs at 7–6 ka (Fig. 6). In the same time interval, SSS rose progressively from 32 to 34.5. Because Pediastrum concentrations and Rd./Md ratios are low, the SSSs might evidence a more distal influence of meltwater, probably in relation with the retreat of the Laurentide Ice Sheet (LIS) until its final collapse around 7–6 ka (e.g., Dyke, 2004; Kaplan and Wolfe, 2006; de Vernal and Hillaire-Marcel, 2006). The winter SSTs estimates and the δ18O in G. bulloides also point to warming during the early Holocene until about 6.5 ka, when the final collapse of the LIS and reduced freshwater discharge enabled the formation of a convection cell in the Labrador Sea (e.g., Dyke, 2004; Kaplan and Wolfe, 2006; de Vernal and Hillaire-Marcel, 2006) thus contributing to the modern AMOC (e.g., Hoogakker et al., 2011).

409

5.2. Long term climate trends at regional scale The climate of the Holocene has been divided into three phases as summarized by Nesje and Dahl (1993). The Preboreal, spanning from 11.6 to 9 ka corresponds to milder climate conditions than those which prevailed during Younger Dryas (e.g., Broecker et al., 1988; Zaragosi et al., 2001). From 9 to 5.7 ka, generally high temperatures are associated with an early-mid Holocene thermal optimum. Finally, a phase of cooling began after 5.7 ka and culminated during the neoglacials of the last millennia. Various studies (e.g., Marchal et al., 2002; Andersen et al., 2004a, 2004b) have illustrated that the Holocene cooling started early, concomitantly with the decrease of summer insolation in the Northern Hemisphere, after 11 ka (Berger and Loutre, 1991). However, other studies illustrate a strong regionalism that indicates the effects of parameters other than the insolation and which include those governing the ocean circulation and the stratification of water masses (e.g., Marchal et al., 2002; Andersen et al., 2004a, 2004b; Solignac et al., 2004; de Vernal and Hillaire-Marcel, 2006). The identification of the climate optimum from core MD95-2002 data is not unequivocal because of the large amplitude SST variations recorded throughout the lower half of the section (Fig. 4). The advection of fresh waters related to the collapse of the LIS until 7–6 ka (e.g., Dyke, 2004; Kaplan and Wolfe, 2006; de Vernal and Hillaire-Marcel, 2006) and decoupling between winter and summer temperatures in response to insolation changes (see Section 5.1) may have hidden the signal of the climate optimum at MD95-2002 site. Nevertheless, a thermal optimum in both winter and summer seems to be reached between 7 and 5.5 ka (Fig. 7). After 5.5 ka, a general cooling trend can be depicted in summer, whereas despite important variations, winter temperatures show relatively high values until about 2.2 ka. Such trends in core MD95–2002 are actually compatible with insolation changes and the classical subdivisions of the Holocene. The thermal optimum recorded in core MD95-2002 seems consistent with other regional records. Based on foraminiferal assemblages, Mojtahid et al. (2013) identified a climatic optimum between 7 and 5.5 ka after an interval marked by strong variability in core KS10b located in southeastern Bay of Biscay (Fig. 2). In core LO09-14, located on Reykjanes Ridge, Andersen et al. (2004a) also detected a thermal optimum that extends from 7.5 to 5 ka based on diatom assemblages. Nearby, in core MD95–2015 from the Gardar Drift, Eynaud et al. (2004) reconstructed a climatic optimum between 9 and 5.7 ka based on dinocyst data. Various authors have however mentioned heterogeneity in the expression of the Holocene thermal optimum and long-term climatic trends, with regard to the amplitude as well as the timing of the events depending upon the location of the sites in North Atlantic (e.g., Solignac et al., 2004, 2006; Andersen et al., 2004b; de Vernal and Hillaire-Marcel, 2006; Giraudeau et al., 2010). Kaplan and Wolfe (2006) have produced a synthetic work on the timing of the climate optimum in North Atlantic. However, the weakness of this compilation consists in the heterogeneity of the data sources that come from different proxies whose response is not necessarily uniform. In order to avoid biases linked to different methodological approaches, de Vernal and Hillaire-Marcel (2006) have compiled SST reconstructions based on the application of MAT to dinocyst assemblages from several cores of the northern North Atlantic. The cores located on the main axis (southwest/northeast) of NAD present the strongest positive anomalies during early Holocene whereas those under the influence of the East Greenland Current (EGC) record negative anomalies. This division in east to west gradients has been highlighted by several authors (e.g., Eynaud et al., 2004; Andersen et al., 2004a; Solignac et al., 2006, 2008; Giraudeau et al., 2010) and might be the result of a most contrasted balance between a stronger EGC and a NAD reinforced on its main path during early to mid-Holocene. This pattern has already been discussed by Thornalley et al. (2009) and Morley and Rosenthal (2014) who invoked strong Subpolar Gyre (SPG) reinforced on its east-west oriented mode. These SPG-

