A Multi-Proxy Reconstruction of Sea Ice Extent and Primary Productivity During Marine Isotope Stage 11 in the Bering Sea

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2020

A multi-proxy reconstruction of sea ice extent and primary productivity during Marine Isotope Stage 11 in the Bering Sea

Natalie Sarah Thompson Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd

Recommended Citation Thompson, Natalie Sarah, "A multi-proxy reconstruction of sea ice extent and primary productivity during Marine Isotope Stage 11 in the Bering Sea" (2020). Graduate Theses and Dissertations. 18237. https://lib.dr.iastate.edu/etd/18237

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. A multi-proxy reconstruction of sea ice extent and primary productivity during Marine Isotope Stage 11 in the Bering Sea

Natalie Thompson

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Co-Majors: Geology; Environmental Science

Program of Study Committee: Beth Caissie, Co-major Professor Alan Wanamaker, Co-major Professor William Gutowski Chris Harding Elizabeth Swanner

The student author, whose presentation of the scholarship herein was approved by the program of study committee, is solely responsible for the content of this dissertation. The Graduate College will ensure this dissertation is globally accessible and will not permit alterations after a degree is conferred.

Iowa State University

Ames, Iowa

2020

DEDICATION

For the many teachers in my life.

Thank you for guiding me, challenging me, and inspiring me.

iii

TABLE OF CONTENTS

Page

LIST OF FIGURES ...... vii

LIST OF TABLES ...... xvi

ACKNOWLEDGMENTS ...... xvii

ABSTRACT ...... xix

CHAPTER 1. INTRODUCTION ...... 1 Motivation ...... 1 The Role of Sea Ice in the Global Climate System ...... 3 Effects of Arctic Sea Ice Decline ...... 4 Dissertation Structure ...... 6

CHAPTER 2. BACKGROUND AND LITERATURE REVIEW ...... 14 Study Area ...... 14 Physiography and Oceanography of the Bering Sea ...... 14 Sea Ice Variability in the Bering Sea ...... 16 Primary Productivity in the Bering Sea ...... 19 Marine Isotope Stage 11 ...... 22 Marine Isotope Stage 11 in the Bering Sea ...... 25 Summary ...... 27 References ...... 28

CHAPTER 3. EVALUATING THE PALEOENVIRONMENTAL SIGNIFICANCE OF SEDIMENT GRAIN SIZE IN BERING SEA SEDIMENTS DURING MIS 11 ...... 42 Abstract ...... 42 Introduction ...... 43 Study Area ...... 45 Location and Oceanographic Setting ...... 45 Sedimentation in the Bering Sea ...... 46 Core Sites ...... 47 Methods ...... 48 Geochronology ...... 48 Grain Size Measurements ...... 48 Statistical Analyses ...... 49 Smear Slide Analyses ...... 51 Results ...... 52 Sediment Composition ...... 52 Grain Size Parameters ...... 53

Size Fractions ...... 53 Statistical Measures ...... 56 Discussion ...... 59 Interpreting Grain Size Parameters ...... 59 Proxy for Sediment Transport and Deposition ...... 59 Distinguishing Sediment Facies ...... 62 Proxy for Ice Rafting ...... 66 Proxy for Paleocurrent Strength ...... 70 Proxy for Paleoproductivity ...... 72 Summary ...... 77 References ...... 78

CHAPTER 4. A MULTI-PROXY RECONSTRUCTION OF CHANGES IN SEA ICE AND PRIMARY PRODUCTIVITY AT IODP SITE U1339 (UMNAK PLATEAU, BERING SEA) DURING MARINE ISOTOPE STAGE 11 ...... 89 Abstract ...... 89 Introduction ...... 90 Background ...... 91 Marine Isotope Stage 11 ...... 91 Study Area ...... 94 Oceanographic Setting ...... 94 Core Site Location ...... 96 Methods ...... 96 Geochronology ...... 96 Sediment Sampling ...... 98 Grain Size and Sedimentological Analyses ...... 99 Published Records ...... 99 Wet Sieving ...... 99 Diatom Analyses ...... 99 Stable Isotope Analyses ...... 100 Results ...... 100 Lithology ...... 100 Diatoms ...... 102 Full Diatom Assemblage ...... 102 Niche Groupings ...... 105 Diatom-based Proxy ...... 109 Geochemistry ...... 109 Carbon Content ...... 109 Organic Matter Source ...... 112 Discussion ...... 113 Sea Ice at the Umnak Plateau ...... 113 Sea Ice History ...... 113 Controls on Sea Ice ...... 115 Primary Productivity at the Umnak Plateau ...... 120 v

Marine Isotope Stage 12 ...... 120 Deglaciation ...... 122 Marine Isotope Stage 11 ...... 123 Comparison of records from the Umnak Plateau ...... 124 Summary ...... 129 References ...... 130

CHAPTER 5. REGIONAL VARIATIONS IN SEA ICE, PRIMARY PRODUCTIVITY, AND OCEAN CIRCULATION DURING MARINE ISOTOPE STAGE 11 IN THE BERING SEA ...... 148 Abstract ...... 148 Introduction ...... 149 Background ...... 153 Marine Isotope Stage 11 ...... 153 Study Area ...... 154 Location and Oceanographic Setting ...... 154 Core Sites ...... 157 Site U1345 ...... 157 Site U1343 ...... 157 Site U1339 ...... 157 Materials and Methods ...... 158 Age Models ...... 158 Diatom Counts ...... 159 Published Records ...... 160 Results and Discussion ...... 160 Lithology ...... 160 Sediment Composition ...... 160 Sediment Grain Size ...... 162 Coarse Interval ...... 164 Laminated Sediments ...... 165 Ice-rafted Material ...... 166 Sea Ice History ...... 168 Productivity Trends ...... 172 Wider Context ...... 175 Summary ...... 177 References ...... 178

CHAPTER 6. GENERAL CONCLUSIONS ...... 190

APPENDIX A SUPPLEMENT TO CHAPTER 4: A MULTI-PROXY RECONSTRUCTION OF CHANGES IN SEA ICE AND PRIMARY PRODUCTIVITY AT IODP SITE U1339 (UMNAK PLATEAU, BERING SEA) DURING MARINE ISOTOPE STAGE 11 ...... 193

APPENDIX B GRAIN SIZE REPEATABILITY ...... 198 B.1. Background ...... 198 B.2. Methods ...... 198

APPENDIX C ASH LAYERS ...... 200 vii

LIST OF FIGURES

Page

Figure 2.1. Map of Beringia, showing place names, bathymetric features, and core sites (yellow dots) referred to in the text. The black dashed line shows the maximum extent of sea ice today (median over the period 1979-2013) (Cavalieri et al., 1996). Currents (shown in blue) are modified from Stabeno et al. (1999). Abbreviations include: Alaska Coastal Current (ACC), Aleutian North Slope Current (ANSC), Bering Slope Current (BSC), Unimak Pass (UkP), Bering Strait (BS), Kamchatka Strait (KS), Shirshov Ridge (SR), Bowers Ridge (BR), and Umnak Plateau (UP). Grey bathymetric shading changes value at -50 m (Bering Strait sill depth), -250 m (shelf/slope break), -1000 m, and -2000 m...... 15

Figure 2.2. Satellite measurements of maximum sea ice extent from 1979-2019 for the Bering Sea (blue line) and the Arctic (grey line). Sea ice data is taken from the Sea Ice Index, National Snow and Ice Data Center (Fetterer et al., 2017, updated daily)...... 18

Figure 2.3. Orbital cycles and global climate records of the past 500 ka. (a) Orbital eccentricity (Laskar et al., 2011); (b) obliquity (Laskar et al., 2004); (c) precession (Laskar et al., 2004); (d) northern hemisphere summer insolation at 65°N (Laskar et al., 2004); (e) benthic foraminiferal δ18O records (the LR04 stack, Lisiecki & Raymo, 2005); (f) compilation of atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015) and (g) relative sea level in meters above or below present (Rohling et al., 2010). Grey bars denote interglacial stages 1 through 13...... 24

Figure 3.1. Map of Beringia, showing place names, bathymetric features, and core sites referred to in the text. The black dashed line shows the maximum extent of sea ice today (median over the period 1979-2013) (Cavalieri et al., 1996). Currents (shown in blue) are modified from Stabeno et al. (1999). Abbreviations include: Alaska Coastal Current (ACC), Aleutian North Slope Current (ANSC), Bering Slope Current (BSC), Bering Strait (BS), Navarin Canyon (NC), Zhemchug Canyon (ZC), and Umnak Plateau (UP). Grey bathymetric shading changes value at -50 m (Bering Strait sill depth), -250 m (shelf/slope break), -1000 m, and -2000 m...... 45

points between Bering Sea δ18O (filled) and magnetic susceptibility (open) records and the global stack. The grey bar shows the duration of MIS 11...... 49

Figure 3.3. Microscope images of sediment composition at Bering Sea core sites. (a) siliciclastic sediment, Site U1345A 15H-1 32 cm; (b) mixed terrigenous and biogenic sediment, U1339C 11H-3 101 cm; and (c) biogenic (diatom) sediment, U1339D 11H-6 135 cm...... 53

Figure 3.4. Grain size distribution plots for representative samples from the Bering Sea, showing unimodal, bimodal, trimodal, and polymodal distributions...... 54

Figure 3.5. % clay (<2 μm, brown), silt (2-63 μm, olive), sand (63-2000 μm, yellow), and gravel (>2000 μm, grey) -sized grains for (a) bulk and (b) terrigenous sediments at IODP sites U1345, U1343 and U1339. The grey panel shows the duration of MIS 11...... 55

Figure 3.6. Downcore variations in median grain size, mean grain size, sorting, and skewness for bulk (solid line) and terrigenous (dashed line) sediments from Bering Sea core sites U1345 (black), U1343 (blue), and U1339 (red) from 430- 365 ka. Sorting and skewness values are related to descriptive terms for sorting (poorly sorted [2-4]; very poorly sorted [4-16]) and skewness (symmetrical [-0.1 to 0.1]; fine skewed [-0.3 to -0.1]; coarse skewed [0.1 to 0.3]) after Folk & Ward (1957). Dots indicate the stratigraphic position of smear slide samples; yellow dots are smear slides from Takahashi et al., 2011. Brown bars show ash layers/accessories, and the grey panel shows the duration of MIS 11...... 58

Figure 3.7. Examples of fine-skewed (U1343A 12H-5 78 cm) and coarse-skewed (U1343C 12H-3 55 cm) samples. Note that the x-axis is reversed to follow convention...... 58

Figure 3.8. Grain size distribution (mean grain size and sorting) for the contouritic intervals at sites U1345 and U1343. The core image shows an example of a sand lens from U1343...... 64

Figure 3.9. Idealized bi-gradational mud-sand contourite facies, showing the complete sequence of divisions C1-C5, linked to variations in bottom current velocity. From Rebesco et al., 2014, based on the original figure by Gonthier, 1984...... 66

Figure 3.10. % terrigenous clay-sized grains, and % terrigenous grains in the >150 µm, and >150 µm size fractions, for sites U1345 (black), U1343 (blue), and U1339 (red). Brown bars depict ash layers and accessories, and the grey panel shows the duration of MIS 11...... 69 ix

Figure 3.11. Grain size distribution plot and light microscope image of IRD (magnification 100x) from Site U1343 (12H-5 41-42 cm, 403.4 ka)...... 69

Figure 3.12. Mean size sortable silt (10-63 μm) at IODP sites U1345 (black), U1343 (blue), and U1339 (red), compared to relative sea level (Rohling et al., 2010). The dashed line represents the Bering Strait sill depth (-50 m). The grey panel shows the duration of MIS 11...... 70

Figure 3.13. Cross-plots of mean size sortable silt versus % sortable silt for (a) sites U1345 (black), U1343 (blue), and U1339 (red), and (b) contourite deposits from sites U1345 (black) and U1343 (blue)...... 71

Figure 3.14. Percent diatoms from smear slide analysis versus mean grain size before (closed circles) and after (open circles) biosilica digestion for sites U1345 (black) and U1339 (red)...... 72

Figure 3.15. Difference in mean grain size between the two sets of samples (bulk minus terrigenous) at sites U1345 (black), U1343 (blue) and U1339 (red). Positive values indicate fining of sediment, negative values indicate coarsening. Numbers on the plot correspond to numbered microscope images, which relate changes in sediment grain size to sediment composition. 1. Coarsening, due to poorly preserved (fragmented) diatoms; 2. Fining, due to the presence of coarse silt- and sand-sized diatoms; 3. Coarsening, due to the presence of coarse silt to sand-sized siliciclastic minerals; and 4. Coarsening, due to the combined effects of low diatom content and high volcanogenic input. The grey panel shows the duration of MIS 11...... 75

Figure 3.16. Light microscope image from Site U1339 (U1339C 11H-3 101 cm), highlighting variations in the size of diatom valves...... 76

Figure 4.1. Orbital cycles and global climate records of the past ~500 ka. (a) Orbital eccentricity (Laskar et al., 2011); (b) northern hemisphere summer insolation at 65°N (Laskar et al., 2004); (c) benthic foraminiferal δ18O records (the LR04 stack, Lisiecki & Raymo, 2005); (d) compilation of atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015, and references therein); and (e) relative sea level in meters above or below present (Rohling et al., 2010). The grey bars denote interglacial marine isotope stages 13-1...... 93

Abbreviations include: Alaska Coastal Current (ACC), Aleutian North Slope Current (ANSC), Bering Slope Current (BSC), Bering Strait (BS), Umnak Plateau (UP), and Unimak Pass (UkP). Grey bathymetric shading changes value at -50 m (Bering Strait sill depth), -250 m (shelf/slope break), -1000 m, and -2000 m...... 95

Figure 4.3. Benthic foraminiferal δ18O record from Site U1339 (red, Cook et al., 2016), aligned with the global LR04 stack (black, Lisiecki & Raymo, 2005). Grey bars denote interglacial stages 11-1...... 97

Figure 4.4. Lithostratigraphic column for site U1339, based on shipboard core descriptions and smear slide analyses (Takahashi et al., 2011; Chapter 3). Column width varies according to the median grain size of bulk sediments (Chapter 3). Colors represent the proportion of diatoms in the sediment: silt/clay with 10- 40% diatoms (olive); silt/clay with 40-60% diatoms (dark green); silt/clay with >60% diatoms (yellow); laminated sediments (pale green). Blue circles indicate the presence and open circles indicate the absence of quartz grains >150 µm in the subset of samples picked for quartz grains. The grey side bar shows the duration of MIS 11...... 101

Figure 4.5. Relative % (area plots) and absolute (line plots) abundances of diatom taxa that make up more than 10% of any assemblage. Note that we use Chaetoceros- free counts in all relative % abundance records, with the obvious exception of Chaetoceros RS: full counts can be found in Appendix A, Fig. A.2. Taxa are grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); North Pacific indicator (orange); summer bloom (light green); other (grey); and Chaetoceros RS (dark green). The final line plot shows the total number of diatom valves per gram of sediment. Note the different scale for absolute abundances of Chaetoceros RS and all diatom valves. The grey panel indicates the duration of MIS 11 (424- 374 ka) and the green bar represents laminated sediments...... 103

Figure 4.6. Margalef (black) and Simpson (green) diversity indices. The grey panel shows the duration of MIS 11 (424-374 ka), and the green bar shows laminated sediments...... 104

Figure 4.7. Relative percent abundances of diatoms grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); North Pacific indicator (orange); high productivity (yellow); and Chaetoceros RS (dark green). Note that we use Chaetoceros-free counts in all relative % abundance records, with the obvious exception of Chaetoceros RS: full counts can be found in Appendix A, Fig. A.3. Also shown is sea ice concentration (including upper and lower limits), derived from a quantitative, diatom-based xi

sea ice proxy (Nesterovich and Caissie, in review). The grey panel shows the duration of MIS 11 (424-374 ka) and the green bar represents laminated sediments...... 107

Figure 4.8 Summary of elemental and isotopic carbon and nitrogen analyses, and related diatom proxies (a) percent total carbon; (b) percent organic carbon; (c) % inorganic carbon; (d) δ13C of organic matter; (e) total diatom abundances (valves per gram); (f) relative % abundance of Chaetoceros RS; and (g) bulk sedimentary δ15N. The grey panel shows the duration of MIS 11 (424-374 ka) and the green bar shows laminated sediments...... 110

Figure 4.9. Cross-plots showing (a) total organic carbon vs. total nitrogen; (b) δ13C vs. TOC/N; and (c) δ13C vs. δ15N. Grey boxes identify endmember values for likely sources of OM to the Bering Sea, which include marine phytoplankton (C/N: 5 to 7 [Redfield, 1963; Meyers, 1994], δ13C: -22 to -19‰, and δ15N: 5 to 8‰ [Walinsky et al., 2009]); ice algae (δ13C: -20.5 to -18.5‰, and δ15N: 3.5 to 5.5‰ [Schubert & Calvert, 2001]); soil (C/N: 10-12, δ13C: -26.5 to -25.5‰, and δ15N: 0 to 1‰ [Walinsky et al., 2009]); C3 vascular plants (C/N: > 20 [Redfield, 1963; Meyers, 1994], δ13C: -27 to -25‰ [Schubert and Calvert, 2001], and δ15N: 0 to 1‰ [Walinsky et al., 2009]); and inorganic (clay-bound) nitrogen (δ15N: 2-4‰ [Schubert & Calvert, 2001])...... 111

Figure 4.10. Summary of sea ice records from Site U1339, compared to global records. (a) mean monthly insolation at 65°N (Laskar et al., 2004); (b) atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015); (c) relative sea level in meters above present (Rohling et al., 2010). The dashed vertical line shows the Bering Strait sill depth; (d) volume % terrigenous clay-sized grains (Chapter 3); (e) IRD proxies, including % terrigenous particles >150 µm (dark blue line; Chapter 3), and occurrence of quartz grains >150 µm (open circles show samples devoid of quartz grains; (f) sea ice concentration from diatom-based proxy (Nesterovich & Caissie, in review). Vertical blue lines indicate the cutoff for unconsolidated ice (15-40% concentration) and consolidated ice (>40% concentration); values less than 15% suggest open ocean conditions; (g) relative % abundance of sea ice diatoms; and (h) relative % abundance of North Pacific indicator species. The grey panel shows the duration of MIS 11, and colored side bars denote the following substages: deglaciation (424-420 ka, turquoise), Peak MIS 11 (430-398 ka, red), and Late MIS 11 (398-374 ka, blue). The green bar represents laminated sediments...... 118

sea level in meters above present (Rohling et al., 2010); The dashed vertical line shows the Bering Strait sill depth; (d) bulk median grain size (Chapter 3); (e) diatom valves per gram of sediment); (f) Chaetoceros per gram of sediment; (g) relative % abundance of Chaetoceros RS; (h) % organic carbon; (i) % inorganic carbon; and (j) δ13C (‰ VPDB). Diatom percent abundances are based on Chaetoceros-free counts. The grey panel shows the duration of MIS 11, and colored side bars denote the following substages: deglaciation (424- 420 ka, turquoise), Peak MIS 11 (430-398 ka, red), and Late MIS 11 (398-374 ka, blue). The green bar represents laminated sediments...... 120

Figure 4.12. Location map of core sites on the Umnak Plateau...... 124

Figure 4.13. Compilation of sea ice proxy records from Umnak Plateau sites U1339 (green), RC14-121 (purple), UMK-3A (yellow), and 51JPC (grey) for MIS 11 and 1 (denoted by grey panels). MIS 11 sea ice diatom records (this study) include all sea ice diatoms (lighter green) and conservative sea ice counts, based on seven species (Fragilariopsis cylindrus, Fragilariopsis oceanica, Fragilariopsis regina-jahniae, Fossula arctica, Nitzschia frigida, Sinerima marigela, and Thalassiosira antarctica) with a strong affinity for sea ice (darker green). MIS 1 sea ice records include % Fragilariopsis species (Caissie et al., 2010), % F. cylindrus (Katsuki & Takahashi, 2005), % Nitzschia cylindra and grunowii (Sancetta & Robinson, 1985), and sea ice concentration from the diatom- based proxy (Nesterovich & Caissie, in review). Green bars show laminated sediments, based on core records from U1339 (MIS 11) and 51JPC (MIS 1)...... 126

Figure 4.14. Compilation of productivity records from Umnak Plateau sites U1339 (green), RC14-121 (purple), UMK-3A (yellow), and 51JPC (grey) for MIS 11, 5, and 1 (denoted by grey panels). Note that records do not exist at all sites for all timeframes. MIS 5 records are from Vaughn & Caissie (2017). MIS 1 records include % organic carbon (Pelto et al., 2018), and % Chaetoceros RS (purple, Sancetta & Robinson, 1983; yellow, Katsuki & Takahashi, 2005; and grey, Caissie et al., 2010). Green bars show laminated sediments, based on core records from U1339 (MIS 11 and 5) and 51JPC (MIS 1)...... 127

Figure 5.1. Orbital cycles and global climate records for MIS 11 (right panel, black) and MIS 1 (left panel, blue). (a) Orbital eccentricity (Laskar et al., 2011); (b) precession (Laskar et al., 2004); (c) northern hemisphere summer insolation at 65°N (Laskar et al., 2004); (d) compilation of atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015, and references therein); (e) benthic foraminiferal δ18O records (the LR04 stack, Lisiecki & Raymo, 2005); and (f) relative sea level in meters above or below present (Rohling et al., 2010). Grey bars show the duration of the interglacial stages 11 and 1. MIS 11 substages include xiii

deglaciation (turquoise bar, 424-420 ka), Peak MIS 11 (red bar, 420-398 ka; Kandiano et al., 2012), and Late MIS 11 (blue bar, 398-374 ka)...... 152

Figure 5.2. Map of Beringia, showing place names, bathymetric features, and core sites referred to in the text. The dashed line shows the maximum extent of sea ice today (median for the period 1979-2013) (Cavalieri et al., 1996). Currents (shown in blue) are modified from Stabeno et al. (1999). Abbreviations include: Alaska Coastal Current (ACC), Aleutian North Slope Current (ANSC), Bering Slope Current (BSC), Bering Strait (BS), Bowers Ridge (BR), Umnak Plateau (UP)...... 156

Figure 5.3. Benthic foraminiferal δ18O values for IODP sites U1345 (black; Cook et al., 2016), U1343 (blue; Asahi et al., 2016) and U1339 (red; Cook et al., 2016), aligned with the LR04 global stack (grey; Lisiecki and Raymo, 2005). Inverted triangles show tie points between Bering Sea δ18O (filled) and magnetic susceptibility (open) records and the global stack. The grey bar shows the duration of MIS 11...... 158

Figure 5.4. Lithostratigraphic columns for IODP sites U1345, U1343, and U1339, based on shipboard core descriptions (Takahashi et al., 2011) and smear slide analyses (Chapter 3). Column width varies according to the median grain size of bulk sediments (Caissie et al., 2016; Chapter 3). Colors signify varying amounts of diatoms relative to terrigenous grains in the sediment. Pale-green bars depict deglacial laminated sediments (Takahashi et al., 2011). Colored side bars show MIS 11 substages: deglaciation (turquoise), Peak MIS 11 (red), and Late MIS 11 (blue)...... 161

Figure 5.5. Median grain size of bulk sediments (solid line) and terrigenous sediments (dashed line), and the volume % terrigenous clay, at sites U1345 (black; Caissie et al., 2016; Chapter 3), U1343 (blue; Chapter 3), and U1339 (red; Chapter 3). Brown bars signify ash layers/accessories, and green bars show deglacial laminations (Takahashi et al., 2011). The grey panel shows the duration of MIS 11, and colored side bars show the following substages: deglaciation (turquoise), Peak MIS 11 (red), and Late MIS 11 (blue)...... 163

Figure 5.6. % terrigenous grains >150 µm (Chapter 3), sea ice concentrations (Nesterovich & Caissie, in review), and % sea ice diatoms (Caissie et al., 2016; Chapter 4), at sites U1345 (black), U1343 (blue), and U1339 (red). Note that we did not conduct full quantitative diatom counts at U1343. Dashed lines show the upper and lower limits of sea ice concentration. Vertical lines on the sea ice concentration plots mark the boundaries between open ocean (<15%), unconsolidated ice (10-40%), and consolidated ice (>40%). The grey panel xiv

shows the duration of MIS 11, and colored side bars show the following substages: deglaciation (turquoise), Peak MIS 11 (red), and Late MIS 11 (blue). .. 168

Figure 5.7. Maps of sea ice concentration at sites U1345 (black), U1343 (blue), and U1339 (red), for key time slices during the MIS 12-10 cycle. Concentrations are derived from the sea ice proxy (Nesterovich & Caissie, in review). The turquoise numbers show sea ice concentration values as a percentage. The white dashed line represents the unconsolidated ice margin, and the turquoise dashed line shows the consolidated ice margin. Note that ice margin positions are only an approximation. Grey bathymetric shading is based on modern-day bathymetry, and changes value at -50 m, -250 m, -1000 m, and -2000 m...... 170

Figure 5.8. Absolute diatom abundances (diatoms per gram of sediment), and relative percent abundances of Chaetoceros RS (high productivity/upwelling indicator), at sites U1345 (black; Caissie et al., 2016), U1343 (blue), and U1339 (red; Chapter 4). Also shown is weight % opal (Kanematsu et al., 2013) for U1345 (black), U1343 (blue), and U1341 (purple). Global eustatic sea level (navy; Rohling et al., 2010), δ15N upwelling index from Site U1343 (orange; Worne et al., 2019), and δ13C proxy for the strength of NPIW (grey; Knudson & Ravelo, 2015) are plotted for reference. The dashed vertical line shows the sill depth of Bering Strait (-50 m). The brown box shows the contourite at sites U1345 and U1343 (408-400 ka). Green bars depict deglacial laminations. The grey panel shows the duration of MIS 11, and colored side bars denote the following substages: deglaciation (turquoise), Peak MIS 11 (red), and Late MIS 11 (blue)...... 174

Figure A.1. Magnetic susceptibility records for sites U1339 and U1345, plotted against composite core depth (m)...... 193

Figure A.2. Relative % (area plots) and absolute (line plots) abundances of diatom taxa that make up more than 10% of any assemblage. Taxa are grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); North Pacific indicator (orange); summer bloom (light green); other (grey); and Chaetoceros RS (dark green). The final line plot shows the total number of diatom valves per gram of sediment. Note the different scale for absolute abundances of Chaetoceros RS and all diatom valves. The grey panel indicates the duration of MIS 11 (424-374 ka) and the green bar represents laminated sediments...... 196

Figure A.3. Relative percent abundances of diatoms grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); xv

North Pacific indicator (orange); high productivity (yellow); and Chaetoceros RS (dark green). Also shown is sea ice concentration (including upper and lower limits), derived from a quantitative, diatom-based sea ice proxy (Nesterovich and Caissie, in review). The grey panel shows the duration of MIS 11 (424-374 ka) and the green bar represents laminated sediments...... 197

Figure B.1. ANOVA boxplot ...... 199

Figure C.1. Ash layers that may provide a means of correlation between age models for the three sites...... 200 xvi

LIST OF TABLES

Page

Table 3.1. Size scale adopted in the GRADISTAT program, compared with those previously used by Udden (1914), Wentworth (1922) and Friedman and Sanders (1978). From Blott & Pye, 2001...... 50

Table 3.2. Statistical formulae used in the calculation of grain size parameters, and suggested descriptive terminology, using geometric (modified) Folk & Ward (1957) graphical measures. (P represents grain diameter, in metric units (µm/mm), at the cumulative percentile value of x). From Blott & Pye, 2001...... 51

Table 3.3. Grain size averages for bulk (top) and terrigenous (bottom) sediments from IODP Sites U1345, U1343, and U1339...... 56

Table 4.1. Age-depth tie points used in the age model for Site U1339...... 97

Table 4.2. Depths and ages of major climate intervals referred to in the text...... 98

Table 4.3. Subset of Bering Sea diatom species observed in this study, grouped by environmental niche (modified from Caissie et al., 2016)...... 106

Table 5.1. Average median grain size of bulk and terrigenous sediments from IODP sites U1345 (Caissie et al., 2016; Chapter 3), U1343, and U1339 (Chapter 3)...... 162

Table A.1. List of diatom taxa identified in this study, including pennate and centric diatoms identified to genus or species level, and morphogenera of Chaetoceros resting spores (RS)...... 194

xvii

ACKNOWLEDGMENTS

I wish to extend my sincere gratitude and appreciation to all those who have supported me throughout my PhD program. First and foremost, my deepest thanks go to my advisor, Beth

Caissie, for introducing me to this exciting field of research. Throughout my time at ISU, she has been a constant source of encouragement, and a truly inspiring mentor. I would also like to thank my committee members, Bill Gutowski, Chris Harding, Betsy Swanner, and Al

Wanamaker, for their helpful feedback on this work.

I am grateful that I was presented with so many opportunities to teach whilst at ISU.

Special thanks go to Cinzia Cervato, for mentoring me as a teaching assistant, giving me free rein to teach Geology 106, and inspiring in me a passion for STEM teaching and outreach.

I would like to thank the staff in the Geology Department (Suzanne Ankerstjerne, Joan

Dodd, DeAnn Frisk, Mark Mathison, and Alison Whale) for their technical and administrative support over the past six years. I would also like to thank my fellow graduate students in the

Geology and Environmental Science Programs, for helping to make my graduate experience so memorable. Above all, a big thank you to Sarah Krueger, for making me so welcome during my first few weeks at ISU. Our time as graduate students may have only overlapped by one semester, but made us lifelong friends. I am grateful to Thomas Harbour, Anna Nesterovich, and Derrick Vaughn, my former colleagues in the Marine Sediments Lab, for their support and friendship, and also to honorary lab member Hannah Carroll, for all her assistance with statistical analyses. In addition, I am indebted to my undergraduate research assistants, including Chris Beck, Courtney Schill, and Kate Staebell, who spent countless hours helping to xviii run grain size samples and count diatom slides. Without their help, I might well still be in the lab!

This work was supported by several lines of funding. I gratefully acknowledge funding from the Integrated Ocean Drilling Program, the Geological Society of America, and the

National Science Foundation, as well as departmental awards including the Bruce Bowen

Fellowship, the Morehouse Graduate Fellowship, and startup funds provided to Dr. Beth

Caissie.

I would never have completed this work, were it not for the support of my friends and family back home, and my new friends and family in the USA. At the risk of this sounding like an

Oscar Acceptance Speech, I am especially grateful to my parents, my sisters, and my lifelong friend, Francesca, for supporting my goals, encouraging my success, and always believing in me.

I would also like to thank Tessie Cat, for being the sweetest companion through long hours of writing. Special thanks go to the Boone Area Humane Society, for providing me with foster kittens (read: emotional support kittens), without which, I would probably have graduated years earlier! Above all, I am indebted to my future husband, John Hoyt, for his love, his constant support, and for keeping me (somewhat) sane over the past few months. I am forever grateful. xix

ABSTRACT

The ongoing decline in Arctic sea ice extent has prompted concerns about the fate of sea ice in the future, and the stability of sea ice-dependent ecosystems. By studying the natural variability of sea ice cover during past warm intervals, we can better understand how sea ice might respond to future warming. Sediments from Integrated Ocean Drilling Program (IODP) sites in the Bering Sea provide a record of paleo sea ice extent and productivity during Marine

Isotope Stage (MIS) 11 (424-374 ka), an interval of prolonged warming that represents a close analog to the Holocene with respect to orbital parameters.

The purpose of this dissertation is twofold; firstly, to examine the effectiveness of sediment grain size as a proxy for the sedimentary history of the Bering Sea, and secondly, to reconstruct variations in sea ice extent and productivity in the Bering Sea during the MIS 12-10 cycle. We show that, with some caveats, grain size parameters, including the volume percent of grains in various size fractions, as well as the statistical measures of mean grain size, sorting, and skewness, can be used to infer past changes in sediment transport, ice rafting and productivity at three core sites in the Bering Sea.

We present a new, high resolution, multi-proxy record of MIS 11 from IODP Site

U1339 (Umnak Plateau, southeastern Bering Sea), based on sediment grain size, diatom assemblages, stable isotopes, and a new diatom-based proxy for sea ice concentration. The record shows that sea ice was more extensive during MIS 11, in comparison to the Holocene, and advanced over the plateau during an interval of high sea level and high global temperatures. As sea ice declined during deglaciation, there was a spike in productivity, but sea ice and productivity trends are not correlated during MIS 11. We then compare the Umnak xx

Plateau record to new and published records from the Bering slope (sites U1345 and U1343).

Sediment grain size, diatom assemblages, and the diatom proxy show that sea ice was present at the slope sites during MIS 11, but there is evidence for considerable regional variability in the timing of ice advance and retreat between the Umnak Plateau and the Bering slope. This variability may be explained by east-west differences in sea level pressure and wind direction, which are controlled by shifts in the geographic position of the Aleutian Low.

CHAPTER 1. INTRODUCTION

Motivation

Climate change has occurred throughout Earth’s history, but the current warming trend is of particular concern, because much of the observed change is the result of human activity, and is progressing at an alarming rate. According to the Intergovernmental Panel on Climate

Change (IPCC) Fifth Assessment (AR5) Synthesis Report, ‘scientific evidence for warming of the climate system is unequivocal, and since the 1950’s, many of these changes are unprecedented over decades to millennia’ (IPCC, 2014). Documented impacts of global climate change include reduced snow cover, glacier and sea ice retreat, rising sea levels, ocean acidification, and an increase in the severity and frequency of natural disasters such as floods, droughts, and wildfires (e.g., Bates et al., 2012; Banholzer et al., 2014; Chu et al., 2016; Roe et al., 2017; Sweet et al., 2017; Boberg et al., 2018; Nerem et al., 2018; Williams et al., 2019).

Climate change is not uniform across the planet. In fact, the Arctic region has warmed at more than twice the average global rate since the mid-1990s (Overland et al., 2019), due to positive feedbacks such as the ice-albedo effect. This phenomenon, known as Arctic amplification, is intrinsically coupled with the extent of Arctic sea ice (Screen & Simmonds,

2010; Serreze et al., 2011; Osborne et al., 2017). The ongoing decline in Arctic sea ice cover has prompted concerns about the uncertain fate of sea ice and made improving our knowledge of sea ice behavior in response to climate change a pressing task. However, considering that direct measurements of sea ice are primarily limited to the past ~40 years of satellite measurements, little is known about the natural variability of sea ice over longer (multi-decadal to multi- millennial) time scales (de Vernal et al., 2013). Without long instrumental records to validate 2 climate models, it is difficult to predict the long-term response of sea ice to a warming climate.

The task is further complicated by the fact that modern climate is influenced by both natural and anthropogenic forcings, which makes it hard to separate the effects of human activity from the internal drivers of Arctic sea ice loss (Meehl et al., 2004; Notz & Marotzke, 2012; Ding et al.,

2017).

We can overcome some of these challenges by using paleoclimate archives (e.g., ice cores, sediment cores, corals, tree rings, speleothems) to build a more comprehensive record of past climate. Unlike many instrumental measurements, proxy-based records extend beyond the pre-industrial period and can therefore be used to gain perspective on natural climate variability over time. This information can help us to understand the role of sea ice in the climate system, place modern sea ice decline into a longer-term context, and better predict future changes in sea ice (de Vernal et al., 2013).

The primary goal of this research is to document the past spatial and temporal variability of Arctic sea ice, and the associated ecological changes that take place as sea ice advances or retreats. We achieve this goal through a multi-proxy examination of three sediment cores from across the Bering Sea during a past interval of climate warming (Marine

Isotope Stage (MIS) 11, 424-374 ka). The sub-arctic Bering Sea is highly sensitive to the effects of climate change, which makes the Bering Sea, its sea ice, and the ecosystem it supports, very important to understanding how global climate change may affect the future of Arctic sea ice.

Going forward, the records produced as part of this work will be used to evaluate past rates of change in sea ice and primary productivity, and ultimately provide a baseline against which current and future change can be measured. 3

The Role of Sea Ice in the Global Climate System

Although sea ice forms mainly in the polar regions, it plays a critical role in regulating

Earth’s climate at both global and regional scales, through positive and negative feedback mechanisms (Goosse et al., 2018). Sea ice exerts a cooling influence on global climate, partly because it acts as a barrier to heat and moisture exchange, effectively insulating the cold atmosphere from the comparatively warm ocean, and also due to its high albedo (Rahmstorf,

2002; Flanner et al., 2011; Thackeray & Hall, 2019). Bare sea ice has an albedo between 0.5 and

0.7, which means that it reflects ~50-70% of incoming solar radiation, in contrast to open ocean, which only reflects ~6% (NSIDC, 2020). However, as sea ice retreats and thins, surface albedo is reduced, which results in increased absorption of solar radiation, amplified surface warming, and further sea ice melt. This positive feedback loop is known as the sea ice-albedo feedback and is a key factor in Arctic climate change (Thackeray & Hall, 2019).

Sea ice also plays an important role in regulating global ocean circulation. As sea ice forms, the process of brine exclusion creates waters that are colder, saltier, and therefore more dense than the surrounding ocean. Differences in temperature, salinity and density drive deep water formation, and contribute to the development of thermohaline circulation (Rudels &

Quadfasel, 1991; Rahmstorf, 2002; Abernathey et al., 2016). In addition, increased freshwater input from melting ice may lead to reduced surface salinity and/or an increase in water column stratification, both of which have the potential to slow down or disrupt ocean circulation. A number of studies (e.g., Abernathey et al., 2016; Jensen et al., 2016; Sévellec et al., 2017; Sun et al., 2018; Liu et al., 2019) have invoked both the thermal and the freshwater effects of sea ice decline as factors in weakening of the Atlantic meridional overturning circulation (AMOC), 4 which has slowed by ~15% since the mid-twentieth century (Rahmstorf et al., 2015; Caesar et al., 2018). Moreover, numerical modelling experiments (Condron et al., 2020) have shown that the export and subsequent melting of thick, paleocrystic sea ice during the last deglaciation could have freshened the Nordic Seas enough to weaken AMOC, triggering episodes of abrupt climate cooling, such as the Younger Dryas (Bradley & England, 2008).

Sea ice is important not only within the climate system; it also supports one of the world’s largest ecosystems. Sea ice itself provides a critical habitat for large communities of ice algae (Arrigo, 2017), whilst seasonal melting of the ice fuels phytoplankton blooms (Perrette et al., 2011). These phytoplankton form the basis of a complex food web that includes primary, secondary, tertiary, and even quaternary consumers, from zooplankton to polar bears. In addition, sea ice cover provides a safe refuge for resting, feeding, and reproduction for a number of marine mammals (walrus, seals, polar bears etc.), along with many seabird species

(Arrigo, 2014). Consequently, changes in sea ice can have a serious impact on ecosystem dynamics (Hunter et al., 2010; Post, 2017; Duffy-Anderson et al., 2019).

Effects of Arctic Sea Ice Decline

The extent of Arctic sea ice has declined dramatically over the course of the modern satellite record (1979-present). Summer sea ice is currently declining at a rate of 12.9% per decade, the 13 lowest summer sea ice extents recorded in the Arctic have all occurred in the past 13 years (2007-2019), and the extent of thick multi-year ice is also declining rapidly.

The proportion of old (> 4 years) Arctic sea ice declined from 33% to 1.2% between 1985 and

2019, whilst the relative proportion of first-year sea ice increased from approximately 35-50% 5 to 70% (Perovich et al., 2019). The IPCC AR5 Synthesis Report states that summer Arctic sea ice loss since 1979 is unprecedented in 150 years based on historical reconstructions (Walsh et al.,

2017), and more than 1,000 years based on paleoclimate evidence (Polyak et al., 2010; Kinnard et al., 2011; Halfar et al., 2013; IPCC, 2014). This decline is strongly linked to increasing surface temperatures, but is also related to the ice-albedo feedback, changes in wind strength, and increased cyclones (Comiso, 2012).

Climate models indicate that the Arctic Ocean may be ice-free in the summer within the next few decades (e.g., Holland et al., 2006; Stroeve et al., 2012; Overland & Wang, 2013;

Stroeve & Notz., 2018). As sea ice retreats in the Arctic, positive feedbacks amplify climate change, with far-reaching effects on global ocean circulation, global carbon cycling, and the atmospheric circulation patterns that link Arctic sea ice to mid-latitude weather (Niebauer et al., 1999; Francis & Vavrus, 2012; Mann et al., 2017). In addition, the continued decline of Arctic sea ice has major implications for the health of marine ecosystems (Post et al., 2013; Arrigo,

2014). Sea ice loss now poses a serious threat to indigenous Arctic mammals, such as seals, walrus, and polar bears (Hunter et al., 2010; Post et al., 2013). Sea ice loss also presents a challenge to the photosynthetic organisms (phytoplankton) that form the basis of marine food webs. Primary productivity in the Arctic is currently undergoing a major shift in response to dwindling sea ice cover, particularly with respect to the extent and timing of the highly productive phytoplankton spring bloom. For example, there is mounting evidence for an overall increase in net primary production (e.g., Arrigo et al., 2008; Kahru et al., 2016; Hill et al., 2018;

Renaut et al., 2018; Frey et al., 2019); satellite measurements of the algal pigment chlorophyll-a show that primary productivity increased in all regions of the Arctic during the period 2003- 6

2019 (Frey et al., 2019). In addition, the spring bloom is expanding northward at a rate of 1 degree of latitude per decade (Renaut et al., 2018), and the length of the growth season has increased by an average of 17 days due to earlier ice retreat and/or later freeze up (Kahru et al.,

2016). Changes in the timing of the spring bloom may cause a mismatch between primary

(phytoplankton) and secondary (zooplankton) producers, with cascading effects on higher trophic levels (Arrigo et al., 2008; Leu et al., 2011; Kahru et al., 2016; Post, 2017; Duffy-

Anderson et al., 2019; Tedesco et al., 2019). As sea ice retreats and thins, competition for light and nutrients between under-ice blooms and the spring bloom may further impact the dynamics of the Arctic ecosystem (Renaut et al., 2018).

Despite recent upwards trends in primary productivity, the response of phytoplankton to sea ice retreat has been both seasonally and spatially variable across the Arctic region (Frey et al., 2019; Duffy-Anderson et al., 2019). The long-term relationship between sea ice and phytoplankton phenology is still not well understood (Hill et al., 2018; Tedesco et al., 2019), and it remains unclear how continued warming and sea ice loss will impact primary producers in the coming decades.

Dissertation Structure

This dissertation is organized into six chapters. Chapter 1 (this chapter) outlines the motivation for the study. Chapter 2 provides background information on sea ice and primary productivity in the Bering Sea, both historic and modern. The following three chapters are written as manuscripts, intended for publication in the near future. 7

Chapter 3 examines the effectiveness and limitations of sediment grain size as a proxy for sediment transport, sea ice rafting, current strength, and paleoproductivity in Bering Sea sediments. Chapter 4 presents a record of changes in sea ice extent and primary productivity throughout MIS 11 at IODP Site U1339 (Umnak Plateau, Bering Sea), using a multi-proxy approach that includes sediment grain size measurements, diatom assemblage counts, and stable isotope analyses. Chapter 5 compares new and published records of sea ice concentration and productivity at three IODP core sites (U1345, U1343, U1339) in the Bering

Sea during the MIS 12-10 cycle, in order to investigate regional variability in sea ice regimes across the Bering Sea. Finally, Chapter 6 summarizes the conclusions drawn from each of the previous chapters, and outlines directions for future work.

