A Thesis Entitled the Deglacial Chronology of the Sturgis Moraine

Home , Fort Wayne Moraine, Valparaiso Moraine

A Thesis

entitled

The Deglacial Chronology of the Sturgis Moraine in South-Central Michigan and

Northeast Indiana

Jennifer M. Horton

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Master of Science Degree in

Geology

______Timothy G. Fisher, PhD., Committee Chair

______Richard Becker, PhD., Committee Member

______James M. Martin-Hayden PhD., Committee Member

______Patricia R. Komuniecki, PhD., Dean College of Graduate Studies

The University of Toledo

August, 2015

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of

The Deglacial Chronology of the Sturgis Moraine in South-Central Michigan and Northeast Indiana

Jennifer M. Horton

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the The Master of Science Degree in Geology

The University of Toledo

August, 2015

Understanding the timing and rate of ice retreat in the Great Lakes Region is critical to understanding any relationship between ice lobes and climate during the Late

Wisconsinan time period. The purpose of this study is to constrain the age of the Sturgis

Moraine, an end moraine of the Saginaw Lobe in south-central Michigan using minimum-limiting radiocarbon and OSL ages.

Previous correlations between till stratigraphy and end moraines suggest that the

Sturgis Moraine formed between 15,500 and 16,100 14C yrs BP, but with little direct supporting chronologic data. To date the Sturgis Moraine, Livingstone sediment cores were collected from three scour lakes within tunnel channels at the distal side of the moraine. Scour rather than kettle lakes are chosen to minimize any organic accumulation lag from meltout of buried ice. Basal ages from gravelly sand in these cores are

13,700±60, 13,750±80, and 13,300±60 14C yrs BP. These ages are similar to basal ages from Clear Lake (13,300±300 14C yrs BP), a kettle lake on the distal side of the

Valparaiso Moraine of the Lake Michigan Lobe, and the Hyre (13,690±50 14C yrs BP), iii

Kenan (13,880±70 14C yrs BP), and Pyles (13,510±160 14C yrs BP) sites from kettle lakes on the distal side of the Fort Wayne Moraine of the Huron Erie Lobe. The minimum- limiting ages for the Sturgis Moraine presented in this study suggest a younger ice margin and smaller Saginaw re-entrant than previously envisioned.

OSL dated sand dunes from within an outwash valley distal of the Saginaw

Moraine (14.3±0.6 and 14.1±0.5 ka) and from dunes that migrated out of the valley bottom (12.6±0.4, 12.3±0.4, 12.4±0.5 and 12.0±0.4 ka) appear to record two separate aeolian activation periods. These ages agree with other sand dune chronologies from northwest Indiana and northwest Ohio suggesting regional variation in climate at these times. The radiocarbon and OSL ages from this study provide a best limiting minimum age of 13,750 ± 80 14C yrs BP (16,320-16,930 cal yrs BP) for deglaciation at the Sturgis

Moraine.

Acknowledgements

I would like to thank the various people who aided with this project: firstly thanks to Mark McDonald and John Dilworth for their assistance in the field and lab, Butch

Berger for his mechanical support and my committee members Dr. Jamie Martin-Hayden and Dr. Richard Becker for their guidance throughout this process. I would also like to thank my fellow University of Toledo graduate geology students; this project wouldn’t have been possible without their friendship and good humor.

The William A. Kneller Graduate Support Fund, and Dr. Lon Ruedisili

Hydrogeology/Environmental Geology Fund made OSL dating possible. OSL dating was processed by Dr. Kenneth Lepper from North Dakota State University. Very special thanks to the Indiana Geologic Survey for providing funding and support for this project.

In particular I would like to thank Marni Karaffa and Henry Loope from the Indiana

Geological Survey for making this project possible through funding and numerous hours of assistance. I would especially like to thank my advisor Timothy Fisher, for engineering the theme of this thesis and for all he has taught me during my duration at

The University of Toledo.

Table of Contents

Abstract ...... iii

Acknowledgements ...... v

Table of Contents ...... vi

List of Tables ...... ix

List of Figures ...... x

1 Introduction ...... 1

1.1 Introduction ...... 1

1.2 Summery of deglacial events ...... 2

1.3 Deglaciation and pro-glacial lake chronology ...... 5

1.4 Study Area ...... 21

1.5 Sand dune chronology of Great Lake Region ...... 23

1.6 Objective and Hypothesis ...... 26

2 Methodology…...... 27

2.1 Introduction ...... 27

2.2 Site selection ...... 27

2.3 Coring ...... 28

2.4 Photograph ...... 29

2.5 Magnetic Susceptibility ...... 29

2.6 Loss-on-Ignition ...... 30 vi

2.7 Grain Size Analysis...... 30

2.8 Radiocarbon dating ...... 30

2.9 Radiocarbon to calendar year conversions ...... 31

2.10 OSL dating ...... 32

2.11 Surficial Mapping ...... 32

3 Results

3.1 Introduction ...... 34

3.2 Surficial Mapping ...... 35

3.3 Coring Sites ...... 38

3.4 Core Lithology ...... 41

3.5 Fennel Lake Core ...... 42

3.6 Wall Lake Sediment Core ...... 44

3.7 Sweet Lake Sediment Core ...... 46

3.8 Radiocarbon Dates and Interpretations ...... 48

3.9 OSL ages ...... 50

4 Discussion ...... 53

4.1 Radiocarbon Ages and Implications ...... 53

4.2 Radiocarbon Ages and Paleoclimate proxies ...... 59

4.3 Sturgis Moraine and radiocarbon chronology of Great Lakes ...... 62

4.4 Sand dune chronology and implications ...... 67

5. Summary and Conclusions ...... 69

5.1 Summary ...... 69

5.2 Conclusion ...... 71

vii

5.3 Future Work ...... 71

References ...... 73

A Loss on Ignition ...... 82

B Magnetic Susceptibility ...... 89

C Grain Size Analysis...... 94

viii

List of Tables

1.1 Summary of ice cover change ...... 3

1.2 Radiocarbon chronology of Great Lakes Region ...... 9

1.3 OSL ages of sand dunes in Great Lakes Region ...... 25

3.1 Coring site location ...... 39

3.2 Radiocarbon ages ...... 48

3.3 OSL ages for Mongo dune field...... 51

4.1 Location of lake sites in Southern Indiana ...... 54

List of Figures

1-1 Glacial lobe of LIS ...... 2

1-2 Ice margins for stadial events in Great Lakes Region ...... 4

1-3 End moraines in MI, IN, and OH ...... 7

1-4 Estimated 14C ages (ka) for deglacial isochrones ...... 8

1-5 Major proglacial lake levels ...... 14

1-6 Minimum limiting ages for deglaciation...... 20

1-7 DEM of Sturgis Moraine ...... 21

1-8 OSL dated aeolian sand dune record in Great Lakes Regions ...... 24

3-1 Surficial geology map of Sturgis Moraine ...... 36

3-2 Coring locations ...... 40

3-3 Stratigraphy for sediment cores ...... 41

3-4 Fennel Lake core stratigraphy ...... 43

3-5 Wall Lake core stratigraphy ...... 45

3-6 Sweet Lake core stratigraphy ...... 47

3-7 Modern lake near Matanuska Glacier ...... 50

3-8 OSL location map of Mongo dune field ...... 52

4-1 Location and basal ages of lake sites in northern Indiana ...... 54

4-2 Sum probability of radiocarbon vs. calibrated ages ...... 58

4-3 Paleoclimate signals and the Sturgis Moraine ...... 60

4-5 Sites that formed during Heinrich Stadial 1 ...... 61

4-6 Location of Lake Maumee 16,400 cal yrs BP...... 63

4-7 Location of Lake Arkona 16,170 cal yrs BP...... 65

4-8 Comparison of sand dunes and Greenland ice cores ...... 68

Chapter 1

Introduction

1.1 Introduction

Interpretation of the deglacial history of the Great Lakes region is complex because of the interaction between ice lobes, and the many retreats and advances of these lobes. Understanding the timing and rate of ice retreat in this area is critical to understanding the relationship between ice lobes and the climate in the Great Lakes region during the Late Wisconsin time period.

Radiocarbon age assignments remain the primary method of correlating glacial sequences and comparing glacial sequences with other records of global climate change.

Accurate correlations of glacial sequences are also essential to test hypotheses of climate change during glacial cycles (Lowell, 1995). The glacial retreat of the Laurentide Ice

Sheet (LIS) in the Great Lakes region has often been reviewed (Leverett and Taylor,

1915; Hough 1958, 1963, 1966; Fullerton, 1980; Mickelson et al., 1982; Karrow, 1989;

Larson and Schaetzl, 2001; Dyke, 2004). While these works synthesize the glacial history in broad terms, a detailed literature review is necessary to explore the supporting data, paying particular attention to the specific radiocarbon dates, if any, that are used to constrain past ice margins.

1.2 Summary of Deglacial Events in the Great Lakes region

This section summarizes the general sequence of deglacial events in the Great Lakes region as presented in the current literature. The next section will critically evaluate the evidence for timing of these events, paying particular attention to events in the Huron-

Erie Lake Plain.

In the Great Lakes region, the LIS was comprised of numerous lobes of ice during deglaciation (Figure 1-1). The movement of these lobes can be categorized into a series of interstadial and stadial periods (Table 1.1).

Figure 1-1. Glacial lobes of the LIS interacting in the Great Lakes region after the last glacial maximum (Kehew et al., 2012).

Table 1.1. Summary of ice cover change through time in the Great Lake region. Adapted from Howard (2010) and Blockland (2013).

Although the radiocarbon database across the Great Lakes region is substantial, only two readvances should be considered well constrained and recorded (Lowell et al.,

1999). The well-dated events (Figure 1-2) are as follows: 1) the readvance that deposited the Two Rivers till over the Two Creeks Forest Bed in Wisconsin (Larson et al., 1994;

Lowell et al., 1999) and above the Cheboygan bryophyte bed in Michigan at about

11,500–11,800 14C yrs BP during the Post-Two Creeks glacial readvance (Black, 1976;

McCartney and Mickelson, 1982; Schneider, 1990; Kaiser, 1994; Larson et al., 1994;

Lowell et al., 1999), and 2) the Marquette Moraine along the southern shore of Lake

Superior dated to 10,000 14C yrs BP at the end of the Younger Dryas cold period (Lowell et al., 1999). These readvances are considered well documented because there is a large

number of ages at each site and with the exception of Lake Gribben, these sites all lie near, but interior to the moraine limit of the readvance. Thus the ages provide a maximum age for the individual moraine, and a time when the ice was expanding, before the moraine was built (Lowell et al., 1999). These readvances are also significant because the timing of these events corresponds to cold or cooling intervals recorded in the Greenland ice cores (Lowell et al., 1999). More time constraints are needed to compare other readvances or ice still stands to the Greenland ice core records. To do this comparison, more minimum-limiting dates are needed to constrain ice margins in areas devoid of dates.

Figure 1-2. Ice margins for stadial events in the Great Lakes region. The stars represent locations of key radiocarbon dates that have been used by multiple authors to reconstruct these ice margins. Relatively few radiocarbon dates are actually used to constrain these ice margins away from the stared study sites. (Adopted from Lowell et al., 1995; Larson and Schaetzl, 2001) 4

1.3 Deglaciation and Pro-glacial Lake Chronology

Since so few deglacial events are constrained by a radiocarbon chronology in the

Great Lake region, the available radiocarbon dates and stratigraphy are reviewed to assess how past researchers made isochrone assignments. A literature review of previous deglacial work shows that more radiocarbon dates are needed to constrain the retreat of the Saginaw lobe and to understand how it controlled drainage from the Lake Erie basin.

At its maximum extent in the southern Great Lakes region the LIS reached southwest

Ohio, central Indiana, and northeastern Illinois. In southern Ohio, the Harwell Moraine marks the outer limit of the LIS and has been dated at 22,730 ± 460 14C yrs BP (Table 2,

Figure 2) (Lowell et al., 1999). From 18,000–16,000 14C yrs BP, ice lobes in the Great

Lakes region oscillated during a slow recession from their LGM extent back into the

Lake Erie basin (Mickelson et al., 1983; Johnson, 1986; Matsch and Schneider, 1986;

Dyke, 2004). The Erie Interstadial is estimated to have occurred between 16,500–15,000

14C yrs BP (Barnett, 1992) or 16,000–14,500 14 C yrs BP (Ridge, 1997) (Table 1.1).

During the Erie Interstade ice pulled back into the Lake Michigan, Lake Huron and Lake

Ontario basins, and the Lake Erie basin was deglaciated. Locally, this nonglacial interval is represented by a layer of lacustrine sand and clays that separate two till layers in the

Malahide Formation (Morner and Dreimanis, 1973; Dreimanis, 1987). Till associated with the readvance from the Erie Interstade position contains glaciolacustrine clay and silt that accumulated in the Lake Erie basin during the ice recession (Karrow et al., 2000).