410

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

Fig. 7. (a) Winter and summer SSTs estimated from dinocysts in core MD95-2002 and (b) insolation at 47°N in January and July from Laskar et al. (2004); (c) SSS estimated from dinocysts in core MD95-2002; (d) sea-ice cover for the HU-090-013-013 core from Solignac et al. (2004); (e) sortable silt data reflecting Iceland-Scoland overflow water (cf. Thornalley et al., 2013); (f) Intensity of the upper-water column stratification south of Iceland (cf. Thornalley et al., 2009); (g) Drift-ice off Northern Iceland (Andrews et al., 2009). Blue bars represent episodes of cooling. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

linked modulations are discussed below in the light of the MD95-2002 hydrological reconstructions, with reference to salinity estimates notably.

5.3. Millennial-scale climate variability at North Atlantic scale Significant variability is observed all along the section. A total of seven major episodes of cooling are recorded at 11.4–11, 10.5–10, 8.9, 7.6, 5.3, 3.3–2.8 ka and 2–1.5 ka (Fig. 7). These cooling pulses were first detected by Denton and Karlén (1973) and Bond et al. (1997), highlighting a strong millennial variability throughout the Holocene, notably in the northern North Atlantic region (e.g., Dansgaard et al., 1993; Andersen et al., 2004b), but also on land, as shown by major advances of Central European and Scandinavian glaciers (e.g., Denton and Karlén, 1973; Solomina et al., 2015), speleothem δ18O records from northwestern Spain (Railsback et al., 2011) together with periods of intensified storminess in northern Europe (Sorrel et al., 2012). There are also numerous marine records that have highlighted variations in sea-surface temperatures along the North Atlantic Drift (NAD) (cf. Solignac et al., 2004) and the Norwegian Atlantic Current (NwAC) (Hald et al., 2007). There are also records showing variations in the strength of the North Atlantic Deep Water (NADW) production, and especially that of the Iceland-Scotland overflow component (Hall et al., 2004; Hoogakker et al., 2011; Thornalley et al., 2013). Unfortunately, the timing of the events seems to differ from one region to another and the regional coherence remains equivocal due to chronological uncertainties.

The comparison of SSSs and SSTs at site MD95-2002 with the sea-ice record of core HU-090-013-013 (P-013; Fig. 7) located on the Greenland Rise (Fig. 1) suggests a strong linkage between the Labrador Sea and the Bay of Biscay as the three curves show a very similar shape. Most maxima of sea-ice cover duration off southwest Greenland coincide with episodes of cooling and low SSSs in the northern Bay of Biscay (Fig. 7). Interestingly, the reconstruction of density anomalies at sub-thermocline depths in the southern Iceland basin by Thornalley et al. (2009) matches the SSS modulations estimated in MD95-2002 from dinocyst assemblages (Fig. 7). Thornalley et al. (2009) have pointed out atmospheric systems as primary triggers of the Holocene millennial variability based on variations of sea-surface salinities and temperatures in the subpolar North Atlantic during Holocene. The consistency of the three records, based on different proxies and from different cores, supports a direct and rapid transfer of anomalies towards the east, thus supporting the hypothesis of an atmospheric dominant forcing via the westerlies. In such a hypothesis, intensified zonal winds across the North Atlantic would have resulted in lower SSTs and more vigorous transport of sea-ice and related meltwater from the Labrador Sea to the Bay of Biscay, impacting on its way the southern Iceland basin (Bond et al., 2001). Such a scenario is consistent with strengthening of the Subpolar Gyre (SPG) under reinforced westerlies, which in turn extends and carries meltwater plumes originating from Greenland further east (Thornalley et al., 2009; Morley and Rosenthal, 2014) thus causing cooling and low salinity pulses recorded in the Northeastern North Atlantic (Bond et al., 2001). Inversely, periods of high SSTs and SSSs at site MD95-2002 correspond to phases of high upper water mass