References

Abernathey, R., Cerovecki, I., Holland, P., Newsom, E., Mazloff, M., & Talley, L.D., 2016. Water- mass transformation by sea ice in the upper branch of the Southern Ocean overturning. Nat. Geosci., 9, 596-601. https://doi.org/10.1038/ngeo2749

Arrigo, K.R., 2014. Sea Ice Ecosystems. Annu. Rev. Mar. Sci., 6 (1), 439-467. https://doi.org/10.1146/annurev-marine-010213-135103

Arrigo, K.R., 2017. Sea ice as a habitat for primary producers. In: Thomas, D.N. (Ed.), Sea Ice (Third Edition). Wiley-Blackwell, Hokboken, NJ, pp. 352-369.

Arrigo, K.R., van Dijken, G., & Pabi, S., 2008. Impact of a shrinking Arctic ice cover on marine primary production. Geophys. Res. Lett., 35 (19), L19603. https://doi.org/10.1029/2008GL035028

Banholzer, S., Kossin, J., & Donner, S., 2014. The Impact of Climate Change on Natural Disasters. In: Singh, A. & Zommers, Z. (Eds.), Reducing Disaster: Early Warning Systems For Climate Change. Springer, Dordrecht, pp. 21-49.

Bates, N.R., Best, M.H.P., Neely, K., Garley, R., Dickson, A.G., & Johnson, R.J., 2012. Detecting anthropogenic carbon dioxide uptake and ocean acidification in the North Atlantic Ocean. Biogeosciences, 9, 2509-2522. https://doi.org/10.5194/bg-9-2509-2012

Boberg, F., Langen, P.L., Mottram, R.H., Christensen, J.H., & Olesen, M., 2018. 21st-century climate change around Kangerlussuaq, west Greenland: From the ice sheet to the shores of Davis Strait. Arct. Antarct. Alp. Res., 50 (1), S100006. https://doi.org/10.1080/15230430.2017.1420862

Bradley, R.S. & England, J.H., 2008. The Younger Dryas and the Sea of Ancient Ice. Quat. Res., 70, 1-10. https://doi.org/10.1016/j.yqres.2008.03.002

Caesar, L., Rahmstorf, S., Robinson, A., Feulner, G., & Saba, V., 2018. Observed fingerprint of a weakening Atlantic Ocean overturning circulation. Nature, 556, 191-196. https://doi.org/10.1038/s41586-018-0006-5

Chu, S.N., Wang, Z.A., Doney, S.C., Lawson, G.L., & Hoering, K.A., 2016. Changes in anthropogenic carbon storage in the Northeast Pacific in the last decade. J. Geophys. Res. Oceans, 121 (7), 4618-4632. https://doi.org/10.1002/2016JC011775

Comiso, J.C., 2012. Large Decadal Decline of the Arctic Multiyear Ice Cover. J. Clim., 25, 1176- 1193. https://doi.org/10.1175/JCLI-D-11-00113.1

Condron, A., Joyce, A.J., & Bradley, R.S., 2020. Arctic sea ice as a driver of deglacial climate. Geology, 48, https://doi.org/10.1130/G47016.1 de Vernal, A., Gersonde, R., Goosse, H., Seidenkrantz, M-S., & Wolff, E.W., 2013. Sea ice in the paleoclimate system: the challenge of reconstructing sea ice from proxies - an introduction. Quat. Sci. Rev., 79, 1-8. https://dx.doi.org/10.1016/j.quascirev.2013.08.009

Ding, Q., Schweiger, A., L’Heureux, M., Battisti, D.S., Po-Chedley, S., Johnson, N.S., Blanchard- Wrigglesworth, E., Harnos, K., Zang, Q., Eastman, R., & Steig, E.J., 2017. Influence of high-latitude atmospheric circulation changes on summertime Arctic sea ice. Nat. Clim. Change, 7, 289-295. https://doi.org/10.1038/nclimate3241

Duffy-Anderson, J.T., Stabeno, P., Andrews III, A.G., Cieciel, K., Deary, A., Farley, E., Fugate, C., Harpold, C., Heintz, R., Kimmel, D., Kuletz, K., Lamb, J., Paquin, M., Porter, S., Rogers, L., Spear, A., & Yasumiishi, E., 2019. Responses of the northern Bering Sea and southeastern Bering Sea pelagic ecosystems following record-breaking low winter sea ice. Geophys. Res. Lett., 46, 9833-9842. https://doi.org/ 10.1029/2019GL083396

Flanner, M.G., Shell, K.M., Barlage, M., Perovich, D.K., & Tschudi, M.A., 2011. Radiative forcing and albedo feedback from the Northern Hemisphere cryosphere between 1979 and 2008. Nat. Geosci., 4, 151-155. https://doi.org/10.1038/NGEO1062

Francis, J.A. & Vavrus, S.J., 2012. Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett., 39 (6), L06801. https://doi.org/10.1029/2012GL051000

Frey, K.E., Comiso, J.C., Cooper, L.W., Grebmeier, J.M., & Stock, L.V., 2019. Arctic Ocean Primary Productivity: The Response of Marine Algae to Climate Warming and Sea Ice Decline. In: Richter-Menge, J., Druckenmiller, M.L., & Jeffries, M. (Eds.), Arctic Report Card 2019. http://www.arctic.noaa.gov/Report-Card

Goosse, H., Kay, J.E., Armour, K.C., Bodas-Salcedo, A., Chepfer, H., Docquier, D., Jonko, A., Kushner, P.J., Lecomte, O., Massonnet, F., Park, H.-S., Pithan, F., Svensson, G., & Vancoppenolle, M., 2018. Quantifying climate feedbacks in polar regions. Nat. Commun., 9, 1919. https://doi.org/10.1038/s41467-018-04173-0

Halfar, J., Adey, W.H., Kronz, A., Hetzinger, S., Edinger, E., & Fitzhugh, W.W., 2013. Arctic sea-ice decline archived by multicentury annual-resolution record from crustose coralline algal proxy. Proc. Natl. Acad. Sci. USA, 110 (49), 19,737-19,741. https://doi.org/10.1073/pnas.1313775110

Hill, V., Ardyna, M., Lee, S.H., & Varela, D.E., 2018. Decadal trends in phytoplankton production in the Pacific Arctic Region from 1950 to 2012. Deep Sea Res. Part II: Top. Stud. Oceanogr., 152, 82-94. https://doi.org/10.1016/j.dsr2.2016.12.015

Holland, M.M., Bitz, C.M. & Tremblay, B., 2006. Future abrupt reductions in the summer Arctic sea ice. Geophys. Res. Lett., 33 (23), 1-5. https://doi.org/10.1029/2006GL028024

Hunter, C.M., Caswell, H., Runge, M.C., Regehr, E.V., Amstrup, S.C., & Stirling, I., 2010. Climate change threatens polar bear populations: a stochastic demographic analysis. Ecology, 91 (10), 2883-2897. https://doi.org/10.1890/09-1641.1

IPCC, 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (Eds.)]. IPCC, Geneva, Switzerland, 151 pp.

Jensen, M.F., Nilsson, J., & Nisancioglu, K.H., 2016. The interaction between sea ice and salinity- dominated ocean circulation: implications for halocline stability and rapid changes of sea ice cover. Clim. Dyn., 47 (9-10), 3301-3317. https://doi.org/10.1007/s00382-016- 3027-5

Kahru, M., Lee, Z., Mitchell, B.G., & Nevison, C.D., 2016. Effects of sea ice cover on satellite- detected primary production in the Arctic Ocean. Biol. Lett., 12 (11), 20160223. https://doi.org/10.1098/rsbl.2016.0223

Kinnard, C., Zdanowicz, C.M., Fisher, D.A., Isaksson, E., de Vernal, A., & Thompson, L.G., 2011. Reconstructed changes in Arctic sea ice over the past 1,450 years. Nature, 479, 509-512. https://doi.org/10.1038/nature10581

Leu, E., Søreide, J.E., Hessen, D.O., Falk-Petersen, S., & Berge, J., 2011. Consequences of changing sea-ice cover for primary and secondary producers in the European Arctic shelf seas: Timing, quantity, and quality. Prog. Oceanogr., 90 (1-4), 18-32. https://doi.org/10.1016/j.pocean.2011.02.004

Liu, W., Fedorox, A., & Sévellec, F., 2019. The Mechanisms of the Atlantic Meridional Overturning Circulation Slowdown Induced by Arctic Sea Ice Decline. J. Clim., 32 (4), 977- 996. https://doi.org/10.1175/JCLI-D-18-0231.1

Mann, M.E., Rahmstorf, S., Kornhuber, K., Steinman, B.A., Miller, S.K., & Coumou, D., 2017. Influence of Anthropogenic Climate Change on Planetary Wave Resonance and Extreme Weather Events. Sci. Rep., 7, 45242. https://doi.org/10.1038/srep45242

Meehl, G., Washington, W., Ammann, C., Arblaster, J., Wigley, T.M.L., & Tebaldi, C., 2004. Combinations of Natural and Anthropogenic Forcings in Twentieth-Century Climate. J. Clim., 17, 3721-3727. https://doi.org/10.1175/1520- 0442(2004)017<3721:CONAAF>2.0.CO;2

Nerem, R.S., Beckley, B.D., Fasullo, J.T., Hamlington, B.D., Masters, D. & Mitchum, G.T., 2018. Climate-change-driven accelerated sea-level rise detected in the altimeter era. Proc. Natl. Acad. Sci. USA, 115 (9), 2022-2025. https://doi.org/10.1073/pnas.1717312115

Niebauer, H.J., Bond, N.A., Yakunin, L.P., & Plotnikov, V.V., 1999. An Update on the Climatology and Sea Ice of the Bering Sea. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 29-60.

NSIDC, 2020. All About Sea Ice. National Snow and Ice Data Center, Boulder, CO. https://nsidc.org/cryosphere/seaice/index.html [accessed 12 February 2020].

Notz, D. & Marotzke, J., 2012. Observations reveal external driver for Arctic sea-ice retreat. Geophys. Res. Lett., 39 (8), L08502. https://doi.org/10.1029/2012GL051094

Osborne, J.M., Screen, J.A., & Collins, M., 2017. Ocean-Atmosphere State Dependence of the Atmospheric Response to Arctic Sea Ice Loss. J. Clim., 30 (5), 1537-1552. https://doi.org/10.1175/JCLI-D-16-0531.1

Overland, J.E. & Wang, M., 2013. When will the summer Arctic be nearly sea ice free? Geophys. Res. Lett., 40 (10), 2097-2101. https://doi.org/10.1002/grl.50316

Overland, J.E., Hanna, E., Hanssen-Bauer, I., Kim, S.-J., Walsh, J.E., Wang, M., Bhatt, U.S., Thoman, R.L., & Ballinger, T.J., 2019. Surface Air Temperature. In: Richter-Menge, J., Druckenmiller, M.L., & Jeffries, M. (Eds.), Arctic Report Card 2019. http://www.arctic.noaa.gov/Report-Card

Perovich, D., Meier, W., Tschudi, M., Farrell, S., Hendricks, S., Gerland, S., Kaleschke, L., Ricker, R., Tian-Kunze, X., Webster, M., & Wood, K., 2019. Sea Ice. In: Richter-Menge, J., Druckenmiller, M.L., & Jeffries, M. (Eds.), Arctic Report Card 2019. http://www.arctic.noaa.gov/Report-Card

Perrette, M., Yool, A., Quartly, G. D., & Popova, E. E., 2011. Near-ubiquity of ice-edge blooms in the Arctic. Biogeosciences, 8 (2), 515-524. https://doi.org/10.5194/bg-8-515-2011

Polyak, L., Alley, R., Andrews, J., Brigham-Grette, J., Cronin, T., Darby, D., Dyke, A., Fitzpatrick, J., Funder, S., Holland, M., Jennings, A., Miller, G., O'Regan, M., Savelle, J., Serreze, M., St. John, K., White, J., & Wolff, E., 2010. History of sea ice in the Arctic. Quat. Sci. Rev., 29, 1757-1778. https://doi.org/10.1016/j.quascirev.2010.02.010

Post, E., 2017. Implications of earlier sea ice melt for phenological cascades in Arctic marine food webs. Food Webs, 13, 60-66. https://doi.org/10.1016/j.fooweb.2016.11.002

Post, E., Bhatt, U.S., Bitz, C.M., Brodie, J.F., Fulton, T.L., Hebblewhite, M., Kerby, J., Kutz, S.J., Stirling, I., & Walker, D.A., 2013. Ecological Consequences of Sea Ice Decline. Science, 341 (6145), 519-524. https://doi.org/10.1126/science.1235225

Rahmstorf, S., 2002. Ocean circulation and climate during the past 120,000 years. Nature, 419 (6903), 207-214. https://doi.org/10.1038/nature01090

Rahmstorf, S., Box, J.E., Feulner, G., Mann, M.E., Robinson, A., Rutherford, S., & Schaffernicht, E.J., 2015. Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation. Nat. Clim. Change, 5, 475-480. https://doi.org/10.1038/nclimate2554

Renaut, S., Devred, E., & Babin, M., 2018. Northward Expansion and Intensification of Phytoplankton Growth During the Early Ice-Free Season in Arctic. Geophys. Res. Lett., 45 (19), 10590-10598. https://doi.org/10.1029/2018GL078995

Roe, G.H., Baker, M.B., & Herla, F., 2017. Centennial glacier retreat as categorical evidence of regional climate change. Nat. Geosci., 10, 95-99. https://doi.org/10.1038/ngeo2863

Rudels, B. & Quadfasel, D., 1991. Convection and deep water formation in the Arctic Ocean— Greenland Sea System. J. Mar. Syst., 2 (3-4), 435-450. https://doi.org/10.1016/0924-7963(91)90045-V

Screen, J.A. & Simmonds, I., 2010. The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464 (7293), 1334-1337. https://doi.org/10.1038/nature09051

Serreze, M.C. & Barry, R.G., 2011. Processes and impacts of Arctic amplification: A research synthesis. Glob. Planet. Change, 77 (1-2), 85-96. https://doi.org/10.1016/j.gloplacha.2011.03.004

Sévellec, F., Fedorov, A., & Liu, W., 2017. Arctic sea-ice decline weakens the Atlantic Meridional Overturning Circulation. Nat. Clim. Change, 7, 604-610. https://doi.org/10.1038/nclimate3353

Stroeve, J.C., Kattsov, V., Barrett, A., Serreze, M., Pavlova, T., Holland, M., & Meier, W.N., 2012. Trends in Arctic sea ice extent from CMIP5, CMIP3 and observations. Geophys. Res. Lett., 39 (16), L16502. https://doi.org/10.1029/2012GL052676

Stroeve, J. & Notz, D., 2018. Changing state of Arctic sea ice across all seasons. Environ. Res. Lett., 13 (10), 103001. https://doi.org/10.1088/1748-9326/aade56

Sun, L., Alexander, M., & Deser, C., 2018. Evolution of the Global Coupled Climate Response to Arctic Sea Ice Loss during 1990–2090 and Its Contribution to Climate Change. J. Clim., 31 (19), 7823-7843. https://doi.org/10.1175/JCLI-D-18-0134.1

Sweet, W.V., Kopp, R.E., Weaver, C.P., Obeysekera, T., Horton, R.M., Thieler, E.R., & Zervas, C., 2017. Global and Regional Sea Level Rise Scenarios for the United States. NOAA Tech. Rep. NOS CO-OPS 083. National Oceanic and Atmospheric Administration, National Ocean Service, Silver Spring, MD, 75 pp.

Tedesco, L., Vichi, M., & Scoccimarro, E., 2019. Sea-ice algal phenology in a warmer Arctic. Science Advances, 5 (5), eeav4830. https://doi.org/10.1126/sciadv.aav4830

Thackeray, C.W. & Hall, A., 2019. An emergent constraint on future Arctic sea-ice albedo feedback. Nat. Clim. Change, 9, 972-978. https://doi.org/10.1038/s41558-019-0619-1

Walsh, J.E., Fetterer, F., Stewart, J.S., & Chapman, W.L., 2017. A database for predicting Arctic sea ice variations back to 1850. Geographical Perspectives on the Arctic, special edition of Geogr. Rev., 107 (1), 89-107. https://doi.org/10.1111/j.1931-0846.2016.12195.x

Williams, A.P., Abatzoglou, J.T., Gershunov, A., Guzman-Morales, J., Bishop, D.A., Balch, J.K., & Lettenmaier, D.P., 2019. Observed Impacts of Anthropogenic Climate Change on Wildfire in California. Earth’s Future, 7 (8), 892-910. https://doi.org/10.1029/2019EF001210

CHAPTER 2. BACKGROUND AND LITERATURE REVIEW

Study Area

Physiography and Oceanography of the Bering Sea

The Bering Sea, located between Russia and Alaska (Fig. 2.1), is the northernmost marginal sea of the Pacific Ocean, and the third largest marginal sea in the world. It comprises a broad (>500 km), shallow continental shelf in the east, a narrow (<100 km) shelf to the west, a steep continental slope and a deep abyssal basin, dissected by several large submarine rises including Shirshov Ridge (SR), Bowers Ridge (BR), and Umnak Plateau (UP) (Fig. 2.1; Ben-

Avraham & Cooper, 1981; Stabeno et al., 1999).

The Bering Sea is divided from the North Pacific Ocean by the Aleutian Island Arc (Fig.

2.1), but water mass exchange with the Pacific occurs through passes in the Aleutian Islands, linking Bering Sea conditions to those of the Pacific (Stabeno et al., 1999). Nutrient-rich intermediate and bottom water from the North Pacific flows into the Bering Sea through these passes and is slightly modified by the mixing of relatively fresh, warm water with the small amounts of bottom water formed in the Bering Sea today (Warner & Roden, 1995). The oxygen and nutrient composition of Bering Sea subsurface water is further modulated by respiration of organic matter in the water column. The Bering Sea hosts a pronounced oxygen minimum zone

(OMZ), which is particularly intense at modern water depths of ~900 m (Cook et al., 2005), with dissolved oxygen concentrations as low as 15 µmol/kg (Roden, 2000). 15

UkP

Surface water masses are derived from inflow of the Alaska Coastal Current (ACC), which enters the Bering Sea through Unimak Pass (52 m deep) and other shallow eastern passes (Ladd et al., 2005), and from inflow of the Alaskan Stream. As the Alaskan Stream flows into the Bering Sea through the many passes in the Aleutian Arc, its forms the Aleutian North 16

Slope Current (ANSC). In turn, the northeastward flow of the ANSC is diverted northwestward as it nears the shallow continental shelf to form the Bering Slope Current (BSC), which flows along the continental slope and supplies critical nutrients to the shelf (Fig. 2.1; Reed & Stabeno,

1999; Stabeno et al., 1999; Stabeno et al., 2009). Cyclonic circulation dominates surface flow, so that much of the Pacific inflow is matched by outflow through the Kamchatka Strait and other passes, but a substantial amount of surface water (~0.8 Sverdrups (Sv)) is also transported northward through the shallow (-50 m) Bering Strait into the Arctic Ocean (Fig. 2.1; Roach et al.,

1995). As the sole gateway between the Bering Sea and the Arctic Ocean, inflow of Pacific water through the Bering Strait is an important source of heat, freshwater and nutrients to the

Arctic (Walsh et al., 1989; Stabeno et al., 1999; Woodgate et al., 2010). The closure of the

Bering Strait at times of lower sea level isolates the North Pacific from the North Atlantic, causing changes in global patterns of circulation, nutrient and salinity distributions (e.g.,

Okazaki et al., 2010; Hu et al., 2012, 2015; Knudson and Ravelo, 2015; Caissie et al., 2016;

Worne et al., 2019). By way of example, Knudson & Ravelo (2015) used a 1.2 Myr record of foraminiferal δ18O and δ13C from IODP Site U1342 as evidence for enhanced ventilation of

North Pacific Intermediate Water (NPIW) during extreme glacial conditions, when sea level was at least 50 m lower than at present and the Bering Strait was closed. Increased sea ice formation and brine rejection within the Bering Sea likely played a large role in NPIW ventilation during such periods (Worne et al., 2019).

Sea Ice Variability in the Bering Sea

The history of sea ice in the Bering Sea extends well into the Pleistocene, and in some cases back to the Pliocene (e.g., Kanematsu et al., 2013; Stroynowski et al., 2015; Kato et al., 17

2016; Onodera et al., 2016; Teraishi et al., 2016; Detlef et al., 2018). At Bowers Ridge (IODP Site

U1340) (Fig. 2.1), for example, siliceous microfossils record a pulse of sea ice during the

Pliocene M2 glacial event at ~3.3 Ma and periodic expansion of sea ice to Bowers Ridge beginning at around 2.9 Ma (Onodera et al., 2016). On shorter timescales, the sea ice margin appears to be relatively stable at the shelf-slope break, even during times of significant climate transition (Caissie et al., 2016), but on longer timescales ice extent has fluctuated considerably, expanding southward during glacial periods, retreating to its present limit (Cook et al., 2005;

Katsuki and Takahashi, 2005; Caissie et al., 2010) or even further north (Kaufman and Brigham-

Grette, 1993) during interglacials. Today, only the northern Bering Sea is seasonally ice-covered

— the southern limit of sea ice extent is about 57°N — but during the Last Glacial Maximum

(LGM), winter sea ice advanced to cover most regions of the Bering Sea. There are several lines of evidence for thick perennial sea ice at Site 51JPC in the Umnak Plateau region of the southeastern Bering Sea during the LGM (Sancetta et al., 1985; Caissie et al., 2010; Pelto et al.,

2018; Nesterovich, 2019), yet Katsuki & Takahashi (2005) only found evidence for seasonal ice in a second core (UMK3A), located slightly further southwest on the Umnak Plateau. In contrast, the south-central part of the Bering Sea remained largely ice free (Katsuki &

Takahashi, 2005; Matul, 2017; Méheust et al., 2018). These discrepant records point to significant regional variability in sea ice regimes across the Bering Sea.

Current conditions in the Bering Sea promote the formation of seasonal sea ice over the continental shelf (Fig. 2.1; Stabeno et al., 2019a). In the fall, as surface waters reach a temperature of -1.7°C (the freezing point for saltwater), ice begins to form in the northern

Bering Sea, and is pushed southward by prevailing north-northeasterly winter winds. Sea ice 18 typically peaks in late March, and melts back in the spring, reaching its minimum extent in

September (Niebauer et al., 1999; Perovich et al., 2019). This annual cycle of sea ice melt and freeze is highly variable over the course of the satellite record (1979-2020), alternating between colder winters with extensive sea ice and warmer winters with reduced sea ice (Fig.

2.2; Stabeno et al., 2012, 2019b; Stabeno & Bell, 2019). Interestingly, prior to 2018 there was no obvious systematic decrease in Bering Sea ice volume; unlike the progressive decline in sea ice extent observed for the Artic as a whole, the Bering Sea record appears more cyclic (Fig. 2.2;

Brown & Arrigo, 2012; Stabeno et al., 2019b). However, during the 2017-2108 season, winter sea ice cover in the Bering Sea reached record low levels, due to persistent warm southerly winds, and the second lowest extent on record was observed the following year (2018-2019)

(Fig. 2.2; Cornwall, 2019; Stabeno & Bell, 2019; Stabeno et al., 2019a, 2019b).

For the past few years, the ice pack in the Bering Sea has advanced later and retreated earlier, resulting in a shorter sea ice season (Wang et al., 2018; Stabeno et al., 2019b). Although sea ice decline is an expected response to climate warming, loss of this magnitude was not predicted to occur before 2030 (Wang & Overland, 2009), and there are concerns that these unexpected changes could indicate a major regime shift in the Bering Sea, with implications for ecosystem stability (Duffy-Anderson et al., 2019; Cornwall, 2019; Stabeno et al., 2019b).

Primary Productivity in the Bering Sea

The Bering Sea has long been recognized as one of the most biologically productive marine ecosystems in the world (e.g., Sambrotto et al., 1984; Loughlin et al., 1999; Sigler et al.,

2010; Brown and Arrigo, 2012, 2013; Stabeno et al., 2019a). Primary production in this region is dominated (both now, and in the past) by the contribution of siliceous microorganisms, primarily diatoms (Takahashi et al., 2000; Okazaki et al., 2005; Aiello & Ravelo, 2012; Ran et al.,

2013; Koizumi & Yamamoto, 2018), that thrive due to extremely high levels of silica in the

Bering Sea. In fact, deep waters of the Bering Sea contain the highest silicate concentrations in the world’s ocean, as much as 40% greater than in the deep North Pacific (Coachman et al.,

1999). As a result, the Bering Sea has been dubbed ‘the sea of silica’ (Tsunogai et al., 1979).

There are several possible explanations for the high silica levels, including: a) extremely high surface productivity, such that high concentrations of dissolved silica are produced at depth by the dissolution of large amounts of marine snow; b) gradual accumulation of dissolved silica over time, due to the long residence time of Bering Sea deep water; or c) high regeneration rates of silicic acid from bottom sediments (Calvert, 1974; Tsunogai et al., 1979; Coachman et al., 1999). 20

In the Bering Sea, the seasonal productivity cycle begins in spring, once light levels are sufficient to penetrate through the ice and induce photosynthesis in sea ice algae. As sea ice retreats, it leaves behind cold, stratified, nutrient-rich waters, which stimulate a massive bloom of phytoplankton (primarily diatoms) (Vancoppenolle et al., 2013). This bloom follows the retreating ice edge northward across the Bering Shelf, until surface nutrients become depleted

(Niebauer et al., 1990; Brown & Arrigo, 2013; Sigler et al., 2014). Typically, over half of the annual net primary production in the Bering Sea occurs within a ~3-month period between May and July; during this time, enough organic matter settles on the seafloor to sustain the entire shelf ecosystem throughout the year (Brown & Arrigo, 2013). Secondary phytoplankton blooms may occur in summer or fall, providing that episodic winds can break down stratification and mix nutrients up into the surface waters (Stabeno et al., 2010; Sigler et al., 2014). Surface sediments in the adjacent Chukchi Sea reflect a pattern of seasonal succession; open waters in summer are dominated by the diatom species Nitzschia longissima, which is replaced by

Proboscia species come autumn (Nesterovich, 2019). The autumn bloom ends once light becomes a limiting factor, and during winter, enhanced vertical mixing and nutrient recharge sets the stage for the next spring phytoplankton bloom (Brown & Arrigo, 2013).

Changes in sea ice extent, as well as the timing of ice retreat, are likely to greatly impact the magnitude and timing of primary production in the Bering Sea, with cascading effects on top predators (Brown and Arrigo, 2013). For example, following the record low sea ice extent in

2017/2018, the spring phytoplankton bloom in the northeastern Bering Sea was delayed until

June, was much smaller than in previous years, and consisted of different species; zooplankton had less fat than usual; small forage fish such as herring and caplen were scarce; and there 21 were a high number of seabird deaths reported (Duffy-Anderson et al., 2019). Models predict that future climate warming and associated sea ice decline will have a strong impact on primary production in the Bering Sea by the end of the 21st century, and that these changes will propagate through the food web, impacting the commercially valuable fisheries in the region

(Hunt et al., 2011; IPCC, 2014; Hermann et al., 2019).

Changes in primary productivity during the last deglaciation (Termination I) have been studied in great detail in the Bering Sea (e.g. Sancetta et al., 1985; Cook et al., 2005; Brunelle et al., 2007; Caissie et al., 2010; Pelto et al., 2018). During the deglacial transition, a widespread decrease in the oxygen concentration of mid-depth waters culminated in the deposition of laminated sediments across the North Pacific and its marginal seas (Cook et al., 2005; Brunelle et al., 2007; Addison et al., 2012; Schlung et al., 2013; Pelto et al., 2018). The exact causes of the dysoxia and subsequent formation of laminated sediments are still debated, but leading hypotheses include changes in the ventilation of intermediate waters (Zheng et al., 2000), increased export productivity (Kuehn et al., 2014), or a combination of the two (Cook et al.,

2005; Matul et al., 2016). Regardless of the cause, many factors indicate high productivity during these laminated intervals, including pulses of CaCO3, a high weight percentage of biogenic opal, and a high percentage of organic carbon (Cook et al., 2005; Okazaki et al., 2005;

Brunelle et al., 2007; Caissie et al., 2010, 2016; Pelto et al., 2018). Additionally, resting spores of

Hyalochaete Chaetoceros, a diatom indicative of upwelling or high productivity environments, dominate Bering Sea diatom assemblages during laminated intervals (Lopes et al., 2006; Caissie et al., 2010). 22

Despite regional differences, the Termination I records display several similarities to

Terminations II (130 ka) and V (424 ka), suggesting that there are certain characteristics of productivity that many, if not all, deglacial transitions have in common. For example, multiple lines of evidence show that primary productivity is lowest during glacials in all areas of the

Bering Sea. Productivity increases during deglaciation, but does not remain high once a warm period stabilizes (Katsuki and Takahashi, 2005; Brunelle et al., 2007; Caissie et al., 2010; Lam et al., 2013; Caissie et al., 2016).

Marine Isotope Stage 11

Marine Isotope Stage (MIS) 11 (~424-374 ka) is recognized as the longest and one of the warmest interglacials of the past 500 ka (Fig. 2.3; Hodell et al., 2000; Berger & Loutre, 2002;

Lisiecki and Raymo, 2005; Rohling et al., 2010), and is further characterized by the highest amplitude deglacial warming of the past 5 Ma (Droxler and Farrell, 2000; Droxler et al., 2003).

The transition from MIS 12 to 11 (Termination V) has been compared to the last deglaciation

(Dickson et al., 2009), and MIS 11 is considered a partial analogue for current and future warming (Droxler and Farrell, 2000; Berger & Loutre, 2002; Loutre & Berger, 2003; Masson-

Delmotte et al., 2006; Bowen, 2010), although it is important to recognize that the natural course of Holocene warming has been disrupted by anthropogenic activity (Ruddiman, 2007;

Palumbo et al., 2019; Cronin et al., 2019).

Various authors use different terms to define Marine Isotope Stage 11 and its substages

(e.g., Prokopenko et al., 2001; Tzedakis et al., 2001; Kandiano et al., 2012; Railsback et al., 2015;

Hrynowiecka et al., 2019). For clarity, we recognize MIS 11 as the warm period from 424-374 ka 23

(Lisiecki & Raymo, 2005), and MIS 11c (420-398 ka) as the interglacial climatic optimum

(Kandiano et al., 2012), with peak warmth centered around 405 ka, consistent with the timing of peak eustatic sea level (~410-400 ka) (Raymo & Mitrovica, 2012).

MIS 11 has been selected as a key time slice for several reasons. Firstly, it is the most recent interglacial period with orbital parameters similar to the Holocene (Fig. 2.3). Like today, the eccentricity of Earth’s orbit during MIS 11 was at a minimum, which reduced the amplitude of precessional change, whereas obliquity was high, leading to warmer summers and a corresponding reduction in ice volume (Berger & Loutre, 2002; Droxler et al., 2003; Loutre &

Berger, 2003). Accordingly, the strong interglacial signal of MIS 11 cannot be explained solely within the context of orbital forcing - the 'MIS 11 problem’ (Imbrie & Imbrie, 1980; Droxler &

Farrell, 2000; Berger & Wefer, 2003). However, Hodell et al. (2000) showed that the sustained interglacial warmth of MIS 11 could be partly explained by the fact that low eccentricity and low amplitude precession resulted in fewer cold substages during the interglacial. In addition, atmospheric CO2 concentrations during MIS 11 averaged 275 ppm for a prolonged period, a value which is comparable to pre-industrial levels (Fig 3; Petit et al., 1999; Siegenthaler et al.,

2005; Lüthi et al., 2008; Bereiter et al., 2015). These relatively high CO2 levels contributed to the extreme warmth recorded in the proxy record; MIS 11 would otherwise have been much cooler, due to relatively low (although stable) insolation values at 65°N (Fig. 2.3; Yin and Berger,

2011).

The spatial pattern and degree of warming during MIS 11 is highly complex. Some paleoclimate records indicate that temperatures were no higher than for any other interglacial of the past 500 ka (e.g., McManus et al., 1999, 2003; Bauch et al., 2000; Hodell et al., 2000). 24

Figure 2.3. Orbital cycles and global climate records of the past 500 ka. (a) Orbital eccentricity (Laskar et al., 2011); (b) obliquity (Laskar et al., 2004); (c) precession (Laskar et al., 2004); (d) northern hemisphere summer insolation at 65°N (Laskar et al., 2004); (e) benthic foraminiferal 18 δ O records (the LR04 stack, Lisiecki & Raymo, 2005); (f) compilation of atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015) and (g) relative sea level in meters above or below present (Rohling et al., 2010). Grey bars denote interglacial stages 1 through 13. 25

Conversely, the extreme and prolonged warmth observed in many other records from this interval has led to MIS 11 being described as a ‘super interglacial’ (Melles et al., 2012). Overall,

MIS 11 is characterized by warmer climatic conditions, especially at high latitudes (de Vernal &

Hillaire-Marcel, 2008; Melles et al., 2012). Sea level during MIS 11c was probably 6-13 m higher than at present (Raymo & Mitrovica, 2012; Dutton et al., 2015), although estimates range from

+20 m (Bowen, 1999; Olson & Hearty, 2009) to 0 m above present sea level (Bowen, 2010;

Rohling et al., 2010, 2014). A sustained sea level highstand of 6-13 m would have required major or even total deglaciation of Greenland and/or the West Antarctic Ice Sheet (Raymo &

Mitrovica, 2012). MIS 11 is further characterized by increased carbonate production at high latitudes, massive expansion of coral reefs, and strong thermohaline circulation (Droxler &

Farrell, 2000). During MIS 11c, Antarctica experienced temperatures 2°C warmer than today

(Jouzel et al., 2007), the Arctic Ocean was seasonally ice-free (Cronin et al., 2019), boreal forest extended across Greenland, which may have been largely ice-free (de Vernal & Hillaire-Marcel,

2008) and large lakes in Siberia were anomalously productive (Prokopenko et al., 2010; Lozhkin

& Anderson, 2013).

Marine Isotope Stage 11 in the Bering Sea

Despite the work done to characterize MIS 11 in the terrestrial realm (e.g., Droxler et al.,

2003; Ashton et al., 2008; Candy, 2009; Dabkowski et al., 2012; Candy et al., 2014; Kleinen et al., 2014; Shi et al., 2016; Tye et al., 2016), as well as the North Atlantic and Nordic Seas (Bauch et al., 2000; Kandiano & Bauch, 2007; Dickson et al., 2009; Poli et al., 2010; Voelker et al., 2010;

Rodrigues et al., 2011; Bauch, 2012; Kandiano et al., 2012, 2017; Candy & McClymont, 2013; 26

Milker et al., 2013), less is known about MIS 11 in the North Pacific region, and especially in the

Bering Sea. In fact, this project is one of the first high resolution studies of MIS 11 in the Bering

Sea. It builds on work by Caissie et al. (2016), in which the authors examine changes in sea ice extent from MIS 12-10 at IODP Site U1345 and find that sea ice diatoms were present (~30% of the assemblage) throughout the entirety of MIS 11 at this site. In addition, low resolution diatom records from other core sites in the Bering Sea indicate that seasonal sea ice advanced over the continental slope (Sites U1345 and U1343) and reached eastern Bowers Ridge (Site

U1340) during this interglacial (Stroynowski et al., 2015; Onodera et al., 2016; Teraishi et al.,

2016). In contrast, winter sea ice barely reaches the northernmost shelf today, and never extends as far south as Bowers Ridge or the Umnak Plateau. Unpublished counts of sea ice diatom species (Fragilariopsis cylindrus, Fragilariopsis oceanica, Fragilariopsis regina-jahniae,

Fossula arctica, Nitzschia frigida, Sinerima marigela, and Thalassiosira antarctica) completed in our lab (Marine Sediments Lab, Iowa State University) show that sea ice diatoms make up between 6 and 10% of the diatom assemblage in modern surface sediments from the slope

Sites U1345 and V21-163. Further south at Site UMK3A, the sea ice diatom Fragilariopsis cylindrus constitutes no more than 4% of the modern assemblage (Katsuki & Takahashi, 2005;

Caissie et al., 2010; Nesterovich & Caissie, in review). As a general guide, less than 3.5% of

Fragilariopsis species indicates perennially ice-free conditions, 13-18% indicates 1-4 months of sea ice cover, and >19% indicates more than 4 months of sea ice cover per year (Caissie et al.,

2010).

Interestingly, MIS 11 was unique in Beringia because tidewater glaciers advanced whilst eustatic sea level was still high, an example of the ‘out-of-phase glaciations’ observed during 27 interglacial periods in this region (Brigham-Grette, 2001). Several lines of evidence (molluscs, pollen, marine microfossils) indicate that ice advanced more than 200 km from the Brooks

Range to Kotzebue Sound (Huston et al., 1990). This glacial advance was driven by solar forcing, and by moisture from the flooded Beringian shelf (Brigham-Grette et al., 2001). Glacial advance during Late MIS 11 is tentatively correlated with the Nome River Glaciation (470 ka +/- 190 kyr), which is widely believed to be the last of the extensive glaciations in Beringia (Brigham-Grette,

2001; Kaufmann et al., 2001; Manley et al., 2001).

Summary

In the face of critical changes in Arctic sea ice extent and sea ice-dependent ecosystems, we reiterate the urgent need to improve our understanding of the processes involved in these changes. To this end, we use proxy records (sediment grain size, diatoms, stable isotopes) from

Bering Sea sediment cores to reconstruct natural variability in sea ice and primary productivity during MIS 11, which is considered one of the best analogs for the (natural) development of

Holocene climate. Our main objective is to fill an important gap in high-resolution proxy records of MIS 11 climate from the North Pacific region. Specifically, we focus on the following questions:

1. How can we use sediment grain size to infer the sedimentary history of the Bering Sea

during MIS 11?

2. What were sea ice conditions in the Bering Sea during Marine Isotope Stage 11, and

most importantly, how did sea ice respond to major warming trends, such as

deglaciation and the peak interglacial warmth of MIS 11c? 28

3. Were sea ice regimes synchronous across the Bering Sea, or do records indicate regional

variability in sea ice and related biological productivity, in keeping with the regional

variability described from LGM records?

4. What similarities/differences do we observe between sea ice and primary productivity

during MIS 11, in comparison to the Holocene?

5. How does the expression of MIS 11 in the Bering Sea compare to the global signal?

References

Addison, J.A., Finney, B.P., Dean, W.E., Davies, M.H., Mix, A.C., Stoner, J.S., & Jaeger, J.M., 2012. Productivity and sedimentary δ15N variability for the last 17,000 years along the northern Gulf of Alaska continental slope: Past Gulf of Alaska Productivity. Paleoceanography, 27, PA1206. https://doi.org/10.1029/2011PA002161

Aiello, I.W. & Ravelo, A.C., 2012. Evolution of marine sedimentation in the Bering Sea since the Pliocene. Geosphere, 8 (6), 1231-1253. https://doi.org/10.1130/GES00710.1

Ashton, N., Lewis, S.G., Parfitt, S.A., Penkman, K.E.H., & Coope, G.R., 2008. New evidence for complex climate change in MIS 11 from Hoxne, Suffolk, UK. Quat. Sci. Rev. 27 (7-8), 652- 668. https://doi.org/10.1016/j.quascirev.2008.01.003

Bauch, H.A., 2012. Interglacial climates and the Atlantic meridional overturning circulation: is there an Arctic controversy? Quat. Sci. Rev., 63, 1-22. https://doi.org/10.1016/j.quascirev.2012.11.023

Bauch, H. A., Erlenkeuser, H., Helmke, J. P., & Thiede, J., 2000. A paleoclimatic evaluation of Marine Oxygen Isotope Stage 11 in the high-northern North Atlantic (Nordic seas). Glob. Planet. Change, 24 (1), 27-39. https://doi.org/0.1016/S0921-8181(99)00067-3

Ben-Avraham, Z. & Cooper, A.K., 1981. Early evolution of the Bering Sea collision of oceanic rises and North Pacific subduction zones. Geol. Soc. Am. Bull., 92 (7), 485-495. https://doi.org/10.1130/0016-7606(1981)92<485:EEOTBS>2.0.CO;2

Bereiter, B., Eggleston, S., Schmitt, J., Nehrbass-Ahles, C., Stocker, T.F., Fischer, H., Kipfstuhl, S., & Chappellaz, J., 2015. Revision of the EPICA Dome C CO2 record from 800 to 600 kyr before present. Geophys. Res. Lett., 42 (2), 542-549. https://doi.org/10.1002/2014GL061957 29

Berger, A. & Loutre, M.F., 2002. An Exceptionally Long Interglacial Ahead? Science, 297 (5585), 1287-1288. https://doi.org/10.1126/science.1076120

Berger, W.H. & Wefer, G., 2003. On the Dynamics of The Ice Ages: Stage-11 Paradox, Mid- Brunhes Climate Shift, and 100-Ky Cycle. In: Droxler, A.W., Poore, R.Z., & Burckle, L.H. (Eds.) Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., pp. 41-59.

Bowen, D.Q., 1999. +23 m Sea Level in Southern Britain. In: Poore, R.Z., Burkle, L., Droxler, A., & McNulty, W.E. (Eds.) Marine Oxygen Isotope Stage 11 and associated Terrestrial Records. U.S. Dept. Interior. U. S. Geological Survey Open File Report, 99-312, pp. 15-17.

Bowen, D Q., 2010. Sea level ∼400,000 years ago (MIS 11): Analogue for present and future sea-level? Clim. Past, 6, 19-29. https://doi.org/10.5194/cp-6-19-2010

Brigham-Grette, J., 2001. New perspectives on Beringian Quaternary paleogeography, stratigraphy, and glacial history. Quat. Sci. Rev., 20, 15-24, 1-3. https://doi.org/10.1016/S0277-3791(00)00134-7

Brigham-Grette, J., Hopkins, D. M., Ivanov, V. F., Basilyan, A. E., Benson, S. L., Heiser, P. A., & Pushkar, V. S., 2001. Last interglacial (isotope stage 5) glacial and sea-level history of coastal Chukotka Peninsula and St. Lawrence Island, Western Beringia. Quat. Sci. Rev., 20, 419-436. https://doi.org/10.1016/S0277-3791(00)00107-4

Brown, Z.W. & Arrigo, K.R., 2012. Contrasting trends in sea ice and primary production in the Bering Sea and Arctic Ocean. J. Mar. Sci., 69, 1180-1193. https://doi.org/10.1093/icesjms/fss113

Brown, Z.W. & Arrigo, K.R., 2013. Sea ice impacts on spring bloom dynamics and net primary production in the Eastern Bering Sea. J. Geophys. Res. Oceans, 118, 43-62. https://doi.org/10.1029/2012JC008034

Brunelle, B.G., Sigman, D.M., Cook, M.S., Keigwin, L.D., Haug, G.H., Plessen, B., Schettler, G., & Jaccard, S.L., 2007. Evidence from diatom-bound nitrogen isotopes for subarctic Pacific stratification during the last ice age and a link to North Pacific denitrification changes. Paleoceanography, 22, PA1215. https://doi.org/10.1029/2005PA001205

Caissie, B. E., Brigham-Grette, J., Lawrence, K. T., Herbert, T. D., & Cook, M. S., 2010. Last Glacial Maximum to Holocene sea surface conditions at Umnak Plateau, Bering Sea, as inferred from diatom, alkenone, and stable isotope records. Paleoceanography, 25 (1), PA1206. https://doi.org/10.1029/2008PA001671

Caissie, B.E., Brigham-Grette, J., Cook, M.S., & Colmenero-Hidalgo, E., 2016. Bering Sea surface water conditions during Marine Isotope Stages 12 to 10 at Navarin Canyon (IODP Site U1345). Clim. Past, 12, 1739-1763. https://doi.org/10.5194/cp-12-1739-2016

Calvert S.E., 1974. Deposition and diagenesis of silica in marine sediments. In: Hsü, K.J. & Jenkyns, H.C. (Eds.), Pelagic sediments: On land and under the sea (Special Publication Number 1 of the International Association of Sedimentologists). Wiley-Blackwell, Oxford, pp. 273-299.

