From land to sea: provenance, composition, and preservation of organic matter in a marine sediment record from the North‐East Greenland shelf spanning the Younger Dryas–Holocene

The organic matter content of marine sediments is often used to infer past changes in ocean conditions. However, the organic carbon pool preserved in coastal sediments is a complex mixture derived from different sources and may not reflect in situ processes. In this study, we combine taxonomic identification of reworked palynomorphs with pyrolysis organic geochemistry and reflected‐light organic petrographic microscopy to investigate the provenance, composition and preservation of organic matter in a marine sediment core retrieved from the NE Greenland shelf. Our study reveals continuous yet variable input of land‐derived organic carbon to the marine environment throughout the late Younger Dryas–Holocene, with the highest input of inert carbon in the late Younger Dryas. Although the sediments contain some recent marine palynomorphs, there is no other evidence of fresh marine organic carbon. In contrast, our results indicate that these shelf sediments represent a significant sink of recycled organic carbon. The results of pyrolysis geochemistry revealed that ~90% of the total organic carbon in the sediments is inert. The organic petrography analyses revealed that >70–84% of the organic carbon in the sediment core is terrigenous. Reworked dinoflagellate cysts showed a continuous provenance of Cretaceous land‐derived material, most likely from the nearby Clavering Island. Our study points to the importance of constraining the organic matter origin, composition and preservation in marine sediments to achieve more accurate palaeoenvironmental reconstructions based on organic proxies.

. It is expected that this will also lead to an increase in the export of old terrigenous organic carbon to marine shelves (Guo & Macdonald 2006).
On time scales of centuries to millennia, biological responses play a key role in the carbon cycle (Sarmiento & Bender 1994).Investigations of past changes in the carbon cycle, ocean conditions and palaeoproductivity are based on analyses of proxy data from marine geological archives.Among these proxy data, total organic carbon (TOC) is often used as an indicator of overall changes in productivity, and organic proxies such as biomarkers are often normalized against TOC contents in sediments (e.g.Belt & M€ uller 2013).However, organic carbon preserved in seafloor sediments constitutes a complex mixture derived from different sources.The organic carbon may be locally produced at the time of deposition (autochthonous) or may be transported to the burial site (allochthonous).Moreover, the allochthonous organic carbon may be of very different origins, ranging from still relatively fresh to old organic carbon eroded from deposits on land.
Furthermore, the organic carbon may have undergone varying degrees of microbial degradation and alteration (Hedges & Keil 1995;Carrie et al. 2012).Therefore, it is challenging to interpret the organic carbon dynamics of a marine system from TOC data alone.Terrestrial material in marine sediments provides information on transport pathways but also on the origin of the material.Among land-derived organic matter, fossil dinoflagellate cysts are abundant in marine deposits.When organic carbon from ancient marine sediments is reworked and preserved, it bears a fingerprint of the time of deposition that may prove vital in assessing its provenance.
Pyrolysis organic geochemistry has long been applied to characterize organic matter in recent sediments (Fahl & Stein 1999;Sanei et al. 2000Sanei et al. , 2001Sanei et al. , 2005;;Carrie et al. 2012).Two recent studies (Strunk et al. 2020;Hansen et al. 2022) have applied the pyrolysis method on Holocene Arctic sediments to investigate the offset of radiocarbon ages derived from bulk sediment, both concluding that a more accurate estimate of the age offsets can be reached when applying pyrolysis organic geochemistry.The pyrolysis method consists of stepwise increases in temperature over time, degrading the more labile organic matter first.It gives an estimate of the TOC as well as information on various hydrocarbon fractions and their degree of degradation (Espitali e et al. 1977;Liebezeit & Wiesner 1990).
Reflected-light organic petrography, measured on organic particles named maceral fragments, estimates the thermal maturity of the preserved carbon from vitrinite reflectance (%VRo), and can furthermore be used as a quantitative way of assessing the abundance of terrigenous and marine macerals in marine sediments with reference to their biological and environmental sources (Boucsein & Stein 2000).Combining pyrolysis with reflected-light organic petrography data can thus, in addition to quantifying preservation, also give indications of a potential provenance.
This study combines an investigation of reworked dinoflagellate cysts and acritarchs with pyrolysis analyses and reflected-light organic petrography to assess the provenance, composition and preservation of organic matter in a well-dated marine sediment record.The marine sediment core (DA17-NG-ST14-171G) was retrieved at 74°N on the NE Greenland shelf and spans the late Younger Dryas to present (12.18 to À0.042 ka BP), a period of known retreat of the nearby Greenland Ice Sheet margin (Funder et al. 2021).Both the temporal coverage of the core and its unique location within the modern-day Sirius Water Polynya are of particular interest for capturing the contribution of landderived organic matter inputs to marine sediments deposited on the shelf.
This work builds upon a previous multiproxy study (Jackson et al. 2022), which includes palynological, micropalaeontological, sedimentological and biogeochemical analyses carried out on the same sediment core.From foraminiferal and dinoflagellate cyst assemblages, a change in ocean circulation across the Younger Dryas/ Early Holocene transition was detected, and a highly stratifiedwater column with a higher-than-present influx of warm, Atlantic-sourced bottom waters under a retreating ice shelf was inferred for the late Younger Dryas.A cooling of bottom waters was reconstructed for the earliest part of the Holocene, following the ice shelf retreat.The onset of the Sirius Water Polynya from ~10 to 8.7 ka BP was linked to higher diatom production (biogenic silica) and from the assemblage composition of dinoflagellate cysts and foraminifera.Based on C:N and the isotopic composition of organic carbon (d 13 C org ), the TOC was attributed mainly to terrigenous inputs.The aim of this study is thus to provide a further investigation into the origin, preservation and composition of the organic matter, characterising land-to-sea fluxes and partitioning the autochthonous (landderived) input of organic carbon to this area of the NE Greenland shelf throughout the Younger Dryas-Holocene.

