Late Glacial to Holocene radiocarbon constraints on North Pacific Intermediate Water ventilation and deglacial atmospheric CO2 sources

Earth and Planetary Science Letters 397 (2014) 57–66

Contents lists available at ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Late Glacial to Holocene radiocarbon constraints on North Pacific Intermediate Water ventilation and deglacial atmospheric CO2 sources Maureen Davies-Walczak a,b,∗ , A.C. Mix a , J.S. Stoner a , J.R. Southon c , M. Cheseby a , C. Xuan a,d a

College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, United States Research School of Earth Sciences, The Australian National University, Australia c Department of Earth System Science, University of California, Irvine, United States d Ocean and Earth Science, National Oceanography Centre Southampton, United Kingdom b

a r t i c l e

i n f o

Article history: Received 12 October 2013 Received in revised form 1 April 2014 Accepted 2 April 2014 Available online 5 May 2014 Editor: G.M. Henderson Keywords: North Pacific ventilation radiocarbon deglaciation

a b s t r a c t Radiocarbon reconstructions of past ocean ventilation rates constrain oceanic sources and sinks of CO2 and mechanisms of subsurface hypoxia. Here, 14 C in coexisting benthic and planktonic foraminifera from a sediment core 682 m deep off Southeast Alaska documents paleoventilation over the past ∼17,000 years. A chronology based on calibrated planktonic foraminiferal dates, consistent with independent terrestrial dates for regional glacial retreat, yields deglacial projection ages moderately greater than those of the Holocene, suggesting comparatively limited ventilation. The observed Holocene increase of apparent ventilation at intermediate depths tracks inundation of the Bering Strait between ∼11,800 and 13,200 years ago, suggesting that flooding of continental shelves and export of lowsalinity surface waters to the Arctic enhanced intermediate water formation in the North Pacific. An abrupt increase in the benthic–planktonic radiocarbon age gradient, implying homogenization of abyssal radiocarbon in deep and intermediate waters, aligns with the younger of two episodes of rapid rise of atmospheric CO2 and depletion of atmospheric 14 C during deglaciation (∼11,500–13,000 years ago), suggesting the North Pacific as a possible pathway for venting of oceanic CO2 to the atmosphere during the second half of the deglacial transition. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Increases in atmospheric CO2 content during the last deglaciation (Termination 1, ∼19,000–11,000 years BP) were accompanied by a decrease in atmospheric 14 C, apparently unsupported by abrupt changes in 14 C production rate (Monnin et al., 2001; Hughen et al., 2006; Broecker and Barker, 2007). Release of CO2 from abyssal waters is hypothesized to explain the apparent introduction of aged carbon to the atmosphere (Toggweiler, 1999; Sigman and Boyle, 2000; Stephens and Keeling, 2000; Monnin et al., 2001; Skinner and Shackleton, 2004; de Boer et al., 2007; Broecker and Barker, 2007; Galbraith et al., 2007; Marchitto et al., 2007; Keeling, 2007; Pena et al., 2008; Stott et al., 2009; Bryan et al., 2010; Basak et al., 2010; Skinner et al., 2010; Burke and Robinson, 2012; Sarnthein et al., 2013). Evidence for two discrete events of 14 C-depletion in some parts of the intermediate depth Indo-Pacific, apparently concomitant with the abrupt rises

*

Corresponding author at: Research School of Earth Sciences, The Australian National University, Australia. E-mail address: [email protected] (M. Davies-Walczak). http://dx.doi.org/10.1016/j.epsl.2014.04.004 0012-821X/© 2014 Elsevier B.V. All rights reserved.

in atmospheric CO2 (Marchitto et al., 2007; Stott et al., 2009; Bryan et al., 2010; Basak et al., 2010) supports this view. The site of release of this aged carbon is not well constrained, but is often presumed to originate around Antarctica; the effects may have been redistributed in the upper ocean via Antarctic Intermediate Water (AAIW). However, the evidence for such transport via intermediate waters is unclear (De Pol-Holz et al., 2010), and a volumetrically sufficient 14 C-depleted glacial watermass has not been found in the deep ocean (Broecker and Clark, 2010; Lund et al., 2011). Simple models that predict the consequences of deep carbon storage are not consistent with evidence from the deepest Pacific (Hain et al., 2011). The two most common approaches to reconstructing paleoventilation are: (1) 14 C age differences between benthic and planktonic foraminifera, placed on a planktonic chronology assuming constant surface reservoir age (e.g. Broecker et al., 1990; Kennett and Ingram, 1995; van Geen et al., 1996; Mix et al., 1999; Keigwin, 2002; McKay et al., 2005; Broecker et al., 2008), and (2) benthic 14 C anomalies relative to known ages and watermass trajectories, placed on a chronology derived by correlation of some measure of local climate change to the layer-counted

58

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Fig. 1. Location of jumbo piston core EW0408-85JC on the Southeastern Alaska margin. Bathymetry contours indicate −200 m (gray), −100 m (gray), and −40 m (black). The inset section shows estimated modern pre-bomb 14 C values (in per mille) from GLODAP gridded data for a North Pacific transect along longitude 145◦ W, delineated on the map by the dashed black line (Key et al., 2004). The red star denotes the latitude and depth of core EW0408-85JC. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

