A multimillion-year-old record of Greenland vegetation and glacial history preserved in sediment beneath 1.4 km of ice at Camp Century
Edited by Mark Thiemens, University of California San Diego, La Jolla, CA, and approved January 27, 2021 (received for review October 23, 2020)
Significance
Understanding Greenland Ice Sheet history is critical for predicting its response to future climate warming and contribution to sea-level rise. We analyzed sediment at the bottom of the Camp Century ice core, collected 120 km from the coast in northwestern Greenland. The sediment, frozen under nearly 1.4 km of ice, contains well-preserved fossil plants and biomolecules sourced from at least two ice-free warm periods in the past few million years. Enriched stable isotopes in pore ice indicate precipitation at lower elevations than present, implying ice-sheet absence. The similarity of cosmogenic isotope ratios in the upper-most sediment to those measured in bedrock near the center of Greenland suggests that the ice sheet melted and re-formed at least once during the past million years.
Abstract
Understanding the history of the Greenland Ice Sheet (GrIS) is critical for determining its sensitivity to warming and contribution to sea level; however, that history is poorly known before the last interglacial. Most knowledge comes from interpretation of marine sediment, an indirect record of past ice-sheet extent and behavior. Subglacial sediment and rock, retrieved at the base of ice cores, provide terrestrial evidence for GrIS behavior during the Pleistocene. Here, we use multiple methods to determine GrIS history from subglacial sediment at the base of the Camp Century ice core collected in 1966. This material contains a stratigraphic record of glaciation and vegetation in northwestern Greenland spanning the Pleistocene. Enriched stable isotopes of pore-ice suggest precipitation at lower elevations implying ice-sheet absence. Plant macrofossils and biomarkers in the sediment indicate that paleo-ecosystems from previous interglacial periods are preserved beneath the GrIS. Cosmogenic 26Al/10Be and luminescence data bracket the burial of the lower-most sediment between <3.2 ± 0.4 Ma and >0.7 to 1.4 Ma. In the upper-most sediment, cosmogenic 26Al/10Be data require exposure within the last 1.0 ± 0.1 My. The unique subglacial sedimentary record from Camp Century documents at least two episodes of ice-free, vegetated conditions, each followed by glaciation. The lower sediment derives from an Early Pleistocene GrIS advance. 26Al/10Be ratios in the upper-most sediment match those in subglacial bedrock from central Greenland, suggesting similar ice-cover histories across the GrIS. We conclude that the GrIS persisted through much of the Pleistocene but melted and reformed at least once since 1.1 Ma.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
The history of the Greenland Ice Sheet (GrIS) and the ecosystems that occupied Greenland during ice-free intervals are poorly known before the Last Glacial Maximum and, particularly, prior to the last interglacial. Most evidence is drawn from marine sediments (1–5), geophysical surveys (6), a few terrestrial deposits (7, 8), and ice-core basal materials (9–12). Because well-dated terrestrial sediment archives are lacking, much of GrIS history is based on offshore records that provide only indirect constraints on ice-sheet extent and terrestrial ecosystems. This fragmentary knowledge of Greenland’s climate history limits our understanding of ice-sheet and ecosystem sensitivity to climate warming.
Knowledge of Greenland glacial history and ecosystems is limited for periods older than ∼1 Ma, and in most cases, only glaciation more extensive than present can be directly constrained. Ice-rafted debris (IRD) in marine sediment indicates that Greenlandic glaciers reached the ocean possibly as early as 45 Ma (13) but certainly by 7.5 Ma (1). Ice cover grew substantially at 3.3 Ma (14) and culminated in an expanded GrIS by 2.7 Ma (1). Cosmogenic isotopic analyses of IRD and other glacial-marine sediment suggest that glacial erosion began stripping the preglacial landscape by 7 Ma and had eroded most shallow regolith by 1.8 to 2.0 Ma (2, 15). Charcoal and organic debris in outwash near the Hiawatha Crater (16) and pollen in marine sediment in Baffin Bay and the Labrador Sea indicates that boreal forests in humid, cool-temperate to sub-Arctic climates during the Late Pliocene transitioned to tundra vegetation in a cold polar climate during the Early Pleistocene (17, 18). Seismic-reflection surveys record multiple GrIS advances to the edge of the continental shelf in Melville Bay after 2.7 Ma (6). Although poorly dated, rare fossil-rich shallow marine and coastal sediments overlying glacial deposits document warm, forested conditions that imply a smaller GrIS for short (<20 ky) periods between ∼2.4 and ∼1.8 Ma (7, 8).
Ice-sheet behavior is better known for the past million years because of analyses of ice-core basal materials sourced from landscapes covered by the present GrIS (Fig. 1). In central Greenland, silty ice at the base of the Greenland Ice Core Project (GRIP) ice core is as old as 950 or 970 ka (9, 19). In the nearby Greenland Ice Sheet Project 2 (GISP2) ice core, stratigraphic reconstructions and δ40Ar dating suggest that the deepest ice is older than 200 ka (20, 21). Silty ice in GISP2 preserves material from a preglacial landscape, which suggests persistent, nonerosive ice cover during much of the Pleistocene (11). Cosmogenic 26Al/10Be data from bedrock in western Greenland (22) and beneath the GISP2 ice core (10) require that ice disappeared from central Greenland at least once within the past 1.1 My. Marine sediment records suggest an ice-free and forested southern Greenland during Marine Isotope Stage (MIS) 11 (374 to 424 ka) (3, 4), while DNA studies of Dye-3 basal ice found flora and fauna from sometime between 450 and 800 ka (9). Preservation of ice from MIS 5e (119 to 127 ka) in all Greenland deep ice cores (9, 12, 19–21, 23, 24), minimal ice-surface lowering at the North Greenland Eemian Ice Drilling (NEEM) ice core site (12) at that time, sediment flux from southern Greenland (25), and coupled climate-ice sheet models (26) all indicate that the GrIS remained largely intact during the last interglacial. Except for Dye-3 and Camp Century, all ice cores have been collected from Greenland’s interior. Less is known about GrIS extent, behavior, and stability near its margins, the portions of the ice sheet that will melt first and contribute the most to sea-level rise under future warming scenarios (27).