Regional setting
The study site is influenced by freshwater runoff from the NE Greenland Ice Sheet (Khan et al. 2014;Stroeve et al. 2014), the cold and ice-loaded East Greenland Current, and the underlying Atlantic Intermediate Water (Fig. 1A; Strass et al. 1993).The Atlantic Intermediate Water is a combination of Arctic Atlantic Water and the relatively warm and saline Recirculating Atlantic Water.The site lies in the Sirius Water Polynya, a wind-driven shelf-and latent heat polynya.Latent heat polynyas are driven by ice motion owing to winds or oceanic currents (Maqueda et al. 2004).Sea ice formed during the freezing Nanna Andreasen et al.

BOREAS
Organic matter in marine sediment from the NE Greenland shelf period is transported away, resulting in brine formation, which in turn controls the mixing of the Young Sound water masses (Dmitrenko et al. 2015;Boone et al. 2018).The Sirius Water is one of three frequently recurring polynyas off the East Greenland coast, together with the Northeast Water and Scoresby Sound polynyas (Pedersen et al. 2010).
The geology of the surrounding area is well known and consists of sedimentary rocks from the Cretaceous, Jurassic, Carboniferous and Upper Permian as well as Palaeogene basalt and allochthonous Caledonian thrust sheets from the Neoproterozoic (Fig. 1C; Surlyk 1978;Stemmerik et al. 1993).Clavering Island and Wollaston Foreland are situated at the mouth of Young Sound (Fig. 1B, C).The hinterland of Young Sound is dominated by sedimentary rocks from the Jurassic and Cretaceous periods.The Cretaceous sediments include 12 formations with a wide variation in depositional environments, ranging from marine mudstones (e.g.Stratumbjerg and Fosdalen formations) to deep marine sandstones (e.g.Østersletten Formation) (Bjerager et al. 2020).Nøhr-Hansen (1993) conducted a comprehensive stratigraphic investigation of the Lower Cretaceous marine sedimentary deposits in NE Greenland and later described the Lower and Upper Cretaceous dinoflagellate cyst stratigraphy of NE Greenland (Nøhr- Hansen et al. 2020).
The Jurassic sediments in NE Greenland include 25 formations with a wide variation in depositional environments ranging from marine mudstones (e.g.Bernbjerg Formation) to deep marine sandstones (e.g.Laugeites Ravine Member) (Surlyk et al. 2021).The Jurassic sediments are most abundant on the northern side of the Young Sound fjord on Wollaston Foreland and Kuhn Island (Fig. 1C), where they are found alongside Cretaceous sediments.South of the Young Sound fjord, on Clavering Island, the sediments are mainly of Cretaceous age.The total areas of exposed Jurassic and Cretaceous sediments are comparable (Fig. 1C).Both the Jurassic and Cretaceous sediments include marine dinoflagellate cysts, which may be used to track the provenance and degree of potential reworking of these sediments.No Jurassic or Cretaceous sediments have been identified north of 75°N (Surlyk 1978;Stemmerik et al. 1993), and thus it is unlikely that any Mesozoic dinoflagellate cysts found in Quaternary marine sediments such as core DA17-NG-ST14-171G can have a northern origin being transported by southward ocean currents.

Sediment core material
The 420-cm-long gravity core DA17-NG-ST14-171G (74°5.413 0N; 19°25.862 0W, 341 m water depth; hereafter referred to as 171G) was retrieved in September 2017 onboard RV 'Dana' during the NorthGreen Expedition (Seidenkrantz et al. 2017).The geochronology of the core is based on 10 benthic foraminiferal 14 C dates, which were calibrated using Marine20 (Heaton et al. 2020) and a reservoir correction of DR = À10AE60 years (previously 140AE60 years using Marine13; McNeely & Brennan 2005;Reimer et al. 2013).One 210 Pb date from the topmost centimetre completes the age model.When the age model is extrapolated from 402.5 cm (deepest date) to 420.5 cm (bottom of the core), the estimated age span is 12.18 to À0.042 ka BP, with a condensed, lowresolution section from ~7.6 to À0.042 ka BP (Jackson et al. 2022).The extremely low sedimentation rate in the top 30 cm of the core (~7.6 to À0.042 ka BP) is discussed in Jackson et al. (2022), and the data points in this interval are treated with caution.
Previous investigations of this sediment record included analysis of total organic carbon (TOC) and nitrogen content measured by Dumas combustion on an elemental analyser, as well as isotopic ratios of d 13 C org and d 15 N on 53 samples (Jackson et al. 2022).Core imaging, radiography, X-ray fluorescence scanning, and biogenic silica, foraminiferal, dinoflagellate cyst and diatom analyses are also reported in Jackson et al. (2022).