δ 18 O record of the Greenland ice cores (e.g. Marchitto et al., 2007; Bryan et al., 2010), or by independent dating, for example based on uranium decay series in favorable samples (e.g., Burke and Robinson, 2012) projected to the atmospheric 14 C history. Here we employ both chronological strategies to evaluate radiocarbon data from jumbo piston core EW0408-85JC, along with its trigger core EW0408-85TC (59◦ 33.32 N, 144◦ 09.21 W, 682 m water depth) and an adjacent multicore EW0408-84MC2 (59◦ 33.30 N, 144◦ 09.16 W, 682 m water depth), recovered from the continental slope of SE Alaska, south of Kayak Island (Fig. 1). Modern sedimentation at this site is hemipelagic, but during the Last Glacial Maximum, the northwestern portion of the Cordilleran Ice Sheet or its outlet glaciers terminated in the ocean (Molnia, 1986); glacially-derived sediments dominate accumulation prior to ∼14,700 cal years BP (Davies et al., 2011), supporting high sedimentation rates that make this site well-suited for reconstruction of radiocarbon while minimizing the smoothing effects of bioturbation. 2. Methods 2.1. Physical properties Physical properties data for the multicore, trigger core, and jumbo piston core from the site were analyzed shipboard on whole-core sections using the Oregon State University GEOTEK multi-sensor core logger (MSCL), and include gamma density, P-wave velocity, and magnetic susceptibility. These measurements are effectively smoothed by the detectors to ∼4 cm for the P-wave velocity data, and ∼6 cm for the magnetic susceptibility data. P-wave velocity values with amplitudes of zero were considered to be spurious due to poor transducer contact, and have been cleaned from the data set. All data are presented on a composite depth scale generated by alignment of the multicore, trigger core, and jumbo piston core physical properties data, described in Davies et al. (2011). A computerized tomographic (CT) scan of EW0408-85JC was logged at the Oregon State University College of Veterinary Medicine using a Toshiba Aquilion 64 Slice. Scans were collected at 120 kVp and 200 mAs. The resulting images were processed with a “sharp” algorithm to generate sagittal and coronal images every 4 mm across the core. Down-core and across-core pixel resolution within each slice is 500 μm. The cores were scanned in ∼60 cm

Fig. 2. Raw planktonic (solid green diamonds) and benthic (open gray squares) radiocarbon dates versus depth in core, plotted with 1-σ measurement uncertainties. The depth range of individual samples is  2 cm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

segments and then joined into a composite image using Adobe Photoshop software (Davies et al., 2011). 2.2. Radiocarbon Radiocarbon analyses were performed at the UC Irvine Keck AMS facility. A total of 78 measurements were performed for 39 samples: 3 benthic/planktonic pairs from EW0408-85TC, and 36 benthic–planktonic pairs from EW0408-85JC (Fig. 2). Benthic and planktonic foraminifera were picked from the >150 μm sediment fraction of jumbo piston core EW0408-85JC. The two predominant planktonic species at this site, N. pachyderma (sinistral) and G. bulloides, were analyzed separately at 555 cmbsf. In this sample, the 14 C age of N. pachyderma was 65 ± 50 years greater than that of G. bulloides (Davies et al., 2011). As this difference approximates the measurement uncertainty for the individual samples, these dates were averaged at 555 cmbsf. For all other depths the species were combined to increase sample size and thus reduce analytical

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

59

uncertainty. Benthic foraminiferal 14 C analyses were run as mixed species, although care was taken to avoid agglutinated (none observed) and deep infaunal (Globobulimina spp and Chilostomella oolina; Zellers et al., 2009, 2011) taxa. 2.3. Stable isotopes Stable isotopic measurements on planktonic foraminifera (Neogloboquadrina pachyderma, sinistral) were carried out at the Oregon State University College of Earth Ocean and Atmospheric Sciences (OSU/CEOAS) Stable Isotope Mass Spectrometer Facility. We performed 177 analyses of δ 18 O and δ 13 C across the last deglacial interval at an average sample interval of 5 cm in the jumbo piston core; this data set as well as further details on our analytical methods are published in Davies et al. (2011). 3. Results and discussion 3.1. Age models To generate Age Model 1, raw radiocarbon dates from planktonic foraminifera were converted to calendar ages using the Marine13 database (Reimer et al., 2013) in the Bayesian radiocarbon chronology program BChron (Haslett and Parnell, 2008; Parnell et al., 2008) (Table A.1). A constant 1-σ  R of 470 ± 80 years reflects modern (pre-bomb) regional surface–ocean reservoir ages of 880 ± 80 years (McNeely et al., 2006). At depth, modern preanthropogenic reservoir ages near the sample site at an equivalent density horizon to 682 m water depth are 1390 ± 100 years (Sabine et al., 2005). The uppermost planktonic sample from the trigger core (22 cmbsf) yielded a raw radiocarbon age of 660 ± 250 (1-σ ), while the coexisting benthic sample yielded a raw radiocarbon age of 1530 ± 45 years; the planktonic date implies negative calendar-corrected surface ocean ages. Excess 210 Pb activity in adjacent multicore EW0408-84MC indicates that the dated level at 22 cmbsf was deposited within the past hundred years (Walinsky et al., 2009). To explain the anomalously young planktonic date relative to known reservoir ages, we infer minor contamination of the planktonic foraminifera with bomb-derived (i.e., post 1962 AD) radiocarbon, and assume a core-top age equivalent to the year of coring (2004 AD, or −54 BP). Age Model 1 is derived from a Bayesian probabilistic model (BChron; Haslett and Parnell, 2008; Parnell et al., 2008) of the remaining calibrated planktonic foraminiferal dates, assuming constant reservoir ages with respect to a changing atmosphere (Fig. 3). To generate tuned Age Model 2, 0 cmbsf was assigned an age reflecting the year of core collection, while the core bottom was assigned an age based on the oldest calibrated planktonic radiocarbon date, after applying the nominal modern reservoir age 880 ± 80 years of (Table A.1). Six correlation tie points of δ 18 O in planktonic foraminifera (Neogloboquadrina pachyderma, sinistral) to δ 18 O in the Greenland NGRIP ice core on the GICC05 chronology (Rasmussen et al., 2006) further constrained Age Model 2. We purposely selected tie points corresponding as closely as possible to those used previously to correlate Baja California core MV99-GC31/PC08 to the Greenland GISP2 chronology (Ortiz et al., 2004; Marchitto et al., 2007) to facilitate comparison of these two intermediate-depth sites (Table A.2; Fig. 4). No significant difference in the EW0408-85JC age–depth relationships resulted from correlating to GISP2 relative to the NGRIP Greenland ice core isotope data on their synchronized GICC05 chronologies (Rasmussen et al., 2006) (Table A.2; Fig. 3). Chronological uncertainties for the tie points in this age model reflect the resolution of the EW0408-85JC δ 18 O record, as well as the uncertainties on the layer-counted GICC05 chronology. A Monte Carlo approach (1000 simulations of the age model, calculated from starting depths and