The timing of the Erie Interstadial is questionable because it has not yet been directly dated and other ice lobes of the LIS such as the James and Des Moines lobes in Iowa do not show evidence of a recession during this time (Dyke, 2004). 5

During the subsequent Port Bruce Stadial (Table 1.1) the Huron-Erie lobe readvanced to the Powell Moraine in Ohio and the Union City Moraine in Indiana by about 14,500 14C yrs BP (Figure 1.3). Spruce wood found in till north of the Powell

Moraine determines a maximum age for the moraine of 14,780±192 14C yrs BP (Table 1-

2) (Ogden and Hay, 1965). Gyttja bulk dated from a sediment core extracted from a bog associated with the Union City Moraine determines a minimum age for the moraine of

14,810±170 14C yrs BP (Table 2) (Shane, 1987) (Table 1-2). According to Barnett

(1992), minimum ages for the recessional Wabash Moraine are 14,050±75 14C yrs BP

(Mickelson et al., 1983) and 14,500±150 14C yrs BP (Shane and Anderson. 1993) (Table

1-2). While there are some ages for the Port Bruce Stadial, most of these ages are from bulk sediment dates which may contain aquatic species that can often be contaminated by old carbon (MacDonald et al., 1991). The ice margin for the Lake Michigan Lobe is loosely constrained at this time and the ice margin for the Saginaw Lobe has not been dated. In the current literature, isochrones are extended from the Huron Erie Lobe across the Saginaw Lobe and Michigan Lobe’s moraines (Figure 1-4). The Tekonsha Moraine system, of the Saginaw Lobe is assigned an age of 15,500 14 C yrs BP (Figure 1-4)

(Fullerton, 1980). However the presence of tunnel channels and drumlins cross cutting the Tekonsha Moraine suggest that the moraine is not associated with the Late Wisconsin deglaciation, and may be much older (Fisher et al., 2005). No radiocarbon dates are available to test this interpretation.

Figure 1-3. End moraines in Michigan, Indiana, and Ohio as mapped by Leverett and Taylor (1915). 7

Figure 1-4. Estimated 14C ages (ka) for deglacial isochrones based on till boundaries and correlations of selected glacial advance limits from Fullerton (1980). Black box is the location of the Sturgis Moraine. According to Fullerton’s age assignments the Sturgis Moraine should be between 16.1-15.5 14C yrs BP. 8

During the retreat of the Huron-Erie Lobe from the Fort Wayne Moraine, meltwater ponded between the receding ice mass and a drainage divide on the Fort Wayne Moraine formed proglacial lakes at successively lower elevations (Leverett and Taylor, 1915;

Dreimanis, 1977). The minimum age for the Fort Wayne Moraine is generally accepted at being around 14,800 14C yrs BP constrained by the date of 14,680 ± 310 14 C yrs BP provided by Shane and Anderson (1993) which established this age (Table 1.2).

However, this date is controversial because it was taken from a bulk sample of gyttja.

Bulk samples used for radiocarbon dating often yield dates much older (1000–2000 years) than those provided by the newer AMS dating method (Fisher et al., 2009). Newer dates taken from lakes associated with the Fort Wayne Moraine (13,690 ± 50 and 13,880

± 70 14C yrs BP) suggest that ice retreated from the moraine about 1,000 years later than originally thought (Glover et al., 2011) (Table 1.2).

The original relative chronology of proglacial lakes in the Lake Erie basin established by Leverett and Taylor (1915), has remained largely unchanged. An absolute chronology of the proglacial lakes in the Lake Erie basin has developed as radiocarbon dating of lake sediments and relevant moraines has been completed (Hough, 1958;

Forsyth, 1960; Dreimanis, 1962, 1966; Totten, 1985; Barnett, 1985; Calkin and Feenstra,

1985)

Table 1.2. Radiocarbon chronology for the Great Lakes Region.

No. Lab Material 14C Age Calib. 2 sigma (cal Stratigraphic Reference No. Age (cal yrs BP) Notes yrs BP) 1. IGSG- Organic 22,730±430 26,891 26,066– Organic mat Lowell, 1999 2837 Mat 27,717 at the base of alluvium 9

above till Interp. Minimum age for Harwell Moraine, southern limit of LIS in Ohio 2. OWU- Wood 14,780±192 17,984 17,524– Wood found Ogden and 83 18,445 in till north of Hay, 1965 the Powell Moraine Interp. Maximum age for the Powell Moraine (Port Bruce Stadial) 3. DIC- Bulk 14,810±170 18,025 17,603– ------Shane, 1987 243 sample 18,445 gyttja/cla y Interp. Minimum age for Union City (Port Bruce Stadial) 4. ISGS- wood 14,050±75 17,079 16,780– ------Mickelson et 348 17,379 al., 1983 Interp. Minimum age for Wabash Moraine 5. ISGS- Bulk 14,500±150 17,632 17,248– ------Shane and 402 sample 18,592 Anderson, 1993 plant litter Interp. Minimum age for Wabash Moraine 6. ISGS- Bulk 14,680±310 17,834 17,077– ------Shane and 1679 sample 18,592 Anderson, 1993 clay/gytt ja Interp. Minimum age for the Fort Wayne Moraine 7. Beta- Twigs 13,690±50 17,016 16,295– Basal Glover et al., 19086 AMS 16,776 section of 2011 4 dating lake core

Interp. Basal date for Hyre Lake site 8. Beta- Twigs 13,800±70 16,792 16,517– Basal Glover et al., 19086 AMS 17,068 section of 2011 4 dating lake core Interp. Basal dates for Kenan Lake site 9. I- Bulk 13,770± 210 16,674 16,056– Unoxidized beds Buckley and 4899 sample 17,293 of sand, silt, clay Willis, 1972 clay/gyttj above sand a overlying unoxidized till in outwash channel Interp. Glacial Lake Maumee III 10. GSC- Bulk 12,190±230 14,199 13,572– Gyttja from 16.0- Lowdon 620 sample 15,036 16.13 m depth below and Blake, clay/gyttj lake level 1968 a Interp. Minimum age for ice retreat from Ingersoll Moraine and Lake Maumee Stage (Fullerton, 1980) Maumee IV (Lowdon and Blake, 1968)

11. W-33 wood 13,600±500 16,395 14,750– Stratigraphic Suess, 1954 17,824 horizon between Arkona and Whittlesey sand Interp. Dates late Port Bruce Stadial (Suess,1954) 10

Arkona: Samples lies right on Port Bruce till at 690 ft. elevation, below the Whittlesey level (Goldthwait,1958) End of Lake Arkona recorded by 10ft of overlying sandy silts and wood found on the Arkona-Whittlesey transition beach (Dreimanis,1977) Arkona III (Fullerton,1980) 12. W- moss 12,570± 500 14,855 13,463– Moss located Farrand et al., 1889 16,247 between two till 1969 layers Interp. Makinaw Interstadial 13. TX- Organic 12,960±250 15,293 14,286– Lacustrine silt Blewett et al., 651 residue 16,301 proximal to the 1993 inner Port Huron Moraine Interp. Maximum age of the Lake Whittlesey 14. Y- Spruce 12,800±250 15,097 14,225– Imbedded in beach Barendsen et 240 wood 15,970 sediments of al., 1957 Whittlesey Interp. Maximum age of the Lake Whittlesey (Barendsen et al., 1957; Calkin,1970; Fullerton, 1980) 15. ISGS- wood 12,770±180 15,080 14,373– ______Hansel and 1418 15,787 Johnson, 1986 Interp. Glenwood II phase of glacial Lake Chicago in Illinois 16. GSC- wood 13,100±110 15,677 15,317– Wood buried by till Gravenor and 2213 16,038 Stapvsky, 1976 Interp. Maximum age for the Wyoming Moraine in Ontario 17. W- wood 12,920±400 15,284 14,027– Peat zone below Rubin and 430 16,541 Lake Whittlesey Alexander, beach gravels; 1958 underlying the peat is sandy alluvium and lake clay at the base of till Interp. Maximum age for Lake Whittlesey (Goldthwait,1958) Ypsilanti (Fullerton,1980) 18. S-31 driftwoo 12,660±440 14,964 13,653– Driftwood from McCallum, d 16,239 Lake Whittlesey 1965 within gravel Interp. Ypsilanti (Fullerton,1980) Arkona or Arkona-Whittlesey transition (Dreimanis,1966:584) 19. Beta2 wood 13,430±90 16,163 15,870– Within littoral facies Campbell et 58945 16,456 al., 2011 Interp. Maximum age for Warren beach 20. ISGS- wood 13,050±100 15,619 15,286– Wood sample from Totten, 1982 473 15,953 basal sand of Warren I Interp. Warren I 21. 1- Wood 12,730 14,943 15,617– From peat overlain Buckley, 3665 (Buckley,1976) ±220 14,146 by alluvium 1976 Organic material (Calkin,1970) Interp. Whittlesey or Early Warren, also minimum date for recession from the Valley Heads Moraine and succeeding Gowanda end moraine (Calkin and McAndrews, 1969) 11

Lake Whittlesey or early Warren 1, also dates the Gowanda Moraine (Calkin, 1970:94) Warren (Lewis, 1969) 22. S-30 Vegetabl 12,000±500 14,201 12,878– Base layer of black McCallum e muck 15,525 carbonaceous muck and Dyck, 1-3 inches thick; 1966 overlies gravelly sand Interp. Warren I, underlying gravelly sand that belongs to Warren I (McCallum and Dyke, 1960) Sand also correlates to a beach presumed to be Whittlesey in age and post Warren muck overlies Lake Warren offshore beach sand and is younger that Warren (Dreimanis, 1966) Warren I (Fullerton, 1980) 23. S-29 wood 11,400± 450 13,665 12,787– From a base layer of McCallum 14,544 black carbonaceous and muck 1-3in thick; Dyck,1966 muck overlies gravelly sand, associated with S-30 Interp. Post Warren, muck overlies Lake Warren offshore or beach sands making it younger than Warren (Dreimanis, 1966) Warren I (Fullerton, 1980) 24. 1- Plant 12,610± 170 14,884 14,218– Sample from layer Lewis, 4040 detritus 15,551 of fine sand and 1969 plant detritus Interp. Early Lake Erie 25. 1- Plant and 11,140± 160 12,999 12,721– Sample from layer Lewis, 4041 shell 13,277 of silt and fine sand 1969 detritus with plant detritus Interp. Early Lake Erie 26. Beta- bryophyt 12,200± 100 14,193 13,766– Between two till Larson et al., 50967 e 14,561 layers 1994 Interp. Correlated with the Two Creeks forest bed deposited during the Two Creeks interstade 27. ISGS- bryophyt 11,400± 290 13,252 12,940– Between two till Larson et al., 233 e 13,565 layers 1994 Interp. Correlated with the Two Creeks forest bed deposited during the Two Creeks interstade 28. ETH- wood 11,050± 80 12,906 12,744– Between two till Larson et al., 9241 13,068 layers 1994 Interp. Correlated with the Two Creeks forest bed deposited during the Two Creeks interstade 29. ETH- Fecal 11,570± 80 13,594 12,700– Between two till Larson et al., 50966 dropping 13,123 layers 1994 s Interp. Correlated with the Two Creeks forest bed deposited during the Two Creeks interstade 30. W-42 wood 13,350±120 13,244 13,000– Two Creeks forest Broecker and 13,448 bed Farrand, 1963 Interp. Readvance of ice causes flood to drown Two Creek area MI

As ice retreated from its Port Bruce position, water was dammed between the retreating ice mass and the Fort Wayne Moraine forming proglacial Lake Maumee

(Leverett and Taylor, 1915; Dreimanis, 1969; Calkin and Feenstra, 1985). Lake Maumee occupied three or four levels designated as Maumee I, II and III, with a possible reoccupation of Maumee II levels known as Maumee IV (Leverett and Taylor, 1915;

Fullerton 1980; Calkin and Feenstra, 1985). Absolute chronology for the Lake Maumee stage is not well developed. Dated plant remains of 13,770±210 14C yrs BP are attributed to the Maumee III stage (Buckley and Willis, 1972; Fullerton, 1980). The plant remains were found in an unoxidized bed of sand, silt and clay overlying till within an outwash channel (Buckley and Willis, 1972). This date has also been used to record abandonment of the Imlay channel (Figure 1-5) at the end of the Maumee Stage. Farrand and Eschman

(1974) argued that the date is attributed to deposition during the existence of Lake

Maumee III and the sample postdates the abandonment of Maumee III. A second radiocarbon date of 12,190±230 14C yrs BP is also attributed to Lake Maumee (Lowdon and Blake, 1968). This date is from a bulk sample of gyttja, interpreted to be a minimum age for ice retreat from the Ingersoll Moraine that correlates to the Lake Maumee shoreline (Lowdon and Blake, 1986). Fullerton (1980) argues that this date represents

Maumee IV as a reoccupation of Maumee II level. Both of these Lake Maumee dates are extracted from bulk samples, and there is a possible contamination from older carbon, making them than older than the minimum age of the events they represent. OSL ages obtained from a Maumee spit, and from fine sand beneath a Lake Whittlesey spit, support an age of approximately 16.8 ka ( ̴13,500 14C yrs BP) for Lake Maumee (Fisher et al.,

2015).

Figure 1-5. Major proglacial lake levels and associated outlets (Larson and Schaetzl, 2001) As ice continued to retreat, Lake Maumee lowered to form Lake Arkona (Hough,

1973), in three distinct stages. Arkona beaches are not well defined and Eschman and

Karrow (1985) attributed this to the fact that Arkona sediments are overrun by the Port

Huron advance in the Erie basin and on the thumb of Michigan. But this explanation does not explain the poorly defined beach sands of the Port Huron advance. Fullerton (1980) associated a radiocarbon date of 13,600±500 14C yrs BP with the lowest Arkona level

(Arkona III) (Table 1.2). Dreimanis (1977) uses it to date the Arkona-Whittlesey transition. While Suess (1954) commented that this age dates the Port Bruce Stadial.