Palynological analysis
Palynological analyses were conducted at the Geological Survey of Denmark and Greenland (GEUS) on 53 samples from core 171G, representing an 8-cm resolution.For each sample, 2-3 g of freeze-dried sediments were processed following standard methods (Mertens et al. 2009).A Lycopodium-spore tablet of approximately 10 000 spores (GEUS Batch no.110118351) was added to each sample prior to processing.The samples were then treated with acid to remove carbonates and silicates.This treatment consisted of adding 2M hydrochloric acid at room temperature (overnight) for carbonate removal, followed by the removal of silicates with hydrofluoric acid (40%) at room temperature for up to 48 h.Additional hydrochloric acid treatment for 72 h was performed when needed for the removal of fluorosilicates (Mertens et al. 2009).The samples were sieved through an 11 lm nylon mesh and rinsed with distilled water until neutralized.The sample residues were swirled and mounted on a microscope slide with glycerol gelatine using a heating plate and sealed with paraffin wax.
The microscope slides were examined under anupright light microscope (Olympus BX51/BX60) with differential interference contrast optics at 4009 and 10009 magnification.Organic-walled dinoflagellate cysts, Halodinium spp., foraminiferal linings, bisaccate and non-bisaccate pollen grains, along with reworked palynomorphs (dinoflagellate cysts, acritarchs, bisaccate pollen and spores) were counted.Of the reworked palynomorphs, dinoflagellate cysts were the most abundant group.There was only one identified acritarch, which is counted as part of the reworked palynomorphs.Reworked pollen and spores were rare and were therefore excluded from further taxonomic identification.When possible, the reworked dinoflagellate cysts were identified to genus or species level following Nøhr- Hansen (1993) and Nøhr-Hansen et al. (2020).All extant dinoflagellate cyst taxa were identified to the lowest possible taxonomic level.A minimum of 100 recent dinoflagellate cysts were counted per slide, excluding the lowermost five core samples, which were nearly barren.It was not possible to reach the recommended minimum of 300 counts per slide (Mertens et al. 2009) owing to generally very low cyst concentrations in the sediments.Fluxes (cysts cm 2 a À1 ) were calculated from the concentration of dinoflagellate cysts (g À1 ) and the mass accumulation rate (g cm À2 a À1 ), therefore, taking the large changes in sedimentation rate into account.The sedimentation rates and fluxes are not presented for the lowermost 18 cm of the core (including three samples), as no dates could be obtained for this interval.

Pyrolysis analyses
Pyrolysis analysis of 30 samples (~50 mg of sediment each) were performed at the Lithospheric Organic Carbon laboratory, Department of Geoscience, Aarhus University.The HAWK Pyrolyser and TOC instrument, Wildcat Technologies, was used following the Rock-Eval 6 â method (Lafargue et al. 1998) combined with the Slice&Dice â peak integration method.Measurements were conducted on core 171G at 8-16 cm resolution for core depths between 416.5 and 32.5 cm and at 4 cm resolution for core depths between 32.5 and 0.5 cm.
The method follows a two-step procedure.The first step is pyrolysis in an inert atmosphere (helium), followed by combustion in an oxic atmosphere (Carrie et al. 2012;Rudra et al. 2021).The heating during the pyrolysis stage causes the organic bonds to break, which results in peaks of S0, S1, S2 and S3, corresponding to the labile organic matter (Fig. S1).In the pyrolysis phase, a low-temperature purge cycle reaches the initial pyrolysis temperature (300 °C), creating the S0 peak.The sample is automatically inserted into the oven, preheated at 300 °C for 3 min, releasing hydrocarbons and creating the S1 peak.The temperature is then ramped at 25 °C min À1 until 650 °C (creating the S2 peak), measured with a flame ionization detector (Rudra et al. 2021).S3 is measured continuously by online infrared detectors during the pyrolysis.Then follows the combustion of residual organic carbon in the oxidation stage, where CO and CO 2 are released and produce the fourth peak, S4 (Carrie et al. 2012).The oxidation phase begins at 300 °C with a ramp of 25 °C min À1 until reaching 850 °C.Combustion of residual organic carbon releases CO and CO 2 (S4: mg CO-CO 2 g À1 ) measured by online infrared detectors (Lafargue et al. 1998;Sanei et al. 2005;Carrie et al. 2012; Fig. S1).
The TOC (wt%) is estimated from the sum of all the organic matter fractions released during pyrolysis (pyrolysable carbon) and oxidation (residual carbon) (Lafargue et al. 1998;Sanei et al. 2005).To prevent the mineral carbon from interfering, the mineral carbon can be determined by the addition of the released CO 2 from the pyrolysis above 400 °C and the decomposition of carbonate during the oxidation phase from 650 to 850 °C (Lafargue et al. 1998;Carrie et al. 2012).From the pyrolysis data, the S1/(S1+S2) fractions are calculated.The calculation of S1/(S1+S2) gives the relation between S1 and the less labile S2.This helps to assess the preservation of organic matter and can be helpful when attributing changes in organic matter to environmental variation vs. preservation issues.Furthermore, the organic matter fractions determine the proportional hydrogen/carbon (H/C) ratio, or hydrogen index (HI; S2/ TOC 9 100) and oxygen index (OI; S3/TOC 9 100), which determines the relative proportion of O/C in the organic matter (Lafargue et al. 1998;Carrie et al. 2012).
The degree of lability of organic matter is defined based on the temperature at which different hydrocarbon fractions are released.The Slice&Dice â method slices the pyrogram during pyrolysis into three fractions and classifies these as labile A, B and C, in order of decreasing lability.Empirical data have shown that the first three labile fractions, A, B, and C, are also soluble in organic acids (Carrie et al. 2012;Souc emarianadin et al. 2018).Therefore, these fractions are considered soluble organic carbon (SOC) (Fig. S1).The hydrocarbons released above 400 °C are considered non-soluble, and hence, they constitute the labile portion of the particulate organic carbon (POC) (Carrie et al. 2012;Souc emarianadin et al. 2018).
The labile-A SOC represents the hydrocarbons from the ultra-labile fraction of organic matter released as a pyrolysis product of S0 (mg HC g À1 ) at 100 °C.The labile-B SOC represents the readily degradable, hydrogen-containing organic carbon released as the pyrolysis product of S1 (mg HC g À1 ) at 300 °C.The labile-C SOC represents the organic carbon concentration of the hydrogen-containing organic carbon and corresponds to the left shoulder of the S2 peak (S2a) released below 400 °C.The labile POC represents the hydrogen-containing organic carbon released during the pyrolysis temperature ramp from 400 to 650 °C and corresponds to the second S2 peak (S2b).This fraction is considered labile because of its hydrocarbon composition and its ability as a proton donor.However, the labile POC fraction is considered insoluble in organic acid and hence tied to particulate organic matter.As an example, biomacromolecules forming algal cell walls are part of the labile POC fraction.The inert POC includes the residual organic carbon fraction as well as highly refractory inertinite organic matter (i.e.char, highly reworked or degraded organic matter) released during oxidation.

