Quaternary Science Reviews 224 (2019) 105973 Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev Postglacial paleoceanography of the western Barents Sea: Implications for alkenone-based sea surface temperatures and primary productivity Magdalena Ła˛ cka a, *, Min Cao b, c, Antoni Rosell-Mele  c, d, Joanna Pawłowska a, Małgorzata Kucharska , Matthias Forwick , Marek Zaja˛ czkowski a a e a  co Institute of Oceanology, Polish Academy of Sciences, Powstan w Warszawy 55, 81-712, Sopot, Poland b Key Laboratory of Karst Environment, School of Geographical Sciences, Southwest University, Chongqing, 400715, China c noma de Barcelona, 08193, Bellaterra, Spain Institute of Environmental Science and Technology (ICTA-UAB), Universitat Auto d  Catalana de Recerca i Estudis Avançats, 08010 Barcelona, Spain Institucio e Department of Geosciences, UiT The Arctic University of Norway in Tromsø, N-9037, Tromsø, Norway a r t i c l e i n f o a b s t r a c t Article history: The increasing influence of Atlantic Water (AW) in the Barents Sea, a process known as “Atlantification”, Received 12 February 2019 is gradually decreasing sea ice cover in the region. Ongoing global climate warming is likely to be one of Received in revised form its drivers, but to further understand the role of natural variability and the biogeochemical impacts of the 26 September 2019 inflow of AW into the western Barents Sea, we reconstructed sea surface temperatures (SSTs) and pri- Accepted 29 September 2019 mary productivity in Storfjordrenna, a climatically sensitive area south of Spitsbergen, between Available online 11 October 2019  approximately 13,950 cal yr BP and 1300 cal yr BP. The alkenone UK 37 proxy has been applied to recon- struct SSTs, and the alkenone accumulation rate in marine sediments has been used to infer changes in Keywords: Alkenones primary productivity. Our data show that the SST increase was concomitant with the progressive loss of Stratification sea ice cover and an increase in primary productivity in the western Barents Sea. We interpret these Holocene changes as reflecting the increasing influence of AW in the area as the ice sheets retreated in Svalbard. Sea ice decrease The transition from the Arctic to the Atlantic domain first occurred after 11,500 cal yr BP, as the Arctic Global warming Front moved eastward of the study site but with considerable variability in surface ocean conditions. North Atlantic Current High SSTs at approximately 6400 cal yr BP may have led to limited winter surface cooling, likely inhib- Arctic iting convective mixing and the return of nutrients to the euphotic zone and/or enhanced organic matter consumption by zooplankton due to an earlier light signal in the ice-free Storfjordrenna. During the late Holocene (3600-1300 cal yr BP), low insolation facilitated sea ice formation and thus brine production. The former may have launched convective water mixing and increased nutrient resupply to the sea surface, consequently enhancing primary productivity in Storfjordrenna. We propose that, on the basis of the paleoceanographic evidence, the modern increasing inflow of warm AW and the disappearance of pack ice on the Eurasian continental shelf are likely to weaken convective water mixing and decrease primary production in the region. © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction et al., 2017). The most striking changes in the extent and thick- ness of sea ice have occurred in the western Barents Sea (Onarheim The effects of ongoing climate warming are especially percep- et al., 2015; Yang et al., 2016). Sea ice decrease in the area is a tible in the Arctic, mainly due to the rapid decline in the extent of reflection of the growing influence of Atlantic water (AW), a process sea ice over the past few decades, which numerous climate models sometimes referred to as “Atlantification” (e.g., Årthun et al., 2012; expect to continue throughout the twenty-first century (e.g., Ding Oziel et al., 2016), which has caused the Barents Sea to reach its highest temperatures over the last decade, since systematic instrumental measurements began (Boitsov et al., 2012). * Corresponding author. The accelerated loss of sea ice in the Barents Sea has a profound E-mail address: mlacka@iopan.gda.pl (M. Ła˛ cka). impact on the global energy budget, atmospheric and oceanic https://doi.org/10.1016/j.quascirev.2019.105973 0277-3791/© 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 2 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 circulation, and the carbon cycle (Polyakov et al., 2017; Serreze and exchanges across the front (Walczowski, 2013; Ła˛ cka et al., 2015b). Barry, 2011). Among the Arctic shelf seas, the Barents Sea seafloor is The study area is the glacial trough of Storfjordrenna, located the largest carbon sink, playing a significant role in controlling south of Spitsbergen (Fig. 1B). Storfjordrenna extends towards the global atmospheric CO2 concentrations and global climate north into Storfjorden, which is located north of a sill of 120 m (Smedsrud et al., 2013). Further decreases in ice cover and increases depth. AW flows into Storfjordrenna in a cyclonic manner parallel in surface water temperature in the western Barents Sea may cause to the trough's southern margin and along the northern slope to- primary productivity to decrease by 15e25% due to reduced wards its mouth (Fer et al., 2003). ArW enters the trough from the nutrient flux into the upper ocean, resulting from decreasing southeast with the ESC (Loeng, 1991). The uppermost 500 m of the advective mixing (Lewandowska et al., 2014; Lind et al., 2018). As water column at the mouth of Storfjordrenna contains relatively the Barents Sea ecosystem supports some of the world's largest warm (4e7  C) and saline (35.2) AW (Fig. 2A). Cold (0.5  C) in- stocks of fish (Dalpadado et al., 2012, 2014), this issue also has termediate water, generated by convection in the Nordic seas, oc- global economic importance. In addition, more organic matter curs beneath the AW (Nilsen et al., 2008). could be buried, relatively speaking, in the seasonally ice-covered In the central part of Storfjordrenna, AW occurs between 50 and northern Arctic regions than in the ice-free areas of the western 160-m water depth (Fig. 2B). However, it is cooler here (max. 4.2  C) Barents Sea (Pathirana et al., 2014; Slagstad et al., 2011). than in the outer part of the Spitsbergen shelf (Fig. 2A). The surface The “Atlantification” of the western Barents Sea is a process that water (SW) reaches 3.8  C, which is comparable to the temperature has occurred in the recent geological past. After the Younger Dryas of the AW at the Storfjordrenna mouth. However, it has a lower (YD), the inflow of AW to the western Barents Sea (Hald et al., 2007; salinity (33.5; Fig. 2B) than the AW at the Storfjordrenna mouth.  Slubowska-Woldengen et al., 2008) and maximum Holocene During the winter-freezing period, which typically lasts from late summer insolation (Berger and Loutre, 1991) caused the increase in November to mid-May, brine-enriched shelf water (BSW) is pro- sea surface temperatures (SSTs) and decrease in sea ice (Rasmussen duced in Storfjorden (Haarpaintner et al., 2001; Skogseth et al., et al., 2014). Sea ice-free conditions remained during most of the 2004). BSW fills Storfjorden to the top of the sill and then begins mid-Holocene as a result of the influence of AW (Ła˛ cka et al., a gravity-driven overflow through Storfjordrenna, further down the 2015b). These conditions coincided with the highest Holocene continental slope (Schauer et al., 2003). Thus, in Storfjordrenna, a primary productivity observed on the northern continental slope of 30-m-thick layer of BSW (temperature of approximately 0  C and the Barents Sea (Wollenburg et al., 2004). salinity slightly over 35.1) usually occurs in the deepest region The objective of this study is to reconstruct the impact of AW immediately above the seafloor (Fig. 2B). inflow to the western Barents Sea, and its influence on primary Northeast from Storfjordrenna, in the northern Barents Sea, AW productivity, from c. 13,950 cal yr BP to c. 1300 cal yr BP. We based flows via the Franz-Victoria Trough and the Kvitøya Trough, from  our reconstruction on the alkenone UK 37 proxy to reconstruct SSTs the Eurasian continental slope (Fig. 1; Pfirman et al., 1994; Lind and (Bendle and Rosell-Mele , 2004) and the alkenone accumulation Ingvaldsen, 2012) below the sea surface dominated by ArW rate in marine sediments to infer the changes in primary produc- (Gammelsrød et al., 2009). ArW occupies the upper 20e100 m of tivity (e.g., Bolton et al., 2010). We compared our data to other the northern Barents Sea, with temperature down to the freezing studies conducted in the western Barents Sea continental slope point (Lind and Ingvaldsen, 2012), whereas AW is found at depths (Martrat et al., 2003), the southwestern Barents Sea (Risebrobakken of 150e350 m, has a salinity of 34.7 and a temperature of 1.1e0.8  C et al., 2010) and the Norwegian Sea (Calvo et al., 2002, Fig. 1; (Fig. 2C) (Klitgaard-Kristensen et al., 2013). The northern Barents Table 1). The comparison provides new information on lateral and Sea experiences seasonal sea ice cover that forms during fall/winter vertical oceanographic gradients, i.e., ocean vertical mixing, ther- (Loeng, 1991). Sea ice breakup occurs during summer, leading to mal stratification, and Arctic front movement in the western open-water conditions in August and September (Belt et al., 2015). Barents Sea since the last deglaciation. Recently, there is observed a decline in sea-ice formation in the northern Barents Sea, due to decrease in sea-ice import through the 2. Oceanographic setting passage between Franz Josef Land and Novaya Zemlya, and a cor- responding loss in freshwater, leading to weakened ocean stratifi- The Barents Sea is an Arctic shelf sea located along the main cation, enhanced vertical mixing and increased upward fluxes of pathway of heat and salt transport within the North Atlantic Cur- heat