Fig. 1.
The Camp Century ice core, collected from the ice-sheet periphery in northwestern Greenland, recovered 3.44 m of frozen sediment from beneath 1,368 m of glacial ice and 14 m of silty ice (28); this is the thickest subglacial sedimentary archive recovered from a Greenland ice core (Figs. 1B and 2A). Collected in 1966, the subglacial sediment was not studied beyond initial reports of sediment provenance (29) and microfossils, including abundant freshwater diatoms and chrysophyte cysts, pollen, and scarce marine fossils (likely windblown), such as diatoms, sponge spicules, and dinoflagellates (30). The basal material remained in frozen storage for decades until it was rediscovered in 2017.
Fig. 2.
Here, we decipher the glacial and ecological history of northwestern Greenland using analytical techniques that were unavailable when this unique sediment archive was retrieved over 50 years ago. We analyze two samples, respectively, from the upper- and lower-most sections of the Camp Century subglacial sediment. We determine the depositional history and characteristics of the sediment using in situ cosmogenic nuclides, infrared-stimulated luminescence (IRSL), geochemistry, optical microscopy, and scanning electron microscopy and energy dispersive X-ray spectroscopy (SEM-EDS) analyses. We infer weathering conditions, paleo-elevation, and paleoclimate from pore-ice composition, and we characterize the paleoecology from macrofossil vegetation and biomarkers. We find that the Camp Century subglacial sediment preserves a unique, multimillion-year-old record of glaciation and vegetation. These data are consistent with persistent ice cover interrupted by at least two periods of ice-sheet loss and regrowth, once in the Early Pleistocene and another in the past 1.1 My.
Results
The Camp Century subglacial sediment includes two distinct units of frozen diamicton separated by an intermediate layer of debris-rich ice (Fig. 2A and SI Appendix, Supplementary Information Text). Permafrost features such as subhorizontal ice lenses are present only in the lower and intermediate unit. The contact between the intermediate ice layer and the upper diamicton is nonconformable and angular. The two samples we analyzed are from the upper (1059-4) and lower (1063-7) diamicton units (Fig. 2 B and C). Both diamictons contain a variety of lithologies, but the mineralogy is primarily quartz with few feldspars and mafic minerals. The lower sample contains a greater percentage of fine sediment (<125 µm) (Fig. 2A and SI Appendix, Fig. S1; SI Appendix includes full description). Pore-ice major ion chemistry and SEM-EDS analysis of grain coatings indicate greater chemical weathering in the lower sample than the upper sample (Fig. 2 D and E and SI Appendix, Fig, S1 and Table S1).
Cosmogenic 26Al/10Be ratios and IRSL confirm that the upper and lower samples have different histories of exposure and burial (Fig. 2F and SI Appendix, Tables S2–S8). In situ 10Be and 26Al are produced in quartz during near-surface exposure to cosmic rays and, when buried, radioactively decay at different rates over time. Relatively low 10Be concentrations and 26Al/10Be ratios less than that expected from surface production (7.3) indicate long burial with limited exposure before or during the burial period (31) (see SI Appendix, Supplementary Information Text for complete details). The 26Al/10Be ratio in the upper sample (4.5 ± 0.3, n = 6, average ± SD, weighted by 26Al/10Be measurement uncertainty) requires exposure within the past 1.0 ± 0.1 My, while 26Al/10Be in the lower sample (1.8 ± 0.4, n = 3) indicates longer burial but no more than 3.2 ± 0.4 My (Fig. 2F and SI Appendix, Table S8). Multiple feldspar luminescence dating approaches indicate the lower sample was last exposed to sunlight before 0.7 to 1.4 Ma (SI Appendix, Fig. S3 and Tables S9 and S10). The upper sample did not produce meaningful IRSL results because of light exposure during storage and subsampling.
Sediment pore-ice stable isotopes are enriched in both samples, which require precipitation at elevations lower than the present ice-surface elevation at Camp Century (1,890 m above sea level) and imply ice-sheet absence (SI Appendix, Table S11). Pore-ice δ18O values are 5.9‰ to 7.5‰ greater than mean Late Holocene precipitation (–29‰) at Camp Century (23), which can be explained by both lower surface elevation and possibly warmer temperatures than present. For example, removal of the present ice sheet (thickness of 1,382 m) at Camp Century and isostatic adjustment would raise the current bedrock elevation by ∼440 m, implying a net surface elevation lowering of ∼950 m. Such an elevation change would account for 5.3 ± 1‰ of the enriched δ18O values in the pore ice, assuming a δ18O/altitude effect of 0.6‰/100 m (32). Temperatures ∼0.2 °C to ∼3 °C warmer than present and/or phase changes within the sediment account for the remaining δ18O enrichment, assuming a δ18O/temperature sensitivity of 0.7 ± 0.1‰/°C (23) (SI Appendix includes full details). The deuterium excess and 17O excess values are consistent with modern precipitation and do not indicate alteration (SI Appendix, Table S11).