Reflectance analysis
Six samples were selected for organic petrography analysis.The samples were chosen to capture the variation within the palynomorph and pyrolysis data throughout the core at the following core depths: 16.5, 144.5, 200.5, 240.5, 368.5 and 400.5 cm.Dried sediment was embedded in a cold-setting epoxy-resin mixture such that it covered a circular surface with a diameter of 2 cm.Samples were subsequently polished at GEUS.Organic petrographic analyses were conducted using white incident light and a fluorescence-inducing blue light Zeiss Axio Image II microscope equipped with the Discus-Fossil system (Hilgers, K€ onigswinter, Germany) at the Lithospheric Organic Carbon Laboratory, Aarhus University.Random reflectance (Ro%) was measured on vitrinite macerals (vitrinite reflectance; VRo%) in white incident light, oil immersion (509 objective).The microscope was calibrated against a KB N-LASF standard (Ro = 1.317%).Fluorescence microscopy was conducted under an excitation bandpass of 365/12 nm and an emission long pass of 397 nm (Rudra et al. 2021).
Reworked palynomorphs were present in all studied samples and identified to genus or species level, whenever possible (Fig. 2; Table S1).Their stratigraphic range was found for both global occurrence and the known occurrences on the east and NE Greenland coast (Table 1; Nøhr- Hansen 1993Hansen , 2012;;Williams et al. 2017, and references therein;Nøhr-Hansen et al. 2020).Taxawith a broad stratigraphical range, such as Oligosphaeridium spp.and Spiniferites spp., were identified and counted but were omitted from the stratigraphic chart (Table 1).The large majority of the identified cysts are of Cretaceous age, and only one species was restricted to the Upper Cretaceous (Table 1).While one species has its first occurrence in the Jurassic, it is also common in the Cretaceous (Williams & Bujak 1985).Thus, the stratigraphic distribution of the reworked dinoflagellate cysts identified in core 171G strongly suggests a predominately Early Cretaceous age, with some Late Cretaceous occurrences.
The relative abundances and fluxes of reworked palynomorphs were variable throughout the core.In particular, the five lowermost samples of the core (12-11.7 ka BP, D1) have a high relative abundance of reworked dinoflagellates (from 55.5 to 100%; Fig. 3C,  zone D1).Here it is important to note that these samples were almost barren of recent dinoflagellate cysts, and the sediments include coarser grain sizes (Ice Rafted Debris, IRD).When focusing on zones D2, D3 and D4, the number of reworked dinoflagellates varies from 3 to 26.5%.Reworked palynomorphs are particularly high in these time intervals: 12.2-11.7 (D1) ka BP, 9.9-9.6 ka BP (within D3) and 9.0-8.5 ka BP (within D3).

Geochemistry
The pyrolysis analyses revealed a large amount of inert carbon and mineral carbon in the sediments (Fig. 4A, B).The inert organic carbon accounts for roughly 90% of the total organic carbon for the entire record, leaving only about 10% of reactive organic carbon (Fig. S2A).As such, these marine sediments contain up to 1.4 wt% carbon (Fig. S3), of which only 0.61-1.08wt% is organic, and only 0.05-0.15wt% is reactive.The lowermost measurement (12.1 ka BP, depth 416.5 cm) is particularly low in the more easily degradable carbon and consists almost entirely of carbon released during oxidation (O-C) and inert organic carbon (Fig. 4A), as well as a high mineral carbon (0.53 wt%) content (Fig. 4B).
The TOC and total carbon content determined in this study from pyrolysis measurements are plotted together with TOC determined from Dumas combustion as reported in Jackson et al. (2022) (Fig. 4C).Variations in TOC are seen throughout the core; however, the causes of these variations are uncertain.The pyrolysis method estimated TOC values to range from 0.61 to 1.08 wt% (Fig. 4), while according to the CS analyser, the TOC ranges from 0.7 to 1.8 wt% (Jackson et al. 2022).The TOC values from the CS analyser show a greater range of values.These differences between the two methods point to either an underestimation of TOC by the pyrolysis method or an overestimation of TOC by the CS analyser.The underestimation of the TOC in the pyrolysis method canvery likely be due to the difference in thermal regimes, as the pyrolysis reaches a maximum temperature of 850 °C and the CS combustion reaches a maximum of ~1700 °C.Although unlikely, we cannot fully exclude In marine sediments, the hydrogen-containing organic carbon fraction (S2, mg HC g À1 sediments) represents mostly a labile, autochthonous fraction.S2 is the sum of the labile-C SOC and labile POC fractions.The depth profile trend of S2 follows that of TOC (pyrolysis).A general increase is seen in zones D1 and D2.The increase continues in D3, where it is generally higher in the middle of the zone.However, S2 shows a different trend than TOC (pyrolysis) in D4, where a small increase in S2 is seen at the top (Fig. 4D).S2 is normalized to TOC (HI, Fig. S2C) and shows a similar trend to S2.The calculated S1/(S1+S2) is relatively stable and varies from 0.25 to 0.9 throughout the core, indicating no major changes in preservation (Fig. S2E).
Most samples plot as terrigenous and reworked organic matter with generally low S2 values relative to TOC (type III and IV organic matter, Fig. 5; Vandenbroucke & Largeau 2007).The regression line intercepts TOC at 0.36 wt%; this value represents the amount of inert organic carbon in the system before the production of autochthonous/algal organic matter.When plotting the hydrogen index vs.oxygen index in a 'van-Krevelevntype' diagram (Fig. S4), all data points plotted in 'IVreworked organic matter' with a hydrogen index lower than 100.