Candy, I., 2009. Terrestrial and freshwater carbonates in Hoxnian interglacial deposits, UK: micromorphology, stable isotopic composition and palaeoenvironmental significance. Proc. Geol. Assoc., 120 (1), 49-57. https://doi.org/10.1016/j.pgeola.2009.03.001

Candy, I. & McClymont, E.L., 2013. Interglacial intensity in the North Atlantic over the last 800000 years: Investigating the complexity of the mid-Brunhes Event. J. Quat. Sci., 28 (4), 343-348. https://doi.org/10.1002/jqs.2632

Candy, I., Schreve, D.C., Sherriff, J., & Tye, G.J., 2014. Marine Isotope Stage 11: Palaeoclimates, palaeoenvironments and its role as an analogue for the current interglacial. Earth Sci. Rev., 128, 18-51. https://doi.org/10.1016/j.earscirev.2013.09.006

Cavalieri, D. J., Parkinson, P., Gloersen, P., & Zwally, H.J., 1996, updated yearly. Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I Passive Microwave Data, Version 1. NASA National Snow and Ice Data Center Distributed Active Archive Center, Boulder, CO. https://doi.org/10.5067/8GQ8LZQVL0VL [accessed March 10, 2015].

Coachman, L.K., Whitledge, T.E., & Goering, J.J., 1999. Silica in Bering Sea Deep and Bottom Water. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 285-309.

Cook, M., Keigwin, L., & Sancetta, C., 2005. The Deglacial history of surface and intermediate water of the Bering Sea. Deep Sea Res. Part II: Topical Studies in Oceanography, 52 (16- 18), 2163-2173. https://doi.org/10.1016/j.dsr2.2005.07.004

Cornwall, W., 2019. Vanishing Bering Sea ice poses climate puzzle. Science, 364 (6441), 616-617. https://doi.org/10.1126/science.364.6441.616

Dabkowski, J., Limondin-Lozouet, N., Antoine, P., Andrews, J.E., Marca-Bell, A., & Robert, V., 2012. Climatic variations in MIS 11 recorded by stable isotopes and trace elements in a 31

French tufa (La Celle, Seine Valley. J. Quat. Sci., 27 (8), 790-799. https://doi.org/10.1002/jqs.2567 de Vernal, A. & Hillaire-Marcel, C., 2008. Natural Variability of Greenland Climate, Vegetation, and Ice Volume During the Past Million Years. Science, 320 (5883), 1622-1625. https://doi.org/10.1126/science.1153929

Detlef, H., Belt, S.T., Sosdian, S.M., Smik, L., Lear, C.H., Hall, I.R., Cabedo-Sanz, P., Husum, K., & Kender, S., 2018. Sea ice dynamics across the Mid-Pleistocene transition in the Bering Sea. Nat. Commun., 9, 941. https://doi.org/10.1038/s41467-018-02845-5

Dickson, A.J., Beer, C.J., Dempsey, C., Maslin, M.A., Bendle, J.A., McClymont, E.L., & Pancost, R.D., 2009. Oceanic forcing of the Marine Isotope Stage 11 interglacial. Nat. Geosci., 2, 428-433. https://doi.org/10.1038/ngeo527

Droxler, A.W. & Farrell, J.W., 2000. Marine isotope stage 11 (MIS 11): new insights for a warm future. Glob. Planet. Change, 24 (1), 1-5. https://doi.org/10.1016/S0921-8181(99)00065-X

Droxler, A.W., Poore, R.Z., & Burckle, L.H., 2003. Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., 240 pp.

Dutton, A., Carlson, A.E., Long, A.J., Milne, G.A., Clark, P.U., DeConto, R., Horton, B.P., Rahmstorf, S., Raymo, M.E., 2015. Sea-level rise due to polar ice-sheet mass loss during past warm periods. Science, 349 (6244), aaa4019. https://doi.org/10.1126/science.aaa4019

Fetterer, F., Knowles, K., Meier, W.N., Savoie, M., & Windnagel, A.K., 2017, updated daily. Sea Ice Index, Version 3. National Snow and Ice Data Center, Boulder, CO. https://doi.org/10.7265/N5K072F8. [accessed 26 February 2020].

Hermann, A.J., Gibson, G.A., Cheng, W., Ortiz, I., Aydin, K., Wang, M., Hollowed, A.B., & Holsman, K.K., 2019. Projected biophysical conditions of the Bering Sea to 2100 under multiple emission scenarios. ICES J. Mar. Sci., 76 (5), 1280-1304. https://doi.org/10.1093/icesjms/fsz043

Hodell, D.A., Charles, C.D., & Ninnemann, U.S., 2000. Comparison of interglacial stages in the South Atlantic sector of the southern Ocean for the past 450 kyr: implications for Marine Isotope Stage (MIS) 11. Glob. Planet. Change, 24 (1), 7-26. https://doi.org/10.1016/S0921-8181(99)00069-7

Hu, A., Meehl, G., Han, W., Abe-Ouchi, A., Morrill, C., Okazaki, Y., & Chikamoto, M., 2012. The Pacific-Atlantic seesaw and the Bering Strait. Geophys. Res. Lett., 39 (3), L03702. https://doi.org/10.1029/2011GL050567

Hu, A.X., Meehl, G.A., Han, W.Q., Otto-Bliesner, B., Abe-Ouchi, A., & Rosenbloom, N., 2015. Effects of the Bering Strait closure on AMOC and global climate under different background climates. Prog. Oceanogr., 132, 174-196. https://doi.org/10.1016/j.pocean.2014.02.004

Hunt, G.L., Jr., Coyle, K.O., Eisner, L.B., Farley, E.V., Heintz, R.A., Mueter, F., Napp, J.M., Overland, J.E., Ressler, P.H., Salo, S., & Stabeno, P.J., 2011. Climate impacts on eastern Bering Sea foodwebs: a synthesis of new data and an assessment of the Oscillating Control Hypothesis, ICES J. Mar. Sci., 68 (6), 1230-1243. https://doi.org/10.1093/icesjms/fsr036

Huston, M.M., Brigham-Grette, J., & Hopkins, D.M., 1990. Paleogeographic significance of middle Pleistocene glaciomarine deposits on Baldwin Peninsula, northwestern Alaska. Ann. Glaciol., 14, 111-114. https://doi.org/10.3189/S0260305500008375

Hrynowiecka, A., Żarski, M., & Drzewicki, W., 2019. The rank of climatic oscillations during MIS 11c (OHO and YHO) and post-interglacial cooling during MIS 11b and MIS 11a in eastern Poland. Geol. Q., 63 (2), 375-394. https://doi.org/10.7306/gq.1470

Imbrie, J. & Imbrie, J.Z., 1980. Modeling the Climatic Response to Orbital Variations. Science, 207 (4434), 943-953. https://doi.org/10.1126/science.207.4434.943

Jouzel, J., Masson-Delmotte, V., Cattani, O., Dreyfus, G., Falourd, S., Hoffmann, G., Minster, B., Nouet, J., Barnola, J.-M., Chappellaz, J.A., Fischer, H., Gallet, J.C., Johnsen, S.J., Leuenberger, M.C., Loulergue, L., Luethi, D., Oerter, H., Parrenin, F., Raisbeck, G.M., Raynaud, D., Schilt, A., Schwander, J., Selmo, E., Souchez, R., Spahni, R., Stauffer, B., Steffensen, J.P., Stenni, B., Stocker, T.F., Tison, J.-L., Werner, M., & Wolff, E.W., 2007. Orbital and millennial Antarctic climate variability over the past 800,000 years. Science, 317 (5839), 793-797. https://doi.org/10.1126/science.1141038

Kandiano, E.S., Bauch, H.A., Fahl, K., Helmke, J.P., Röhl, U., Pérez-Folgado, M., & Cacho, I., 2012. The meridional temperature gradient in the eastern North Atlantic during MIS 11 and its link to the ocean–atmosphere system. Palaeogeogr. Palaeoclimatol. Palaeoecol., 333- 334, 24-39. https://doi.org/10.1016/j.palaeo.2012.03.005

Kandiano, E., van der Meer, M., Schouten, S., Fahl, K., Damsté, J.S.S., & Bauch, H.A., 2017. Response of the North Atlantic surface and intermediate ocean structure to climate warming of MIS 11. Sci. Rep., 7, 46192. https://doi.org/10.1038/srep46192

Kanematsu, Y., Takahashi, K., Kim, S., Asahi, H., & Khim, B.-K., 2013. Changes in biogenic opal productivity with Milankovitch cycles during the last 1.3 Ma at IODP Expedition 323 Sites U1341, U1343, and U1345 in the Bering Sea. Quat. Int., 310, 213-220. https://doi.org/10.1016/j.quaint.2013.06.003

Kato, Y., Onodera, J., Suto, I., Teraishi, A., & Takahashi, K., 2016. Pliocene and Pleistocene paleoceanography in the western subarctic Pacific based on diatom analyses of ODP Leg 145 Hole 884B and IODP Expedition 323 Holes U1341B and U1343E. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 29-37. https://doi.org/10.1016/j.dsr2.2015.04.004

Katsuki, K. & Takahashi, K., 2005. Diatoms as paleoenvironmental proxies for seasonal productivity, sea-ice and surface circulation in the Bering Sea during the late Quaternary. Deep Sea Res. Part II: Top. Stud. Oceanogr., 52 (16-18), 2110-2130. https://doi.org/10.1016/j.dsr2.2005.07.001

Kaufman, D.S. & Brigham-Grette, J., 1993. Aminostratigraphic correlations and paleotemperature implications, Pliocene–Pleistocene high sea level deposits, northwestern Alaska. Quat. Sci. Rev., 12 (1), 21-33. https://doi.org/10.1016/0277-3791(93)90046-O

Kaufman, D.S., Manley, W.F., Forman, S.L., & Layer, P.W., 2001. Pre-Late-Wisconsin glacial history, coastal Ahklun Mountains, southwestern Alaska-new amino acid, thermoluminescence, and 40Ar/39Ar results. Quat. Sci. Rev., 20 (1-3), 337-352. https://doi.org/10.1016/S0277-3791(00)00112-8

Kleinen, T., Hildenbrandt, S., Prange, M., Rachmayani, R., Müller, S., Bezrukov, E., Brovkin, V., & Tarasov, P.E., 2014. The climate and vegetation of Marine Isotope Stage 11 – Model results and proxy-based reconstructions at global and regional scale. Quat. Int., 348, 247-265. https://doi.org/10.1016/j.quaint.2013.12.028

Knudson, K.P. & Ravelo, A.C., 2015. North Pacific Intermediate Water circulation enhanced by the closure of the Bering Strait. Paleoceanography, 30 (10), 1287-1304. https://doi.org/10.1002/2015PA002840

Koizumi, I. & Yamamoto, H., 2018. Diatom ooze and diatomite–diatomaceous sediments in and around the North Pacific Ocean. JAMSTEC-R, 27, 26-46. https://doi.org/10.5918/jamstecr.27.26

Kuehn, H., Lembke-Jene, L., Gersonde, R., Esper, O., Lamy, F., Arz, H., Kuhn, G., & Tiedemann, R., 2014. Laminated sediments in the Bering Sea reveal atmospheric teleconnections to Greenland climate on millennial to decadal timescales during the last deglaciation. Clim. Past, 10, 2215-2236. https://doi.org/10.5194/cp-10-2215-2014

Lam, P.J., Robinson, L.F., Blusztajn, J., Li, C., Cook, M.S., McManus, J.F., & Keigwin, L.D., 2013. Transient stratification as the cause of the North Pacific productivity spike during deglaciation. Nat. Geosci., 6 (8), 622-626. https://doi.org/10.1038/ngeo1873

Laskar, J., Fienga, A., Gastineau, M., & Manche, H., 2011. La2010: a new orbital solution for the long-term motion of the Earth. Astron. Astrophys., 532, A89. https://doi.org/10.1051/0004-6361/201116836

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., & Levrard, B., 2004. A long-term numerical solution for the insolation qualities of the Earth. Astron. Astrophys., 428 (1), 261-285. https://doi.org/10.1051/0004-6361:20041335

Lisiecki, L. E. & Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20, PA1003. https://doi.org/10.1029/2004PA001071

Lopes, C., Mix, A.C., & Abrantes, F., 2006. Diatoms in northeast Pacific surface sediments as paleoceanographic proxies. Mar. Micropaleontol., 60, 45-65. https://doi.org/10.1016/j.marmicro.2006.02.010

Loughlin, T.R., Sukhanova, I.N., Sinclaie, E.H., & Ferrero, R.C., 1999. Summary of Biology and Ecosystem Dynamics in the Bering Sea. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 387-407.

Loutre, M.F. & Berger, A., 2003. Marine Isotope Stage 11 as an analogue for the present interglacial. Glob. Planet. Change, 36, 209-217. https://doi.org/10.1016/S0921- 8181(02)00186-8

Lozhkin, A.V. & Anderson, P.M., 2013. Vegetation responses to interglacial warming in the Arctic: examples from Lake El'gygytgyn, Far East Russian Arctic. Clim. Past, 9, 1211-1219. https://doi.org/10.5194/cp-9-1211-2013

Lüthi, D., le Floch, M., Bereiter, B., Blunier, T., Barnola, J.M., Siegenthaler, U., Raynaud, D., Jouzel, J., Fischer, H., Kawamura, K., & Stocker, T.F., 2008. High-resolution carbon 35

dioxide concentration record 650,000–800,000 years before present. Nature, 453 (7193), 379-382. https://doi.org/10.1038/nature06949

Manley, W.F., Kaufman, D.S., & Briner, J.P., 2001. Pleistocene glacial history of the southern Ahklun Mountains, southwestern Alaska: Soil-development, morphometric, and radiocarbon constraints. Quat. Sci. Rev., 20 (1-3), 353-370. https://doi.org/10.1016/S0277-3791(00)00111-6

Masson-Delmotte, V., Dreyfus, G., Braconnot, P., Johnsen, S., Jouzel, J., Kageyama, M., Landais, A., Loutre, M.-F., Nouet, J., Parrenin, F., Raynaud, D., Stenni, B., & Tuenter, E., 2006. Past temperature reconstructions from deep ice cores: relevance for future climate change. Clim. Past, 2, 399-448. https://doi.org/10.5194/cp-2-145-2006

Matul, A.G., 2017. Probable limits of sea ice extent in the northwestern Subarctic Pacific during the last glacial maximum. Oceanology, 57 (5), 700-706. https://doi.org/10.1134/S0001437017050113

Matul, A., Abelmann, A., Khusid, T., Chekhovskaya, M., Kaiser, A., Nürnberg, D., & Tiedemann, R., 2016. Late Quaternary changes of the oxygen conditions in the bottom and intermediate waters on the western Kamchatka continental slope, the Sea of Okhotsk. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 184-190. https://doi.org/10.1016/j.dsr2.2013.03.023

McManus, J.F., Oppo, D.W., & Cullen, J.L., 1999. A 0.5-Million-Year Record of Millennial-Scale Climate Variability in the North Atlantic. Science, 283 (5404), 971-975. https://doi.org/10.1126/science.283.5404.971

McManus, J., Oppo, D., Cullen, J., & Healey, S., 2003. Marine Isotope Stage 11 (Mis 11): Analog for Holocene and Future Climate? In: Droxler, A.W., Poore, R.Z., & Burckle, R.H. (Eds.), Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., pp. 69-85.

Méheust, M., Stein, R., Fahl, K., & Gersonde, R., 2018. Sea-ice variability in the subarctic North Pacific and adjacent Bering Sea during the past 25 ka: new insights from IP25 and Ukʹ37 proxy records. arktos, 4 (8), 1-19. https://doi.org/10.1007/s41063-018-0043-1

Melles, M., Brigham-Grette, J., Minyuk, P.S., Nowaczyk, N.R., Wennrich, V., DeConto, R.M., Anderson, P.A., Andreev, A.A., Coletti, A., Cook, T.L., Haltia-Hovi, E., Kukkonen, M., Lozhkin, A.V., Rosén, P., Tarasov, P.E., Vogel, H., & Wagner, B., 2012. 2.8 Million years of Arctic climate change from Lake El'gygytgyn, NE Russia. Science, 337 (6092), 315-320. https://doi.org/10.1126/science.1222135

Milker, Y., Rachmayani, R., Weinkauf, M.F.G., Prange, M., Raitzsch, M., Schulz, M., & Kucera, M., 2013. Global and regional sea surface temperature trends during Marine Isotope Stage 11. Clim. Past, 9, 2231-2252. https://doi.org/10.5194/cp-9-2231-2013

Nesterovich, A., 2019. Quantitative diatom-based proxy for sea ice extent in the Bering and Chukchi seas. PhD Thesis, Iowa State University, 17277. https://lib.dr.iastate.edu/etd/17277

Nesterovich, A. & Caissie, B.E., in review. A diatom-based quantitative sea ice proxy for the Bering and Chukchi Seas. Prog. Oceanogr.

Niebauer, H.J., Alexander, V., & Henrichs, S.M., 1990. Physical and biological oceanographic interaction in the spring bloom at the Bering Sea marginal ice edge zone. J. Geophys. Res., 95 (C12), 22229-22241. https://doi.org/10.1029/JC095iC12p22229

Okazaki, Y., Takahashi, K., Katsuki, K., Ono, A., Hori, J., Sakamoto, T., Uchida, M., Shibata, Y., Ikehara, M., & Aoki, K., 2005. Late Quaternary paleoceanographic changes in the southwestern Okhotsk Sea: Evidence from geochemical, radiolarian, and diatom records. Deep Sea Res. Part II: Top. Stud. Oceanogr., 52, 2332-2350. https://doi.org/10.1016/j.dsr2.2005.07.007

Okazaki, Y., Timmermann, A., Menviel, L., Harada, N., Abe-Ouchi. A., Chikamoto, M.O., Mouchet, A., & Asahi, H., 2010. Deepwater formation in the North Pacific during the last glacial termination. Science, 329 (5988), 200-204, https://doi.org/10.1126/science.1190612

Olson, S.L. & Hearty, P.J., 2009. A sustained +21-m sea-level highstand during MIS 11 (400 ka): direct fossil and sedimentary evidence from Bermuda. Quat. Sci. Rev., 28 (3), 271-285. https://doi.org/10.1016/j.quascirev.2008.11.001

Onodera, J., Takahashi, K., & Nagatomo, R., 2016. Diatoms, silicoflagellates, and ebridians at Site U1341 on the western slope of Bowers Ridge, IODP Expedition 323. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 8-17. https://doi.org/10.1016/j.dsr2.2013.03.025

Palumbo, E., Voelker, A.H.L., Flores, J.A., & Amore, O.F., 2019. Surface-ocean dynamics during eccentricity minima: A comparison between interglacial Marine Isotope Stage (MIS) 1 and MIS 11 on the Iberian Margin. Glob. Planet. Change, 172, 242-255. https://doi.org/10.1016/j.gloplacha.2018.10.011

Pelto, B.M., Caissie, B.E., Petsch, S.T., & Brigham-Grette, J., 2018. Oceanographic and Climatic Change in the Bering Sea, Last Glacial Maximum to Holocene. Paleoceanogr. Paleoclimatol., 33, (1), 93-111. https://doi.org/10.1002/2017PA003265

Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola, J.-M., Basile, I., Bender, M., Chappellaz, J., Davis, M., Delaygue, G., Delmotte, M., Kotlyakov, V.M., Legrand, M., Lipenkov, V.Y., Lorius, C., Pépin, L., Ritz, C., Saltzman, E., & Stievenard, M., 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature, 399 (6735), 429-436. https://doi.org/10.1038/20859

Poli, M.S., Meyers, P.A., & Thunell, R.C., 2010. The western North Atlantic record of MIS 13 to 10: Changes in primary productivity, organic carbon accumulation and benthic foraminiferal assemblages in sediments from the Blake Outer Ridge (ODP Site 1058). Palaeogeogr. Palaeoclimatol. Palaeoecol., 295 (1-2), 89-101. https://doi.org/10.1016/j.palaeo.2010.05.018

Prokopenko, A.A., Bezrulova, E.V., Khursevich, G.K., Solotchina, E.P., Kuzmin, M.I., & Tarasov, P.E., 2010. Climate in continental interior Asia during the longest interglacial of the past 500 000 years: the new MIS 11 records from Lake Baikal, SE Siberia. Clim. Past, 6, 31-48. https://doi.org/10.5194/cp-6-31-2010

Prokopenko, A.A., Karabanov, E.B., Williams, D.F., Kuzmin, M.I., Shackleton, N.J., Crowhurst, S.J., Peck, J.A., Gvozdkov, A.N., & King, J.W., 2001. Biogenic silica record of the Lake Baikal response to climatic forcing during the Brunhes chron. Quat. Res., 55 (2), 123- 132. https://doi.org/10.1006/qres.2000.2212

Railsback, L., Gibbard, P., Head, M., Voarintsoa, N.R., & Toucanne, S., 2015. An optimized scheme of lettered marine isotope substages for the last 1.0 million years, and the climatostratigraphic nature of isotope stages and substages. Quat. Sci. Rev., 111, 94- 106. https://doi.org/10.1016/j.quascirev.2015.01.012

Ran, L.H., Chen, J.F., Jin, H.Y., Li, H., Lu, W., & Wang, K., 2013. Diatom distribution of surface sediment in the Bering Sea and Chukchi Sea. Adv. Polar. Sci., 24 (2), 106-112. https://doi.org/10.3724/SP.J.1085.2013.00106

Raymo, M.E. & Mitrovica, J.X., 2012. Collapse of polar ice sheets during the stage 11 interglacial. Nature, 483 (7390), 453-456. https://doi.org/10.1038/nature10891

Reed, R.K. & Stabeno, P.J., 1999. The Aleutian North Slope Current. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 177-191.

Roach, A.T., Aagaard, K., Pease, C.H., Salo, S.A., Weingartner, T., Pavlov, V., & Kulakov, M., 1995. Direct measurements of transport and water properties through the Bering Strait. J. Geophys. Res. Oceans, 100, 18,443-18,458.

Roden, G.I., 2000. Flow and water property structures between the Bering Sea 16 and Fiji in the summer of 1993. J. Geophys. Res., 105 (C12), 28,595-28,612. https://doi.org/10.1029/1999JC000030

Rodrigues, T., Voelker, A.H.L., Grimalt, J.O., Abrantes, F., & Naughton, F., 2011. Iberian Margin sea surface temperature during MIS 15 to 9 (580–300 ka): glacial suborbital variability versus interglacial stability. Paleoceanography, 26, PA1204. http://dx.doi.org/10.1029/2010PA001927

Rohling, E.J., Foster, G.L., Grant, K.M., Marino, G., Roberts, A.P., Tamisiea, M.E., & Williams, F., 2014. Sea-level and deep-sea-temperature variability over the past 5.3 million years. Nature, 508 (7497), 477-482. https://doi.org/10.1038/nature13230

Rohling, E.J., Braun, K., Grant, K., Kucera, M., Roberts, A.P., Siddall, M., & Trommer, G., 2010. Comparison between Holocene and Marine Isotope Stage-11 sea-level histories. Earth Planet. Sci. Lett., 291 (1), 97-105. https://doi.org/10.1016/j.epsl.2009.12.054

Ruddiman, W.F., 2007. The early anthropogenic hypothesis: Challenges and responses. Rev. Geophys., 45, RG4001. https://doi.org/10.1029/2006RG000207

Sambrotto, R.N., Goering, J.J., & and McRoy, C.P., 1984. Large yearly production of phytoplankton in the western Bering Strait. Science, 225, 1147-1150. https://doi.org/10.1126/science.225.4667.1147

Sancetta, C., Heusser, L., Labeyrie, L., Naidu, A.S., & Robinson, S.W., 1985. Wisconsin— Holocene Paleoenvironment of the Bering Sea: Evidence from Diatoms, Pollen, Oxygen Isotopes and Clay Minerals. Mar. Geol., 62 (1-2), 55-68. https://doi.org/10.1016/0025- 3227(84)90054-9

Schlung, S.A., Ravelo, A.C, Aiello, I.W., Andreasen, D.H., Cook, M.S., Drake, M., Dyez, K.A., Guilderson, T.P., LaRiviere, J.P., Stroynowski, Z., & Takahashi, K., 2013. Millennial-scale climate change and intermediate water circulation in the Bering Sea from 90 ka: A high- resolution record from IODP Site U1340. Paleoceanography, 28 (1), 54-67. https://doi.org/10.1029/2012PA002365

Shi, P., Yang, T., Tian, Q., & Li., C., 2016. A warmer but drier Marine Isotope Stage 11 during the past 650 ka as revealed by the thickest loess on the western Chinese Loess Plateau. J. Arid Land, 8, 315-330. https://doi.org/10.1007/s40333-016-0123-7

Siegenthaler, U., Stocker, T.F., Monnin, E., Lüthi, D., Schwander, J., Stauffer, B., Raynaud, D., Barnola, J.M., Fischer, H., Masson-Delmotte, V., & Jouzel, J., 2005. Stable carbon cycle- climate relationship during the late Pleistocene. Science, 310 (5752), 1313-1317. https://doi.org/10.1126/science.1120130

Sigler, M.F., Harvey, H.R., Ashjian, J., Lomas, M.W., Napp, J.M., Stabeno, P.J., & Van Pelt, T.I., 2010. How Does Climate Change Affect the Bering Sea Ecosystem? EOS, 91 (48), 457- 458. https://doi.org/10.1029/2010EO480001

Sigler, M.F., Stabeno, P.J., Eisner, L.B., Napp, J.M., & Mueter F.J., 2014. Spring and fall phytoplankton blooms in a productive subarctic ecosystem, the eastern Bering Sea, during 1995-2011. Deep Sea Res. Part II: Top. Stud. Oceanogr., 109, 71-83. https://doi.org/10.1016/j.dsr2.2013.12.007

Stabeno, P.J. & Bell, S.W., 2019. Extreme Conditions in the Bering Sea (2017–2018): Record- Breaking Low Sea-Ice Extent. Geophys. Res Lett., 46 (15), 8952-8959. https://doi.org/10.1029/2019GL083816

Stabeno, P.J., Bell, S.W., Bond, N.A., Kimmel, D.G., Mordy, C.W., & Sullivan, M.E., 2019b. Distributed biological observatory region 1: Physics, chemistry and plankton in the northern Bering Sea. Deep Sea Res. Part II: Top. Stud. Oceanogr., 162, 8-21. https://doi.org/10.1016/j.dsr2.2018.11.006

Stabeno, P.J., Kachel, N.B., Moore, S.E., Napp, J.M., Sigler, M., Yamaguchi, A., & Zerbini, A.N., 2012. Comparison of warm and cold years on the southeastern Bering Sea shelf and some implications for the ecosystem. Deep Sea Res. Part II: Top. Stud. Oceanogr., 65-70, 31-45. https://doi.org/10.1016/j.dsr2.2012.02.020

Stabeno, P.J., Ladd, C., & Reed, R.K., 2009. Observations of the Aleutian North Slope Current, Bering Sea, 1996–2001. J. Geophys. Res. Oceans, 114, C05015. https://doi.org/10.1029/2007JC004705

Stabeno, P., Napp, J., Mordy, C., & Whitledge, T., 2010. Factors influencing physical structure and lower trophic levels of the eastern Bering Sea shelf in 2005: sea ice, tides, and winds. Prog. Oceanogr., 85 (3-4), 180-196. https://doi.org/10.1016/j.pocean.2010.02.010

Stabeno, P.J., Schumacher, J.D., & Ohtani, K., 1999. The physical oceanography of the Bering Sea. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 1-28. 40

Stabeno, P.J., Thoman, R.L., & and Wood, K., 2019a. Recent Warming in the Bering Sea and Its Impact on the Ecosystem. In: Richter-Menge, J., Druckenmiller, M.L., & Jeffries, M. (Eds.), Arctic Report Card 2019. http://www.arctic.noaa.gov/Report-Card.

Stroynowski, Z., Ravelo, A. C., & Andreasen, D., 2015. A Pliocene to recent history of the Bering Sea at Site U1340A, IODP Expedition 323. Paleoceanography, 30 (12), 1641-1656. https://doi.org/10.1002/2015PA002866

Takahashi, K., Fujitani, N., Yanada, N., & Maita, Y., 2000. Long-term biogenic particle fluxes in the Bering Sea and the central subarctic Pacific Ocean, 1990-1995. Deep Sea Res. Part I: Oceanogr. Res. Pap., 47 (9), 1723-1759. https://doi.org/10.1016/S0967-0637(00)00002-9

Teraishi, A., Suto, I., Onodera, J., & Takahashi, K., 2016. Diatom, silicoflagellate and ebridian biostratigraphy and paleoceanography in IODP 323 Hole U1343E at the Bering slope site. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 18-28. https://doi.org/10.1016/j.dsr2.2013.03.026

Tsunogai, S., Kusakabe, M., Iizumi, H., Koike, I., & Hattori, A., 1979. Hydrographic features of the deep water of the Bering Sea—the sea of Silica. Deep Sea Res. Part I: Oceanogr. Res. Pap., 26 (6), 641-659. https://doi.org/10.1016/0198-0149(79)90038-4

Tzedakis, P.T., Andrieu, V., de Beaulieu, J.-L., Birks, H.J.B., Crowhurst, S., Follieri, M., Hooghiemstra, H., Magri, D., Reille, M., Sadori, L., Shackleton, N.J., & Wijmstra, T.A., 2001. Establishing a terrestrial chronological framework as a basis for biostratigraphical comparisons. Quat. Sci. Rev., 20 (16-17), 1583-1592. https://doi.org/10.1016/S0277- 3791(01)00025-7

Vancoppenolle, M., Meiners, K.M., Michel, C., Bopp, L., Brabant, F., Carnat, G., Delille, B., Lannuzel, D., Madec, G., Moreau, S., Tison, J.-L., & van der Merwe, P., 2013. Role of sea ice in global biogeochemical cycles: emerging views and challenges. Quat. Sci. Rev., 79, 207-230. https://doi.org/10.1016/j.quascirev.2013.04.011

Voelker, A.H.L., Rodrigues, T., Billups, K., Oppo, D., McManus, J., Stein, R., Hefter, J., & Grimalt, J.O., 2010. Variations in mid-latitude North Atlantic surface water properties during the mid-Brunhes (MIS 9–14) and their implications for the thermohaline circulation. Clim. Past, 6, 531-552. https://doi.org/10.5194/cp-6-531-2010

Walsh, J.J., McRoy, C.P., Blackburn, T.H., Coachman, L.K., Goering, J.J., Henriksen, K., Andersen, P., Nihoul, J.J., Parker, P.L., Springer, A.M., Tripp, R.B., Whitledge, T.E., & Wirick, C.D., 1989. The role of Bering Strait in the carbon/nitrogen fluxes of polar marine ecosystems. In: Rey, L. & Alexander, V. (Eds.), Proceedings of the Sixth Conference of the Comité Arctique International. Brill, Leiden, Netherlands, pp. 90-120.

Wang, M. & Overland, J.E., 2009. A sea ice free summer Arctic within 30 years? Geophys. Res. Lett., 36 (7), L07052. https://doi.org/10.1029/2009GL037820

Wang, M., Yang, Q., Overland, J., & Stabeno, P., 2018. Sea-ice cover timing in the Pacific Arctic: The present and projections to mid-century by selected CMIP5 models. Deep Sea Res. Part II: Top. Stud. Oceanogr., 152, 22-34. https://doi.org/10.1016/j.dsr2.2017.11.017

Warner, M. & Roden, G., 1995. Chlorofluorocarbon evidence for recent ventilation of the deep Bering Sea. Nature, 373, 409-412. https://doi.org/10.1038/373409a0

Woodgate, R.A., Weingartner, T., & Lindsay, R., 2010. The 2007 Bering Strait oceanic heat flux and anomalous Arctic sea-ice retreat. Geophys. Res. Lett., 37, L01602. https://doi.org/10.1029/2009GL041621

Worne, S., Kender, S., Swann, G.E.A., Leng, M.J., & Ravelo, A.C., 2019. Coupled climate and subarctic Pacific nutrient upwelling over the last 850,000 years. Earth Planet. Sci. Lett., 552, 87-97. https://doi.org/10.1016/j.epsl.2019.06.028

Yin, Q. & Berger, A., 2011. Individual contribution of insolation and CO2 to the interglacial climates of the past 800,000 years. Clim. Dyn., 38 (3), 709-724. https://doi.org/10.1007/s00382-011-1013-5

Zheng, Y., Van Geen, A., Anderson, R.F., Gardner, J.V., & Dean, W.E., 2000. Intensification of the northeast Pacific oxygen minimum zone during the Bolling-Allerod warm period. Paleoceanography, 15 (5), 528–536, 2000. https://doi.org/10.1029/1999PA000473

CHAPTER 3. EVALUATING THE PALEOENVIRONMENTAL SIGNIFICANCE OF SEDIMENT GRAIN

SIZE IN BERING SEA SEDIMENTS DURING MIS 11

Modified from a manuscript to be submitted to Quaternary Science Reviews

Natalie S. Thompsona & Beth E. Caissiea,b

aDepartment of Geological & Atmospheric Science, Iowa State University, Ames, IA bCurrently at: US Geological Survey, Menlo Park, CA

Abstract

Grain size is an important textural property of sediments and is widely used in paleoenvironmental studies as a means to infer changes in the sedimentary environment.

However, grain size parameters are not always easy to interpret, without a full understanding of the factors that influence grain size. Here, we measure grain size in sediment cores from the

Bering slope and the Umnak Plateau, and review the effectiveness of different grain size parameters as proxies for sediment transport, current strength, and productivity, during a past warm interval (Marine Isotope Stage 11, 424-374 ka).

In general, sediments in the Bering Sea are hemipelagic, making them ideal deposits for paleoenvironmental reconstructions, but there is strong evidence in the grain size distribution for contourite deposits between ~408-400 ka at the slope sites, suggesting a change in bottom current transport at this time. We show that the grain size of coarse (>150 µm) terrigenous sediment can be used effectively as a proxy for ice rafting, although it is not possible to distinguish between iceberg and sea ice rafting processes, based on grain size alone. We find that the mean grain size of bulk sediments can be used to infer changes in productivity on 43 glacial-interglacial timescales, but the size and preservation of diatom valves also exert a control on mean grain size. Lastly, we show that the mean size of sortable silt (10-63 µm) is not a valid proxy for bottom current strength in the Bering Sea, because the input of ice-rafted silt confounds the sortable silt signal.

Introduction

Marine sediments provide valuable records of Earth’s climate history over long (millions of years) time scales and are thus a highly valuable tool in paleoceanographic and paleoclimatic reconstructions. In particular, the textural properties of marine sediments can help us to gain a deeper understanding of the different sedimentary processes that operate under changing environmental conditions (Poppe et al., 2006; Rothwell & Rack, 2006).

Grain size is a fundamental property of ocean sediments that is widely used in paleoceanographic studies, most notably as a proxy for sediment source, transport, and depositional processes. For example, variations in the grain size of marine sediments have been related to changes in paleoproductivity (Warner & Domack, 2002; Muhong et al., 2005; Aiello &

Ravelo, 2012), fluvial discharge (Weltje & Prins, 2003; Briceño-Zuluaga et al., 2016; Carlin et al.,

2019), aeolian dust input (Weltje & Prins, 2003; Holz et al., 2007; Serno et al., 2014; Stuut et al.,

2014; Briceño-Zuluaga et al., 2016), and the strength and flow speed of paleo currents (e.g.,

McCave et al., 1995, 2017; Tegez et al., 2014; Voigt et al., 2016; Hoffman et al., 2019). In addition, grain size is key to interpreting sediment composition (Sval’nov & Alekseeva, 2006;

Aiello & Ravelo, 2012), and can also be used to help distinguish between different types of deep-water sediment facies (e.g., Stow & Tabrez, 1998; Stow & Smillie, 2020). 44

Of particular importance in high latitude settings is the information that grain size can provide about glacial and sea ice extent. Numerous studies have used sediment grain size records to infer the ice rafting history of a region (e.g., Nürnberg et al., 1994; Wolf-Welling et al., 1996; Andrews & Principato, 2002; Sakamoto et al., 2005; St. John et al., 2008; deGelleke et al., 2013; O’Regan et al., 2014; Kim et al., 2018). By way of example, St. John et al. (2008) used grain size to reconstruct the Cenozoic history of ice rafting in the Arctic Ocean, and showed that ice was present in the Arctic Ocean as early as the middle Eocene, whilst Sakamoto et al. (2005) used the grain size distribution of terrigenous particles to reconstruct sea ice expansion events in the Sea of Okhotsk over the past 100 ka.

Whilst grain size has the potential to be an extremely useful proxy, it is not always easy to interpret because of the multiple factors that influence grain size. Bering Sea sediments are a mix of biogenic, terrigenous, and volcanogenic sediments that reflect a wide range of transport and depositional processes, including vertical settling, winds, sea ice, icebergs, gravity flows, bottom currents, and other transport mechanisms. Here we show that Bering Sea grain size records can successfully be used to infer the sedimentary history of the Bering Sea during a past warm interval (Marine Isotope Stage 11; 424-374 ka). We measure grain size in sediments from three core sites on the Bering slope and Umnak Plateau, which were obtained during Integrated

Ocean Drilling Program (IODP) Expedition 323 in 2009. We calculate a number of grain size parameters for the sediments, and review the limitations and effectiveness of these parameters to serve as a proxy for sediment facies, transport mechanisms such as ice rafting, bottom current strength, and paleoproductivity.

Study Area

Location and Oceanographic Setting

The Bering Sea, located between Russia and Alaska, is a marginal sea of the North Pacific

Ocean. It comprises a broad (>500 km), shallow continental shelf in the east, a narrow (<100 46 km) shelf to the west, a steep continental slope, and a deep abyssal basin, dissected by several large submarine rises (Fig. 3.1; Ben-Avraham & Cooper, 1981; Stabeno et al., 1999). To the north, the shallow (-50 m) Bering Strait serves as the only connection between the Pacific and

Arctic Oceans (Fig. 3.1).

Water mass exchange with the North Pacific occurs via passes in the Aleutian Islands, linking Bering Sea conditions to those of the Pacific (Stabeno et al., 1999). Nutrient-rich currents, derived from inflow of the Alaska Coastal Current and the Alaskan Stream, flow along the continental slope and supply critical nutrients to the shelf (Fig. 3.1). These nutrient-rich surface waters contribute to the extremely high rates of biological productivity in the Bering

Sea, which is one of the most productive marine ecosystems in the world (Loughlin et al., 1999;

Sigler et al., 2010; Brown and Arrigo, 2012, 2013; Stabeno et al., 2019).

Sedimentation in the Bering Sea

Sediments in the Bering Sea are primarily a mix of biogenic and fine-grained siliciclastic particles, whilst secondary components of the sediment include volcanogenic material, sand- sized siliciclastics, and ice-rafted debris (IRD) (Takahashi et al., 2011; Aiello & Ravelo, 2012). The biogenic materials largely comprise siliceous diatom frustules, with lesser contributions from other microfossils, including foraminifera, radiolarians, and sponge spicules (Takahashi et al.,

2011; Aiello & Ravelo, 2012). Terrigenous sediments in the Bering Sea are mostly derived from the surrounding landmasses, in particular, the Alaskan mainland, the Alaskan Peninsula, and eastern Siberia, and to a lesser extent, from the Aleutian Arc, which is the primary source of volcanogenic input to the region (Naidu & Mowatt, 1983; Asahara et al., 2012; Nagashima et al., 2012). These sediments are transported to the Bering shelf by aeolian and fluvial processes, 47 as well as ice rafting. In the North Pacific today, wind-blown aerosols are largely restricted to the vicinity of their source environments, and aeolian input of desert dust from the Asian continent is minimal (0.5-1 g m2 yr-1) (Mahowald et al., 2005; Serno et al., 2014), although there is evidence for enhanced dust fluxes during past glacial periods (Riethdorf et al., 2013; Shaffer

& Lambert, 2018). Major rivers, including the Yukon, Kuskokwim and Anadyr, discharge millions of tons of sediment — mostly clay, silt, and fine sand-sized siliciclastics — to the continental shelf; in particular, the Yukon River provides ~63% of the total sediment load to the Bering Sea

(Riethdorf et al., 2013). In addition, icebergs and sea ice may entrain and transport a mix of coarse and fine terrigenous material far offshore (e.g. Nürnberg et al., 1994; Lisitzin, 2002;

Darby & Zimmerman, 2008; Darby et al., 2011; Mager et al., 2013; St John et al., 2015).

Sediment dispersal across the shelf is further influenced by wave energy, storms, and ocean currents.

Core Sites

Sediment cores used in this study were retrieved from the Bering slope and the Umnak

Plateau during IODP Expedition 323 to the Bering Sea in 2009 (Fig. 3.1; Takahashi et al., 2011).

Site U1345 is the northernmost of the three core sites and is located on an interfluve ridge near the shelf-slope break, just south of Navarin Canyon, at a water depth of 1008 m (Fig. 3.1;

Takahashi et al., 2011). Site U1343 is situated on a topographic high (-1953 m), which is isolated from the Bering Shelf by the Zhemchug Canyon (Fig. 3.1; Takahashi et al., 2011). Both of these sites are located relatively close to the modern limit of winter sea ice on the northern shelf (Fig.

3.1) and are situated within the highly productive Bering Sea Green Belt (Springer, 1996;

Takahashi et al., 2011). Site U1339 is located in the southeastern Bering Sea, on the northwest 48 flank of Umnak Plateau, at a water depth of 1870 m. Today, Site U1339 is influenced by the relatively warm Alaskan Stream, which inhibits the formation of sea ice in the region, but there is evidence for substantial sea ice cover at Site U1339 and other regions of the Bering Sea in the past (e.g., Sancetta et al., 1985; Cook et al., 2005; Katsuki & Takahashi, 2005; Caissie et al.,

2010, 2016; Onodera et al., 2016; Méheust et al., 2018; Pelto et al., 2018; Nesterovich and

Caissie, in review).

Methods

Geochronology

Shipboard bio- and magnetostratigraphy were used to place the sediment core records in the correct stratigraphic position (Takahashi et al., 2011). Published benthic foraminiferal oxygen isotope measurements (Asahi et al., 2016; Cook et al., 2016) were used to align the records to the global benthic stack (LR04; Lisiecki & Raymo, 2005), and dates for each sample were obtained using linear interpolation between tie points (Fig. 3.2). The age model for Site

U1345 was subsequently improved by adding an extra tie point based on magnetic susceptibility (Caissie et al., 2016).

Grain Size Measurements

Sediment was sampled periodically from the primary splices at core sites U1345 (n = 97),

U1343 (n =239) and U1339 (n = 296), and the grain size of these samples was measured using a

Malvern Mastersizer Laser 3000, equipped with a Hydro MV dispersion tank. This instrument uses the principle of laser diffraction to determine the volume distribution of particles in 101 size bins ranging from 0.01 to 3500 µm. 49

Freeze-dried bulk sediments were massed to approximately 0.025 g and treated with the deflocculant sodium hexametaphosphate (SHMP) prior to being analyzed. To obtain the grain size of just the terrigenous fraction, a second set of samples were sequentially treated with 30% H2O2, 10% HCl, and 1 M NaOH to remove the organic, carbonate, and siliceous biogenic material, respectively, before grain size was measured (Sakamoto et al., 2005). Bulk grain size records for Site U1345 were published previously by Caissie et al. (2016); we used the same samples to measure the grain size of the terrigenous fraction only.