Fig. 3. Age–depth relationships based on Marine13-calibrated planktonic Age Model 1 (solid green line; Reimer et al., 2013), NGRIP-tuned Age Model 2 (solid blue line; Rasmussen et al., 2006), and the GISPII-tuned version of Age Model 2 (dashed blue line; Rasmussen et al., 2006). Tie points used in generating Age Model 2 (solid blue circles, mostly obscured by nearly identical open blue circles) and the GISPIItuned (open blue circles) age model, are plotted with estimated 1-σ uncertainties. Calibrated planktonic (solid green diamonds) dates are also shown with 1-σ uncertainties for Age Model 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

calendar ages selected randomly from within the 1-σ distribution of the sampling and chronological uncertainty) was then applied to generate a 1-σ uncertainty envelope for the tuned age–depth model (Table A.3). We propose Age Models 1 and 2 as reasonable options, acknowledging that the assumption of constant sediment accumulation rates between tie points in Age Model 2 is probably not realistic. More detailed correlation to Greenland would require a δ 18 O reconstruction of higher resolution than recovered from this site. We evaluated an age model (not shown) based on interpolation of surface–ocean reservoir ages between points of correlation to Greenland (following a strategy similar to that of Skinner et al., 2010) but this approach yielded a result intermediate to those of Age Models 1 and 2, and would not change our general conclusions. We also considered tuning the age models based on radiocarbon “plateaus” with respect to depth (Lund and Mix, 1998; Sarnthein et al., 2007). Sedimentation rates are sufficiently variable on the continental margin that radiocarbon plateau tuning is not warranted here. 3.2. Projection ages Projection age calculation requires an independent chronology; here we evaluate the implications of Age Models 1 and 2. The projection method accounts for changes in atmospheric radiocarbon but not for the reservoir ages of the watermass source region(s) or changes in the mixing ratio of source regions with differing initial reservoir ages (Adkins and Boyle, 1997). Nonetheless, this method remains instructive where subsurface watermasses have experienced prolonged isolation from the atmosphere, as long as the potential mechanisms contributing to the projection age are considered in the interpretation. All projection ages were calculated here with respect to the IntCal13 atmospheric reconstruction (Reimer et al., 2013). Benthic foraminiferal radiocarbon ages were converted into subsurface ocean 14 C values (14 C deep ) at site EW0408-85JC for Age Models 1 and 2 following standard methods (Stuiver and Polach, 1977) (Table A.3; Fig. 5). A Monte Carlo approach (1000

60

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Fig. 4. EW0408-85JC N. pachyderma left-coiling δ 18 O on the two potential age models: (a) Age Model 1 (solid green diamonds) assumes a constant surface ocean reservoir age for radiocarbon dates while (b) Age Model 2 (solid blue diamonds) is based on correlation (following Marchitto et al., 2007) to (c) Greenland ice-core NGRIP δ 18 O (thin light blue line, 20-year running average overlain in bold) on the GICC05 timescale (Rasmussen et al., 2006). Dashed lines show temporal relationship between isotopic features on the different age models. Shaded bars demarcate the Bølling, Allerød, and Younger–Dryas climate events as defined in NGRIP. Age control points with 1-σ uncertainties are shown for each age model as open diamonds; dates from the trigger core, for which there is no isotopic data, are delineated by a central dot. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

simulations for each dated sample including both chronological and radiocarbon analytical uncertainty) was applied to calculate a 1-σ uncertainty on the projection age intersection with the atmospheric calibration curve (Fig. 5). For Age Model 2, we also calculated apparent surface–ocean projection ages from planktonic 14 C values (Table A.3). This calculation does not reflect time of isolation from the atmosphere, but it is useful in evaluating possible artifacts in the benthic age model. Planktonic 14 C is enriched relative to the atmosphere younger than ∼8000 cal years BP on Age Model 2, resulting in negative projection ages; we reject this result as an artifact of tuning. 3.3. Northeast Pacific paleoventilation Age Model 1 yields mean benthic projection ages for Holocene time (<11,700 cal years BP) of 1880 ± 180 years (±1-σ ), and 2000 ± 340 years for the prior interval back to ∼17,400 cal years BP (Fig. 6). Relatively high benthic projection ages (i.e., >2000 years) occur intermittently during the glacial transition, with the highest values observed between ∼14,600 and 11,900 cal years BP. The greatest benthic projection age of 2720 ± 130 years occurs at 12,290 ± 105 cal years BP and coincides with anomalously high benthic–planktonic 14 C age differences of up to 1420 ± 40 years (Fig. 6), consistent with the presence of an aged watermass in the intermediate-depth North Pacific at that time. In Age Model 2, the correlation to Greenland produces two distinct intervals of high benthic projection ages (Fig. 6). The older and larger excursion implies an increase in benthic projection ages to 3000 years between 15,770 ± 60 and 14,800 ± 60 cal years BP, with a peak of 3500 ± 80 years at 15,120 ± 60 cal years BP. In the younger excursion benthic projection ages are >2500 years between 14,030 ± 500 and 12,450 ± 230 cal years BP, with a peak of 3130 ± 290 years at 12,840 ± 290 cal years BP, similarly timed to, although larger in magnitude than, the large ventilation–age anomaly reconstructed on Age Model 1 and supported by the

uncalibrated benthic–planktonic 14 C data. These two events are reminiscent of (although not identical in timing to) benthic projection age anomalies inferred off Baja California and Oman based on an age model similarly tuned to the Greenland GISP2 ice core (Marchitto et al., 2007; Bryan et al., 2010) (Fig. 5; Fig. 6). Note, however, that the anomalously high benthic projection ages found by this method in the Gulf of Alaska site (and also in the Baja site) occur mostly in the intervals interpolated between the correlation points, and thus may be artifacts of this interpolation. The planktonic foraminifera 14 C data off Alaska, on Age Model 2, also produce anomalously high near-surface water projection ages, >2500 years, during the older of the two events (Table A.1; Fig. 6). If correct, this interpretation would imply that a hypothetical radiocarbon-depleted deep glacial watermass spread in essentially unaltered form to the high North Pacific, vented to the atmosphere here, and contributed to the abrupt drop in atmospheric 14 C observed during Termination 1. However, such anomalously old apparent reservoir ages for planktonic foraminifera are inconsistent with other regional reconstructions based on plateau tuning (Sarnthein et al., 2007) and are implausibly of the same amplitude as the benthic events in spite of the dominance of watermass downwelling on this margin. Simple ocean models (Hain et al., 2011) suggest that such a scenario is unlikely; rapid dilution of intermediate-depth radiocarbon anomalies by strong ocean mixing near the sea surface and relatively rapid isotopic exchange of carbon with the atmosphere would almost certainly attenuate the surface ocean reservoir-age anomalies relative to those at depth. Age Model 2 would imply rapid glacier retreat at ∼15,000 cal years BP, preceded by relatively constant sedimentation rates of ∼150 cm/ky within the glacial–marine unit and followed by an abrupt reduction to ∼20 cm/ky. This is inconsistent with cosmogenic isotope exposure dates and terrestrial radiocarbon constraints that suggest earlier glacier retreat (Briner et al., 2005; Shennan, 2009), as well as with sedimentological changes in core EW0408-85JC. The MSCL physical properties data sets, along with