Reflectance analysis
Macerals are typically divided into groups of inertinite, liptinite and vitrinite, which are defined by their reflectance and appearance.Here the appearance was not assessed thoroughly, and the macerals are grouped according to reflectance values.Histograms of the six samples (Fig. S5) show a similar pattern throughout the core, and the reflectance measurements are divided into three groups based on the deconvolution of the %VRo distribution histogram within the context of the three representative thermal maturity ranges: Group 1 (representing immature vitrinite at 0.3-0.6%VRo),Group 2 (representing the wide range of catagenesis from the early oil window to the dry gas window at 0.6-1.3%VRo),and Group 3 (most likely the carbonized or highly mineralized organic matter/inertinite >1.3 %VRo) (Tissot & Welte 1984).In fluorescent light, the grain particles point to poor sorting showing large sandy particles transported and sedimented along with silt and clay fractions (Fig. 6A, B).Organic matter is predominantly nonfluorescing (Figs 6, S6E), which is attributed to the highly degraded nature of the organic matter and is in agreement with the low hydrogen index.The reflectance measurement in Group 1 represents 16-30.4% of the total reflectance, Group 2 represents 51.5-71.0%and Group 3 represents 11.1-27.2%(Fig. 7).

Discussion
Marine sediments are generally considered an important carbon reservoir (Barber et al. 2017).The organic matter content and composition of marine sediments deposited over time can provide a valuable archive of past changes in ocean conditions, and organic proxies are routinely used in palaeoenvironmental reconstructions.However, in near-shore shelf environments, the organic matter preserved in sediment records often contains a mixture of terrigenous (allochthonous) and marine (autochthonous) sources that require careful consideration (Khan et al. 2014;Stroeve et al. 2014;Syring et al. 2020a).Here a significant terrigenous source to the core site was found, and the contribution of land-derived organic matter and its preservation in the marine record is discussed along with the possible implications of our findings for palaeoenvironmental studies.

Provenance of reworked organic matter
The reworked palynomorphs found throughout the core were identified as Cretaceous dinoflagellate cysts As the Jurassic sediments of the region are known to contain dinoflagellate cysts (Surlyk et al. 2021), it would have been expected that some of these cysts would be identified if Jurassic sediments were a significant source area; however, this is not the case, and the palynology clearly indicates a single-source: Cretaceous sediments.
Our reflected-light organic petrography analyses revealed a narrow population of a low reflectance group, Group 1 (0.3-0.6%VRo), which comprises the only wellpreserved source of immature vitrinite that can be used for source determination.The %VRo represents the thermally immature coal or sedimentary organic matter deposits buried shallower than 2 km and exposed to a maximum burial temperature of below 60 °C.The Group 2 higher reflectance could potentially be derived from degradation during transport or intense oxidation/ degradation within the water column and during precipitation (Taylor et al. 1998).Therefore, this group cannot be reliably used for source tracing.Group 3 is attributed to inertinite, which does not necessarily relate to maximum geological burial and hence is excluded from the provenance discussion (Taylor et al. 1998).
Sedimentary rocks from the Jurassic and the Cretaceous periods dominate the Young Sound hinterland.Piasecki et al. (2020) found sediments of the Lower Cretaceous to be immature, corresponding to a value of <0.6%VRo.Petersen & Vosgerau (1999) reported the immature reflectance of 0.30-0.53%VRofor the Middle Jurassic coal-bearing Muslingebjerg Formation of Hochstetter Foreland and Kuhn Island, respectively.Vitrinite reflectance measurements on samples from Hochstetter Foreland (Bojesen-Koefoed et al. 1996) revealed values of 0.37-0.44%VRo.The published %VRo reflectance is the value of the lowest reflectance group, therefore, representing the least altered in situ vitrinites.
The low reflectance measurements found throughout the core are consistent with both the Jurassic and the Cretaceous hinterland.However, while the reflectance measurement is linked to regional maturity, the stratigraphic distribution of the palynomorphs is better constrained in time.Thus, the palynomorphs, supported by the reflectance analysis, suggest that the Cretaceous outcrops in the Young Sound hinterland formed a major sedimentary source.The Cretaceous outcrops are found alongside Jurassic sediments on Clavering Island, Wollaston Foreland and Kuhn Island (Fig. 1C).However, on Clavering Island, the sediments are mainly of Cretaceous age.As no mix of Jurassic and Cretaceous cysts is found, we attribute the provenance of the allochthonous material to Cretaceous sediments on Clavering Island.