and salt (Lind et al., 2018). rent (NAC) entering the Arctic (Smedsrud et al., 2013; Rudels et al., 2015, Fig. 1). Thus, it is influenced by two main water masses, warm 3. Materials and methods and saline AW and colder and fresher Arctic Water (ArW). AW (T > 3  C, S > 35.0; Loeng, 1991) is transported northwards by the 3.1. Sediment cores NAC, following the continental slope of Norway (Fig. 1A). The topographically steered flow of NAC bifurcates into two branches c. A 4.25-m-long gravity core JM09-020-GC (76.31489 N, 72 N (Fig. 1A). One branch flows into the southern Barents Sea (i.e., 19.69957 E) was retrieved from a 253-m water depth in Storfjor- North Cape Current), whereas the other branch continues north- drenna, with the R/V Jan Mayen (now: Helmer Hanssen; UiT The wards along the western Barents Sea slope and western Svalbard Arctic University of Norway) in November 2009 (Fig. 1). The coring margin into the Arctic Ocean as the West Spitsbergen Current site is located inside Storfjordrenna in a 3-km broad flat plain (<1% (Aagaard et al., 1987) (Fig. 1). The ArW (T < 0  C, S < 34.4), formed inclination); its location was selected after an echo-acoustic by mixing of AW and polar waters in the Arctic Ocean, enters the investigation in an area whose sediments were not under direct Barents Sea from the north and is carried southward by the East influence of brine-enriched waters. Conductivity-temperature- Spitsbergen Current (ESC) (Loeng, 1991). At the boundary between depth (CTD) measurements were performed in August 2015 at Arctic and Atlantic water, a sharp gradient in terms of temperature, the coring site (Figs. 1 and 2). salinity, and sea-ice distribution is formed called the Arctic Front The new results from core JM09-020-GC were compared to (AF; Hopkins, 1991). The AF determines the position of the marginal previously published alkenone data (Table 1). Core PSh-5159N ice zone and surface productivity in the summer season (Smith and (Risebrobakken et al., 2010) was retrieved from the SW Barents Sakshaug, 1990). At present, the AF is located east of our study area; Sea (Fig. 1) at a water depth of 422 m. Its basal age is c. 15,000 cal yr however, the frontal zone changes its position, and water mixes and BP. Core M23258-2 (Martrat et al., 2003) was recovered from the M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 3 Fig. 1. Simplified map showing place names and core locations mentioned in the text, the dominant present-day surface water circulation in the Nordic and Barents Seas. The location of the studied core JM09-020 and the locations of the cores discussed in this paper are marked with dots: MD95-2011 (Calvo et al., 2002), PSh-5159N (Risebrobakken et al., 2010), M23258-2 (Martrat et al., 2003), and JM09-KA11 (Belt et al., 2015). Conductivity-temperature-depth (CTD) stations (A, B, and C) are marked with a black X. Red arrows represent Atlantic Water (AW), blue arrows represent Arctic Water (ArW), and royal blue dashed lines represent the Arctic Front (AF). The remaining abbreviations are as follows: NAC - North Atlantic Current, NCC- North Cape Current, WSC - West Spitsbergen Current, and ESC - East Spitsbergen Current. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Table 1 Positions of the cores and references to the original publications discussed in this paper. Core Location Reference Latitude (N) Longitude (E) Water depth (m) JM09-020-GC Storfjordrenna This study 76.31 19.70 253 MD95-2011 NE Norwegian Sea Calvo et al. (2002) 66.97 7.63 1048 PSh-5159N SW Barents Sea Risebrobakken et al. (2010) 71.36 22.65 422 M23258 W Barents Sea Martrat et al. (2003) 74.99 13.97 1768 4 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 Fig. 2. Temperature and salinity profiles from A) the continental slope at the mouth of Storfjordrenna measured in August 2012 (1300-m water depth) (selected from the World Ocean Database https://www.nodc.noaa.gov/OC5/SELECT/dbsearch/dbsearch.html); B) Storfjordrenna (JM09-020 GC) measured in August 2015 at the inner part of the trough (253- m water depth; this study); C) the Kvitøya Trough (NP05-49) measured in August 2005 (300-m water depth) (Klitgaard-Kristensen et al., 2013). The colors on the conductivity- temperature-depth (CTD) profiles indicate the prevailing water masses. Abbreviations: SW - surface water, AW - Atlantic Water, ArW- Arctic Water, IW - intermediate water, and BSW - brine-enriched shelf water. The locations of the CTD stations are indicated in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Barents Sea continental slope (Fig. 1) at a depth of 1768 m, and it 1-5 cm intervals were ultrasonically extracted with dichloro- spans over 15,000 cal yr BP. Core MD 95-2011 was collected in the methane/methanol (DCM/MeOH 3:1, v/v). An internal standard (2- eastern Norwegian Sea at a water depth of 1048 m (Fig. 1) and was nonadecanone, C19H38O, 3.5 ng/mL) was added to each test tube dated to c. 13,800 cal yr BP (Calvo et al., 2002). prior to the extraction of the biomarkers. The samples were extracted three more times using the same procedure, combined 3.2. Chronology and treated with acid-activated copper to remove elemental sulfur. The total extracts were cleaned with open-column chromatography A lithological description and a chronology of core JM09-020-GC using hexane:DCM (1:1, v/v; 6 mL) and 6 mL of DCM (alkenone were published by Ła˛ cka et al. (2015b). The AMS 14C dates were fraction). converted into calibrated ages using the Marine13 calibration curve Long-chain alkenones were quantified on an Agilent 7890A gas (Reimer et al., 2013) and a DR 105 ± 24 (Mangerud et al., 2006) in chromatograph with a flame ionization detector (GC-FID). The the calibration program Calib 7.1 (Stuiver and Reimer, 1993). The compounds were eluted through an Agilent HP-1 capillary column basal age of the sediment core was 13,950 cal yr BP. The lithological with a length of 60 m, an internal diameter of 0.25 mm, and a film and micropaleontological investigation of the sediment core thickness of 0.25 mm. External standards were used to identify the revealed that the core apparently represents continuous sedi- alkenones in the GC-FID based on their retention time. The quan- mentation, with no signs of redeposition and/or lateral sediment tification of alkenones was assessed using known concentrations of transport. The uppermost c. 40 cm of sediments were lost during internal standards added prior to extraction. A solvent blank and coring; therefore, the age model for the sediment surface is cut off sediment standard from Fram Strait (Rueda, 2013) were extracted at 1300 cal yr BP (Ła˛ cka et al., 2015b). The time resolution of the with each batch of samples and analyzed in the same way as the record was between 10 and 400 years, depending on the core depth rest of the samples. This routine allowed for the control of possible with the lowest resolution identified during the mid-Holocene contamination during the preparative analysis of the samples and (Supplement 1). The previously published age models for cores helped to recognize the alkenone peaks. MD95-2011, PSh-5159N and M23258 have been recalibrated using Alkenones are used widely to calculate the UK 37 indices in the Marine13 radiocarbon calibration curve (Reimer et al., 2013) in North Atlantic region (Ła˛ cka et al., 2015a and references therein). Calib 7.1 (Stuiver and Reimer, 1993). The degree of alkenone unsaturation (the number of double bonds between the carbon atoms) provides an established way to 3.3. Alkenone analysis reconstruct past ocean conditions (Brassell et al., 1986; Herbert, 2001; Marlowe et al., 1984; Volkman et al., 1980). However, ac- The alkenone analysis methodology is consistent with the cording to Rosell-Mele (1998), when the %C37:4 value constitutes standard methodology developed by Bendle and Rosell-Mele  more than 5% of the total %C37 alkenone value and UK 37 is negative, (2004). Individual subsamples for biomarker analysis collected at the calculated temperatures are unreliable. In the case of JM09020- M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 5 GC, the values generally exceeded 5%. An alternative index to es- dropped below the detection limit (Fig. 3D). Thus, SST estimates are timate SSTs for samples when C37:4 values are significantly high was missing. At approximately 12,600 cal yr BP, Storfjordrenna was proposed by Bendle and Rosell-Mele  (2004) as follows: covered by cold (T < 1  C) surface water (Fig. 3B) with a high contribution of C37:4 (c. 58%). Between 12,500 cal yr BP and  C37:2 11,700 cal yr BP, the SSTs were slightly higher (average 2  C; Fig. 3B), UK 37 ¼ C37:2 þ C37:3 þ C37:4 and the alkenone flux remained low (average 0.3 ng cm2 a1;  Fig. 3D). After 11,700 cal yr BP, the alkenone flux increased (up to SSTs can be reconstructed using the alkenone-derived UK 37 index , 2004) according to the following equation 4 ng cm2 a1; Fig. 3D) with a simultaneous decrease in %C37:4 (Bendle and Rosell-Mele (Fig. 3A). by Müller et al. (1998): A distinct decrease in %C37:4 (from 50% to approximately 25%;  UK Fig. 3A) occurred c. 11,500 cal yr BP. This decrease was followed by a 37  0:044 Tð CÞ ¼ substantial SST increase, from an average of 2  C to an average of 0:033 8.4  C; however, SST oscillated significantly from 3  C to 12.5  C The statistical error of the regression is 1.5  C. According to (Fig. 3B). At the same time, the alkenone flux remained relatively Müller et al. (1998) and Conte et al. (2006), the global core-top high (average 1.6 ng cm2 a1; Fig. 3D). calibration of alkenone-based SST revealed that the best fit is ob- Between 9200 cal yr BP and 3400 cal yr BP, SSTs in Storfjor- tained with annual mean SSTs. The same relationship was also drenna remained highly variable (from 3  C to almost 13  C; established for polar and frontal regions (e.g., Müller and Fischer, Fig. 3B). The %C37:4 varied between 7% and 38% (Fig. 3A), and the 2004). In the study area, alkenone producers can bloom in sum- alkenone flux was very low (between 0.08 and 0.61 ng cm2 a1). mer, as well as in autumn when stratification of the upper water After 3400 cal yr BP, the SSTs in Storfjordrenna ranged between column breaks down, i.e., the sedimentary signal would potentially 0.2  C and 10  C (Fig. 3B), with a rapid, concomitant increase in the integrate a wider range of temperatures than just those from the alkenone flux (from 0.5 to 3 ng cm2 a1). Towards the upper part warmest season. of the core, the alkenone flux increased even more, reaching 12 ng To facilitate the comparison of our results with the published cm2 a1 c. 1500 cal yr BP (Fig. 3D). At the same time, the %C37:4 alkenone SST records in Calvo et al. (2002) and Martrat et al. (2003), varied between 9% and 25%; (Fig. 3A). we have recalculated the UK 37 -based SSTs according to the Bendle  and Rosell-Mele  (2004) UK 37 model. The SST data after 5. Discussion Risebrobakken et al. (2010) are presented in the original form, as the raw data are unavailable. 