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

stratification south of Iceland (Fig. 7), probably related to the advection of warmer and saltier water masses at the site as the result of westward contraction of the SPG that enabled a northward extension of the Subtropical Gyre (STG) in the Northeastern North Atlantic (e.g. Sorrel et al., 2012; Staines-Urías et al., 2013; and references therein). These extensions/contractions of the SPG implying episodes of surface water freshening are likely to have played a major role on sea-surface conditions over the North Atlantic and on the AMOC, which recorded episodic weakening notably in the Iceland-Scotland Overflow Water (ISOW) production (Hall et al., 2004; Hoogakker et al., 2011; Thornalley et al., 2013). Therefore, the overall data tend to demonstrate close relationship between sea-ice cover in the Labrador Sea, sea-surface conditions in the Bay of Biscay, changes in the upper-water column stratification south of Iceland and the strength of the ISOW during the Holocene (Fig. 7). Two features are worth mentioning when comparing ISOW strength to MD95-2002 SSSs. The first one is recorded between 11 and 10 ka. It corresponds to high ISOW production and very low sea-ice cover duration at P-013 whereas the MD95-2002 SSS shows the lowest values of the entire section. This has already been discussed with reference to the core location subject to important meltwater advection from the proximal European continent (see Section 5.1). The second feature concerns the interval after 5 ka when, despite concordant oscillations with P-013 sea-ice, the SSS values recorded in core MD95–2002 are the highest of the entire section (Fig. 7). This might be related to the end of meltwater discharge from the LIS after its final collapse around

411

6.5 ka (Fig. 6) (e.g., Dyke, 2004; Kaplan and Wolfe, 2006; de Vernal and Hillaire-Marcel, 2006). This might also relate to a contraction of the SPG during warm episodes, which resulted in weaker influence of the NAD in the Bay of Biscay, compensated by enhanced effect of the more saline and warmer Slope Current (SC) during the late Holocene (Fig. 1). Solignac et al. (2008) and Mojtahid et al. (2013) also evoked higher contributions of the SC along the proximal continental margins from the study of core HM03-133-25 and KS10b, respectively located in the Faroe-Shetland Channel (FSC) and the southeastern Bay of Biscay. 5.4. The importance of the atmospheric forcing Although, the intensity of the westerlies seems to have played a major role in the west-east freshwater transfer in the North Atlantic they cannot explain by themselves the initial and synchronous variations of sea-ice cover duration recorded at site P-013, more extensive sea-ice cover being recorded concomitantly to low SSS events in the bay of Biscay, and thus to positive density anomalies at sub-thermocline depth (Fig. 7). A possible linkage with the Arctic Oscillation (AO)/NAO mode of variability has already been proposed in previous studies (e.g. Andrews et al., 2009; Giraudeau et al., 2010; Staines-Urías et al., 2013; Morley and Rosenthal, 2014) with positive mode being accompanied by enhanced strength of the East Greenland Current (EGC) and export of Arctic sea ice along the East Greenland margins (Fig. 7), thus affecting the freshwater budget of the northern North Atlantic and AMOC (cf. Thornalley

Fig. 8. (a) Winter and summer SSTs estimated from dinocysts in core MD95-2002 and (b) insolation at 47°N in January and July from Laskar et al. (2004); (c) SSS estimated from dinocysts in core MD95-2002; (d) Drift-ice off Northern Iceland (Andrews et al., 2009); (e) Forest evolution in the SE Iberia Peninsula (Fletcher et al., 2012). Blue bars represent episodes of cooling. Grey bars represent episodes of forest declines in the Western Mediterranean (Fletcher et al., 2012). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