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Fig. 5. Benthic 14 C values calculated using (a) Age Model 1 (solid green diamonds) and (b) Age Model 2 (hollow blue diamonds), plotted with ±1-σ uncertainties. The IntCal13 atmospheric radiocarbon calibration curve (black; Reimer et al., 2013) with ±1-σ uncertainty window (gray) is shown on both panels. Dashed lines indicate projection age decay pathways. Benthic 14 C data in the Age Model 2 panel are superimposed upon results from Baja California MV99-GC31/PC08 (pale red triangles; Marchitto et al., 2007), shown on the published GISPII-tuned timescale. On both panels the modern pre-anthropogenic 14 C at an equivalent density horizon to 682 m water depth is shown as a red star (Sabine et al., 2005). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the CT-scan image, show a clear transition from glacier-proximal diamict to hemipelagic sedimentation at 831 cmbsf, identified in both the tuned and calibrated age models as the transition to warm/fresh regional surface ocean conditions at ∼14,700 cal years BP (Fig. 7). In calibrated-planktonic Age Model 1, sediment accumulation rates begin to decrease ∼70 cm deeper in the core, at 904 cmbsf. P-wave velocity, magnetic susceptibility, and gamma density all increase, and CT-scans document a shift to a coarser matrix material in the interval 904–831 cmbsf relative to deeper (older) intervals (Fig. 7). We attribute the decrease in sedimentation rate in this interval for Age Model 1 to a reduction in fine-grained sediment delivery and/or winnowing at the site during glacial retreat. The large projection age anomalies produced by Age Model 2 for both benthic and planktic foraminifera in this interval are most likely artifacts produced by the assumption of constant sediment accumulation rate through this interval of changing lithology. We thus reject the projection age anomalies in Age Model 2 as an artifact of a forced long-distance correlation of regional climate variations with the Greenland Ice core. Although it is plausible that some climatic events in Alaska and Greenland are coeval (for example, the abrupt warming associated with the onset of the Bølling

61

interstade appears to be synchronous), correlation for the purposes of age-model determination should be undertaken cautiously and with support from corroborating evidence. Age Model 1 implies that sediment accumulation rates in the glacial–marine diamict were >500 cm/ky, an order of magnitude greater than average Holocene sedimentation, but fell to ∼50 cm/ky by 17,010 ± 150 cal years BP, prior to the end of glacial–marine sedimentation (Fig. 7). This apparent decrease in sedimentation rate within the diamict is associated with changes in P-wave velocity, sediment density, and magnetic susceptibility that evidence a reduction of fine-grained particles in this interval (Fig. 7). Glacial–marine sediment accumulation rates today are observed to fall exponentially with distance from ice terminus (Cowan and Powell, 1991). Such a decrease in sedimentation rate within the interval of glacial–marine sedimentation in core EW0408-85JC implies sequential retreat of the marine margin of the Bering Glacier between ∼17,000 and ∼15,000 cal years BP, consistent with regional land-based dates (Briner et al., 2005; Shennan, 2009), and pullback onto land by 14,820 ± 110 cal years BP. This suggests that initial retreat of ice here was essentially synchronous with that of the southern Cordilleran Ice Sheet (Porter and Swanson, 1998; Cosma and Hendy, 2008), as well as with glaciers in Europe, South America, and New Zealand during the so-called “Mystery Interval” (Denton et al., 1999; Broecker and Barker, 2007), and likely associated with the initial rapid deglacial rise of atmospheric CO2 and the warming that followed (Shakun et al., 2012). Age Model 1 thus suggests an increase in the apparent 14 C age of subsurface watermasses during the deglaciation (between ∼14,600 and 11,900 cal years BP). We find no evidence for ventilation events greater than today, as has been inferred in the western Pacific (Okazaki et al., 2010, 2012, 2014; Max et al., 2014). The likely implication is a reduction of ventilation or an incursion of older water masses during the deglacial transition. This assumes a relatively constant preformed 14 C value in intermediate sources waters. An alternate interpretation could be an increase in the preformed age of the source waters around Antarctica or elsewhere (Marchitto et al., 2007; Skinner et al., 2010; Lund et al., 2011), although Lund (2013) discounts this hypothesis. The comparison of projection ages based on Age Models 1 and 2 illustrates the sensitivity of projection age calculations to age model assumptions. We present these age models not as specific alternatives, but as commonly-employed simplistic approaches among the range of options that must be considered when interpreting regional radiocarbon anomalies. An alternate way to evaluate oceanic radiocarbon information is based on benthic– planktonic age differences. Between ∼795 and ∼780 cmbsf in the core the uncalibrated benthic–planktonic 14 C age differences increase to over 1000 years, reaching a maximum of 1420 ± 40 years at 790 cmbsf (dated at 12,290 ± 105 cal years BP on calibrated Age Model 1, or 12,450 ± 230 cal years BP on tuned Age Model 2; Fig. 6). This event is coincident with anomalously high projection ages on both age models, and with the younger of two projection age anomalies inferred off Baja California (Marchitto et al., 2007). The existence of this event is independent of the details of chronology or the projection calculations and it is sufficiently brief (<1500 yr) that it is unlikely to be an artifact of variations in radiocarbon production rates; it likely documents the incursion of a relatively 14 C-deficient “old” watermass into the shallow subsurface North Pacific during the second half of Termination 1. The incursion of an older watermass into the shallow subsurface North Pacific between ∼12,500 and ∼12,000 cal years BP corresponds to the younger of two deglacial episodes of rapid rise in atmospheric CO2 (Monnin et al., 2001, timescale of Lemieux-Dudon et al., 2010; Fig. 8). At this time, the benthic–