5.1. Bølling-Allerød (13,950e12,800 cal yr BP) The total abundance of C37:4 alkenones in the record was used as a tracer to infer variations in the water mass type as follows (Bendle Grounded ice retreated from Storfjordrenna during the B-A et al., 2005): warming, c. 13,950 cal yr BP (Ła˛ cka et al., 2015b), as a consequence of the overall warming caused by the increase in the northern C37:4 %C37:4 ¼ x100 hemisphere insolation (Berger and Loutre, 1991). The SSTs at the C37:2 þ C37:3 þ C37:4 study site during the B-A varied between 2  C and 4  C (Fig. 3B), In the subpolar and polar regions of the Nordic Seas, there is an which is comparable to modern SST values in Storfjordrenna increased proportion of the C37:4 alkenone relative to the C37:3 and (Fig. 2B). The modern-like conditions have also been noted on the C37:2 alkenones (Rosell-Mele  et al., 1994, Rosell-Mele, 1998). High % Barents Sea continental slope, where SST reached 11  C during the C37:4 values are associated with ArW, whereas low values are B-A (Martrat et al., 2003, Fig. 4A), which is comparable to the related to AW (Bendle et al., 2005). Hence, %C37:4 can also be used as modern summer SST in this region (Trudnowska et al., 2016). an indicator of the AF position in the Nordic Seas (Rosell-Mele et al., Furthermore, the low PIIIIP25 (sea ice proxy) in the Kveithola 1998). Trough, south of Storfjordrenna (see the location of the JM09-KA11 The absolute quantification of alkenones (C37:2þC37:3þC37:4) was core in Fig. 1), indicate stable ice edge or marginal ice zone condi- assessed using known concentrations of internal standards added tions at that time (Fig. 4D) (Belt et al., 2015). The sea ice conditions prior to extraction and expressed as ƩC37 ng g1 of dry sediment. in Storfjordrenna during the B-A could be driven by prevailing The estimated level of detection was 3 ng g1, using an average southwesterly winds. The modern sea ice occurrence in Storfjor- sediment sample of 1 g dry weight. Concentration versus depth drenna arises from seasonal ice advected from the Arctic Ocean and profiles can be potentially misleading with respect to quantitative the Barents Sea (Hendricks et al., 2011) and is strictly connected to interpretation of alkenone data, since they do not adequately the prevailing wind direction. According to Skogseth et al. (2004), reflect the influences of any changes in bulk sediment properties or the amount of sea ice in the western Barents Sea is high during accumulation rates. Therefore, alkenone concentrations were con- winters dominated by northeasterly winds and when temperatures verted to annual fluxes, by combining individual sediment con- are low, whereas in winters dominated by southwesterly winds, the centrations with sediment densities and accumulation rates inflow of the AW increases SSTs, and the sea ice decreases. derived from dry bulk density data and age/depth models, Although the B-A SSTs were comparable to contemporary respectively. values, the alkenone concentration and flux at the study site were low (Fig. 3D), with a high contribution of C37:4 (between 13% and 79%; Fig. 3A). This inverse relationship between ƩC37 and C37:4 in- 4. Results dicates that fewer alkenones were produced in the fresher ArW. High %C37:4 values have been found in modern surface sediments in The SSTs in the lowermost part of the core (13,950e12,800 cal yr regions where surface temperatures and salinity are both low BP) varied between 0.5  C and 5.5  C (Fig. 3B). Both the alkenone (Bendle and Rosell-Mele , 2004; Bendle et al., 2005; Harada et al., flux and concentration (ƩC37) were low (0.3e1.4 ng cm2 a1 and 7- 2006), e.g., high %C37:4 values (up to 70%) have been noted in the 32 ng g1 sed, respectively; Fig. 3D), with a high contribution of waters of the East Greenland Current, where the sea ice cover C37:4 (between 13% and 79%; Fig. 3A). reached approximately 80% (Bendle et al., 2005). Phytoplankton At approximately 12,800 cal yr BP, the alkenone concentrations productivity is much lower in these ArWs than in AWs (Andreassen 6 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 Fig. 3. Proxy records from core JM09-020 GC A) %C37:4; B) UK 37 -based sea surface temperatures (SSTs; right scale) and stable oxygen isotope data (red line; left scale; Ła ˛ cka et al., 2015b); C) stable carbon isotope data (burgundy line; left scale; Ła˛ cka et al., 2015b) and mass accumulation rates (g cm2 a1; violet shading; right scale; Ła˛ cka et al., 2015b); D) 1 2 1 total alkenone concentrations (ƩC37; ng g ; black line; left scale) and total alkenone flux (ng cm a ; green shading; right scale). The black triangles on the x axis denote the AMS 14 C converted to calibrated radiocarbon ages (after Ła˛cka et al., 2015b). The Arctic Front transition at 11,500 cal yr BP is marked. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) et al., 1996). However, the low concentration of alkenones could today. This difference was caused by the proximity of the retreating also result from their degradation and/or dilution by the high Svalbard Barents Sea Ice Sheet (SBSIS) front, which was located in sediment supply (e.g., Hoefs et al., 1998). A high sediment accu- the inner basin of Storfjorden at that time (e.g., Rasmussen and mulation rate in Storfjordrenna (43 g cm2kyr1) during this in- Thomsen, 2014) and released large amounts of turbid meltwater terval (Fig. 3C; Ła˛ cka et al., 2015b) supports the latter (Ła˛ cka et al., 2015b), limiting primary productivity (Fig. 4C). This interpretation. process is observed today in glaciated fjords on Svalbard, where, Much lower alkenone concentrations during the B-A were also despite relatively high SSTs, productivity remains low (Kubiszyn noted on the Barents Sea continental slope (Martrat et al., 2003), et al., 2014; Piwosz et al., 2009; Zaja˛ czkowski, 2008). We indicating that the AF was located further to the west than it is conclude that the oceanographic conditions in Storfjordrenna M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 7 1 Fig. 4. Comparison of A) UK 37 -based sea surface temperatures (SSTs; right scale); B) %C37:4; C) total alkenone concentrations (ƩC37; ng g ) from the current study (JM09-020 GC; black line), western Barents Sea (Martrat et al., 2003) (M23258; green line), the southwestern Barents Sea (Risebrobakken et al., 2010) (PSh-5159N; purple line), and the north- eastern Norwegian Sea (Calvo et al., 2002) (MD95-2011; blue line); and D) summer insolation (June) at 80 N (Berger and Loutre, 1991; black line, left scale) and IP25 (orange line, right scale), and PIIIIP25 (purple line, right scale) records from the core JM09-KA11 obtained from the western Barents Sea (after Belt et al., 2015) showing the sea ice conditions. The black triangles on the x axis denote the AMS 14C converted to calibrated radiocarbon ages (after Ła˛ cka et al., 2015b). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 8 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 during the B-A were similar to modern conditions in the innermost therein). Nevertheless, AW influenced numerous locations in the parts of the glaciated fjords of western Svalbard. North Atlantic region throughout the YD (Bartels et al., 2017; Ła˛ cka et al., 2015b; Pearce et al., 2013; Rasmussen et al., 2007), leading to 5.2. Younger Dryas (12,800e11,700 cal yr BP) variable sea ice conditions in the Nordic Seas during the latter part of YD (Bakke et al., 2009; Cabedo-Sanz et al., 2013). Our data The onset of the YD (12,800 cal yr BP) in Storfjordrenna was indicate that heavy sea ice conditions prevailed in Storfjordrenna characterized by low alkenone concentrations, occasionally only at the very beginning of the YD. In the later part of YD, the remaining below the detection limit (Fig. 3D). This extremely low oceanography of Storfjordrenna was modified by interactions be- alkenone signal indicates a reduction in primary productivity that tween Arctic and Atlantic waters, as indicated by the variability in persisted for several decades. Similarly, low or absent primary SST (Fig. 3B) and %C37:4 (Fig. 3A). The similar pattern in alkenone production is observed beneath modern pack ice in the Arctic record has been also observed on the continental slope of the Ocean (Antoniades et al., 2011; Darby et al., 2006), suggesting that Barents Sea, where after 12,200 cal yr BP ArW contribution sea ice covered the area. Another explanation for the low or lack of decreased concomitant with the productivity increase (Fig. 4 B and alkenone signal might be postdepositional degradation. However, C). Contradictory, the PIIIIP25 signal from Kveithola Trough pointed Ła˛ cka et al. (2015b) noticed, based on the absence of ice-rafted to the continuous presence of consistently long seasonal sea ice debris in this interval, that the beginning of YD in Storfjordrenna cover (Belt et al., 2015). was characterized by temporary polar conditions and the formation of perennial pack ice in Storfjorden that locked icebergs proximal to 5.3. Early Holocene (11,700-9200 cal yr BP) their calving fronts, thus preventing their movement over the coring