Terrestrial plant macrofossils are abundant in both samples and preserve a record of the vegetation that occupied northwestern Greenland during ice-free intervals (Fig. 3 A–J and SI Appendix, Figs. S4–S7). The upper sample includes twigs (possibly including Empetrum), moss leaves and stems of Tomentypnum nitens and Polytrichum juniperinum, and sclerotia of the fungus Cenococcum geophilum (Fig. 3 A–H and SI Appendix, Figs. S4–S6), which are consistent with a tundra ecosystem but could coexist with boreal forest (9). In the lower sample, macrofossils such as bryophyte stems are present (Fig. 3 I and J and SI Appendix, Fig. S7) but not as well preserved and thus difficult to identify. Mixed woody tissue from the upper and lower samples yield δ13C ratios of −26.7 ± 0.1‰ and −29.6 ± 0.1‰, δ15N ratios of 2.4 ± 0.8‰ and −2.3 ± 0.8‰, and C/N ratios ranging from 15.0 to 23.0 and 47.4 to 53.5, respectively, which are consistent with tundra as well as boreal vegetation (SI Appendix, Table S12) (33). A twig from the upper sample yielded a radiocarbon age >50 14C ka (SI Appendix, Table S13).
Fig. 3.
Well-preserved leaf waxes in Camp Century basal sediment resemble those of modern Greenland tundra ecosystems. Concentrations are higher in the upper sample (Fig. 3K and SI Appendix, Tables S14 and S15). Both samples contain C20 to C32 n-alkanoic acids with minimal interference from unsaturated compounds and a high even-to-odd chain length index (>3.45), indicating minimal postdepositional alteration. The n-alkanoic acids are dominated by C24, similar to modern lake sediments in Greenland and soils in the boreal forest regions of Canada (34, 35). The midchain n-alkanes (C21 to C25) have a low odd-to-even chain length index (≤1.0) and coelute with an unidentified complex mixture, likely derived from hydrocarbon-based drilling fluid (SI Appendix, Fig. S8). Long-chain n-alkanes (C27 to C33) have a high odd-to-even chain length index (>4.5) and are dominated by C27 and C29, similar to modern shrubs in Greenland (36).
Discussion
The Camp Century subglacial sediment provides a unique stratigraphic record of ice-free ecosystems and ice cover in Greenland. Plant macrofossils and lipid biomarkers in both samples are direct evidence for the preservation of terrestrial paleo-ecosystems beneath the ice sheet. The greatly enriched δ18O of pore ice in both samples requires that precipitation fell at much lower elevations, demonstrating ice-sheet absence and possibly a warmer climate and/or water phase changes in the sediment. Based on differences in stratigraphy, weathering, 26Al/10Be ratios between the upper and lower sediment, and the lower diamicton IRSL minimum burial age (>0.7 to 1.4 Ma), this sedimentary sequence represents at least two episodes of subaerial exposure and subsequent burial. Abundant subhorizontal ice lenses are present only in the lower diamicton, which suggests an active permafrost environment before the emplacement of the upper diamicton unit. The nonconformable contact between the intermediate debris-rich ice layer and the overlying upper diamicton suggests a second glacial advance that incorporated and transported sediment and vegetation from a younger ice-free event.
The Camp Century lower diamicton, buried after 3.2 ± 0.4 Ma (26Al/10Be maximum age) but before 0.7 to 1.4 Ma (IRSL minimum age), may derive from the growth of the GrIS during the Early Pleistocene (Fig. 4). The cosmogenic 26Al/10Be limiting burial ages fit with the timing of GrIS expansion to the edge of the continental shelf at ∼2.7 Ma (Fig. 4E) (1, 6). The presence of extensive grain coatings and high solute concentrations in pore ice are remnants of a weathered surface landscape (2, 15) (Fig. 2). Given the burial age range, macrofossils and lipid biomarkers in the lower diamicton add to growing evidence that paleo-ecosystems from the Early Pleistocene, and perhaps earlier, are preserved both in present ice-free areas (7, 8) as well as below the GrIS (16). The lower diamicton represents the oldest terrestrial record for an ice-free ecosystem recovered beneath the GrIS.
Fig. 4.
Exposure of the Camp Century upper diamicton within the last 1.0 ± 0.1 My is consistent with paleoclimate records, including basal materials in other ice cores, which suggest a smaller or possibly absent GrIS within the last ∼1 My. Ice absence in the northwestern GrIS periphery at Camp Century since 1.0 ± 0.1 Ma agrees with the timing of bedrock exposure below the GrIS summit (<1.1 Ma) (10), which together mandate that much of Greenland was ice free within the last 1.1 My (Fig. 4G and SI Appendix, Table S16). The maximum burial duration for the GISP2 bedrock and the Camp Century upper diamicton are consistent with GRIP basal ice ages (950 to 970 ka) (9, 19). These age constraints require that the GrIS was smaller during the last 1.1 My, most likely during a Pleistocene superinterglacial; however, the precise timing and extent remains uncertain. During MIS 31 (1.06 to 1.09 Ma), global mean sea level reached its Pleistocene maximum (37), and paleotemperatures in northeastern Siberia were elevated (38), consistent with a greatly reduced GrIS (Fig. 4 A–C). A smaller-than-present GrIS during the long MIS 11 (374 to 424 ka) interglacial is also possible, as offshore records suggest a more ice-free and forested southern Greenland during this time (3, 4, 9) and sea-level records require significant reduction of the GrIS (39). In contrast, ice-core and geologic evidence (9, 12, 19–21, 23–25) along with modeling studies (26) suggest that the GrIS was mostly intact during MIS 5e (119 to 125 ka).