The dated piece of wood was found stratigraphically between deposits of Arkona and

Whittlesey sediments. Driftwood can float in lakes for an unknown number of years or be transported into the lake, which makes this date at best a maximum age for Lake

Whittlesey.

During the Makinaw Interstadial, following the Port Bruce Stadial, ice retreated northeastward to an unknown position. During this interstadial, falling water levels of

Lake Arkona are recorded across the thumb of Michigan by outlet channels (Eschman and Karrow, 1985). Water continued to recede to an even lower stage known as Lake

Ypsilanti with drainage presumably into the Lake Ontario basin (Kunkle, 1963;

Dreimanis, 1977; Calkin and Feenstra, 1985). An interpreted buried river channel near

Ypsilanti, MI based on water-well log data documents this stage. This channel cuts into till deposited during the Port Bruce Stadial. In the Lake Erie basin a channel is covered by lake sediments from Lake Whittlesey and Warren (Dreimanis, 1977; Calkin, 1985;

Fullerton, 1980). This river channel could also possibly be a tunnel channel from when the area was last covered by ice.

Following the Makinaw Interstadial, the LIS readavanced during the Port Huron

Stadial and built the very prominent Port Huron end moraines in Michigan. Organics within silt deposited on the proximal margin of the northwest section of the inner Port

Huron system give a minimum age of 12,960±300 14C yrs BP for the inner Port Huron

Moraine (Blewett et al., 1993) (Table 1.1). This date is compatible with other features associated with the Port Huron Stadial, such as organic material within glacial Lake

Whittlesey in Ohio and New York (12,800±250 14C yrs BP) (Barendsen et al., 1957), the

Glenwood II phase of glacial Lake Chicago in Illinois (12,770±180 14C yrs BP) (Hansel

and Johnson, 1986) and the wood buried in till associated with the Wyoming Moraine in

Ontario (13,100±110 14C yrs BP) (Gravenor and Stupavsky, 1976) (Table 1.2).

During the Port Huron Stadial, lake levels rose to the Lake Whittlesey stage

(Leverett and Taylor, 1915; Hough, 1958, 1963; Dreimanis, 1977). Lake Whittlesey is also associated with the Port Huron Moraines in western New York when they intersect the Whittlesey strandlines in the Lake Erie basin (Muller and Prest, 1985). A maximum age for Lake Whittlesey is 12,800±250 14C yrs BP from wood found within beach sediments (Barendsen and Deevey, 1957). Wood dated at 12,920±400 14C yrs BP from peat beneath beach gravels is another maximum age for the beach (Rubin and Alexander,

1958) (Table 1.2). This date has also been interpreted as evidence for a lower lake level

(Fullerton, 1980). Driftwood dated on top of Lake Whittlesey gravel (12,660±440 14C yrs BP) provides another maximum age or possible limiting age for the lake (McCallum,

1955) (Table 1.2). However, this date has also been used to date the Ypsilanti lake level

(Fullerton, 1980) and the Arkona-Whittlesey transition, and just the Arkona stage

(Dreimanis, 1966).

As ice retreated from the Port Huron Moraine, the Ubly Channel (Figure 1-5) opened and Lake Whittlesey drained into Lake Saginaw (Calkin and Feenstra, 1985) and water level dropped to the Lake Warren level (Calkin and Feenstra, 1985). Lake Warren can be divided into three stages, Warren I, II, and III (Leverett and Taylor, 1915; Calkin and Feenstra, 1985; Muller and Prest, 1985). The formation of the Marilla Moraine in western New York is correlated to the strandline of Lake Warren I (Muller and Prest,

1985). Geomorphic and stratigraphic evidence suggests that Whittlesey levels dropped to form Lake Warren I when ice retreated to the Marilla Moraine (Calkin, 1970). 16

A piece of driftwood from the eastern edge of the littoral unit of the Oak Opening

Ridge in Toledo, Ohio was dated at 13,430± 90 14C yrs BP (Campbell et al., 2011) (Table

1.2). This date is older than several previous dates and is considered a non-limiting maximum age for Lake Warren. Four OSL ages from the upper Warren Beach face overlap at one standard error, and average to 14.2±0.3 ka (Higley et al., 2014). Previous to these OSL dates, a wood sample from basal Warren I sand was used to date Lake

Warren at 13,050±100 14C yrs BP (15,619 cal yrs BP) (Totten 1985). A piece of wood

(Bickley, 1976) and organic material (Calkin, 1970), from a horizon of silty peat overlain by stream deposits was dated at 12,730 ±220 14C yrs BP and is interpreted as part of the recessional sequence between Lakes Whittlesey and Warren (Calkin, 1970: 94), and as a maximum Lake Warren (Calkin, 1970). Based on the dates of 12,000±500 14C yrs BP and 11,400±450 14C yrs BP from organic material lying directly above Lake Warren sediment, Dreimanis (1966) and Calkin and Feenstra (1985) determined that the Warren level was abandoned at 12,000 14C yrs BP.

Numerous, short-lived lake levels (Wayne, Grassmere, Lundy, and Early

Algonquin) developed as water levels incrementally dropped below the Warren level before the Erie Lobe retreated north of the Niagara escarpment and water levels dropped to the Early Lake Erie level. This ice retreat is correlated to ice retreat from the Port

Huron Moraine (Hough, 1963). Two dates from organics in the Pelee Basin in western

Lake Erie (12,650±170 and 11,140±160 14C yrs BP) provides timing for the end of the proglacial lake stages (Lewis, 1969).

Ice recession in the Lake Erie basin occurred earlier than in the Lake Michigan basin. Following the Port Huron Stadial the Two Creeks Interstade occurred in the Lake 17

Michigan basin (Figure 1-2). A bryophyte bed near Cheboygan Michigan was thought to have been associated with the Mackinaw interstadial by providing age constraint (Farrand et al., 1969) but was reevaluated and found to be younger and associated with the Two

Creeks Interstade (Larson et al., 1994). The bryophyte bed was redated (12,200±100,

11,400±290, 11,050±80, and 11,750±80 14C yrs BP) (Larson et al., 1994). These ages correspond with the age of the Two Creeks forest bed in Wisconsin (̴11,700 14C yrs BP)

(Rech et al., 2011). The forest was drowned when the Michigan Lobe readvanced forcing water level of Lake Calumet to rise.

The last readvance of ice into the Great Lakes region was during the Younger

Dryas cold period, when the Lake Superior Lobe expanded south onto land from the lake basin. Several authors have reported wood ranging in age from 11,000–10,000 14C yrs

BP, associated with a red till representing this event (Black, 1976; Clayton and Moran,

1983; Hack, 1965). In Marquette, Michigan near Lake Gribben, prograding outwash sediments buried a boreal forest as the Outer Marquette Moraine was being built (Drexler et al., 1983). Radiocarbon ages from the wood average 10,040±65 14C yrs BP (Table 1.2)

(Lowell et al., 1999). Following this readvance the ice margin retreated northward and by approximately 9,000 14C yrs BP it had withdrawn from the Great Lakes watershed

(Barnett 1992; Karrow et al., 2000).

Unfortunately there is little to no age control for the retreat of the Saginaw lobe during these multiple stadial and interstadial events (Figure 1-6). Understanding the deglacial chronology of the Saginaw lobe is important to both understanding the relationship of the various lobes of the LIS and to understanding the lake stages of ancestral Lake Erie. The ice margin of Saginaw lobe controlled the drainage of Lakes 18

Maumee and Arkona. According to Eschman and Karrow (1985), as the Saginaw lobe retreated northward off the “thumb of Michigan” the drainage of Lake Maumee switched from the southern outlet of the Fort Wayne Moraine to the Imlay Channel, that cuts north and then west across the axis of the “thumb”. This change in outlets led to the various stages of Lake Maumee. Late in the history of Lake Maumee, drainage was also directed into the Saginaw lowlands as the ice retreated even farther northward (Leverett and

Taylor, 1959, Eschman and Karrow, 1985). Lake Maumee was then replaced by glacial

Lake Arkona when the Saginaw ice margin finally retreated far enough north across the

“thumb” to allow for the merging of water in the Huron and Erie basins with that in the

Saginaw lowlands (Leverett and Taylor 1915, Fullerton, 1980). The chronology for these various lake stages is lacking and there is no age control on the Saginaw lobe ice margin during Lake Maumee or glacial Lake Arkona.

Figure 1-6. Minimum limiting ages for deglaciation in central MI, northern IN, and northern OH. The red outline is the Sturgis Moraine, which is completely devoid of dates. The radiocarbon ages do not support Fullerton’s (1980) age assignments for the moraine in central Michigan. They tend to be around 1,000-2,000 years younger than his ages. There are very few dates for the Saginaw Lobe retreat.

1.4 Study Area: The Sturgis Moraine

One area within the Great Lakes region that is mostly without a deglacial chronology is the Saginaw lobe and all of its end moraines, including the Sturgis

Moraine. The moraine lies in south central Michigan and extends slightly into northeast

Indiana (Figure 1-7).

Figure 1-7. Digital elevation model (DEM) of the Sturgis Moraine area. The yellow star in the inset map is the location of the Sturgis Moraine with the Great Lakes Region.

The Sturgis Moraine is a well-developed and prominent moraine at the distal end of a drumlinized till plain and network of tunnel channels (Fisher and Taylor, 2002).

Total relief is 25 meters and it’s dominantly composed of glaciofluvial sediment (Colgan

et al., 2005). Thick alluvial fans extend from the moraine, and tunnel channels cut through it with some extending beyond it (Colgan et al., 2005; Kehew and Kozlowski,

2007; Kehew et al., 2012). The moraine was first described by Leverett and Taylor

(1915). They observed the distinct ridge of the moraine and the great outwash plans associated with it that leads down into the St. Josephs River. The Sturgis Moraine is a product of the Saginaw Lobe, which flowed southwesterly across the interior of southern

Michigan, eroding the Saginaw Bay of Lake Huron. Several authors (Leverett and

Taylor, 1915; Fullerton, 1980; Larson and Schaetzl, 2001) suggested that the Lake

Michigan and Huron-Erie Lobes developed rapidly, streaming flow, while the Saginaw

Lobe advanced into south-central Michigan first, but weakened or stagnated as the flanking lobes grew stronger and encroached upon the recently deglaciated Saginaw Lobe terrain. This conceptual idea has not been tested and a chronology is not available to test it.

The available deglacial chronology of the Sturgis Moraine comes from isochrones by Fullerton (1980) (Figure 1-4). His reconstructions show that the Sturgis Moraine did not readvance into the interlobate region of Indiana after the Erie Interstade. He correlated traceable stratigraphic units of till between Great Lake basins to connect the

Tekonsha Moraine of the Saginaw Lobe to the Union City Moraine of the Huron-Erie

Lobe (Figure 1-4). Using radiocarbon dates from the Union City Moraine it is inferred that the Sturgis Moraine was ice free sometime before 15,000 14C yrs BP. Another till correlation constructed by Monaghan et al. (1986) places the ice of the Michigan Lobe at the West Chicago Moraine; the outermost ridge of the Valparaiso Moraine System in

Illinois, when ice of the Saginaw Lobe was at the Sturgis Moraine (Figure 1-4). 22

There is little to no limiting age control on the Sturgis Moraine. As discussed above, the only assigned age for the Sturgis Moraine is from Fullerton’s (1980) isochrones. These isochrones are poorly constrained and only a few bulk sediment dates, processed before the 1980s are used to extrapolate the ages of each moraine. More recently published minimum-limiting ages (Glover et al., 2011) associated with these isochrones suggest that Fullerton’s (1980) assigned ages are too old (Figure 1-6). Age control on the Sturgis Moraine is critical to understanding the interaction between the

Michigan, Huron-Erie, and Saginaw lobes.

1.5 Sand Dune Chronology of the southern Great Lakes Region

While radiocarbon dating of organic material is the most common way to develop records of climate change and deglaciation, other dating techniques are equally as valuable. In areas where proxies such as high-resolution ice, coral, cave, marine, or lacustrine sediment stratigraphies are lacking, optically stimulated luminescence (OSL) dating of aeolian sand dunes can serve as proxies for dry, cold and windy events at the close of the last ice age.

The chronology of inland-forested dunes of the southern Great Lakes region is limited and only a few studies have used OSL samples to date sand dune activity (Figure

1-8). The recent work of Fisher et al. (2015) reported ages clustered around (11.8–12.4 ka) and older building on the previous work of Campbell et al. (2011) who reported nine ages that clustered around the Younger Dryas time period and younger. Most ages between 13–8 ka center on colder, drier periods of climate (Campbell et al., 2011). Dates

from the Fair Oaks dunes in Northwest Indiana also record dune activity during the

Younger Dryas (Kilibarda and Blockland, 2010) (Table 1.3).

Dune activity clustered around 12 and 14 ka appears to coincide with periods of cooler climate recorded in both vegetation pollen records and the Greenland Ice Cores.