62

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Fig. 6. Comparison of North Pacific ventilation scenarios: (a) benthic minus planktonic foraminiferal raw radiocarbon ages on the chronology of Age Model 1 (open green diamonds), (b) benthic projection ages reconstructed using Age Model 1 (solid green diamonds), (c) benthic (solid blue circles) and planktonic (pale blue squares) projection ages reconstructed using tuned Age Model 2, and (d) benthic projection ages from Baja California MV99-GC31/PC08 (solid red triangles; Marchitto et al., 2007). All plotted uncertainties are ±1-σ . Note that planktonic projection ages for Age Model 2 are only shown beyond 8000 cal years BP; at younger ages tuning produces 14 C values close to or greater than those of the coeval atmosphere, which must be spurious. Two deglacial intervals of benthic hypoxia identified in core EW0408-85JC are delineated by green bars (Davies et al., 2011). The current estimated range for the timing of Beringian shelf inundation is shown as a gray bar (Elias et al., 1996; Cook et al., 2005; Keigwin et al., 2006; England and Furze, 2008). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

planktonic age difference at core EW0408-85JC converges to nearly the same value as deeper Northeast Pacific sites (Lund et al., 2011). In contrast, we find no equivalent evidence for similar homogenization of subsurface radiocarbon between intermediate and deep waters associated with the older event of rapid CO2 rise beginning at ∼18,500 cal years BP. Independent of calculations of projection age, the homogenization of benthic–planktonic age differences across a broad range of depths during the younger CO2 rise implies that the North Pacific could be part of the pathway transferring “aged” carbon from the deep ocean to the atmosphere (in turn contributing to a relatively rapid reduction in the 14 C signature of atmospheric carbon) during the late stages of the deglacial transition. This does not imply that venting of an aged watermass at the sea surface actually occurred near the site of core EW0408-85JC. If the 14 C-deficient subsurface watermass was also present at the sea surface, then benthic–planktonic age differences would be reduced; strong haline stratification and convergent downwelling along the continental margin likely precludes net release of marine CO2 to the atmosphere here. If the older watermass observed in core EW0408-85JC vented CO2 to the atmosphere in the North Pacific, a more likely place for that to occur would be in the western Pacific or Bering Sea, where winter

overturn of the watermass occurs today. Unlike the highly stratified Gulf of Alaska, the venting region would likely record low benthic–planktonic age differences because of anomalously high surface–ocean reservoir ages. Such low age differences have been observed in the Okhotsk and Bering Sea during deglaciation (Max et al., 2014) although they have been interpreted as evidence of more vigorous intermediate water circulation (accompanied by high benthic δ 13 C) rather than as anomalously old surface–ocean reservoir ages. Our finding of modestly reduced ventilation during late glacial time relative to modern is supported by evidence for greater watercolumn stratification during times of low sea level when the Bering Strait was closed (Zahn et al., 1991; Sigman et al., 2004). The well documented events of regional deglacial hypoxia from 14,800–13,000 cal years BP and 11,200–10,800 cal years BP are not related to abrupt stagnation events, but more likely are associated with high production and export of organic matter from near-surface waters (Davies et al., 2011; Cook et al., 2005). It is plausible, however, that the general reduction in ventilation during the deglacial interval may have pre-conditioned the system to be more sensitive to hypoxia caused by peaks in productivity.

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

63

Fig. 7. The interval of EW0408-85JC between 600 and 1200 cmbsf, showing (a) Age Model 1 (solid green line; calibrated dates shown as solid green diamonds), and Age Model 2 (dashed blue line; tie points shown as solid blue circles), (b) sediment accumulation rates implied by Age Model 1 (solid green line) and by Age Model 2 (dashed blue line), (c) P-wave velocity (orange solid line; caps indicate intervals of poor acoustic coupling, usually at core section boundaries), (d) MSCL magnetic susceptibility (solid gray line), (e) gamma density (solid red line, gaps occur at core section boundaries), and (f) the false-color CT-scan image of the deglacial section of the core (“warmer” colors indicate higher density; see Davies et al., 2011). A light gray bar highlights the interval between 831 and 904 cmbsf, in which sediment accumulation decrease relative to peak glacial values in Age Model 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

On Age Model 1, benthic projection ages decreased (implying more effective or faster ventilation) starting at 11,510 ± 115 cal years BP (Fig. 6); Holocene (0–11,700 years BP) benthic projection ages average 1880 ± 180 years (standard error of mean, n = 18). The apparent increase in ventilation occurs at about the same time as the flooding of the Beringian continental shelf and the opening of a connection to the Arctic Oceans via the Bering Strait. Radiocarbon dates of ∼11,500 14 C years (with a reservoir age correction of 750 years) associated with the appearance of a Pacific bivalve in the Arctic Ocean, suggest that Bering Strait opened by ∼13,200 cal years BP (England and Furze, 2008). Radiocarbon dates of 11,000 ± 60 14 C years on nowsubmerged terrestrial peats indicate Bering Strait opening prior to ∼12,900 cal years BP (Elias et al., 1996). A marine foraminiferal

record of transition from estuarine to fully marine conditions north of Bering Strait is dated at ∼10,900 ± 140 14 C years (with a reservoir age correction of 300 years) suggests Bering Strait opening near ∼11,800 cal years BP (Cook et al., 2005; Keigwin et al., 2006). Although flooding of the Bering shelves may have been a gradual process, multiple lines of evidence suggest that the marine connection between the Arctic and Pacific oceans likely occurred at some point between ∼13,200–11,800 cal years BP. Today, ventilation of North Pacific Intermediate Water depends on the formation of Dense Shelf Water in the Sea of Okhotsk (Talley, 1991; Watanabe and Wakatsuchi, 1998). Northward export of buoyant low salinity surface waters via the Bering Strait helps this process by weakening haline stratification in the North Pacific (Emile-Geay et al., 2003; de Boer and Nof, 2004). The rise of