site. Therefore, we suggest that Storfjordrenna was covered At the beginning of the early Holocene (11,700 cal yr BP), the by perennial sea ice at the onset of the YD, limiting light penetra- alkenone concentrations in Storfjordrenna increased and the %C37:4 tion into the surface water and subsequently restricting phyto- decreased, indicating the increasing influence of AW in the trough, plankton growth. a reduction in sea ice, as well as an increase in marine productivity At c. 12,600 cal yr BP, Storfjordrenna was covered by cold (Fig. 3C). A similar oceanographic change was noted in the south- (T < 1  C) surface water (Fig. 3B) of Arctic origin, as seen from the western Barents Sea by Risebrobakken et al. (2010) and in the high values of %C37:4 (Fig. 3A). The surface productivity at the study Norwegian Sea by Calvo et al. (2002). In the record after Martrat site increased slightly (Fig. 3D). However, Ła˛ cka et al. (2015b) noted et al. (2003), further decrease in %C37:4 and increase in productiv- a lighter d18O signal from benthic foraminifera, indicating the ity was observed. The transition from prolonged seasonal ice cover presence of warmer waters at the bottom (Fig. 3B). The nearly to ice-edge conditions is also supported by the PIIIIP25 signal from continuous presence of warmer and more saline AW during the the western Barents Sea, which indicates that sea ice cover at that beginning of the YD was also noted in the subsurface waters of the time was variable (Belt et al., 2015). southern Nordic Seas by Knudsen et al. (2004) and Rasmussen et al. Approximately 11,500 cal yr BP, a rapid transition from an Arctic (2011). The difference between cold, less saline surface water and to an Atlantic water-dominated environment occurred in Storfjor- warmer but high saline bottom water indicates stratification, which drenna, suggesting that AF passed our study site and moved east- was essential for the presence of sea ice, since the pycnocline ward. This is illustrated by a further decrease in %C37:4 (from 50% to protected the cold sea surface from the heat stored in the AW approximately 25%; Fig. 3A) and a substantial SST increase, from an below. According to Rasmussen et al. (2014), the pycnocline over average of 2  C prevailing during the YD to 3e12.5  C at the the western Spitsbergen shelf occurred at depths between 100 m beginning of the Holocene (Fig. 3B). Today, such high SSTs (average and 150 m at this time, because of the freshwater supply from the 8.4  C; Fig. 3B) are observed in the core of the West Spitsbergen decaying SBSIS. Our data may indicate that the oceanographic Current flowing west of Spitsbergen (e.g., Trudnowska et al., 2016); conditions of Storfjordrenna at this time were similar to modern however, they are less variable than the SSTs during the early Ho- conditions in the northern Barents Sea (Fig. 1), with strong water locene. A significant increase of SSTs shows AW appearance in the column stratification, prolonged seasonal sea ice cover (Fig. 2C) and upper part of the water column. However, the high amplitudes in lower marine productivity (Fig. 3D). However, as the lighter d18O SST prevailing throughout the entire Holocene indicate the can also point to the changes in bottom water salinity, more studies continuous proximity of the AF to our study site (Fig. 3B). The SST are needed to confirm our hypothesis. Nevertheless, the severe variability presented in Fig. 3B corresponds to the pulsatory inflow seasonal sea ice cover throughout the YD also occurred 150 km of ArW shown in Fig. 3A. The location of the AF closer to the south of our coring site, in the Kveithola Trough, as indicated by Spitsbergen coast at approximately 11,500 cal yr BP is also high PIIIIP25 values (Belt et al., 2015) (Fig. 4D). Further west, the confirmed by the increase in productivity in Storfjordrenna (higher continental margin remained ice free; however, the SST decreased alkenone flux and lighter d13C; Fig. 3C and D). The presence of a to 7  C, and the alkenone production also decreased significantly highly productive frontal zone at our study site is also visible in the (Martrat et al., 2003, Fig. 4C). higher abundance of the AF proximity proxy, i.e., the benthic Between 12,500 cal yr BP and 11,700 cal yr BP, the SST in Storf- foraminifera Nonionellina labradorica, in Storfjordrenna (Ła˛ cka jordrenna increased (average 2  C); however, surface productivity et al., 2015b). The timing of the AF transition in Storfjordrenna remained low (Fig. 3B and D). This might have resulted from the was the same as that on the Barents Sea continental slope (Martrat continuous delivery of turbid waters from the decaying SBSIS et al., 2003) and that farther south in the eastern Norwegian Sea (Ła˛ cka et al., 2015b), which strongly limited light penetration and (Calvo et al., 2002) (Fig. 4B), where %C37:4 decreased to modern consequently led to decreased primary productivity and/or alke- values characteristic of the AW domain, i.e., 5% (Rosell-Mele et al., none signal dilution (compare with Hoefs et al., 1998). At the same 1998). Previously, Ła˛ cka et al. (2015b) based on d18O concluded time, the mineral particles in the surface waters absorbed an that the transition from the Arctic to the Atlantic domain occurred amount of sunlight energy (e.g., Kara et al., 2005) that was close to at approximately 9600 cal yr BP (Fig. 3B), because the lighter d18O is the maximum Holocene values (Fig. 4D), thereby elevating the SST characterized for the ArW of Barents Sea origin (Duplessy et al., in Storfjordrenna. A widespread hypothesis regarding the mecha- 2005). Indeed, the d18O values are still shifted towards the lighter nism of the YD cold spell suggests that the YD cooling resulted from values, but this could be also caused by the continuous production a slowdown in the AMOC (e.g., Ritz et al. (2013) and references of brine in Storfjordrenna. Although the AW dominated at the M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 9 surface, sea ice formed during the polar night, and lower d18O brine surface productivity in the Norwegian and Svalbard shelf areas was delivered to the bottom of Storfjordrenna (Fig. 3B). (Fig. 4C). The low productivity in the region is also confirmed by Early Holocene SSTs in Storfjordrenna were variable and reached high d13C and the low benthic foraminifera flux in Storfjordrenna the Holocene temperature maximum (Fig. 3B). Because the alke- (Fig. 3C) (Ła˛ cka et al., 2015b), as well as low benthic foraminifera none record is susceptible to temperature changes related to orbital flux further south, in Kveithola Trough (Groot et al., 2014). Two forcing (Risebrobakken et al., 2011), we suggest that the SST basic factors regulate primary productivity in the ocean: sunlight changes observed during the early Holocene in Storfjordrenna and and nutrient availability. During the polar day, insolation is high, on the adjacent continental shelf resulted from both AW inflow and but light penetration throughout the water column can be limited a peak in the northern hemisphere summer insolation (Berger and by water turbidity. According to Ła˛ cka et al. (2015b), the sediment Loutre, 1991) (Fig. 4D). The presence of warm AW was found in the flux was low (approximately 0.019e0.002 g cm2 yr1; Fig. 3C) in early Holocene at the northern tip of Svalbard (Bartels et al., 2017) Storfjordrenna during the mid-Holocene because of the reduced and even farther in the northwestern Barents Sea (Ivanova et al., glaciers on Svalbard. Therefore, we suggest that the most important 2019). The appearance of AW at the sea surface in Storfjordrenna factor limiting primary productivity in the study area was nutrient suggests that the contribution of turbid meltwater diminished, as availability. According to Behrenfeld et al. (2006), reduced nutrient confirmed by the reduced sediment accumulation rate at the core flux into the upper oceans can be caused by enhanced thermal site during the early Holocene (Fig. 3C; Ła˛ cka et al., 2015b). Indeed, vertical stratification. The high SST in Storfjordrenna that prevailed according to Forwick and Vorren (2009) and Hughes et al. (2016), approximately 6400 cal yr BP (Fig. 3A) most likely limited the sur- after the YD-early Holocene transition, the final deglaciation of face cooling during the winter and consequently inhibited Svalbard and Scandinavia occurred, and after 10,000 cal yr BP, the convective water mixing and the return of nutrients to the euphotic glaciers in western Svalbard were even smaller than those of the zone. Wollenburg et al. (2004) found that productivity increased in present day. the northern Barents Sea (at latitude 81 N) during the mid- Holocene. The mid-Holocene productivity maximum (seen as 5.4. Mid-Holocene (9200-3400 cal yr BP) maximum planktic and benthic foraminifera abundances) was also  noted by Slubowska et al. (2005) in a sediment core obtained from SSTs in Storfjordrenna varied significantly during the mid- the northern Svalbard continental margin at 80 N. Thus, we sug- Holocene (Fig. 3B), although in the western Barents Sea continen- gest that the high productivity zone at that time shifted northward tal margin (Martrat et al., 2003) and in the Norwegian Sea (Calvo to the region where Arctic pack ice appeared seasonally, causing sea et al., 2002) they remained stable (10  C and 11.5  C, respectively) surface cooling in winter and inducing convective water mixing. (Fig. 4A). Over the European continental slope, warm AW domi- Our suggestion is supported by the northward and eastward shift of nated at the surface during this period, as confirmed by the absence the ice edge and phytoplankton blooms in the modern Barents Sea or low values of C37:4 alkenone (Fig. 4B; Calvo et al., 2002; Martrat observed over the last 17 years (Oziel et al., 2017). et al., 2003). Conversely, high %C37:4 (10-20%) values were found in An alternative explanation of the low primary productivity in Storfjordrenna sediments (Fig. 3A; this study) and in the south- Storfjordrenna could be