New insights into sea ice changes over the past 2.2 kyr in Disko Bugt, West Greenland
- 261 Downloads
Past sea ice conditions and open water phytoplankton production were reconstructed from a sediment core taken in Disko Bugt, West Greenland, using the sea ice biomarker IP25 and other specific phytoplankton biomarker (i.e., brassicasterol, dinosterol, HBI III) records. Our biomarker record indicates that Disko Bugt experienced a gradual expansion of seasonal sea ice during the last 2.2 kyr. Maximum sea ice extent was reached during the Little Ice Age around 0.2 kyr BP. Superimposed on this longer term trend, we find short-term oscillations in open water primary production and terrigenous input, which may be related to the Atlantic Multidecadal Oscillation and solar activity changes as potential climatic trigger mechanisms. A direct sample-to-sample multiproxy comparison of our new biomarker record with microfossil (i.e., benthic foraminifera, dinocysts, and diatoms) and other geochemical records (i.e., alkenone biomarkers) indicates that different proxies are influenced by the complex environmental system with pronounced seasonal changes and strong oceanographic gradients, e.g., freshwater inflow from the Greenland Ice Sheet. Differences in sea ice reconstructions may indicate that the IP25 record reflects only the relatively short sea ice season (spring), whereas other microfossil reconstructions may reflect a longer (spring–autumn) interval.
KeywordsBaffin Bay Disko Bugt Late Holocene IP25 HBI III Brassicasterol PIP25
Sea ice plays an important role in the climate system by influencing the Earth’s energy balance as well as the exchange between the ocean and atmosphere. It is considered as one of the critical components in climate models , and the underestimation of the recent sea ice loss [25, 118, 134] displays the need to extend our knowledge of natural sea ice dynamics. As reliable satellite observations are only available for the past ~ 40 years , high-resolution proxy reconstructions from regions characterized by highly dynamic sea ice conditions may provide vital information about pre-industrial sea ice variability.
An area that displays such seasonal sea ice variability is the Baffin Bay, especially the Disko Bugt area off central West Greenland (Fig. 1). In autumn, the sea ice margin of the so-called ‘Westice’ expands southwards and reaches as far as Kangerlussaq, close to Disko Bugt, during winter (Fig. 2; ). In summer, the ice margin retreats as far north as Nares Strait and leaves Disko Bugt ice free (Fig. 2; ). Furthermore, the ecological setting of Disko Bugt is highly affected by outlet glaciers of the Greenland Ice Sheet and associated meltwater input, causing a strong seasonal stratification .
Due to the importance of this area, several studies have focused on paleoceanographic changes (e.g., [2, 27, 60, 62, 71, 77, 80, 92, 97, 98, 105, 117]) and variations of the Greenland Ice Sheet (e.g., [147, 153]) during the mid-to-late Holocene. Within these studies, first insights into changes in sea ice conditions throughout the Holocene have been gained. However, the application of different microfossil proxies displays different inferences of sea ice conditions. This demonstrates the need for more direct proxy reconstructions of sea ice.
We provide a high-resolution sea ice reconstruction based on the sea ice proxy IP25 and contribute important information about the natural variability of sea ice in this highly sensitive region over the past 2.2 kyr. Furthermore, we compare microfossil and geochemical proxies to provide a better understanding of the relationship of those proxies to sea ice and other environmental factors.
Application of biomarkers for environmental reconstructions
Biomarker proxy for sea ice
A highly branched isoprenoid (HBI) alkene with 25 carbon atoms (i.e., IP25) has been proven as a useful proxy for sea ice reconstructions in the Northern Hemisphere . Produced by specific sea ice diatoms [9, 18], IP25 serves as a presence/absence indicator for spring sea ice. For more detailed interpretations of past sea ice conditions, Müller et al.  have introduced the so-called ‘PIP25 index’ (see “Materials and methods” for calculation). This index combines open water phytoplankton biomarker proxies with the sea ice proxy IP25. This approach also avoids the misleading interpretations concerning the absence of IP25, which may result either from a lack of sea ice or a permanent, thick sea ice cover with limited light penetration for ice algae growth . As open water phytoplankton marker serve brassicasterol and dinosterol produced by a wide range of algae groups (e.g., [144, 145]). These sterols have commonly been used in the original work by Müller et al. , to calculate the PBIP25 and PDIP25, respectively (for calculation see Chap. 3.3.). When applying the PIP25 index for sea ice reconstruction, however, individual biomarker concentrations still need to be considered, as in-phase changes to high (associated with marginal sea ice conditions) or low (associated with permanent sea ice cover) concentrations of IP25 and the open water phytoplankton marker may result in a similar PIP25 index .
Despite these limitations, comparisons with satellite-derived values from the East Greenland Shelf and Fram Strait , Barents Sea , and the central Arctic Ocean [150, 152] verified that the IP25/PIP25 approach represents modern sea ice concentration quite well in.
Furthermore, this approach has been successfully applied to reconstruct past sea ice conditions for the Holocene time period in North Atlantic regions, i.e., the Barents Sea [12, 13, 57, 139], the East Greenland and North Iceland Shelves [1, 3, 21, 28, 59, 73, 149], the Fram Strait [22, 81, 84] and Labrador Sea , the Arctic Ocean and Eurasian marginal seas [34, 43, 100], the Canadian Arctic Archipelago [11, 99, 138] and the North Pacific . On Quaternary interglacial and glacial time scales, sea ice conditions have been reconstruct based on the IP25/PIP25 approach in the North Atlantic [23, 39, 82], the Arctic Ocean [44, 64, 128, 129, 130, 131, 151] and the North Pacific . The oldest studies so far applied the IP25/PIP25 approach in Pliocene sediments from the Iceland Sea  and Miocene sediments from the central Arctic Ocean .
Recently, Smik et al.  have introduced a tri-unsaturated HBI alkene (HBI III, C25:3, Z-isomer) as a possible more suitable open water phytoplankton biomarker for the PIP25 calculation (resulting in the PIIIIP25 index). HBI III is produced by marine open water diatoms, presumably by diatoms of the genera Pleurosigma and Rhizosolenia [7, 112] and highest abundances have been observed close to the marginal ice zone [6, 8; 8]. So far, however, the relation of sedimentary HBI III concentrations to modern sea ice conditions and its use in the PIP25 calculation (resulting in the PIIIIP25 index) have only been tested in sedimentary records from the Barents Sea and its northern continental margin [8, 125] and in East Greenland fjord sediments . For paleo sea ice reconstructions, the PIIIIP25 index revealed reliable sea ice reconstructions from the Central Arctic Ocean [64, 132] and Barents Sea .