Sand dune activity clustered around 12 ka occurred during the Younger Dryas cold period. However, the climate conditions during the Young Preboreal time period (8ka) recorded by aeolian activity are not recorded in the available pollen records. The limited data suggests that sand dune reactivation is sensitive to cooler, drier episodes of climate.

More samples from different sites are needed to expand this OSL dated chronology.

Figure 1-8. The current OSL dated aeolian sand dune record in the Great Lakes region. All dates are present thousand years BP rounded to the nearest hundred. Black box is the location of the Mongo dune field (Table 1.3). 24

Table 1.3. OSL Ages of Sand Dunes in the southern Great Lakes Region

# Sample Burial H2O K2O U Th Cosmic Dose De No. of Age Source Depth (%) (%) (ppm) (ppm) (Gy) Rate (Gy) Aliquot (ka) (M) (Gy/k a)

A UNL24 5.3 3.3 1.46 0.85 1.94 0.11 1.60± 18.30 35 11.5± (Kilibarda and 69 0.06 ±5.75 0.7 Blockland, 2010) B UNL24 5 4.4 1.42 0.6 1.59 0.11 1.47± 16.47 28 11.2± (Kilibarda and 70 0.06 ±0.60 0.7 Blockland, 2010) C UNL24 1.7 5.1 1.59 1.05 2.05 0.17 1.78± 22.22 38 12.5± (Kilibarda and 71 0.07 ±0.70 0.7 Blockland, 2010) D UNL24 4.2 5.3 1.45 0.63 1.78 0.12 1.51± 18.85 40 12.5±. (Kilibarda and 72 0.06 ±0.64 08 Blockland, 2010) E UNL24 1.3 16.2 1.34 0.63 2.48 0.18 1.35± 5.32± 42 3.93± (Kilibarda and 73 0.06 0.20 0.26 Blockland, 2010) F UNL24 0.9 4.1 1.35 0.52 1.72 0.19 1.48± 0.32± 40 0.22± (Kilibarda and 74 0.06 0.05 0.04 Blockland, 2010) G UNL19 0.96 1.1 1.10 0.6 1.8 0.19 1.35± 16.94 40 12.5± (Campbell et al., 13 0.06 ±0.72 0.8 2011) H UNL19 0.95 1.4 1.14 0.6 1.9 0.19 1.39± 17.11 41 12.3± (Campbell et al., 14 0.06 ±0.68 0.8 2011) I UNL19 0.60 2.3 1.41 0.7 1.8 0.19 1.62± 14.16 46 8.76± (Campbell et al., 15 0.07 ±0.56 0.56 2011) J UNL19 0.90 1.9 1.19 0.6 1.7 0.19 1.41± 17.66 48 12.5± (Campbell et al., 16 0.06 ±0.66 0.7 2011) K UNL21 0.66 5.8 1.42 1.52 2.76 0.20 1.81± 18.35 48 10.2± (Campbell et al., 02 0.07 ±0.55 0.6 2011) L UNL21 0.65 4.2 1.65 0.81 2.05 0.20 1.81± 18.14 46 10.0± (Campbell et al., 03 0.07 ±0.57 0.6 2011) M UNL21 0.80 6.1 1.44 1.10 2.23 0.19 1.68± 14.78 47 8.78± (Campbell et al., 04 0.06 ±0.39 0.48 2011) N UNL21 0.56 7.8 1.55 0.65 1.77 0.20 1.61± 18.91 39 11.7± (Campbell et al., 05 0.06 ±0.56 0.7 2011) O UNL21 0.61 4.6 1.38 0.75 1.87 0.20 1.57± 1.24± 75 0.79± (Campbell et al., 06 0.06 0.06 0.06 2011) P UNL21 2.05 3.4 1.53 0.99 1.82 0.16 1.38± 19.46 44 14.1± (Campbell et al., 06 0.07 ±0.65 1.0 2011) Q UNL24 0.75 3.3 1.16 0.66 1.96 0.19 1.37± 13.16 37 9.63± (Campbell et al., 51 0.05 ±0.68 0.69 2011) R UNL34 0.85 2.1 1.51 1.51 3.6 0.19 2.02 ± 23.79 51 11.8 ± (Fisher et al.,2015) 55 0.07 ± 0.81 0.6 T UNL34 1.24 3.5 1.14 0.94 2.19 0.18 1.71 ± 27.86 54 12.9 ± (Fisher et al.,2015) 57 0.06 ± 1.17 0.6 U UNL34 1.5 2.5 1.19 0.72 2.28 0.18 1.61 ± 19.83 55 12.3 ± (Fisher et al.,2015) 58 0.06 ± 0.64 0.6 V UNL34 1.0 14.9 1.42 1.52 2.97 0.19 1.75 ± 26.34 50 15.1 ± (Fisher et al.,2015) 88 0.07 ± 0.73 0.8

1.6 Objectives and Hypothesis

The primary objective of this study is to determine the deglacial age of the Sturgis

Moraine. One of the goals is to generate a surficial geology map of the study area. The hypothesis to be tested is that the Sturgis Moraine is between 16,700–15,00014C yrs BP and sand dunes dated from the Mongo Dune field will be significantly younger and both will correspond with cooling events in the Greenland ice core records.

Chapter 2

Methods

2.1 Introduction

To test the hypothesis, sediment cores from scour lakes associated with the

Sturgis Moraine were collected and analyzed. Radiocarbon dated organic material from the bottom of these cores provides the minimum-age chronology for the moraine. Cores were collected using a Livingstone corer and were processed at the University of Toledo

Glacial Lacustrine and Sediment Stratigraphy Lab (GLASS Lab). Analysis included photography, core descriptions, grain size analysis and magnetic susceptibility.

Terrestrial macrofossils were radiocarbon dated to provide the minimum-age chronology for the moraine. Core stratigraphy was examined to better understand the depositional environment from which the terrestrial macrofossils were collected.

2.2 Site Selection

Three lake sites to be cored were chosen based on their proximity to the distal side of the moraine. It is assumed that the lake basin was deglaciated immediately as ice receded from the Sturgis Moraine and that organic material collected from the basal sediment will provide a close minimum age for the moraine. Often kettle lakes are chosen for dating moraines but kettle lakes were avoided, because with kettles there is potential that the buried ice could take a significant amount of time (thousands of years)

to melt (Florin and Wright, 1969). This lag times means that radiocarbon dates are not accurate limiting minimum ages for deglaciation. To avoid kettle lakes, longer narrower lakes interpreted to be scour basins within a tunnel channels were chosen. Their elongated shapes indicate formation by fluid scour and not by ice block melt out.

2.3 Coring

Cores were extracted using the Livingstone coring method (Livingstone, 1955).

Two sediment cores were extracted from the frozen surfaces on Fennel, Wall, and Sweet

Lakes in January 2014. A Livingstone hydraulic system developed by the University of

Toledo was used to extract sediment cores more evenly and deeply from the lake bottom.

The corer was erected close to each basin’s center (deepest water depth). The first core

(A-core) from the study site was taken in 1- meter increments, or thrusts, starting at the sediment/water contact. Once refusal was reached in the first hole, the hydraulic system was moved laterally 1–2 meters from the initial hole, and the second core (B-core) was similarly recovered. The starting depth for the B-core was generally 150 cm lower in order to overlap the gaps that occur in the A-core.

During coring field notes were taken that included an overview of the material recovered, sedimentary breaks, the calculated depth, and the total length of the thrust.

Each thrust section was wrapped in plastic wrap and encased in a 2” PVC pipe for transport. Cores were stored in a 4°C cooler until processed.

2.4 Photography

Cores were photographed immediately after being spilt as the material often contains structures and colors that change after exposure to light and oxidation. The photographs were used to record sedimentary structures and to correlate between thrust sections.

Photos were taken in the University of Toledo Geology Wet Lab. Cores were placed on a wooden board with a meter sick and placed below the lens of a Nikon camera mounted on a copy stand. Photos were taken at 20 cm intervals. Attached to the copy stand were tungsten work lamps angled at 45 from the work surface.

Images were downloaded from the camera card and compiled with Adobe

Photoshop 7.0. For each thrust, the jpegs files were stitched together, and then flattened into one layer.

2.5 Magnetic Susceptibility

Magnetic susceptibility (MS) offers a relative measure of iron, magnetite, and other magnetic sediments deposited into the lake basin over time. Materials that exhibit high susceptibility tend to be wind-blown and fluvial sediments, and glacially derived diamicton (Dearing, 1999). Sediment rich in organics and water exhibits little to no susceptibility.

Following the methodology from Dearing (1999), the magnetic susceptibility throughout the sequence was recorded at 1-cm intervals. A Barrington magnetic susceptibility instrument with a MS-2 probe was used, which reports values in units (x10-

5 SI). At each 1-cm interval, a reading was taken in three positions along the rounded edge of exposed core: closest to the operator, at the top of the rounded outside of the 29

thrust, and in the back, away from the operator (referred to as frontal, medial, distal, respectively).

2.6 Loss-on-Ignition

Loss-on-Ignition (LOI) values were determined using the methods outlined by

Heiri et al. (2001). Samples were removed from the spilt side of each thrust. A sample volume of 0.5 cm3 was taken at 1-cm intervals. The samples were dehydrated in pre- weighted porcelain crucibles at 110°C. Once devoid of residual water, the crucibles were cooled and weighed. All samples were heated to 550°C for one hour in a muffle furnace in order to burn off organic matter. After a second round of cooling and weighing crucibles, the samples were returned to the oven and fired to 1000°C for two hours to remove carbonate and reweighed again.

2.7 Grain Size Analysis

To determine the origin or sediments within the cores, grain size analysis was determined using a Malvern Mastersizer 2000 laser diffraction particle size analyzer equipped with a 2000MU dispersion unit. Grain size analysis methods followed those found in the Malvern Instruments Operation Guide (1999) and User’s Manual (2007),

Campbell (2009), and Hanes (2011). Samples to be analyzed were chosen where changes in lithology were apparent, and at various locations throughout the cores.

2.8 Radiocarbon Dating

Basal radiocarbon ages are the critical data for testing the hypothesis of this study.

One sample for analysis was collected during initial examination of the core where a wooden log was located. For the rest of the samples, the macrofossils submitted for dating were obtained from the base of the lower-most thrust from each lake. Sediment in 30

2-cm thick packages was removed systematically staring from the bottom of the core.

The sediment was digested in 50–60 mL 10% KOH at temperature of 60–80C.

Materials were then passed through a 345 m Mesh Sieve (US Standard #45) in order to separate out large particles. Macrofossils found in this manner were then washed and cleaned with distilled water and a size 000 artist’s paintbrush. Once extracted, macrofossils were stored in a small vial of distilled water with a drop of 10% HCL to prevent new organic growth.

In preparation for submission, selected samples were dehydrated at 70–80C on a watch glass. Once dried, the samples were individually weighed on an analytical balance. Material meeting the minimum weight requirement (5–10 mg) were prepared and packaged for submission to the accelerator mass spectrometry (AMS) facility, Woods

Hole Institute. The Woods Hole Institute followed the procedures outlined in Karlen

(1964), Olsson (1970), Stuvier and Polach (1977) and Stuiver (1980) to determine and report the radiocarbon data. The radiocarbon data included the radiocarbon age with error and the δ13C value.

2.9 Radiocarbon to Calendar Year Conversions

All radiocarbon dates from these study sites were converted to calendar years BP

(cal yrs BP) using the conversion program Calib 7.0 and the intcal 04.14c dataset available online through the Queen’s University of Belfast (Stuiver and Reimer, 1993).

The 68% (1) and 95% (2) mean probability values generated with Calib 7.0 were used.

Dates were rounded to the nearest decade.

2.10 OSL Dating

Optically stimulated luminescence (OSL) dating was used to determine absolute ages for sand dunes in the Mongo Dune Field, south of the Sturgis Moraine near the town of Mongo, Indiana (Figure 6). OSL determines the age when quartz or feldspar grains were last exposed to sunlight. Sample collection was based on procedures outlined in

Aitkens (1998). With an aluminum irrigation tube, samples were collected from the C- horizon of hand-dug soil pits. Soils pits were dug at the crest of each sand dune. Dr.

Kenneth Lepper at the Luminescence Geochronology Laboratory at North Dakota State

University performed the analyses of the OSL samples following procedures outlined in

Lepper et al. (2007).

2.11 Surficial Mapping

The map in this study is a representation of the surficial geology of the Sturgis

Moraine. The type of surficial material was determined from soil descriptions and water- well log stratigraphy. Soil descriptions were downloaded from the USDA’s SSURGO dataset and manipulated in ArcGIS. Water-well log information was downloaded from the Indiana DNR water-well record database (http://www.in.gov/dnr/water/3595.htm) and

Michigan’s DEQ water-well record retrieval system (http://www.deq.state.mi.us/well- logs). Landform identification and classification was interpreted from high resolution

DEMs and LiDAR imagery generated in ArcGIS. The extent of individual landforms was determined by breaks in elevation and confirmed by the parent material of the soil within the outlined landform. Landform boundaries and sedimentary units were traced and color-coded in Adobe Illustrator. Map design was modeled after the Indiana Geological

Survey’s digital compilation of northern Indiana Surficial Geology Maps in Steuben and 32

Lagrange Counties (Fleming et al., 1997; Brown et al., 1998; Karaffa, 2009). The map legend used was from the Indiana Geologic Survey’s Miscellaneous Map 49 (Gray,

1989).