the earlier light signal in the spring in the western Barents Sea (Risebrobakken et al., 2010), which could ice-free water and the earlier appearance of mesozooplanktonic suggest a continuous ArW contribution (Fig. 4B) at those sites. A organisms in the sea surface. According to Zaja˛ czkowski et al. feasible explanation of the SST variability on the western shelf of (2010), in the ice-free Adventfjorden (western Spitsbergen), the the Barents Sea could be the alternating influence of the ESC with open-water spring period enhanced primary productivity. How- the NAC. ever, the majority of the particulate organic matter was then At approximately 6400 cal yr BP, the SST in Storfjordrenna consumed by zooplankton, and the organic matter sedimentation reached a peak of almost 13  C (Fig. 3B) and the %C37:4 decreased was much lower in this period than in the ice-covered years. (Fig. 3A), indicating significant AW inflow. Based on benthic fora- Additional investigations are needed to verify these two scenarios. minifera assemblages, especially the occurrence of Melonis bar- leeanum, Ła˛ cka et al. (2015b) suggested that around that time, the 5.5. Late Holocene (3400-1300 cal yr BP) Storfjordrenna sea environment was similar to that of the contemporary Norwegian fjords, which are dominated by AW with After 3400 cal yr BP, the SSTs in Storfjordrenna decreased and temperatures of 6e8  C. Our new alkenone data confirm this ranged from 10  C to approximately 0  C (Fig. 3A). Total alkenone finding. Additionally, the finding is supported by the occurrence of flux increased rapidly (Fig. 3D). Throughout the entire late Holo- Mytilus edulis on the western and eastern coasts of Svalbard at this cene, further gradual surface water cooling occurred, reflecting time (Salvigsen, 2002; Mangerud and Svendsen, 2018), as M. edulis increases in alkenone flux. A similar alkenone signal pattern was is a thermophilous mollusk spawning at temperatures above 8- detected in both the southwestern Barents Sea (Risebrobakken 10  C (Thorarinsdo ttir and Gunnarsson, 2003). The mid-Holocene is et al., 2010) and the Norwegian Sea (Calvo et al., 2002) (Fig. 4A often referred to as the Holocene Thermal Maximum (HTM) in and C). In general, this cooling trend corresponded to further de- marine and terrestrial records (Risebrobakken et al., 2011). The high creases in northern hemisphere insolation (Fig. 4D; Berger and SST predominated throughout the European Arctic, reducing sea ice Loutre, 1991). We suggest that the cooler surface of Storfjor- formation up to the northern margins of the Svalbard shelf (Müller drenna froze periodically at that time and produced brine, which et al., 2012) and in the central Arctic Ocean (Polyak et al., 2010). launched convective water mixing and increased nutrient resupply However, the SST maximum occurred when the summer insolation to the sea surface. Consequently, primary production was enhanced had already started to decrease, after a maximum peak during the in the area, as is supported by other productivity proxies such as early Holocene (Berger and Loutre, 1991, Fig. 4D). This indicates higher benthic foraminifera flux (Ła˛ cka et al., 2015b). Simulta- that, at this time, the main heat source was the enhanced inflow of neously, diminished surface water productivity was noted in the AW, and insolation played a secondary role in the heating of inner Storfjorden, stemming from dense, packed sea ice cover Storfjordrenna surface waters. (Knies et al., 2017). South of Storfjordrenna, in the Kveithola During the mid-Holocene, the alkenone flux in Storfjordrenna Trough, PIIIIP25 was mainly absent throughout the late Holocene, was very low (Fig. 3D). In the other records discussed here, the reflecting the predominantly ice-free ocean conditions (Belt et al., alkenone concentration also decreased, which suggests low sea 2015) (Fig. 4D). According to Risebrobakken and Berben (2018), 10 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 the AF in Kveithola Trough was shifted towards inner part of the causing SST variability. trough at that time. We suggest that the difference in the location of The early Holocene in Storfjordrenna (11,700e9200 cal yr BP) the AF was caused by other local environmental conditions. was characterized by a transition from an Arctic to an Atlantic Late Holocene near-bottom conditions in Storfjordrenna were domain. The alkenone records can be interpreted to show that at characterized by variable oceanographic conditions and higher approximately 11,500 cal yr BP, the AF passed the study site and salinity and temperature gradients because of the proximity of the moved eastward. Early Holocene warming was driven by the AF that resulted from the reduced inflow of AW (Ła˛ cka et al., increasing northern hemisphere insolation and final decay of the 2015b). This finding is supported by our new data on SSTs and SBSIS, leading to a shift from ice-sheet proximal to ice-sheet distal other alkenone records published by Calvo et al. (2002) and Martrat conditions at the study site. The high SSTs between 6400 and et al. (2003) (Fig. 4A). SSTs decreased (c. 1-2  C), and marine pro- 3400 cal yr BP maintained water stratification in winter as well as ductivity increased. Moreover, an increased %C37:4 (Fig. 3A) in- inhibited convective water mixing and the return of nutrients to dicates a greater contribution of ArW, suggesting continuous sea ice the euphotic zone and/or enhanced organic matter consumption by formation inducing the convective replenishing of nutrients from zooplankton, due to earlier light signals in the ice-free Storfjor- the bottom to the surface. drenna. The high productivity zone at this time shifted from Storfjordrenna to the northern edge of the Eurasian shelf, where 6. Summary and conclusions Arctic pack ice enabled sea surface cooling in winter and induced convective water mixing. During late the Holocene (3400- Alkenone analyses in sediment core JM-09-020GC from Storf- 1300 cal yr BP), low insolation facilitated sea ice formation and jordrenna were performed with the purpose of reconstructing SSTs, brine production, which launched convective water mixing and the relative influences of AW and ArW, the location of the AF, as increased nutrient resupply to the sea surface, consequently well as the marine primary productivity in the western Barents Sea enhancing the primary productivity in Storfjordrenna. between approximately 13,950 cal yr BP and 1300 cal yr BP. The Based on past changes in Arctic oceanography, combined with proposed sequence of climatic and oceanographic events in the observations in the modern Barents Sea (e.g., Oziel et al., 2017), we time period studied are as follows: suggest that the increasing inflow of warm AW and the disap- The SSTs in Storfjordrenna during the B-A warming were com- pearance of pack ice on the Eurasian continental shelf may weaken parable to modern SST values. They were probably the result of convective water mixing during the polar night and may limit increased insolation and the reinvigoration of the AMOC. However, spring/summer primary production in the region. we argue that the constant delivery of turbid meltwater from the Svalbard glaciers caused a significant decrease in marine produc- Acknowledgments tivity, unlike the present day situation in the core site location. At the onset of the YD (12,800 cal yr BP), Storfjordrenna was We acknowledge the National Science Centre in Poland for its covered by perennial sea ice, causing light limitations in the surface financial support through project no. 2016/21/B/ST10/02308 water and a subsequent reduction in phytoplankton growth. At (alkenone analysis) and 2012/05/N/ST10/03696 (other analyses) approximately 12,600 cal yr BP, Storfjordrenna was covered by cold and the Spanish research Ministry Maria de Maeztu award MDM- surface water, which we proposed caused water column stratifi- 2015-0552. We also extend our thanks to the captain and crew of cation and was essential for sea ice formation. Thus, a marked the R/V Jan Mayen and the cruise participants, in particular Steinar pycnocline could have isolated the cold sea surface from the heat Iversen, for their help at sea. Eva Calvo and Belen Martrat are stored in the AW below. After 12,500 cal yr BP, the surface condi- gratefully acknowledged for providing their alkenone data. We are tions in Storfjordrenna gradually ameliorated, and AW appeared at  ska for data processing. We also very grateful to Agnieszka Promin the surface, leading to a transition from the long seasonal ice cover acknowledge two anonymous reviewers for their constructive to sea ice edge conditions. Although the YD is regarded as a stadial comments on the manuscript. period, the present study shows that heavy sea ice conditions prevailed in Storfjordrenna only at the onset of the YD. In the Appendix A middle part of the YD, the oceanography of Storfjordrenna was modified by interactions between Arctic and Atlantic waters, Depth [cm] Age BP C37:4 C37:3 C37:2 %C37:4 UK37* SST UK37* Muller 5 1290.71421 88.41 366.76 144.11 14.75270324 0.2404719 5.953693928 9 1332.42654 13.49 60.78 29.16 13.04263753 0.281929808 7.20999417 13 1401.94709 82.87 235.51 17.4 24.6798499 0.051819644 0.236958903 17 1457.56353 20.75 101.29 56.19 11.64226 0.31526679 8.220205761 21 1513.17997 99.95 340.12 190.77 15.84395409 0.302406315 7.830494406 25 1554.8923 82.34 317.44 96.96 16.57607602 0.195192656 4.58159564 29 1624.41285 88.28 441.85 210.32 11.92247957 0.284043487 7.274045063 33 1680.02929 9.87 61.69 34.47 9.308686221 0.325096671 8.518080932 37 1721.74162 39.06 295.7 65 9.770862518 0.162597559 3.59386541 41 1791.26217 100.87 443.2 130.81 14.94636083 0.193827051 4.540213659 45 1846.87861 62.17 217.28 92.86 16.69845022 0.249415809 6.224721497 53 1958.11149 44.84 188.21 32.79 16.86728859 0.123344869 2.404389973 57 2013.72793 53.22 238.87 118.52 12.96120406 0.288643725 7.413446218 61 2069.34437 55.28 247.76 59.69 15.23998566 0.16455766 3.653262423 65 2124.96081 26.49 53.8 39.8 22.05845616 0.331418103 8.709639488 