Biomarker proxies for organic carbon sources
Specific biomarkers have been associated with terrigenous and marine organic matter input in paleoenvironmental reconstructions (e.g., [32, 75, 127]). Long chain n-alkanes (C25, C27, C29, C31) and lignin are established proxies for the reconstruction of organic matter input from vascular plants (e.g., [101, 154]). Despite being found in a number of diatoms [102, 142], β-sitosterol and campesterol are especially associated with terrestrial sources, i.e., vascular land plants [111, 143, 144] and have been successfully applied for biomarker reconstructions of terrigenous input (e.g., [31, 32, 33, 44, 150, 151]).
For marine organic matter production reconstructions, specific short-chained n-alkanes (C17, C19; [15, 101, 154]) and short-chain fatty acids (e.g., [29, 31, 86, 145]) have been successfully applied. Here, we use the sterols brassicasterol and dinosterol, which have been established as reliable phytoplankton biomarker proxies (e.g., [144, 145]).
Baffin Bay is strongly influenced by the West Greenland Current (WGC), which is composed of Atlantic/Irminger Current (IC) derived waters and waters of Polar/East Greenland Current (EGC) origin (Fig. 1; ). The WGC inflow to Disko Bugt occurs mainly in the southern part of the bay, whereas the outflow occurs predominantly via the Vaigat Strait in the northern part of the bay (Fig. 1). The WGC penetrates deeper parts of the fjords and affects subglacial melting of tidewater glaciers (Fig. 1; [40, 108]).
The bay is characterized by seasonal sea ice and is influenced by two different types of sea ice . Multiyear ice, the so-called ‘Storis’ originates in the Arctic Ocean and is transported, along with polar waters, by the EGC around Cape Farewell and then carried northward with the WGC . The amount of ‘Storis’ reaching Baffin Bay is determined by the sea ice formation in the Arctic Ocean, outflow intensity via Fram Strait and conditions in the Greenland Sea . The second kind of sea ice is first-year ice, the ‘Westice’, which starts to form in northern Baffin Bay in September, expands southwards and reaches its maximum extent in March/April. Then, almost the entire Baffin Bay, except eastern Davis Strait, is covered by sea ice [16, 54].
Our study site, the Disko Bugt, is a large open bay (40,000 km2) off central West Greenland (Fig. 1). It is located at the southern boundary of seasonal ice coverage in Baffin Bay and is highly sensitive to changes in sea ice extent. At present, Disko Bugt is covered by seasonal land-fast sea ice for 3–5 months per year with a mean thickness of ca. 70 cm [20, 87, 106]. Over the last 20 years, milder winter temperatures have caused a reduction of the seasonal sea ice cover of Disko Bugt from 5 to 3 months per year [87, 122].
Disko Bugt can be considered as a highly productive low arctic marine environment. Phytoplankton activity in Godthåbsfjord, south of our study site, is characterized by three annual maxima, during spring, mid-summer, and autumn, with seasonal changes in the phytoplankton community typical for low Arctic environments . Haptophytes and diatoms are the dominant phytoplankton species throughout the year, with haptophytes dominating from March–July and diatoms being dominant in July . Other phytoplankton groups, i.e., dinoflagellates, ciliates, chrysiophytes, and silicoflagellates, are also observed during winter .
Surface water salinities and temperatures in Disko Bugt are influenced by seasonal changes of meltwater fluxes and icebergs from outlet glaciers (e.g., Jakobshavn Isbrae) and sea ice cover [20, 36, 40]. Sea ice cover limits the light availability in surface waters and thus phytoplankton production . In spring, the combination of sea ice break-up, which improves light conditions, and a remaining stratification of the water column, with nutrient enriched surface waters, favour the spring plankton bloom . During summer meltwater inflow from the Greenland Ice Sheet and melting sea ice create a fresh water lens in the upper 20–30 m of the water column . A strong seasonal shift in stratification is caused by enhanced wind activity during autumn and favors a well-mixed upper water layer .
Materials and methods
Sediments were collected in June 2007 from southwestern Disko Bugt (Fig. 1; 68° 38′N, 53°49′W; water depth: 855 m) from Egedesminde Dyb, during Cruise MSM05/03 of the R/V Maria S. Merian . Sediments consist of olive brown to olive grey organic rich silty clay with occasional shell fragments. The gravity core (GC 343310-5-1; 940 cm) was sampled at 1 cm steps; the multi core (MUC 343310-2-2; 32 cm) was sampled at 0.5 cm. This study uses the upper 580.5 cm of the gravity core and five selected samples of the multi-core in a composite record referred to as Core 343310. Sediments were stored at 4 °C, and samples were taken and freeze-dried in 2011 and stored at 4 °C until geochemical analysis in 2016.
A robust age model was established based on 20 AMS 14C dates for the gravity core and 10 AMS 14C dates for the multi core, from benthic foraminifera and mollusc shells (Supplementary Fig. 2; [71, 98]). Ages were calibrated with Marine 09  using OxCal 4.1  and a marine reservoir age of ΔR = 140 ± 35 years. In addition, the age model of the multi-core was complemented with 210Pb/137Cs measurements . Due to gravity coring disturbance and sediment loss, there is a gap of ~ 100 years between the gravity and multi core. Further details about the age control are given by Perner et al.  and Lloyd et al. .
The analysis of surface sediments and Core 343310 was carried out on freeze-dried and homogenised sediment. The sediment was analysed for total organic carbon (TOC) content and concentrations of IP25, HBI III, brassicasterol, dinosterol, β-sitosterol and campesterol (Table 1).
Compounds and their source used in this study
Sea ice diatoms; Pleurosigma stuxbergii var. rhomboides; Haslea crucigeroides (and/or Haslea spicula); Haslea kjellmanii
polar and sub-polar marine diatoms, i.e., Rhizosolenia
Marine and freshwater phytoplankton
4 α, 23, 24 trimethyl-5α-cholest-22E-en- 3β -ol
Marine and freshwater phytoplankton (dinoflagellates)
Predominantly terrestrial plants
Predominantly terrestrial plants
Homogenised subsamples (0.1 g) were analysed for TOC content with a carbon–sulphur determinator (ELTRA CS-125).