412

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413

et al., 2009, 2013). This is corroborated by the strong correlation between the episodes of cooling and low SSSs recorded at site MD952002 and the periods of increased drift-ice advection southward along the EGC (Andrews et al., 2009). Furthermore, several authors studying the Holocene climate variability over the Iberian Peninsula (Bernárdez et al., 2008; Fletcher et al., 2012; Chabaud et al., 2014) have shown that the episodes of dryness over the Peninsula were related to positive mode of the NAO and associated with cold and wet episodes over the northeastern North Atlantic. Six of the seven episodes of forest declines in the western Mediterranean induced by dry winters (Fletcher et al., 2012) (Fig. 8) are concomitant with episodes of reduced stratification south of Iceland (cf. Fletcher et al., 2012) and the cooling pulses recorded at site MD95-2002. Hence, it seems clear that the NAO mode of variability and its impact on the SPG behavior play a key role in the millennial-scale variability of the Holocene climate. 6. Conclusion The study of the sea-surface conditions recorded from core MD952002 during the Holocene illustrates the complexity of the ocean-atmosphere-cryosphere interactions in the boreal Atlantic as also previously highlighted at other key locations (e.g., Bond et al., 1997, 2001; Hall et al., 2004; Andersen et al., 2004b; Solignac et al., 2004, 2006; Giraudeau et al., 2010; Morley and Rosenthal, 2014; Thornalley et al., 2013). Beyond the findings of other studies, the present paper shed light on a few regional characteristics of hydrographical changes. First, evidence for a strong stratification of the upper-water column in the Bay of Biscay, especially until 7 ka as the site was submitted to freshwater advection from both proximal European continent and distal LIS. Second, the reconstructions of sea-surface conditions highlights variations during the Holocene that correspond to the classical paleoclimatic subdivisions of the Holocene with a climatic optimum between 7 and 5.5 ka, although its identification is equivocal because its signal was partly hidden by the meltwater inputs of the remnant LIS. At last, the record shows that the long-term trends are superimposed to large amplitude changes at centennial to millennial time-scales: Seven episodes of cooling and low salinity have been identified at site MD95-2002 and correspond to episodes of enhanced sea-ice cover in the Labrador Sea. Enhanced westerlies at these periods are likely to have lowered SSTs at both sites and favored the advection of sea-ice and related meltwater from south Greenland to the northern Bay of Biscay. Such phenomenon, probably linked to the NAO variability, might have had a huge impact on the AMOC as weakenings in the ISOW production are also recorded during these very same periods. Author contributions J.Z., F.E. and A. dV. designed this study focussed on the MD95-2002 Holocene section. J.Z. and F.E. performed dinocyst analyses. All authors contributed to hydrographical quantification calculations and establishment of the age model. J.Z. performed the new foraminiferal analyses (stable isotopes) on the core. Each author contributed to discussions and interpretation of the results, and wrote the manuscript. Acknowledgments Core MD95-2002 was recovered thanks to financial support from MENRT, CNRS, IPEV and European Union MAST programme ENAM2. This work was supported by the funds from the Natural Sciences and Engineering Research Council of Canada (NSERC) (38340) and the Fonds de Recherche du Québec – Nature et Technologies (FRQNT) and through French projects: ANR HAMOC, INSU TS/INTERRVIE & LEFE/IMAGO ICE-BIO-RAM, ARTEMIS 14C facilities, and has benefited from the European Union's Seventh Framework programme (FP7/2007–2013) under grant 243908, “Past4Future - Climate change-Learning from the