69 2180.57725 47.97 230.31 115.14 12.1930761 0.292664328 7.535282673 73 2236.19369 55.93 246.19 121.68 13.19726286 0.287116564 7.367168619 77 2291.81013 6.02 29.57 21.68 10.51161166 0.378557709 10.1381124 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 11 (continued ) Depth [cm] Age BP C37:4 C37:3 C37:2 %C37:4 UK37* SST UK37* Muller 81 2391.42853 33.15 51.14 50.15 24.65783993 0.37302886 9.970571529 85 2505.71425 51.79 258.85 81.99 13.19053562 0.208822556 4.994622897 89 2619.99997 71.56 300.61 138.88 14.00254378 0.271754231 6.901643378 93 2734.28569 46.66 198.48 109.15 13.17000198 0.308080951 8.00245305 97 2848.57141 79.58 274.74 75.11 18.53154181 0.174906271 3.9668567 105 3077.14285 60.72 266.53 126.31 13.38742394 0.278485757 7.105629004 109 3191.42857 64.84 286.15 133.91 13.37182924 0.276160033 7.035152515 113 3571.5625 24.37 152.48 75.53 9.656074174 0.299270941 7.73548305 117 4040.3125 31.59 59.93 41.02 23.83431417 0.309491474 8.04519619 121 4509.0625 23.52 102.36 40.52 14.13461538 0.243509615 6.045745921 125 4977.8125 14.75 64.03 13.61 15.96493127 0.147310315 3.130615605 129 5446.5625 9.56 70.14 44.27 7.711543115 0.357102525 9.487955297 130 5563.75 28.4 143.78 81.35 11.20183016 0.320869325 8.389979549 131 5680.9375 24.36 13.69 26.06 37.99719233 0.406488847 10.98451052 132 5798.125 33.54 104.93 66.67 16.34980989 0.324997563 8.515077656 133 5798.125 23.75 119.52 73.26 10.96845703 0.338336489 8.919287551 135 6149.6875 18.71 98.34 57.35 10.72821101 0.328841743 8.631567973 137 6384.0625 11.37 39.68 44.77 11.86599875 0.467230223 12.82515828 141 6852.8125 18.94 72.48 15.27 17.75236667 0.143124941 3.003786104 145 7327 19.09 46.88 30.85 19.71700062 0.318632514 8.322197392 149 7803 9.67 78.97 56.09 6.681406757 0.38754923 10.41058272 151 8041 24.16 62.69 38.5 19.27403271 0.307140008 7.973939636 152 8160 8.84 21.5 6.69 23.87253578 0.180664326 4.141343219 153 8352 7.71 61.89 39.34 7.077290252 0.361116211 9.609582144 154 8544 5.34 29.72 17.84 10.09451796 0.337240076 8.886062897 157 9120 13.62 72.37 20.01 12.8490566 0.188773585 4.38707833 161 9204.60732 13.53 60.1 35.16 12.43680485 0.32319147 8.46034757 165 9289.21464 17.33 60.67 40.34 14.64424539 0.340882204 8.996430419 169 9373.82196 4.9 51.96 35.1 5.328403654 0.38168769 10.23296031 177 9543.0366 11.8 56.78 36.61 11.21779637 0.348036886 9.213238959 181 9627.64392 3.55 12.28 7.97 14.91596639 0.33487395 8.814362108 185 9712.25124 8.77 15.63 5.46 29.37039518 0.182853315 4.207676226 193 9881.46588 16.87 121.05 73.36 7.9846649 0.347216963 9.188392826 197 9966.0732 10.33 72.89 41.59 8.276580402 0.333226504 8.764439524 201 10050.68052 16.04 108.87 82.82 7.721561643 0.398690608 10.74820024 213 10304.50248 22.87 100.44 71.16 11.76016866 0.365917622 9.755079462 217 10389.1098 27.37 89.79 64.48 15.0682669 0.354987888 9.423875398 221 10473.71712 4.78 53.62 28.55 5.497412306 0.328349626 8.61665534 229 10642.93176 10.49 80.52 15.78 9.823017136 0.147766645 3.144443782 237 10812.1464 4.73 27.75 27.78 7.849319615 0.461002323 12.63643404 245 10981.36104 8.2 13.72 8.59 26.87643396 0.281547034 7.198394962 256 11136.36364 5.917 48.63 36.19 6.521044337 0.398845014 10.7528792 273 11290.90909 5.29 27.6 12.7 11.6034218 0.278569862 7.108177631 274 11300 19.82 51.16 15.98 22.79208832 0.183762649 4.235231803 275 11425 3.32 8.56 1.15 25.47966232 0.088257866 1.341147469 277 11675 6.38 9.21 3.42 33.56128353 0.179905313 4.118342818 279 11925 11.61 10.64 1.72 48.43554443 0.071756362 0.841101882 281 12175 2.46 1.59 1.03 48.42519685 0.202755906 4.810785016 283 12425 6.64 7.61 2.44 39.78430198 0.146195327 3.096828077 285 12577.08333 3.27 2.32 0 58.49731664 0 1.333333333 287 12631.25 2.96 2.52 0 54.01459854 0 1.333333333 291 12739.58333 292 12766.66667 296 12875 4.67 8.75 1.64 31.00929615 0.108897742 1.966598253 300 12983.33333 47.05 8.14 3.82 79.73224877 0.064734791 0.628326991 301 13010.41667 22.03 8.86 3.75 63.59699769 0.108256351 1.947162153 303 13064.58333 20.88 22.95 6.44 41.53570718 0.128108216 2.548733807 306 13145.83333 7.23 37.2 9.75 13.34440753 0.179955703 4.119869794 309 13227.08333 8.3 12.06 3.29 35.09513742 0.139112051 2.882183356 311 13281.25 10.51 16.2 3.49 34.8013245 0.115562914 2.168573149 313 13335.41667 2.28 4.68 1.1 28.28784119 0.136476427 2.802315964 315 13389.58333 10 17.62 6.33 29.455081 0.186450663 4.31668675 317 13443.75 5.2 18.13 5.73 17.89401239 0.197178252 4.641765209 319 13497.91667 8.25 11.44 2.73 36.79750223 0.12176628 2.356553943 320 13525 2.93 6.16 2.61 25.04273504 0.223076923 5.426573427 321 13552.08333 6.92 5.18 1.58 50.58479532 0.115497076 2.166578061 328 13741.66667 12.35 17.53 5.74 34.67153285 0.161145424 3.549861331 333 13862.5 2.64 2.81 1.4 38.54014599 0.204379562 4.859986729 349 13966.07143 3.64 5.42 2.25 32.18390805 0.198938992 4.695120971 353 13973.21429 25.66 26.85 9.39 41.453958 0.151696284 3.263523768 361 13987.5 41.34 45.06 21.96 38.15060908 0.202657807 4.807812343 12 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 Appendix B Depth [cm] Age BP Concentration [ng/g] Flux C37:4 C37:3 C37:2 C37SUM 5 1290.71421 53.04666856 220.0587734 86.75531339 359.8607554 0.399911 13 1401.94709 65.03106385 184.8131513 13.65440462 263.4986198 6.424431 17 1457.56353 27.87734803 136.081763 75.23887993 239.1979909 5.831951 21 1513.17997 77.77767941 264.6697781 147.9558687 490.4033262 11.95666 25 1554.8923 49.34041767 190.2188752 58.48846907 298.0477619 7.266783 29 1624.41285 57.05254478 285.5535445 135.9230994 478.5291887 11.66715 37 1721.74162 29.72679445 225.0438587 49.79834231 304.5689954 7.425779 41 1791.26217 66.72572082 293.1777483 86.53109488 446.434564 10.88464 45 1846.87861 42.43731238 148.315574 64.02020454 254.7730909 6.211692 53 1958.11149 40.03573422 168.0447265 29.17921096 237.2596717 5.784692 57 2013.72793 33.97158726 152.476382 75.90630523 262.3542745 6.396531 61 2069.34437 35.28689835 158.1527123 38.35594782 231.7955585 5.65147 65 2124.96081 15.01101939 30.48670605 22.4781846 67.97591004 1.657339 69 2180.57725 37.40676185 179.5945658 90.38415654 307.3854842 7.494449 73 2236.19369 40.80960124 179.6337516 88.78441406 309.2277669 7.539366 81 2391.42853 18.24908823 28.15259041 27.60759501 74.00927365 0.765834 85 2505.71425 38.52618225 192.556522 61.39834127 292.4810456 3.026539 89 2619.99997 43.44250786 182.4937435 84.02996795 309.9662193 3.207472 93 2734.28569 54.29214597 230.945245 127.4269401 412.6643311 4.270173 97 2848.57141 56.63742218 195.5336186 53.81247833 305.9835191 3.16626 105 3077.14285 39.27497179 172.3972041 81.4276287 293.0998046 3.032942 109 3191.42857 40.58523168 179.1095627 82.9799613 302.6747557 3.132021 113 3571.5625 16.59153093 103.8111053 51.25076475 171.653401 0.528979 117 4040.3125 18.40218485 34.91114081 24.05476438 77.36809004 0.238423 121 4509.0625 30.79533865 134.0225708 35.72294199 200.5408514 0.618 125 4977.8125 10.20589518 44.30396395 9.448490902 63.95835002 0.197098 129 5446.5625 7.884813505 57.84945808 36.39091613 102.1251877 0.314716 130 5563.75 14.18922277 71.8354384 40.77960975 126.8042709 0.390769 131 5680.9375 17.510013 9.840397289 18.6070963 45.95750659 0.141626 132 5798.125 20.88423652 65.33640245 41.78993428 128.0105733 0.394486 133 5798.125 17.31400309 87.13135364 53.58534508 158.0307018 0.486998 135 6149.6875 11.4440418 60.15003049 34.84448736 106.4385597 0.328008 137 6384.0625 12.47831699 43.54789957 32.91981958 88.94603614 0.274102 139 6618.4375 11.19929201 53.22689577 13.86206074 78.28824851 0.241258 141 6852.8125 12.94514574 49.53876259 10.5063459 72.99025423 0.224932 145 7327 12.12583424 29.77784752 19.66102285 61.56470461 0.199847 149 7803 6.97993157 57.00157147 40.48649036 104.4679934 0.339117 151 8041 13.85355565 35.94699518 22.29699964 72.09755048 0.158642 152 8160 8.842603666 21.50633245 6.736583558 37.08551967 0.081602 153 8352 4.528609886 36.35222644 23.03004685 63.91088317 0.140628 157 9120 8.599867886 45.6954801 12.46614623 66.76149421 0.1469 161 9204.60732 9.536243161 42.35980887 24.69894027 76.5949923 1.469983 165 9289.21464 9.052845678 31.69279557 21.07280985 61.8184511 1.186397 169 9373.82196 4.383258505 46.48043101 31.39844358 82.26213309 1.578744 177 9543.0366 7.996511123 38.47812725 24.97491136 71.44954974 1.371233 181 9627.64392 4.613664168 15.95937915 10.39252763 30.96557095 0.59428 185 9712.25124 6.117922304 10.90343508 3.82157377 20.84293115 0.40001 193 9881.46588 8.235218342 59.09147483 35.93061073 103.2573039 1.981676 197 9966.0732 9.854906876 69.53767301 39.94172451 119.3343044 2.29022 201 10050.68052 10.17330995 69.0503899 52.70336918 131.927069 2.531895 213 10304.50248 10.20487748 44.81757298 31.96415721 86.98660768 1.669415 217 10389.1098 14.18583956 46.53805386 33.19711054 93.92100396 1.802497 229 10642.93176 5.840208141 44.82874733 8.873219191 59.54217467 1.142711 237 10812.1464 2.973365047 17.44416069 17.34659911 37.76412485 0.724755 245 10981.36104 4.345921937 7.271469388 4.582969013 16.20036034 0.310911 256 11136.36364 5.00027305 41.09570364 20.49064065 66.58661733 2.83411 273 11290.90909 5.279225257 14.72625993 3.398749343 23.40423453 0.996149 274 11300 9.532718171 24.60614842 7.737052907 41.8759195 1.782354 277 11675 31.64814907 45.68643463 17.07809524 94.41267894 4.018463 279 11925 5.354861784 4.907470231 0.785379728 11.04771174 0.470221 281 12175 2.098636993 1.356436105 0.875768611 4.330841709 0.184333 283 12425 5.225739741 5.989138468 1.920301953 13.13518016 0.559069 285 12577.08333 3.173916974 2.251831003 0 5.425747976 0.230935 287 12631.25 1.753015669 1.492432259 0 3.245447928 0.138135 291 12739.58333 292 12766.66667 296 12875 3.132830539 5.8698645 1.092845782 10.09554082 0.429694 300 12983.33333 25.90820206 4.482311684 2.103492707 32.49400645 1.383034 301 13010.41667 12.98241677 5.22125341 2.209898452 20.41356864 0.868858 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 13 (continued ) Depth [cm] Age BP Concentration [ng/g] Flux C37:4 C37:3 C37:2 C37SUM 306 13145.83333 2.890517667 14.87237306 3.885007666 21.64789839 0.921394 309 13227.08333 4.026166618 5.850068604 1.595914238 11.47214946 0.488286 311 13281.25 5.815876481 8.964528924 1.905497317 16.68590272 0.710198 313 13335.41667 4.724536762 9.697733353 2.28697971 16.70924982 0.711192 315 13389.58333 5.361766013 9.447431714 3.393997886 18.20319561 0.774778 317 13443.75 2.581143775 8.999257045 2.84422189 14.42462271 0.613952 319 13497.91667 4.501449818 6.242010414 1.489570667 12.2330309 0.520671 320 13525 3.143706097 6.609293364 2.800366182 12.55336564 0.534306 321 13552.08333 3.693124811 2.764506722 0.851660198 7.309291731 0.311104 328 13741.66667 6.067274671 8.612091092 2.801132165 17.48049793 0.744018 349 13966.07143 2.647012014 3.941429977 1.647110498 8.235552488 0.350528 353 13973.21429 10.416686 10.89976691 3.862698776 25.17915169 1.071694 361 13987.5 7.943709193 8.658527727 4.233801001 20.83603792 0.886839 Appendix C. Supplementary data Ottersen, G., 2014. Productivity in the Barents Sea-response to recent climate variability. PLoS One 9 (5), e95273. Darby, D.A., Polyak, L., Bauch, H.A., 2006. Past glacial and interglacial conditions in Supplementary data to this article can be found online at the Arctic Ocean and marginal seasea review. Prog. Oceanogr. 