Prior to the extraction, two internal standards, 7-HND (7-hexylnonadecane, 0.076 µg/sample) and cholesterol-D6 (cholest-5-en-3β-ol-D6, 11 µg/sample), were added for quantification purposes. The samples were extracted with an Accelerated Solvent Extractor (DIONEX, ASE 200; 100 °C, 5 min, 1000 psi). About 3–5 g of sediment was extracted using dichloromethane:methanol (2:1 vol/vol) as solvent.
The extracts were separated into different fractions by open-column chromatography, with SiO2 as stationary phase. As solvent n-hexane (5 ml) was used for IP25, and ethylacetate:n-hexane (20:80 vol/vol; 7 ml) for sterols. The sterol fraction was silylated using 200 µl BSTFA (60 °C, 2 h).
Hydrocarbon concentrations were identified with a gas chromatograph Agilent Technologies 7890 GC (30 m HP-1MS column, 0.25 mm in diameter and 0.25 µm film thickness) coupled to an Agilent Technologies 5977 A mass selective detector. Sterol concentrations were identified with a gas chromatograph Agilent Technologies 6850 GC (30 m HP-1MS column, 0.25 mm in diameter, and 0.25 µm film thickness) coupled to an Agilent Technologies 5975 A mass selective detector. The individual compounds were identified by comparing the retention times with those of reference compounds. IP25 and HBI III (IP25: m/z 350; HBI III: m/z 348) were quantified in relation with the abundant fragment ion m/z 266 of the internal standard 7-hexylnonadecane. The sterols were quantified as trimethylsilyl ethers (brassicasterol: m/z 470, campesterol:m/z 472, β-sitosterol: m/z 486, dinosterol: m/z 500) with respect to the molecular ion of cholest-5-en-3b-ol-D6 (ion m/z 464). A detailed description of the quantification methods is given by Fahl and Stein . The concentrations of all biomarkers have been normalised to the amount of sediment and TOC. As both show nearly the same signal and variability, we only present the data normalised to the amount of TOC referred to as µg/gTOC in the main text. Data normalised to the amount of sediment (µg/gSed) are shown in Supplementary Fig. 3. Instrument stability was controlled by several reruns of external standards (cholesterol-D6, 7-HND, reference sediment) several times during one analytical sequence and by replicate analyses for random samples.
For a statistic analysis of short-term oscillations, a MatLab code was applied to detrend the original data, and the BTuckey function of AnalySeries  was applied to produce a spectrum of the oscillations (confidence 90%, bandwidth: 0.1).
All data are available on https://doi.org/10.1594/PANGAEA.885107.
Throughout the last 2.2 kyr the TOC contents in Core 343310 showed strong fluctuations on decadal scales ranging from 2.6 to 1.6%, averaging around ~ 2% (Fig. 3f). Overall, IP25 concentrations remained relatively low and constant until 1.2 kyr BP, followed by a gradual increase (Fig. 3e). The lowest concentrations, around 0.06 µg/gTOC, are observed in the lowermost core section from 2.2 to 1.5 kyr BP (Fig. 3e). In the upper core section, IP25 concentrations increased and reached high and variable concentrations around 0.4 kyr BP, with the highest concentrations ~ 0.2 kyr BP (~ 0.9 µg/gTOC; Fig. 3e). HBI III concentrations were low (~ 0.05 µg/gTOC) from 2.2 to 0.2 kyr BP, after which concentrations increased to maximum values of 1.8 µg/gTOC (Fig. 3d). Phytoplankton sterols, brassicasterol, and dinosterol showed strong decadal fluctuations ranging from 13 to 86 µg/gTOC and 3 to 71 µg/gTOC, respectively (Fig. 3b, c). Phases with enhanced sterols concentrations occured from 1.8 to 1.6 kyr (50 µg/gTOC and 55 µg/gTOC, respectively), from 0.8 to 0.7 kyr BP (86 µg/gTOC and 71 µg/gTOC, respectively), and from 0.15 kyr BP onwards (81 µg/gTOC and 50 µg/gTOC, respectively; Fig. 3b, c). Campesterol + β-sitosterol showed the lowest concentrations (~ 50 µg/gTOC) from 2.2 to 1.2 kyr BP, with a pronounced peak ~ 1.4 kyr BP (reaching 187 µg/gTOC; Fig. 3a). From 1.0 to 0.6 kyr BP, the highest concentrations were observed with maximum values of ~ 300 µg/gTOC. After 0.6 kyr, concentrations decreased and reach values of ~ 75 µg/gTOC (Fig. 3a).
During the last 0.1 kyr, all biomarker concentrations showed an increase, brassicasterol and HBI III reach maximum values in the uppermost sample (80 µg/gTOC and 1.8 µg/gTOC, respectively; Fig. 3b, d).
Superimposed on the general trend, brassicasterol, dinosterol, and campesterol + β-sitosterol concentrations show a high variability on a decadal time scale (Fig. 3a–c).
Sea ice variability in Disko Bugt over the past 2.2 kyr BP
Our biomarker record provides a new and more direct sea ice reconstruction than the previous microfossil reconstructions. This helps to improve the interpretation of ecological changes and to identify the driving mechanisms of sea ice changes in Disko Bugt over the past two millennia, a time span with several well-known short-term climate events (e.g., [67, 68, 72]). Furthermore, paleoenvironmental conditions have previously been reconstructed applying a variety of microfossil proxies and alkenone geochemistry on Core 343310 [2, 60, 80, 98, 105]. This wide range of proxy records provides an ideal basis for comparing microfossil and geochemical proxies, which were derived from the same sediment material.
In general, our biomarker-based sea ice reconstruction indicates a gradual increase of seasonal sea ice concentration over the past 2.2 kyr BP (Fig. 4c, d). We use the PIP25 index to receive more information on past sea ice conditions. The phytoplankton sterols brassicasterol and dinosterol have a similar relationship to IP25 as indicated in the initial study of Müller et al.  and can be related to their classifications regarding sea ice conditions (Fig. 5). The resulting PBIP25 (using brassicasterol) and PDIP25 (using dinosterol) indices reveal very similar results for sea ice reconstructions (see Supplementary Fig. 4). The relationship of the phytoplankton marker HBI III and IP25, however, seems to be more complex, which hampers a clear categorization based on Müller et al. . Hence, a direct relation of the resulting PIIIIP25 index (shown in Supplementary Fig. 4) to sea ice concentration in Baffin Bay remains uncertain. Therefore, we focus on the PDIP25 and PBIP25 index for indications on sea ice concentration in the following discussion. As they show nearly identical results, only the PDIP25 index it used for interpretations.