64

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Fig. 8. (a) Atmospheric CO2 from EPICA Dome C (Monnin et al., 2001; Flückiger et al., 2002), here presented on the updated NGRIP-correlated chronology of Lemieux-Dudon et al. (2010) after interpolation to 250 yr resolution and smoothing to 750 yr resolution (black), as well as the rate of change of CO2 per kyr (dCO2 /dt), illustrating the two deglacial episodes of rapid global CO2 rise (gray). We compare the rate of change in CO2 to the uncalibrated benthic–planktonic 14 C content, reflecting the radiocarbon gradient between the surface and intermediate-depth North Pacific, on (b) calibrated Age Model 1 and (c) tuned Age Model 2. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

sea level that flooded the Beringian shelf and opened the Bering Strait for northward export of low-salinity surface waters may have conditioned the North Pacific for more effective ventilation of intermediate water during Holocene time. 4. Conclusions Evaluated on a plausible chronology provided by the calibrated planktonic radiocarbon dates (Age Model 1), projected ventilation ages for intermediate waters of the Northeast Pacific were on average equivalent or greater during glacial and deglacial time (2000 ± 340 years between ∼17,400 and 11,900 cal years BP, standard error of mean, n = 20), than during Holocene time (1830 ± 170 years, standard error of mean, n = 18). The apparent decrease in subsurface ventilation ages to Holocene values follows inundation of the Bering shelves and opening of Bering Strait, underscoring the importance of flooded continental shelves and Arctic export of low-salinity surface water to the North Pacific formation of intermediate waters. Prior to ∼14,500 cal years BP we find no evidence for an aged intermediate watermass in the Northeast Pacific. However, independent of the choice of age model, high benthic–planktonic radiocarbon age differences indicate near-homogenization of abyssal

radiocarbon between intermediate and deep waters (Lund et al., 2011), coincident with the younger of two deglacial episodes of rapid rise in atmospheric CO2 (Monnin et al., 2001). This finding likely implicates the high latitude North Pacific (although not the Gulf of Alaska) as a source of venting of aged carbon from the deep ocean to the atmosphere during the younger event of deglacial CO2 rise around 12,500–12,000 cal years BP. Acknowledgements The authors acknowledge the crew and science party of R/V Maurice Ewing cruise EW0408, the staff of the OSU Marine Geology Repository, the OSU CEOAS Stable Isotope and the UC Irvine Keck AMS laboratories. We thank N. Pisias, A. Schmittner, E. Brook, S. Lancaster, S. Fallon, and four anonymous reviewers for comments and suggestions. This work was funded by NSF grants AGS-0602395 (Project PaleoVar) and EAR-0711584. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.epsl.2014.04.004.

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

References Adkins, J.F., Boyle, E.A., 1997. Changing atmospheric 14 C and the record of deep water paleoventilation ages. Paleoceanography 12, 337–344. Basak, C., Martin, E.E., Horikawa, K., Marchitto, T., 2010. Southern Ocean source of 14 C-depleted carbon in the North Pacific Ocean during the last deglaciation. Nat. Geosci. 3, 770–773. Briner, J.P., Kaufman, D.S., Manley, W.F., Finkel, R.C., Caffee, M.W., 2005. Cosmogenic exposure dating of late Pleistocene moraine stabilization in Alaska. Geol. Soc. Am. Bull. 117, 1108–1120. Broecker, W.S., Barker, S., 2007. A 190h drop in atmosphere’s. 14 C during the ‘Mystery Interval’ (17.5–14.5 kyr). Earth Planet. Sci. Lett. 256, 90–99. Broecker, W.S., Clark, E., 2010. Search for a glacial-age 14 C-depleted ocean reservoir. Geophys. Res. Lett. 37. LI3606. Broecker, W., Peng, T.-H., Trumbore, S., Bonani, G., 1990. The distribution of radiocarbon in the glacial ocean. Glob. Biogeochem. Cycles 4, 103–117. Broecker, W., Clark, E., Barker, S., 2008. Near constancy of the Pacific Ocean surface to mid-depth radiocarbon age difference over the last 20 kyr. Earth Planet. Sci. Lett. 274, 322–326. Bryan, S.P., Marchitto, T.M., Lehman, S.J., 2010. The release of 14 C-depleted carbon from the deep ocean during the last deglaciation: evidence from the Arabian Sea. Earth Planet. Sci. Lett. 298, 244–254. Burke, A., Robinson, L.F., 2012. The Southern Ocean’s role in carbon exchange during the last deglaciation. Science 335, 557–561. Cook, M.S., Keigwin, L.D., Sancetta, C.A., 2005. The deglacial history of surface and intermediate water of the Bering Sea. Deep-Sea Res., Part 2, Top. Stud. Oceanogr. 52, 2163–2173. Cosma, T., Hendy, I.L., 2008. Pleistocene glacimarine sedimentation on the continental slope off Vancouver Island, British Columbia. Mar. Geol. 255, 45–54. Cowan, E.A., Powell, R.D., 1991. Ice-proximal sediment accumulation rates in a temperate glacial fjord, southeastern Alaska. In: Anderson, J.B., Ashley, G.M. (Eds.), Glacial Marine Sedimentation: Paleoclimatic Significance. In: Spec. Pap., Geol. Soc. Am., vol. 261, pp. 61–73. Davies, M.H., Mix, A.C., Stoner, J.S., Addison, J.A., Jaeger, J., Finney, B., Wiest, J., 2011. The deglacial transition on the Southeastern Alaska Margin: meltwater input, sealevel rise, marine productivity, and sedimentary anoxia. Paleoceanography 26, PA2223. de Boer, A.M., Nof, D., 2004. The Bering Strait’s grip on the Northern Hemisphere climate. Deep-Sea Res., Part 1, Oceanogr. Res. Pap. 51, 1347–1366. de Boer, A.M., Sigman, D.M., Toggweiler, J.R., Russell, J.L., 2007. Effect of global ocean temperature on deep ocean ventilation. Paleoceanography 22, PA2210. De Pol-Holz, R., Keigwin, L., Southon, J., Hebbeln, D., Mohtadi, M., 2010. No signature of abyssal carbon in intermediate waters off Chile during deglaciation. Nat. Geosci. 3, 192–195. Denton, G.H., Heusser, C.L., Lowell, T.V., Moreno, P.I., Anderson, B.G., Heusser, L.E., Schlüchter, C., Marchant, D.R., 1999. Interhemispheric linkage of paleoclimate during the last glaciation. Geogr. Ann., Ser. A, Phys. Geogr. 81A, 107–153. Elias, S.A., Short, S.K., Nelson, C.H., Birks, H.H., 1996. Life and times of the Bering land bridge. Nature 382, 60–63. Emile-Geay, J., Cane, M.A., Naik, N., Seager, R., Clement, A.C., van Geen, A., 2003. Warren revisited: atmospheric freshwater fluxes and “Why is no deep water formed in the North Pacific”. J. Geophys. Res. 108, 3178. England, J.H., Furze, M.F.A., 2008. New evidence from the western Canadian Arctic archipelago for the resubmergence of Bering Strait. Quat. Res. 70, 60–67. Flückiger, J., Monnin, E., Stauffer, B., Schwander, J., Stocker, T.F., Chappellaz, J., Raynaud, D., Barnola, J.-M., 2002. High resolution Holocene N2 O ice core record and its relationship with CH4 and CO2 . Glob. Biogeochem. Cycles 16, GB1020. Galbraith, E.D., Jaccard, S.L., Pedersen, T.F., Sigman, D.M., Haug, G.H., Cook, M., Southon, J.R., Francois, R., 2007. Carbon dioxide release from the North Pacific abyss during the last deglaciation. Nature 499, 890–893. Hain, M.P., Sigman, D.M., Haug, G.H., 2011. Shortcomings of the isolated abyssal reservoir model for deglacial radiocarbon changes in the mid-depth Indo-Pacific Ocean. Geophys. Res. Lett. 38, L04604. Haslett, J., Parnell, A., 2008. A simple monotone process with application to radiocarbon-dated depth chronologies. J. R. Stat. Soc., Ser. C, Appl. Stat. 57, 399–418. Hughen, K., Southon, J., Lehman, S., Bertrand, C., Turnbull, J., 2006. Marine-derived 14 C calibration and activity record for the past 50,000 years updated from the Cariaco Basin. Quat. Sci. Rev. 25, 3216–3227. Keeling, R.F., 2007. Deglaciation mysteries. Science 316, 1440–1441. Keigwin, L.D., 2002. Late Pleistocene–Holocene paleoceanography and ventilation of the Gulf of California. J. Oceanogr. 58, 421–432. Keigwin, L.D., Donnelly, J.P., Cook, M.S., Driscoll, N.W., Brigham-Grette, J., 2006. Rapid sea-level rise and Holocene climate in the Chukchi Sea. Geology 34, 861–864. Kennett, J.P., Ingram, B.L., 1995. A 20,000-year record of ocean circulation and climate change from the Santa Barbara basin. Nature 377, 510–514. Key, R.M., Kozyr, A., Sabine, C.L., Lee, K., Wanninkhof, R., Bullister, J.L., Feely, R.A., Millero, F.J., Mordy, C., Peng, T.-H., 2004. A global ocean carbon climatology: results from GLODAP. Glob. Biogeochem. Cycles 18, GB4031.