The similarity of 26Al/10Be burial histories between the Camp Century upper diamicton and GISP2 subglacial bedrock (10) suggests a similar history of ice cover and absence between the GrIS margin and interior (SI Appendix, Fig. S9 and Table S16). The slightly higher 26Al/10Be (4.5 ± 0.3) in the Camp Century upper diamicton than GISP2 bedrock (4.1 to 4.2), and thus slightly shorter burial duration, is consistent with the ice-marginal position of Camp Century. Nevertheless, the similarity not only indicates GrIS presence for much of the Pleistocene but also suggests at least one episode of greatly reduced ice-sheet extent, likely in response to prolonged interglacial warmth. For much of the Pleistocene, the GrIS persisted through interglacial periods while other northern hemisphere ice sheets vanished; however, our data show that the GrIS disappeared from Camp Century at some point in the last 1.1 My. The Pleistocene resilience of the GrIS and its ability to melt beyond some warming threshold is important information for assessing the risk of sea-level rise under future climatic conditions.
Analysis of the basal silty ice and the remaining archive of subglacial sediment from Camp Century will further resolve the cryostratigraphy, how sediments were emplaced, and the sensitivity of the northwestern GrIS margin to past warming and the types of ecosystems that develop under warmer, ice-free conditions in Greenland. Precise age control of past ice-free episodes recorded in ice-core basal materials, from both existing archives and future drilling campaigns, is key to identifying the climate thresholds that permitted GrIS collapse in the past and, thus, under future warming scenarios.
Materials and Methods
Materials.
The subglacial sediment from the Camp Century ice core was collected in 1966 (28). The sediment was stored frozen, initially at University at Buffalo from 1966, until it was transferred to Niels Bohr Institute (NBI) in 1994 and 1996. The samples are kept at −30 °C as ∼10 cm thick subsamples stored in individual glass jars. In the summer of 2019, two samples (1059-4 and 1063-7) were cut at NBI using a dedicated cold-room–hosted diamond-wire saw designed for cutting heterogeneous material (40). Loose material from 1059-4 was kept as a replicate sample (1059-4 Debris).
Sedimentological Description.
Basal sediment was inventoried, photographed, and described in October 2019 in the NBI freezer. The dimensions and masses of each subsample were measured to determine the bulk density (∼5% uncertainty). We refined the preliminary characterization of the subglacial sediment (29) by describing the granulometry, texture, clast shape, and presence of particular structures, such as ice lenses or deformation features. Our observations and the presence of mixed lithologies indicate that these Camp Century sediments are probably glacial diamictons consisting of glacial and/or periglacial sediments reworked by subsequent ice advances after initial deposition. A full sediment description is in SI Appendix.
Sample Processing.
Sample material was sent frozen to the University of Vermont (UVM). We subsampled frozen fragments of original samples for IRSL and SEM-EDS. Samples were thawed at 4 °C inside sample bags. Pore-ice meltwater was decanted from sample bags directly into 50 mL plastic vials. The remaining sediment was transferred into clean, unused plastic bottles and centrifuged, and then meltwater was pipetted into 50 mL plastic centrifuge tubes. Pore-ice melt samples were stored in 50 mL test tubes, capped, sealed, and refrigerated at 4 °C.
Sediment was dried at 65 °C, photographed, assigned a soil color, and measured for mass. Sediment was wet sieved with deionized water into <125, 125 to 250, 250 to 500, 500 to 850, 850 to 2,000, and >2,000 μm grain size fractions. During wet sieving, woody tissue and macrofossils were isolated using a dedicated disposable pipette, dried on a paper filter over a vacuum, and stored frozen. The 250 to 500 and 500 to 850 μm fractions were density separated using lithium polytungstate to isolate the felsic mineral fraction (<2.85 g/cm3).
Major Ion Chemistry.
Pore-ice meltwater was aliquoted by volume (1059-4: 4 mL; 1063-7: 0.5 mL) and massed. The samples were diluted to 10 mL (1059-4) and 8 mL (1063-7), remassed, transferred to 10 mL syringes, and filtered through 0.45 µm nylon filters back into the initial vials. Major cations and anions were measured using atomic absorption spectroscopy (PerkinElmer PinAAcle 900H) and ion chromatography (Metrohm 883/863 Ion Chromatograph) in the Environmental Analysis Laboratory at Williams College. Si was measured with a Technicon Autoanalyzer II utilizing the reduction of silicomolybdate to “molybdenum blue” with ascorbic acid. Laboratory pH and alkalinity were determined using a Fisher Scientific 320 pH meter, a Radiometer-analytical TIM840 auto-titrator, and double endpoint Gran titration.
SEM-EDS Analyses of Grain Coatings.
Thawed sample materials (∼100 mg) were sprinkled onto clean paper and embedded in epoxy (EPO-TEK 301). Epoxy plugs were polished using a decreasing grit size to 1 μm and carbon sputter-coated prior to analysis in backscattered electron mode using a TESCAN VEGA3 scanning electron microscope coupled with an Oxford Instruments AZtec Elemental Mapping EDS in the Geology Department at Middlebury College. EDS maps were acquired at 20 keV for a minimum of 3 min and tricolor plots generated using the Gatan digital micrograph 3.1 software.
Cosmogenic 10Be and 26Al.
Quartz purification and beryllium and aluminum extraction were performed independently at UVM (250 to 500 μm and 500 to 850 μm) and Lamont–Doherty Earth Observatory (LDEO), Columbia University (>2,000 μm). At UVM, quartz from the felsic separates of the 250 to 500 and 500 to 850 μm fractions was purified using acid etches (41), and then Be and Al were isolated and extracted using standard procedures (42) specialized for low-concentration samples (43). Two procedural blanks were prepared with the samples during extraction to estimate 10Be and 26Al backgrounds from laboratory sample processing and accelerator mass spectrometry analysis. At LDEO, the >2,000 μm samples were crushed, and quartz was isolated and mostly separated from feldspar by froth flotation. The quartz fraction was subsequently leached in 1% HF/3% HNO3 on a shaker table. Be and Al were isolated and extracted using LDEO procedures (10, 44) for low-level concentrations.