Organic carbon composition
The TOC values found in this study ranged from 0.61 to 1.08 wt% (pyrolysis).Although these would be globally considered rather low values, they are comparable with those reported from modern Arctic sediments.In Birgel et al. (2004), the distribution of TOC in surface sediments from the Norwegian-Greenland Sea and Fram Strait ranges from 0.5 to 1.5%.There is a general trend of high TOC values towards the eastern part of the Fram Strait.
On the contrary, towards the west, lower values (<0.75%) correspond to the area of permanent ice cover except for a local maximum with TOC values >1.25% appearing in the region of the Northeast Water Polynya and reflecting higher marine productivity (Birgel et al. 2004).As 171G is retrieved at 74°N and within the modern-day Sirius Water Polynya, it would be expected that TOC values are higher than for other shelf sites further north and/or not under the influence of polynya formation.Indeed, TOC values reported for a marine sediment core at 79N (PS100/270) and spanning the Holocene were <0.1-0.5% (Syring et al. 2020a), whereas a Holocene record from the Northeast Water Polynya (core PS93/025) reported TOC values of 0.5-0.9%(Syring et al. 2020b).However, a continuous dominance of inert organic carbon is seen throughout the record.Approximately 90% of the organic matter in sediment core 171G is found to be inert, and the sum of labile POC and all the labile SOC fractions indicates only a very small amount of fresh material (0.016-0.1 wt%; Fig. S3).The high amount of inert organic carbon can result from poor preservation of fresh organic carbon and/or input of allochthonous marine or terrestrial organic carbon.Moreover, the analyses revealed a large mineral carbon content (0.5 wt%).
A continuous high terrigenous signal is seen throughout the core.Out of a total of 1.08 wt% TOC, 0.36 wt% is attributed to highly degraded organic matter (Fig. 5), probably derived from land, although the total terrestrial input is probably more.The S2 vs. TOC plot further revealed a depositional environment heavily influenced Fig. 5. Organic matter origin of the marine sediments from core 171G, as inferred from S2 plotted against TOC in wt% from sediment core 171G retrieved on the NE Greenland shelf.A composition analysis with the best fit of linear regression is conducted for all data points (black line).A slope of 1.08 9 100 = 108 represents the overall hydrogen index (HI) of the system (S2/TOC 9 100).The HI of 108 mg HC/TOC is low and represents type IV, highly degrading/high level of reworked organic matter input.The regression line intersects the x-axis at 0.36 wt% TOC.This is regarded as the baseline inertinite (inert organic matter) content of the sediment supplied by detrital influx (Sanei et al. 2014).by a source-based inert organic carbon, and the clustering of samples leads to a classification of the organic matter as type III 'terrestrial organic matter' or type IV 'reworked organic matter' and is consistent with the reworked palynomorphs and reworked terrestrial particles observed in the palynological slides.The clustering of samples furthermore shows no significant changes in the overall composition of the organic matter downcore.This is also confirmed by the consistency in the reflectance measurements of organic maceral particles and fluorescent light observations, showing the continuous presence of highly degraded organic matter throughout the core.The relation between inert and labile organic matter, together with the plot of S2 vs. TOC and reflectance measurements, all point to an environment heavily influenced by terrigenous carbon.

Preservation
A consistent level of preservation of labile organic carbon throughout the record is indicated by the relatively stable ratio between S1 and S2 (Fig. S2E).However, the S1/(S1+S2) indexdoes not indicate whether the preservation is high or low.The large proportion of inert organic carbon (~90%) is consistent with either poor preservation of autochthonous marine carbon and/ or a high input of allochthonous organic carbon.From the composition analysis (Fig. 5), a large amount of allochthonous material is determined, suggesting that the high amount of inert carbon is land derived.