Figure 3.2. Benthic foraminiferal δ18O values for IODP sites U1345 (black; Cook et al., 2016, U1343 (blue; Asahi et al., 2016) and U1339 (red; Cook et al., 2016) compared to the LR04 stack (grey; Lisiecki and Raymo, 2005). Inverted triangles show tie points between Bering Sea δ18O (filled) and magnetic susceptibility (open) records and the global stack. The grey bar shows the duration of MIS 11.

Statistical Analyses

The raw dataset provided by the Malvern software includes the volume % of grains in

109 bin sizes, as well as the 10th (Dx10), 50th (Dx50) and 90th (Dx90) percentiles. We also calculated the volume distribution of grains in the following size fractions: clay (<2 µm); silt (2- 50

63 µm); sand (63-2000 µm); gravel (>2000 µm); ice-rafted debris (>150 µm; >250 µm), and sortable silt (10-63 µm), based on a modified Udden-Wentworth scale (Table 3.1).

Table 3.1. Size scale adopted in the GRADISTAT program, compared with those previously used by Udden (1914), Wentworth (1922) and Friedman and Sanders (1978). From Blott & Pye, 2001.

Additional grain size parameters, including mean size, sorting and skewness, were calculated in GRADISTAT, a program that allows for rapid analysis of grain size statistics by a variety of methods (Blott & Pye, 2001). Here, we use the geometric (modified) Folk & Ward

(1957) graphical measure, which provides a robust basis for routine comparisons of compositionally variable sediments (Blott & Pye, 2001). A further advantage of the Folk and

Ward method is the opportunity to convert parameter values to descriptive terms for the sediment (see Table 3.2).

Smear Slide Analyses

Smear slide analyses were conducted on a subset of samples from the three sites (n =

55), in order to quantify the relative contribution of terrigenous and biogenic material, and elucidate the relationship between sediment composition and grain size. The biogenic and 52 mineral components in 10 random fields of view were identified, using a Nikon ECLIPSE Ni transmitted light microscope at magnifications of 100x and 400x, and the relative percent abundance of each component was visually estimated.

Results

Sediment Composition

Takahashi et al. (2011) used shipboard core descriptions and low-resolution smear slide counts to show that sediments in the Bering Sea are primarily a mix of terrigenous and biogenic particles. Results from our higher resolution smear slide counts support these findings. We identified three main sediment compositions: siliciclastic (>60% siliciclastic material); mixed

(sub-equal proportions of terrigenous and biogenic material); and biogenic (>60% biogenic material) (Fig. 3.3). The terrigenous sediments mainly consist of clay- to silt-sized siliciclastics.

Volcanogenic material, mostly in the form of tephra shards, and coarse (>150 µm) siliciclastic minerals and rock fragments typically make up a small (<10%) proportion of the sediment.

Biogenic sediments largely comprise siliceous diatom frustules (Fig. 3.3c), although other microfossils, including sponge spicules, radiolarians, and foraminifera, are present in much lower abundances.

At Site U1345, sediments are more siliciclastic, with diatom content ranging from as little as 2%, to a maximum of 78% (average 26.5%). In contrast, sediments at Site U1339 are more biogenic in composition; diatom content varies from 10 to 95% (average 58%). Sediments at Site U1343 represent something of an intermediate composition between U1345 and U1339, with a more equal mix of terrigenous and biogenic sedimentation. 53

Grain Size Parameters

Size Fractions

Sediments in the Bering Sea contain a wide range of particle sizes, from fine clays to coarse sands and gravel. For simplicity, we grouped sediments into four size classes (clay, silt, sand, and gravel), based on a modified Udden-Wentworth scale (Table 3.1). Because the proportion of grains in each size class varies between samples, grain size distribution plots show a range of modal compositions, including polymodal, trimodal, bimodal, and rarely, unimodal, although typically, there is always one dominant mode (Fig. 3.4).

Figure 3.5 shows downcore variations in the volume % of grains in each size class for bulk and terrigenous samples. At all three sites, the majority of particles fall within the silt-sized

(2-63 μm) fraction. The volume % of silt-sized grains is slightly higher for bulk sediments, averaging 84.2, 78.7, and 77.7% at sites U1345, U1343, and U1339, respectively, compared to

71.9, 73.9, and 74.6% for terrigenous sediments (Table 3.3). In bulk sediments, the proportion of silt-sized grains is highest at Site U1345, and lowest at U1339; when biogenic material is dissolved, this trend is reversed (Fig. 3.5; Table 3.3). Sand (63-2000 μm) is the second most abundant size fraction, followed by clay (<2 μm), then gravel (>2000 μm). In bulk sediment 54 samples, the volume % clay is relatively low at all sites, particularly at Site U1345, and the % sand is lowest at U1345 and highest at U1339. After removing biogenics, the volume % of clay- sized grains increases at all sites, and is highest at U1339, and lowest at U1345. In contrast, the

% of sand-sized particles is highest at U1345, and lowest at U1339 (Fig. 5.5; Table 3.3)

Figure 3.4. Grain size distribution plots for representative samples from the Bering Sea, showing unimodal, bimodal, trimodal, and polymodal distributions. 55

Table 3.3. Grain size averages for bulk (top) and terrigenous (bottom) sediments from IODP Sites U1345, U1343, and U1339.

Bulk Sediment Median Mean % Clay % Silt % Sand Sorting Skewness Site n (μm) (μm) (>2 μm) (2-63 μm) (63-2000 μm) U1345 97 20.8 20.3 2.9 -0.01 1.1 84.2 14.5 U1343 239 17.3 16.8 3.7 -0.03 5.8 78.7 14.9 U1339 297 19.7 19.1 3.6 -0.05 4.3 77.5 18.1

Terrigenous Sediment Median Mean % Clay % Silt % Sand Sorting Skewness Site n (μm) (μm) (>2 μm) (2-63 μm) (63-2000 μm) U1345 89 21.7 20.7 4.8 -0.02 6.2 71.9 20.2 U1343 242 17.4 17.4 4.4 0.02 7.3 73.9 17.9 U1339 290 12.6 13.6 4.8 0.11 9.6 74.6 15

Statistical Measures

Here, we use the graphic statistical measures of mean, sorting, and skewness, to describe grain size distributions (Table 3.2; Fig. 3.6; Folk & Ward, 1957; Blott & Pye, 2001). It should be noted that Blott & Pye (2001) urge caution when using grain size statistics to analyze multimodal sediments, suggesting that descriptors provided by the GRADISTAT software (e.g., mode, median, distribution spread) may be more reliable. However, as the above statistics are routinely used in studies of multimodal marine sediments (e.g., Cronan, 1972; Schlee, 1973;

McLaren, 1981; Martins & Barboza, 2005; Warrier et al., 2016), we chose to include these measures in our analyses.

Mean grain size was computed by averaging particle sizes at the 16th, 50th, and 84th percentile values (Table 3.2; Folk & Ward, 1957). Sorting, or standard deviation, is a measure of the spread of particles around the average. We applied the inclusive standard deviation of Folk 57

& Ward (1957), which includes the 5th and 95th percentiles, to define a spread of 1.65 standard deviations on either side of the mean (Table 3.2). Skewness is a measure of the asymmetry of the grain size distribution. In a normal (symmetrical or near-symmetrical) distribution, the mean, median, and mode all coincide. By convention, skewness is measured in phi (ɸ) units, where a positive skew indicates a tail in the direction of the fine particles, with the mean and median shifted towards finer grain sizes, whilst a negative skew indicates a tail in the direction of the coarse particles (McManus et al., 1988; Blott & Pye, 2001; Fig. 3.7). However, logarithmic and geometric skewness parameters are inversely related, so to avoid confusion, we replace the terms ‘positive’ and ‘negative’, with ‘fine-skewed’, indicating an excess of fine particles, and coarse-skewed, indicating a tail of coarse particles (Fig. 3.7; Blott & Pye, 2001). Kurtosis - the expression of sorting in the tails relative to the central distribution - is not widely used in grain size interpretations, except to measure the non-normality of a distribution (Blatt et al., 1982;

McLaren, 1981; McManus et al., 1988), so we chose not to examine this parameter.

Mean grain size values also fall mostly within the silt-sized fraction, except for intervals of coarsening centered around 403.5 ka at sites U1345 and U1343 (Fig. 3.6). Overall, mean grain size is highest at the northernmost site (U1345) (Table. 3.3). The mean grain size of bulk sediments is higher at Site U1339 than at U1343, but for terrigenous sediments, mean grain size is higher at U1343 (Table 3.3). In general, mean grain size at sites U1345 and U1343 is similar for both bulk and terrigenous sediments, but at Site U1339, bulk sediments are coarser than terrigenous sediments (Table 3.3; Fig. 3.6).

Figure 3.6. Downcore variations in median grain size, mean grain size, sorting, and skewness for bulk (solid line) and terrigenous (dashed line) sediments from Bering Sea core sites U1345 (black), U1343 (blue), and U1339 (red) from 430-365 ka. Sorting and skewness values are related to descriptive terms for sorting (poorly sorted [2-4]; very poorly sorted [4-16]) and skewness (symmetrical [-0.1 to 0.1]; fine skewed [-0.3 to -0.1]; coarse skewed [0.1 to 0.3]) after Folk & Ward (1957). Dots indicate the stratigraphic position of smear slide samples; yellow dots are smear slides from Takahashi et al., 2011. Brown bars show ash layers/accessories, and the grey panel shows the duration of MIS 11.

Figure 3.7. Examples of fine-skewed (U1343A 12H-5 78 cm) and coarse-skewed (U1343C 12H-3 55 cm) samples. Note that the x-axis is reversed to follow convention.

Sediments range from poorly to very poorly sorted (Fig. 3.6), which makes sense, given the broad range of particle sizes. Typically, bulk sediments are better sorted, and more symmetrical, than terrigenous, especially at Site U1345 (Table 3.3; Fig 3.6), perhaps because the 59 biogenic particles included in bulk samples have a smaller range of sizes (typically silt to fine sand). At all three sites, terrigenous sediments can mostly be described as very poorly sorted

(Table 3.3; Fig. 3.6). Overall, terrigenous sediments have a more symmetrical distribution at sites U1345 (60%) and U1343 (75%), compared to U1339, where only 51% of the samples have a symmetrical or near-symmetrical distribution. Average skewness values (0.11) for Site U1339 indicate that terrigenous sediments at this site are more coarse-skewed (Table 3.3; Fig. 3.6). As not all samples are symmetrical, there is some variation between measures of central tendency

(mean, median, mode), although mean and median grain size plots display broadly similar trends (Fig. 3.6).

Discussion

Interpreting Grain Size Parameters

Downcore variations in grain size distribution can be used to infer changes in the sedimentary environment (e.g., McLaren, 1981; Aiello & Ravelo, 2012; Wang et al., 2015;

Warrier et al., 2016; Vaughn & Caissie, 2017; Pelto et al., 2018; Stow & Smillie, 2020). In this section, we test whether grain size is effective as a proxy for siliciclastic input, sediment transport, ice rafting, paleocurrent strength, and paleoproductivity in the Bering Sea.

Proxy for Sediment Transport and Deposition

Deposits on the Bering shelf are derived from various continental sources, and transported to the shelf by several agents, including wind, rivers, and ice. Accordingly, shelf sediments comprise a broad range of size classes, from clay, to coarse sand, gravel (2-64 mm), and infrequently, boulders (>64 mm) (Table 3.1). Mean grain size on the Bering shelf is a 60 function of the energy of the depositional environment, and generally decreases with increasing depth and distance from the shore. Interestingly, the Bering shelf is one of the few continental shelves where sediment is graded and in equilibrium with the present-day current regime (Sharma, 1972, 1975; Richwine et al., 2018).

Shelf sediments may be transported beyond the shelf-slope break by various processes, including bottom currents, downslope transport (e.g., debris flows, turbidity flows), and ice rafting. The characteristics of these slope deposits are inherited from the source material, but may be somewhat altered by the processes occurring during transport and deposition

(McLaren, 1981). Sediments from our study sites on the Bering slope contain a mix of different size classes, and have a multimodal distribution, reflecting the diverse sediment inputs to the shelf. In addition, the sediments are poorly to very poorly sorted. The grain size distribution suggests that icebergs and/or sea ice are likely transport agents, because ice-rafted debris (IRD) typically consists of poorly sorted sediment with a wide range of grain sizes (von Huene et al.,

1973; Krissek et al., 1985; Reimnitz et al., 1998; Lisitzin, 2002; Sakamoto et al., 2005). In addition, we cannot ignore the role of current action, which may transport finer grains as suspended load, and coarser grains via bed load traction across the seafloor (Stow & Smillie,

2020), although it should be noted that bottom currents rarely transport particles >63 µm

(Masson et al., 2004; McCave & Hall, 2006). Turbidity currents can transport much coarser material, and in fact, are one of the most important ways by which fine-, medium-, and coarse- grained materials are transported from the shelf into deeper water (Stow & Smillie, 2020).

However, the characteristic features of turbidite deposits (e.g., moderate to good sorting, 61 normal grading, abrupt changes in grain size, and clear erosive surfaces), are not apparent at any of our core sites.

The skewness of a deposit reflects the ability of the transport agent to selectively remove finer or coarser material. A symmetrical or near-symmetrical distribution may indicate a low-energy environment, whereas high skewness values may indicate a large amount of sediment reworking (Cadigan, 1961). Sediments may become fine-skewed due to the removal of fines by winnowing, or through selective deposition of grains in transport. Alternatively, they may become coarse-skewed, due to total deposition of the transported sediment (McLaren,

1981; Martins, 2003). In general, Bering Sea sediments are either near symmetrical, or slightly coarse-skewed (Fig. 3.6), suggesting a relatively low energy depositional environment. At sites

U1345 and U1343, however, there is an interval from ~408-400 ka where sediments become more fine-skewed (Fig. 3.6), which may reflect a change in energy conditions at this time.

Several factors complicate the use of skewness as a proxy for sediment transport processes. Firstly, skewness values often reflect the grain size characteristics of the source material, rather than the energy of the transport medium (Andrews & van der Lingen, 1969;

McLaren, 1981). Secondly, skewness can result from subequal mixing of different grain size populations in a multimodal distribution (Folk & Ward, 1957). Lastly, skewness is a metric best suited to well sorted, unimodal sediments; indeed, the primary application of skewness in grain size studies seems to be in distinguishing between adjacent deposits of well sorted sediments, such as dune sands and beach sands (e.g., Friedman, 1961, 1979; Sevon, 1965; McLaren, 1981;

Martins, 2003; Kaspar-Zubillaga & Carranza-Edwards, 2005; Lopez et al., 2020). Cadigan (1961) showed that asymmetry in poorly sorted sediments produces much lower skewness values than 62 in well-sorted sediments. As sediments from the Bering Sea are both poorly sorted and multimodal, skewness may have a limited application in this study

Distinguishing Sediment Facies

Downcore variations in grain size may be linked to a range of processes that operate in deep water to erode, transport and deposit sediments, including gravity-driven, current-driven, and vertical settling processes, each of which produce a distinctive type of deposit or sediment facies. In general, sediments at the three core sites fit the definition of a hemipelagic facies.

Hemipelagites are fine-grained (mean 5-35 µm), poorly sorted deposits that comprise a mix of biogenic (>10%) and terrigenous and/or volcanogenic material (>10%), in which at least 40% of the terrigenous fraction is silt-sized or larger. Hemipelagic sediments are further characterized by a wide range of grain size classes, and typically have a multimodal distribution (Stow &

Smillie, 2020). Hemipelagites are typical of outer shelf and slope settings and are deposited by a combination of vertical settling and very slow lateral advection in a low energy environment

(Stow et al., 1998). Hemipelagic deposition is a continuous process that occurs under more or less steady state conditions, although sedimentation rates may vary in response to changes in biogenic and terrigenous inputs. As such, undisturbed (hemi)pelagic deposits are ideal candidates for paleoceanographic studies.

At sites U1345 and U1343, there is a coarse interval centered on 403.5 ka (Fig. 3.6, 3.8).

Sediments from this interval do not match the signature of hemipelagic deposits, and thus indicate a change in sediment facies. Based on grain size distribution, and sedimentary structures, we suggest that this anomalous interval may be a contourite, a sediment deposited or reworked by the persistent action of bottom currents. Specifically, sediments from this 63 interval fit the description of a fine-grained, muddy or silty contourite (Fig. 3.9; Rebesco et al.,

2014; Stow & Smillie, 2020). Fine-grained contourites are a mix of biogenic and terrigenous sediment, characterized by poor sorting, unimodal distribution, and a bi-gradational unit in the range of 0.5 to 3 m thick (Rebesco et al., 2014; Stow & Smillie, 2020). The ‘coarse interval’ at sites U1345 and U1343 is in fact a ~2 m thick bi-gradational sequence, with a coarsening upward unit (~408 to 403.5 ka) followed by a fining upward unit (403.5 to ~400 ka) (Fig. 3.8), much like the idealized mud-sand contourite depicted in Figure 3.9 (Rebesco et al., 2014).

Evidence for a contourite is particularly strong at U1345; samples from the coarsest part of the deposit at this site represent the only unimodal grain size distribution in the entire record. The coarsest deposits at U1343 are not unimodal, likely due to input of poorly sorted IRD during this interval (Fig. 3.8). Irregular sandy lenses, sand mottles, sand layers, and indistinct dark laminations occur both before and after the bi-gradational sequence at both sites (Fig. 3.8).

Typically, lenses of coarser material are one of the few sedimentary structures preserved in contourite facies, due to the pervasive bioturbation associated with contourites (Stow et al.,

2002; Stow & Faugères, 2008; Stow & Smillie, 2020).

Historically, contourite processes have been linked to bottom current dynamics, as highlighted in Figure 3.9 (Gonthier et al., 1984), however, sediment supply is also a major control on the formation of contourite facies (e.g., Michels et al., 2001; Mulder et al., 2013;

Rebesco et al., 2014). For example, Mulder et al. (2013) showed that the same facies succession can be derived in one of two ways: either by the selective removal of fine-grained material through winnowing, or from a new supply of coarser grains, which could include material transported from upstream contourite drifts, as well as more remote sources, such as rivers. At 64 high latitudes, the process of ice rafting may provide a further supply of coarse, terrigenous particles. As sediments at sites U1345 and U1343 become more fine-skewed during the contourite, we propose that winnowing is the main control on the development of contourite facies on the Bering slope.

Figure 3.8. Grain size distribution (mean grain size and sorting) for the contouritic intervals at sites U1345 and U1343. The core image shows an example of a sand lens from U1343.

Although evidence for contourite deposits at sites U1345 and U1343 is strong, it is somewhat unusual to find a single contourite facies. Most contourites are formed by the action of bottom currents, which are strongly influenced by thermohaline and wind-induced circulation; as a result, contourite sequences are believed to reflect regular or sub-regular periodicities in mean bottom current velocities, driven by larger-scale changes in oceanic and 65 atmospheric circulation, climate, and sea level (Stow et al., 2002; Stow & Smillie, 2020). Stow et al. (2002) calculated periodicities for contourite drifts of terrigenous to mixed composition in various locations, including the Gulf of Cadiz, Rockall Margin, West Shetlands, and Norwegian

Margin, and obtained periodicities between 5,000 and 20,000 years. In addition, periodicities of

20,000-40,000 years have been observed in bioclastic contourites, analogous to the

Milankovitch cyclicity recognized in many pelagic and hemipelagic deposits, whilst cycles between 20 and 200 ka have been estimated for ancient contourite systems (Hüneke & Stow,

2008; Stow & Faugères, 2008). It is possible that contourites on the Bering slope have longer periodicities, such as the 100 ka cycle, but we are unable to resolve such periodicities in a

~60,000 year record. Alternatively, bottom currents can also operate as part of upwelling and downwelling systems, up- and down-canyon currents, internal tides and waves, and seafloor polishing and spillover. Such processes may produce isolated examples of contourite facies

(Faugères & Stow, 2008; Hüneke, 2016).

The presence of contourite sequences raises some concerns about the validity of the existing age models, given that we do not know exactly how the processes responsible for contourite formation might influence sedimentation rates, and the input of older, reworked sediment to the core sites. However, it has been shown that deposition in contourite drifts is mostly continuous, thus enabling good age control (e.g., Voelker et al., 2006; Toucanne et al.,

2007; Mulder et al., 2013). There is no evidence for significantly older, reworked material in our deposits; oxygen isotope values of benthic foraminifera from within the contourites are consistent with those from the hemipelagic deposits. Although sedimentation rates may be 66 more variable within the contourite sequences, small-scale variations in sedimentation rates should not affect our interpretation of the core records on orbital or millennial timescales.

Proxy for Ice Rafting

Ice rafting is the process by which glaciers and/or sea ice entrain and transport terrigenous and near-shore particles far from land. As the ice melts, it releases the entrained sediment, which settles through the water column, and accumulates on the seafloor

(Dowdeswell, 2009). Ice rafting is an important mode of sediment transport at high latitudes

(Ruddiman, 1977; Gilbert, 1990; Andrews, 2000; Kennish, 2002; Dowdeswell, 2009), and results 67 in deposits of poorly sorted, coarse-grained siliciclastic material, known as ice-rafted debris

(IRD). In reality, the process of ice rafting can transport a large range of particle sizes, from fine- grained clays, to coarse sands, gravel, and dropstones (e.g., von Huene et al., 1973; Krissek et al., 1985; Reimnitz et al., 1998; Lisitzin, 2002; Sakamoto et al., 2005), but we typically examine the coarse fraction in IRD studies, because ice rafting is one of the only mechanisms capable of transporting coarse (sand-sized and larger) grains into the open ocean. Other possible mechanisms include turbidity currents, and volcanic activity. We see no evidence for turbidites at any of the core sites, but volcanogenic material, mostly in the form of fine ash, has been identified from visual core descriptions, core images, and smear slide analyses at all three sites

(Fig. 3.6, 3.10; Takahashi et al., 2011). Ash is most abundant at Site U1339, due to its proximity to the main volcanic source of the Aleutian Arc, and is a minor component of the sediments at

U1345 and U1343 (Fig. 3.6, 3.10; Takahashi et al., 2011). However, there is no consistent relationship between the volume % of coarse terrigenous grains and the presence of ash layers or accessories (Fig. 3.10), which suggests that volcanogenic material is not a major control on grain size. Considering that other transport mechanisms can mostly be ruled out, we feel confident interpreting the bulk of the coarse terrigenous grains as IRD.

IRD is defined as the weight percent, volume percent, or number of terrigenous grains within a given size fraction (Dowdeswell, 2009). Here, we examine the volume percent of particles in the >150 µm and >250 µm size fractions (Fig. 3.10). Regardless of which measure we analyze, IRD is present at all sites throughout almost the entire record. The volume % of grains

>150 µm averages 6.3%, 5.5%, and 5.4% at sites U1345, U1343 and U1339, respectively; for the

>250 µm size fraction, average values are 4.4%, 3.4%, and 3.6%, although there is considerable 68 downcore variability at all sites (Fig. 3.10). The highest volume % of coarse grains at sites U1345 and U1343 falls within the interval that we define as a contourite. In particular, at site U1343, the volume % of grains >250 µm is notably high (22.1%) at 403.39 ka (Fig. 3.10). During this interval, we are unable to distinguish between IRD and coarse contourite deposits, based on grain size alone. However, smear slide analysis of the coarsest sample from Site U1343 revealed the presence of very coarse siliciclastic grains, interpreted as IRD (Fig. 3.11). In addition, the grain size distribution of this sample is very poorly sorted and resembles the grain size distribution of sediments from the Sea of Okhotsk that were deposited under melting sea ice

(Fig. 3.11; Sakamoto et al., 2005). Ice-rafted material is a passive input into fine-grained contourite deposits, and is not subsequently reworked to any great extent by bottom currents

(Rebesco et al., 2014), therefore, ice rafting may introduce coarser sediments into a fine- grained contourite deposit, independent of the processes acting to form the contourite facies

(e.g., Howe et al., 2007; Lucchi & Rebesco, 2007).

Overall, the IRD records presented here suggest that iceberg and/or sea ice rafting was an important source of sediment delivery to the Bering slope throughout MIS 11. Distinguishing between iceberg and sea ice rafting processes is critical to understanding the ice rafting history of an area, but it can be hard to distinguish iceberg-rafted debris (IRD) from sea ice-rafted debris (SIRD), based on grain size alone (St John et al., 2015). Although iceberg-rafted material is often considered to be coarser than that transported by sea ice, under certain conditions, both icebergs and sea ice are capable of transporting a comparable fraction of coarse terrigenous material far offshore (e.g., Nürnberg et al., 1994; Lisitzin, 2002; Darby &

Zimmerman, 2008; Darby et al., 2011; Mager et al., 2013; St John et al., 2015). However, it 69 should be noted that tidewater glaciers are uncommon in the Bering Sea, especially during interglacial periods (Caissie et al., 2016), so it is likely that most of the IRD in our records was transported by sea ice.

Figure 3.11. Grain size distribution plot and light microscope image of IRD (magnification 100x) from Site U1343 (12H-5 41-42 cm, 403.4 ka). 70

Proxy for Paleocurrent Strength

The mean size of sortable silt (10-63 μm) is used as a proxy for the flow speed of near- bottom currents in the deep ocean, with coarser deposits reflecting higher flow speeds.

(McCave et al., 1995, 2017). This particular size fraction is used, because at the upper limit, ocean currents rarely transport grains larger than 63 μm (Masson, 2004; McCave and Hall,

2006), and at the lower limit, particles less than 10 μm tend to behave cohesively (McCave et al., 1995; McCave & Hall, 2006). To test whether sortable silt is an effective proxy for current strength at our core sites, we examined downcore variations in the mean size of sortable silt

(SS) (Fig. 3.12), as well as the relationship between SS and the percent sortable silt (SS%). Under 71 a current-sorted regime, SS is positively correlated with SS%, whereas unsorted sediments show no correlation (Roberts et al., 2017; McCave & Andrews, 2019).

The mean size of sortable silt (calculated for the terrigenous fraction) decreases from north to south, from an average of 28.1 μm at Site U1345, to 26.6 μm at U1343, and 25.5 μm at

U1339. At all sites, SS increases following deglaciation, in line with sea level rise and flooding of the Bering Shelf (Fig. 3.12). At U1339, SS stays relatively consistent throughout MIS 11, but at sites U1345 and U1343, sortable silt size is more variable, with consistently higher values (>30

μm) from ~407 to 401 ka (Fig. 3.12). This increase in SS corresponds to the period of contourite formation, during which time we might reasonably expect to see an increase in current sorting.

However, our records suggest the opposite. There is little to no positive correlation between SS and SS% at any of our core sites, and in fact, there is a strong negative correlation (R2 = -0.87) 72 during the contouritic interval at Site U1343 (Fig. 3.13), likely due to contamination by unsorted

IRD. Several studies (e.g., Jonkers et al., 2015; Wu et al., 2018) have suggested that the sortable silt proxy is not effective in regions with IRD input, because the input of ice-rafted silt may complicate the interpretation of paleocurrent strength. McCave and Andrews (2019) cautioned that the sortable silt proxy should only be used to infer current strength if it can be proven, based on a positive correlation between SS and SS%, that the sediment is indeed sorted by currents. As our records indicate a lack of current sorting (Fig. 3.13), we are unable to use sortable silt as a proxy for paleocurrent strength in the Bering Sea.

Figure 3.14. Percent diatoms from smear slide analysis versus mean grain size before (closed circles) and after (open circles) biosilica digestion for sites U1345 (black) and U1339 (red).

Proxy for Paleoproductivity

Our records show that silt is the predominant size class in Bering Sea sediments, and that the volume % of silt-sized grains is greater for bulk than terrigenous samples (Table 3.3; 73

Fig. 3.5). This suggests that the grain size distribution of bulk sediments may be influenced by the abundance of silt-sized biogenic material. In the Bering Sea, diatom frustules are the main component of biogenic sediments, and most diatom species in our samples fall within the silt- sized fraction. Therefore, biosilica digestion has the effect of reducing the volume % of particles in the silt size class, which in turn can affect median/mean grain size. Aiello & Ravelo (2012) showed that ~40% of variability in the mean grain size of Bering Sea sediments can be explained by variations in the abundance and preservation of diatom valves, and to a lesser extent, sponge spicules, and suggested that mean grain size provides a rough indication of trends in diatom productivity. With this in mind, we examined the relationship between diatom content

(referring to the relative % abundance of diatom valves in comparison to other sediment components) and mean grain size in a subset of samples from sites U1345 and U1339 (Fig.

3.14). At Site U1339, diatom content explains ~23% of variability in the mean grain size of bulk sediments, whereas at Site U1345, there is virtually no correlation between diatom content and mean grain size (Fig. 3.14). We postulate that differences between the two sites are linked to differences in sediment composition; sediments at U1339 have a higher diatom content, whilst biogenic material at Site U1345 is heavily diluted by siliciclastic input, so that diatom content has less of an influence on grain size (Aiello & Ravelo, 2012; Kanematsu et al., 2013). Our findings differ from those of Aiello and Ravelo (2012), in part because their analyses included sediments from Bowers Ridge in the southern Bering Sea, where sediments are less diluted by terrigenous matter, but not sediments from U1345. It is also possible that diatom content has more of an influence on grain size over glacial-interglacial timescales, and less so at the finer resolution of this study. 74

Based on the relationship between diatom content and mean grain size (Fig. 3.14), it is clear that diatom content is not a major control on bulk grain size, especially at Site U1345.

How, then, do we account for the differences in mean grain size between bulk and terrigenous samples shown in figure 3.15? Smear slide analyses show that there are additional biogenic controls on mean grain size, including the preservation of diatom valves, and the size of diatoms within the assemblage (Fig. 3.15, 3.16). If the diatom valves are poorly preserved, or fragmented, bulk mean grain size is likely to be smaller, because the sample is dominated by clay and/or fine silt-sized diatom fragments, as opposed to larger silt-sized and even sand-sized whole diatom valves. Accordingly, sediment will become coarser (mean grain size increases) when the finer-grained biogenic material is dissolved (Fig. 3.15). Conversely, if the diatoms are well-preserved, sediment may become finer after removing biogenic components in the coarse silt range (Fig. 3.15). For similar reasons, the size of diatom valves in the assemblage can also influence the mean grain size of bulk sediments (Fig. 3.16). For example, an assemblage dominated by small pennate diatoms, such as Neodenticula seminae, is likely to be finer- grained than a sediment containing large, centric diatoms, such as Coscinodiscus species, which can be up to ~200 µm or more in diameter (Sancetta, 1987). Of course, the degree of coarsening or fining also depends on the grain size distribution of siliciclastic material. In sediments with sub-equal proportions of well-preserved, silt-sized diatoms and silt-sized siliciclastics, for example, we might not expect much difference in mean grain size between the bulk and terrigenous samples. However, if the sediment were comprised of well-preserved, silt- sized diatoms, and abundant silt- to sand-sized siliciclastic material, the mean grain size of the terrigenous fraction is likely to be coarser than that of the bulk sediment. 75

Aiello & Ravelo (2012) discussed the importance of diatom preservation on the mean grain size of bulk sediments at several sites across the Bering Sea. They found that the mean grain size of bulk sediments was coarser during the Holocene and the Last Interglacial, due to 76 relatively high abundances of well-preserved, whole, silt-sized diatom valves. In contrast, samples from the Last Glacial Maximum contained poorly preserved, fragmented diatoms, and abundant clay-sized particles. Likewise, our smear slide analyses showed more fragmented diatoms, corresponding to a slightly lower mean grain size, during the glacial periods (MIS 12 and 10) within our study timeframe.

To summarize, mean grain size is not a direct or straightforward proxy for paleoproductivity, but at the very least, downcore variations in the mean grain size of bulk sediments can be used to infer changes in productivity that occur on glacial-interglacial cycles

(Aiello & Ravelo, 2012). In addition, when combined with smear slide analyses, grain size records may provide valuable information on changes in diatom abundance, preservation, and assemblage composition.

Figure 3.16. Light microscope image from Site U1339 (U1339C 11H-3 101 cm), highlighting variations in the size of diatom valves.

Summary

In this study, we aimed to show how grain size parameters, including the volume percent of grains in various size fractions, as well as the statistical measures of mean grain size, sorting, and skewness, can be used to infer past oceanographic conditions at three core sites in the Bering Sea. At all three sites, sediments are a poorly sorted, multimodal mix of siliciclastic and biogenic sediments, with minor volcanogenic input. The southernmost site, U1339, has more biogenic and volcanogenic grains, whilst Site U1345, located on the northern slope, has the most siliciclastic input. Sediments at Site U1343 represent an intermediate composition between the other two sites.

Sediments in the Bering Sea can primarily be classified as hemipelagites, making them ideal deposits for paleoenvironmental reconstructions. There is no evidence for turbidite deposition at any site; however, there is strong evidence for contourite facies deposited between ~408-400 ka at sites U1345 and U1343. Contourites are associated with sediment reworking through the action of bottom currents, and there is evidence for winnowing of finer sediments during this interval, which may complicate paleoclimate interpretations.

The prevalence of very coarse (>150 um) grains throughout all three records indicates that ice rafting is a consistent source of sediment transport to the Bering slope, but we are unable to distinguish between iceberg and sea ice rafting processes, based on grain size alone. In general, mean grain size can be used to infer productivity on glacial-interglacial timescales, however, both the size and the preservation of diatom valves also exert a control on mean grain size. Finally, the mean size of sortable silt is not a valid proxy for bottom current 78 strength in the Bering Sea, most likely because the input of ice-rafted silt confounds the sortable silt signal.

References

Aiello, I.W. & Ravelo, A.C., 2012. Evolution of marine sedimentation in the Bering Sea since the Pliocene. Geosphere, 8 (6), 1231-1253. https://doi.org/10.1130/GES00710.1

Andrews, J.T., 2000. Icebergs and Iceberg Rafted Detritus (IRD) in the North Atlantic: Facts and Assumptions. Oceanography, 13 (3), 100-108. https://doi.org/10.5670/oceanog.2000.19

Andrews, J.T. & Principato, S.M., 2002. Grain-size characteristics and provenance of ice- proximal glacial marine sediments. In: Dowdeswell, J.D. & Ò’Cofaigh, C. (Eds.), Glacier- influenced Sedimentation on High-latitude Continental Margins. Geol. Soc. London, Spec. Publ., 203, pp. 305-324.https://doi.org/10.1144/GSL.SP.2002.203.01.16

Andrews P.B. & van der Lingen, G.J., 1969. Environmentally significant sedimentologic characteristics of beach sands. New Zealand J. Geol. Geophys., 12 (1), 119-137. https://doi.org/10.1080/00288306.1969.10420230

Asahara, Y., Takeuchi, F., Nagashima, K., Harada, N., Yamamoto, K., Oguri, K., & Tadai, O., 2012. Provenance of terrigenous detritus of the surface sediments in the Bering and Chukchi Seas as derived from Sr and Nd isotopes: Implications for recent climate change in the Arctic regions. Deep Sea Res. Part II: Top. Stud. Oceanogr., 61-64, 155-171. https://doi.org/10.1016/j.dsr2.2011.12.004

Asahi, H., Kender, S., Ikehara, M., Sakamoto, T., Takahashi, K., Ravelo, A.C., Alvarez Zarikan, C.A., Khim, B.K., & Leng, M.J., 2016. Orbital-scale benthic foraminiferal oxygen isotope stratigraphy at the northern Bering Sea Slope Site U1343 (IODP Expedition 323) and its Pleistocene paleoceanographic significance. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 66-83. https://doi.org/10.1016/j.dsr2.2014.01.004

Ben-Avraham, Z. & Cooper, A.K., 1981. Early evolution of the Bering Sea collision of oceanic rises and North Pacific subduction zones. Geol. Soc. Am., Bull., 92 (7), 485-495. https://doi.org/10.1130/0016-7606(1981)92<485:EEOTBS>2.0.CO;2

Blott, S.J. & Pye, K., 2001. GRADISTAT: A Grain Size Distribution And Statistics Package For The Analysis Of Unconsolidated Sediments. Earth Surf. Process. Landforms, 26, 1237-1248. https://doi.org/10.1002/esp.261

Briceño Zuluaga, F., Sifeddine, A. Caquineau, S., Cardich, J., Salvatteci, R., Gutiérrez, D., Ortlieb, L., Velazco, F., Boucher, H., & Machado, C., 2016. Terrigenous material supply to the Peruvian central continental shelf (Pisco, 14° S) during the last 1000 years: Paleoclimatic implications. Clim. Past, 12, 787-798. https://doi.org/10.5194/cp-12-787-2016

Brown, Z.W. & Arrigo, K.R., 2012. Contrasting trends in sea ice and primary production in the Bering Sea and Arctic Ocean. J. Mar. Sci., 69, 1180-1193. https://doi.org/10.1093/icesjms/fss113

Cadigan, R.A., 1961. Geologic Interpretation of Grain-Size Distribution Measurements of Colorado Plateau Sedimentary Rocks. J. Geol., 69 (2), 121-144.

Carlin, J., Addison, J., Wagner, A., Schwartz, V., Hayward, J., & Severin, V., 2019. Variability in Shelf Sedimentation in Response to Fluvial Sediment Supply and Coastal Erosion Over the Past 1,000 Years in Monterey Bay, CA, United States. Front. Earth. Sci., 7, 113. https://doi.org/10.3389/feart.2019.00113

Cook, M., Keigwin, L., & Sancetta, C., 2005. The Deglacial history of surface and intermediate water of the Bering Sea. Deep Sea Res. Part II: Top. Stud. Oceanogr., 52 (16-18), 2163- 2173. https://doi.org/10.1016/j.dsr2.2005.07.004

Cook, M., Ravelo, A., Mix, A., Nesbitt, I, & Miller, N., 2016. Tracing subarctic Pacific water masses with benthic foraminiferal stable isotopes during the LGM and late Pleistocene. Deep Sea Res. Part II: Top. Stud. Oceanogr., 125-126, 84-95. https://doi.org/10.1016/j.dsr2.2016.02.006

Cronan, D.S., 1972. Skewness and Kurtosis in Polymodal Sediments from the Irish Sea. J. Sed. Res., 42 (1), 102-106. http://doi.org/10.1306/74D724A0-2B21-11D7-8648000102C1865D

Darby, D.A., Myers, W.B., Jakobsson, M., & Rigor, I., 2011. Modern dirty sea ice characteristics and sources: the role of anchor ice. J. Geophys. Res., 116 (C9), C09008. https://doi.org/10.1029/2010JC006675

Darby, D. A. & Zimmerman, P., 2008. Ice-rafted detritus events in the Arctic during the last glacial interval, and the timing of the Innuitian and Laurentide ice sheet calving events. Polar Res., 27 (2), 114-127. https://doi.org/10.1111/j.1751-8369.2008.00057.x deGelleke, L., Hill, P., Kienast, M., & Piper, D., 2013. Sediment dynamics during Heinrich event H1 inferred from grain size. Mar. Geol., 336, 160-169. https://doi.org/10.1016/j.margeo.2012.12.007

Dowdeswell J.A., 2009. Ice-Rafted Debris (IRD). In: Gornitz V. (Ed.), Encyclopedia of Paleoclimatology and Ancient Environments. https://doi.org/10.1007/978-1-4020-4411-3_113

Faugères, J.-C. & Stow, D.A.V., 2008. Contourite drifts: nature, evolution and controls. In: Rebesco, M. & Camerlenghi, A. (Eds.), Contourites: Developments in Sedimentology, 60. Elsevier, Amsterdam, pp. 257-288.

Folk, R.L. & Ward W.C., 1957. Brazos River bar: a study in the significance of grain size parameters. J. Sed. Petrol., 27 (1), 3-26. https://doi.org/10.1306/74D70646-2B21-11D7-8648000102C1865D

Friedman, G.M., 1961. Distinction between dune, beach, and river sands from their textural characteristics. J. Sed. Res., 31 (4), 514-529. https://doi.org/10.1306/74D70BCD-2B21-11D7-8648000102C1865D

Friedman, G.M., 1979. Differences in size distributions of populations of particles among sands of various origins. Sedimentology, 26, 3-32.

Friedman, G.M. & Sanders, J.E., 1978. Principles of Sedimentology. Wiley, New York, 792 pp.

Gilbert, R., 1990. Rafting in glacimarine environments. In: Dowdeswell, J.A. & Scourse, J.D. (Eds.), Glacimarine Environments: Processes and Sediments. Geol. Soc. London, Spec. Publ., 53, 105-120.

Gonthier, E., Faugères, J.-C., & Stow, D.A.V., 1984. Contourite facies of the Faro Drift, Gulf of Cadiz. In: Stow, D.A.V. & Piper, D.J.W. (Eds.), Fine Grained Sediments, Deep-water Processes and Facies. Geol. Soc. London, Spec. Publ., 15, pp. 275-291.

Hoffman, S.S., Dalsing, R.E., & Murphy, S.C., 2019. Sortable silt records of intermediate-depth circulation and sedimentation in the Southwest Labrador Sea since the Last Glacial Maximum. Quat. Sci. Rev., 206, 99-110. https://doi.org/10.1016/j.quascirev.2018.12.028 81

Holz, C., Stuut, J.-B., Henrich, R., & Meggers, H., 2007. Variability in terrigenous sedimentation processes off northwest Africa and its relation to climate changes: Inferences from grain-size distributions of a Holocene marine sediment record. Sediment. Geol., 202, 499-508. https://doi.org/10.1016/j.sedgeo.2007.03.015

Howe, J.A., Shimmield, T.M., Harland, R., & Eyles, N., 2007. Late Quaternary contourites and glaciomarine sedimentation in the Fram Strait. Sedimentology, 55 (1), 179-200. https://doi.org/10.1111/j.1365-3091.2007.00897.x

Hüneke, H., 2016. Contourites. In: Harff, J., Meschede, M., Petersen, S., & Thiede J. (Eds.) Encyclopedia of Marine Geosciences. https://doi.org/10.1007/978-94-007-6238-1_49

Hüneke, H. & Stow, D.A.V., 2008. Identification of Ancient Contourites: Problems and Palaeoceanographic Significance. In: Rebesco, M. & Camerlenghi, A. (Eds.), Contourites: Developments in Sedimentology, 60. Elsevier, Amsterdam, pp. 323-344.

Jonkers, L., Barker, S., Hall, I.R., & Prins, M.A., 2015. Correcting for the influence of ice-rafted detritus on grain size-based paleocurrent speed estimates. Paleoceanogr. Paleoclimatol., 30 (10), 1347-1357. https://doi.org/10.1002/2015PA002830

Kasper Zubillaga, J. & Carranza-Edwards, A., 2005. Grain size discrimination between sands of desert and coastal dunes from northwestern Mexico. Rev. Mex. Cienc. Geol, 22 (3), 383- 390.

Kim, S., Yoo, K.-C., Lee, Il, J., Lee, M.K., Kim, K., Yoon, H., & Moon, H., 2018. Relationship between magnetic susceptibility and sediment grain size since the last glacial period in the Southern Ocean off the northern Antarctic Peninsula – Linkages between the cryosphere and atmospheric circulation. Palaeogeogr. Palaeoclimatol. Palaeoecol., 505, 359-370. https://doi.org/10.1016/j.palaeo.2018.06.016

Krissek, L.A., Morley, J.J., & Lofland, D.K., 1985. The occurrence, abundance, and composition of ice-rafted debris in sediments from Deep Sea Drilling Project Sites 579 and 580, northwest Pacific. In Heath, G.R., Burckle, L.H., et al., Init. Repts. DSDP, 86. Washington (U.S. Govt. Printing Office), 647-655. 82

Lisiecki, L. E. & Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20, PA1003. https://doi.org/10.1029/2004PA001071

Lisitzin, A.P., 2002. Sea-Ice and Iceberg Sedimentation in the Ocean: Recent and Past. Springer, New York, 562 pp.