The 10Be/9Be ratios in all samples and blanks were measured at the Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory (LLNL), and normalized to primary standard 07KNSTD3110 (assumed 10Be/9Be ratio: 2,850 × 10−15) (SI Appendix, Table S2) (45). At UVM, four blanks (∼240 μg 9Be) are included in the in situ blank correction: two processed side by side with samples in this project and two from another batch of low-concentration samples from a separate project processed in the same hood and analyzed together at LLNL. For samples prepared at UVM, we used an average blank 10Be/9Be ratio of 7.4 ± 2.4 × 10−16, subtracted the blank ratio from the sample ratios, and propagated uncertainties in quadrature (SI Appendix, Table S3). At LDEO, four blanks (∼187 μg 9Be) are included, yielding an average 10Be/9Be ratio of 4.1 ± 1.9 × 10−16 (SI Appendix, Table S4). Blank subtraction and error propagation were calculated using the UVM procedure.
The 27Al/26Al ratios in all samples and blanks were measured at the Purdue Rare Isotope Measurement Laboratory and normalized against primary standard KNSTD (assumed ratio: 1.818 × 10−12) (SI Appendix, Table S5) (46). For samples prepared at UVM, we calculated the blank correction by averaging the two batch blanks. To constrain the uncertainty, we used the precision based on counting statistics because the two blanks agreed more closely than statistically allowable. We used a blank ratio of 1.8 ± 0.3 × 10−15, subtracted the blank ratio from the sample ratios, and propagated uncertainties in quadrature (SI Appendix, Table S6). We applied a consistent strategy for the LDEO samples with two blanks (average 9.8 ± 3.1 × 10−16) (SI Appendix, Table S7).
IRSL.
Luminescence dating estimates the time since sediment was last exposed to light, which resets the luminescence signal (47). Samples were sent frozen to the Utah State University Luminescence Laboratory. The outer 2 mm of each sample was shaved off to remove light-exposed sediments and processed to isolate the 150 to 350 μm of the quartz and potassium feldspar fractions. Sample 1059-4 (USU-3195) produced little natural luminescence signal, indicating bleaching by light exposure during storage and subsampling. Measurements of the quartz fraction of 1063-7 (USU-3196) indicated that the natural signal was saturated and beyond dating range. All remaining measurements were conducted on the feldspar fraction of 1063-7.
IRSL measurements followed the single-aliquot regenerative-dose method for feldspars (48) and methods to reduce the effects of anomalous fading (loss of signal with time). The first analyses followed the high temperature post-infrared IRSL (p-IRSL) method with infrared stimulation at 225 °C (p-IRSL225) (49, 50) and included fading correction (47, 51). The second measurements utilized a modified p-IRSL protocol with multiple elevated temperatures (MET) (52), in which equivalent dose (De) values were calculated at progressively higher temperatures (50 to 300 °C at 50 °C temperature steps). Dose-rate values for each sample were determined using inductively coupled plasma mass spectrometry and inductively coupled plasma atomic emission spectroscopy on subsamples from each core segment (SI Appendix, Table S9).
Results from both IRSL measurement techniques are presented in SI Appendix, Table S10, and De distributions are presented in SI Appendix, Fig. S3. Dose–response measurements indicate that the natural luminescence signals were near or beyond saturation and results should be considered minimum age estimates. The fading corrected p-IRSL225 results suggest the basal sample was last exposed to light at least 0.70 My, and the MET p-IRSL200, 250 results suggest the sample was deposited prior to 1.4 Ma. SI Appendix includes a detailed IRSL methods.
Pore-Ice Stable Isotopes.
We used established laser spectroscopy methods (53, 54) to analyze 1 mL aliquots of undiluted pore-ice meltwater for δ18O, δD, and δ17O at the University of Washington. Samples were calibrated against in-house reference waters previously calibrated against Vienna Standard Mean Ocean Water and Standard Light Antarctic Precipitation, using Greenland Ice Sheet Precipitation as a quality standard. The reference waters used were Seattle water and Vostok water, with WW (West Antarctic water) used as a quality check. WW has δD, δ17O, and δ18O, similar to that of modern Greenland summit snow. Values for all standards, and details of the measurement and calibration procedure, are given in refs. 53–55.
Macrofossil and Pollen Extraction and Examination.
Separate size fractions were rinsed with distilled water into a clean Petri dish and examined under a dissectoscope (10 to 60×) and then at 100 to 200× for moss leaf detail and photographs at LDEO. Macrofossils were better preserved and more easily identified in 1059-4 than 1063-7. Pollen extraction and examination used standard techniques using screens of 7 and 150 μm, acetolysis, and mounting in silicone oil (56). No pollen was found.
Organic Geochemistry.
Bulk woody tissue was analyzed for δ15N, δ13C, total organic nitrogen (TON), and total organic carbon (TOC) at the University of Washington using continuous-flow mass spectrometry. Samples were flash combusted at 1,000 °C with excess oxygen in a Costech ECS 4010 Elemental Analyzer following established methods (57, 58). TON and TOC are calibrated with a glutamic acid standard with known N and C concentrations. Internal laboratory reference materials (glutamic acid GA1 [δ13C = −28.3‰, δ15N = −4.6‰], GA2 [δ13C = −13.7‰, δ15N = −5.7‰], and Salmon [δ13C = −21.33‰, δ15N = +11.3‰]) were interspersed with samples for calibration. All data are on to the Air-N2 scale, for δ15N, and to the Vienna Pee Dee Belemnite scale, for δ13C. Precision and accuracy are determined for each run using one of the three references as an unknown.