Chapter 3

Results

3.1 Introduction

This study used radiocarbon dated basal ages from lacustrine sediment cores and

OSL dated sand dunes to constrain the timing of Quaternary events in the study area.

The oldest radiocarbon age from all lakes is the best (most limiting) minimum age for deglaciation of the Sturgis Moraine because sedimentation within the lake could only start after ice retreated from the moraine. The basal age for Fennel Lake is 13,700±60

14C years BP (16,530 cal yrs BP), Sweet Lake is 13,300±60 14C years BP (16,000 cal yrs

BP), and Wall Lake 13,750±80 14C years BP (16610 cal yrs BP). The oldest minimum age for deglaciation of the Sturgis Moraine is 13,750±80 C14 years BP (16610 cal yrs

BP). OSL dated sand dunes documented two Aeolian events in the Mongo Dune field.

One event is centered on ̴12.2 ka and an older event on ̴14.1 ka. The oldest sand dune

(14.4±0.5 ka) can be used as a minimum age for deglaciation of the outwash valley and the Sturgis Moraine because the outwash has to be deposited before the sand dune was activated. Although these techniques are independently significant, their value is reinforced when they are placed in a regional context provided by detailed surficial mapping. 34

3.2 Surficial Mapping

Surficial sediments of the Sturgis Moraine can be classified into four different types of sediment, which are outwash, till, aeolian, and alluvium deposits (Figure 3-1).

Deposits of outwash include undifferentiated outwash, sourced from the Saginaw or

Huron-Erie lobe, thick outwash fans of sand and gravel that extend to the southwest of the Sturgis Moraine, and outwash within an outwash valley sourced from the Huron Erie lobe in the southeast corner of the map. The undifferentiated outwash fills in the non- distinct low areas between the Sturgis Moraine and the drumlinized till plains to the northeast on the map. The large outwash fans that extend beyond the Sturgis Moraine slope gently to the south-southwest. Well logs indicate a sand and gravel thickness of 20–

60 m. Linear, northeast-trending collapsed zones mark places where the outwash fans were deposited when ice still filled tunnel channels. The lowermost parts of the fans are cross-cut by a younger outwash valley scoured from the Huron-Erie Lobe to the east.

This outwash valley is referred to as the Lima Plain in the quadrangle report for

LaGrange County (Fleming et al., 1998). The Lima Plain consists of medium to coarse– grain sand and gravel outwash. The upper stratigraphy is mainly coarse gravel with sporadic lenses of sand. The modern channels of the Pigeon and Fawn Rivers traverse this valley.

Figure 3-1. Surficial geology map of the Sturgis Moraine and surrounding topography.

The moraine topography, the drumlinized till plains to the north, and the older till plains to the south consist of sandy loamy till. The Sturgis Moraine trends northwest- southeast and diminishes in size to the east, where it extends into northernmost Indiana.

The moraine consists of stacked slabs of diamiction interbedded with glaciofluvial sediments (Flemming et al., 1998). The crest of the moraine is hummocky and mantled with ice-walled lake plains with a total relief of about 25 m. The drumlinized till plains to the north of the moraine were described and mapped in more detail by Fisher and

Taylor (2002), Fisher et al. (2005), and Kehew et al. (2012). Fisher and Taylor (2002),

Fisher et al. (2005) and Kehew et al. (2012) refer to the drumlinized till plains as upland blocks separated by lowlands intersected by a network of tunnel channels. The drumlins are assumed to record the last ice flow direction, (northeast to southwest) perpendicular to the orientation of the Sturgis Moraine. Mapping the older till plains to the south of the

Sturgis Moraine was outside the scope of this project. It was previously mapped as gravel outwash and sandy diamicton by Flemming et al. (1998).

Sand dunes located in the southeast section of the mapped area, near the town of

Mongo, IN are made up of well-sorted, fine-to-medium-grained aeolian sand. All dunes in the study area were deposited by a westerly wind and are presumably sourced from the sand of the Lima Plain. The parabolic dunes are greater than 50 m in length, and 3 m in relief.

The alluvium deposits in the mapped area consist of alluvium within modern river valleys. The largest rivers in the mapped area, from to north to south are Prairie, Fawn, and Pigeon Rivers. These rivers are incised into outwash plains. The channels of these

relatively fast-flowing rivers are underlain mainly by gravelly alluvium atop outwash

(Flemming et al., 1998).

A complex of tunnel channels provides a constraint on the evolution of the subglacial landscape of the study area. The term tunnel channel is adapted by Clayton et al. (1999) and implies that the lateral depression in the landscape (often a modern valley) formed from a subglacial channel at bank full discharge, where the size of the channel matches the size of the valley. The term tunnel channel is genetic rather than descriptive because these channels reflect subglacial hydrologic conditions and bank full discharge.

The Sturgis Moraine is cross-cut by two generations of tunnel channels (Figure 1-3). The younger generation of tunnel channels (dark green) terminates in the thick outwash fans

(pink) extending in front of the moraine. These northeast-southwest trending channels usually contain a series of closed-basin depressions formed from ice collapse sometime after deposition of the outwash fans. An older generation of tunnel channels (light green) is similar in appearance to the younger generation, but they extend to the southwest beyond the Lima Plain. This older generation of channels can be traced to the La Grange

Moraine (Figure 1-3), which is located south of the Sturgis Moraine and was not mapped in this study.

3.3 Coring Sites

Sediment cores were extracted from Fennel, Wall and Sweet Lakes (Figure 3-2)

(Table 3.1). All lakes are located within tunnel channels that cross cut the Sturgis

Moraine. The elongated shape and irregular bathymetry of these lakes suggests that they were formed by glaciofluvial scour. Sedimentation into these closed basins started once the ice margin had retreated from the location. The closed basin nature of these lakes 38

(Figure 3-2) means sediment and terrestrial organics deposited into these lakes must either be wind-transported or washed in from the surrounding slopes. Basal radiocarbon dates from these depressions will provide a minimum age for deglaciation of the Sturgis

Moraine because the ages must postdate ice retreat from the overlying land surface. The tunnel channels that the lakes are located in increase in elevation from northeast to southwest towards the distal end of moraine. As ice pulled back from the moraine these channels would become inactive as they became detached from the subglacial and ice marginal meltwater system. Since these lakes are closed basins located near the crest of the moraine (the highest elevation in the area) they would have formed fairly quickly after ice retreated from the moraine, making them excellent locations to collect minimum ages for deglaciation.

Table 3.1. Coring Site Locations

Lake Latitude Longitude UTM (mE) UTM(mN) Elevation(m) Water Name Depth (m) Sweet 41°48’12.87” 85°21’13.24” 16T 4629286 271 7.01 636774 Fennel 41°45’04.74” 85°16’40.67” 16T 4623556 279 13.7 643179 Wall 41°43’41.40” 85°12’13.95” 16T 4621076 287 10.5 649515

Figure 3-2. Coring locations. Lakes from left to right are Sweet, Fennel and Wall Lake.

3.4 Core Lithology

The stratigraphic sequence recovered for all three sites is shown in Figure 3-3 and classification of sediment facies is guided by Schnurrenberger et al. (2003). The overall succession of sedimentary units was similar for each site. Generally, the cores transitioned from a clastic facies of sand or pebble sand into a finer-grained facies of silty sand and finally into an organic-rich sapropel facies. Since the samples for radiocarbon ages dating in this study were extracted from the basal section of each core, the lithology for those sections are described in greater detail below.

Figure 3-3. Stratigraphy for sites examined in depth for this study. Facies key is the same for all subsequent logs.

3.5 Fennel Lake Core

The lacustrine sequence recovered at Fennel Lake (Figure 3-4), is 6.5 m thick. A more detailed description of the basal section of the Fennel Lake core is shown in Figure

3-4. This basal section includes a facies of fine-grained sand at a depth of 645–650 cm below the water/sediment contact. Terrestrial macrofossils used for radiocarbon dating were extracted from this unit. Above the sand is about 90 cm of dark grey (5Y 4/10) silty sand containing a log with intact bark. The next facies up section of the sand is a 50 cm facies of fine-grained silt. Above this facies there is ̴ 55 cm thick facies of sandy silt that is a lighter grey color (5Y 6/0) than the facies below. This facies contains laminations of very-fine-grained silt with evidence of bioturbation. Up section of the sandy silt facies is a 4.5 m of dark brown (5Y 1/2), gelatinous sapropel. The sapropel is highly organic, as recorded in the LOI values above the contact, and the abundance of aquatic plant debris observed throughout the facies. The net organic material is extremely low in the basal sand facies and silt-rich faces (except for the log). The transition up section into sapropel is also marked by a visible change in color (Figure 3-4). The carbonate content remains relatively constant ( ̴̴ 40%) throughout the entire core. The magnetic susceptibility remains high ( ̴10 x10-5 SI) throughout the basal section implying a higher-energy environment.

Figure 3-4. Lacustrine stratigraphy and radiocarbon dates for the basal section of the Fennel Lake core. LOI 500C is the black line and LOI 1000C is the grey line. The red line is the magnetic susceptibility reading.

3.6 Wall Lake Sediment Core

The lacustrine sequence recovered at Wall Lake (Figure 3-3), is 6.7 m thick. A more detailed description of the basal section of the core is shown in Figure 3-5. This basal section includes a facies of fine-grained, pebbly sand at a depth of 670–660 cm below the sediment/water contact. Terrestrial macrofossils used for radiocarbon dating were extracted from this unit. The next facies up-section of the sand is 10 cm of fine, silty sand. The material is dark grey in color (5Y 4/0) and contains terrestrial macrofossils. Above this facies there is 90 cm of fine-grained silt that is a lighter grey color (5Y 6/0) than the facies below. This facies contains laminations of very fine- grained sandy silt. These light grey laminations are almost white in color (5Y 9/0) and are repeated throughout the basal section (Figure 3-5). Up-section of the silt is a 3.5 m thick facies of dark brown, gelatinous sapropel. Across the contact between these facies there is a rapid increase in organic content and an abundance of aquatic plant debris.

Similar to the Fennel Lake lacustrine sediment core, the net organic material remains extremely low ( 5%) in the pebbly sand and silt-rich facies, but rapidly transitions up section into the sapropel facies (30%). The carbonate content remains relatively constant throughout the entire core ( ̴̴ 40%). The MS reading remains high for the bottom

1.5 m indicating coarse sediment. The MS value decreases in the fine-grained silt facies.

Figure 3-5. Lacustrine stratigraphy for the basal section of the Wall Lake core. LOI 550C is the black line and LOI 1000C is the grey line. Red line is the magnetic susceptibility reading.

3.7 Sweet Lake Sediment Core

The lacustrine sediment sequence recovered at Sweet Lake (Figure 3-6) is 5.5 m thick. A more detailed description of the basal section of the core is shown in Figure 13.

This basal section includes a facies of fine-grained sand at a depth of 550-545 cm below the water/sediment contact. Terrestrial macrofossils used for radiocarbon dating were extracted from this unit. Above this facies is 5 cm of fine-grained, silty sand, dark grey in color (5Y 4/0). Above this facies is 150 cm of fine- grained silt that is lighter grey in color than the sediment below (5Y 6/0). This facies contains laminations of very-fine grained, sandy silt. These thin laminations (less than 1 cm thick) differ from the laminations in the Wall Lake Core because they are a darker grey color (5Y 3/0), darker than the silt facies. Up-section of the silt is a facies of dark brown, gelatinous sapropel about 3.5 m thick. Similar to the Wall Lake core the contact between these facies is documented by a rapid increase in organics and the abundance of aquatic plant debris.

Similar to the other cores, the net organic material remains extremely low in the initial sandy, then silt-rich facies. The carbonate content remains relatively constant throughout the entire core ( ̴̴ 40%). The MS reading pulses through the basal section. It is high during the pebbly sand facies and decreased up section into fine-grained silty facies.

Figure 3-6. Stratigraphy of the basal section of the Sweet Lake core. Black line is LOI 500°C and the grey line is LOI 1000°C. The red line is the magnetic susceptibility reading.

3.8 Radiocarbon Dates and Interpretations

All radiocarbon and calibrated age conversions for each lake are reported in Table

5. The oldest minimum age for each lake is the best minimum age for lake formation.

The basal age for Fennel Lake is 13,700±60 14C years BP (16,530 cal yrs BP), Sweet

Lake is 13,300±60 14C years BP (16,000 cal yrs BP), and Wall Lake 13,750±80 14C years

BP (16610 cal yrs BP). The oldest age from all lakes is the best (most limiting) minimum age for deglaciation of the Sturgis Moraine because sedimentation within the lake could only start after ice retreated from the moraine. The oldest minimum age for deglaciation of the Sturgis Moraine is 13,750±80 14C years BP (16,320–16,930 cal yrs

BP).