Minor marine contributions to the sedimentary organic matter pool
The pyrolysis data and reflectance measurements show organic matter that is heavily influenced by allochthonous contributions.Organic petrography analyses revealed large amounts of inert terrigenous particulate organic matter and enabled a qualitative assessment of the coal fragments found in the palynological slides.A few fragments of fresher particulate organic matter were found in fluorescent light but with no indications of marine origin.Marine macerals were not detected during the petrographic microscopy; however, the appearance was not assessed thoroughly enough to fully exclude the possible presence of small particles such as marine liptodetrinite or lamalginite (Boucsein & Stein 2000).If any marine material was present in the reflectance measurements, it would have had a low reflectance and thus would belong in the low reflectance group (Group 1).Therefore, the sum of Group 2 and Group 3 has been calculated to represent the minimum terrestrial input.The minimum sum of Group 2 and Group 3 is found at sample depth 240.5 cm and is 70%, and the highest sum of Group 2 and Group 3 is measured at sample depth 368.5 cm and is 84%.It is, therefore, concluded that the minimum terrestrial input varies between 70 and 84% (Fig. 7).Nevertheless, it is known that most of Group 1 is not marine, and the terrestrial input is, therefore, certainly higher but cannot be precisely quantified.
The only indication of marine inputs to the organic matter pool is from our palynological analyses, which reveal very low concentrations of recent marine dinoflagellates throughout the core.Previous studies from the Arctic region have shown similar patterns to what we report here for NE Greenland.ATOC content of around 1% was observed in most of the investigated Holocene cores from this region (Boucsein & Stein 2000;Stein & Fahl 2004), with HI values of <100 mg HC g À1 OC and falling into the Krevelen-type III field.The C:N corrected for inorganic nitrogen was about 11-15 and the d 13 C was between À25.5 and À26.5&, all pointing to the dominance of terrigenous organic matter (Boucsein & Stein 2000;Stein & Fahl 2004) and in line with our values for 171G: C:N, 9.5-16; d 13 C org , À25 and À27.7&; and HI, 27-91 mg HC g À1 TOC (Jackson et al. 2022).In fact, sediment studies, including organic geochemical bulk parameters, maceral composition and biomarker data, have shown that the terrigenous supply is the most important factor controlling the organic carbon deposition in the Laptev Sea continental margin throughout the Holocene (Stein & Fahl 2004).Quantitative estimates of terrigenous and marine macerals revealed that 95% of the organic matter in the inner Laptev Sea sediments was terrigenous material and that at the mouth of the Lena River, values were as high as 97-99%.
When considering temporal changes in the TOC data in 171G, both methods show an increasing trend from 12.2 to 10.6 ka BP (zones D1, D2 and the beginning of D3) (Fig. 4C).Values of S2 and HI are also increasing in D1 and D2 (Fig. S2).S2 represents the hydrogen-bearing, more labile fraction of the organic carbon, and HI is S2 normalized to TOC, thus both show a signal of more labile organic matter.There is some support for an increase in marine-derived organic matter at this point from d 13 C org and C:N (Jackson et al. 2022), despite a strong indication of overall terrigenous influence in the core (C:N ranging from $ 9.5 to 16 and d 13 C org values fluctuating between À25 and À27.7&).This interpretation is supported by the decreasing IRD content of the sediments over the same period, as would be expected when transitioning from an ice sheet-proximal setting in the late Younger Dryas/Early Holocene to a more marine environment.A possibly more terrigenous influenced TOC (CS) signal is seen from 10.7 to 8.2 ka BP (D3 and D4), consistent with the d 13 C org decrease (~27.7&).
While TOC concentrations in core 171G are dominated by terrigenous inputs, a temporal signal of autochthonous primary productivity might be deciphered when combined with information on the lability of hydrocarbons obtained from the pyrolysis method and preservation.An increase in TOC, S2 and HI is seen from the bottom of the core transitioning into zone D3 (12.2-10.7 ka BP) (Figs 4, S2).Here, both d 13 C org and C:N show a slightly less terrigenous signal, accompanied by higher fluxes of recent dinoflagellate cysts.When combining these lines of evidence, an increasing trend in marine productivity is interpreted, and zones D1 and D2 are seen as transitioning from an ice sheet-proximal setting to a more coastal marine environment (from ~10.7 ka BP).In the organic carbon composition, the labile-C SOC and labile POC (light and dark green) represent S2.Therefore, the changes seen in S2 represent changes in this minor fraction of the total organic carbon composition.It can, therefore, not be excluded that this minor change could instead represent an increased input of labile organic carbon from land.However, evidence from biogenic silica analysis, recent dinoflagellate cysts and foraminifera assemblages from the same core all indicate slightly higher production in this period.Furthermore, heterotrophic dinoflagellates argue for the presence of other living microalgae.While we conclude that the TOC in these shelf sediments is almost exclusively derived from land, when combining evidence from other proxies (micropaleontology, pyrolysis, C:N, d 13 C org ) we argue that slight temporal changes in marine export production might be detected.
North-East Greenland shelf sediments as a sink of recycled organic carbon from land The palynological slides revealed a low abundance of 'fresh' land-sourced material such as pollen, spores and Halodinium spp., but large amounts of coal fragments.This further confirms the high amount of terrigenous supply and gives good indications of low amounts of fresh terrestrial organic matter.The coal fragments were further investigated by organic petrography; however, the relative amounts of coal fragments and fresh organic material were not assessed during palynological analyses.Hilton et al. (2015) found that organic carbon-rich, high-latitude soils represent an important geological sink of CO 2 .That study differs in many ways from ours (e.g. the studied soils were organic carbon-rich and had high sedimentation rates).It still, however, illustrates how terrestrially derived carbon can represent a carbon sink.The significance of the recycled terrestrial organic matter in 171G is illustrated in the organic carbon composition, where it is clear that the inert carbon makes up the largest fraction.The labile carbon corresponds to a minor fraction, and the fresh and marine organic carbon is far from as significant as the reworked organic carbon.
Despite temporal fluctuations, all data clearly support the interpretation that marine autochthonous contributions to the total organic carbon pool in this sediment record are minor to negligible and that shelf sediments on this region of the NE Greenland shelf represent a sink of recycled carbon, mainly originating from nearby terrestrial deposits of Cretaceous age.

Implications for palaeoenvironmental studies: biomarkers and bulk sediment dating
Uncertainties in the interpretations of marine records of organic matter alone can, to some extent, result from unknown provenance and unknown input from multiple source areas.Sedimentary formations imprint the depositional environment and the thermal geological history.This study shows that pyrolysis analysis can be a helpful tool when interpreting organic matter pools preserved in marine sediment core records.Reworked microfossils are common in coastal sediments, and by identifying their taxonomic and stratigraphic affinity, alongside reflectance measurements, the provenance and source area can be elucidated and, thereby, it is possible to gain a better understanding of the palaeoenvironment and the processes at play.Our results clearly show that the TOC contents in marine sediments cannot simply be attributed to in situ marine organic matter production.On the contrary, we provide an example of a record containing predominantly terrigenous inert organic carbon.This can have severe consequences for the interpretation of biomarker proxy records and bulk sediment dating, as discussed below.
Biomarker records from the Fram Strait/NE Greenland region have been used to reconstruct productivity and sea-ice conditions during the Holocene (e.g.M€ uller et al. 2009;Syring et al. 2020a, b).Biomarker concentrations normalized against TOC have been proved to be more robust for regional comparisons (Kolling et al. 2020), and normalization is often done to account for changes in sediment composition downcore (e.g.Berben et al. 2017;Saini et al. 2020;Syring et al. 2020a).Belt & M€ uller (2013) address normalized concentrations relative to productivity indicators such as TOC.Using TOC in marine sediments to normalize biomarker concentrations can be highly problematic for shelf environments, as shown here, given the risk that downcore variations will be misinterpreted by terrestrial inputs of organic carbon.Thus, it is highly advisable to obtain knowledge on the provenance of organic matter in a given record, before biomarker datasets can be fully interpreted.
Radiocarbon dating of marine sediments in the Arctic is sometimes managed by bulk organic matter owing to the low availability of calcareous fossils.Strunk et al. (2020) used three lake sediment cores from East Greenland and Southeast Greenland and carried out paired bulk-and macrofossil dating throughout the cores to compare the age offset between the two methods.They were able to distinguish three main sources of organic carbon in dated bulk samples, thus obtaining better age constraints.Hansen et al. (2022) investigated the age offsets between radiocarbon dates from benthic foraminifera and bulk organic matter in a Holocene marine sediment core from Baffin Bay.Organic pyrolysis (TOC, HI, OI) was used to conclude that a reduced age offset was seen when higher percentages of autochthonous organic matter and lower percentages of older allochthonous organic matter were present.In this study, a large amount of allochthonous carbon would have offset the age.These examples further illustrate the usefulness of pyrolysis analyses for palaeoenvironmental studies.