past climate”. Thanks are due to Claude Hillaire-Marcel for his most valuable guidance and Maryse Henry and Jean-François Hélie (GEOTOP) for their help in laboratory. We are grateful to the anonymous reviewers of the journal for their helpful and constructive comments on the manuscript. References Andersen, C., Koc, N., Moros, M., 2004a. A highly unstable Holocene climate in the subpolar North Atlantic: evidence from diatoms. Quat. Sci. Rev. 23 (20), 2155–2166. Andersen, C., Koc, N., Jennings, A., Andrews, J., 2004b. Nonuniform response of the major surface currents in the Nordic seas to insolation forcing: implications for the Holocene climate variability. Paleoceanography 19 (2). Andrews, J.T., Darby, D., Eberle, D., Jennings, A.E., Moros, M., Ogilvie, A., 2009. A robust, multisite Holocene history of drift ice off northern Iceland: implications for North Atlantic climate. The Holocene 19 (1), 71–77. Auffret, G., Zaragosi, S., Dennielou, B., Cortijo, E., Van Rooij, D., Grousset, F., ... Siegert, M., 2002. Terrigenous fluxes at the Celtic margin during the last glacial cycle. Mar. Geol. 188 (1), 79–108. Bassinot, F., Labeyrie, L., 1996. Les rapports de campagnea la mera bord du MarionDufresne: campagne IMAGES MD 101 (du 29-05-95 au 11-07-95). IFRTP Editions, Brest. Berger, A., 1978. Long-term variations of caloric insolation resulting from the Earth's orbital elements. Quat. Res. 9 (2), 139–167. Berger, A., Loutre, M.-F., 1991. Insolation values for the climate of the last 10 million years. Quat. Sci. Rev. 10 (4), 297–317. Bernárdez, P., González-Álvarez, R., Francés, G., Prego, R., Bárcena, M., Romero, O.E., 2008. Late Holocene history of the rainfall in the NW Iberian peninsula—evidence from a marine record. J. Mar. Syst. 72 (1), 366–382. Blaauw, M., 2010. Methods and code for ‘classical’ age-modelling of radiocarbon sequences. Quat. Geochronol. 5 (5), 512–518. Bond, G., Showers, W., Cheseby, M., Lotti, R., Almasi, P., Priore, P., ... Bonani, G., 1997. A pervasive millennial-scale cycle in North Atlantic Holocene and glacial climates. Science 278 (5341), 1257–1266. Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W., ... Bonani, G., 2001. Persistent solar influence on North Atlantic climate during the Holocene. Science 294 (5549), 2130–2136. Bourillet, J.-F., Zaragosi, S., Mulder, T., 2006. The French Atlantic margin and deep-sea submarine systems. Geo-Mar. Lett. 26 (6), 311–315. Broecker, W.S., Andree, M., Wolfli, W., Oeschger, H., Bonani, G., Kennett, J., Peteet, D., 1988. The chronology of the last deglaciation: implications to the cause of the Younger Dryas event. Paleoceanography 3 (1), 1–19. Carton, J.A., Hakkinen, S., 2011. Introduction to: Atlantic Meridional Overturning Circulation (AMOC). Deep-Sea Res. II Top. Stud. Oceanogr. 