The Roman Warm Period (2.2–1.7 kyr BP)
Based on previous proxy reconstructions from Core 343310 the time period from 2.2 to 1.7 kyr BP has been characterized by reduced sea ice conditions (Fig. 7e, f; ) and a gradual surface and subsurface warming associated to high contributions of Atlantic derived waters to the WGC (Fig. 4n; ). Strong Atlantic Water inflow has been related to favour subsurface melting of Greenland outlet glaciers causing the observed high meltwater contribution to the marine environment which has been associated with the reconstructed reduction in SSTs at our core site (Fig. 4k, l; ).
Our biomarker data suggest low sea ice algae productivity and reduced sea ice conditions at the mouth of Disko Bugt, which is evident from low IP25 concentration and relatively low PDIP25 values (Fig. 4c, d). This strengthens findings from dinocyst sea ice reconstructions, characterizing this time period by winter-only sea ice . Further supporting previous assumptions of Krawczyk et al. , peaks in open water phytoplankton sterols, i.e., brassicasterol and dinosterol, around 1.75 kyr BP (Fig. 4h, i), may be related to variable environmental conditions and a strong spring/summer stratification. A pronounced stratification of the upper water column is associated with favourable nutrient conditions, beneficial conditions for phytoplankton blooms , which may be represented in high phytoplankton concentrations during this time.
The period between 2.2 and 1.2 kyr BP, with lower than modern sea ice conditions in Disko Bugt (Fig. 6b), coincides with generally warm conditions over the Greenland Ice Sheet (Fig. 4b; ). A phase of relative atmospheric warming has been observed in NW Europe, termed the Roman Warm Period (RWP; 1.65–1.94 kyr BP; ). This characteristic warming and reduction of sea ice during the RWP has been recorded in several records along the West Greenland coast (e.g., [30, 58, 67, 117]), the Canadian Arctic  as well as the East Greenland Shelf, North Iceland and the northern North Atlantic [21, 49, 78, 95, 96, 109, 123, 146, 79, 21, 95, 96], which indicates that this warming is at least of a wider regional scale in the northern Hemisphere.
The Dark Ages Cold Period (1.7–1.2 kyr BP)
Based on foraminiferal assemblages, the core site at Disko Bugt was characterized by a gradual cooling and an increasing contribution of polar EGC waters advected to the WGC during the interval from 1.8 kyr BP onwards (Fig. 4n, m; ). However, surface water proxies, i.e., diatoms and dinocysts, reveal contradictory surface water conditions. Whilst diatom assemblages indicate a gradual warming of surface waters around 1.7 kyr BP and only a minor increase in sea ice conditions (Fig. 4e, f; [60, 61]) dinocyst reconstructions point towards a gradual cooling with reduced sea ice conditions (Fig. 4g) and relatively high phytoplankton productivity . However, recent dinocyst reconstructions indicate a major change in surface water conditions at ~ 1.5 kyr BP towards cooler conditions and prolonged sea ice cover (up to 8 month/year; ).
Our new biomarker record supports a slight increase in sea ice conditions at the core site from 1.7 to 1.2 kyr BP, indicated by a minor increase in IP25 concentrations and PIP25 values (Fig. 4c, d), similar to findings from diatom assemblages (Fig. 4e, f; [60, 61]). Increased sedimentation rates and higher terrigenous sterol concentrations are observed during this interval (Figs. 3g, 4j). This may be related to a re-advance of the Greenland Ice Sheet (Fig. 4; [69, 116]) with higher glacial activity resulting in an increased input of terrigenous material. Open water phytoplankton sterol concentrations do not indicate major changes in open water phytoplankton production, compared to the previous RWP (Fig. 4h, i), which is contradictory to findings from dinocyst assemblages . However, as previously discussed, phytoplankton proxies and their signals may be strongly influenced by water mass composition and water column structure/stratification, caused especially by changes in meltwater supply  which could account for the observed differences.
The minor increase of sea ice conditions and terrigenous input (Fig. 4c, d, j) correlates with the NW European Dark Ages Cold Period (DACP, 1.65–1.15 kyr BP; ). The findings from our new biomarker record strengthen earlier assumptions of several Disko Bugt records, which characterise the DACP as a phase of surface water cooling and propose an increase in sea ice [77, 92, 117, 120], which has been related to an increased EGC contribution and a relative reduction of IC waters to the WGC [97, 98].
The Medieval Climate Anomaly (1.2–0.7 kyr BP)
The period from 1.2 to 0.7 kyr BP was characterized by a slight subsurface warming . Reconstructions of surface water conditions reveal inconsistent results; diatom assemblages point towards a cooling and increasing sea ice concentrations during this period (Fig. 4e, f; ), whereas dinocyst only indicate cool SSTs and relatively high sea ice conditions until 0.9 kyr BP followed by warmer SSTs and decreasing sea ice conditions until 0.7 kyr BP (Fig. 4g; ). However, both surface water proxies point towards high but variable freshwater input from adjacent glaciers and melting sea ice which is supported by alkenone reconstructions [60, 80, 105].
Our biomarker records show a gradual increase in IP25 concentration and the PIP25 index, which may point towards a gradual increase in sea ice algae productivity and sea ice concentration from 1.2 to 0.7 kyr BP (Fig. 4c, d). This indicates a minor increase in seasonal sea ice conditions, conceivably related to a temporal extent of ‘Westice’, an extension of land-fast ice out of Disko Bugt or increased drift ice transport by the EGC. An increase in sea ice conditions may have caused a longer period of favourable light conditions and sea ice habitat for sea ice algae bloom (Fig. 6a, b), which supports findings from dinocyst reconstructions .
HBI III concentrations show a minor increase, however, remaining at very low concentrations (Fig. 3d). Based on assumptions of Belt et al. , such low levels may indicate that a prolonged, high productivity ice edge was not established. The role of the HBI III is not yet fully understood, which makes it difficult to interpret this slight increase in concentration.
Open water phytoplankton productivity, as indicated by brassicasterol and dinosterol concentrations (Fig. 4h, i), remains relatively high and variable, which supports findings from dinocyst reconstructions . Variable phytoplankton production may be related to the continuous unstable environmental conditions as indicated by alkenone reconstructions showing variable meltwater inflow and associated changes in surface SST (Fig. 4k, l; ). These highly variable seasonal environmental conditions may have affected diatoms and dinocysts and caused the inconsistent (sea ice) signals reflected in their paleorecords from Core 343310 (Fig. 4 e–g; [2, 60, 80, 105]).