Replicate TOC and TON analyses of bulk woody tissue were performed at the UVM Environmental Stable Isotope Facility by combusting samples in sealed tin capsules and analyzing the gas released in a CE Instruments NC 2500 elemental analyzer calibrated with Organic Analytical Standard B-2150 (6.72% C and 0.50% N) and National Institute of Standards and Technology Peach Leaves (46.34% C and 2.9% N). The precision of the analyzer is ∼1% of the quantity measured for TOC and ∼0.5% for TON.
Radiocarbon.
A twig and aliquot of woody tissue from 1059-4 were analyzed for 14Corganic at the Keck-Carbon Cycle Accelerator Mass Spectrometer, University of California Irvine. The samples (plus secondary standards and blanks of known-age and 14C-free wood) were sonicated in acetone, methanol, and Milli-Q water to remove possible contamination from drilling fluid and then treated with acid-base-acid (1 N HCl and 1 N NaOH, 75 °C) prior to combustion. 14C-dead blanks and standards covering a range of sizes were prepared and run with the two samples to estimate the masses of the modern and dead blanks. One-sigma uncertainty values for the resulting blank corrections are based on observed blank scatter within and between runs and are ±50% for samples <100 μg. The 1059-4 wood sample disintegrated during the solvent sonication, and only a small mass of sample (56 μg) was analyzed, which may have resulted in modern 14C contamination that yielded a finite age.
Lipid Biomarkers.
We extracted, purified, and analyzed the n-alkanoic acids and n-alkanes from the <125 µm fraction in the University at Buffalo Organic and Stable Isotope Biogeochemistry Laboratory following published procedures (59). We extracted free lipids from homogenized, dried samples on an Accelerated Solvent Extractor (Dionex ASE-200) using methylene chloride (DCM):methanol 9:1 (volume:volume), flushed three times at 10 min each. We added cis-eicosenoic acid and hexatriacontane internal standards to each sample and purified the samples using flash column chromatography. We methylated the acid fraction with acidified methanol at 60 °C for 8 h. We quantified the resulting fatty acid methyl ester (FAME) peak areas on a Thermo Trace 1310 Gas Chromatograph with flame ionization detector. We converted peak areas to compound mass by comparison with an external calibration curve established for a C28 FAME standard (for FAMEs) and a C27 alkane standard (for alkanes) and normalized compound mass by dry mass of the extracted sediment.
Data Availability
All study data are included in the article and/or SI Appendix.
Acknowledgments
We thank J. Racela, A. J. Schauer, O. Cowling, B. Buck, J. Goes, B. Linsley, M. Kirk, and L. Williamson for analytic assistance and the Geology Department at Middlebury College for SEM-EDS access. Ice-core samples have been curated by the University of Copenhagen since 1994. We thank the editor and two anonymous reviewers for constructive comments. Funding support is as follows: Gund Institute for Environment (A.J.C and P.R.B.), NSF-EAR 1735676 (P.R.B.), NSF-ARC-1023191 (P.R.B.), NASA Goddard Institute for Space Studies and LDEO (D.M.P.), Belgian Fonds National de la Recherche Scientifique—Fonds de la Recherche Fondamentale Collective grant no. 2.4601.12F (J.-L.T.), and NSF-EAR 1652274 (E.K.T.). This work was prepared by LLNL under contract DE‐AC52‐07NA27344 and LLNL‐JRNL‐811066 (A.J.H.).
Supporting Information
Appendix (PDF)
- Download
- 2.87 MB
References
1
H. C. Larsen et al.; ODP Leg 152 Scientific Party, Seven million years of glaciation in Greenland. Science 264, 952–955 (1994).
2
P. R. Bierman, J. D. Shakun, L. B. Corbett, S. R. Zimmerman, D. H. Rood, A persistent and dynamic East Greenland Ice Sheet over the past 7.5 million years. Nature 540, 256–260 (2016).
3
A. de Vernal, C. Hillaire-Marcel, Natural variability of Greenland climate, vegetation, and ice volume during the past million years. Science 320, 1622–1625 (2008).
4
A. V. Reyes et al., South Greenland ice-sheet collapse during marine isotope Stage 11. Nature 510, 525–528 (2014).
5
R. G. Hatfield et al., Interglacial responses of the southern Greenland ice sheet over the last 430,000 years determined using particle-size specific magnetic and isotopic tracers. Earth Planet. Sci. Lett. 454, 225–236 (2016).
6
P. C. Knutz et al., Eleven phases of Greenland Ice Sheet shelf-edge advance over the past 2.7 million years. Nat. Geosci. 12, 361–368 (2019).
7
S. Funder et al., Late Pliocene Greenland - The Kap København formation in North Greenland. Bull. Geol. Soc. Den. 48, 117–134 (2001).
8
O. Bennike et al., Early Pleistocene sediments on Store Koldewey, northeast Greenland. Boreas 39, 603–619 (2010).
9
E. Willerslev et al., Ancient biomolecules from deep ice cores reveal a forested southern Greenland. Science 317, 111–114 (2007).
10
J. M. Schaefer et al., Greenland was nearly ice-free for extended periods during the Pleistocene. Nature 540, 252–255 (2016).
11
P. R. Bierman et al., Preservation of a preglacial landscape under the center of the Greenland Ice Sheet. Science 344, 402–405 (2014).