58 (17), 1741–1743. Chabaud, L., Goñi, M.F.S., Desprat, S., Rossignol, L., 2014. Land–sea climatic variability in the eastern North Atlantic subtropical region over the last 14,200 years: atmospheric and oceanic processes at different timescales. The Holocene 24 (7), 787–797. Conkright, M.E., Locarnini, R.A., Garcia, H.E., O'Brien, T.D., Boyer, T.P., Stephens, C., Antonov, J.I., 2002. World Ocean Atlas 2001: Objective Analyses, Data Statistics, and Figures: CD-ROM documentation: US Department of Commerce, National Oceanic and Atmospheric Administration. National Oceanographic Data Center, Ocean Climate Laboratory. Coplen, T.B., 1988. Normalization of oxygen and hydrogen isotope data. Chem. Geol. Isot. Geosci. 72 (4), 293–297. Cossa, D., Cotté-Krief, M.-H., Mason, R.P., Bretaudeau-Sanjuan, J., 2004. Total mercury in the water column near the shelf edge of the European continental margin. Mar. Chem. 90 (1), 21–29. Dansgaard, W., Johnsen, S., Clausen, H., Dahl-Jensen, D., Gundestrup, N., Hammer, C., ... Jouzel, J., 1993. Evidence for general instability of past climate from a 250-kyr icecore record. Nature 364 (6434), 218–220. de Vernal, A., Hillaire-Marcel, C., 2006. Provincialism in trends and high frequency changes in the northwest North Atlantic during the Holocene. Glob. Planet. Chang. 54 (3), 263–290. de Vernal, A., Hillaire-Marcel, C., Bilodeau, G., 1996. Reduced Meltwater Outflow from the Laurentide Ice Margin during the Younger Dryas. de Vernal, A., Henry, M., Bilodeau, G., 1999. Techniques de préparation et d'analyse en micropaléontologie. Les cahiers du GEOTOP. 3, p. 41. de Vernal, A., Hillaire-Marcel, C., Rochon, A., Fréchette, B., Henry, M., Solignac, S., Bonnet, S., 2013. Dinocyst-based reconstructions of sea ice cover concentration during the Holocene in the Arctic Ocean, the northern North Atlantic Ocean and its adjacent seas. Quat. Sci. Rev. 79, 111–121. Denton, G.H., Karlén, W., 1973. Holocene climatic variations—their pattern and possible cause. Quat. Res. 3 (2), 155IN1175–1174IN2205. Duplessy, J.-C., Shackleton, N.J., Matthews, R.K., Prell, W., Ruddiman, W.F., Caralp, M., Hendy, C.H., 1984. 13 C record of benthic foraminifera in the last interglacial ocean: implications for the carbon cycle and the global deep water circulation. Quat. Res. 21 (2), 225–243. Dyke, A.S., 2004. An outline of North American deglaciation with emphasis on central and northern Canada. Quaternary Glaciations: Extent and Chronology. 2, pp. 373–424. Dyke, A.S., Moore, A., Robertson, L., 2003. Deglaciation of North America. Geological Survey of Canada, Ottawa, ON. Eynaud, F., 1999. Kystes de Dinoflagellés et Evolution paléoclimatique et paléohydrologique de l'Atlantique Nord au cours du Dernier Cycle Climatique du Quaternaire.