65

Lemieux-Dudon, B., Blayo, E., Petit, J.-R., Waelbroeck, C., Svensson, A., Ritz, C., Barnola, J.-M., Narcisi, B.M., Parrenin, F., 2010. Consistent dating for Antarctic and Greenland ice cores. Quat. Sci. Rev. 29, 8–20. Lund, D.C., 2013. Deep Pacific ventilation aged during the last deglaciation: evaluating the influence of diffusive mixing and source region reservoir age. Earth Planet. Sci. Lett. 381, 52–62. Lund, D.C., Mix, A.C., 1998. Millennial-scale deep water oscillations: reflections of the North Atlantic in the deep Pacific from 10 to 60 ka. Paleoceanography 13, 10–19. Lund, D.C., Mix, A.C., Southon, J., 2011. Increased ventilation age of the deep north– east Pacific Ocean during the last deglaciation. Nat. Geosci. 4, 771–774. Marchitto, T.M., Lehman, S.J., Ortiz, J.D., Flückiger, J., van Geen, A., 2007. Marine radiocarbon evidence for the mechanisms of deglacial atmospheric CO2 rise. Science 316, 1456–1459. Max, L., Lembke-Jene, L., Riethdorf, J.-R., Tiedemann, R., Nürnberg, D., Kühn, H., Mackensen, A., 2014. Pulses of enhanced North Pacific Intermediate Water ventilation from the Okhotsk Sea and Bering Sea during the last deglaciation. Clim. Past 10, 591–605. McKay, J.L., Pedersen, T.F., Southon, J., 2005. Intensification of the oxygen minimum zone in the northeast Pacific off Vancouver Island during the last deglaciation: ventilation and/or export production? Paleoceanography 20, PA4002. McNeely, R., Dyke, A.S., Southon, J.R., 2006. Canadian marine reservoir ages, preliminary data assessment. Geol. Surv. Can. Open File 5049, 3 pp. Mix, A.C., Lund, D.C., Pisias, N.G., Bodén, P., Bornmalm, L., Lyle, M., Pike, J., 1999. Rapid climate oscillations in the Northeast Pacific during the last deglaciation reflect Northern and Southern Hemisphere sources. In: Clark, P.U., et al. (Eds.), Mechanisms of Global Climate Change at Millennial Time Scales. In: Geophys. Monogr., vol. 112. American Geophysical Union, pp. 127–148. Molnia, B.F., 1986. Late Wisconsin glacial history of the Alaska continental margin. In: Hamilton, T.D., Reed, K.M., Thorson, R.M. (Eds.), Glaciation in Alaska: The Geologic Record, pp. 99–144. Monnin, E., Indermühle, A., Dällenbach, A., Flückiger, J., Stauffer, B., Stocker, T.F., Raynaud, D., Barnola, J.-M., 2001. Atmospheric CO2 concentrations over the Last Glacial Termination. Science 291, 112–114. Okazaki, Y., Kimoto, K., Asahi, H., Sato, M., Nakamura, Y., Harada, N., 2014. Glacial to deglacial ventilation and productivity changes in the southern Okhotsk Sea. Palaeogeogr. Palaeoclimatol. Palaeoecol. 395, 53–66. Okazaki, Y., Sagawa, T., Asahi, H., Horikawa, K., Onodera, J., 2012. Ventilation changes in the western North Pacific since the last glacial period. Clim. Past 8, 17–24. Okazaki, Y., Timmermann, A., Menviel, L., Harada, N., Abe-Ouchi, A., Chikamoto, M.O., Mouchet, A., Asahi, A., 2010. Deepwater formation in the North Pacific during the Last Glacial Termination. Science 329, 200–204. Ortiz, J.D., O’Connell, S.B., Del Viscio, J., Dean, W., Carriquiry, J.D., Marchitto, T., Zheng, Y., van Geen, A., 2004. Enhanced marine productivity off western North America during warm climate intervals of the past 52 k.y. Geology 32, 521–524. Parnell, A.C., Haslett, J., Allen, J.R.M., Buck, C.E., Huntley, B., 2008. A flexible approach to assessing synchroneity of past events using Bayesian reconstructions of sedimentation history. Quat. Sci. Rev. 27, 1872–1885. Pena, L.D., Cacho, I., Ferretti, P., Hall, M.A., 2008. El Ninõ – Southern Oscillation – like variability during glacial terminations and interlatitudinal teleconnections. Paleoceanography 23, PA3101. Porter, S.C., Swanson, T.W., 1998. Radiocarbon age constraints on rates of advance and retreat of the Puget Lobe of the Cordilleran Ice Sheet during the last glaciation. Quat. Res. 50, 205–213. Rasmussen, S.O., Andersen, K.K., Svensson, A.M., Steffensen, J.P., Vinther, B.M., Clausen, H.B., Siggard-Andersen, M.-L., Johnson, S.J., Larsen, L.B., Dahl-Jensen, D., Bigler, M., Röthlisberger, R., Fischer, H., Goto-Azuma, K., Hansson, M.E., Ruth, U., 2006. A new Greenland ice core chronology for the last glacial termination. J. Geophys. Res. 111, D06102. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., Hatté, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 54, 1869–1887. Sabine, C.L., Key, R.M., Kozyr, A., Feely, R.A., Wanninkhof, R., Millero, F.J., Peng, T.-H., Bullister, J.L., Lee, K., 2005. Global ocean data analysis project: results and data. Rep. ORNL/CDIAC-145/NDP-083 Carbon Dioxide Inf. Anal. Cent. Oak Ridge Natl. Lab., U.S. Dep. of Energy, Oak Ridge, Tenn. Sarnthein, M., Grootes, P.M., Kennett, J.P., Nadeau, M.-J., 2007. 14 C reservoir ages show deglacial changes in ocean currents and carbon cycle. AGU Geophys. Monogr. Ser. 173, 175–196. Sarnthein, M., Schneider, B., Grootes, P.M., 2013. Peak glacial 14 C ventilation ages suggest major draw-down of carbon into the abyssal ocean. Clim. Past 9, 2595–2614. Shakun, J.D., Clark, P.U., He, F., Marcott, S.A., Mix, A.C., Liu, Z., Otto-Bliesner, B., Schmittner, A., Bard, E., 2012. Global warming preceded by increasing carbon dioxide concentrations during the last deglaciation. Nature 484, 49–54. Shennan, I., 2009. Late Quaternary sea-level changes and palaeoseismology of the Bering Glacier region, Alaska. Quat. Sci. Rev. 28, 1762–1773.