This phase of slightly enhanced sea ice conditions corresponds with an atmospheric warming over Greenland (Fig. 4b; ) and encompasses the European Medieval Climate Anomaly (MCA; 1.0–0.75 kyr BP; ). Nevertheless, our biomarker record does not indicate a reduction of sea ice that can be associated with the atmospheric warming during the MCA. At Disko Bugt, such a warming has only been recorded in bottom water proxies, i.e., benthic foraminifera (Fig. 4m, n; ), whereas surface proxies, such as dinoflagellates and diatoms, support a cooling trend (Fig. 4e–g, e.g., 60, 77, 117, 105]). An extent of sea ice during the MCA, indicated in our biomarker data, may be linked to a general change in oceanography, e.g., an intensification of Arctic outflow via the EGC and Baffin Island Current (BIC), transporting colder, polar water masses to the area. Evidence for a strengthening of the EGC and a southward migration of the sub-arctic front as well as a reduction in SSTs during this time has been found at Reykjanes Ridge [95, 126]. Furthermore, a lack of surface warming during the MCA in the Labrador Sea has been attributed to a constant strong Labrador Current (LC) along the Canadian coast over the past 1.6 kyr [55, 107]. The LC originates from the BIC and WGC, which may point towards a strong Arctic contribution from both currents. The combination of enhanced cold Arctic water inflow from the Northwest and the South to Baffin Bay may have caused the south-eastward expansion of ‘Westice’ . Sea ice reconstruction from the Canadian Arctic, located at the gateway from the Arctic Ocean to Baffin Bay, supports a gradual increase of sea ice algae productivity during this phase .
The Little Ice Age (0.7–0.2 kyr BP)
For the period from 0.7 to 0.2 kyr BP, benthic foraminiferal assemblages indicate increasing contributions of EGC to the WGC, accompanied with a subsurface cooling at the investigated core site (Fig. 4m, n; ). Surface conditions are, based on dinocysts, characterized by cooler SSTs, extensive sea ice cover (Fig. 4g), reduced phytoplankton production, and lowered terrigenous input . Further alkenone reconstructions point towards low meltwater inflow and cold SSTs (Fig. 4k, l; ). Conversely, diatom reconstructions indicate a reducing trend in SSTs and sea ice over the course of the last 0.7 kyr (Fig. 4e, f; [60, 61]).
Our biomarker record supports harsher sea ice conditions, possibly similar to conditions as observed today (Fig. 6b), indicated by strong increased in IP25 concentration and the PDIP25 index (Fig. 4c, d). Based on assumptions of Belt et al. , increasing HBI III concentrations (Fig. 3d) may indicate the presence of an ice edge at or close to our core site during spring. Relatively stable but variable phytoplankton sterol concentrations support the previous findings of variable environmental conditions . The relative reduction of meltwater inflow and a subsequent increase SSTs, as reconstructed from alkenones , has been associated with the signal preserved in diatom reconstructions [60, 61, 80]. This demonstrates again that different proxies seem to be affected in a different way by certain environmental conditions.
This time interval is characterized by a decrease of air temperatures over Greenland (Fig. 4b; ) and advances of the Greenland Ice Sheet and its outlet glaciers, e.g., Jakobshavn Isbræ (Fig. 4; [69, 71]). This atmospheric cooling likely favoured the development of maximum sea ice concentrations (Fig. 4d) and possibly the establishment of an ice edge until late spring, with favourable light conditions for IP25 production, in the Disko Bugt area (Fig. 6a). The strongest increase in sea ice algae production and sea ice concentration, indicated by our biomarker record (Fig. 4c, d), correlates with the maximum of the Little Ice Age (LIA; ) around 0.2 kyr BP. Such a cooling and increase in sea ice conditions has been recorded along the West Greenland coast [27, 50, 110, 119], North Iceland [21, 52, 73] and the Canadian Arctic Archipelago [11, 138] and seems to represent a widespread climatic phenomenon.
The extent of sea ice during the LIA may be related to a general increase in subsurface EGC contribution to the WGC , causing a general cooling of subsurface water masses in the area. In addition, a dominant westward Transpolar Drift, generating a strong cold LC inflow into Baffin Bay , may have favoured the temporal expansion of ‘Westice’. The onset of the widely recorded LIA has been associated with shifts in the NAO (Fig. 4a, e.g., [123, 137]) and an increase in volcanic activity, which may have caused a short time reduction in atmospheric temperature, which itself may set off a sea ice feedback [58, 76].
The last century
A reduction in sea ice concentrations and a strong increase in phytoplankton productivity are observed after 0.2 kyr BP (Fig. 4c, d, h, i). The youngest samples seem to indicate a strong increase of ice algae and phytoplankton productivity (Fig. 4c, d, h, i). Findings from microfossils from our study site support a general reduction in sea ice during the last 0.2 kyr (Fig. 4e, f, g; [2, 60, 107]) and are in line with a general increase in atmospheric temperatures observed at the coast of Disko Bugt, e.g., in Ilulisaat . This warming has been associated with a decrease in mean annual sea ice cover since the 1990s . Due to the low sample resolution in our biomarker record, a detailed interpretation is not possible for this time period.
Cyclicity and driving mechanisms of sea ice conditions in Disko Bugt
Microfossil records from Disko Bugt and West Greenland sites found in parts of the late Holocene an antiphase correlation to the North Atlantic Oscillation (NAO) modes observed in the North Atlantic, i.e., a cooling during the positive NAO mode associated with the MCA and a warming during the negative NAO mode correlating with the LIA (Fig. 4a, e, f, g; [60, 63, 105, 117]). Our IP25 sea ice record does not reflect such an anti-correlation in sea ice variability (Fig. 4a, c, d). It rather reflects a general increase in temporal sea ice extend parallel decreasing solar summer insolation associated with the Neoglacial cooling. The cooling culminates in the Little Ice Age. It should be noted that NAO reconstructions are still subject to discussion and the exact timing of late Holocene shifts in this atmospheric system are debated [35, 90, 91, 137].
We find evidence for the sensitivity of open water phytoplankton productivity and terrigenous input into Disko Bugt to the general oceanic circulation changes applying spectral analysis (Fig. 7). We identify a 63–90-year cycle in phytoplankton and terrigenous biomarker concentrations during the last 2.2 kyr (90% confidence; Fig. 7). In line with our findings, dinocyst reconstructions from Core 343310  and a study focussing on the driving mechanisms of sea ice formation changes off West Greenland in a diatom record  both identify a very similar cyclicity, i.e., 60–70 and 50–70 years, respectively. This cycle has previously been associated with the Atlantic Multidecadal Oscillation (AMO) frequency of ~ 50 to 70 years . In our IP25 sea ice record, we could not find evidence for such a cyclicity, which may be related to the relatively short sea ice season (spring) at the entrance of Disko Bugt during the investigated time period, whereas other phytoplankton biomarkers and microfossils reflect a longer season (spring, summer, and autumn; ).