Conclusions
We investigated the provenance, composition and preservation of organic matter in a well-dated marine sediment record from the NE Greenland shelf and found that the large majority (>70-84%) of the organic matter preserved in these sediments originates from land and about 90% of the organic carbon is inert.
Stratigraphic age determination of reworked palynomorphs (mainly dinoflagellate cysts), combined with pyrolysis organic geochemistry and reflectance-light organic petrographic microscopy, revealed that Cretaceous deposits on Clavering Island have been the main source of organic carbon to the shelf sediments throughout the late Younger Dryas and until present.These shelf sediments, located beneath the present-day Sirius Water Polynya, act as a sink of recycled carbon originating from the hinterland of the Young Sound region.Although a large amount of recycled organic carbon is buried, the core site is also a sink for more fresh organic carbon; however, the proportion of fresh terrigenous and marine organic carbon was not determined in this study.
Our findings have implications for palaeoenvironmental studies aimed at reconstructing climate and ocean conditions using organic proxy records (such as biomarkers), and efforts to date bulk sediments.The TOC content of coastal marine sediments influenced by terrestrial inputs cannot be interpreted as representing in situ marine productivity and thus, discerning the source of the organic carbon is necessary to avoid misinterpreting palaeoenvironmental records.
Fig. 1. A. Map of the study region with major oceanic currents: East Greenland Current (EGC; blue line), Atlantic Intermediate Water (AIW; orange dashed line), Return Atlantic Current (RAC; red line), North Atlantic Current (NAC), West Spitsbergen Current (WSC), the Irminger Current (IC) and the West Greenland Current (WGC).Figure modified from Jackson et al. (2022).B. Satellite image of the Sirius Water Polynya at the mouth of Young Sound (YS) (image dated 01.04.2021 and downloaded from NASAWorldview/MODIS. C. Geological map of the hinterland of Young Sound.Map adapted from Nøhr-Hansen et al. (2020).The outer line in the box in panel A refers to the location of the satellite image, and the inner line of the box refers to the geological map.The orange dot is the location of core DA17-NG-ST14-171G.

Table 1 .
Stratigraphic chart of reworked dinoflagellate cysts found in core 171G, showing the stratigraphic range of known occurrences (light yellow) and known occurrences from the East Greenland coast (orange)(Nøhr-Hansen 1993, 2012; Williams et al. 2017, and references herein;Nøhr-Hansen et al. 2020).

Fig. 3 .
Fig.3.Sedimentological and palynological properties of sediment core 171G.A. Grain size measurements from 53 samples at an 8 cm resolution are displayed as cumulative sand content; light brown is sand grains >63 lm, and dark brown is sand grains >150 lm.B. Sedimentation rate in centimetres per thousand years (cm ka À1 ).Median (black), the grey area represents 95% uncertainty.The lower 18 cm are not presented as no dates were obtained for this interval.The same accounts for the fluxes.C. Percentage of reworked (red) and recent palynomorphs (yellow).D. Flux of reworked palynomorphs.E. Flux of Quaternary dinoflagellate cysts.The five lowermost samples are marked with stars; here, fewer than five recent cysts were counted.Note the two y-axes: the first y-axis shows age in years BP, note the broken axis; the second y-axis shows the depth in the core (cm).Dinoflagellate cyst zones from cluster analysis byJackson et al. (2022) are represented as D1-D4.

A
Fig.4. A. Total organic carbon composition as an percentage of sediment core 171G.The organic carbon composition is composed of labile-A, -B and -C soluble organic carbon (SOC), labile particulate organic carbon (POC), carbon released during oxidation (O-C) and inert organic carbon (OC).B. Mineral carbon.C. Total organic carbon (TOC; wt%) of pyrolysis (blue) and CS (red) measurements and pyrolysis total carbon (wt%) (green).D. S2 measured in mg HC g À1 from pyrolysis.S2 is the sum of labile-C SOC and labile POC, which is illustrated in the legend of A. Note the two y-axes: the first y-axis shows scaled age in ka BP; the second y-axis shows the depth in the core (cm).The depths of reflectance measurements are marked with red dots at the depth axis.
Fig. 6. A. Mosaic of photomicrographs showing organic matter in fluorescent light from the sample at core depth 200.5 cm (9.6 ka BP).Small orange spots of fluorescing organic matter detritus can be observed, and almost no 'fresh' microalgae are present.B. Photomicrograph showing organic matter in fluorescent light showing variation in grain size from sample depth 400.5 cm (11.8 ka BP). C. Photomicrograph showing a fragment of vitrinite maceral in white light (left) and fluorescent light (right) from sample depth 200.5 cm (9.6 ka BP).White light, immersion oil, 109 objective and 509 magnification.A-C.Scale bar = 50 lm.