Table 3.2: Radiocarbon Ages

Lab Material Glass Lab # δ13C Radiocarbon Age Mean Two ơ Number Dated Age Error Probability range (14C years BP) (yr) (Calendar (Calendar years) years)

OS- wood log FenLK1401 -24.89 13,450 60 16180 15,960- 111979 16,400 OS- plant/wood FenLK1403 -25.54 13,700 60 16530 16,289- 111980 16,810 OS- plant/wood FenLK1404 -24.5 13,550 55 16320 16,120- 111981 16,550 OS- plant/wood SweetLK1401 -25.95 12,000 55 13860 16,290- 111982 14,020 OS- plant/wood SweetLK1402 -25.81 13,150 60 15800 15,566- 111983 16,030 OS- plant/wood SweetLK1403 -23.95 13,300 60 16000 16,120- 111984 16,550 OS- plant/wood WallLK1401 -26.52 13,650 60 16450 16,290- 111985 16,810 OS- plant/wood WallLK1402 -25.86 13,750 80 16610 16,320- 111986 16,930 OS- plant/wood WallLK1403 -29.78 13,700 60 16530 16,220- 111987 16,720

The facies of each basal section of the cores were dated. In Fennel Lake the sand facies was deposited at 13,700±60 14C years BP (16,530 cal yrs BP), the sandy-silt facies was deposited at 13,550±55 14C years BP (16320 cal yrs BP) and the silty facies was deposited at 13,450±60 14C years BP (16180 cal yrs BP). These dates are all within a 100

14C year span of each other; which implies a high sedimentation rate of 3cm/year.

Rapid deposition of clastic material, high MS reading and low organic content in the basal sections of the cores is characteristic of unstable slopes adjacent to the lake in close contact with an ice margin (Figure 3-7). In this study, each core bottom terminated in sand or in pebble sand facies and transitioned to organic-rich sediments around 2 m.

Although the cores did not penetrate through the basal facies, the low organic content and high MS reading represents high-energy conditions adjacent to the glacier and high sedimentations rates such as fluvial input, bottom currents of final flow in the tunnel channel before the lake formed. Sedimentation rates of modern glaciers are typically high. For example, in Sunwapta Lake, Alberta as of 1974 between 2 and 11 m of sediment was deposited in the lake basin 34–14 years following deglaciation (Gilbert and

Shaw, 1981). This rapid deposition of clastic material dilutes any organic matter developing within the lake. As ice melts back and a source of glacial meltwater is cut off, and the landscape stabilizes, then sapropel is allowed to accumulate in the lake basin.

Figure 3-7. Modern day example of sediment rich, unvegetated sloped around a recently deglaciated lake basin in contact with the Matanuska Glacier in southern Alaska. The melting glacier creates an unstable environment and meltwater brings large amount of sediment into the basin. Circled, crouching person for scale (photo was taken by the author of this study the summer 2014).

3.9 OSL Ages

All OSL ages for this study are reported in Table 6. Location of OSL samples from within the C horizon is show on Figure 14. Equivalent dose distributions obtained from these samples were symmetric (M/n  1.05) and had equivalent dose rates well within acceptable limits in accordance with procedures outlined in Lepper et al. (2000,

2007). These OSL ages represent the last time the sand dune was active. There appear to be two aeolian events documented in the Mongo Dune field. One event is centered on ̴12.2 ka and an older event on ̴14.1 ka. The older dunes are located within the outwash

valley near the Pigeon River, while the younger dunes are higher in elevation above the outwash valley (Figure 3-8). The oldest sand dune (14.3±0.6 ka) can be used as a minimum age for deglaciation of the outwash valley and the Sturgis Moraine because the outwash has to be deposited before the sand dune was activated.

Table 3.3. OSL ages for samples collected from the Mongo Dune Field

a c Sample ID NDe M/m Equivalent Dose Rate Age±se Uncert. (all data) Doseb (J/Kg/ka) (ka) (ka) (Gy) Mongo1401 91/95 1.01 14.429±0.465 1.141±0.109 12.6±0.4 1.3 Mongo1402 90/94 0.99 14.093±0.501 1.144±0.111 12.3±0.4 1.3 Mongo1403 90/95 1.02 13.555±0.500 1.097±0.111 12.4±0.5 1.3 Mongo1404 86/95 0.97 13.747±0.551 0.959±0.092 14.3±0.6 1.5 Mongo1405 90/96 0.98 12.836±0.478 0.911±0.100 14.1±0.5 1.6 Mongo1407 95/96 1.02 13.737±4.700 1.147±0.117 12.0±0.4 1.3 a Number of aliquots used for OSL De calculation/number of aliquots from which OSL data was collected b Equivalent doses are the mean and standard error of the OSL De distributions c Presented as calculated OSL age ± std.err. (uncertainty); convention proposed in Lepper et al. (2011).

Figure 3-8. A) is the location map of Mongo dune field and sample locations. B) is soil pit descriptions and OSL ages of the Mongo dunes. Red star is depth in the soil pit at which the OSL sample was taken.

Chapter 4

Discussion

4.1 Radiocarbon Ages and Implications

The minimum age for deglaciation of the Sturgis Moraine is 13,750±80 14C years

BP or 16,320–16,930 cal yrs BP. Although the radiocarbon chronology data for deglaciation of the Saginaw lobe in northern Indiana and southern Michigan is very limited, when the minimum ages from this study are compared to the chronology of the adjacent lobes, the results suggest a more contemporaneous ice margin retreat than what was originally proposed in the literature (Figure 4-1). Table 4.1 presents the regional network of minimum deglacial ages and their locations. Clear Lake is associated with the deglaciation of the Valparaiso Moraine, a moraine of the Michigan Lobe, while the Hyre,

Kenan, Pyles and Ladd Lake sites are associated with deglaciation of the Fort Wayne

Moraine, a product of the Huron-Erie Lobe (Figure 4-1).

Table 4.1. Locations of bog and lake sites in southern Indiana, including basal dates.

Site Site Lat., Lab Basal age Median Age 2ơ Reference no. name Long number (14C yrs BP) age range Probability (cal yr (cal yrs BP) BP) 1 Clear 41.650, IU-143 13,300±300 16,100 16,920– 1.000 Bailey, Lake -86.530 15,100 1972 2 Ladd 41.417, ISGS- 14,680±310 17,860 18,569– 1.000 Shane; Lake -84.750 1679 17,123 Anderson 1993 3 Hyre 41.267,- Beta- 13,690±50 16,830 17,003– 1.000 Glover et 84.947 19405 16,662 al., 2011 4 Kena 41.367,- Beta- 13,880±70 16,950 17,174– 1.000 Glover et n 83.841 190864 16,771 al., 2011 5. Pyle 40.667,- ISGS- 13,510±160 16,600 16,993– 0.9804 Shane; 84.883 1055 15,853 Anderson 1993

Figure 4-1. The location and basal ages (ka 14C yrs BP) of lake sites in northern Indiana in yellow. Blue dashed line is the new ice margin at 16,600 cal yrs BP proposed in this study. The yellow line is the approximate location of the Valparaiso Moraine and the green lines are the approximate location of the Fort Wayne and Wabash Moraine. Image shows that minimum-limiting ages for the Sturgis Moraine are similar to minimum-limiting ages for the Fort Wayne and Valparaiso Moraine. These ages suggest a younger and coincident ice margin (dash blue line) for this area. 54

The distribution of all 14 basal ages in tables 3-2 and 4-1 ranges from 14,680±310

14C years BP to 13,300±60 14C years BP. This distribution is characteristic of minimum- limiting datasets. Another example of broad distributions in minimum-limiting datasets is found in Bromley et al. (2014). Bromley et al. (2014) dated moraines in the western sector of Rannoch Moor, a former ice cap in Scotland using similar methods to this study.

Bromley et al. (2014) generated a chronology that ranged from 9,140±180 14C years BP to 10,555±65 14C years BP. These broad distributions are characteristic of minimum- limiting datasets and confirm that basal ages do not represent a single event

(deglaciation) but a process: the progressive postglacial colonization by plants.

Plant colonization of landscapes is not an individual event but a prolonged process, which occurs over considerable timescales. As is evident in newly deglaciated terrains today, initial vegetation of the Sturgis Moraine likely would have been sparse and discontinuous and poorly represented in the sediment record. This can be observed by the low organic content in the basal section of the lacusturine cores collected in this study.

There is also an unknown time lag between deglaciation and the onset of plant colonization. The lag cannot be assessed by the coring method used in this study. The amount of time it would have taken a pioneer species to move into the newly deglaciated landscape is unknown. This time lag could have been even longer if remnant buried ice remained in the tunnel channel depressions. Modern analogues of vegetation colonization from retreating glaciers in the Arctic suggest that this time lag is short and may be within the dating error. Jones and Henry (2003) determined that the woody

shrub Salix Arctic and mosses Polytrichum-Pogonotum sspp developed on a terrain in as little as 33 years after deglaciation on Ellesmere Island. Moreau et al. (2008) sampled

300 plots in the 1 km long forefield of the glacier Midtre Lovenbreen in Spitsbergen deglaciated within the past 100 years. They found 16 species of vascular plants on outwash deposits and 15 species on moraines, with 12 of those species colonizing moraines in less than 30 years. Similar to sediment cores in northern Alberta (Fisher et al., 2009), the basal sections of the Sturgis Moraine cores have a high and rapid sedimentation rate that buried basal, and dated macrofossils. Such results may imply that the time lag between deglaciation and vegetation colonization at the Sturgis Moraine was short.

In order to provide a minimum-limiting age for deglaciation it is necessary to obtain the oldest organics from as close to the basal facies as possible. Collecting the deepest sediments from the surface of the lake basin is often challenging. The combination of these factors can result in an age distribution reflecting basal ages that are younger than the onset of plant colonization. For this reason the earliest probable age for the onset of plant growth for each moraine is represented by the oldest radiocarbon age.

For the Sturgis Moraine the oldest radiocarbon age is 13,750±80 14C years BP with a mean calibrated age of 16,610 cal yrs BP and 2 range of 16,32016,930 cal yrs

BP. There is a 98% (2 range) probability that the deglaciation of the Sturgis Moraine was completed before 16,930 cal yrs, the oldest possible age of the 2 range. An older age exists for the Fort Wayne Moraine, which is 14,680±310 14C years BP from the Ladd

Lake site with a 2 range of 17,12318, 569 cal yrs BP. This date should be used with 56

caution because it is a bulk sediment radiocarbon age, and may contain aquatic species contaminated by old carbon as discussed in MacDonald et al. (1991). For this reason the next oldest date should be considered the earliest probable age for the onset of plant growth on the Fort Wayne Moraine. The next oldest age is 13,690±50 14C years BP from the Hyre Site. This age has a mean calibrated age of 16,830 and a 2 range of

17,00316,662 cal yrs BP. This means the minimum limiting age for the Fort Wayne

Moraine is 16,662 cal yrs BP, the oldest possible age of the 2 range.

The oldest radiocarbon age for the Valparaiso Moraine is 13,300±300 14C years

BP from the Clear Lake site (Bailey, 1972). This age has a mean calibrated age of 16,100 cal yrs with a 2 range of 15,10016,920 cal yrs BP. This minimum age requires that deglaciation there is a 98% (2 range) probability that the Valparaiso Moraine had to have been completed before 16,920 cal yrs BP, the oldest possible age of the 2 range.

Unfortunately, this age is one of the only ages for the deglacation of the Valparaiso

Moraine. More basal ages are needed to increase confidence that this radiocarbon age is the best possible minimum age for the moraine.

In order to assess whether the similar ages of these sites are statistically significant a summed probability diagram for all radiocarbon dates in Table 3-2 and

Table 4-1 was generated (Figure 4-2). Calibrated to calendar years this distribution has a possible 2 range of 18,56913,869 cal yrs BP. The three populations of calibrated ages in Figure 16 overlap at the 2 range suggesting that their ages are indistinguishable from each other, and that the ice margins are similar in age.

Figure 4-2. The sum probability of radiocarbon ages vs. calibrated ages. Calibrated using the InterCal09 curve. The new ages for the Sturgis Moraine agree with the previously published radiocarbon ages in northern Indiana and their similar ages suggests that all three lobes of the LIS had retreated from northern Indiana by 16,920 cal yrs BP, the oldest minimum age of the 2 range for the Sturgis Moraine.

4.2 Radiocarbon ages and other paleoclimate proxies

The cluster of basal ages around 16,600 cal yrs BP presented in Table 3-2 and

Table 4-1 suggest widespread deglaciation in northern Indiana at this time. The summed probability diagram for basal dates from the Sturgis, Valparaiso, and Fort Wayne moraines (Figure 4-3) corresponds with warming recorded in Midwest terrestrial transition dates (Glover et al., 2011). These ages occur during the Heinrich Stadial and close to the peak of Heinrich Event 1 (16,800 cal yrs) (Hemming, 2004; Barker et al.,

2009). The Midwest terrestrial transition dates consists of a radiocarbon dated record showing significant organic, sedimentological or pollen changes in core sequences from small terrestrial basins, located in central Indiana and Ohio (Figure 4-4). Glover et al.

(2011) dated the first major sedimentary transition up section of the lowermost part of the sediment which represents the shift from silt or clay accumulation to material of greater organic content. These transitions indicate a switch from an unstable glacial environment to a more stable environment within the basin. In particular, the decline of Picea (spruce) species is seen as a transition to warmer climates for the region (Shane and Anderson,

1993). The transitions in each basin occur around 16,800 cal yrs BP and represent a regional warming, switching from a glacial climate.

Also, the basal radiocarbon data set in Glover et al. (2011) suggests a warming event for the Midwest region at 16,800 cal yrs BP but they had low confidence that this event was real because of a lack of basal ages around this time. The new ages presented in this study add support to Glover’s suggested warming event around 16,800 cal yrs BP.