AMO variability has been linked to solar activity changes . Changes in incoming radiation may influence sea ice extent and the Greenland Ice Sheet behaviour  and consequently affect the freshwater discharge/inflow  and nutrient availability to the area. A self-amplifying system may have caused the environmental changes observed in Disko Bugt area as follows: solar triggered Arctic sea ice melt  increases freshwater supply towards the North Atlantic causing a reduction in sub-polar gyre activity and AMO [41, 115] as described by Sha et al. . This may in turn cause distinct changes in WGC composition and meltwater supply from the Greenland Ice Sheet that affects phytoplankton blooms in West Greenland.
However, others argue for volcanic eruptions being the main trigger for the aforementioned sea ice feedback mechanism (e.g., [58, 76, 124]) that led to cold events, such as the LIA, during the late Holocene. It remains unclear whether there is a single mechanism that drives the sea ice extent in the Northern Hemisphere or if a combination of external and internal forcing’s and feedbacks, that may differ for certain time intervals, caused changes in atmospheric and oceanic systems that may have led to an expansion/reduction of sea ice extent during the late Holocene in Baffin Bay (cf. ).
Environmental reconstructions: biomarkers vs. microfossils
For Core 343310, a great variety of proxies have been compiled from the same sample set (for an overview see ), which provides ideal conditions for a direct proxy-to-proxy comparison that allows to gain further information about the correlation between microfossils and geochemical sea ice proxies and environmental factors controlling the signal formation (Fig. 8).
Regarding sea ice-related proxies, we find implications for a negative correlation between IP25 and the Atlantic/IC benthic foraminifera group (r2 = 0.54; Fig. 8a). The negative correlation can be explained by the oceanographic setting of Disko Bugt. The regional oceanography is controlled by the composition of the WGC, higher contributions of IC derived Atlantic waters, and associated subsurface warming may influence sea ice conditions, and, therefore, IP25 production. Furthermore, a stronger IC/weaker EGC contribution to the WGC off West Greenland has been associated with atmospheric and oceanic warming [97, 98, 117].
Sea ice diatoms have commonly been used for sea ice reconstructions along West Greenland (e.g., [50, 60, 63, 77]). Recently, Krawczyk et al.  found a correlation of sea-ice-associated diatoms to satellite-derived April sea ice conditions in surface sediments, collected along the West Greenland shelf. In our downcore record, IP25 and sea-ice-associated diatoms do not show a direct correlation (r2 = 0.12; Fig. 8b). In the previous studies, the diatoms included in the sea-ice-associated assemblage were only indirectly connected to the ice habitat, mainly by environmental conditions found close to sea ice, e.g., meltwater, low SSTs , whereas IP25 is produced by diatoms living inside the sea ice . These different habitats, different bloom seasons (spring-only, spring–autumn), and different factors controlling the abundance of the specific diatoms may cause the different signals recorded by the different proxies.
No correlation between PDIP25 and the MAT diatoms sea ice reconstruction (Fig. 8c; r2 = 0.01; ) could be found. Hence, we assume that this may be caused by different environmental factors controlling sea ice associated and IP25 producing diatoms. It may be possible that sea-ice-associated diatoms and IP25 producers reflect different seasonal signals (e.g., spring-only vs. spring–autumn) or underlie different deposition patterns, e.g., one species living mostly under the ice and another inside the ice.
The differences in the proxy records and lack of correlation, especially in regard to sea ice proxies, may be due to different seasons and annual time intervals reflected in each proxy. The relatively short sea ice season observed over the past 2.2 kyr in Disko Bugt area may cause our new biomarker record to reflect the rather short (peak winter/spring) annual time period. Microfossils, such as dinoflagellates, diatoms, and foraminifera, however, may represent a longer period (spring–early summer/autumn).
Besides ecological factors, issues such as selective and no uniform depositional and post-depositional degradation of biomarkers  and species-selective dissolution/preservation as well as under-representation of sea ice species affecting the diatom assemblages  have to be respected when comparing biomarkers, i.e., IP25, and diatoms.
Late Holocene sea ice conditions, open water phytoplankton productivity and terrigenous input in the southwest Disko Bugt area were reconstructed with a biomarker approach providing direct insights into sea ice variability and its driving mechanisms. Furthermore, a multiproxy comparison indicates different mechanisms behind phytoplankton and sea ice algae blooms and possibly different seasons reflected by different sea ice reconstructions.
We find that the Disko Bugt area was influenced by seasonal sea ice over the last 2.2 kyr BP. The overall sea ice trend indicates a development from a reduced sea ice cover during early spring, with sea ice algae productivity hampered by light availability to a gradual extend of the sea ice season from 1.2 kyr BP onwards. This change in sea ice extend is parallel to decreasing Northern Hemisphere atmospheric temperatures and culminates in the Little Ice Age around 0.2 kyr. We assume that modern conditions, with sea ice present until late spring and the presence of a stable ice edge at Disko Bugt, established around that time. A strong persistent Baffin Island Current and East Greenland Current may have contributed to this gradual increase in ‘Westice’ extent during the last 2.2 kyr BP.
The general trend in sea ice extent at the entrance of Disko Bugt, seems to follow the global the longer term atmospheric cooling trend in the Northern Hemisphere associated to decreasing solar insolation. Superimposed on longer term trends, phytoplankton productivity and terrigenous input show a strong short-term multidecadal periodicity at 63–90 years, indicating a rapid response to water mass properties such as salinity, and temperature. The observed cyclicity seems to be connected to the Atlantic Multidecadal Oscillation and to solar activity.
Financial support for this study was provided by the Deutsche Forschungsgemeinschaft through ‘ArcTrain’ (GRK 1904). We wish to thank the captain, crew and science party of the R/V Maria S. Merian expedition MSM 05/03. We would like to thank Walter Luttmer for laboratory support. Thanks to Simon Belt and colleagues (Biogeochemistry Research Centre, University of Plymouth) for providing the internal standard for IP25 analysis. Kerstin Perner has been funded through the DFG Grant PE2071/2–2. We thank two anonymous reviewers for the comments, which clearly led to an improved version of the original manuscript.