Lopez, O.M., Hegy, M.C., & Missime, T.M., 2020. Statistical comparisons of grain size characteristics, hydraulic conductivity, and porosity of barchan desert dunes to coastal dunes. Aeolian Res., 43, 100576. https://doi.org/10.1016/j.aeolia.2020.100576

Lucchi, R.G. & Rebesco, M., 2007. Glacial contourites on the Antarctic Peninsula margin: insight for palaeoenvironmental and palaeoclimatic conditions. In: Viana, A.R. & Rebesco, A.M. (Eds.), Economic and Palaeoceanographic Significance of Contourite Deposits. Geol. Soc. London., Spec. Publ., 276, 111-128.

Mager, S.M., Smith, I.J., Kempema, E.W., Thomson, B.J., & Leonard, G.H., 2013. Anchor ice in polar oceans. Progr. Phys. Geogr., 37 (4), 468-483. https://doi.org/10.1177/0309133313479815

Mahowald, N., Baker, A., Bergametti, G., Brooks, N., Duce, R., Jickells, T., Kubilay, N., Prospero, J., & Tegen, I., 2005. The atmospheric global dust cycle and iron inputs to the ocean. Global Biogeochem. Cycles, 19 (4), GB4025. https://doi.org/10.1029/2004GB002402

Martins, L.R., 2003. Recent Sediments and Grain-Size Analysis. Gravel, 1, 90-105.

Martins, L. & Barboza, E., 2005. Sand-Gravel Marine Deposits and Grain-Size Properties. Gravel, 3, 59-70.

Masson, D.G., Wynn, R.B., & Bett, B.J., 2004. Sedimentary environment of the Faroe-Shetland and Faroe Bank Channels, north-east Atlantic, and the use of bedforms as indicators of bottom current velocity in the deep ocean. Sedimentology, 51 (6), 1207-1241. https://doi.org/10.1111/j.1365-3091.2004.00668.x

McCave, I.N. & Andrews, J.T., 2019. Distinguishing current effects in sediments delivered to the ocean by ice. I. Principles, methods and examples. Quat. Sci. Rev., 212, 92-107. https://doi.org/10.1016/j.quascirev.2019.03.031

McCave, I.N. & Hall, I.R., 2006. Size sorting in marine muds: processes, pitfalls and prospects for paleoflow-speed proxies. Geochem. Geophys. Geosyst., 7 (10), Q10N05. https://doi.org/10.1029/2006GC001284

McCave, I.N., Manighetti, B., & Robinson, S.G., 1995. Sortable silt and fine sediment size/composition slicing: parameters for palaeocurrent speed and palaeoceanography. Paleoceanography, 10, 593-610. https://doi.org/10.1016/j.dsr.2017.07.003

McCave, I.N., Thornalley, D.J.R., & Hall, I.R., 2017. Relation of sortable silt grain-size to deep-sea current speeds: Calibration of the ‘Mud Current Meter’. Deep Sea Res. Part I: Oceanogr. Res. Pap., 127, 1-12. https://doi.org/10.1016/j.dsr.2017.07.003

McLaren, P., 1981. An interpretation of trends in grain size measures. J. Sed. Res., 51 (2), 611- 624. https://doi.org/10.1306/212F7CF2-2B24-11D7-8648000102C1865D

McManus, J., 1988. Grain size determination and interpretation. In: Tucker, M. (Ed.), Techniques in Sedimentology. Blackwell Scientific Publications, Palo Alto, CA, pp. 63-85.

Michels, K.H., Rogenhagen, J., & Kuhn, G, 2001. Recognition of contour-current influence in mixed contourite-turbidite sequences of the western Weddell Sea, Antarctica. Mar. Geophys. Res., 22, 465-485. https://doi.org/10.1023/A:1016303817273

Muhong, C., Fan, Z., Jun, L., Shangbin, X., Wen, Y., Zhong, C., Rong, X., Gangjian, W., & Lanlan, Z., 2005. Original component of grain size index in core sediment from southwestern slope of the South China Sea and its paleoenvironmental implication. Chin. Sci. Bull., 50, 896-902. https://doi.org/10.1007/BF02897384

Mulder, T., Hassan, R., Ducassou, E., Zaragosi, S., Gonthier, E., Hanquiez, V., Marchès, E., & Toucanne, S., 2013. Contourites in the Gulf of Cadiz: a cautionary note on potentially ambiguous indicators of bottom current velocity. Geo-Mar. Lett., 33, 357-367. https://doi.org/10.1007/s00367-013-0332-4

Nagashima, K., Asahara, Y., Takeuchi, F., Harada, N., & Tada, R., 2012. Contribution of detrital materials from the Yukon River to the continental shelf sediments of the Bering Sea based on the ESR intensity and crystallinity of quartz. Deep Sea Res. Part II: Top. Stud. Oceanogr., 61-64, 145-154. https://doi.org/10.1016/j.dsr2.2011.12.001

Naidu, A.S. & Mowatt, T.C., 1983. Sources and dispersal patterns of clay minerals in surface sediments from the continental shelf areas off Alaska. Geol. Soc. Am., Bull., 94 (7), 841- 854. https://doi.org/10.1130/0016-7606(1983)94%3C841:SADPOC%3E2.0.CO;2 84

Nesterovich, A. & Caissie, B.E., in review. A diatom-based quantitative sea ice proxy for the Bering and Chukchi Seas. Prog. Oceanogr.

Nürnberg, D., Wollenberg, I., Dethleff, D., Eicken, H., Kassens, H., Letzig, T., Reimnitz, E., & Thiede, J., 1994. Sediments in Arctic sea ice: Implications for entrainment, transport and release. Mar. Geol., 119 (3-4), 185-214. https://doi.org/10.1016/0025-3227(94)90181-3

O'Regan, M., Sellén, E., & Jakobsson, M., 2014. Middle to late Quaternary grain size variations and sea-ice rafting on the Lomonosov Ridge. Polar Res., 33 (1), 23672. https://doi.org/10.3402/polar.v33.23672

Poppe, L.J., Williams, S.J., & Paskevich, V.F., 2006. Textural analysis of marine sediments at the USGS Woods Hole Science Center: methodology and data on DVD. Proceedings of the Joint Eighth Federal Interagency Sedimentation Conference, Reno, Nev., April 2-6, 2006, pp. 905-911.

Rebesco, M., Hernández-Molina, F.J., Van Rooij, D., & Wahlin, A., 2014. Contourites and associated sediments controlled by deep-water circulation processes: State-of-the-art and future considerations. Mar. Geol., 352, 111-154. https://doi.org/10.1016/j.margeo.2014.03.011

Reimnitz, E., McCormick, M., Bischof J., & Darby, D., 1998. Comparing sea-ice sediment load with Beaufort Sea shelf deposits: Is entrainment selective? J. Sed. Res., 68, 777-787. https://doi.org/10.2110/jsr.68.777

Richwine, K.A., Smith, K.R., & McConnaughey, R.A., 2018. Surficial sediments of the eastern Bering Sea continental shelf: EBSSED-2 database documentation. U.S. Dep. Commer., NOAA Tech. Memo., NMFS-AFSC-377, 48 pp.

Riethdorf, J.R., Nürnberg, D., Max, L., Tiedemann, R., Gorbarenko, S.A., & Malakhov, M.I., 2013. Millennial-scale variability of marine productivity and terrigenous matter supply in the western Bering Sea over the past 180 kyr. Clim. Past, 9, 1345-1373. https://doi.org/10.5194/cp-9-1345-2013

Rothwell, R.G. & Rack., F.R., 2006. New Techniques in Sediment Core Analysis: an introduction. In: Rothwell, R.G. (Ed)., New Techniques in Sediment Core Analysis. Geol. Soc. London, Spec. Publ., 267, 1-29. https://doi.org/10.1144/GSL.SP.2006.267.01.01

Ruddiman, W.F., 1977. Late Quaternary deposition of ice-rafted sand in the subpolar North Atlantic (lat 40° to 65°N). Geol. Soc. Am., Bull., 88, 1813-1827. https://doi.org/10.1130/0016-7606(1977)88%3C1813:LQDOIS%3E2.0.CO;2

Sakamoto, T., Ikehara, M., Aoki, K., Iijima, K., Kimura, N., Nakatsuka, T., & Wakatsuchi, M., 2005. Ice-rafted debris (IRD)-based sea-ice expansion events during the past 100 kyrs in the Okhotsk Sea. Deep Sea Res. Part II: Top. Stud. Oceanogr., 52 (16-18), 2275-2301. https://doi.org/10.1016/j.dsr2.2005.08.007

Shaffer, G. & Lambert, F., 2018. In and out of glacial extremes by way of dust−climate feedbacks. Proc. Natl. Acad. Sci. USA, 115 (9), 2026-2031. https://doi.org/10.1073/pnas.1708174115

Schlee, J., 1973. Atlantic Continental Shelf and Slope of the United States: Sediment Texture of the Northeastern Part. U.S. Geological Survey Professional Paper 529-L, 64 pp.

Serno, S., Winckler, G., Anderson, R., Hayes, C., McGee, D., Machalett, B., Ren, H., Straub, S., Gersonde, R., & Haug, G., 2014. Eolian dust input to the Subarctic North Pacific. Earth Planet. Sci. Lett., 387, 252-263. https://doi.org/10.1016/j.epsl.2013.11.008

Sevon, W.D., 1966. Distinction of New Zealand beach, dune, and river sands by their grain size distribution characteristics. New Zealand J. Geol. Geophys., 9 (3), 212-223. https://doi.org/10.1080/00288306.1966.10422810

Sharma, G.D., 1975. Contemporary epicontinental sedimentation and shelf grading in the southeast Bering Sea. In: Forbes, R.B (Ed.), Contributions to the Geology of the Bering Sea Basin and Adjacent Regions. Geol. Soc. Am., Spec. Pap., 151, 33-48.

Springer, A., Mcroy, C., & Flint, M., 1996. The Bering Sea Green Belt: Shelf-edge processes and ecosystem production. Fish. Oceanogr., 5 (3-4), 205-223. https://doi.org/10.1111/j.1365-2419.1996.tb00118.x 86

St John, K., Passchier, S., Tantillo, B., Darby, D., & Kearns, L., 2015. Microfeatures of modern sea-ice-rafted sediment and implications for paleo-sea-ice reconstructions. Ann. Glaciol., 56 (69), 83-93. https://doi.org/10.3189/2015AoG69A586

Stabeno, P.J., Thoman, R.L., & and Wood, K., 2019. Recent Warming in the Bering Sea and Its Impact on the Ecosystem. In: Richter-Menge, J., Druckenmiller, M.L., & Jeffries, M. (Eds.), Arctic Report Card 2019. http://www.arctic.noaa.gov/Report-Card.

Stow, D.A.V. & Faugères, J.C., 2008. Contourite facies and the facies model. In: Rebesco, M. & Camerlenghi, A. (Eds.), Contourites: Developments in Sedimentology, 60. Elsevier, Amsterdam, pp. 223-256.

Stow, D.A.V., Faugères, J.-C., Howe, J.A., Pudsey, C.J., & Viana, A.R., 2002. Bottom currents, contourites and deep-sea sediment drifts: Current state-of-the-art. In: Stow, D.A.V., Pudsey, C.J., Howe, J.A., Faugères, J.-C., & Viana, A.R (Eds.), Deep-Water Contourite Systems: Modern Drifts and Ancient Series, Seismic and Sedimentary Characteristics. Geol. Soc. London, Memoirs, 22, pp. 7-20.

Stow, D. & Smillie, Z., 2020. Distinguishing between Deep-Water Sediment Facies: Turbidites, Contourites and Hemipelagites. Geosciences, 10 (2), 68. https://doi.org/10.3390/geosciences10020068

Stow, D.A.V. & Tabrez, A.R., 1998. Hemipelagites: processes, facies and model. In: Stoker, M.S., Evans, D., & Cramp, A. (Eds.), Geological Processes on Continental Margins: Sedimentation, Mass Wasting and Stability. Geol. Soc. London., Spec. Publ., 129, 317- 337.

Stuut, J.-B.W., Temmesfeld, F., & De Deckker, 2014. A 550 ka record of aeolian activity near North West Cape, Australia: inferences from grain-size distributions and bulk chemistry of SE Indian Ocean deep-sea sediments. Quat. Sci. Rev., 83, 83-94. https://doi.org/10.1016/j.quascirev.2013.11.003

Sval’nov, V.N. & Alekseeva, T.N., 2006. Characteristics of the grain-size composition of deep- water oceanic sediments. Lithol. Miner. Resour., 41 (3), 201-214. https://doi.org/10.1134/S0024490206030011

Takahashi, K., Ravelo, A.C., Alvarez Zarikian, C. A., and Expedition 323 Scientists, 2011. Proceedings of the Integrated Ocean Drilling Program, 323, Tokyo. doi:10.2204/iodp.proc.323.2011

Tegez, A.D., Jansen, E., & Telford, R.J., 2014. The role of the northward-directed (sub)surface limb of the Atlantic Meridional Overturning Circulation during the 8.2 ka event. Clim. Past, 10, 1887-1904. https://doi.org/10.5194/cp-10-1887-2014

Toucanne, S., Mulder, T., Schönfeld, J., Hanquiez, V., Gonthier, E., Duprat, J., Cremer, M., & Zaragosi, S., 2007. Contourites of the Gulf of Cadiz: a high-resolution record of the paleocirculation of the Mediterranean outflow water during the last 50,000 years. Palaeogeogr. Palaeoclimatol. Palaeoecol., 246, 354-366. https://doi.org/10.1016/j.palaeo.2006.10.007

Udden, J.A., 1914. Mechanical composition of clastic sediments. Geol. Soc. Am., Bull., 25 (1), 655-744. https://doi.org/10.1130/GSAB-25-655

Vaughn, D.R. & Caissie, B.E., 2017. Effects of sea-level, sea-ice extent, and nutrient availability on primary production at the Umnak Plateau, Bering Sea (IODP Site U1339) during Marine Isotope Stage (MIS) 5. Palaeogeogr. Palaeoclimatol. Palaeoecol., 485, 283-292. https://doi.org/10.1016/j.palaeo.2017.06.020

Voelker, A.H.L., Lebreiro, S.M., Schonfeld, J., Cacho, I., Erlenkeuser, H., & Abrantes, F., 2006. Mediterranean outflow strengthening during northern hemisphere coolings: a salt source for the glacial Atlantic? Earth Planet. Sci. Lett., 245 (1-2), 39-55. https://doi.org/10.1016/j.epsl.2006.03.014

Voigt, I., Chiessi, C.M., Piola, A.R., & Heinrich, R., 2016. Holocene changes in Antarctic Intermediate Water flow strength in the Southwest Atlantic. Palaeogeogr. Palaeoclimatol. Palaeoecol., 463, 60-67. https://doi.org/10.1016/j.palaeo.2016.09.018 von Huene, R., Larson, E., and Crouch, J., 1973. Preliminary study of ice rafted erratics as indicators of glacial advances in the Gulf of Alaska. In Kulm, L.D., von Huene, R., et al., Init. Repts. DSDP, 18. Washington (U.S. Govt. Printing Office), 835-842.

Wang, R., Biskaborn, B.K., Ramisch, A., Ren, J., Zhang, Y., Gersonde, R., & Diekmann, B., 2016. Modern modes of provenance and dispersal of terrigenous sediments in the North Pacific and Bering Sea: implications and perspectives for palaeoenvironmental reconstructions. Geo-Mar. Lett., 36 (4), 259-270. https://doi.org/10.1007/s00367-016-0445-7

Warner, N. & Domack, E., 2002. Millennial- to decadal scale paleoenvironmental change during the Holocene in the Palmer Deep, Antarctica, as recorded by particle size analysis. Paleoceanography, 17 (3), 8004. https://doi.org/10.1029/2000PA000602

Warrier, A.K., Pednekar, H., Mahesh, B.S., Mohan, R., & Gazi, S., 2016. Sediment grain size and surface textural observations of quartz grains in late quaternary lacustrine sediments 88

from Schirmacher Oasis, East Antarctica: Paleoenvironmental significance. Polar Science, 10 (1), 89-100. https://doi.org/10.1016/j.polar.2015.12.005

Weltje, G. & Prins, M., 2003. Muddled or mixed? Inferring paleoclimate from size distributions of deep-sea clastics. Sediment. Geol., 162, 39-62. https://doi.org/10.1016/S0037-0738(03)00235-5

Wentworth, C.K., 1922. A scale of grade and class terms for clastic sediments. J. Geol., 30 (5), 377-392.

Wolf-Welling, T., Cremer, M., Oconnell, S., Winkler, A., & Thiede, J., 1996. Cenozoic Arctic Gateway Paleoclimate Variability: Indications from Changes in Coarse-Fraction Composition. In: Thiede, J., Myhre, A.M., Firth, J.V., Johnson, G.L., & Ruddiman, W.F. (Eds.), Proceedings of the Ocean Drilling Program, Scientific Results, 151, pp. 515-567.

Wu, L., Wang, R., Xiao, W., Krijgsman, W., Li, Q., Ge, S., & Ma, T., 2018. Late Quaternary deep stratification-climate coupling in the Southern Ocean: implications for changes in abyssal carbon storage. Geochem. Geophys. Geosyst., 19 (2), 379-395. https://doi.org/10.1002/2017GC007250

CHAPTER 4. A MULTI-PROXY RECONSTRUCTION OF CHANGES IN SEA ICE AND PRIMARY PRODUCTIVITY AT IODP SITE U1339 (UMNAK PLATEAU, BERING SEA) DURING MARINE ISOTOPE STAGE 11

Modified from a manuscript to be submitted to Palaeogeography, Palaeoclimatology, Palaeoecology

Natalie S. Thompsona & Beth E. Caissiea,b

aDepartment of Geological & Atmospheric Science, Iowa State University, Ames, IA bCurrently at: US Geological Survey, Menlo Park, CA

Abstract

The ongoing decline in Arctic sea ice extent has prompted concerns about the fate of sea ice in the future, and the stability of sea ice-dependent ecosystems. Climate records from past warm intervals can help us to better understand the interplay between sea ice and primary productivity. Marine Isotope Stage (MIS) 11 (424-374 ka) is an interval of prolonged warming that represents a close analog to the Holocene with respect to orbital parameters. Here, we present a multi-proxy record for MIS 12 to 10 (430-368 ka) from the Integrated Ocean Drilling

Program (IODP) Site U1339 on the Umnak Plateau in the southeastern Bering Sea.

Sediment grain size, diatom assemblages, and a new diatom-based proxy for sea ice show that MIS 12 is characterized by extensive seasonal sea ice, similar to conditions during the

Last Glacial Maximum. Sea ice declines during deglaciation, but does not remain low throughout MIS 11. In fact, diatoms and ice rafted debris indicate that ice re-advanced across the plateau during the interglacial climatic optimum, when sea level and global temperatures were still high. In contrast, sea ice is not present at the Umnak Plateau today. We suggest that, 90 during MIS 11, sea ice at the Umnak Plateau may have persisted due to a more easterly and/or weaker Aleutian Low.

Laminated sediments at the boundary between MIS 12 and 11 point toward an interval of enhanced seasonal productivity during deglaciation. However, absolute diatom abundances and total organic carbon indicate that productivity does not remain high once the interglacial stabilizes. The gradual decline in productivity over the course of MIS 11 occurs independently of fluctuations in sea ice concentration and upwelling, and instead may be a response to increasing nutrient limitation.

Introduction

Sea ice is a critical component of the global climate system. However, the extent of

Arctic sea ice has declined substantially over the course of the modern satellite record (1979- present). The 13 lowest summer sea ice minima have occurred over the past 13 years (2007-

2019), and the extent of thick multi-year ice is also declining rapidly, which further contributes to the downward trend in minimum ice extent (IPCC, 2014; Osborne et al., 2018; Perovich et al.,

2019). Climate models indicate that the Arctic Ocean may become seasonally ice-free within the next few decades (e.g., Holland et al., 2006; Wang & Overland, 2009; Stroeve et al., 2012;

Overland & Wang, 2013; Stroeve & Notz, 2018; Screen & Deser, 2019). Alarmingly, the magnitude and rate of Arctic sea ice loss today is unprecedented in at least the last 1,500 years

(Polyak et al., 2010; Kinnard et al., 2011; Halfar et al., 2013; IPCC, 2014).

The ongoing decline in Arctic sea ice extent has prompted concerns about the fate of sea ice in the future and the stability of sea ice-dependent ecosystems. Although satellite data 91 only extend back to 1979, and observations of sea ice prior to the 1970s are sparse at best, we can use proxy records to examine changes in sea ice during previous warm intervals, in order to better understand and predict current/future change. Sediment cores from the Bering Sea provide us with sub-millennial-scale resolution records of natural variability in sea ice extent and associated changes in primary productivity during an interval where climate warmed dramatically: the transition from Marine Isotope Stage 12 to the prolonged warmth of Marine

Isotope Stage (MIS) 11 at ~424 ka. MIS 11 is the most recent interglacial where orbital parameters and insolation values were similar to modern values, and so has long been considered a suitable analogue for pre-industrial warming during the Holocene (Droxler &

Farrell, 2000; Berger & Loutre, 2002; Loutre & Berger, 2003; Masson-Delmotte et al., 2006;

Bowen, 2010). In this paper, we document changes in sea ice extent and primary productivity throughout MIS 11 at Integrated Ocean Drilling Program (IODP) Site U1339 (Umnak Plateau,

Bering Sea), using a multi-proxy approach that includes sediment grain size measurements, diatom assemblage counts, and stable isotope analyses. The sea ice records produced as part of this work will help to put modern sea ice decline into a longer-term context, and will also help to elucidate the relationship between diatom productivity and sea ice extent.

Background

Marine Isotope Stage 11

Throughout the Pleistocene, there have been interglacial periods of exceptional warmth known as ‘super-interglacials’ (Melles et al., 2012). MIS 11 (424-374 ka) is one such period, and has, in fact, been recognized as the longest (and potentially the warmest) interglacial of the 92 past 500 ka (Fig. 4.1; Hodell et al., 2000; Berger & Loutre, 2002; Lisiecki & Raymo, 2005; Rohling et al., 2010). In addition, it is the most recent interglacial period when orbital parameters were similar to today; eccentricity was low, obliquity was high and the amplitude of precessional changes was dampened (Berger & Loutre, 2002; Droxler et al., 2003; Loutre & Berger, 2003).

Relatively high atmospheric CO2 concentrations of 275 ppm (Petit et al., 1999; Siegenthaler et al., 2005; Lüthi et al., 2008; Bereiter et al., 2015) contributed to the extreme warmth recorded in the proxy record; MIS 11 would otherwise have been much cooler, due to low insolation values throughout (Fig. 4.1; Yin & Berger, 2011). The interglacial climatic optimum (MIS 11c) lasted in the order of 10-30 ka and was characterized by higher than present sea level (Raymo

& Mitrovica, 2012; Dutton et al., 2015), enhanced carbonate plankton blooms at high latitudes and the expansion of coral reefs (Droxler & Farrell, 2000). During MIS 11c, Antarctic temperatures were up to 2°C warmer than today (Jouzel et al., 2007), boreal forest extended across southern Greenland (de Vernal & Hillaire-Marcel, 2008; Kleinen et al., 2014) and large lakes in Siberia were anomalously productive (Prokopenko et al., 2010; Lozhkin and Anderson,

2013).

Despite the work done to characterize the warmth of MIS 11 on land (e.g., Droxler et al.,

2003; Ashton et al., 2008; Candy, 2009; Candy et al., 2014; Kleinen et al., 2014; Tye et al., 2016), as well as in the North Atlantic (e.g., Kandiano & Bauch, 2007; Dickson et al., 2009; Poli et al.,

2010; Voelker et al., 2010; Bauch, 2012; Kandiano et al., 2012, 2017; Milker et al., 2013), less is known about this interval from the North Pacific region. This project is one of the first studies to describe sub-millennial-scale changes in sea ice extent in the Bering Sea during MIS 11. It builds on recent work by Caissie et al. (2016) on orbital- and millennial-scale changes in productivity 93 and sea ice extent at the Bering Sea shelf-slope break (IODP Site U1345) from MIS 12 to 10, and on work by Vaughn and Caissie (2017), in which the authors investigate the effects of sea ice extent on primary productivity in the southeastern Bering Sea (Site U1339) during MIS 5.

Study Area

Oceanographic Setting

Located between Russia and Alaska, the Bering Sea is the northernmost marginal sea of the Pacific Ocean, and the third largest marginal sea in the world. The Bering Sea is divided from the North Pacific by the Aleutian Island Arc (Fig. 4.2), but water mass exchange occurs through passes in the Aleutian Islands, linking Bering Sea conditions to those of the Pacific

(Stabeno et al., 1999). Warm, nutrient-rich surface currents, derived from inflow of the Alaskan

Stream (Fig. 4.2), fuel the highly productive Bering Sea ecosystem. As the sole gateway between the Bering Sea and the Arctic Ocean, inflow of Pacific water through the Bering Strait (Fig. 4.2) is an important source of heat, freshwater and nutrients to the Arctic (Walsh et al., 1989; Stabeno et al., 1999; Grebmeier et al., 2006; Woodgate et al., 2010). Closure of the Bering Strait during glacial lowstands isolates the North Pacific from the North Atlantic, causing major changes in global ocean circulation, salinity, and nutrient distributions (e.g., Okazaki et al., 2010; Hu et al.,

2012, 2015; Knudson & Ravelo, 2015; Caissie et al., 2016).

Because of its unique topography and location, the Bering Sea is well situated for the study of past changes in sea ice extent. Sediment core records dating back to the Pliocene hint at significant regional variability in sea ice regimes across the Bering Sea (e.g., Cook et al., 2005;

Katsuki & Takahashi, 2005; Caissie et al., 2010, 2016; Takahashi et al., 2011; Kanematsu et al.,

2013; Stroynowski et al., 2015; Onodera et al., 2016; Teraishi et al., 2016; Detlef et al., 2018).

On shorter timescales, for example, sea ice appears to be relatively stable at the shelf-slope break, even during times of significant climate transition (Caissie et al., 2016), but elsewhere in the Bering Sea ice has fluctuated considerably, expanding southward during glacial periods, 95 retreating to its present limit (Cook et al., 2005; Katsuki & Takahashi, 2005; Caissie et al., 2010) or even further north (Kaufmann and Brigham-Grette, 1993) during interglacials.

Figure 4.2. Map of Beringia, showing the position of IODP Site U1339 (green star), and other sites referred to in the text. The dashed line shows the maximum extent of sea ice today (median over the period 1979-2013) (Cavalieri et al., 1996). Currents (shown in blue) are modified from Stabeno et al. (1999). Abbreviations include: Alaska Coastal Current (ACC), Aleutian North Slope Current (ANSC), Bering Slope Current (BSC), Bering Strait (BS), Umnak Plateau (UP), and Unimak Pass (UkP). Grey bathymetric shading changes value at -50 m (Bering Strait sill depth), -250 m (shelf/slope break), -1000 m, and -2000 m.

Core Site Location

Integrated Ocean Drilling Program (IODP) Site U1339 is located on the Umnak Plateau in the southeastern Bering Sea, at a water depth of approximately 1870 m (Fig. 4.2; Takahashi et al., 2011). The Umnak Plateau is erosionally isolated from the continental slope by the submarine Bering Canyon (Scholl et al., 1968). Surface and intermediate water flow in the region is influenced by the Alaska Coastal Current (ACC), and by the Aleutian North Slope

Current (ANSC), which is derived from inflow of the Alaskan Stream (AS), and which provides an important nutrient source for marine ecosystems in the southeastern Bering Sea (Reed &

Stabeno, 1999; Stabeno et al., 1999, 2009; Mordy et al., 2005). These relatively warm currents inhibit the formation of sea ice in the Umnak Plateau region today (Fig. 4.2). However, there are several lines of evidence for substantial sea ice coverage at the Umnak Plateau during the Last

Glacial Maximum (LGM) and previous glacial intervals (e.g., Sancetta et al., 1985; Katsuki &

Takahashi, 2005; Caissie et al., 2010; Vaughn & Caissie, 2017; Pelto et al., 2018; Nesterovich,

2019).

Methods

Geochronology

Cook et al. (2016) constructed an age model for site U1339 by aligning the δ18O values of the benthic foraminifera species Uvigerina peregrina, Uvigerina senticosa and Elphidium cf. batialis to the Lisiecki & Raymo (2005) global benthic foraminiferal stack (LR04) (Table 4.1; Fig.

4.3). They identified Marine Isotope Stages by selecting depths at the mid-point of transitions 97 between glacials and interglacials in comparison to LR04. Dates for individual samples were obtained through linear interpolation between tie points (Table 4.1).

Table 4.1. Age-depth tie points used in the age model for Site U1339.

Global Stack (Lisiecki et al., 2005) Site U1339 (Cook et al., 2016) Stage End (ka) Start (ka) Top (CCSF-m) Bottom (CCSF-m) M MIS 1 0 14 -0.5 ± 1 3.5 ± 0.5 MIS 5 71 130 20.5 ± 1 34.6 ± 1 MIS 7 191 243 54.2 ± 1 70.5 ± 1.5 MIS 9 300 337 83.2 ± 2 91.8 ± 1 MIS 11 374 424 98.5 ± 1 112 ± 1.5 MIS 13 478 533 123.7 ± 2 135.9 ± 1.5

Figure 4.3. Benthic foraminiferal δ18O record from Site U1339 (red, Cook et al., 2016), aligned with the global LR04 stack (black, Lisiecki & Raymo, 2005). Grey bars denote interglacial stages 11-1.

The δ18O record for Site U1339 is only tied to the global stack at the beginning and end of MIS 11, and so we have low confidence in the exact timing of peak interglacial conditions.

We attempted to improve the existing age model by aligning the U1339 magnetic susceptibility record to the well-dated record from site U1343 (Asahi et al., 2016), using the intra-site 98 correlation age modeling (iscam) program in MATLAB (Fohlmeister, 2012; Appendix A, Fig. A.1).

However, this approach did not yield reliable tie points, so we use the original age model in this work. Considering the oxygen isotope sampling resolution, in addition to the stated error of 4 kyr in the LR04 dataset, the uncertainty of the age model for site U1339 is probably several thousand years (Lisiecki & Raymo, 2005; Cook et al., 2016). Therefore, we urge caution when comparing millennial-scale changes at this site with high resolution MIS 11 records from other sites.

Sediment Sampling

Sediment cores used in this work were retrieved from the Umnak Plateau Site U1339 during IODP Expedition 323 to the Bering Sea in 2009 (Fig. 4.2), using an advanced piston corer

(Takahashi et al., 2011). As this study is focused on records from MIS 11, sediments from the primary splice (holes U1339C and U1339D) were sampled from depths of 96.87 to 113.3 m below the seafloor (CCSF-m). The sampling interval spans from 430 to 365 ka, and encompasses the entirety of MIS 11, as well as parts of the glacial stages 12 and 10 (Table 4.1). Table 4.2 shows the depths and ages of major climate intervals referred to in the text (Lisiecki et al.,

2005; Kandiano et al., 2012; Cook et al., 2016; PAGES Past Interglacials Working Group, 2016).

Table 4.2. Depths and ages of major climate intervals referred to in the text.

Interval Top (CCSF-m) Bottom (CCSF-m) End (ka) Start (ka) MIS 10 96.87 98.5 365 374 Late MIS 11 98.5 104.98 374 398 Peak MIS 11 (11c) 104.98 110.92 398 420 Deglaciation/Early MIS 11 110.92 112 420 424 Termination V 112 424 MIS 12 112 113.3 424 430 99

Grain Size and Sedimentological Analyses

Published Records

In this study, we use previously published grain size records and smear slide data from

Site U1339 (Chapter 3). Grain size was measured every ~5.5 cm, for an average temporal resolution of 220 years. Measurements were made on bulk sediments, and on the terrigenous fraction of the sediment, after removing organic, biogenic and carbonate material (see Chapter

3 for detailed methodology). In addition, we refer to shipboard core descriptions (Takahashi et al., 2011) and smear slide counts (Chapter 3) in our interpretations. Shipboard core descriptions were used as a basis to characterize primary sediment lithologies, whilst smear slide counts enabled a more in-depth analysis of sediment composition.

Wet Sieving

A subset of samples (n=34) were washed over 63 μm and 150 μm sieves, and oven-dried at 60°C. Quartz grains (a proxy for ice rafting) were picked from the >150 μm size fraction, using a Nikon stereomicroscope. We also recorded the number of foraminifera in both the 63-150 μm and the >150 μm size fractions.

Diatom Analyses

57 quantitative diatom slides (~1100-year temporal resolution) were prepared following the revised methods of Warnock and Scherer (2015) and counted based on the methods of

Schrader and Gersonde (1978). At least 300 diatom valves (plus Chaetoceros resting spores) in at least 3 random transects were counted, using a Nikon ECLIPSE Ni transmitted light microscope at a magnification of 1000x. Transect lengths were measured with a stage micrometer. Where possible, diatoms were identified to species level, using published 100 taxonomic descriptions and images (Fryxell & Hasle, 1972; Koizumi, 1973; Sancetta, 1982;

Medlin & Priddle, 1990; Witkowski et al., 2000; Onodera and Takahashi, 2007; Lundholm and

Hasle, 2010). Species were then grouped according to their preferred ecological niche (modified from Caissie et al., 2016, and references therein). In addition, a quantitative, diatom-based proxy for sea ice concentration (Nesterovich and Caissie, in review) was applied to the diatom record.

Stable Isotope Analyses

Carbon and nitrogen isotopes from 60 bulk (non-acidified) and 60 organic (acidified) sediment samples (~1000-year temporal resolution) were measured in the Stable Isotope Lab at

Iowa State University using a Delta XL Plus Mass Spectrometer in continuous flow mode coupled with a Costech Elemental Analyzer. Prior to analysis, samples were ground into a fine powder using a mortar and pestle, and one set of samples was treated with 1 M HCl to remove inorganic carbon. Reference standards (Caffeine [IAEA-600], IAEA-N2, Cellulose [IAEA-CH-3}, and lab standard Acetanilide were used for isotopic corrections. The combined uncertainty is ±

0.08-0.11 ‰ (VPDB) for δ13C and ± 0.13-0.21 ‰ (air) for δ15N.

Results

Lithology

Sedimentation rates for marine isotope stages 12 through 10 were calculated based on the age-depth tie points in Table 4.1 (Cook et al., 2016). Sedimentation rates at Site U1339 range from 18 to 27 cm/kyr−1, and are highest during MIS 11, although the difference in average sedimentation rates between the glacial and interglacial periods is fairly small. 101

Figure 4.4. Lithostratigraphic column for site U1339, based on shipboard core descriptions and smear slide analyses (Takahashi et al., 2011; Chapter 3). Column width varies according to the median grain size of bulk sediments (Chapter 3). Colors represent the proportion of diatoms in the sediment: silt/clay with 10-40% diatoms (olive); silt/clay with 40-60% diatoms (dark green); silt/clay with >60% diatoms (yellow); laminated sediments (pale green). Blue circles indicate the presence and open circles indicate the absence of quartz grains >150 µm in the subset of samples picked for quartz grains. The grey side bar shows the duration of MIS 11.

Sediments at Site U1339 are a mix of biogenic, siliciclastic and volcaniclastic (mostly fine ash) components (Fig. 4.4). Sediment is predominantly biogenic, composed largely of siliceous diatom frustules, with lesser amounts of other microfossils such as foraminifera, radiolarians and silicoflagellates. Foraminifera typically have low abundances in the Bering Sea (Takahashi et al., 2011), but are present in large numbers at Site U1339 from ~422 to 418 ka. The biogenic materials are mixed with small amounts of tephra and varying amounts of siliciclastic 102 sediments, mostly clay and silt, but also including sand-sized minerals and rock fragments, gravel, and pebble-sized clasts that are interpreted as ice-rafted debris (IRD) (Fig. 4.4; Takahashi et al., 2011; Chapter 3). Quartz grains >150 μm are consistently present throughout the record, save for the interval from 421 to 418.5 ka (Fig. 4.4). It should be noted that samples from this interval contain large numbers of foraminifera, which may have diluted the IRD signal.

Throughout most of Marine Isotope Stage 11, sediments alternate between biogenic material (silt/clay with >60% diatoms, or diatom ooze) and subequal proportions of biogenic and siliciclastic material (silt/clay with 40-60% diatoms). However, after ~379 ka, the biogenic content of the sediment decreases, with sediments composed of silt/clay with 10-40% diatoms

(Fig. 4.4). Ash layers are observed at ~421.5 and ~411 ka, but volcanogenic material at Site

U1339 is more frequently present as an accessory or trace lithology (Takahashi et al., 2011).

During deglaciation, laminated sediments are deposited over a ~2000-year interval, from 422.3 to 420.2 ka (Fig. 4.4). The laminated sediments consist of alternations between mm- scale dark (more terrigenous) and light (diatom-rich) laminae (Harbour, 2019). Such features are most likely to be preserved under hypoxic or anoxic conditions, when oxygen concentrations at the sediment-water interface are low enough to inhibit sediment mixing by benthic organisms (Calvert, 1964; Kemp, 1996; Pike & Kemp, 1999; van Geen et al., 2003;

Schimmelmann et al., 2016).

Diatoms

Full Diatom Assemblage

More than 90 diatom taxa were identified across 57 samples (Appendix A, Table A.1).

The assemblage is dominated by Chaetoceros and Thalassiosira antarctica resting spores (RS), 103 with smaller amounts of Fragilariopsis cylindrus, Thalassiosira gravida, Thalassiosira nordenskioeldii, Dentonula confervacaea, Actinocyclus curvatulus, Shionodiscus trifultus,

Shionodiscus oestrupii, Neodenticula seminae, Thalassionema nitzschioides, Coscinodiscus spp.,

Rhizosolenia hebetata, and Thalassiosira jouseae (Fig. 4.5). The number of taxa per sample ranges from 29 to 51, with an average of 37 taxa.

Figure 4.5. Relative % (area plots) and absolute (line plots) abundances of diatom taxa that make up more than 10% of any assemblage. Note that we use Chaetoceros-free counts in all relative % abundance records, with the obvious exception of Chaetoceros RS: full counts can be found in Appendix A, Fig. A.2. Taxa are grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); North Pacific indicator (orange); summer bloom (light green); other (grey); and Chaetoceros RS (dark green). The final line plot shows the total number of diatom valves per gram of sediment. Note the different scale for absolute abundances of Chaetoceros RS and all diatom valves. The grey panel indicates the duration of MIS 11 (424-374 ka) and the green bar represents laminated sediments.

104

The Margalef diversity index, a measure of species richness (Maurer and McGill, 2011), shows a decrease in the number of taxa during deglaciation (Fig. 4.6). Overall, species richness is highest during glacial stages 12 and 10, and increases gradually throughout MIS 11, although there is considerable downcore variability. Species evenness, as reflected by the Simpson diversity index, is highly variable throughout the core (Fig. 4.6). A Simpson value of 1 indicates that each taxon is present in equal abundance, whereas values close to 0 indicate that a single species dominates the assemblage, which could signify a strong dissolution signal (Maurer and

McGill, 2011). In general, the Simpson index is closer to 1 throughout the record (mean value of

0.83), but there are several periods where Simpson values decrease, most notably during the laminated interval from ~422-420 ka, and again between ~403.5-400 ka (Fig. 4.6). These declines in species evenness reflect intervals where Chaetoceros RS dominate the assemblage.

Figure 4.6. Margalef (black) and Simpson (green) diversity indices. The grey panel shows the duration of MIS 11 (424-374 ka), and the green bar shows laminated sediments.

105

Absolute diatom abundances (total number of valves per gram of sediment) vary from

106 to 108 diatoms per gram, with values 1-2 orders of magnitude higher (108) during the laminated interval, compared to massive intervals (Fig. 4.5). Diatom abundances increase during deglaciation and reach a peak towards the end of the laminated interval (420.74 ka), then gradually decline throughout the record. In general, diatom abundances are higher during

MIS 11, and lowest during glacial stages 12 and 10 (Fig. 4.5). For the most part, absolute abundances follow the same trend as relative abundances (Fig. 4.5); for this reason, we chose to focus on relative percent abundances in our interpretations.

Niche Groupings

Diatoms are valuable tools in paleoenvironmental reconstructions, because individual species have specific and unique ecological requirements with respect to temperature, salinity and other environmental variables. In this study, we group together diatoms with similar environmental preferences to infer paleoceanographic conditions at IODP Site U1339 from MIS

12 to 10 (Table 4.3; modified from Caissie et al., 2016). We use Chaetoceros-free counts in the majority of our interpretations, because the extremely high abundances of Chaetoceros RS in sediments at Site U1339 dominate the assemblage, and mask trends in other, less abundant species (see also Sjunneskog & Scherer, 1997; Leventer et al., 2002; Allen et al., 2005). Full counts can be found in Appendix A, Figs. A.2 and A.3).

The full diatom assemblage is clearly dominated by Chaetoceros RS, which make up between 10 and 72% of the assemblage. Resting spores of Chaetoceros are more readily preserved in sediments than the vegetative cells, which are weakly silicified, easily dissolved, and easily fragmented (Kanaya & Koizumi, 1966). Relative % abundances of Chaetoceros RS are 106 highest during the laminated interval, with a second peak from 403.5 to 401.9 ka. Total abundances of Chaetoceros RS are strongly coupled with total diatom abundances (correlation coefficient of 0.9) (Fig. 4.5).

Table 4.3. Subset of Bering Sea diatom species observed in this study, grouped by environmental niche (modified from Caissie et al., 2016).

Sea Ice Dicothermal Summer Bloom Freshwater Bacterosira bathyomphala Actinocyclus curvatulus Coscinodiscus spp. Aulacoseira spp. Fossula arctica Shionodiscus trifultus Proboscia curvirostris Eunotia spp.

Fragilariopsis cylindrus Rhizosolenia hebetata Lindavia cf. ocellata Fragilariopsis nana Rhizosolenia spp. Lindavia radiosa Warmer Water Fragilariopsis oceanica Asteromphalus robustus Fragilariopsis reginae-jahniae Azpetia tabularis High Productivity Neritic Porosira glacialis Shionodiscus oestrupii Chaetoceros RS Actinoptychus spp. Sinerima marigela Thalassiosira binata Odontella aurita Cymatosira spp. Synedropsis hyperborea Thalassiosira eccentrica Thalassiosira pacifica Dentonula confervacaea Thalassiosira angulata Thalassiosira symmetrica Thalassiothrix longissima Diploneis smithii

Thalassiosira antarctica RS Diploneis spp.

Thalassiosira decipiens North Pacific Entomoneis spp. Thalassiosira gravida Neodenticula seminae Melosira sol Thalassiosira hyalina Thalassionema nitzschioides Paralia sulcata Thalassiosira nordenskioeldii Stephanopyxis turris Thalassiosira baltica

In Chaetoceros-free counts, sea ice-associated diatoms, including Bacterosira bathyomphala, Fragilariopsis cylindrus, Fragilariopsis reginae-jahniae, Fossula arctica,

Thalassiosira antarctica, Thalassiosira gravida, Thalassiosira hyalina, and Thalassiosira nordenskioeldii, are the most abundant niche group. Sea ice diatoms comprise a large, albeit variable, proportion of the assemblage, ranging from 7% to a maximum of 62% (Fig. 4.7). There is a marked decline in sea ice diatoms at the onset of MIS 11, and another short-lived decline

~400 ka, but the percentage of sea ice species does not remain low throughout MIS 11 (Fig. 107

4.7). From 417 to ~402 ka, the relative percent abundance of the sea ice diatom group increases, with a peak centered on the interval from 403.5 to 401.9 ka (Fig. 4.7). A second, short-lived peak occurs during late MIS 11 (~391 ka), but for the most part, the % abundance of sea ice diatoms is relatively stable after ~398 ka (Fig. 4.7).