Figure 4-3. Summary of climate events during the last 30,000 years adapted from Glover et al. (2013). Climate events and their dates that are well-documented in the Northern Hemisphere are labeled on the top (Denton et al., 2010). T1=Termination 1 (22,00018,000 cal yrs), HS1=Heinrich Stadial 1 (18,00014,600 cal yrs; Barker et al., 2009), B/A=Bolling-Allerod (14,70012,700 cal yrs), and YD= Younger Dryas (12,80011,500 cal yrs). The black line represents the approximate minimum age for deglaciation of the Sturgis Moraine (16,660 cal yrs BP). A) Probability summation of the basal ages from the Sturgis, Fort Wayne and Valparaiso moraines. B) Probability summation of transition dates from the Midwest (Glover et al., 2011). C) Greenland ice core oxygen isotope data (Anderson et al., 2006; and Rasmussen et al., 2006). D) Ice-rafted debris (Bard et al., 2000). The minimum ages of the Sturgis, Fort Wayne and Valparaiso Moraines aligns with the ages of the Midwest terrestrial transition dates. This alignment indicates a climate warming in Indiana, Ohio and Michigan at this time that caused deglaciation of these moraines and a switch from a glacial to nonglacial environments within basins in central Indiana and Ohio. The Greenland ice core records do not show warming for this time period. This implies that climate forcing affected the Greenland Ice Sheet and the LIS differently and the Greenland ice core records may not be an accurate way to understand the retreat of the LIS. The increase of ice rafted debris shortly after the deglaciation of the Sturgis, Fort Wayne and Valparaiso moraines suggests a more widespread melting of ice around this time. 60

Figure 4-4. Sites that have formed and/or undergone an environmental transition during Heinrich Stadial 1 from Glover et al. (2011). Black box is the location of Sturgis Moraine. It should be noted that both radiocarbon databases presented in this study and in

Glover et al. (2013) found discrepancies between the apparent onset of deglaciation in the

Midwest and the Greenland isotope records. This problem is not unusual and is discussed in detail elsewhere (Berger and Loutre, 1991; Clarke et al., 2009; Glover et al.,

2013). This discrepancy hints at a limitation of the Greenland oxygen isotope record as a direct proxy for ice sheet forcing everywhere. In North America variations in northern summer isolation values may have forced the beginning of the ice-sheet melting (Clark et

al., 2009; Denton et al., 2010) while full-glacial conditions remained in Greenland

(Svensson et al., 2008).

4.3 New Age Control on the Sturgis Moraine and the current radiocarbon chronology of the Great Lakes Region The ice margin presented in this study (Figure 4-1) is very different from the ice margin for 16,610 cal yrs BP in the current literature (Figure 4-5). According to the current literature (Larson and Schaetzl, 2001), the ice margin for the Saginaw Lobe at this time has retreated far enough northward that Lake Maumee expanded into the southern part of the Huron basin and drained north and the west across the “thumb” of southern

Michigan via the Imlay Channel (Leverett and Taylor, 1915, Eschman and Karrow,

1985). Initially Lake Maumee drained southwest into the Wabash Valley across the Fort

Wayne Moraine through the Fort Wayne outlet allowing for a late deglaciation across the thumb of Michigan. The change from southwest drainage to drainage through the Imlay

Channel was originally used to explain several lake phases known as Maumee I, II, and

III (Leverett and Taylor, 1915). However, Bleuer and Moore (1972) demonstrated that the Fort Wayne outlet sill is low enough for all stages of Lake Maumee to drain out the

Fort Wayne outlet. There is little age control on Lake Maumee. The timing for Lake

Maumee has been assigned from by one radiocarbon date of 13,770 ± 210 14C years BP

(16,395 cal yrs) (Table 2) (Buckley and Willis, 1972). There is little information about this radiocarbon age in the literature beyond it is from plant remains, leaves, and twigs from the Weaver Drain site in Lapeer Co., MI (Figure 4-5). The dated organic matter was found within unoxidized beds of sands; silts, and clay above poorly-sorted coarse sand overlying till in presumable the Imlay Channel (Buckley and Willis, 1972). Farrand

commented in Buckley and Willis (1972) that although organic matter is post-till in age, modern ecology and range of species suggest the time lag between deposition of till and plant material was short; and the date represents ice retreat during Lake Maumee III. The

Maumee shoreline on the Defiance Moraine was also OSL dated at 16.7±0.8, 16.3±0.9,

15.4±0.7, and 16.8±1.1 ka (Fisher et al., 2015 in press).

Figure 4-5. Location and general extent of LIS ice margin and Lake Maumee and glacial Lake Chicago around 16,400 cal yrs BP as presented in the current literature (Larson and Schaetzl, 2001). The dotted blue line represents the Sturgis Moraine (16,600 cal yrs). The star is the location of the Weaver Drain site where the one radiocarbon date of 13,770 ± 210 14C years BP, 16,395 cal yrs BP, which is used to constrain Lake Maumee is located (Buckley and Willis,1972). The inset map is from Larson and Schaetzl, 2001). 63

The current age control presented in this study from the Sturgis Moraine also conflicts with the timing of glacial Lake Arkona, which succeeded Lake Maumee. Lake

Arkona occurred when the ice margin retreated far enough north across the “thumb” of

Michigan to allow for the merging of water in the Huron and Erie basins with water in the Saginaw lowlands (Figure 4-6) (Leverett and Taylor, 1915; Fullerton, 1980). Lake

Arkona drained into Grand Valley via the outlet near Maple Rapids (Eschman and

Karrow, 1985; Barnett, 1985, 1992). The age control on Lake Arkona is also minimal and is constrained by one radiocarbon date of 13,600±500 C14 years BP (2 range14,75017,824 cal yrs BP) from driftwood found between deposits of Arkona and

Whittlesey sand in Cleveland, Ohio (Suess, 1954). Dates from driftwood should be used with caution because the amount of time the wood was transported is unknown, and it was one of the first radiocarbon dates to be determined, and should be redated to minimize the error of ± 500 years.

Figure 4-6. Location and general extent of LIS margin and glacial Lake Arkona and glacial Lake Chicago around 16,170 cal yrs BP as presented in the current literature (Larson and Schaetzl, 2001). The dotted blue line represents the Sturgis Moraine (16,600 cal yrs). Lake Arkona is dated at 13,600±500 (16,174 cal yrs) from driftwood found in Cleveland, Ohio (not pictured) to the east of this map (Suess,1954). The inset map is from Larson and Schaetzl, 2001).

The new age control for the Sturgis Moraine presented in this study appears to conflict with the timing of Lake Maumee and glacial Lake Arkona. While there are very few radiocarbon dates to constrain lakes Maumee and Arkona, the few available dates are similar to the minimum age for the Sturgis Moraine. There are several possibilities for this overlap. If the age control for lakes Maumee and Arkona and the Sturgis Moraine are all correct, along with their interpretation, the Saginaw lobe would have had less than 400 years to retreat from the Sturgis Moraine to the Saginaw Bay lowlands (350 km), to allow for drainage through the Imlay channel, which is about 870 m/a. There is no age control on the Saginaw Lobe north of the Sturgis Moraine to generate a rate of retreat. Elsewhere though, authors have reconstructed the rate of retreat for the LIS. In New England,

Ridge et al. (1999) employed varve techniques to track the ice sheet margin and found a constant retreat rate of 155 m/a. Along the southern portion of the Des Moines Lobe,

Lepper et al. (2007) used radiocarbon ages from maximum stratigraphic positions and basal organics from lakes to propose rates from 105 to 270 m/a. In the Fort McMurray,

Alberta, Fisher et al. (2009) found rates from 23 to nearly 300 m/s. Lowell et al., 2009 found similar rates of retreat in Thunder Bay, Ontario ranging from (150 to 250 m/a).

Based on these estimates a rapid retreat of the Saginaw lobe is possible, but probably not as rapid as 870 m/a.

Another possibility for the timing of Lake Arkona is that it is younger than originally predicted. Radiocarbon dating technology and methods have greatly progressed since the sample for Lake Arkona was first processed. Redating this site to take advantage of technological advances in radiocarbon dating (e.g. dating small

samples of in situ terrestrial macrofossils) and OSL dating of strandlines should better constrain the age of Lake Arkona. Regardless, more age control on the various shorelines of ancestral Lake Erie is needed to understand the timing of the various lake stages and how the glacial lakes evolved with the retreat of the Saginaw lobe.

4.4 Sand Dune Chronology and Implications

There appears to be two aeolian events documented in the Mongo Dune Field.

One event is centered on 12.2 ka and an older event on 14.1 ka. The older dunes are located within the Lima Plain outwash valley near the Pigeon Creek, while the younger have migrated on the upland surface above the outwash valley (Figure 3-8). One possible explanation for why the older dunes were not reactivated is that higher elevations are more susceptible to desiccation and loss of vegetation because of lower groundwater tables, and the dunes in the valley bottom were more shielded from the wind and likely remained vegetated from a higher groundwater table. desiccation and loss of vegetation because of lower groundwater tables and the dunes at the lower elevation where shielded from the wind and had access to a higher groundwater table.

The sand dunes in this study were the same age as dunes in northwest Ohio dated by Campbell et al. (2011) and Fisher et al. (2015). Fourteen out of the 22 dates on dunes overlap at one sigma error with the Younger Dryas Stadial (Figure 4-7). The Younger

Dryas cold period has been described as cooler and windier (Alley, 2000) in both North

America (Thorson and Schile, 1950; Leavitt et al., 2006) and Europe (e.g., Rebollan and

Perez-Gonzalez, 2008). Younger Dryas aged dunes are also reported from central

Wisconsin (Rawling et al., 2008) and northwest Indiana (Kilibarda and Blockland, 2011),

and New England (Thorson and Schile, 1995). As suggested in Campbell et al. (2011) their data implies that during stadials in the Great Lakes region, landscape instability was driven by arid conditions and intense winds. In general, OSL-dated sand dunes and beach ridges (Figure 4-7) in northwest Ohio and the dated sand dunes from this study record dune mobility and deflation processes overlapping in times with cooler periods and episodes of abrupt climate change as recorded in the Greenland ice cores (Fisher et al., 2015). Fisher at al. (2015) suggest that dune mobility may have been caused by lake level drops in ancestral Lake Erie, but the dated dunes in this study are not located close to an ancient shoreline. This suggests that climate was more of a driving factor in dune mobility in the Mongo Dune Field.

Figure 4-7. Comparison of littoral and aeolian events in northwest Ohio and northeast Indiana with Greenland ice-cores climate proxy data (Rasmussen et al., 2006). Earlier work was by Campbell et al., (2011) and Higley et al. (2014). Sand dunes Mongo 14-1, 14-2, 14-3, 14-7 correspond with the Younger Dryas cold period and support earlier work that the timing of Greenland stadial corresponds with aeloian activity in the Midwest. Older sand dunes (Mongo 14-4 and 14-5) correspond with older dunes in northwest Ohio and littoral deposits of Lake Warren (Fisher et al., 2015). 68

Chapter 5

Summary and Conclusion

5.1 Summary

Understanding the timing and rate of ice retreat in the Great Lakes Region is critical to understanding the relationship between ice lobes and climate during the Late

Wisconsinan time period. The primary purpose of this study was to constrain the age of the Sturgis Moraine, and end moraine of the Saginaw Lobe in south-central Michigan using minimum-limiting radiocarbon and OSL ages.

Previous correlations between till stratigraphy and end moraines suggest that the

Sturgis Moraine formed between 15,500 and 16,100 14C yrs BP, but with little direct supporting chronologic data. To test this hypothesis, Livingstone sediment cores were collected from three scour lakes, within tunnel channels at the distal side of the Sturgis

Moraine. The sediment cores were photographed and described in the G.L.A.S.S. Lab, including photographs, loss on ignition, magnetic susceptibility, grain size analysis, and sedimentology. Terrestrial macrofossils from the basal section of each core were extracted for radiocarbon dating.

The oldest minimum age for each lake is the most limiting age for lake formation.

The basal age for Fennel Lake is 13,700±60 14C yrs BP, Sweet Lake 13,300±6014C yrs 69

BP and Wall Lake is 13,750±6014C yrs BP. The oldest most limiting minimum age for deglaciation of the Sturgis Moraine is 13,750±80 14C yrs BP or 16,32016,930 cal yrs

BP. These ages are similar to basal ages from Clear Lake (13,300±300 14C yrs BP), a kettle lake on the distal side of the Valparaiso Moraine of the Lake Michigan Lobe, and the Hyre (13,690 ± 50 14C yrs BP), Kenan (13,880±70 14C yrs BP) and Pyles

(13,510±160 14C yrs BP) sites from kettle lakes on the distal side of the Fort Wayne

Moraine of the Huron- Erie Lobe. The minimum-limiting ages for the Sturgis Moraine presented in this study suggest a younger ice margin by about 2,000 radiocarbon years and more contemporaneous ice retreat with adjacent lobes than originally predicted.

Therefore the hypothesis that the Sturgis Moraine was deglaciated between 15,500 and

16,100 14C yrs BP is rejected.