Compliance with ethical standards
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
- 2.Allan E, Vernal A, Knudsen MF, Hillaire-Marcel C, Moros M, Ribeiro S, Ouellet-Bernier M, Seidenkrantz M (2018) Late Holocene sea-surface instabilities in the Disko Bugt area, west Greenland, in phase with δ18O-oscillations at Camp Century. Paleoceanogr Paleoclimatology 1–17. https://doi.org/10.1002/2017PA003289 CrossRefGoogle Scholar
- 4.Andersen OGN (1981) The annual cycle of temperature, salinity, currents and water masses in Disko Bugt and adjacent waters, West Greenland. Comm Sci Res Greenl 5:33Google Scholar
- 7.Belt ST, Brown TA, Smik L, Tatarek A, Wiktor J, Stowasser G, Husum K (2017) Identification of C25 highly branched isoprenoid (HBI) alkenes in diatoms of the genus Rhizosolenia in polar and sub-polar marine phytoplankton. Org Geochem 110:65–72. https://doi.org/10.1016/j.orggeochem.2017.05.007 CrossRefGoogle Scholar
- 8.Belt ST, Cabedo-Sanz P, Smik L, Navarro-Rodriguez A, Berben SMP, Knies J, Husum K (2015) Identification of paleo Arctic winter sea ice limits and the marginal ice zone: optimised biomarker-based reconstructions of late Quaternary Arctic sea ice. Earth Planet Sci Lett 431:127–139. https://doi.org/10.1016/j.epsl.2015.09.020 CrossRefGoogle Scholar
- 11.Belt ST, Vare LL, Massé G, Manners HR, Price JC, MacLachlan SE, Schmidt S (2010) Striking similarities in temporal changes to spring sea ice occurrence across the central Canadian Arctic Archipelago over the last 7000 years. Quat Sci Rev 29(25–26):3489–3504. https://doi.org/10.1016/j.quascirev.2010.06.041 CrossRefGoogle Scholar
- 14.Bianchi TS (2007) Biogeochemistry of estuaries. Oxford University Press, New YorkGoogle Scholar
- 16.Boertmann D, Mosbech A, Schiedek D, Dünweber M (2013), Disko West. A strategic environmental impact assessment of hydrocarbon activities, Aarhus University, DCE—Danish Centre for Environment and Energy, pp. 306, Scientific Report from DCE—Danish Centre for Environment and Energy No. 71Google Scholar
- 21.Cabedo-Sanz P, Belt ST, Jennings AE, Andrews JT, Geirsdóttir Á (2016) Variability in drift ice export from the Arctic Ocean to the North Icelandic Shelf over the last 8000 years: A multi-proxy evaluation. Quat Sci Rev 146:99–115. https://doi.org/10.1016/j.quascirev.2016.06.012 CrossRefGoogle Scholar
- 24.Cavalieri DJ, Parkinson CL, Gloersen P, Zwally HJ (1996). Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I-SSMIS Passive Microwave Data, Version 1. updated yearly; accessed August 2017. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center, accessed A. https://doi.org/10.5067/8GQ8LZQVL0VLGoogle Scholar
- 25.Cavalieri DJ, Gloersen P, Parkinson CL, Comiso JC, Zwally HJ (1997). Observed hemispheric asymmetry in global sea ice changes. Science 278(5340), 1104–1106. from http://www.sciencemag.org/cgi/content/abstract/278/5340/1104 CrossRefGoogle Scholar
- 37.Harff J, Dietrich R, Endler R, Hentzsch B, Jensen JB, Krauss KA, Leipe N, Lloyd T, Mikkelsen J, Perner NM,M, Richter K, Risgaard-Petersen A, Rysgaard N, Richter S, Sandgren T, Sheshenko P, Snowball V, Waniek I, Weinrebe J, Witkowski WA (2007) Deglaciation history, Coastal Development, and environmental change during the Holocene in western Merian’, Greenland. Cruise report MSM 05/03 R/V “Maria S. Merian.” Baltic Sea Research Institute, RostockGoogle Scholar
- 41.Holland, M., Bitz, C., Eby, M., Weaver, A. (2001). The role of Ice-Ocean interactions in the variability of the North Atlantic thermohaline circulation. J Clim 14, 656–675. https://doi.org/10.1175/1520-0442(2001)014<0656:TROIOI>2.0.CO;2 CrossRefGoogle Scholar
- 47.Jakobsson M, Macnab R, Mayer L, Anderson R, Edwards M, Hatzky J, Schenke HW, Johnson P (2008) An improved bathymetric portrayal of the Arctic Ocean: Implications for ocean modeling and geological, geophysical and oceanographic analyses. Geophys Res Lett. https://doi.org/10.1029/2008GL033520 CrossRefGoogle Scholar
- 51.Jensen LM, Christensen TR (2014). Nuuk Ecological Research Operations. In: 7 th Annual Report. Aarhus University, DCE—Danish Centre for Environment and Energy, p 94Google Scholar
- 53.Johns L, Wraige EJ, Belt ST, Lewis CA, Massé G, Robert JM, Rowland SJ (1999) Identification of a C25 highly branched isoprenoid (HBI) diene in Antarctic sediments, Antarctic sea-ice diatoms and cultured diatoms. Org Geochem 30:1471–1475. https://doi.org/10.1016/S0146-6380(99)00112-6 CrossRefGoogle Scholar
- 54.Juul-Pedersen T, Arendt KE, Mortensen J, Krawczyk D, Rysgaard S, Retzel A (2014), Nuuk Basic: The MarineBasis programme, in Nuuk Ecological Research Operations, 7th Annual Report, 2013, edited by L. MGoogle Scholar
- 61.Krawczyk DW, Witkowski A, Moros M, Lloyd JM, Høyer JL, Miettinen A, Kuijpers A (2017) Quantitative reconstruction of Holocene sea ice and sea surface temperature off West Greenland from the first regional diatom data set. Paleoceanography 32(1):18–40. https://doi.org/10.1002/2016PA003003 CrossRefGoogle Scholar
- 62.Krawczyk DW, Witkowski A, Wroniecki M, Waniek J, Kurzydłowski KJ, Płociński T (2012). Reinterpretation of two diatom species from the West Greenland margin—Thalassiosira kushirensis and Thalassiosira antarctica var. borealis—hydrological consequences. Mar Micropaleontol. https://doi.org/10.1016/j.marmicro.2012.02.004 CrossRefGoogle Scholar