Neritic and freshwater diatoms reflect shelf to basin transport. Freshwater diatoms make up a relatively small (0-2.3%) proportion of the assemblage at Site U1339, whereas neritic 108 diatoms range from 2 to 42%. On average, neritic diatoms comprise <10% of the assemblage, but there is an interval from ~395 to 392 ka where the % abundance of the neritic species D. confervacaea remains consistently high (>20%) for approximately 3000 years (Figs. 4.5, 4.7).

The dicothermal layer is a cold layer of water located between seasonally warm surface and warmer deep waters. In the Bering Sea today, the formation of the dicothermal layer is associated with melting sea ice. Overall, the percentage of the dicothermal indicator species A. curvatulus and S. trifultus is lowest during stages 12 and 10, increases during deglaciation, and is highly variable throughout MIS 11, ranging from 3.3 to 21.9 % of the assemblage.

Diatoms associated with warmer water (mostly S. oestrupii) are present in relatively low abundances (<10%) for most of the record (Figs. 4.5, 4.7). Two species (N. seminae and T. nitzschioides) are used here specifically as tracers of warm, nutrient-rich North Pacific Water

(NPW) (Abrantes et al., 2007; Miettinen, 2013; Lopes & Mix, 2018). Although North Pacific indicator species are more abundant than warm water diatoms, the two groups follow broadly similar trends (Fig. 4.7). Relative percent abundances of North Pacific indicator species increase at the onset of MIS 11, with an abrupt decline from 403.5 to 401.9 ka, consistent with a drop in the percentage of warm water species. This is followed by a peak in North Pacific indicator species and a corresponding increase in warm water species around 400 ka (Fig. 4.7).

High productivity indicator species (mainly Odontella aurita and Thalassiothrix longissima) are present in low abundances throughout most of the record (Fig. 4.7).

Chaetoceros RS are also a taxon associated with high productivity environments, but are presented separately from other high productivity indicators, because of their association with highly productive upwelling systems (Abrantes, 1988; Abrantes et al., 2007). 109

Diatom-based Proxy

We used a quantitative, diatom-based proxy, based on a 5 species General Additive

Model (Nesterovich and Caissie, in review), to estimate spring (March, April, May, June) sea ice concentrations at Site U1339. For the most part, sea ice concentrations follow the same general trend as the sea ice diatom group. Sea ice concentration is highly variable throughout the record, ranging from 7 to 89% ice cover. Sea ice concentration is mostly high (>50%) during late

MIS 12, and declines with the onset of deglacial warming, reaching its lowest concentration around 413 ka, although this is by no means a linear decline (Fig. 4.7). There is a peak in sea ice concentration from 403.5 to 401.9 ka, consistent with an increase in sea ice diatoms during the same interval (Fig. 4.7).

Geochemistry

Carbon Content

Total carbon follows roughly the same trend as absolute diatom abundances, with the highest values occurring during the laminated interval, and an overall decline throughout MIS

11 (Fig. 4.8), indicating a strong positive relationship between diatom productivity and carbon export at Site U1339. Total organic carbon (TOC) values are highest (1.5%) during the laminated interval, compared to 0.7% for the entire record. TOC declines after deglaciation, reaching a low of 0.4% by 416 ka. Overall, TOC is consistently lower during MIS 12 and 10, and considerably more variable during MIS 11 (Fig. 4.8).

Inorganic carbon (i.e., the % CaCO3), is also highest during the laminated interval, consistent with an increase in foraminifera, and other calcareous microorganisms (Takahashi et al., 2011). For the most part, the % inorganic carbon (IC) is lower than the % TOC, save for the 110 interval from 391.6 to 391.1 ka, where there is a peak in IC at the same time that TOC drops to its lowest value, and δ13C values become more negative. It is possible that this shift could be attributed to the presence of an authigenic carbonate nodule, as authigenic dolomite, which is depleted in δ13C (Lein, 2004), occurs as an accessory lithology at Site U1339 (Takahashi et al.,

2011).

111

112

Organic Matter Source

Cross-plots of carbon, nitrogen, and their isotopes can be used to help identify sources of organic matter (OM) to the study site. The majority of samples comprise a mixture of terrigenous OM from soils and C3 plants, and marine OM from phytoplankton (Fig. 4.9). δ13C values range from -24.8‰ to -20.4‰ VPDB, with an average value of 23.4‰ (Figs. 4.8, 4.9).

During the laminated interval, δ13C values plot closer to those of marine phytoplankton (-20 to -

22‰) and ice algae (-18.5 to -20.5‰) and remain high (>-23‰) for a period of ~6000 years

(423.7-417.6 ka). There is no substantial difference between the glacial and interglacial stages in terms of average δ13C, although the record is considerably more variable throughout MIS 11

(Fig. 4.8). Whilst OM is never fully terrestrial, there are several discrete intervals during MIS 11 when δ13C values plot close to those of terrestrial OM (-25 to -27‰) (Figs. 4.8, 4.9).

Bulk sedimentary δ15N values range from 0.8 to 7.3‰ (Fig. 4.8g). On average, δ15N is slightly lower during MIS 12 and 10 (4.7 and 4.3‰, respectively), compared to an average of

5‰ during MIS 11. After ~387 ka, there are several large and irregular excursions towards both high and low δ15N values (Fig. 4.8). Bulk sediments may contain a significant fraction of inorganic nitrogen, adsorbed onto clay particles or incorporated into the crystal lattice of potassium-rich clays such as illite. This is a common occurrence in Arctic sediments, which typically have a high illite content (Schubert and Calvert, 2001; Drenzek et al., 2007; Stein,

2008). The presence of inorganic nitrogen complicates the interpretation of nitrogen records, because it lowers C/N ratios and δ15N values. A regression analysis (Fig. 4.9a) shows that total organic carbon is significantly correlated with total nitrogen (R2 = 0.71), but the positive y- intercept (0.05) indicates that sediments at Site U1339 contain a sizable fraction of inorganic 113 nitrogen (Fig. 4.9a; Schubert & Calvert, 2001). The y-intercept remains positive when plotted separately for MIS 12, 11, and 10, so no single interval is without the influence of inorganic nitrogen. The presence of inorganic nitrogen is also seen in the cross plot of δ13C versus organic

C/N (Fig. 4.9b). Without the complicating factor of inorganic nitrogen, we would expect to see a negatively sloping trend line, reflecting mixing between the marine and terrestrial end members. Instead, the samples have a positive slope, due to the low C/N ratios. The low (2-

4‰) δ15N values associated with clay-bound nitrogen may result in samples having a lower δ15N signal than expected, in comparison to mean ocean δ15N of ~5‰ (Sigman et al., 2000).

Conversely, the process of denitrification, which is widespread today in the Bering Sea, can leave the nitrate pool enriched in δ15N, which ultimately results in higher (>5.5‰) sedimentary

δ15N values (Brunelle et al., 2007). The highly variable nitrogen values at site U1339 are likely driven by a combination of denitrification and the presence of inorganic nitrogen. We therefore conclude that bulk sedimentary nitrogen is not a reliable proxy for organic carbon source

(marine vs. terrestrial) or nitrogen utilization at Site U1339.

Discussion

Sea Ice at the Umnak Plateau

Sea Ice History

MIS 12 (474-424 ka) is considered one of the most severe glaciations of the Quaternary

(McManus et al., 1999; Stein et al., 2009; Rodrigues et al., 2017; Koutsodendris et al., 2019).

During Late MIS 12 (430-424 ka in this record), insolation, atmospheric CO2, and relative sea level begin to increase, although sea level does not cross the -50 m threshold until the 114 deglaciation (Fig. 4.10). Both diatom assemblages and the diatom proxy indicate a period of relatively high ice cover (>60%) during this time, implying a longer duration of seasonal sea ice.

Grain size analyses (Fig. 4.10; Chapter 3) show that the relative proportion of terrigenous clay is highest during MIS 12, whereas the relative proportion of coarse (>150 µm) terrigenous grains, interpreted as IRD, is low (Fig. 4.10). Sea ice is known to transport terrigenous clay, so it is tempting to attribute the increase in clay-sized grains to enhanced ice-rafting during glacials.

However, it has been shown that glacial periods were both drier and dustier (McGee et al.,

2010), which could also result in increased clay during glacials through enhanced aeolian transport. Low proportions of coarse IRD may indicate thick, perennial ice cover at Site U1339 during MIS 12, with reduced melting and thus limited deposition of sediments entrained within the ice. If this were the case, however, we would expect to see much lower sedimentation rates during MIS 12 (e.g., Nørgaard-Pedersen et al., 2003). Instead, MIS 12 was likely characterized by extensive seasonal sea ice, similar to the conditions at Umnak Plateau Site 51JPC during the

LGM (Sancetta et al., 1985; Caissie et al., 2010). In addition, the IRD content of sea ice at Site

U1339 might reveal a relationship between sea ice rafting and sea level. During MIS 12, when sea level was 120-140 m below present (Rohling et al., 1998, 2014; Spratt & Lisiecki, 2016), the area of shallow continental shelf over which sea ice forms was significantly reduced. We suggest that sea ice likely formed over deeper waters, and subsequently entrained less coarse terrigenous material. At the same time, the arid, windy, and dusty glacial climate conditions allowed for increased transport of fine-grained wind-blown particles to the site (McGee et al.,

2010). 115

Termination V (~424 ka) marks the transition from MIS 12 to 11 and is followed by a period of deglaciation that occurs in line with rising eccentricity, insolation, and sea level (Fig.

4.10). Deglaciation at Site U1339 occurs between 424 and 420 ka and is marked by rapid sea ice retreat. This is seen as a simultaneous decline in estimated sea ice concentrations and the relative percent abundance of sea ice diatoms, including a marked decrease in T. antarctica RS, a species associated with thick pack ice (Fig. 4.5, 4.10; Horner, 1985).

The interglacial climatic optimum, hereafter referred to as MIS 11c, extends from 420 to

398 ka at Site U1339 (Kandiano et al., 2012; PAGES Past Interglacials Working Group, 2016), with peak warmth centered around 405 ka, consistent with the timing of peak eustatic sea level

(~410-400 ka) (Raymo & Mitrovica, 2012). The deglacial sea ice retreat continues into MIS 11, with sea ice reaching its minimum extent around 417-413 ka (Fig. 4.10). From 417.5 to 410.5 ka, sea ice concentrations indicate a shift toward unconsolidated ice (15-40%) and even open- water conditions (<15%) (Fig. 4.10). However, sea ice does not remain low throughout MIS 11.

In fact, it re-advances during MIS 11c, at a time when insolation, global temperatures, and sea level were at their highest. This is shown by the consistent presence of IRD, a gradual increase in the relative percent abundance of sea ice diatoms, and an increase in sea ice concentrations, from 417 to 401.9 ka (Fig. 4.10). Following a short-lived drop in sea ice at 400 ka, average sea ice concentrations of 61% signal a period of relatively high seasonal sea ice cover that persists throughout Late MIS 11 (398-374 ka), and into MIS 10 (Fig. 4.10).

Controls on Sea Ice

Bering Sea climate is strongly influenced by variations in the inflow of North Pacific waters (Stabeno et al., 1999, 2017; Detlef et al., 2018). Today, relatively warm currents, derived 116 from inflow of the Alaskan Stream and the Alaska Coastal Current, inhibit sea ice formation in the Umnak Plateau region, but there is evidence that glacial reductions in North Pacific Water allowed for the expansion of sea ice at both the Umnak Plateau and the Bering Sea slope sites

(e.g., Sancetta et al., 1985; Caissie et al., 2010; Teraishi et al., 2016; Stroynowski et al., 2017;

Vaughn & Caissie, 2017; Detlef et al., 2018). The strong, inverse correlation (-0.76) between relative % abundances of sea ice diatoms and North Pacific indicator species at Site U1339 is further evidence for this relationship. During MIS 12 and 10, when sea ice is more extensive,

NPW is reduced, reflected by low relative % abundances of the North Pacific indicator species

N. seminae and T. nitzschioides (Fig. 4.10). It is probable that low sea level during glacial stages

12 and 10 led to closure of Unimak Pass (-70 m) and other shallow eastern passes (Stabeno et al., 2002; Ladd et al., 2005), thereby restricting inflow of the Alaska Coastal Current and the distal end of the Alaskan Stream, which would have isolated the southeastern Bering Sea, and promoted sea ice growth. Conversely, during deglaciation, as sea level rose above -50 m, the influx of warm Pacific waters at Site U1339 led to sea ice retreat and melting (Fig. 4.10).

However, this somewhat simplistic interpretation is challenged by the findings of Sancetta &

Silvestri (1984), who showed that abundances of N. seminae fluctuate on glacial-interglacial timescales after 450 ka, roughly in phase with the global oxygen isotope record. Therefore, low abundances of North Pacific species in our records from MIS 12 and 10 may be partially explained by overall lower abundances of N. seminae in the Pacific Ocean during glacials.

If sea level were the main control on North Pacific inflow to the Bering Sea, we might reasonably expect to see increasing inflow, and a corresponding decline in sea ice, when sea level was at its highest during MIS 11c. Our records, however, show the opposite; not only is 117

North Pacific inflow highly variable throughout MIS 11, the influence of NPW is reduced during the interglacial climatic optimum, at the same time that sea ice advances (Fig. 4.10). Although mesoscale variations in Alaskan Stream transport have been shown to impact climate and circulation on the Bering shelf (e.g., Ezer & Oey, 2010), it is unlikely that such variations could have reduced NPW inflow to near-glacial levels during MIS 11. Instead, we propose that the interplay between sea ice and North Pacific Water at the Umnak Plateau during MIS 11 was controlled, at least in part, by changes in the geographic position of the wintertime Aleutian

Low, a semi-permanent low pressure system centered over the Aleutian Islands (Niebauer et al., 1999). Today, sea ice extent in the Bering Sea is related to wind intensity, storm tracks, and the frequency of storms, all of which are controlled by the Aleutian Low (Overland & Pease,

1982; Niebauer et al., 1999; Rodionov et al., 2007; Katsuki et al., 2009). When the Aleutian Low is centered over the western Bering Sea, southerly winds inhibit the southward expansion of sea ice in the eastern Bering Sea; when the Aleutian Low is shifted further east to the Gulf of

Alaska, sea ice can advance (Rogers, 1981; Overland & Pease, 1982; Stabeno et al., 1999). Since wind stress at the sea surface is an important driver of ocean circulation, changes in the strength of the Aleutian Low have also been shown to influence ocean circulation in the region of the North Pacific subarctic gyre, of which the AS forms the northern boundary current (e.g.,

Stabeno et al., 1999; Pickart et al., 2009; Wills et al., 2019). Accordingly, we suggest that, during

MIS 11, the strength of the Aleutian Low exerted a control on the advection of North Pacific

Water to Site U1339, with increased transport of NPW occurring when the Aleutian Low intensified, due to the influence of wind stress curl.

118

In order to account for long-term trends in sea ice extent, the Aleutian Low would have to remain consistently weak, or consistently strong, over long (millennial) timescales. Decadal- scale variability in the Aleutian Low is recognized from instrumental data, but less is known 119 about long-term variability in the Aleutian Low. Anderson et al. (2005) used oxygen isotope records from Jellybean Lake (southern Yukon Territory) to investigate Holocene changes in the strength and position of the Aleutian Low. Their data showed a predominantly weaker and/or westward Aleutian Low from 7000 to 4500 years BP, and a stronger or more eastward shift between 4500 and 3000 BP, suggesting that the Aleutian Low may persist in a single state for several thousand years.

A large increase in the relative percent abundance of neritic diatoms from 395 to 392 ka

(Fig. 4.7) may also be related to the position of the Aleutian Low. At this time, the neritic group is dominated by D. confervacaea (Fig. 4.5), a species that thrives in cold water bays, and also appears to grow well in nutrient-deficient environments (Guillard & Ryther, 1962). If optimal growth conditions existed during Late MIS 11, this may have produced large blooms of D. confervacaea in coastal regions. If the Aleutian Low was shifted further to the east, sea ice formed along the coast could have entrained large numbers of D. confervacaea, then been blown southward toward the Umnak Plateau by northerly winds. Alternatively, the neritic diatoms could have been transported to the plateau by icebergs from tidewater glaciers.

However, considering the location of Site U1339, sea ice is more likely to transport terrigenous sediment to the site. Similar to the LGM, glacial ice in Beringia was likely present in the form of small, mountain-valley glaciers, which would not have produced large numbers of icebergs

(Glushkova, 2001; Elias & Brigham-Grette, 2013). Consistent with this, sediments typical of glacial IRD, such as dropstones, are rare at Site U1339.

120

Figure 4.11. Summary of productivity records from Site U1339, compared to global records. (a) mean monthly insolation at 65°N (Laskar et al., 2004); (b) atmospheric CO2 from Antarctic ice cores (Bereiter et al., 2015); (c) relative sea level in meters above present (Rohling et al., 2010); The dashed vertical line shows the Bering Strait sill depth; (d) bulk median grain size (Chapter 3); (e) diatom valves per gram of sediment); (f) Chaetoceros per gram of sediment; (g) relative % abundance of Chaetoceros RS; (h) % organic carbon; (i) % inorganic carbon; and (j) δ13C (‰ VPDB). Diatom percent abundances are based on Chaetoceros-free counts. The grey panel shows the duration of MIS 11, and colored side bars denote the following substages: deglaciation (424-420 ka, turquoise), Peak MIS 11 (430-398 ka, red), and Late MIS 11 (398-374 ka, blue). The green bar represents laminated sediments.

Primary Productivity at the Umnak Plateau

Marine Isotope Stage 12

On glacial-interglacial time scales, changes in median grain size provide a rough estimate of changes in productivity; around 23% of grain size variability at Site U1339 can be explained by variations in diatom content (Chapter 3). Bulk mean grain size records are in agreement with 121 absolute diatom abundances, and organic carbon %, all of which are consistently low during

MIS 12, and higher during MIS 11 (Fig. 4.11). This is consistent with other glacial-interglacial records from the Bering Sea and North Pacific (e.g., Sancetta et al., 1985; Brunelle et al., 2007;

Galbraith et al., 2008; Caissie et al., 2010; Riethdorf et al., 2013; Vaughn & Caissie, 2017; Worne et al., 2019), which suggest that glacial periods in this region are marked by low primary productivity. Today, productivity in the Bering Sea is strongly dependent upon the availability of nutrients and light, and is therefore highly seasonal (Arrigo, 2014). In particular, increased light availability and stratification during spring sea ice melt drives the hugely productive spring phytoplankton bloom (e.g., Niebauer et al., 1990; Sigler et al., 2010; Brown & Arrigo, 2013;

Vancoppenolle et al., 2013; Leu et al., 2015). If, as we suggest, sea ice was more extensive during MIS 12, algal blooms at Site U1339 would have been limited to in-ice or below-ice production, with the spring bloom occurring in the marginal ice zone further to the south.

Increased sea ice cover may have further restricted productivity; firstly, by limiting the effects of wind forcing, thereby reducing the formation of eddies, which drive upwelling of deep, nutrient-rich waters (Rainville et al., 2011); and secondly, by strengthening water column stratification, thus limiting the flux of nutrients into the photic zone (Sigman et al., 2004). In addition, glacial reductions in primary productivity were driven, at least in part, by reduced upwelling of nutrients, due to the expansion of sea ice and concurrent expansion of low- nutrient North Pacific Intermediate Water (Worne et al., 2019), whilst restricted inflow through the Aleutian Island Passes may have caused Bering Sea circulation to weaken, further reducing nutrient supply to the shelf edge (Kanematsu et al., 2013). 122

Deglaciation

During deglaciation, as sea level rose above -50 m, Unimak Pass and the Bering Strait re- opened, which led to flooding of the continental shelf, and a subsequent influx of nutrients to the Bering Sea (Shiga & Koizumi, 1999; Bertrand et al., 2000; Poli et al., 2010). The combination of rising sea level, the collapse of glacial North Pacific Intermediate Water, declining sea ice cover and increased nutrient supply drove rapidly increasing productivity, culminating in the deposition of laminated sediments between 422.3 and 420.2 ka (Fig. 4.11). Laminated sediments are widespread features of deglacial transitions; during the last deglaciation, for example, laminae were simultaneously deposited across the North Pacific and its marginal seas

(e.g., Cook et al., 2005; Brunelle et al., 2007; Caissie et al., 2010; Addison et al., 2012; Schlung et al., 2013; Pelto et al., 2018). Laminated sediments have also been reported from other sites in the Bering Sea during Early MIS 11 (e.g., Takahashi et al., 2011; Caissie et al., 2016). As laminated sediments typically form under low oxygen conditions, the presence of laminated sediments at Site U1339 would have required expansion of the oxygen minimum zone, currently located between -700 and -1600 m, (Takahashi et al., 2011), to a water depth of at least 1800 m.

Although the exact causes of deglacial anoxia are still debated, leading hypotheses include changes in the ventilation of intermediate waters (Zheng et al., 2000), increased export productivity (Kuehn et al., 2014), or a combination of the two (Cook et al., 2005; Matul et al.,

2016). Several lines of evidence point toward extremely high productivity at U1339 during the laminated interval, including an increase in absolute diatom abundances by 2 orders of magnitude over MIS 12, and a marked increase in bulk median grain size, which implies a 123 greater contribution from silt-sized biogenic material, i.e., diatoms (Fig. 4.11; Aiello & Ravelo,

2012; Chapter 3). Consistent with this, a peak in δ13C values suggests that OM is primarily sourced from marine phytoplankton during this time (Fig. 4.11). In addition, an increase in both organic and inorganic carbon during the laminated interval indicates high rates of productivity across multiple phytoplankton groups, including calcareous organisms such as foraminifera and nannofossils (Takahashi et al., 2011).

Marine Isotope Stage 11

In general, primary productivity declines over the course of MIS 11, as shown by gradually decreasing diatom abundances and an overall decrease in organic carbon (Fig. 4.11).

Median grain size remains higher throughout MIS 11, not because productivity is consistently high, but because the grain size of bulk sediments is also influenced by the input of coarser, ice- rafted material.

During MIS 11, when sea ice is high, productivity can be either low or high (Figs. 4.10,

4.11), which indicates that the relationship between sea ice and productivity is far more complex than the assumption that sea ice inhibits productivity. As described previously, there is a massive flux of nutrients during deglaciation, when sea level rise floods the Bering shelf; however, this influx of nutrients is not sustained throughout MIS 11, and it may be that productivity declines as nutrients become gradually depleted during MIS 11. Pelto et al. (2018) hypothesized that productivity at the Umnak Plateau remains high until the Bering Strait is fully submerged, because until that point, nutrients remain trapped in the Bering Sea. Once the strait is flooded, however, Bering shelf waters can transport nutrients from the Bering Sea northward into the Chukchi Sea, as they do today (Cooper et al., 1996; Tremblay et al., 2015). 124

The resting spores of Chaetoceros are indicators of both high productivity and upwelling, because Chaetoceros only blooms when nutrient levels are high enough to sustain multiple diatom blooms each year, or when there is an input of nutrients later in the year to support a second bloom (Abrantes et al., 1998, 2007). At Site U1339, Chaetoceros RS tell a different story to other productivity indicators. Absolute abundances of Chaetoceros RS are closely coupled with absolute diatom abundances (Fig. 4.11), simply because Chaetoceros RS dominate the assemblage, but the relative % abundance of Chaetoceros RS fluctuates throughout MIS 11, independent of both sea ice and productivity. We conclude that downcore variations in Chaetoceros RS likely reflect changes in upwelling, separate from productivity.

Figure 4.12. Location map of core sites on the Umnak Plateau.

Comparison of records from the Umnak Plateau

During the Last Glacial Maximum (30-19 ka; Lambeck et al., 2002), conditions at the

Umnak Plateau were similar in many ways to those observed during MIS 12. Under low sea 125 level, the Aleutian basin was partially isolated from the North Pacific, and the associated reduction in North Pacific inflow allowed for the development of thick, extensive sea ice over the deep waters of the Umnak Plateau (e.g., Sancetta & Robinson, 1983; Sancetta et al., 1985;

Caissie et al., 2010; Pelto et al., 2018; Nesterovich & Caissie, in review). Based on a maximum in sea ice diatoms (Nitzschia cylindra and Nitzschia grunowii) at Site RC14-121 (Fig. 4.12, 4.13),

Sancetta et al. (1985) proposed that, during the LGM, the Umnak Plateau was covered by sea ice throughout the year. In contrast, Kastsuki and Takahashi (2005) did not find evidence for perennial sea ice at a different Umnak Plateau site (UMK-3A; Fig. 4.12). Instead, they suggested that the presence of F. cylindrus, which is common in regions influenced by seasonal sea ice today, was evidence for seasonal sea ice during the LGM (Fig. 4.13). At the nearby Site 51JPC

(Fig. 4.12), Caissie et al. (2010) found evidence for thick, extensive ice cover, that persisted for at least six months per year, based on high % abundances of Fragilariopsis species and

Thalassiosira antarctica. In addition, Nesterovich & Caissie (in review) showed that sea ice concentrations at this site were high (average of 74.6%) during the LGM, further evidence for longer annual duration of sea ice cover (Fig. 4.13). Likewise, we find similar evidence (high % abundance of sea ice diatoms, in particular F. cylindrus and T. antarctica, as well as high sea ice concentrations) for extensive, seasonal sea ice cover at Site U1339 during Late Stage 12 (Fig.

4.5; Fig. 4.13). To enable a more robust comparison between diatom records from MIS 11 and

MIS 1, most of which focus on a single species or species group, we present both full sea ice diatom counts, and conservative counts in Figure 4.13. The conservative counts are based on seven species (Fragilariopsis cylindrus, Fragilariopsis oceanica, Fragilariopsis regina-jahniae,

Fossula arctica, Nitzschia frigida, Sinerima marigela, and Thalassiosira antarctica) with a strong 126 affinity for sea ice. Whilst overall abundances are somewhat lower for conservative counts, the two records display like trends.

Figure 4.13. Compilation of sea ice proxy records from Umnak Plateau sites U1339 (green), RC14-121 (purple), UMK-3A (yellow), and 51JPC (grey) for MIS 11 and 1 (denoted by grey panels). MIS 11 sea ice diatom records (this study) include all sea ice diatoms (lighter green) and conservative sea ice counts, based on seven species (Fragilariopsis cylindrus, Fragilariopsis oceanica, Fragilariopsis regina-jahniae, Fossula arctica, Nitzschia frigida, Sinerima marigela, and Thalassiosira antarctica) with a strong affinity for sea ice (darker green). MIS 1 sea ice records include % Fragilariopsis species (Caissie et al., 2010), % F. cylindrus (Katsuki & Takahashi, 2005), % Nitzschia cylindra and grunowii (Sancetta & Robinson, 1985), and sea ice concentration from the diatom-based proxy (Nesterovich & Caissie, in review). Green bars show laminated sediments, based on core records from U1339 (MIS 11) and 51JPC (MIS 1).

Although sea ice declined during the transition from MIS 12 to MIS 11, seasonal sea ice remained present at the Umnak Plateau (Site U1339) throughout most of the interglacial period, except for an excursion towards unconsolidated ice, and possibly ice-free conditions, for a few thousand years following deglaciation (Fig. 4.13). Whilst there are few records of MIS 5 127 on the Umnak Plateau, grain size data from Site U1339 (Vaughn & Caissie, 2017) point toward consistent input of IRD throughout MIS 5, save for a ~5000-year interval centered on 100 ka, which may correspond to the interstadial 5c. In contrast, the last deglaciation marked a transition from extensive sea ice cover to ice-free conditions across the Umnak Plateau. By 8 ka, modern oceanographic conditions were established in the Bering Sea (Sancetta, 1979), and both diatom assemblages and the diatom proxy indicate that sea ice has not advanced as far south as the Umnak Plateau since this time (Fig. 4.13; Sancetta & Robinson, 1983; Caissie et al.,

2010; Caissie & Nesterovich, in review).

Similar to MIS 11, laminated sediments were deposited across the Umnak Plateau region at the onset of MIS 5 and MIS 1 (Fig. 4.14). Again, several lines of evidence indicate 128 enhanced productivity during the laminated intervals; in fact, productivity is higher during deglaciation at all Umnak Plateau sites (Fig. 4.14; Caissie et al., 2010; Vaughn & Caissie, 2017;

Pelto et al., 2018), but a unique feature of the MIS 1 laminae is that they form two distinct sets

(Fig. 4.14; Caissie et al., 2010). The first set, deposited from 15.1-13.7 cal ka BP, are associated with the Bølling-Allerød, and the second set (11.5-11 cal ka BP) with the Pre-Boreal; the two are separated by a return to near-glacial conditions during the Younger Dryas. Interestingly, a

Younger Dryas-like reversal during Early MIS 11 has also been observed in records from the

North Atlantic (Voelker et al., 2010), Antarctica (EPICA Community Members, 2004), and Siberia

(Vogel et al., 2013), but there is no evidence for such an event at Site U1339.

At Site 51JPC, Caissie et al. (2010) found evidence that the last deglaciation was marked by a complete turnover in the diatom assemblage. A shift from sea ice diatoms to high productivity species occurred during the Bølling-Allerød, perhaps marking the transition from more extensive sea ice cover to marginal ice zone conditions. A second shift to more cosmopolitan species, including high abundances of the North Pacific diatom N. seminae, occurred during the Pre-Boreal. At UMK-3A, Katsuki & Takahashi (2005) also observed a rapid assemblage change during deglaciation, to an assemblage in which sea ice species and North

Pacific species coexisted. During MIS 11, there is no evidence for an assemblage turnover like that described by Caissie et al. (2010). However, sea ice diatoms and North Pacific species coexist in the assemblage throughout most of MIS 11 (Fig. 4.10), although relative percent abundances of N. seminae are considerably lower during MIS 11 than during the Holocene

(Caissie et al., 2010). This may suggest that the influence of warm, North Pacific waters was not as strong during MIS 11, which allowed seasonal sea ice to persist in the Umnak Plateau region. 129

During MIS 11, and especially MIS 5, relative % abundances of Chaetoceros RS are much higher than during the Holocene. (Fig. 4.14). Caissie et al. (2010) proposed that decreasing trends in Chaetoceros at 51JPC indicate reduced diatom productivity since the early Holocene

Thermal Maximum (11.5-9 cal ka BP). We note, however, that relative % abundances of

Chaetoceros RS at Site U1339 do not decrease in line with productivity, and that fluctuations in

Chaetoceros are more likely to reflect changes in upwelling. It could thus be argued that upwelling was more intense during MIS 11 and 5, in comparison to the Holocene. Williams et al.

(2011) used the δ13C values of coralline algae to show that upwelling in the northern North

Pacific and the Bering Sea increases when the Aleutian low intensifies. If the Aleutian Low was stronger during MIS 11, this could help to explain why Chaetoceros RS are so much more abundant during this time. Alternatively, since Chaetoceros are one of the last species to bloom in a seasonal succession (Saitoh and Taniguchi, 1978; von Quillfeldt, 2001), it could point toward a longer growing season during MIS 11.

Summary

In this study, we described changes in sea ice extent and primary productivity at the

Umnak Plateau (Site U1339) during the MIS 12-10 cycle. Similar to the LGM, MIS 12 was likely characterized by extensive, seasonal sea ice over the Umnak Plateau. We do not find evidence for perennial ice in this region, but high sea ice concentrations indicate a longer annual duration of seasonal ice.

Sea ice declined during deglaciation, and at the beginning of MIS 11, with a shift toward unconsolidated ice, and possibly open water conditions, starting around 417.5 ka. However, sea 130 ice did not remain low throughout MIS 11. Instead, we see evidence that sea ice re-advanced over the Umnak Plateau during the latter half of MIS 11c, when global temperatures and sea level were still high, but the influence of warm North Pacific water was reduced. In contrast, sea ice has been absent from the Umnak Plateau region since the Holocene Thermal Maximum (~8 ka). We suggested that the interplay between sea ice extent and North Pacific Water during

MIS 11 was controlled, at least in part, by persistent changes in the geographic position and intensity of the Aleutian Low.

In general, productivity at Site U1339 was higher during MIS 11, and lowest during glacial stages 12 and 10. Similar to the last deglaciation, laminated sediments were deposited during the transition from MIS 12 to 11. Multiple lines of evidence indicate a spike in productivity during the deposition of the laminated sediments. The gradual decline in productivity over the course of MIS 11 occurred independently of fluctuating sea ice, which suggests that productivity was controlled by more complex factors than simply changing sea ice extent. We proposed that nutrient availability may have become a limiting factor as MIS 11 progressed.

References

Abrantes, F., 1988. Diatom assemblages as upwelling indicators in surface sediments off Portugal. Mar. Geol., 85 (1), 15-39. https://doi.org/10.1016/0025-3227(88)90082-5

Abrantes, F., Lopes, C., Mix, A., & Pisias, N., 2007. Diatoms in Southeast Pacific surface sediments reflect environmental properties. Quat. Sci. Rev., 26 (1-2), 155-169. https://doi.org/10.1016/j.quascirev.2006.02.022

Addison, J., Finney, B., Dean, W., Walczak, M., Mix, A., Stoner, J., & Jaeger, J., 2012. Productivity and sedimentary δ15N variability for the last 17,000 years along the northern Gulf of 131

Alaska continental slope. Paleoceanography, 27, PA1206. https://doi.org/10.1029/2011PA002161

Aiello, I.W. & Ravelo, A.C., 2012. Evolution of marine sedimentation in the Bering Sea since the Pliocene. Geosphere, 8 (6), 1231-1253. https://doi.org/10.1130/GES00710.1

Allen, C.S., Pike, J., Pudsey, C.J., & Leventer, A., 2005. Submillennial variations in ocean conditions during deglaciation based on diatom assemblages from the southwest Atlantic. Paleoceanogr. Paleoclimatol., 20 (2), PA2012. https://doi.org/10.1029/2004PA001055

Anderson, L., Abbott, M.B., Finney, B.P., & Burns, S.J., 2005. Regional atmospheric circulation change in the North Pacific during the Holocene inferred from lacustrine carbonate oxygen isotopes, Yukon Territory, Canada. Quat. Res., 64 (1), 21-35. https://doi.org/10.1016/j.yqres.2005.03.005

Arrigo, K.R., 2014. Sea Ice Ecosystems. Annu. Rev. Mar. Sci., 6, 439-467. https://doi.org/10.1146/annurev-marine-010213-135103

Berger, A. & Loutre, M.F., 2002. An Exceptionally Long Interglacial Ahead? Science, 297 (5585), 1287-1288. https://doi.org/10.1126/science.1076120

Bertrand, P., Pedersen, T. F., Martinez, P., Calvert, S., & Shimmield, G., 2000. Sea level impact on nutrient cycling in coastal upwelling areas during deglaciation: evidence from 132

nitrogen isotopes. Global Biogeochem. Cy., 14 (1), 341-355. https://doi.org/10.1029/1999GB900099

Bowen, D Q., 2010. Sea level ∼400,000 years ago (MIS 11): Analogue for present and future sea-level? Clim. Past, 6, 19-29. https://doi.org/10.5194/cp-6-19-2010

Calvert, S.E., 1964. Factors Affecting Distribution of Laminated Diatomaceous Sediments in Gulf of California. In: van Andel, T.H. & Shor, G.G., Jr., (Eds.), Marine Geology of the Gulf of California: a symposium. AAPG Memoir, 3, 311-330.

Cooper, L.W., Whitledge, T.E., Grebmeier, J.M., & Weingartner, T., 1997. The nutrient, salinity, and stable oxygen isotope composition of Bering and Chukchi Seas waters in and near the Bering Strait. J. Geophys. Res., 102 (C6), 12,563-12,573. https://doi.org/10.1029/97JC00015 de Vernal, A. & Hillaire-Marcel, C., 2008. Natural Variability of Greenland Climate, Vegetation, and Ice Volume During the Past Million Years. Science, 320 (5883), 1622-1625. https://doi.org/10.1126/science.1153929

Drenzek, N.J., Montlucon, D.B., Yunker, M.B., Macdonald, R.W., & Eglinton, T.I., 2007. Constraints on the origin of sedimentary organic carbon in the Beaufort Sea from coupled molecular 13C and 14C measurements. Mar. Chem., 103 (1-2), 146-162. https://doi.org/10.1016/j.marchem.2006.06.017

Droxler, A.W. & Farrell, J.W., 2000. Marine isotope stage 11 (MIS 11): new insights for a warm future. Glob. Planet. Change, 24 (1), 1-5. https://doi.org/10.1016/S0921-8181(99)00065- X

Droxler, A.W., Poore, R.Z., & Burckle, L.H., 2003. Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., 240 pp.

Dutton, A., Carlson, A.E., Long, A.J., Milne, G.A., Clark, P.U., DeConto, R., Horton, B.P., Rahmstorf, S., & Raymo, M.E., 2015. Sea-level rise due to polar ice-sheet mass loss during past warm periods. Science, 349 (6244), aaa4019. https://doi.org/10.1126/science.aaa4019

Eicken, H., Gradinger, R., Gaylord, A., Mahoney, A., Rigor, I., & Melling, H., 2005. Sediment transport by sea ice in the Chukchi and Beaufort seas: Increasing importance due to 134

changing ice conditions? Deep Sea Res. Part II: Top. Stud. Oceanogr., 52, 3281-3302. https://doi.org/10.1016/j.dsr2.2005.10.006

Elias, S. & Brigham-Grette, J., 2013. Late Pleistocene events in Beringia. Encyclopedia of Quaternary Science (Second Edition), 191-201. https://doi.org/10.1016/B978-0-444- 53643-3.00116-3

EPICA Community Members, 2004. Eight glacial cycles from an Antarctic ice core. Nature, 429, 623-628. https://doi.org/10.1038/nature02599

Ezer, T. & Oey, L.-Y., 2010. The role of the Alaskan Stream in modulating the Bering Sea climate. J. Geophys. Res. Oceans, 15, C04025. https://doi.org/10.1029/2009JC005830

Fohlmeister, J., 2012. A statistical approach to construct composite climate records of dated archives. Quat. Geochronol., 14, 48-56. https://doi.org/10.1016/j.quageo.2012.06.007

Fryxell, G.A. & Hasle, G.R., 1972. Thalassiosira eccentrica (Ehreb.) Cleve, T. symmetrica sp. nov., and some related centric diatoms. J. Phycol., 8 (4), 297-317.

Galbraith, E.D., Keinast, M., Kaccard, S.L., Pedersen, T.F., Brunelle, B.G., Sigman, D.M., & Kiefer, T., 2008. Consistent relationship between global climate and surface nitrate utilization in the western subarctic Pacific throughout the last 500 ka. Paleoceanography, 23 (2), PA2212. https://doi.org/10.1029/2007PA001518

Glushkova, O.Y., 2001. Geomorphological correlation of Late Pleistocene glacial complexes of Western and Eastern Beringia. Quat. Sci. Rev., 20 (1-3), 405-417. https://doi.org/10.1016/S0277-3791(00)00108-6

Grebmeier, J., Overland, J., Moore, S., Farley, E., Carmack, E., Cooper, L., Frey, K., Helle, J., Mclaughlin, F., & Mcnutt, L., 2006. A Major Ecosystem Shift in the Northern Bering Sea. Science, 311 (5766), 1461-1464. https://doi.org/10.1126/science.1121365

Guillard, R.R.L. & Ryther, J.H., 1962. Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt, and Detonula confervacaea (cleve) Gran. Can. J. Microbiol., 8 (2), 229-239. https://doi.org/10.1139/m62-029

Harbour, T.A., 2019. Arctic sedimentation on interannual and million year time scales: Laminations and ice rafted debris. Masters Thesis, Iowa State University, 17022. https://lib.dr.iastate.edu/etd/17022 135

Holland, M.M., Bitz, C.M. & Tremblay, B., 2006. Future abrupt reductions in the summer Arctic sea ice. Geophys. Res. Lett., 33 (23), 1-5. https://doi.org/10.1029/2006GL028024

Horner, R.A., 1985. Sea-Ice Biota. CRC Press, Boca Raton, FL., 215 pp.

Hu, A.X., Meehl, G.A., Han, W.Q., Otto-Bliestner, B., Abe-Ouchi, A., & Rosenbloom, N., 2015. Effects of the Bering Strait closure on AMOC and global climate under different background climates. Prog. Oceanogr., 132, 174-196. https://doi.org/10.1016/j.pocean.2014.02.004 IPCC, 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (Eds.)]. IPCC, Geneva, Switzerland, 151 pp.

Kanaya, T. & Koizumi, I., 1966. Interpretation of diatom thanatocoenoses from the North Pacific applied to a study of core V20-130. Tohoku Univ Sci Reports (Series 2), 37 (2), 89-130.

Kandiano, E.S. & Bauch, H.A., 2007. Phase relationship and surface water mass change in the NorthEast Atlantic during Marine Isotope Stage 11 (MIS 11). Quat. Res., 68 (3), 445-455. https://doi.org/10.1016/j.yqres.2007.07.009

136

Katsuki, K., Khim, B.K., Itaki, T., Harada, N., Sakai, H., Ikeda, T., Takahashi, K., Okazaki, Y., & Asahi, N., 2009. Land-sea linkage of Holocene paleoclimate on the Southern Bering Continental Shelf. The Holocene, 19 (5), 745-754. https://doi.org/10.1177/0959683609105298

Kemp, A.E.S., 1996. Laminated sediments as palaeo-indicators. Geol. Soc. London, Spec. Publ., 116, 7-10. https://doi.org/10.1144/GSL.SP.1996.116.01.01

Koizumi, I., 1973. The Late Cenozoic diatoms of Sites 183-193, Leg 19 Deep-Sea Drilling Project. Initial Reports of the Deep-Sea Drilling Project, 19, 805-855.