J. Zumaque et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 468 (2017) 403–413 Eynaud, F., Turon, J., Duprat, J., 2004. Comparison of the Holocene and Eemian palaeoenvironments in the south Icelandic Basin: dinoflagellate cysts as proxies for the North Atlantic surface circulation. Rev. Palaeobot. Palynol. 128 (1), 55–79. Eynaud, F., Zaragosi, S., Scourse, J., Mojtahid, M., Bourillet, J.-F., Hall, I.R., ... Reijonen, A., 2007. Deglacial laminated facies on the NW European continental margin: The hydrographic significance of British-Irish Ice Sheet deglaciation and Fleuve Manche paleoriver discharges. Geochem. Geophys. Geosyst. 8 (6). Eynaud, F., Malaizé, B., Zaragosi, S., Vernal, A., Scourse, J., Pujol, C., ... Toucanne, S., 2012. New constraints on European glacial freshwater releases to the North Atlantic Ocean. Geophys. Res. Lett. 39 (15). Farmer, E.J., Chapman, M.R., Andrews, J.E., 2008. Centennial-scale Holocene North Atlantic surface temperatures from Mg/Ca ratios in Globigerina bulloides. Geochem. Geophys. Geosyst. 9 (12). Fletcher, W.J., Debret, M., Goñi, M.F.S., 2012. Mid-Holocene emergence of a low-frequency millennial oscillation in western Mediterranean climate: implications for past dynamics of the North Atlantic atmospheric westerlies. The Holocene 0959683612460783. Frew, R.D., Dennis, P.F., Heywood, K.J., Meredith, M.P., Boswell, S.M., 2000. The oxygen isotope composition of water masses in the northern North Atlantic. Deep-Sea Res. I Oceanogr. Res. Pap. 47 (12), 2265–2286. Giraudeau, J., Grelaud, M., Solignac, S., Andrews, J., Moros, M., Jansen, E., 2010. Millennialscale variability in Atlantic water advection to the Nordic seas derived from Holocene coccolith concentration records. Quat. Sci. Rev. 29 (9), 1276–1287. Guiot, J., de Vernal, A., 2007. Transfer functions: methods for quantitative paleoceanography based on microfossils. Developments in Marine Geology. 1, pp. 523–563. Hald, M., Andersson, C., Ebbesen, H., Jansen, E., Klitgaard-Kristensen, D., Risebrobakken, B., ... Telford, R.J., 2007. Variations in temperature and extent of Atlantic water in the northern North Atlantic during the Holocene. Quat. Sci. Rev. 26 (25), 3423–3440. Hall, I.R., Bianchi, G.G., Evans, J.R., 2004. Centennial to millennial scale Holocene climatedeep water linkage in the North Atlantic. Quat. Sci. Rev. 23 (14), 1529–1536. Hoogakker, B.A., Chapman, M.R., McCave, I.N., Hillaire-Marcel, C., Ellison, C.R., Hall, I.R., Telford, R.J., 2011. Dynamics of North Atlantic deep water masses during the Holocene. Paleoceanography 26 (4). http://www.r-project.org/. Ivy-Ochs, S., 2015. Glacier variations in the European Alps at the end of the last glaciation. Cuadernos de investigación geográfica 41 (2), 295–315. Ivy-Ochs, S., Kerschner, H., Reuther, A., Maisch, M., Sailer, R., Schaefer, J., ... Schlüchter, C., 2006. The timing of glacier advances in the northern European Alps based on surface exposure dating with cosmogenic 10Be, 26Al, 36Cl, and 21Ne. Geol. Soc. Am. Spec. Pap. 415, 43–60. Kaplan, M.R., Wolfe, A.P., 2006. Spatial and temporal variability of Holocene temperature in the North Atlantic region. Quat. Res. 65 (2), 223–231. Kroopnick, P., 1985. The distribution of 13 C of ΣCO 2 in the world oceans. Deep Sea Res. Part A 32 (1), 57–84. Kucera, M., Rosell-Melé, A., Schneider, R., Waelbroeck, C., Weinelt, M., 2005. Multiproxy approach for the reconstruction of the glacial ocean surface (MARGO). Quat. Sci. Rev. 24 (7), 813–819. Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A., Levrard, B., 2004. A long-term numerical solution for the insolation quantities of the earth. Astron. Astrophys. 428 (1), 261–285. Lazure, P., Dumas, F., Vrignaud, C., 2008. Circulation on the Armorican shelf (Bay of Biscay) in autumn. J. Mar. Syst. 72 (1), 218–237. Marchal, O., Cacho, I., Stocker, T.F., Grimalt, J.O., Calvo, E., Martrat, B., ... van Kreveld, S., 2002. Apparent long-term cooling of the sea surface in the northeast Atlantic and Mediterranean during the Holocene. Quat. Sci. Rev. 21 (4), 455–483. Mertens, K.N., Verhoeven, K., Verleye, T., Louwye, S., Amorim, A., Ribeiro, S., ... González, C., 2009. Determining the absolute abundance of dinoflagellate cysts in recent marine sediments: the Lycopodium marker-grain method put to the test. Rev. Palaeobot. Palynol. 157 (3), 238–252. Mojtahid, M., Jorissen, F., Garcia, J., Schiebel, R., Michel, E., Eynaud, F., ... Siccha, M., 2013. High resolution Holocene record in the southeastern Bay of Biscay: global versus regional climate signals. Palaeogeogr. Palaeoclimatol. Palaeoecol. 377, 28–44. Morley, A., Rosenthal, Y., 2014. Ocean-atmosphere climate shift during the mid-to-late Holocene transition. Earth Planet. Sci. Lett. 388, 18–26. National Oceanographic Data Centre (NODC), 2001. World Ocean Atlas. http://www. nodc.noaa.gov/OC5/WODOl/pr_wodOl.html. Naughton, F., Goñi, M.S., Desprat, S., Turon, J.-L., Duprat, J., Malaizé, B., ... Freitas, M., 2007. Present-day and past (last 25000 years) marine pollen signal off western Iberia. Mar. Micropaleontol. 62 (2), 91–114.