- 64.Kremer A, Stein R, Fahl K, Ji Z, Yang Z, Wiers S, Matthiessen J, Forwick M, Löwemark L, O’Regan M, Chen J, Snowball I (2018) Changes in sea ice cover and ice sheet extent at the Yermak Plateau during the last 160 ka—reconstructions from biomarker records. Quat Sci Rev 182:93–108. https://doi.org/10.1016/j.quascirev.2017.12.016 CrossRefGoogle Scholar
- 65.Lamb HH (1977) Climate: present, past and future. Vol 2. Methuen and Co. Ltd., LondonGoogle Scholar
- 79.Moros M, Jansen E, Oppo DW, Giraudeau J, Kuijpers A (2012). Reconstruction of the late-Holocene changes in the Sub-Arctic Front position at the Reykjanes Ridge,north Atlantic. Holocene 22(8):877–886. http://www.scopus.com/inward/record.url?eid=2-s2.0-84863557938&partnerID=40&md5=75bf62fce6e5b761b6b2f5a2fa104ec5 CrossRefGoogle Scholar
- 80.Moros M, Lloyd JM, Perner K, Krawczyk D, Blanz T, de Vernal A, Jansen E (2016) Surface and sub-surface multi-proxy reconstruction of middle to late Holocene palaeoceanographic changes in Disko Bugt, West Greenland. Quat Sci Rev 132:146–160. https://doi.org/10.1016/j.quascirev.2015.11.017 CrossRefGoogle Scholar
- 83.Müller J, Wagner A, Fahl K, Stein R, Prange M, Lohmann G (2011) Towards quantitative sea ice reconstructions in the northern North Atlantic: a combined biomarker and numerical modelling approach. Earth Planet Sci Lett 306(3–4):137–148. https://doi.org/10.1016/j.epsl.2011.04.011 CrossRefGoogle Scholar
- 89.NSIDC, Boulder. http://nsidc.org/ (accessed August 2017)Google Scholar
- 101.Prahl FG, Muehlhausen LA (1989) Lipid biomarkers as geochemical tools for paleoceanographic study. Prod Ocean Present Past 27:1–289Google Scholar
- 104.Ribeiro S, Sejr MK, Limoges A, Heikkilä M, Andersen TJ, Tallberg P, Weckström K, Husum K, Forwick M, Dalsgaard T, Massé G, Seidenkrantz MS, Rysgaard S (2017) Sea ice and primary production proxies in surface sediments from a High Arctic Greenland fjord: spatial distribution and implications for palaeoenvironmental studies. Ambio 46:106–118. https://doi.org/10.1007/s13280-016-0894-2 CrossRefGoogle Scholar
- 106.Ribergaard MH, Kliem N, Jespersen M (2006) HYCOM for the North Atlantic Ocean with Special Emphasis onWest GreenlandWater. Technical Report 06 – 0. http://www.dmi.dk/dmi/tr06-07
- 109.Risebrobakken B (2003). A high-resolution study of Holocene paleoclimatic and paleoceanographic changes in the Nordic Seas. Paleoceanography 18(1):1017. http://www.agu.org/pubs/crossref/2003/2002PA000764.shtml CrossRefGoogle Scholar
- 115.Schmith T, Hansen C (2003) Fram strait ice export during the nineteenth and twentieth centuries reconstructed from a multiyear sea ice index from southwestern Greenland. J Clim 16(16):2782–2791. https://doi.org/10.1175/1520-0442(2003)016<2782:FSIEDT>2.0.CO;2 CrossRefGoogle Scholar
- 117.Seidenkrantz MS, Roncaglia L, Fischel A, Heilmann-Clausen C, Kuijpers A, Moros M (2008) Variable North Atlantic climate seesaw patterns documented by a late Holocene marine record from Disko Bugt, West Greenland. Mar Micropaleontol 68(1–2):66–83. https://doi.org/10.1016/j.marmicro.2008.01.006 CrossRefGoogle Scholar
- 122.Sha L, Jiang H, Seidenkrantz MS, Knudsen KL, Olsen J, Kuijpers A, Liu Y (2014) A diatom-based sea-ice reconstruction for the Vaigat Strait (Disko Bugt, West Greenland) over the last 5000 year. Palaeogeogr Palaeoclimatol Palaeoecol 403:66–79. https://doi.org/10.1016/j.palaeo.2014.03.028 CrossRefGoogle Scholar
- 125.Smik L, Cabedo-Sanz P, Belt ST (2016) Semi-quantitative estimates of paleo Arctic sea ice concentration based on source-specific highly branched isoprenoid alkenes: a further development of the PIP25 index. Org Geochem 92:63–69. https://doi.org/10.1016/j.orggeochem.2015.12.007 CrossRefGoogle Scholar
- 126.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(3–4):319–334. https://doi.org/10.1016/j.quascirev.2003.06.003 CrossRefGoogle Scholar
- 129.Stein R, Fahl K (2012) A first southern Lomonosov Ridge (Arctic Ocean) 60 ka IP25 sea-ice record. Polarforschung 82:83–86Google Scholar
- 130.Stein R, Fahl K, Müller J (2012) Proxy reconstruction of Cenozoic Arctic Ocean sea-ice history—from IRD to IP25-. Polarforschung 82(1):37–71Google Scholar
- 132.Stein R, Fahl K, Schade I, Manerung A, Wassmuth S, Niessen F, Nam S, Il (2017) Holocene variability in sea ice cover, primary production, and Pacific-Water inflow and climate change in the Chukchi and East Siberian Seas (Arctic Ocean). J Quat Sci 32:362–379. https://doi.org/10.1002/jqs.2929 CrossRefGoogle Scholar
- 133.Stein R, Fahl K, Schreck M, Knorr G, Niessen F, Forwick M, Lohmann G (2016). Evidence for ice-free summers in the late Miocene central Arctic Ocean. Nat Commun, 7:11148. http://www.nature.com/ncomms/2016/160404/ncomms11148/full/ncomms11148.html CrossRefGoogle Scholar
- 147.Weidick A, Bennike O, Grafisk S (2007) Quaternary glaciation history and glaciology of Jakobshavn Isbrae and the Disko Bugt region, West Greenland: a review. Geol Surv Den Greenl Bull 14:26–49Google Scholar
- 155.Yruela I, Barbe A, Grimalt JO (1990) Determination of double bond position and geometry in linear and highly branched hydrocarbons and fatty acids from gas chromatography-mass spectrometry of epoxides and diols generated by stereospecific resin hydration. J Chromatogr Sci. https://doi.org/10.1093/chromsci/28.8.421 CrossRefGoogle Scholar