137

Koutsodendris, A., Kousis, I., Peyron, O., Wagner, B., & Pross, J., 2019. The Marine Isotope Stage 12 pollen record from Lake Ohrid (SE Europe): Investigating short-term climate change under extreme glacial conditions. Quat. Sci. Rev., 221, 105873. https://doi.org/10.1016/j.quascirev.2019.105873

Ladd, C., Hunt, G.L., Jr., Mordy, C.W., Salo, S.A., & Stabeno, P.J., 2005. Marine environment of the eastern and central Aleutian Islands. Fish. Oceanogr., 14 (s1), 22-38. https://doi.org/10.1111/j.1365-2419.2005.00373.x

Lambeck, K., Yokoyama, Y., & Purcell, T., 2002. Into and Out of the Last Glacial Maximum: Sea- level Change during Oxygen Isotope Stages 3 and 2. Quat. Sci. Rev., 21, 343-360. https://doi.org/10.1016/S0277-3791(01)00071-3

Laskar, J. Fienga, A., Gastineau, M., & Manche, H., 2011. La2010: a new orbital solution for the long-term motion of the Earth. Astron. Astrophys., 532, A89. https://doi.org/10.1051/0004-6361/201116836

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., Joutel, F., & Levrard, B., 2004. A long-term numerical solution for the insolation qualities of the Earth. Astron. Astrophys., 428 (1), 261-285. https://doi.org/10.1051/0004-6361:20041335

Lein, A.Y., 2004. Authigenic Carbonate Formation in the Ocean. Lithol. Miner. Resour., 39, 1-30. https://doi.org/10.1023/B:LIMI.0000010767.52720.8f

Leu, E., Mundy, C., Assmy, P., Campbell, K., Gabrielsen, T., Gosselin, M., Juul-Pedersen, T., & Gradinger, R., 2015. Arctic spring awakening - Steering principles behind the phenology of vernal ice algal blooms. Prog. Oceanogr., 139, 151-170. https://doi.org/10.1016/j.pocean.2015.07.012

Leventer, A., Domack, E., Barkoukis, A., McAndrews, B., & Murray, J., 2001. Laminations from the Palmer Deep: A diatom-based interpretation. Paleoceanogr. Paleoclimatol., 17 (3). https://doi.org/10.1029/2001PA000624

Lisiecki, L. E. & Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20, PA1003. https://doi.org/10.1029/2004PA001071

138

Lopes, C. & Mix, A.C., 2018. North Pacific Paleotemperature and Paleoproductivity Reconstructions Based on Diatom Species. Paleoceanogr. Paleoclimatol., 33 (7), 703- 715. https://doi.org/10.1029/2018PA003352

Loutre, M.F. & Berger, A., 2003. Marine Isotope Stage 11 as an analogue for the present interglacial. Glob. Planet. Change, 36, 209-217. https://doi.org/10.1016/S0921-8181(02)00186-8

Lundholm, N. & Hasle, G.R., 2010. Fragilariopsis (Bacillariophyceae) of the Northern Hemisphere - Morphology, taxonomy, phylogeny and distribution, with a description of F. pacifica sp. nov. Phycologia, 49 (5), 438-460. https://doi.org/10.2216/09-97.1

Lüthi, D., le Floch, M., Bereiter, B., Blunier, T., Barnola, J.M., Siegenthaler, U., Raynaud, D., Jouzel, J., Fischer, H., Kawamura, K., & Stocker, T.F., 2008. High-resolution carbon dioxide concentration record 650,000–800,000 years before present. Nature, 453 (7193), 379-382. https://doi.org/10.1038/nature06949 Masson-Delmotte, V., Dreyfus, G., Braconnot, P., Johnsen, S., Jouzel, J., Kageyama, M., Landais, A., Loutre, M.-F., Nouet, J., Parrenin, F., Raynaud, D., Stenni, B., & Tuenter, E., 2006. Past temperature reconstructions from deep ice cores: relevance for future climate change. Clim. Past, 2, 399-448. https://doi.org/10.5194/cp-2-145-2006

Maurer, B.A. & McGill, B.J., 2011. Chapter 5: Measurement of species diversity. In: Magurran, A. & McGill, B.J. (Eds.), Biological Diversity: Frontiers in Measurement and Assessment. Oxford University Press, New York, pp. 55-65.

McGee, D., Broecker, W., & Winckler, G., 2010. Gustiness: The driver of glacial dustiness? Quat. Sci. Rev., 29 (17-18), 2340-2350. https://doi.org/10.1016/j.quascirev.2010.06.009

McManus, J.F., Oppo, D.W., & Cullen, J.L., 1999. A 0.5-million-year record of millennial-scale climate variability in the North Atlantic. Science, 283 (5404), 971-975. https://doi.org/10.1126/science.283.5404.971

Medlin, L.K. & Priddle, J., 1990. Polar Marine Diatoms. British Antarctic Survey, Natural Environment Research Council, Cambridge, 214 pp. 139

Meyers, P.A., 1994. Preservation of elemental and isotopic source identification of sedimentary organic matter. Chem. Geol., 114 (3-4), 289-302. https://doi.org/10.1016/0009-2541(94)90059-0

Miettinen, A., Koç, N., & Husum, K., 2013. Appearance of the Pacific diatom Neodenticula seminae in the northern Nordic Seas — An indication of changes in Arctic sea ice and ocean circulation. Mar. Micropaleontol., 99, 2-7. https://doi.org/10.1016/j.marmicro.2012.06.002

Mordy, C.W., Stabeno, P.J., Ladd, C., Zeeman, S., Wisegarver, D.P., Salo, S.A., & Hunt, G.L., Jr., 2005. Nutrients and primary production along the eastern Aleutian Island Archipelago. Fish. Oceanogr., 14 (s1), 55-76. https://doi.org/10.1111/j.1365-2419.2005.00364.x

Nesterovich, A., 2019. Quantitative diatom-based proxy for sea ice extent in the Bering and Chukchi seas. PhD Thesis, Iowa State University, 17277. https://lib.dr.iastate.edu/etd/17277

Nesterovich, A. & Caissie, B.E., in review. A diatom-based quantitative sea ice proxy for the Bering and Chukchi Seas. Prog. Oceanogr.

Niebauer, H.J., Alexander, V., & Henrichs, S.M., 1990. Physical and biological oceanographic interaction in the spring bloom at the Bering Sea marginal ice edge zone. J. Geophys. Res., 95, 22,229-22,241. https://doi.org/10.1029/JC095iC12p22229

Niebauer, H.J., Bond, N.A., & Plotnikov, V.V., 1999. An update on the climatology and sea ice of the Bering Sea. In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 29-60.

Nørgaard-Pedersen, N., Spielhagen, R.F., Erlenkeuser, H., Grootes, P.M., Heinemeier, J., & Knies, J., 2003. Arctic Ocean during the Last Glacial Maximum: Atlantic and polar domains of surface water mass distribution and ice cover. Paleoceanography, 18 (3), 1063. https://doi.org/10.1029/2002PA000781

140

Onodera, J. & Takahashi, K., 2007. Diatoms and siliceous flagellates (silicoflagellates, ebridians, and endoskeletal dinoflagellate Actiniscus) from the Subarctic Pacific. Bulletin of the Faculty of Sciences, Kyushu University: Series D, Earth and Planetary Sciences, 31 (4), 105-136.

Osborne, E., Richter-Menge, J., & Jeffries, M. (Eds.), 2018. Arctic Report Card 2018. https://arctic.noaa.gov/Report-Card/Report-Card-2018

Overland, J.E. & Pease, C.W., 1982. Cyclone Climatology of the Bering Sea and Its Relation To Sea Ice Extent. Mon. Weather Rev., 110, 5-13. https://doi.org/10.1175/1520-0493(1982)110%3C0005:CCOTBS%3E2.0.CO;2 Overland, J.E. & Wang, M., 2013. When will the summer Arctic be nearly sea ice free? Geophys. Res. Lett., 40 (10), 2097-2101. https://doi.org/10.1002/grl.50316

Past Interglacials Working Group of PAGES, 2016. Interglacials of the last 800,000 years. Rev. Geophys., 54 (1), 162-219. https://doi.org/10.1002/2015RG000482

Pickart, R.S., Moore, G.W.K., MacDonald, A.M., Renfrew, I.A., Walsh, J.E., & Kessler, W.S., 2009. Seasonal evolution of Aleutian low pressure systems: Implications for the North Pacific 141

subpolar circulation. J. Phys. Oceanogr., 39, 1317-1339. https://doi.org/10.1175/2008JPO3891.1

Pike, J. & Kemp, A.E.S., 1999. Diatom mats in Gulf of California sediments: Implications for the paleoenvironmental interpretation of laminated sediments and silica burial. Geology, 27 (4), 311-314. https://doi.org/10.1130/0091-7613(1999)027%3C0311:DMIGOC%3E2.3.CO;2

Rainville, L., Lee, C., & Woodgate, R., 2011. Impact of Wind-Driven Mixing in the Arctic Ocean. Oceanography, 24, 136-145. https://doi.org/10.5670/oceanog.2011.65

Raymo, M.E. & Mitrovica, J.X., 2012. Collapse of polar ice sheets during the stage 11 interglacial. Nature, 483 (7390), 453-456. https://doi.org/10.1038/nature10891

Redfield, A.C., Ketchum, B.H., & Richards, F.A., 1963. The influence of organisms on the composition of seawater. In: Hill, M.N. (Ed.), The sea: ideas and observations on progress in the study of the seas. Wiley Interscience, New York, pp. 26-77.

142

Rodionov, S.N., Bond, N.A., & Overland, J.E., 2007. The Aleutian Low, storm tracks, and winter climate variability in the Bering Sea. Deep Sea Res. Part II: Top. Stud. Oceanogr., 54 (23- 26), 2560-2577. https://doi.org/10.1016/j.dsr2.2007.08.002

Rodrigues, T., Alonso-García, M., Hodell, D.A., Rufino, M., Naughton, F., Grimalt, J., Voelker, A., & Abrantes, F., 2017. A 1-Ma record of sea surface temperature and extreme cooling events in the North Atlantic: A perspective from the Iberian Margin. Quat. Sci. Rev., 172, 118-130. https://doi.org/10.1016/j.quascirev.2017.07.004

Rogers, J.C., 1981. The north Pacific oscillation. J. Climatol., 1, 39-57. https://doi.org/10.1002/joc.3370010106

Rohling, E.J., Fenton, M., Jorissen, F.J., Bertrand, P., Ganssen, G., & Caulet, J.P., 1998. Magnitudes of sea-level lowstands of the past 500,000 years. Nature, 394, 162- 165. https://doi.org/10.1038/28134

Rohling, E.J., Foster, G.L., Grant, K.M., Marino, G., Roberts, A.P., Tamisiea, M.E., & Williams, F., 2014. Sea-level and deep-sea-temperature variability over the past 5.3 million years. Nature 508, 477-482 (2014). https://doi.org/10.1038/nature13230

Saitoh, K. & Taniguchi, A., 1978. Phytoplankton communities in the Bering Sea and adjacent seas II: Spring and summer communities in seasonally ice-covered areas. Astarte, 11 (1), 27-35.

Sakamoto, T., Ikehara, M., Aoki, K., Iijima, K., Nimurra, K., Nakatsuka, T., & Wakatsuchi, M., 2005. Ice-rafted debris (IRD)-based sea-ice expansion events during the past 100 kyrs in the Okhotsk Sea. Deep Sea Res. Part II: Top. Stud. Oceanogr., 52 (16-18), 2275-2301. https://doi.org/10.1016/j.dsr2.2005.08.007

Sancetta, C., 1979. Use of semiquantitative microfossil data for paleoceanography. Geology, 7 (2), 88-92. https://doi.org/10.1130/0091-7613(1979)7%3C88:UOSMDF%3E2.0.CO;2

Sancetta, C.A., 1982. Distribution of diatom species in surface sediments of the Bering and Okhotsk seas. Micropaleontology, 28 (3), 221-257.

143

Sancetta, C., & Robinson, S.W., 1983. Diatom evidence on Wisconsin and Holocene events in the Bering Sea. Quat. Res., 20, 232-245. https://doi.org/10.1016/0033-5894(83)90079-0

Sancetta, C. & Silvestri, S., 1984. Diatom stratigraphy of the late Pleistocene (Brunhes) subarctic Pacific. Mar. Micropaleontol., 9 (3), 263-274. https://doi.org/10.1016/0377-8398(84)90016-1

Schimmelmann, A., Lange, C., Schieber, J., Francus, P., Ojala, A., & Zolitschka, B., 2016. Varves in marine sediments: A review. Earth Sci. Rev., 159, 215-246. https://doi.org/10.1016/j.earscirev.2016.04.009

Schlung, S., Ravelo, A., Aiello, I., Andreasen, D., Cook, M., Drake, M., Dyez, K., Guilderson, T., Lariviere, J., Stroynowski, Z., & Takahashi, K., 2013. Millennial-scale climate change and intermediate water circulation in the Bering Sea from 90 ka: A high-resolution record from IODP Site U1340. Paleoceanography, 28, 54-67. https://doi.org/10.1029/2012PA002365

Scholl, D.W., Buffington, E.C., & Hopkins, M., 1968. Geologic history of the continental margin of North America in the Bering Sea. Mar. Geol., 6 (4), 279-330. https://doi.org/10.1016/0025-3227(68)90021-2 Schrader, H.J. & Gersonde, R., 1978. Diatoms and silicoflagellates. In: Zachariasse, W.J., Riedel, W.R., Sanfilippo, A., Schmidt, R.R., Brolsma, M.J., Schrader, H.J., Gersonde, R., Drooger, M.M., and Broekman, J.A. (Eds.), Micropaleontological counting methods and techniques - an exercise on an eight metres section of the Lower Pliocene of Capo Rossello, Sicily. Utrecht Micropaleontol. Bull., Netherlands, pp. 129-176.

Schubert, C.J. & Calvert, S.E., 2001. Nitrogen and carbon isotopic composition of marine and terrestrial organic matter in Arctic Ocean sediments: implications for nutrient utilization and organic matter composition. Deep Sea Res. Part I: Oceanogr. Res. Pap., 48 (3), 789- 810. https://doi.org/10.1016/S0967-0637(00)00069-8

Screen, J.A. & Deser, C., 2019. Pacific Ocean Variability Influences the Time of Emergence of a Seasonally Ice-Free Arctic Ocean. Geophys. Res. Lett., 46 (4), 2222-2231. https://doi.org/10.1029/2018GL081393

Shiga, K. & Koizumi, I., 1999. Latest Quaternary oceanographic changes in the Okhotsk Sea based on diatom records. Mar. Micropaleontol., 38 (2), 91-117. https://doi.org/10.1016/S0377-8398(99)00041-9

144

Sigman, D.M., Jaccard, S.L., & Haug, G.H., 2004. Polar ocean stratification in a cold climate. Nature, 428, 59-63. https://doi.org/10.1038/nature02357

Sigman, D.M., Altabet, M.A., McCorkle, D.C., Francois, R. & Fischer, G., 2000. The δ15N of nitrate in the Southern Ocean: nitrogen cycling and circulation in the ocean interior. J. Geophys. Res., 105 (C8), 19,599-19,614. https://doi.org/10.1029/2000JC000265

Sjunneskog, C., & Scherer, R., 1997. High-resolution diatom record in bioturbated antarctic sediments. Antarct. J. U.S., 32 (5), 34-36.

Spratt, R.M. & Lisiecki, L.E., 2016. A Late Pleistocene sea level stack. Clim. Past, 12, 1079-1092. https://doi.org/10.5194/cp-12-1079-2016

Stabeno, P.J., Duffy-Anderson, J.T., Eisner, L.B., Farley, E.V., Heintz, R.A., & Mordy, C.W., 2017. Return of warm conditions in the southeastern Bering Sea: Physics to fluorescence. PloS One, 12 (9), e0185464. https://doi.org/10.1371/journal.pone.0185464 Stabeno, P.J., Ladd, C., & Reed, R.K., 2009. Observations of the Aleutian North Slope Current, Bering Sea, 1996–2001. J. Geophys. Res. Oceans, 114, C05015. https://doi.org/10.1029/2007JC004705

Stabeno, P., Reed, R., & Napp, J., 2002. Transport through Unimak Pass, Alaska. Deep Sea Res. Part II: Top. Stud. Oceanogr., 49, 5919-5930. https://doi.org/10.1016/S0967-0645(02)00326-0

Stein, R., 2008. Arctic Ocean Sediments: Processes, Proxies, and Paleoenvironment. Elsevier, Oxford, UK, 592 pp.

Stein, R., Hefter, J., Grützner, J., Voelker, A., & Naafs, B.D.A., 2009. Variability of surface-water characteristics and Heinrich-like Events in the Pleistocene mid-latitude North Atlantic Ocean: Biomarker and XRD records from IODP Site U1313 (MIS 16 – 9). Paleoceanography, 24 (2), PA2203. https://doi.org/10.1029/2008PA001639

145

Stroeve, J. & Notz, D., 2018. Changing state of Arctic sea ice across all seasons. Environ. Res. Lett., 13 (10), 103001. https://doi.org/10.1088/1748-9326/aade56

Stroynowski, Z., Abrantes, F., & Bruno, E., 2017. The response of the Bering Sea Gateway during the Mid-Pleistocene Transition. Palaeogeogr. Palaeoclimatol. Palaeoecol., 485, 974-985. https://doi.org/10.1016/j.palaeo.2017.08.023

Takahashi, K., Ravelo, A.C., Alvarez Zarikian, C. A., and Expedition 323 Scientists, 2011. Proceedings of the Integrated Ocean Drilling Program, 323, Tokyo. doi:10.2204/iodp.proc.323.2011

Tremblay, J.-É., Anderson, L.G., Matrai, P., Coupel, P., Bélanger, S., Michel, C., & Reigstad, M., 2015. Global and regional drivers of nutrient supply, primary production and CO2 drawdown in the changing Arctic Ocean. Prog. Oceanogr., 139, 171-196. https://doi.org/10.1016/j.pocean.2015.08.009

Tye, G.J., Sherri, J., Candy, I., Coxon, P., Palmer, A., McClymont, E.L., & Schreve, D.C., 2016. The 18O stratigraphy of the Hoxnian lacustrine sequence at Marks Tey, Essex, UK: implications for the climatic structure of MIS 11 in Britain. J. Quat. Sci., 31 (2), 75-92. https://doi.org/10.1002/jqs.2840 van Geen, A., Zheng, Y., Bernhard, J.M., Cannariato, K.G., Carriquiry, J., Dean, W.E., Eakins, B.W., Ortiz, J.D., & Pike, J., 2003. On the preservation of laminated sediments along the western margin of North America. Paleoceanogr. Paleoclimatol., 18 (4), 1098. https://doi.org/10.1029/2003PA000911

Vancoppenolle, M., Bopp, L., Madec, G., Dunne, J., Ilyina, T., Halloran, P., & Steiner, N., 2013. Future Arctic Primary Productivity from CMIP5 Simulations: Uncertain Outcome, but Consistent Mechanisms. Global Biogeochem. Cy., 27 (3), 605-619. https://doi.org/10.1002/gbc.20055

Vogel, H., Meyer-Jacob, C., Melles, M., Brigham-Grette, J., Andreev, A.A., Wennrich, V., Tarasov, P. E., & Rosén, P., 2013. Detailed insight into Arctic climatic variability during MIS 11c at Lake El’gygytgyn, NE Russia. Clim. Past, 9, 1467-1479. https://doi.org/10.5194/cp-9- 1467-2013 von Quillfeldt, C.H., Ambrose, W.G., Jr., & Clough, L.M., 2003. High number of diatom species in first-year ice from the Chukchi Sea, Polar Biol., 26, 806-818. https://doi.org/10.1007/s00300-003-0549-1

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 (13), 1565-1579. https://doi.org/10.1016/j.csr.2009.04.006

Wang, M. & Overland, J.E., 2009. A sea ice free summer Arctic within 30 years? Geophys. Res. Lett., 36 (7), L07052. https://doi.org/10.1029/2009GL037820

Warnock, J. & Scherer, R., 2015. A revised method for determining the absolute abundance of diatoms. J. Paleolimnol., 53, 157-163. https://doi.org/10.1007/s10933-014-9808-0

Williams, B., Halfar, J., Steneck, R., Wortmann, U.G., Hetzinger, S., Adey, W., Lebednik, P., & Joachimski, M., 2011. Twentieth Century δ13C Variability in Surface Water Dissolved Inorganic Carbon Recorded by Coralline Algae in the Northern North Pacific Ocean and the Bering Sea. Biogeosciences, 8, 165-174. https://doi.org/10.5194/bg-8-165-2011

Wills, R.C.J., Battisti, D.S., Proistosescu, C., Thompson, L., Hartmann, D.L., & Armour, K.C., 2019. Ocean Circulation Signatures of North Pacific Decadal Variability. Geophys. Res. Lett., 46 (3), 1690-1701. https://doi.org/10.1029/2018GL080716

Witkowski, A., Lange-Bertalot, H., & Metzeltin, D., 2000. Diatom Flora of Marine Coasts I. Iconographia Diatomologica, 7. A.R.G. Gantner, Königstein, Germany, 925 pp.

Woodgate, R.A., Weingartner, T., & Lindsay, R., 2010. The 2007 Bering Strait oceanic heat 147

flux and anomalous Arctic sea-ice retreat. Geophys. Res. Lett., 37, L01602. https://doi.org/10.1029/2009GL041621

Worne, S., Kender, S., Swann, G., Leng, M., & Ravelo, A., 2019. Coupled climate and subarctic Pacific nutrient upwelling over the last 850,000 years. Earth Planet. Sci. Lett., 522, 87-97. https://doi.org/10.1016/j.epsl.2019.06.028

148

CHAPTER 5. REGIONAL VARIATIONS IN SEA ICE, PRIMARY PRODUCTIVITY, AND OCEAN CIRCULATION DURING MARINE ISOTOPE STAGE 11 IN THE BERING SEA

Modified from a manuscript to be submitted to Climate of the Past

Natalie S. Thompsona & Beth E. Caissiea,b

aDepartment of Geological & Atmospheric Science, Iowa State University, Ames, IA bCurrently at: US Geological Survey, Menlo Park, CA

Abstract

Marine Isotope Stage (MIS) 11 (424-374 ka) has long been considered one of the more suitable analogs for the natural development of Holocene climate, because it is the most recent interglacial period with similar orbital boundary conditions. However, there is significant global as well as regional variability in the response of the climate system to the interglacial warming of MIS 11. Here, we examine new and published sediment core records from three IODP sites

(U1345, U1343, and U1339) across the Bering Sea to investigate regional changes in paleoceanographic conditions during Marine Isotope Stages 12-10.

Sediment grain size, diatom assemblages, and a new diatom-based proxy for sea ice show that sea ice was present over the Bering slope and the Umnak Plateau (southeastern

Bering Sea) during MIS 11, but today, none of these sites are seasonally ice-covered. This suggests that sea ice regimes in the Bering Sea during MIS 11 were very different to those of the Holocene. We also note significant regional differences in the response of sea ice to MIS 11 warming. At the Umnak Plateau, sea ice concentrations decline during deglaciation, but remain relatively high at the slope sites until the interglacial climatic optimum (~ 410 ka). Sea ice re- advances over the Umnak Plateau during peak interglacial warmth, at the same time that it 149 declines over the slope sites. Late MIS 11 (398-374 ka) is characterized by relatively high concentrations of seasonal sea ice at the Umnak Plateau, whilst sea ice at the slope sites fluctuates between periods of consolidated (>40%) and unconsolidated (15-40%) ice cover. This east-west dichotomy may be explained by regional differences in sea level pressure and wind direction, which are driven, at least in part, by changes in the geographic position of the

Aleutian Low.

Introduction

In the face of rapid climate change, improving our knowledge of climate system dynamics has become ever more critical. Our understanding of climate during the Holocene

(Marine Isotope Stage 1) may be improved through the investigation of past interglacial climate analogs. Although the natural course of Holocene warming has been disrupted by anthropogenic activity, such that past interglacials cannot be used as straightforward analogs for the present climate, several interglacials of the past 800 ka warrant further study as partial analogs for the evolution of Holocene climate (Yin & Berger, 2015; Past Interglacials Working

Group of PAGES, 2016). These include the Marine Isotope Substages 5e (Kukla et al., 1997), 9e

(Ruddiman, 2007), 11c (Droxler et al., 2003; Loutre & Berger, 2003; McManus et al., 2003;

Kandiano et al., 2012; Candy et al., 2014; Yin & Berger, 2015; Cronin et al., 2019) and 19c

(Tzedakis et al., 2009, 2012; Pol et al., 2010; Tzedakis, 2010; Yin & Berger, 2015; Vavrus et al.,

2018).

In particular, Marine Isotope Stage (MIS) 11 (424-374 ka) has long been considered one of the more suitable analogs for pre-industrial warming, because it is the most recent 150 interglacial period with high greenhouse gas concentrations (Fig. 5.1; Petit et al., 1999;

Siegenthaler et al., 2005; Lüthi et al., 2008; Bereiter et al., 2015) and orbital parameters similar to modern (Fig. 5.1; Berger and Loutre, 2002, 2003; Loutre and Berger, 2003). Globally, MIS 11 is recognized as an interval of prolonged warmth, with reduced global ice volume, and sea level estimated in the range of 6-13 m above present sea level (a.p.s.l.) (Fig. 5.1; Lisiecki & Raymo,

2005; de Vernal & Hillaire-Marcel, 2008; Raymo & Mitrovica, 2012). However, there is considerable regional variability in the expression of MIS 11. For example, during the interglacial climatic optimum, hereafter referred to as MIS 11c, pollen records suggest that boreal forest covered at least the southern part of Greenland, which may have been largely ice- free (de Vernal & Hillaire-Marcel, 2008; Raymo & Mitrovica, 2012; Robinson et al., 2017). In addition, sediments from Lake El'gygytgyn in northeastern Russia signify that maximum summer temperatures were 4-5°C higher than during MIS 1 (Melles et al., 2012), and sea surface temperature (SST) reconstructions from the subtropical Northeast Atlantic and the western Mediterranean indicate that average SSTs were 1-2°C warmer than today (Kandiano et al., 2012). In contrast, other paleoclimate records (e.g., McManus et al., 1999, 2003; Bauch et al., 2000; Hodell et al., 2000) suggest that MIS 11c was no warmer than the present day, whilst

Kandiano et al. (2012, 2016) showed evidence for colder SSTs in the polar Nordic Seas throughout the entirety of MIS 11. Clearly, the response of the climate system to interglacial warming during MIS 11 was highly variable, which highlights the importance of investigating past interglacials at a regional scale, so as to ensure that specific regional effects do not go unnoticed. 151

Abundant high resolution paleoclimate records from the North Atlantic (e.g., Bauch et al., 2000; Poli et al., 2000; de Abreu et al., 2005; Kandiano & Bauch, 2007; Lawrence et al.,

2009, 2010; Voelker et al., 2010; Rodrigues et al., 2011; Bauch, 2012; Kandiano et al., 2012,

2016, 2017; Candy & McClymont, 2013; Palumbo et al., 2013) have allowed for in-depth investigations of past oceanographic conditions during MIS 11 in this region. In contrast, there are relatively few MIS 11 records from the North Pacific and its marginal seas (Caissie et al.,

2016). In this study, we examine new and published sediment core records from three sites across the Bering Sea (Fig. 5.2) to investigate regional changes in paleoceanographic conditions during the MIS 12-10 cycle. Specifically, we use a multi-proxy approach to reconstruct past changes in sea ice, primary productivity, and ocean currents, and to place these changes in both a global and a regional context. Our primary objective is to fill a major gap in records of MIS 11 climate from the Bering Sea. To this end, we address the following questions: What were sea ice conditions during MIS 11, and how do they compare to modern (Holocene) sea ice conditions in the Bering Sea? Do the three sites show different regional responses to interglacial warming during MIS 11? How do paleoceanographic conditions in the Bering Sea compare to those in other regions, including the northern North Atlantic, the Nordic Seas, and the Arctic Ocean, and further, how does MIS 11 in the Bering Sea compare to the global signal?

152

153

Background

Marine Isotope Stage 11

Marine Isotope Stage 11 and the Holocene share a number of similarities, including similar orbital configurations, namely, low eccentricity, dampened precession, and low amplitude changes in boreal summer insolation (Fig. 5.1; Berger & Loutre, 2002; Droxler et al.,

2003; Loutre & Berger, 2003; Laskar et al., 2004, 2011). Atmospheric CO2 concentrations during

MIS 11 averaged 275 ppm, comparable to pre-industrial levels of pCO2 (Fig 5.1; Petit et al.,

1999; Siegenthaler et al., 2005; Lüthi et al., 2008; Bereiter et al., 2015). Relatively high pCO2 during MIS 11 resulted in an exceptionally long warm period (Fig. 5.1; Berger & Loutre, 2002;

Lisiecki and Raymo, 2005; Rohling et al., 2010; Melles et al., 2012), with peak interglacial warmth (MIS 11c) lasting in the order of 20-30 ka (Droxler et al., 2003; McManus et al., 2003;

Raynaud et al., 2005; Kandiano et al., 2012; Candy et al., 2014; Past Interglacial Working Group of PAGES, 2016), longer than other Quaternary interglacials, and considerably longer than the elapsed duration of the Holocene (Past Interglacial Working Group of PAGES, 2016). Likewise, high concentrations of anthropogenic CO2 have already extended the natural duration of the

Holocene (Tzedakis et al., 2012; IPCC, 2014; Candy et al., 2014; Ganopolski et al., 2016), and climate models indicate that the next glacial inception may now not occur for tens of millennia

(Ganopolski et al., 2016; Past Interglacial Working Group of PAGES, 2016). Although we cannot expect MIS 11 and Holocene climate to evolve in exactly the same way, MIS 11 records nonetheless provide a unique opportunity to investigate how the climate system operates during an extended interglacial period. 154

As previously described, the magnitude and spatial pattern of warming during MIS 11 is highly complex (Candy et al., 2014; Past Interglacials Working Group of PAGES, 2016). The history of MIS 11 in the Bering Sea is not well constrained, because to date there have been very few high-resolution paleoclimate studies of this region. However, existing records indicate a unique response to MIS 11 warming in Beringia; tidewater glaciers advanced whilst eustatic sea level was still high, an example of the ‘out-of-phase glaciations’ peculiar to Beringia

(Brigham-Grette, 2001; Caissie et al., 2016), and sea ice in the northern Bering Sea was more prevalent than it is today (e.g., Stroynowski et al., 2015; Caissie et al., 2016; Onodera et al.,

2016; Teraishi et al., 2016).

Study Area

Location and Oceanographic Setting

The Bering Sea, located between Russia and Alaska, is a marginal sea of the North

Pacific. It comprises a broad (>500 km), shallow continental shelf in the east, a narrow (<100 km) shelf to the west, a steep continental slope, and a deep abyssal basin, dissected by several large submarine rises (Fig. 5.2; Ben-Avraham & Cooper, 1981; Stabeno et al., 1999).

Today, only the northern continental shelf is seasonally ice covered, but the location of the sea ice margin has fluctuated considerably over glacial-interglacial cycles (e.g., Sancetta et al., 1985; Kaufman and Brigham-Grette, 1993; Cook et al., 2005; Katsuki and Takahashi, 2005;

Caissie et al., 2010, 2016; Onodera et al., 2016; Méheust et al., 2018; Pelto et al., 2018). In the northernmost part of the Bering Sea, the shallow (-50 m a.p.s.l.) Bering Strait serves as the sole gateway between the Pacific and Arctic Oceans (Fig. 5.2). The closure of the Bering Strait at 155 times of lower sea level would have had significant impacts on global patterns of ocean circulation, nutrient and salinity distributions (e.g., Okazaki et al., 2010; Hu et al., 2012, 2015;

Knudson and Ravelo, 2015; Caissie et al., 2016; Worne et al., 2019). Water mass exchange with the North Pacific occurs via passes in the Aleutian Islands, linking Bering Sea conditions to those of the Pacific (Stabeno et al., 1999). Surface water masses are derived from inflow of the Alaska

Coastal Current and the Alaskan Stream. A part of the Alaskan Stream flows eastward along the north slope of the Aleutian Islands, becoming the Aleutian North Slope Current (ANSC). The northeastward flow of the ANSC turns northwestward as it nears the shallow continental shelf to form the Bering Slope Current (BSC), which flows along the continental slope and supplies critical nutrients to the shelf (Fig. 5.2). These nutrient-rich surface waters contribute to the extremely high rates of biological productivity in the Bering Sea, one of the most productive marine ecosystems in the world (Loughlin et al., 1999; Sigler et al., 2010; Brown and Arrigo,

2012, 2013; Stabeno et al., 2019). The highest rates of primary production occur within the

Bering Sea ‘Green Belt’ along the shelf-slope break, where shelf edge processes cause stratification, upwelling, and onwelling of nutrient-rich waters (Springer et al., 1996).

In this study, we compare both new and published sediment core records from

Integrated Ocean Drilling Program (IODP) Site U1343 (Fig. 5.2; Chapter 3) with two previously published records from sites U1345 (Fig. 5.2; Caissie et al., 2016) and U1339 (Fig. 5.2; Chapter 3;

Chapter 4). These cores were retrieved during IODP Expedition 323 to the Bering Sea in 2009

(Takahashi et al., 2011). All three sites are located on topographic highs, to minimize the risk of downslope transport. The north to southeast transect of the core sites allows us the 156 opportunity to compare changes in paleoceanographic conditions along the Bering slope with those in the southeastern Bering Sea.

157

Core Sites

Site U1345

Site U1345 (Fig. 5.2) is the northernmost of the three core sites. It is located on an interfluve ridge near the shelf-slope break, just south of Navarin Canyon, one of the largest submarine canyons in the world (Normack & Carlson, 2003). Sediments were retrieved from a water depth of 1008 m, placing this site within the modern-day oxygen minimum zone (OMZ)

(Roden, 2000; Takahashi et al., 2011). Site U1345 is closest to the maximum extent of present- day seasonal sea ice cover (Fig. 5.2), and is situated within the highly productive Green Belt

(Takahashi et al., 2011).

Site U1343

Site U1343 is situated on a topographic high (-1953 m) on the northern slope, isolated from the Bering Shelf by the Zhemchug Canyon, the world’s deepest submarine canyon

(Normack & Carlson, 2003). Like U1345, this site is located relatively close to the modern seasonal sea ice limit, and within the Bering Sea Green Belt (Fig. 5.2; Takahashi et al., 2011).

Site U1339

Site U1339 is located in the southeastern Bering Sea, on the northwest side of Umnak

Plateau, at a water depth of 1870 m. The Umnak Plateau is separated from the continental shelf by the Bering Canyon, the longest canyon in the Bering Sea (Normack & Carlson, 2003).

Today, this site is influenced by the relatively warm Alaskan Stream, which inhibits the formation of sea ice in the region. However, there is evidence for substantial sea ice cover at

Site U1339 in the past (e.g., Sancetta et al., 1985; Katsuki & Takahashi, 2005; Caissie et al.,

2010; Pelto et al., 2018). 158

Materials and Methods

Age Models

The shipboard age model, derived from bio- and magnetostratigraphy, was used to place the sediment core records in the correct stratigraphic position (Takahashi et al., 2011).

Published benthic foraminiferal oxygen isotope measurements (Asahi et al., 2016; Cook et al.,

2016) were used to align the records to the Lisiecki & Raymo (2005) global stack (LR04), and dates for each sample were obtained using linear interpolation between tie points (Fig. 5.3).

Site U1343 is well aligned to the LR04 stack during MIS 11 (Fig. 5.3; Asahi et al., 2016), but U1345 and U1339 are only tied at the beginning and end of MIS 11, which makes it difficult to constrain the exact timing of peak interglacial conditions. Caissie et al. (2016) added an extra tie point to the U1345 record by connecting inflection points in magnetic susceptibility between

U1343 and U1345, which shifted the age model 1.5 ka younger, and allowed for more 159 confidence in the timing of peak interglacial conditions. We applied this technique to U1339 but were unable to match the magnetic susceptibility records (Appendix A, Fig. A.1), so we use the original age model from Cook et al. (2016) in this work. Taking into account oxygen isotope sampling resolution, as well as the stated error in the LR04 stack of 4 kyr (Lisiecki & Raymo,

2005), uncertainty in the age model could be up to 5 kyr.

Diatom Counts

Diatom counts were carried out on 32 samples (~1,500 year resolution) from Site

U1343. Quantitative diatom slides were prepared following the revised methods of Warnock and Scherer (2015) and counted based on the methods of Schrader & Gersonde (1978). At least

300 diatom valves in at least 3 random transects were counted, using a Nikon ECLIPSE Ni transmitted light microscope at a magnification of 1000x. Transect lengths were measured with a stage micrometer. Only the 5 species used in a quantitative, diatom-based proxy for spring

(March, April, May, June) sea ice concentration (Fragilariopsis cylindrus, Fragilariopsis regina- jahniae, Neodenticula seminae, Paralia sulcata, and Sinerima marigela) were identified, whilst all other valves were recorded as either Chaetoceros resting spores (RS), centric diatoms, or pennate diatoms. The sea ice proxy is a Generalized Additive Model based on the distribution of the above five species in surface sediments from the Bering and Chukchi Seas, each of which has a unique association with sea ice (Nesterovich & Caissie, in review). Spring sea ice concentrations were obtained by uploading the diatom counts to the web app at http://seaiceproxy.geol.iastate.edu. 160

Published Records

In addition to the original analyses presented here, we refer to shipboard core descriptions (Takahashi et al., 2011) and smear slide counts (Chapter 3) in our interpretations.

These analyses were used as a basis to characterize primary sediment lithologies and elucidate the relationship between sediment composition and grain size.

Grain size records were previously published in Chapter 3. Grain size was measured on bulk sediments, and on the terrigenous fraction of the sediments, after removing organics, carbonates, and biogenics (see Chapter 3 for detailed methodology). Note that bulk grain size for Site U1345 was first published by Caissie et al. (2016).

Diatom records are based on published counts from sites U1345 (Caissie et al., 2016) and U1339 (Chapter 4). Sea ice concentrations for these sites were obtained by uploading the counts to the sea ice proxy (Nesterovich, 2019; Nesterovich & Caissie, in review; Chapter 4).

Results and Discussion

Lithology

Sediment Composition

Across the Bering Sea, sediments are primarily a mix of siliciclastic and biogenic particles

(Takahashi et al., 2011; Aiello & Ravelo, 2012). The siliciclastic sediments mostly consist of clay- to silt-sized grains, whilst biogenic materials largely comprise siliceous diatom frustules, with lesser contributions from other microfossils. Secondary components of the sediment include volcanogenic material and ice-rafted material (Takahashi et al., 2011; Chapter 3). Volcanogenic 161 material (mostly fine ash) is most abundant at U1339, and decreases with distance from the main volcanic sources in the Aleutian Arc (Takahashi et al., 2011).

Sediments are more siliciclastic at the northernmost Site U1345, because this site is closest to the continental shelf (Fig. 5.4). In contrast, sediments at the Umnak Plateau site

(U1339) are more biogenic, alternating for the most part between diatom ooze (>60% diatoms) and diatom mud (40-60% diatoms) (Fig. 5.4). Sediments at U1343 represent something of an 162 intermediate composition between U1345 and U1339 (Fig. 5.4; Takahashi et al., 2011; Chapter

3). It should be noted that the higher relative % of biogenic sedimentation at U1339 does not necessarily imply higher rates of productivity at this site. In fact, productivity is probably lowest at Site U1339, because the two slope sites are located in the highly productive Green Belt. At present, primary production in the Green Belt is 60% greater than on the adjacent outer shelf

(Springer et al., 1996). Assuming that oceanographic conditions were similar during MIS 11, we would expect higher productivity at sites U1345 and U1343. Therefore, the difference in diatom content between the sites is probably related to dilution by lithogenic material (e.g., Aiello et al., 2012; Kanematsu et al., 2013). There is also the possibility that more extensive sea ice cover at these sites (especially U1345, which is the most northerly) limits biological productivity beneath the ice.

Table 5.1. Average median grain size of bulk and terrigenous sediments from IODP sites U1345 (Caissie et al., 2016; Chapter 3), U1343, and U1339 (Chapter 3).

Median grain size (µm) Median grain size (µm) (bulk sediment) (terrigenous sediment)

Site Full record MIS 12 MIS 11 MIS 10 Full record MIS 12 MIS 11 MIS 10 U1345 20.8 15.7 22 18.1 21.7 15.2 23.7 13.4 U1343 17.3 11.9 18.2 12.4 17.4 12.3 18.3 12.4 U1339 19.7 13.7 20.3 17.7 12.6 8.7 12.7 10.9

Sediment Grain Size

Overall, the average median grain size of bulk sediments is highest at Site U1345 (20.8

µm), and lowest at Site U1343 (17.3 µm) (Table 5.1; Caissie et al., 2016; Chapter 3). For the terrigenous fraction, average median grain size is highest at Site U1345 (21.7 µm), and lowest at 163

U1339 (12.6 µm) (Table 5.1; Chapter 3). These trends reflect differences in the relative contribution of siliciclastic and biogenic input between the sites as described above (Chapter 3).

Bulk median grain size at all sites is higher during MIS 11, relative to the glacial stages 12 and 10 (Table 5.1; Fig. 5.5). On glacial-interglacial time scales, an increase in bulk grain size can be partly explained by the generally higher rates of primary productivity during interglacials

(Aiello & Ravelo, 2012; Chapter 3). The median grain size of terrigenous sediments is also higher 164 during MIS 11, although less obviously so at Site U1339 (Table 5.1; Fig. 5.5). This could be attributed to one of several factors, including increased sea ice melt and subsequent deposition of entrained sediments, remobilization of shelf sediments due to deglacial flooding of the shelf, or a change in sea ice sediment entrainment processes.

For the most part, silt (2-63 µm) is the predominant grain size, reflecting the dominant input of both silt-sized siliciclastics and silt-sized diatoms (Takahashi et al., 2011; Chapter 3).

Sand (63-2000 μm) is the second most abundant size fraction, followed by clay (<2 μm) (Fig.

5.5; Chapter 3). At all three sites, terrigenous clay is most abundant during the glacial stages

(MIS 12 and 10), with clay-sized grains at a minimum during MIS 11c (Table 5.1; Fig. 5.5). It has been shown that past glacial periods were both drier and dustier (McGee et al., 2010).

Accordingly, the glacial increase in clay likely reflects enhanced aeolian transport during MIS 12 and 10.

Coarse Interval

Anomalously coarse sediments are deposited around 403.5 ka at sites U1345 and U1343

(Fig. 5.5). These sediments form part of a bi-gradational sequence, with a coarsening upward unit from 408-403.5 ka, followed by a fining upward unit from 403.5 to 400 ka. In addition, sand mottles, sand lenses, sandy pockets, and indistinct or diffuse dark blue laminations are scattered throughout both cores (Takahashi et al., 2011). These sedimentary structures are associated with bottom current strength, and may provide evidence for alternating flow conditions along the Bering slope (Rebesco et al., 2014). Based on the grain size distribution, we interpret the coarse interval as a mud-sand contourite facies (Gonthier et al., 1984; Rebesco et al., 2014; Stow & Smillie, 2020; Chapter 3). Contourites are sediments deposited or 165 substantially reworked by the action of subsurface contour currents, and are ubiquitous features of ocean basins. To date, contourites have not been widely described in the Bering Sea

(Rebesco et al., 2014), although Pelto et al. (2018) noted a possible contourite in Holocene sediments from Site 3JPC on the northern Bering slope (Fig. 5.2). They suggested that contourites might form at this site if the Bering Slope Current increased in strength in line with increasing northward flow.

There are potential problems associated with using contourites in paleoclimate studies, including that contourites may be associated with winnowing of fine sediments, input of more distal coarse sediments, or both, and also that contour currents are capable of reworking turbidite deposits (Veeken & van Moerkerken, 2013; Stow & Smillie, 2020). For these reasons, we do not feel confident using grain size records from the contourite facies in our paleoclimate interpretations.

Laminated Sediments

Laminated sediments are widespread features of deglaciation in the North Pacific and its marginal seas (e.g., Zheng et al., 2000; Cook et al., 2005; Itaki et al., 2009; Caissie et al., 2010,

2016; Davies et al., 2011; Schlung et al., 2013; Kuehn, 2014; Matul et al., 2016; Pelto et al.,

2018; Chapter 4). Consistent with this, laminated sediments occur at all three sites during the transition from MIS 12 to MIS 11 (Fig. 5.4; Takahashi et al., 2011). These deglacial laminations consist of well-preserved, mm-scale alternations between dark (more terrigenous) and light

(diatom-rich) laminae (Harbour, 2019).

Laminations are preserved under dysoxic or anoxic conditions, when oxygen concentrations at the sediment-water interface are low enough to inhibit sediment mixing by 166 benthic organisms (Calvert, 1964; Kemp, 1996; Pike & Kemp, 1999; van Geen et al., 2003;

Schimmelmann et al., 2016). At Site U1345, a ~3.5 m thick laminated interval spans approximately 12 kyr; its presence indicates that bottom waters at this site were dysoxic for an extended period following deglaciation. In contrast, laminated sediments at sites U1343 and

U1339 persist for shorter intervals of 4000 years and 2000 years, respectively (Fig. 5.4;

Takahashi et al., 2011).