413

Nesje, A., Dahl, S.O., 1993. Late Glacial and Holocene glacier fluctuations and climate variations in western Norway: a review. Quat. Sci. Rev. 12 (4), 255–261. Penaud, A., Eynaud, F., Turon, J., Zaragosi, S., Malaizé, B., Toucanne, S., Bourillet, J.-F., 2009. What forced the collapse of European ice sheets during the last two glacial periods (150 ka BP and 18 ka cal BP)? Palynological evidence. Palaeogeogr. Palaeoclimatol. Palaeoecol. 281 (1), 66–78. Radi, T., de Vernal, A., 2008. Dinocysts as proxy of primary productivity in mid–high latitudes of the northern hemisphere. Mar. Micropaleontol. 68 (1), 84–114. Railsback, L.B., Liang, F., Romaní, J.R.V., Grandal-d'Anglade, A., Rodríguez, M.V., Fidalgo, L.S., ... Edwards, R.L., 2011. Petrographic and isotopic evidence for Holocene longterm climate change and shorter-term environmental shifts from a stalagmite from the Serra do Courel of northwestern Spain, and implications for climatic history across Europe and the Mediterranean. Palaeogeogr. Palaeoclimatol. Palaeoecol. 305 (1), 172–184. Ravelo, A.C., Hillaire-Marcel, C., 2007. Chapter eighteen the use of oxygen and carbon isotopes of foraminifera in paleoceanography. Developments in Marine Geology. 1, pp. 735–764. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., ... Friedrich, M., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55 (4), 1869–1887. Rochon, A., Vernal, A.d., Turon, J.-L., Matthießen, J., Head, M., 1999. Distribution of recent dinoflagellate cysts in surface sediments from the North Atlantic Ocean and adjacent seas in relation to sea-surface parameters. American Association of Stratigraphic Palynologists Contribution Series. 35, pp. 1–146. Schiebel, R., Hemleben, C., 2005. Modern planktic foraminifera. Paläontol. Z. 79 (1), 135–148. Solignac, S., de Vernal, A., Hillaire-Marcel, C., 2004. Holocene sea-surface conditions in the North Atlantic—contrasted trends and regimes in the western and eastern sectors (Labrador Sea vs. Iceland Basin). Quat. Sci. Rev. 23 (3), 319–334. Solignac, S., Giraudeau, J., de Vernal, A., 2006. Holocene sea surface conditions in the western North Atlantic: spatial and temporal heterogeneities. Paleoceanography 21 (2). Solignac, S., Grelaud, M., de Vernal, A., Giraudeau, J., Moros, M., McCave, I.N., Hoogakker, B., 2008. Reorganization of the upper ocean circulation in the mid-Holocene in the northeastern Atlantic. Can. J. Earth Sci. 45 (11), 1417–1433. Solomina, O.N., Bradley, R.S., Hodgson, D.A., Ivy-Ochs, S., Jomelli, V., Mackintosh, A.N., ... Wiles, G.C., 2015. Holocene glacier fluctuations. Quat. Sci. Rev. 111, 9–34. Sorrel, P., Debret, M., Billeaud, I., Jaccard, S.L., McManus, J.F., Tessier, B., 2012. Persistent non-solar forcing of Holocene storm dynamics in coastal sedimentary archives. Nat. Geosci. 5 (12), 892–896. Staines-Urías, F., Kuijpers, A., Korte, C., 2013. Evolution of subpolar North Atlantic surface circulation since the Early Holocene inferred from planktic foraminifera faunal and stable isotope records. Quat. Sci. Rev. 76:66–81. http://dx.doi.org/10.1016/j. quascirev.2013.06.016. Sutton, R., Allen, M., 1997. Decadal predictability of North Atlantic sea surface temperature and climate. Nature 388 (6642), 563–567. Thornalley, D.J., Elderfield, H., McCave, I.N., 2009. Holocene oscillations in temperature and salinity of the surface subpolar North Atlantic. Nature 457 (7230), 711–714. Thornalley, D.J., Blaschek, M., Davies, F.J., Praetorius, S., Oppo, D.W., McManus, J.F., ... McCave, I.N., 2013. Long-term Variations in Iceland–Scotland Overflow Strength During the Holocene. Toucanne, S., Zaragosi, S., Eynaud, F., Bourillet, J.-F., Lericolais, G., Gibbard, P., 2011. Comments to Westaway and Bridgland–‘Causes, consequences and chronology of largemagnitude palaeoflows in Middle and Late Pleistocene river systems of Northwest Europe’. Earth Surf. Process. Landf. 36 (10), 1410–1413. van Aken, H.M., 2000. The hydrography of the mid-latitude Northeast Atlantic Ocean: I: the deep water masses. Deep-Sea Res. I Oceanogr. Res. Pap. 47 (5), 757–788. Zaragosi, S., Auffret, G., Faugères, J.-C., Garlan, T., Pujol, C., Cortijo, E., 2000. Physiography and recent sediment distribution of the Celtic Deep-Sea Fan, Bay of Biscay. Mar. Geol. 169 (1), 207–237. Zaragosi, S., Eynaud, F., Pujol, C., Auffret, G., Turon, J.-L., Garlan, T., 2001. Initiation of the European deglaciation as recorded in the northwestern Bay of Biscay slope environments (Meriadzek Terrace and Trevelyan Escarpment): a multi-proxy approach. Earth Planet. Sci. Lett. 188 (3), 493–507. Zaragosi, S., Bourillet, J.-F., Eynaud, F., Toucanne, S., Denhard, B., Van Toer, A., Lanfumey, V., 2006. The impact of the last European deglaciation on the deep-sea turbidite systems of the Celtic-Armorican margin (Bay of Biscay). Geo-Mar. Lett. 26 (6), 317–329. Zumaque, J., 2014. Les reconstitutions paléo-hydrographiques de surface dans le Golfe de Gascogne au cours de l'Holocène.