12
NEEM community members, Eemian interglacial reconstructed from a Greenland folded ice core. Nature 493, 489–494 (2013).
13
A. Tripati, D. Darby, Evidence for ephemeral middle Eocene to early Oligocene Greenland glacial ice and pan-Arctic sea ice. Nat. Commun. 9, 1038 (2018).
14
K. Blake-Mizen et al., Southern Greenland glaciation and western boundary undercurrent evolution recorded on Eirik Drift during the late Pliocene intensification of Northern Hemisphere glaciation. Quat. Sci. Rev. 209, 40–51 (2019).
15
A. J. Christ et al., The northwestern Greenland Ice Sheet during the Early Pleistocene was similar to today. Geophys. Res. Lett. 47, e2019GL085176 (2020).
16
A. A. Garde et al., Pleistocene organic matter modified by the Hiawatha impact, northwest Greenland. Geology 48, 867–871 (2020).
17
A. de Vernal, P. J. Mudie, Late Pliocene to Holocene palynostratigraphy at ODP site 645, Baffin Bay. Baffin Bay Labrador Sea 105, 387–399 (1989).
18
A. de Vernal, P. J. Mudie, Pliocene and Pleistocene palynostratigraphy at ODP Sites 646 and 647, eastern and southern Labrador Sea. Baffin Bay Labrador Sea 105, 401–422 (1989).
19
A. M. Yau, M. L. Bender, T. Blunier, J. Jouzel, Setting a chronology for the basal ice at Dye-3 and GRIP: Implications for the long-term stability of the Greenland Ice Sheet. Earth Planet. Sci. Lett. 451, 1–9 (2016).
20
M. Suwa, J. C. von Fischer, M. L. Bender, A. Landais, E. J. Brook, Chronology reconstruction for the disturbed bottom section of the GISP2 and the GRIP ice cores: Implications for Termination II in Greenland. J. Geophys. Res. Atmos. 111, 1–12 (2006).
21
M. L. Bender, E. Burgess, R. B. Alley, B. Barnett, G. D. Clow, On the nature of the dirty ice at the bottom of the GISP2 ice core. Earth Planet. Sci. Lett. 299, 466–473 (2010).
22
A. Strunk et al., One million years of glaciation and denudation history in west Greenland. Nat. Commun. 8, 14199 (2017).
23
S. J. Johnsen et al., Oxygen isotope and palaeotemperature records from six Greenland ice-core stations: Camp Century, Dye-3, GRIP, GISP2, Renland and NorthGRIP. J. Quat. Sci. 16, 299–307 (2001).
24
A. Svensson et al., Annual layering in the NGRIP ice core during the Eemian. Clim. Past 7, 1427–1437 (2011).
25
E. J. Colville et al., Sr-Nd-Pb isotope evidence for ice-sheet presence on southern Greenland during the Last Interglacial. Science 333, 620–623 (2011).
26
M. M. Helsen, W. J. van de Berg, R. S. W. van de Wal, M. R. van den Broeke, J. Oerlemans, Coupled regional climate-ice-sheet simulation shows limited Greenland ice loss during the Eemian. Clim. Past 9, 1773–1788 (2013).
27
A. Aschwanden et al., Contribution of the Greenland Ice Sheet to sea level over the next millennium. Sci. Adv. 5, eaav9396 (2019).
28
B. L. Hansen, C. C. Langway Jr., Deep core drilling in ice and core analysis at Camp Century, Greenland, 1961-66. Antarct. J. US. 1, 207–208 (1966).
29
J. Fountain, T. M. Usselman, J. Wooden, C. C. Langway Jr., Evidence of the bedrock beneath the Greenland Ice Sheet near Camp Century, Greenland. J. Glaciol. 27, 193–197 (1981).
30
D. M. Harwood, Do diatoms beneath the Greenland Ice Sheet indicate interglacials warmer than present? Arctic 39, 304–308 (1986).
31
L. B. Corbett et al., Cosmogenic 26Al/10Be surface production ratio in Greenland. Geophys. Res. Lett. 44, 1350–1359 (2017).
32
B. M. Vinther et al., Holocene thinning of the Greenland ice sheet. Nature 461, 385–388 (2009).
33
S. M. Schaeffer, E. Sharp, J. P. Schimel, J. M. Welker, Soil-plant N processes in a High Arctic ecosystem, NW Greenland are altered by long-term experimental warming and higher rainfall. Glob. Change Biol. 19, 3529–3539 (2013).
34
E. K. Thomas, J. P. Briner, J. J. Ryan-Henry, Y. Huang, A major increase in winter snowfall during the middle Holocene on western Greenland caused by reduced sea ice in Baffin Bay and the Labrador Sea. Geophys. Res. Lett. 43, 5302–5308 (2016).
35
A. Bakkelund, T. J. Porter, D. G. Froese, S. J. Feakins, Net fractionation of hydrogen isotopes in n-alkanoic acids from soils in the northern boreal forest. Org. Geochem. 125, 1–13 (2018).
36
M. A. Berke, A. Cartagena Sierra, R. T. Bush, D. Cheah, K. O’Connor, Controls on leaf wax fractionation and δ2H values in tundra vascular plants from western Greenland. Geochim. Cosmochim. Acta 244, 565–583 (2019).
37
K. G. Miller et al., Cenozoic sea-level and cryospheric evolution from deep-sea geochemical and continental margin records. Sci. Adv. 6, eaaz1346 (2020).
38
M. Melles et al., 2.8 million years of Arctic climate change from Lake El’gygytgyn, NE Russia. Science 337, 315–320 (2012).