In the Bering Sea today, dissolved oxygen concentrations are lowest between -700 and -

1600 m (Rodin, 2000; Takahashi et al., 2011). Site U1345 (water depth 1007 m) is located within the modern OMZ, but the presence of laminated sediments at U1343 (-1953 m) and U1339 (-

1870 m) suggests expansion of the OMZ during MIS 11. Differences in water depth probably explain why the laminated interval lasts so much longer at U1345, whilst U1343 and U1339 record more transient deepening of the OMZ. In addition, we note that the laminated interval is shorter at U13339 than at U1343, even though the two sites are located at similar depths.

Although it is still not exactly clear what causes laminations — whether they are related to a change in the ventilation of intermediate waters (Zheng et al., 2000), an increase in export productivity (Mix et al., 1999; Ortiz et al., 2004; Davies et al., 2011), or a combination of the two factors (Cook et al., 2005; Matul et al., 2016) — it is possible that bottom waters at Site U1343 were dysoxic for longer than those at Site U1339, due to the higher rates of productivity within the Green Belt.

Ice-rafted Material

Coarse terrigenous grains are present at all three sites throughout almost the entire record (Fig. 5.6; Chapter 3). Coarse terrigenous grains in sediments from the mid- to high- 167 latitude oceans are commonly interpreted as ice-rafted debris (IRD), because ice rafting is the only mechanism capable of transporting coarse sediments to the open ocean (Lisitzin, 2002).

Smear slide analyses show that the coarse grains at our study sites largely consist of sand-sized siliciclastic minerals and rock fragments, consistent with the nature of ice-rafted material in the

Bering Sea (Takahashi et al., 2011; Chapter 3). As glacial ice in Beringia is typically restricted to mountain-valley glaciers (e.g., Glushkova, 2001; Elias & Brigham-Grette, 2013), which do not produce large numbers of icebergs, IRD in the Bering Sea is most likely to be transported by sea ice. However, several lines of evidence (molluscs, pollen, marine microfossils) indicate that tidewater glaciers advanced during Late MIS 11 (Huston et al., 1990; Brigham-Grette, 2001), whilst global eustatic sea level was still high, so iceberg rafting remains a possible source of sediment delivery to the core sites.

The relative abundance of IRD averages 6.3%, 5.5%, and 5.4% at sites U1345, U1343 and

U1339, respectively, although there is considerable downcore variability, particularly at Site

U1339 (Fig. 5.6; Chapter 3). Interestingly, the proportion of IRD is lower at all sites during MIS

12 and 10. As discussed in Chapter 4, this trend may indicate a relationship between sea level and ice rafting, in that when sea level is low during glacials, sea ice forms over deeper waters, and thus entrains less coarse material. The volume % of coarse terrigenous grains at U1345 and

U1343 is highest from~405-402 ka, an interval which falls within the timeframe of peak interglacial warmth. Unfortunately, these coarse grains occur during the contourite, which complicates the interpretation of ice rafting, because we cannot be certain whether the coarse grains are related to bottom current activity, ice rafting, or a combination of these two processes. Overall, the IRD records presented here suggest that sea ice rafting was a consistent 168 source of sediment delivery to the Bering slope and the Umnak Plateau throughout MIS 11, although it is not possible to quantify changes in sea ice extent, based on IRD records alone.

Sea Ice History

At all sites, spring sea ice concentrations are high (>60%) during MIS 12 (Figs. 5.6, 5.7), implying that MIS 12 was a period marked by extensive seasonal sea ice cover. Conditions during MIS 12 may have been similar to the Last Glacial Maximum (LGM, 30-19 ka; Lambeck et al., 2002), when the cold glacial climate allowed for thick, extensive, and possibly perennial sea 169 ice cover in the Bering Sea. During the LGM, sea ice advanced across the northern slope, covered much of the central Bering Sea, and extended at least as far south as the eastern

Bowers Ridge (e.g., Sancetta et al., 1985; Katsuki & Takahashi, 2005; Caissie et al., 2010; Pelto et al., 2018). Caissie et al. (2010) showed that Umnak Plateau Site 51JPC (Fig. 5.2) may have been ice-covered for >8 months per year during the LGM, as evidenced by high relative % abundances of Fragilariopsis species. This interpretation is supported by high average spring sea concentrations (74.6%) at the site (Nesterovich, 2019; Nesterovich & Caissie, in review).

Although there are few high resolution studies of MIS 11 from the Bering Sea, several low resolution studies (e.g., Aiello & Ravelo, 2012; Stroynowski et al., 2015; Onodera et al.,

2016; Teraishi et al., 2016; Detlef et al., 2018) also provide some insight into paleoceanographic conditions in the Bering Sea during the MIS 12-10 cycle. In particular, low values of the molecular biomarker IP25 and low opal mass accumulation rates (MAR) at Site U1343 indicate a prolonged period (~33 kyr) of extensive sea ice cover during Late MIS 12 (Detlef et al., 2018).

In general, sea ice concentrations at sites U1345 and U1343 remain relatively high until

~410 ka (Figs. 5.6, 5.7; Nesterovich & Caissie, in review). This period is followed by a gradual decline in sea ice concentrations at both sites. At U1345, there is a shift towards unconsolidated ice (15-40% concentration), and even ice-free conditions (<15%) from 408-392 ka (Nesterovich & Caissie, in review), whilst at U1343, there is a shorter excursion towards unconsolidated ice around 403 ka (Figs. 5.6, 5.7). In contrast, sea ice declines much earlier at

U1339, with a shift towards unconsolidated ice/open water from ~417 to 410 ka. After 410 ka, however, sea ice re-advances over the Umnak Plateau, at the same time it declines over the slope sites (Figs. 5.6, 5.7). 170

Figure 5.7. Maps of sea ice concentration at sites U1345 (black), U1343 (blue), and U1339 (red), for key time slices during the MIS 12-10 cycle. Concentrations are derived from the sea ice proxy (Nesterovich & Caissie, in review). The turquoise numbers show sea ice concentration values as a percentage. The white dashed line represents the unconsolidated ice margin, and the turquoise dashed line shows the consolidated ice margin. Note that ice margin positions are only an approximation. Grey bathymetric shading is based on modern- day bathymetry, and changes value at -50 m, -250 m, -1000 m, and -2000 m.

171

It is unclear whether the original diatom record at sites U1345 and U1343 was also influenced by bottom currents during the contouritic interval (~408-400 ka), either through winnowing of smaller diatom species, or by the dilution of sea ice species through transport of different species to the site. However, diatom counts from Site U1345 indicate a major shift in the assemblage within the contourite, with Chaetoceros RS replaced by neritic, freshwater, and dicothermal species (Caissie et al., 2016). This provides evidence for increased shelf to basin transport during the contourite (Caissie et al., 2016). Similarly, at U1343, Chaetoceros RS decrease during the contouritic interval (Fig. 5.8), but as we do not have full assemblage counts for Site U1343, we are unable to determine which species become dominant at this time.

Sea ice concentration increases at all sites during Late MIS 11, as sea level and atmospheric CO2 decline (Figs. 5.6, 5.7). This is in keeping with work by Detlef et al. (2018), who showed that a broad gradual increase of IP25 and HBI III, accompanied by a high opal MAR, occurs at Site U1343 during cooling phases of interglacials, indicative of seasonal sea ice of increasing duration. Sea ice concentrations are relatively high at Site U1339 during Late MIS 11, but alternate between periods of consolidated and unconsolidated ice at the slope sites (Figs.

5.6, 5.7).

The differences in sea ice cover between the Bering slope sites and the Umnak Plateau site may be explained by regional differences in sea level pressure and wind direction, related to the position of the Aleutian Low. The Aleutian Low is a wintertime low pressure system centered near the Aleutian Islands (Niebauer et al., 1999). In general, when the Aleutian Low is displaced westward, southerly winds inhibit the southward expansion of sea ice in the Bering

Sea, but when the Aleutian Low is shifted further east towards the Gulf of Alaska, sea ice can 172 advance (Rogers, 1981; Overland & Pease, 1982; Stabeno et al., 1999). It is possible that, if the

Aleutian Low were not shifted as far to the east, sea level pressure could be low over the Gulf of Alaska and the Umnak Plateau, and higher in the rest of the Bering Sea. Under such a scenario, southerly winds would inhibit sea ice advance at the Umnak Plateau, but not at the slope sites.

Like the IRD records, the sea ice concentration records presented here suggest that seasonal sea ice was almost consistently present in some parts of the Bering Sea during MIS 11

(Fig. 5.6). In contrast, none of the core sites are seasonally ice-covered today. Sea ice diatoms make up less than 10% of the assemblage in modern surface sediments from the northern slope region (sites U1345 and V21-163; Fig. 5.2), and ~4% at the Umnak Plateau (Katsuki &

Takahashi, 2005; Caissie et al., 2010; Nesterovich & Caissie, in review). During MIS 11, however, relative percent abundances of sea ice diatoms averaged ~30% at the two slope sites (Caissie et al. 2016; Teraishi et al., 2016), 24% at the Umnak Plateau (Chapter 4), and 10-20% at the eastern Bowers Ridge (Site U1340; Fig. 5.2; Stroynowski et al., 2015). Sea ice was not present at western Bowers Ridge (Site U1341; Fig. 5.2), as evidenced by low (<2%) abundances of sea ice diatoms (Katsuki & Takahashi, 2005; Onodera et al., 2016). Instead, high abundances of dicothermal species indicate that the site was influenced by highly stratified waters, most likely derived from seasonal melting and retreat of sea ice to the north and east.

Productivity Trends

In general, primary productivity in the Bering Sea varies on glacial-interglacial timescales, with higher rates of productivity during interglacials (e.g., Katsuki & Takahashi, 173

2005; Aiello et al., 2012; Kanematsu et al., 2013; Iwasaki et al., 2016; Worne et al., 2019).

Consistent with this, diatom abundances at all three sites are higher during MIS 11, and lower during the glacial stages (Fig. 5.8). Kanematsu et al. (2013) used biogenic opal content to infer

Milankovitch-scale changes in surface water productivity at three sites (U1345, U1343, and

U1341) across the Bering Sea over the past 1.3 Ma (Fig. 5.2). Low opal concentrations at Sites

U1345 and U1343 during glacial periods (Fig. 5.8) were attributed to increased sea ice cover, and dilution by lithogenic material. In addition, Worne et al. (2019) presented an upwelling index, based on δ15N and opal mass accumulation rates, to show that glacial reductions in primary productivity at the Bering slope (Site U1343) were driven, at least in part, by reduced upwelling of nutrients, due to the expansion of sea ice and concurrent expansion of low- nutrient North Pacific Intermediate Water (NPIW).

During deglaciation, as sea level rose, the Bering Slope Current strengthened, glacial NPIW collapsed, and upwelling increased rapidly (Worne et al., 2019). This process helps to explain the productivity spikes observed during deglaciation in the Bering Sea, and wider North Pacific region (e.g., Cook et al., 2005; Brunelle et al., 2007; Caissie et al., 2010, 2016; Khim et al., 2011; Lam et al., 2013; Kuehn et al., 2014; Pelto et al., 2018). Kanematsu et al. (2013) showed that, during MIS 11, opal concentrations at the slope sites were the highest recorded in the past 1 Ma. Similarly, several lines of evidence point toward an interval of enhanced productivity at the onset of MIS 11 at all three sites, including an increase in diatom abundances by 1-2 orders of magnitude, an increase in the relative % abundance of Chaetoceros RS, and the presence of laminations (Fig. 5.8). As described previously, laminated sediments form under low oxygen conditions. We suggest that rapid upwelling during deglaciation, in addition to an influx of nutrients from flooding of the Bering shelf, allowed for extremely high rates of productivity, which led to a downward migration of the OMZ, more dysoxic bottom waters, and the subsequent preservation of laminations. This adds weight to the argument that increased 174 export productivity is at least partially responsible for the occurrence of laminated sediments.

Chaetoceros resting spores (RS) are a taxon indicative of high productivity environments but are also strongly associated with upwelling systems (Abrantes, 1998, 2007). In general, relative % abundances of Chaetoceros RS at the two slope sites track with changes in the upwelling index for Site U1343 (Fig. 5.8; Worne et al., 2019), which remains high throughout

MIS 11c, and declines during Late MIS 11. Relative % abundances of Chaetoceros RS are also strongly coupled with absolute diatom abundances (Fig. 5.8). This suggests that, during MIS 11, high productivity at the slope sites was driven by upwelling. In the Bering Sea today, high 175 productivity within the Green Belt is closely related to upwelling associated with eddies in the

Bering Slope Current (Springer et al., 1996). Based on the diatom records presented here, we propose that conditions similar to those of the modern-day Green Belt were also present at the slope sites during MIS 11. In contrast, the relative % abundance of Chaetoceros RS at Site U1339 is much more variable throughout MIS 11, and is not coupled with either the upwelling index or diatom abundances (Fig. 5.8). This suggests that, similar to today, upwelling regimes at the

Umnak Plateau were different to those along the Bering Slope.

Wider Context

Globally, MIS 11 is characterized by warmer climatic conditions, especially at high latitudes (de Vernal & Hillaire-Marcel, 2008; Melles et al., 2012). Sea level during MIS 11c was probably 6-13 m higher than at present (Raymo & Mitrovica, 2012; Dutton et al., 2015), which would have required major or even total deglaciation of Greenland and/or the West Antarctic

Ice Sheet (Raymo & Mitrovica, 2012). There is evidence for strong warming in the northern hemisphere terrestrial realm (e.g., de Vernal & Hillaire-Marcel, 2008; Prokopenko et al., 2010;

Tzedakis, 2010; Melles et al., 2012; Lozhkin and Anderson, 2013; Candy et al., 2014; Kleinen et al., 2014; Tye et al., 2016), but marine records do not reflect this warming as strongly. In fact,

MIS 11 is not the warmest interglacial in most marine records. In addition, the climate response to interglacial warming during MIS 11 was highly variable in the marine realm. For example, sea surface temperature (SST) reconstructions from the subtropical Northeast Atlantic and the western Mediterranean showed that average SSTs were 1-2°C warmer than today (Kandiano et al., 2012). In contrast, surface waters in the Nordic Seas were both colder and fresher during 176

MIS 11 than they are today, and showed evidence for multiple, high magnitude cold excursions, attributed to freshwater forcing (Kandiano et al., 2012, 2016; Doherty & Thibodeau, 2018).

Like the Nordic Seas, the Bering Sea does not reflect strong warming during MIS 11.

Relative percent abundances of warm water species at Site U1345 (Caissie et al., 2016) and

U1339 (Chapter 4) suggest that SSTs during MIS 11c were not much warmer than during MIS

12. This contrasts with records from the Arctic Ocean, which indicate high SSTs (~8-10 °C near the Mendeleev Ridge) between 410 and 400 ka (Cronin et al., 2019). There is evidence that tidewater glaciers in Beringia advanced whilst eustatic sea level was still high, an example of the

‘out-of-phase glaciations’ observed during interglacial periods in this region. This glacial advance was driven by solar forcing, and by moisture from the flooded Beringian shelf

(Brigham-Grette et al., 2001). Swanger et al. (2017) also documented glacier advance during

MIS 11 in the McMurdo Dry Valleys of Antarctica, which they attributed to increased atmospheric temperatures and precipitation.

In the Arctic Ocean, low abundances of the sub-sea ice dwelling ostracod species

Acetabulastoma indicate a seasonally ice-free Arctic during MIS 11 (Cronin et al., 2019). This transition from perennial to seasonal sea ice is unique to MIS 11, and suggests that sea ice in the Arctic responded strongly to interglacial warming during MIS 11. In contrast, seasonal sea ice remained present in some parts of the Bering Sea throughout MIS 11, although today, sea ice is restricted to the northern shelf. In particular, sea ice advanced over the Umnak Platea during the latter half of MIS 11c (~410-398 ka), when global sea level was high, and much of the world was experiencing peak interglacial warmth. At the same time, however, sea ice concentrations were reduced at the slope sites, and Site U1345, which is closest to the marginal 177 ice zone today, recorded near ice-free conditions between 404 and 402 ka (Figs. 5.6, 5.7). These discrepant records indicate significant regional as well as global variability in the response of the climate system to the interglacial warming of MIS 11.

Summary

In this study, we described trends in sea ice and productivity at three sites in the Bering

Sea during MIS 11. In addition, we noted strong evidence for contourite sequences deposited between 408 and 400 ka at the Bering slope sites (U1345 and U1343). As contourites are associated with sediment reworking through the action of bottom currents, their presence may complicate paleoclimate interpretations from this interval.

At all three sites, MIS 12 was a period marked by extensive seasonal sea ice cover, perhaps similar to conditions during the Last Glacial Maximum. Laminated sediments occurred at all sites during the transition from MIS 12 to MIS 11, suggesting widespread deglacial expansion of the OMZ. Productivity spiked during deglaciation, but did not remain as high once the interglacial period stabilized.

The IRD records presented here suggest that sea ice rafting was a consistent source of sediment delivery to the Bering slope and the Umnak Plateau throughout MIS 11. However, there were regional differences in the response of sea ice to interglacial warming, perhaps related to differences in sea level pressure and wind direction associated with changes in the position of the Aleutian Low. At the Umnak Plateau, sea ice concentrations decreased during deglaciation/early MIS 11, but remained relatively high at the slope sites until approximately

410 ka. Sea ice re-advanced over the Umnak Plateau during peak interglacial warmth, at the 178 same time that sea ice at the slope sites declined. Late MIS 11 was characterized by relatively high sea ice concentration at U1339, and alternations between consolidated and unconsolidated ice at the slope sites. In contrast, none of these sites are seasonally ice-covered today. This suggests a fundamental difference in sea ice regimes between MIS 11 and the

Holocene.

References

Abrantes, F., 1988. Diatom assemblages as upwelling indicators in surface sediments off Portugal. Mar. Geol., 85 (1), 15-39. https://doi.org/10.1016/0025-3227(88)90082-5

Aiello, I.W. & Ravelo, A.C., 2012. Evolution of marine sedimentation in the Bering Sea since the Pliocene. Geosphere, 8 (6), 1231-1253. https://doi.org/10.1130/GES00710.1

Bauch, H. A., Erlenkeuser, H., Helmke, J. P., & Thiede, J., 2000. A paleoclimatic evaluation of Marine Oxygen Isotope Stage 11 in the high northern North Atlantic (Nordic seas). Glob. Planet. Change, 24 (1), 27-39. https://doi.org/0.1016/S0921-8181(99)00067-3

Ben-Avraham, Z. & Cooper, A.K., 1981. Early evolution of the Bering Sea collision of oceanic rises and North Pacific subduction zones. Geol. Soc. Am. Bull., 92(7), 485-495. https://doi.org/10.1130/0016-7606(1981)92<485:EEOTBS>2.0.CO;2

Berger, A. & Loutre, M.F., 2002. An Exceptionally Long Interglacial Ahead? Science, 297 (5585), 1287-1288. https://doi.org/10.1126/science.1076120 179

Berger, A. & Loutre, M.F., 2003. Climate 400,000 Years Ago, a Key to the Future? In: Droxler, A.W., Poore, R.Z., & Burckle, L.H. (Eds.), Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., pp. 17-26.

Bereiter, B., Eggleston, S., Schmitt, J.,Nehrbass-Ahles, C., Stocker, T.F., Fischer, H., Kipfstuhl, S., & Chappellaz, J., 2015. Revision of the EPICA Dome C CO2 record from 800 to 600 kyr before present. Geophys. Res. Lett., 42 (2), 542-549. https://doi.org/10.1002/2014GL061957

Brigham-Grette, J., 2001. New perspectives on Beringian Quaternary paleogeography, stratigraphy, and glacial history. Quat. Sci. Rev., 20, 15-24, 1-3. https://doi.org/10.1016/S0277-3791(00)00134-7

Brown, Z.W. & Arrigo, K.R., 2012. Contrasting trends in sea ice and primary production in the Bering Sea and Arctic Ocean. J. Mar. Sci., 69, 1180-1193. https://doi.org/10.1093/icesjms/fss113

Brown, Z.W. & Arrigo, K.R.,2013. Sea ice impacts on spring bloom dynamics and net primary production in the Eastern Bering Sea. J. Geophys. Res. Oceans, 118, 43-62. https://doi.org/10.1029/2012JC008034

Cronin, T.M., Keller, K.J., Farmer, J.R., Schaller, M.F., O'Regan, M., Poirier, R., Coxall, H., Dwyer, G.S., Bauch, H., Kindstedt, I.G., Jakobsson, M., Marzen, R., & Santin, E., 2019. Interglacial paleoclimate in the Arctic. Paleoceanogr. Paleoclimatol., 34 (12), 1959-1979. https://doi.org/10.1029/2019PA003708 de Abreu, L., Abrantes, F.F., Shackleton, N.J., Tzedakis, P.C., McManus, J.F., Oppo, D.W., & Hall, M.A., 2005. Ocean climate variability in the eastern North Atlantic during interglacial marine isotope stage 11: a partial analogue for the Holocene? Paleoceanography, 20 (3), PA3009. https://doi.org/10.1029/2004PA001091 de Vernal, A. & Hillaire-Marcel, C., 2008. Natural Variability of Greenland Climate, Vegetation, and Ice Volume During the Past Million Years. Science, 320 (5883), 1622-1625. https://doi.org/10.1126/science.1153929

Doherty, J.M. & Thibodeau, B., 2018. Cold Water in a Warm World: Investigating the Origin of the Nordic Seas' Unique Surface Properties During MIS 11. Front. Mar. Sci., 5, 251. https://doi.org/10.3389/fmars.2018.00251

Droxler, A.W., Alley, R.B., Howard, W.R., Poore, R.Z., & Burckle, L.H., 2003. Unique and Exceptionally Long Interglacial Marine Isotope Stage 11: Window Into Earth Warm Future Climate. In: Droxler, A.W., Poore, R.Z., & Burckle, L.H. (Eds.), Earth's Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question. Geophys. Monogr. Ser., 137, AGU, Washington, D.C., pp. 1-14.

Dutton, A., Carlson, A.E., Long, A.J., Milne, G.A., Clark, P.U., DeConto, R., Horton, B.P., Rahmstorf, S., Raymo, M.E., 2015. Sea-level rise due to polar ice-sheet mass loss during 181

past warm periods. Science, 349 (6244), aaa4019. https://doi.org/10.1126/science.aaa4019

Elias, S. & Brigham-Grette, J., 2013. Late Pleistocene events in Beringia. Encyclopedia of Quaternary Science (Second Edition), 191-201. https://doi.org/10.1016/B978-0-444-53643-3.00116-3

Ganopolski, A., Winkelmann, R., & Schellnhuber, H., 2016. Critical insolation–CO2 relation for diagnosing past and future glacial inception. Nature, 529, 200-203. https://doi.org/10.1038/nature16494

Hu, A.X., Meehl, G.A., Han, W.Q., Otto-Bliestner, B., Abe-Ouchi, A., &Rosenbloom, N., 2015. Effects of the Bering Strait closure on AMOC and global climate under different background climates. Prog. Oceanogr., 132, 174-196. https://doi.org/10.1016/j.pocean.2014.02.004 Huston, M.M., Brigham-Grette, J., & Hopkins, D.M., 1990. Paleogeographic significance of middle Pleistocene glaciomarine deposits on Baldwin Peninsula, northwestern Alaska. Ann. Glaciol., 14, 111-114. https://doi.org/10.3189/S0260305500008375

Iwasaki, S., Takahashi, K., Kanematsu, Y., Asahi, H., Onodera, J., & Ravelo, A. C., 2016. Paleoproductivity and paleoceanography of the last 4.3 Myrs at IODP Expedition 323 Site U1341 in the Bering Sea based on biogenic opal content. Deep Sea Res. Part II Top. Stud. Oceanogr., 125-126, 145-154. https://doi.org/10.1016/j.dsr2.2015.04.005 182

Kandiano, E.S. & Bauch, H.A., 2007. Phase relationship and surface water mass change in the NorthEast Atlantic during Marine Isotope stage 11 (MIS 11). Quat. Res., 68 (3), 445-455. https://doi.org/10.1016/j.yqres.2007.07.009

Kandiano, E.S., van der Meer, M.T.J., Bauch, H.A., Helmke, J., Damsté, J.S.S., & Schouten, S., 2016. A cold and fresh ocean surface in the Nordic seas during MIS 11: Significance for the future ocean: Cold and Fresh Nordic Seas During MIS 11. Geophys. Res. Lett., 43 (20), 10,929-10,937. https://doi.org/10.1002/2016GL070294

Kaufman, D.S. & Brigham-Grette, J., 1993. Aminostratigraphic correlations and paleotemperature implications, Pliocene–Pleistocene high sea level deposits,

northwestern Alaska. Quat. Sci. Rev., 12 (1), 21-33. https://doi.org/10.1016/0277-3791(93)90046-O

Khim, B.-K., Kim, S., Uchida, M., & Itaki, T., 2011). High organic carbon deposition in the northern margin of the Aleutian Basin (Bering Sea) before the last deglaciation. Ocean Sci., 45, 203-211. https://doi.org/10.1007/s12601-010-0019-y

183

Kukla, G., McManus, J.F., Rousseau, D.D., & Chuine, I., 1997. How long and how stable was the last interglacial? Quat. Sci. Rev., 16 (6), 605-612. doi:10.1016/s0277-3791(96)00114-x.

Lam, P., Robinson, L., Blusztajn, J., Li, C., Cook, M.S., McManus, J.F., & Keigwin, L.D., 2013. Transient stratification as the cause of the North Pacific productivity spike during deglaciation. Nature Geosci., 6, 622-626. https://doi.org/10.1038/ngeo1873

Lawrence, K.T., Herbert, T.D., Brown, C.M., Raymo, M.E., & Haywood, A.M., 2009. High- amplitude variations in North Atlantic sea surface temperature during the Pliocene warm period. Paleoceanography, 24 (2), PA2218. https://doi.org/10.1029/2008PA001669

Lawrence, K.T., Sosdian, S., White, H.E., & Rosenthal, Y., 2010. North Atlantic climatic evolution through the Plio-Pleistocene climate transitions. Earth Planet. Sci. Lett., 300 (3-4), 329- 342. https://doi.org/10.1016/j.epsl.2010.10.013

Lisiecki, L.E. & Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20, PA1003. https://doi.org/10.1029/2004PA001071

Lisitzin, A.P., 2002. Sea-ice and iceberg sedimentation in the ocean: recent and past. Springer- Verlag, Berlin, 563 pp.

184

Loutre, M.F. & Berger, A., 2003. Marine Isotope Stage 11 as an analogue for the present interglacial. Glob. Planet. Change, 36, 209-217. https://doi.org/10.1016/S0921-8181(02)00186-8

McGee, D., Broecker, W.S., & Winckler, G., 2010. Gustiness: The driver of glacial dustiness? Quat. Sci. Rev., 29 (17-18), 2340-2350. https://doi.org/10.1016/j.quascirev.2010.06.009

Nesterovich, A., 2019. Quantitative diatom-based proxy for sea ice extent in the Bering and Chukchi seas. PhD Thesis, Iowa State University, 17277. https://lib.dr.iastate.edu/etd/17277 185

Nesterovich, A. & Caissie, B.E., in review. A diatom-based quantitative sea ice proxy for the Bering and Chukchi Seas. Prog. Oceanogr.

Normack, W.R. & Carlson, P.R., 2003. Giant submarine canyons: is size any clue to their importance in the rock record? In: Chan, M.A. & Archer, A.W. (Eds.), Extreme depositional environments: Mega end members in geologic time. Geological Society of America Special Paper, 370, 175-190. https://doi.org/10.1130/0-8137-2370-1.175

Palumbo, E., Flores, J.A., Perugia, C., Petrillo, Z., Voelker, A.H.L., & Amore, F.O., 2013. Millennial scale coccolithophore paleoproductivity and surface water changes between 445 and 360 ka (Marine Isotope Stages 12/11) in the Northeast Atlantic. Palaeogeogr. Palaeoclimatol. Palaeoecol., 383-384, 27-41. https://doi.org/10.1016/j.palaeo.2013.04.024

Past Interglacials Working Group of PAGES, 2016. Interglacials of the last 800,000 years. Rev. Geophys., 54 (1), 162-219. https://doi.org/10.1002/2015RG000482

186

Pol, K., Masson-Delmotte, V., Johnsen, S., Bigler, M., Cattani, O., Durand, G., Falourd, S., Jouzel, J., Minster, B., Parrenin, F., Ritz, C., Steen-Larsen, H.C., & Stenni, B., 2010. New MIS 19 EPICA Dome C high resolution deuterium data: Hints for a problematic preservation of climate variability at sub-millennial scale in the “oldest ice”. Earth Planet. Sci. Lett., 298 (1-2), 95-103, https://doi.org/10.1016/j.epsl.2010.07.030

Poli, M.S., Thunell, R.C., & Rio, D., 2000. Millennial-scale changes in north Atlantic Deep Water circulation during marine isotope stages 11 and 12: linkage to Antarctic climate. Geology, 28 (9), 807-810. https://doi.org/10.1130/0091-7613(2000)28<807:MCINAD>2.0.CO;2

Raymo, M.E. & Mitrovica, J.X., 2012. Collapse of polar ice sheets during the stage 11 interglacial. Nature, 483 (7390), 453-456. https://doi.org/10.1038/nature10891

Raynaud, D., Barnola, J., Souchez, R., Lorrain, R., Petit, J.-R, Duval, P., & Lipenkov, V.Y., 2005. The record for marine isotopic stage 11. Nature, 436, 39-40. https://doi.org/10.1038/43639b

Robinson, A., Alvarez-Solas, J., Calov, R., Ganopolski, A., & Montoya, M., 2017. MIS-11 duration key to disappearance of the Greenland ice sheet. Nat. Commun., 8, 16008. https://doi.org/10.1038/ncomms16008 Roden, G., 2000. Flow and water property structures between the Bering Sea and Fiji in the summer of 1993. J. Geophys. Res., 105 (C12), 28595-28612. https://doi.org/10.1029/1999JC000030

Rogers, J.C., 1981. The north Pacific oscillation. J. Climatol., 1, 39-57. https://doi.org/10.1002/joc.3370010106

187

Rohling, E.J., Braun, K., Grant, K., Kucera, M., Roberts, A.P., Siddall, M., & Trommer, G., 2010. Comparison between Holocene and Marine Isotope Stage-11 sea-level histories. Earth Planet. Sci. Lett., 291 (1-4), 97-105. https://doi.org/10.1016/j.epsl.2009.12.054

Ruddiman, W. F. (2007). The early anthropogenic hypothesis: Challenges and responses. Rev. Geophys., 45, RG4001. https://doi.org/10.1029/2006RG000207

Sancetta, C., Heusser, L., Labeyrie, L., Naidu, A.S., & Robinson, S.W., 1985. Wisconsin— Holocene Paleoenvironment of the Bering Sea: Evidence from Diatoms, Pollen, Oxygen Isotopes and Clay Minerals. Mar. Geol., 62(1-2), 55-68. https://doi.org/10.1016/0025-3227(84)90054-9

Schrader, H.J. & Gersonde, R., 1978. Diatoms and silicoflagellates. In: Zachariasse, W.J., Riedel, W.R., Sanfilippo, A., Schmidt, R.R., Brolsma, M.J., Schrader, H.J., Gersonde, R., Drooger, M.M., and Broekman, J.A. (Eds.), Micropaleontological counting methods and techniques - an exercise on an eight metres section of the Lower Pliocene of Capo Rossello, Sicily. Utrecht Micropaleontol. Bull., Netherlands, pp. 129-176.

Stabeno, P.J., Ladd, C., & Reed, R.K., 2009. Observations of the Aleutian North Slope Current, Bering Sea, 1996–2001. J. Geophys. Res. Oceans, 114, C05015. https://doi.org/10.1029/2007JC004705

Stabeno, P.J., Schumacher, J.D., & Ohtani, K., 1999. The physical oceanography of the Bering Sea.In: Loughlin, T.R. & Ohtani, K. (Eds.), Dynamics of the Bering Sea. University of Alaska Sea Grant, AK-SG-99-03, Fairbanks, pp. 1-28.

188

Stein, R., 2008. Arctic Ocean Sediments: Processes, Proxies, and Paleoenvironment. Elsevier, Oxford, UK, 592 pp.

Stow, D. & Smillie, Z., 2020. Distinguishing between Deep-Water Sediment Facies: Turbidites, Contourites and Hemipelagites. Geosciences, 10 (2), 68. https://doi.org/10.3390/geosciences10020068

Swanger, K., Lamp, J., Winckler, G., Schaefer, J.M., & Marchant, D.R., 2017. Glacier advance during Marine Isotope Stage 11 in the McMurdo Dry Valleys of Antarctica. Sci. Rep., 7, 41433. https://doi.org/10.1038/srep41433

Takahashi, K., Ravelo, A.C., Alvarez Zarikian, C. A., and Expedition 323 Scientists, 2011. Proceedings of the Integrated Ocean Drilling Program, 323, Tokyo. doi:10.2204/iodp.proc.323.2011

Tye, G.J., Sherri, J., Candy, I., Coxon, P., Palmer, A., McClymont, E.L., & Schreve, D.C., 2016. The 18O stratigraphy of the Hoxnian lacustrine sequence at Marks Tey, Essex, UK: implications for the climatic structure of MIS 11 in Britain. J. Quat. Sci., 31 (2), 75-92. https://doi.org/10.1002/jqs.2840 Tzedakis, P. C., 2010. The MIS 11-MIS 1 analogy, southern European vegetation, atmospheric methane and the 'early anthropogenic hypothesis'. Clim. Past, 6 (2), 131-144. https://doi.org/10.5194/cp-6-131-2010

Tzedakis, P.C., Channell, J.E.T., Hodell, D.A., Kleiven, H.F., & Skinner, L.C., 2012. Determining the natural length of the current interglacial. Nat. Geosci., 5 (2), 138-141. https://doi.org/10.1038/ngeo1358

Tzedakis, P.C., Raynaud, D., McManus, J.F., Berger, A., Brovkin, V., & Kiefer, T., 2009. Interglacial diversity. Nat. Geosci., 2 (11), 751-755. https://doi.org/10.1038/ngeo660

189

Vavrus, S.J., He, F., Kutzbach, J.E., Ruddiman, W.F., & Tzedakis, P.C., 2018. Glacial Inception in Marine Isotope Stage 19: An Orbital Analog for a Natural Holocene Climate. Sci. Rep., 8, 10213. https://doi.org/10.1038/s41598-018-28419-5

Veeken, P.C.H. & van Moerkerken, B., 2013. Depositional environment and lithofacies interpretation. In: Veeken, P.C.H. & van Moerkerken, B. (Eds.), Seismic Stratigraphy and Depositional Facies Models. EAGE Publications, Houton, The Netherlands, pp. 215-374.

Warnock, J. & Scherer, R., 2015. A revised method for determining the absolute abundance of diatoms. J. Paleolimnol., 53, 157-163. https://doi.org/10.1007/s10933-014-9808-0

Yin, Q. & Berger, A., 2015. Interglacial analogues of the Holocene and its natural near future. Quat. Sci. Rev., 120, 28-46. https://doi.org/10.1016/j.quascirev.2015.04.008

190

CHAPTER 6. GENERAL CONCLUSIONS

In this dissertation, I used multiple proxies (sediment grain size, diatoms, and bulk sedimentary isotopes) from Bering Sea sediment cores to investigate natural variability in sea ice and primary productivity during Marine Isotope Stage 11, which is considered one of the best analogs for the (natural) development of Holocene climate. The main objective was to fill an important gap in high-resolution proxy records of MIS 11 climate from the North Pacific region.

Marine sediments provide valuable records of Earth’s climate history over long timescales, but are only useful if we know how to interpret them. In Chapter 3, I evaluated a number of grain size parameters to discern whether grain size is an effective proxy for past sedimentary processes in the Bering Sea, and found that grain size can be used effectively to infer sediment transport processes, ice rafting, and paleoproductivity. Sortable silt is not a useful proxy in the Bering Sea, because it is not possible to distinguish between current-sorted silt and ice-rafted silt. In addition, I was able to use grain size distributions to infer the presence of contourite sequences at sites U1345 and U1343. Contourites were not previously identified at these sites, but it is important to be aware of them, because contourites are associated with sediment reworking through the action of bottom currents, thus, their presence might limit paleoclimate interpretations. There is evidence for bottom current activity throughout MIS 11 at both of the slope sites, but to the best of my knowledge, there is only one contourite sequence. However, it is likely that contourites reflect periodic changes in bottom current strength, possibly linked to Milankovitch cyclicity, and grain size records may provide the key to this. 191

In chapter 4, I presented a high-resolution record of changes in sea ice and productivity at the Umnak Plateau (Site U1339). Diatom assemblages at Site U1339 reflect changes in sea ice, upwelling, and North Pacific inflow, and there is general agreement between diatom and grain size records, which show evidence for persistent sea ice cover at the Umnak Plateau, even during peak MIS 11. Productivity spikes during deglaciation, but gradually declines throughout

MIS 11, independent of fluctuations in sea ice. Nutrient availability may have become a limiting factor on productivity as the interglacial progressed. Unfortunately, I was unable to use δ15N values to infer changes in nitrogen utilization, because the record was influenced by low δ15N values from inorganic nitrogen. Going forward, it would be more useful to measure the δ15N of diatom-bound nitrogen, rather than bulk sedimentary nitrogen, in sediments known to have a high illite content.

In chapter 5, I compared the sea ice and productivity records from the Umnak Plateau to records from two sites on the Bering slope. Sea ice was present at all three sites during MIS 11, but sea ice at the slope sites responded differently to interglacial warming. Today, Site U1345 is closest to the maximum limit of sea ice extent on the Bering shelf, but during MIS 11, sea ice advanced across the Umnak Plateau at the same time it declined over the slope. This points towards significant regional variability in sea ice regimes across the Bering Sea, and also suggests that there are fundamental differences between MIS 11 and the Holocene.

Several avenues for further work suggested themselves during the writing of this dissertation, including a test of grain size repeatability (Appendix B), improvement of the age models for sites U1339 and U1345 (Appendix A, Fig. A.1), and construction of an age model for

Bowers Ridge Site U1340, in the hope that we could include analyses from this site in our 192 regional compilation. Unfortunately, this last project had to be abandoned, because of the extensive soft sediment deformation in cores from Site U1340. One final avenue for further work is suggested by the limitations of the age models used in this study. The age model for

U1345 is relatively high resolution, but sites U1345 and U1339 are only tied to the global stack at the beginning and end of MIS 11. However, there are ash layers at all three sites, which are dated differently, but which look remarkably similar in core images (Appendix C). If it can be shown that these ash layers are from the same eruption, they could be used to improve agreement between the age models.

193

APPENDIX A SUPPLEMENT TO CHAPTER 4: A MULTI-PROXY RECONSTRUCTION OF CHANGES IN SEA ICE AND PRIMARY PRODUCTIVITY AT IODP SITE U1339 (UMNAK PLATEAU, BERING SEA) DURING MARINE ISOTOPE STAGE 11

Figure A.1. Magnetic susceptibility records for sites U1339 and U1345, plotted against composite core depth (m).

194

Table A.1. List of diatom taxa identified in this study, including pennate and centric diatoms identified to genus or species level, and morphogenera of Chaetoceros resting spores (RS).

PENNATE DIATOMS CENTRIC DIATOMS CHAETOCEROS RS

Acanthes spp. Actinocyclus curvatulus Dispinodiscus

Amphora spp. Actinocyclus ochotensis Gemellodiscus

Caloneis spp. Actinocyclus spp. Hyaline

Cocconeis costata Aulacoseira spp. Liradiscus

Cocconeis spp. Azpetia tabularis Syndendrium

Cymatosira spp. Azpetia spp. Vallodiscus

Delphineis kippae Actinoptychus spp. Xanthiopyxis

Delphineis surrirella Asteromphalus robustus

Delphineis spp. Asteromphalus spp.

Diploneis smithii Bacterosira bathyomphalus

Diploneis spp. Coscinodiscus spp.

Entomoneis spp. C. insignis (reworked)

Eunotia spp. Cyclotella spp.

Fragilariopsis cylindrus Dentonula confervacea

Fragilariopsis oceanica Lindavia cf. ocellata

Fragilariopsis reginae-jahniae Lindavia radiosa

Fragilariopsis pacifica Melosira sol

Fragilariopsis atlantica Odontella aurita

Fragilariopsis nana Pseudopodosira elegans

Fragilariopsis pseudonana Porosira glacialis

Fragilariopsis humboldtianarum Paralia sulcata

Fragilariopsis spp. Proboscia curvirostris

Fragilaria spp. Podosira spp.

Fossula arctica Rhizosolenia hebetata

Gomphonemopsis spp. Rhizosolenia spp.

Grammatophora spp. Shionodiscus oestrupii 195

Table A1. (continued) Nitzschia spp. Shionodiscus trifultus

Neodenticula seminae Sinerima marigela

Pinnularia spp. Stephanopyxis turris

Raphoneis amphiceros Thalassiosira angulata

Stauroneis livingstonii Thalassiosira antarctica

Synedra spp. Thalassiosira baltica

Synedropsis hyperborea Thalassiosira binata

Thalassiothrix longissima Thalassiosira bulbosa

Thalassionema nitzschioides Thalassiosira decipiens

Unidentified Naviculoid pennates Thalassiosira eccentrica

Unidentified Nitzschioid pennates Thalassiosira gravida

Thalassiosira hyalina

Thalassiosira hyperborea

Thalassiosira jouseae

Thalassiosira kushirensis

Thalassiosira leptopus

Thalassiosira lineata

Thalassiosira nordenskioldii

Thalassiosira pacifica

Thalassiosira subtilis

Thalassiosira symmetrica

Thalassiosira tealata

Thalassiosira spp. 196

197

Figure A.3. Relative percent abundances of diatoms grouped by environmental niche: sea ice (dark blue); neritic (brown); dicothermal (light blue); warmer water (red); North Pacific indicator (orange); high productivity (yellow); and Chaetoceros RS (dark green). Also shown is sea ice concentration (including upper and lower limits), derived from a quantitative, diatom- based sea ice proxy (Nesterovich and Caissie, in review). The grey panel shows the duration of MIS 11 (424-374 ka) and the green bar represents laminated sediments.

198

APPENDIX B GRAIN SIZE REPEATABILITY

B.1. Background

One of the benefits of laser diffraction particle size analysis is that it works with very small sample sizes, but this can also be considered a drawback, since small samples may not be representative of the grain size distribution of the full sample. In addition, sampling techniques might bias against coarser particles. To test the repeatability of grain size measurements, multiple measurements were run on a small batch (n = 23) of samples from IODP Site U1340.

Analysis of variance (ANOVA) with a Bonferroni correction suggested that measurements were similar within run, but varied between samples (Table C.1; Fig. C.1). I plan to write these results up as a short note with Beth Caissie and Hannah Carroll as co-authors.

B.2. Methods

Grain size was measured using a Malvern Mastersizer Laser 3000, equipped with a

Hydro MV dispersion tank. This instrument uses the principle of laser diffraction to determine the volume distribution of particles in 101 size bins ranging from 0.01 to 3500 µm.

Freeze-dried bulk sediments from Site U1340 were massed to approximately 0.025 g and treated with the deflocculant sodium hexametaphosphate (SHMP) prior to being analyzed.

Each sample was sub-sampled three times, and replicate measurements were performed on each of the subsamples. Mean grain size was calculated in GRADISTAT, and statistical analyses

(ANOVA with a Bonferroni post hoc) were performed in R.

199

Figure B.1. ANOVA boxplot

200

APPENDIX C ASH LAYERS

Figure C.1. Ash layers that may provide a means of correlation between age models for the three sites.