71, 129e144. https://doi.org/10.1016/j.quascirev.2019.105973. Ding, Q., Schweiger, A., L'Heureux, M., Battisti, D.S., Po-Chedley, S., Johnson, N.C., Blanchard-Wrigglesworth, E., Harnos, K., Zhang, Q., Eastman, R., 2017. Influence of high-latitude atmospheric circulation changes on summertime Arctic sea ice. References Nat. Clim. Chang. 7, 289e295. Duplessy, J.C., Cortijo, E., Ivanova, E., Khusid, T., Labeyrie, L., Levitan, M., Murdmaa, I., Aagaard, K., Foldvik, A., Hillman, S.R., 1987. The West Spitsbergen Current: dispo- Paterne, M., 2005. Paleoceanography of the Barents Sea during the Holocene. sition and water mass transformation. J. Geophys. Res.: Oceans 92 (C4), Paleoceanography 20, PA4004. 3778e3784. Fer, I., Skogseth, R., Haugan, P.M., Jaccard, P., 2003. Observations of the Storfjorden Andreassen, I., No € thig, E.-M., Wassmann, P., 1996. Vertical particle flux on the shelf overflow. Deep-Sea Res. Part I Oceanogr. 50, 1283e1303. off northern Spitsbergen, Norway. Mar. Ecol. Prog. Ser. 137, 215e228. Forwick, M., Vorren, T.O., 2009. LateWeichselian and Holocene sedimentary envi- Antoniades, D., Francus, P., Pienitz, R., St-Onge, G., Vincent, W.F., 2011. Holocene ronments and ice rafting in Isfjorden, Spitsbergen. Palaeogeogr. Palaeoclimatol. dynamics of the Arctic's largest ice shelf. Proc. Natl. Acad. Sci. U. S. A. 108, Palaeoecol. 280, 258e274. 18899e18904. Gammelsrød, T., Leikvin, Ø., Lien, V., Budgell, W.P., Loeng, H., Maslowski, W., 2009. Årthun, M., Eldevik, T., Smedsrud, L.H., Skagseth, Ø., Ingvaldsen, R., 2012. Quanti- Mass and heat transports in the NE Barents Sea: observations and models. fying the influence of Atlantic heat on Barents Sea ice variability and retreat. J. Mar. Syst. 75, 56e69. J. Clim. 25, 4736e4743. Groot, D.E., Aagaard-Sørensen, S., Husum, K., 2014. Reconstruction of Atlantic water Bakke, J., Lie, Ø., Heegaard, E., Dokken, T., Haug, G.H., Birks, H.H., Dulski, P., Nilsen, T., variability during the Holocene in the western Barents Sea. Clim. Past 10 (1), 2009. Rapid oceanic and atmospheric changes during the Younger Dryas cold 51e62. period. Nat. Geosci. 2 (3), 202. Haarpaintner, J., Gascard, J.-C., Haugan, P.M., 2001. Ice production and brine for- Bartels, M., Titschack, J., Fahl, K., Stein, R., Seidenkrantz, M.-S., Hillaire-Marcel, C., mation in Storfjorden, Svalbard. J. Geophys. Res. 106, 013. Hebbeln, D., 2017. Atlantic Water advection vs. glacier dynamics in northern Hald, M., Andersson, C., Ebbesen, H., Jansen, E., Klitgaard-Kristensen, D., Spitsbergen since early deglaciation. Clim. Past 13, 1717e1749. Risebrobakken, B., Salomonsen, G.R., Sarnthein, M., Sejrup, H.P., Telford, R.J., Behrenfeld, M.J., O'Malley, R.T., Siegel, D.A., McClain, C.R., Sarmiento, J.L., 2007. Variations in temperature and extent of Atlantic water in the northern Feldman, G.C., Boss, E.S., 2006. Climate-driven trends in contemporary ocean North Atlantic during the Holocene. Quat. Sci. Rev. 26, 3423e3440. productivity. Nature 444, 752e755. Harada, N., Ahagon, N., Sakamoto, T., Uchida, M., Ikehara, M., Shibata, Y., 2006. Rapid Belt, S.T., Cabedo-Sanz, P., Smik, L., Navarro-Rodriguez, A., Berben, S.M., Knies, J., fluctuation of alkenone temperature in the southwestern Okhotsk Sea during Husum, K., 2015. Identification of paleo Arctic winter sea ice limits and the the past 120 ky. Glob. Planet. Chang. 53, 29e46. marginal ice zone: optimised biomarker-based reconstructions of late Quater- Hendricks, S., Gerland, S., Smedsrud, L., Haas, C., Pfaffhuber, A., Nilsen, F., 2011. Sea- nary Arctic sea ice. Earth Planet. Sci. Lett. 431, 127e139. ice thickness variability in Storfjorden, Svalbard. Ann. Glaciol. 52, 61e68. , A., 2004. Distributions of UK K0 0 Herbert, T., 2001. Review of alkenone calibrations (culture, water column, and Bendle, J., Rosell-Mele 37 UK37 and U37 UK37 in the surface waters and sediments of the Nordic Seas: implications for paleo- sediments). Geochem. Geophys. Geosyst. 2, 2000GC000055. ceanography. Geochem. Geophys. Geosyst. 5 (11), Q11013. Hoefs, M.J., Versteegh, G.J., Rijpstra, W.I.C., de Leeuw, J.W., Damste , J.S.S., 1998. Bendle, J., Rosell-Mele , A., Ziveri, P., 2005. Variability of unusual distributions of Postdepositional oxic degradation of alkenones: implications for the mea- alkenones in the surface waters of the Nordic seas. Paleoceanography 20, surement of palaeo sea surface temperatures. Paleoceanography 13 (1), 42e49. PA2001. Hopkins, T.S., 1991. The GIN Sea: a synthesis of its physical oceanography and Berger, A., Loutre, M.-F., 1991. Insolation values for the climate of the last 10 million literature review, 1972e1985. Earth Sci. Rev. 30, 175e318. years. Quat. Sci. Rev. 10, 297e317. Hughes, A.L., Gyllencreutz, R., Lohne, Ø.S., Mangerud, J., Svendsen, J.I., 2016. The last Boitsov, V.D., Karsakov, A.L., Trofimov, A.G., 2012. Atlantic water temperature and Eurasian ice sheetsea chronological database and time-slice reconstruction. climate in the Barents Sea, 2000e2009. ICES J. Mar. Sci. 69, 833e840. DATED-1. Boreas 45, 1e45. Bolton, C.T., Lawrence, K.T., Gibbs, S.J., Wilson, P.A., Cleaveland, L.C., Herbert, T.D., Ivanova, E., Murdmaa, I., de Vernal, A., Risebrobakken, B., Peyve, A., Brice, C., 2010. Glacialeinterglacial productivity changes recorded by alkenones and Seitkalieva, E., Pisarev, S., 2019. Postglacial paleoceanography and paleoenvir- microfossils in late Pliocene eastern equatorial Pacific and Atlantic upwelling onments in the northwestern Barents Sea. Quat. Res. 1e20. zones. Earth Planet. Sci. Lett. 295, 401e411. Kara, A.B., Wallcraft, A.J., Hurlburt, H.E., 2005. Sea surface temperature sensitivity to Brassell, S., Eglinton, G., Marlowe, I., Pflaumann, U., Sarnthein, M., 1986. Molecular water turbidity from simulations of the turbid Black Sea using HYCOM. J. Phys. stratigraphy: a new tool for climatic assessment. Nature 320, 129e133. Oceanogr. 35 (1), 33e54. Cabedo-Sanz, P., Belt, S.T., Knies, J., Husum, K., 2013. Identification of contrasting Knies, J., Pathirana, I., Cabedo-Sanz, P., Banica, A., Fabian, K., Rasmussen, T.L., seasonal sea ice conditions during the Younger Dryas. Quat. Sci. Rev. 79, 74e86. Forwick, M., Belt, S.T., 2017. Sea-ice dynamics in an Arctic coastal polynya Calvo, E., Grimalt, J., Jansen, E., 2002. High resolution UK during the past 6500 years. Arktos 3, 1. 37 U 37 K sea surface tem- perature reconstruction in the Norwegian Sea during the Holocene. Quat. Sci. Knudsen, K.L., Eiríksson, J., Jansen, E., Jiang, H., Rytter, F., Gudmundsdo ttir, E.R., Rev. 21, 1385e1394. 2004. Palaeoceanographic changes off North Iceland through the last 1200 Conte, M.H., Sicre, M.A., Rühlemann, C., Weber, J.C., Schulte, S., Schulz-Bull, D., years: foraminifera, stable isotopes, diatoms and ice rafted debris. Quat. Sci. Rev. Blanz, T., 02006. Global temperature calibration of the alkenone unsaturation 23, 2231e2246. index (UK Kristensen, D.K., Rasmussen, T.L., Koç, N., 2013. Palaeoceanographic changes in the 37 UK' 37) in surface waters and comparison with surface sediments. Geochem. Geophys. Geosyst. 7 (2), 2006Q02005. northern Barents Sea during the last 16 000 yearsenew constraints on the last Dalpadado, P., Ingvaldsen, R.B., Stige, L.C., Bogstad, B., Knutsen, T., Ottersen, G., deglaciation of the SvalbardeBarents Sea ice sheet. Boreas 42, 798e813. Ellertsen, B., 2012. Climate effects on Barents Sea ecosystem dynamics. ICES J. Kubiszyn, A., Piwosz, K., Wiktor Jr., J., Wiktor, J., 2014. The effect of inter-annual Mar. Sci. 69, 1303e1316. Atlantic water inflow variability on the planktonic protist community struc- Dalpadado, P., Arrigo, K.R., Hjøllo, S.S., Rey, F., Ingvaldsen, R.B., Sperfeld, E., ture in the West Spitsbergen waters during the summer. J. Plankton Res. 36, 14 M. Ła˛ cka et al. / Quaternary Science Reviews 224 (2019) 105973 1190e1203. Risebrobakken, B., Moros, M., Ivanova, E.V., Chistyakova, N., Rosenberg, R., 2010. Lewandowska, A.M., Boyce, D.G., Hofmann, M., Matthiessen, B., Sommer, U., Climate and oceanographic variability in the SW Barents Sea during the Ho- Worm, B., 2014. Effects of sea surface warming on marine plankton. Ecol. Lett. locene. Holocene 20 (4), 609e621. 17, 614e623. Risebrobakken, B., Dokken, T., Smedsrud, L.H., Andersson, C., Jansen, E., Moros, M., Lind, S., Ingvaldsen, R.B., 2012. Variability and impacts of Atlantic water entering the Ivanova, E.V., 2011. Early Holocene temperature variability in the Nordic Seas: Barents Sea from the North. Deep-Sea Res. Part I Oceanogr. 62, 70e88. the role of oceanic heat advection versus changes in orbital forcing. Paleo- Lind, S., Ingvaldsen, R.B., Furevik, T., 2018. Arctic warming hotspot in the northern ceanography 26, PA4206. Barents Sea linked to declining sea-ice import. Nat. Clim. Chang. 8, 634e639. Ritz, S.P., Stocker, T.F., Grimalt, J.O., Menviel, L., Timmermann, A., 2013. Estimated Loeng, H., 1991. Features of the physical oceanographic conditions of the Barents strength of the Atlantic overturning circulation during the last deglaciation. Sea. Polar Res. 10, 5e18. Nat. Geosci. 