39
A. Dutton et al., Sea-level rise due to polar ice-sheet mass loss during past warm periods. Science 349, aaa4019 (2015).
40
J.-L. Tison, Instruments and methods: Diamond wire-saw cutting technique for investigating textures and fabrics of debris-laden ice and brittle ice. J. Glaciol. 40, 410–414 (1994).
41
C. P. Kohl, K. Nishiizumi, Chemical isolation of quartz for measurement of in-situ -produced cosmogenic nuclides. Geochim. Cosmochim. Acta 56, 3583–3587 (1992).
42
L. B. Corbett, P. R. Bierman, D. H. Rood, An approach for optimizing in situ cosmogenic 10Be sample preparation. Quat. Geochronol. 33, 24–34 (2016).
43
J. D. Shakun et al., Minimal East Antarctic ice sheet retreat onto land during the past eight million years. Nature 558, 284–287 (2018).
44
J. M. Schaefer et al., High-frequency Holocene glacier fluctuations in New Zealand differ from the northern signature. Science 324, 622–625 (2009).
45
K. Nishiizumi et al., Absolute calibration of 10Be AMS standards. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 258, 403–413 (2007).
46
K. Nishiizumi, Preparation of 26Al AMS standards. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 223–224, 388–392 (2004).
47
D. J. Huntley, M. Lamothe, Ubiquity of anomalous fading in K-feldspars and the measurement and correction for it in optical dating. Can. J. Earth Sci. 38, 1093–1106 (2001).
48
J. Wallinga, A. Murray, A. Wintle, The single-aliquot regenerative-dose (SAR) protocol applied to coarse-grain feldspar. Radiat. Meas. 32, 529–533 (2000).
49
K. J. Thomsen, A. S. Murray, M. Jain, L. Bøtter-Jensen, Laboratory fading rates of various luminescence signals from feldspar-rich sediment extracts. Radiat. Meas. 43, 1474–1486 (2008).
50
J. P. Buylaert, A. S. Murray, K. J. Thomsen, M. Jain, Testing the potential of an elevated temperature IRSL signal from K-feldspar. Radiat. Meas. 44, 560–565 (2009).
51
M. Auclair, M. Lamothe, S. Huot, Measurement of anomalous fading for feldspar IRSL using SAR. Radiat. Meas. 37, 487–492 (2003).
52
B. Li, Z. Jacobs, R. G. Roberts, S.-H. Li, Review and assessment of the potential of post-IR IRSL dating methods to circumvent the problem of anomalous fading in feldspar luminescence. Geochronometria 41, 178–201 (2014).
53
E. J. Steig et al., Calibrated high-precision 17O-excess measurements using cavity ring-down spectroscopy with laser-current-tuned cavity resonance. Atmos. Meas. Tech. 7, 2421–2435 (2014).
54
A. J. Schauer, S. W. Schoenemann, E. J. Steig, Routine high-precision analysis of triple water-isotope ratios using cavity ring-down spectroscopy. Rapid Commun. Mass Spectrom. 30, 2059–2069 (2016).
55
S. W. Schoenemann, A. J. Schauer, E. J. Steig, Measurement of SLAP2 and GISP δ17O and proposed VSMOW-SLAP normalization for δ17O and 17O(excess). Rapid Commun. Mass Spectrom. 27, 582–590 (2013).
56
K. Faegri, J. Iversen, P. E. Kaland, K. Krzywinski Textbook of Pollen Analysis (John Wiley and Sons, ed. 4, 1989).
57
A. Barrie, J. E. Davies, A. J. Park, C. T. Workman, Continuous-flow stable isotope analysis for biologists. Spectroscopy (Springf.) 4, 42–52 (1989).
58
D. J. Verardo, P. N. Froelich, A. McIntyre, Determination of organic carbon and nitrogen in marine sediments using the Carlo Erba NA-1500 Analyzer. Deep. Res. 37, 157–165 (1990).
59
E. K. Thomas, K. V. Hollister, A. A. Cluett, M. C. Corcoran, Reconstructing Arctic precipitation seasonality using aquatic leaf wax δ2H in lakes with contrasting residence times. Paleoceanogr. Paleoclimatology 35, e2020PA003886 (2020).
60
P.-H. Blard, M. Lupker, M. Rousseau, J. Tesson, Two MATLAB programs for computing paleo-elevations and burial ages from paired-cosmogenic nuclides. MethodsX 6, 1547–1556 (2019).
Information & Authors
Information
Published in
Classifications
Copyright
© 2021. Published under the PNAS license.
Data Availability
All study data are included in the article and/or SI Appendix.
Submission history
Published online: March 15, 2021
Published in issue: March 30, 2021
Keywords
Acknowledgments
We thank J. Racela, A. J. Schauer, O. Cowling, B. Buck, J. Goes, B. Linsley, M. Kirk, and L. Williamson for analytic assistance and the Geology Department at Middlebury College for SEM-EDS access. Ice-core samples have been curated by the University of Copenhagen since 1994. We thank the editor and two anonymous reviewers for constructive comments. Funding support is as follows: Gund Institute for Environment (A.J.C and P.R.B.), NSF-EAR 1735676 (P.R.B.), NSF-ARC-1023191 (P.R.B.), NASA Goddard Institute for Space Studies and LDEO (D.M.P.), Belgian Fonds National de la Recherche Scientifique—Fonds de la Recherche Fondamentale Collective grant no. 2.4601.12F (J.-L.T.), and NSF-EAR 1652274 (E.K.T.). This work was prepared by LLNL under contract DE‐AC52‐07NA27344 and LLNL‐JRNL‐811066 (A.J.H.).
Notes
This article is a PNAS Direct Submission.
Authors
Competing Interests
The authors declare no competing interest.
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFGet Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to get full access to it.