66

M. Davies-Walczak et al. / Earth and Planetary Science Letters 397 (2014) 57–66

Sigman, D.M., Boyle, E.A., 2000. Glacial/interglacial variations in atmospheric carbon dioxide. Nature 407, 859–869. Sigman, D.M., Jaccard, S.L., Haug, G.H., 2004. Polar ocean stratification in a cold climate. Nature 428, 59–63. Skinner, L.C., Shackleton, N.J., 2004. Rapid transient changes in northeast Atlantic deep water ventilation age across Termination 1. Paleoceanography 19. PA2005. Skinner, L.C., Fallon, S., Waelbroeck, C., Michel, E., Barker, S., 2010. Ventilation of the deep southern ocean and deglacial CO2 rise. Science 328, 1147–1151. Stephens, B.B., Keeling, R.F., 2000. The influence of Antarctic sea ice on glacial– interglacial CO2 variations. Nature 404, 171–173. Stott, L., Southon, J., Timmermann, A., Koutavas, A., 2009. Radiocarbon age anomaly at intermediate water depth in the Pacific Ocean during the last deglaciation. Paleoceanography 25, PA2223. Stuiver, M., Polach, H.A., 1977. Discussion: reporting of 14 C data. Radiocarbon 19, 355–363. Talley, L.D., 1991. An Okhotsk Sea water anomaly: implications for ventilation in the North Pacific. Deep-Sea Res., A, Oceanogr. Res. Pap. 38, S171–S190. Toggweiler, J.R., 1999. Variation of atmospheric CO2 by ventilation of the ocean’s deepest water. Paleoceanography 14, 571–588.

van Geen, A., Fairbanks, R.G., Dartnell, P., McGann, M., Gardner, J.V., Kashgarian, M., 1996. Ventilation changes in the northeast Pacific during the Last Deglaciation. Paleoceanography 11, 519–528. Walinsky, S.E., Prahl, F.G., Mix, A.C., Finney, B.P., Jaeger, J.M., Rosen, G.P., 2009. Distribution and composition of organic matter in surface sediments of coastal southeast Alaska. Cont. Shelf Res. 29, 1565–1579. Watanabe, T., Wakatsuchi, M., 1998. Formation of 26.8–26.9 σ θ water in the Kuril Basin of the Sea of Okhotsk as a possible origin of North Pacific Intermediate Water. J. Geophys. Res. 103, 2849–2865. Zahn, R., Pedersen, T.F., Bornhold, B.D., Mix, A.C., 1991. Water mass conversion in the glacial subarctic Pacific (54◦ N, 148◦ W). Physical constraints and the benthic– planktonic stable isotope record. Paleoceanography 6, 543–560. Zellers, S.D., Mueller, K., Hill, D., 2011. Variation in Foraminiferal distributions across the Pleistocene/Holocene transition off the Kayak Slope, Northern Gulf of Alaska. In: PACLIM Workshop, Asilomar State Conference Grounds. Pacific Grove, CA. Zellers, S.D., Ullrich, A.D., Jaeger, J.M., 2009. Characterization of late Neogene benthic Foraminifera in a fjord to slope transect, northern Gulf of Alaska. Abstr. – North Am. Paleontol. Conv. 9, 247.