6, 208e212. Mangerud, J., Bondevik, S., Gulliksen, S., Hufthammer, A.K., Høisæter, T., 2006. Rosell-Mele , A., 1998. Interhemispheric appraisal of the value of alkenone indices as Marine 14C reservoir ages for 19th century whales and molluscs from the North temperature and salinity proxies in high-latitude locations. Paleoceanography Atlantic. Quat. Sci. Rev. 25, 3228e3245. 13 (6), 694e703. Mangerud, J., Svendsen, J.I., 2018. The Holocene thermal maximum around Sval- Rosell-Mele , A., Carter, J., Eglinton, G., 1994. Distributions of long-chain alkenones bard, Arctic North Atlantic; molluscs show early and exceptional warmth. Ho- and alkyl alkenoates in marine surface sediments from the North East Atlantic. locene 28 (1), 65e83. Org. Geochem. 22, 501e509. Marlowe, I., Brassell, S., Eglinton, G., Green, J., 1984. Long chain unsaturated ketones Rosell-Mele, A., Weinelt, M., Sarnthein, M., Koc, N., Jansen, E., 1998. Variability of the and esters in living algae and marine sediments. Org. Geochem. 6, 135e141. Arctic front during the last climatic cycle: application of a novel molecular Martrat, B., Grimalt, J.O., Villanueva, J., van Kreveld, S., Sarnthein, M., 2003. Climatic proxy. Terra Nova Oxford 10, 86e89. dependence of the organic matter contributions in the north eastern Norwe- Rudels, B., Korhonen, M., Schauer, U., Pisarev, S., Rabe, B., Wisotzki, A., 2015. Cir- gian Sea over the last 15,000 years. Org. Geochem. 34, 1057e1070. culation and transformation of Atlantic water in the Eurasian basin and the Müller, Peter J., Fischer, Gerhard, 2004. C37-Alkenones as paleotemperature tool: contribution of the Fram Strait inflow branch to the Arctic ocean heat budget. fundamentals based on sediment traps and surface sediments from the south Prog. Oceanogr. 132, 128e152. Atlantic ocean. In: Wefer, G., Mulitza, S., Ratmeyer, V. (Eds.), The South Atlantic Rueda, G., 2013. Reconstructing Climate Variability in the North Atlantic during the in the Late Quaternary: Reconstruction of Material Budgets and Current Sys- Late Holocene: an Integrated Biomarker Approach the Autonomous University tems. Springer, Berlin, Heidelberg, New York, pp. 167e193. of Barcelona. The Autonomous University of Barcelona, Barcelona, p. 179. Müller, P.J., Kirst, G., Ruhland, G., Von Storch, 0I., Rosell-Mele, A., 1998. Calibration of Salvigsen, O., 2002. Radiocarbon-dated Mytilus edulis and Modiolus modiolus from the alkenone paleotemperature index UK 0 37 U 37 K based on core-tops from the northern Svalbard: climatic implications. Norsk Geografisk Tidsskrift- eastern South Atlantic and the global ocean (60 N-60 S). Geochem. Cosmo- Norwegian J. Geog. 56, 56e61. chim. Acta 62, 1757e1772. Schauer, U., Rudels, B., Fer, I., Haugan, P., Skogseth, R., Bjo €rk, G., Winsor, P., 2003. Müller, J., Werner, K., Stein, R., Fahl, K., Moros, M., Jansen, E., 2012. Holocene cooling Return of Deep Shelf/Slope Convection in the Western Barents Sea, Seventh culminates in sea ice oscillations in Fram Strait. Quat. Sci. Rev. 47, 1e14. Conference on Polar Meteorology and oceanography and Joint Symposium on Nilsen, F., Cottier, F., Skogseth, R., Mattsson, S., 2008. Fjordeshelf exchanges High-Latitude Climate Variations. The American Meteorological Society. controlled by ice and brine production: the interannual variation of Atlantic Serreze, M.C., Barry, R.G., 2011. Processes and impacts of Arctic amplification: a Water in Isfjorden, Svalbard. Cont. Shelf Res. 28, 1838e1853. research synthesis. Glob. Planet. Chang. 77, 85e96. Onarheim, I.H., Eldevik, T., Årthun, M., Ingvaldsen, R.B., Smedsrud, L.H., 2015. Skillful Skogseth, R., Haugan, P., Haarpaintner, J., 2004. Ice and brine production in Storf- prediction of Barents Sea ice cover. Geophys. Res. Lett. 42, 5364e5371. jorden from four winters of satellite and in situ observations and modeling. Oziel, L., Sirven, J., Gascard, J.-C., 2016. The Barents Sea frontal zones and water J. Geophys. Res.: Oceans 109, C10008. masses variability (1980e2011). Ocean Sci. 12, 169e184. Slagstad, D., Ellingsen, I., Wassmann, P., 2011. Evaluating primary and secondary Oziel, L., Neukermans, G., Ardyna, M., Lancelot, C., Tison, J.L., Wassmann, P., production in an Arctic Ocean void of summer sea ice: an experimental simu- Sirven, J., Ruiz-Pino, D., Gascard, J.C., 2017. Role for Atlantic inflows and sea ice lation approach. Prog. Oceanogr. 90, 117e131. loss on shifting phytoplankton blooms in the Barents Sea. J. Geophys. Res.:  Slubowska, M.A., Koç, N., Rasmussen, T.L., Klitgaard-Kristensen, D., 2005. Changes in Oceans 122, 5121e5139. the flow of Atlantic water into the Arctic Ocean since the last deglaciation: Pathirana, I., Knies, J., Felix, M., Mann, U., 2014. Towards an improved organic car- evidence from the northern Svalbard continental margin, 80 . N. Paleoceanogr. bon budget for the western Barents Sea shelf. Clim. Past 10, 569e587. 20 (4), PA4014. Pearce, C., Seidenkrantz, M.-S., Kuijpers, A., Masse , G., Reynisson, N.F.,  Slubowska-Woldengen, M., Koç, N., Rasmussen, T.L., Klitgaard-Kristensen, D., Kristiansen, S.M., 2013. Ocean lead at the termination of the Younger Dryas cold Hald, M., Jennings, A.E., 2008. Time-slice reconstructions of ocean circulation spell. Nat. Commun. 4, 1664. changes on the continental shelf in the Nordic and Barents Seas during the last Pfirman, S., Bauch, D., Gammelsrød, T., 1994. The Northern Barents Sea: Water Mass 16,000 cal yr BP. Quat. Sci. Rev. 27, 1476e1492. Distribution and Modification. AGU, American Geophysical Union. Smedsrud, L.H., Esau, I., Ingvaldsen, R.B., Eldevik, T., Haugan, P.M., Li, C., Lien, V.S., Piwosz, K., Walkusz, W., Hapter, R., Wieczorek, P., Hop, H., Wiktor, J., 2009. Com- Olsen, A., Omar, A.M., Otterå, O.H., 2013. The role of the Barents Sea in the Arctic parison of productivity and phytoplankton in a warm (Kongsfjorden) and a cold climate system. Rev. Geophys. 51, 415e449. (Hornsund) Spitsbergen fjord in mid-summer 2002. Polar Biol. 32, 549e559. Smith, W.O., Sakshaug, E., 1990. Polar phytoplankton. In: Smith, W.O. (Ed.), Polar Polyak, L., Alley, R.B., Andrews, J.T., Brigham-Grette, J., Cronin, T.M., Darby, D.A., Oceanography, Part B: Chemistry, Biology and Geology. Academic Press, New Dyke, A.S., Fitzpatrick, J.J., Funder, S., Holland, M., 2010. History of sea ice in the York, pp. 447e525. Arctic. Quat. Sci. Rev. 29, 1757e1778. Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB radiocarbon Polyakov, I.V., Pnyushkov, A.V., Alkire, M.B., Ashik, I.M., Baumann, T.M., calibration program. Radiocarbon 35, 215e230. Carmack, E.C., Goszczko, I., Guthrie, J., Ivanov, V.V., Kanzow, T., 2017. Greater role Thorarinsdo ttir, G.G., Gunnarsson, K., 2003. Reproductive cycles of Mytilus edulis L. for Atlantic inflows on sea-ice loss in the Eurasian basin of the Arctic ocean. on the west and east coasts of Iceland. Polar Res. 22, 217e223. Science 356 (6335), 285e291. Trudnowska, E., Gluchowska, M., Beszczynska-Mo €ller, A., Blachowiak-Samolyk, K., Rasmussen, T.L., Thomsen, E., 2014. Brine formation in relation to climate changes Kwasniewski, S., 2016. Plankton patchiness in the polar front region of the west and ice retreat during the last 15,000 years in Storfjorden, Svalbard, 76e78 N. Spitsbergen shelf. Mar. Ecol. Prog. Ser. 560, 1e18. Paleoceanography 29, 911e929. Volkman, J.K., Eglinton, G., Corner, E.D., Forsberg, T., 1980. Long-chain alkenes and  Rasmussen, T.L., Thomsen, E., Slubowska, M.A., Jessen, S., Solheim, A., Koç, N., 2007. alkenones in the marine coccolithophorid Emiliania huxleyi. Phytochem 19, Paleoceanographic evolution of the SW Svalbard margin (76 N) since 20,000 2619e2622. 14C yr BP. Quat. Res. 67, 100e114. Walczowski, W., 2013. Frontal structures in the west Spitsbergen current margins. Rasmussen, T.L., Thomsen, E., Nielsen, T., Wastegård, S., 2011. Atlantic surface water Ocean Sci. 9 (6), 957e975. inflow to the Nordic seas during the PleistoceneeHolocene transition (midelate Wollenburg, J.E., Knies, J., Mackensen, A., 2004. High-resolution paleoproductivity Younger Dryas and Pre-Boreal periods, 12 450e10 000 a BP). J. Quat. Sci. 26, fluctuations during the past 24 kyr as indicated by benthic foraminifera in the 723e733. marginal Arctic Ocean. Palaeogeogr. Palaeoclimatol. Palaeoecol. 204, 209e238.  Rasmussen, T.L., Thomsen, E., Skirbekk, K., Slubowska-Woldengen, M., Yang, X.-Y., Yuan, X., Ting, M., 2016. Dynamical link between the BarentseKara Sea Kristensen, D.K., Koç, N., 2014. Spatial and temporal distribution of Holocene Ice and the Arctic oscillation. J. Clim. 29, 5103e5122. temperature maxima in the northern Nordic seas: interplay of Atlantic-, Arctic- Zaja˛ czkowski, M., 2008. Sediment supply and fluxes in glacial and outwash fjords, and polar water masses. Quat. Sci. Rev. 92, 280e291. Kongsfjorden and Adventfjorden. Svalbard. Polish Polar Res. 29, 59e72. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Zaja˛ czkowski, M., Nygård, H., Hegseth, E.N., Berge, J., 2010. Vertical flux of partic- Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., ulate matter in an Arctic fjord: the case of lack of the sea-ice cover in Adven- Haflidason, H., Hajdas, I., HattẶ, S.C.,  Heaton, T.J., Hogg, A.G., Hughen, K.A., tfjorden 2006e2007. Polar Biol. 33 (2), 223e239. Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Ła˛ cka, M., Pawłowska, J., Zaja˛ czkowski, M., 2015a. New Methods in the Recon- Scott, E.M., Southon, J.R., Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and struction of Arctic Marine Palaeoenvironments, Impact of Climate Changes on MARINE13 radiocarbon age calibration curves 0e50,000 years cal BP. Radio- Marine Environments. Springer, pp. 127e148. carbon 55, 1869e1887. Ła˛ cka, M., Zaja˛ czkowski, M., Forwick, M., Szczucin  ski, W., 2015b. Late Weichselian Risebrobakken, B., Berben, S.M.P., 2018. Early Holocene establishment of the Barents and Holocene palaeoceanography of Storfjordrenna, southern Svalbard. Clim. Sea Arctic front. Front. Earth Sci. 6, 166. Past 11, 587e603.