The new age control for the Sturgis Moraine presented in this study does not support Fullerton’s (1980) age assignments for the Sturgis Moraine and other Saginaw

Lobe Moraines in south-central Michigan. The new minimum age for the Sturgis

Moraine means the Fullerton’s (1980) age assignments should be rejected because it is

2,000 radiocarbon years older than the original age assignment. Age assignments for

Saginaw Lobe moraines north of the Sturgis Moraine also need to be reevaluated because these moraines must have minimum ages younger than 13,750±80 14C yrs BP or

16,32016,930 cal yrs BP.

In order to better understand the relative chronology of glacial landforms associated with the Sturgis Moraine a surficial sediment map of the area was generated.

The type of surficial material for this map was determined using soil descriptions and water-well log stratigraphy. Six OSL-samples were collected and processed from the 70

Mongo Dune Field, near Mongo IN. There appear to be two aeolian events documented in the Mongo Dune field. One is centered on  12.2 ka and an older event on  14.1 ka.

The younger event corresponds with the Younger Dryas cold period and the older event is the same age as older dunes in northwest Ohio. These older dune ages are also a minimum age for the Sturgis Moraine, although this is not a very limiting minimum age.

5.2 Conclusions

The primary objective of this study is to determine the deglacial age of the Sturgis

Moraine by testing the hypothesis that the moraine is between 16,70015,000 14C yrs BP.

Minimum limiting radiocarbon age for the Sturgis Moraine is 13,750 ± 80 14C yrs BP

(16,32016,930 cal yrs BP). Therefore this hypothesis is rejected. This age suggests a younger ice margin, by about 2000 yrs, and more contemporaneous ice retreat of adjacent lobes than originally predicted. Dated sand dunes from the Mongo Dune Field, record to be two aeolian events at 12.2 ka and 14.1 ka. The older dunes are a minimum age for the Sturgis Moraine, although not a very limiting minimum age.

5.3 Future Work

The deglacial chronology of the Sturgis Moraine can be improved in several ways. A near absolute age of the moraine should be obtainable if OSL samples are taken from the outwash fans on the distal end of the moraine, this approach was attempted in this project but the outwash material was too coarse. It is suggested instead to sample from gravel pits located on the very distal edge of the outwash fans where fine grained sand is most likely located. The minimum-limiting ages presented in this study could be compared to ages extracted from ice-walled lake plains associated with the moraine. The

sediment architecture of ice walled-lake plains, the lower contact is inset into a matrixsupported diamicton, and their curved shaped suggests that the basal ages from these landforms dates when active glacial ice switched to stagnating conditions. The minimum-limiting ages presented in this study could also be compared to basal ages from kettle lakes from within the outwash fans. Glover’s (2011) minimum-limiting ages where extracted from kettle lakes and it is a popular method to date ice margins. Basal ages from kettle lakes from within the Sturgis Moraine outwash fans can also be directly compared to minimum-limiting ages from scour lakes presented in this study and this could also be an exercise to compare the various methods of dating different geomorphic environments. At this point, there no age control on any other moraine of the Saginaw

Lobe. Other moraines of the Saginaw Lobe need to be dated in order to better understand the rate of retreat for this lobe.

References

Aitken, M.J., 1998, An introduction to optical dating: the dating of Quaternary sediments by the use of photon-stimulated luminescence, Oxford University Press, New York.

Andersen, K.K., Svensson, A., Johnsen, S.J., Rasmussen, S.O., Bigler, M., Rothlisberger, R., Ruth, U., Siggard-Anersen, M.L, Steffensen, J.P., Dahl-Jensen, D., Vinther, B.M., Clausen, H.B., 2006, The Greenland ice core chronology, 15-42 kyrs. Part 1: constructing the time scale: Quaternary Science Reviews, v. 25, p. 3246-3257.

Bard, E., Rostek, F., Turon, J.-L., and Gendreau, S., 2000, Hydrological impact of Heinrich events in the subtropical northeast Atlantic: Science, v. 289, no. 5483, p. 1321-1324.

Barendsen, G. W., Deevey, E. S., and Gralenski, L. J., 1957, Yale natural radiocarbon measurements III: Science See Saiensu, v. 126.

Barker, S., Diz, P., Vautravers, M. J., Pike, J., Knorr, G., Hall, I. R., and Broecker, W. S., 2009, Interhemispheric Atlantic seesaw response during the last deglaciation: Nature, v. 457, no. 7233, p. 1097-1102.

Barnett, P.J., 1985, Glacial Retreat and Lake Levels, North Central Lake Erie Basin, Ontario, in Calin, P.E., ed., Quaternary evolution of the Great Lakes: Geological Association of Canada Special Paper, v. 30.

Barnett, P., 1992, Quaternary geology of Ontario: Geology of Ontario Special Paper, v. 4, no. Part 2, p. 1011-1088.

Black, R., 1976, Quaternary geology of Wisconsin and contiguous upper Michigan: Quaternary Stratigraphy of North America: Stroudesbourg, Pennsylvania, Dowden, Hutchinson and Ross, p. 93-117.

Berger, A., and Loutre, M.F., 1991, Insolation values for the climate of the last 10 million years: Quaternary Science Reviews, v. 10, no. 4, p. 297-317.

Blewett, W. L., Winters, H. A., and Rieck, R. L., 1993, New age control on the Port Huron moraine in northern Michigan: Physical Geography, v. 14, no. 2, p. 131- 138.

Blockland, J. D., 2013, The surficial Geology of Fulton County, Ohio, insight into the Late Pleistocene-Early Holocene glaciated landscape of the Huron-Erie lake plain, Fulton County Ohio, USA: University of Toledo MS Thesis.

Buckley, J.M., Willis, E.H., 1972, Isotopes' Radiocarbon Measurements IX: Radiocarbon, v. 14, p. 114-139.

Buckley, J.M., 1976, Isotopes' Radiocarbon Measurements XI: Radiocarbon, v. 18, p. 172-189.

Broecker, W. S., and Farrand, W. R., 1963, Radiocarbon age of the Two Creeks forest bed, Wisconsin: Geological Society of America Bulletin, v. 74, no. 6, p. 795-802.

Bromley, G.R., Putama, A.E., Rademakera, K.M., Wincklerb, G., Lowell, T.V., Schaeferb, J.M., Halla, B., Birkela, S.D., Harold, W.B., 2014, Younger Dryas deglaciation of Scotland driven by warming summers: Proceedings of the National Academy of Sciences, v. 111.17, p. 6215-6219.

Calkin, P.E., and McAndrews, J.H., 1969, Dating late glacial recession and vegetation in the Erie Basin, northwestern New York: Geological Society of America Abstracts with Programs for 1969, v. 1, p. 5.

Calkin, P.E., 1970, Strand lines and chronology of the glacial great lakes in Northwestern New York: The Ohio Journal of Science, v. 70, no. 2, p. 78-96.

Calkin, P.E., Brett, C.E., and Krajewski, J.L., 1975, Ancestral Niagara river deposits- stratigraphy and paleontology: Geological Society of America Abstracts with Programs, v. 7, no. 1, p. 37.

Calkin, P. E., and Feenstra, B. H., 1985, Evolution of the Erie-basin great lakes: Quaternary Evolution of the Great Lakes, v. 30, p. 149-170.

Clark, P.U., Dyck, A.S., Skakun, J.D., Carlson, A.E., Clark, J., Wohlfarth, B., Mitrovica, J.X., Hostetler, S.W., McCabe, A.M., 2009, The glacial maximum: Science, v. 325, no. 5941, p. 710-711

Campbell, M. C., 2009, The evolution and chronology of the Lake Warren shoreline in the oak openings region of Ohio, USA: University of Toledo MS Thesis.

Campbell, M. C., Fisher, T. G., and Goble, R. J., 2011, Terrestrial sensitivity to abrupt cooling recorded by aeolian activity in northwest Ohio, USA: Quaternary Research, v. 75, no. 3, p. 411-416.

Clayton, L., 1983, Chronology of Lake Agassiz drainage to Lake Superior: Glacial Lake Agassiz, v. 26, , p. 291-307.

Clayton, L., and Moran, S. R., 1982, Chronology of late wisconsinan glaciation in middle North America: Quaternary Science Reviews, v. 1, no. 1, p. 55-82.

Clayton, L., Attig, J.W., Mickelson, D.M., 1999, Tunnel channels formed in Wisconsin during the last glaciation: Mickelson, D.M., Attig, J.W. (Eds.), Glacial Processes Past and Present, Geological Society of America Special Paper, v. 337, p. 69–82.

Coates, D. R., Landry, S. O., and Dlipe, W., 1971, Mastodon bone age and geomorphic relations in the Susquehanna Valley: Geological Society of America Bulletin, v. 82, no. 7, p. 2005-2010.

Colgan, P.M., Mickelson, D.M., Cutler, P.M., 2005, Ice-marginal terrestrial landsystems: southern laurentide ice sheet margin: Evans, D.J.A. (Ed.), Glacial Landsystems, p. 111-142.

Dearing, J, 1999, Magnetic susceptibility, Environmental magnetism: a practical guide: Quaternary Research Association London, v. 6, p. 35-62.

Denton, G. H., Heusser, C., Lowel, T., Moreno, P. I., Andersen, B. G., Heusser, L. E., Schlühter, C., and Marchant, D. R., 1999, Interhemispheric linkage of paleoclimate during the last glaciation: Geografiska Annaler Series A Physical Geography, v. 81, no. 2, p. 107-153.

Dreimanis, A., 1962, Quantitative gasometric determination of calcite and dolomite by using chittick apparatus: Journal of Sedimentary Research, v. 32, no. 3.

Dreimanis, A., 1966, Lake Arkona-Whittlesey and post-Warren radiocarbon dates from" Ridgetown Island" in southwestern Ontario: The Ohio Journal of Science, v. 66, no. 6, p. 582-586.

Dreimanis, A., 1977, Late Wisconsin glacial retreat in the Great Lakes Region, North America : Annals of the New York Academy of Sciences, v. 288, no. 1, p. 70-89.

Dreimanis, A., 1987, The Port Talbot interstadial site, southwestern Ontario: Geological Society of America Centennial Field Guide–Northeastern Section, p. 345-348.

Drexler, C. W., Farrand, W. R., and Hughes, J. D., 1983, Correlation of glacial lakes in the Superior basin with eastward discharge events from Lake Agassiz: Glacial Lake Agassiz, v. 26, p. 261-290.

Dyke, A. S., 2004, An outline of North American deglaciation with emphasis on central and northern Canada: Quaternary Glaciations Extent and chronology, v. 2, p. 373- 424.

Eschman, D. F., and Karrow, P. F., 1985, Huron basin glacial lakes: a review: Quaternary Evolution of the Great Lakes, Volume 30, Geological Association of Canada, p. 79-93.

Farrand, W. R., 1969, The quaternary history of Lake Superior, in Proceedings 12th Conference. Great Lakes Res, p. 181-197.

Farrand, W.R., and Eshman, D.F., 1974, Glaciation of the southern Peninsula of Michigan-a review: Michigan Acadamician, v. 7, p. 31-56.

Fisher, T. G., and Taylor, L. D., 2002, Sedimentary and stratigraphic evidence for subglacial flooding, south-central Michigan, USA: Quaternary International, v. 90, no. 1, p. 87-115.

Fisher, T. G., Jol, H. M., and Boudreau, A. M., 2005, Saginaw Lobe tunnel channels (Laurentide Ice Sheet) and their significance in south-central Michigan, USA: Quaternary Science Reviews, v. 24, no. 22, p. 2375-2391.

Fisher, T. G., Waterson, N., Lowell, T. V., and Hajdas, I., 2009, Deglaciation ages and meltwater routing in the Fort McMurray region, northeastern Alberta and northwestern Saskatchewan, Canada: Quaternary Science Reviews, v. 28, no. 17, p. 1608-1624.

Fisher, T.G., Blockland, J.D., Anderson, B., Krantz, D.E., Stierman, D.J., Golbe, R., 2015, Testing lake level fluctuations in Ancestral Lake Eriem northwest Ohio: Ohio Journal of Science, in press

Florin, M.-B., Wright, H.E.J., 1969, Diatom evidence for the persistence of stagnant glacial ice in Minnesota: Geological Society of America Bulletin, v. 80, p. 695– 704.

Forsyth, J. L., and Carney, F. L., 1960, The beach ridges of northern Ohio: Division of Geological Survey, Professional Paper, v. 504, p. 13.

Fullerton, D. S., 1980, Preliminary correlation of post-Erie interstadial events (16,000- 10,000 radiocarbon years before present), central and eastern great lakes region, and Hudson, Champlain, and St. Lawrence lowlands: Geological Survey Professional Paper, v. 1089.

Glover, K. C., Lowell, T. V., Wiles, G. C., Pair, D., Applegate, P., and Hajdas, I., 2011, Deglaciation, basin formation and post-glacial climate change from a regional 76

network of sediment core sites in Ohio and eastern Indiana: Quaternary Research, v. 76, no. 3, p. 401-410.

Gravenor, C., and Stupavsky, M., 1976, Magnetic, physical, and lithologic properties and age of till exposed along the east coast of Lake Huron, Ontario: Canadian Journal of Earth Sciences, v. 13, no. 12, p. 1655-1666.

Goldthwait, R.P., 1958, Wisconsin age forests in western Ohio, age and glacial events: Ohio Journal of Science, v. 58, p. 24-27.

Hack, J. T., 1965, Postglacial drainage evolution and stream geometry in the Ontonagon area, Michigan, The United States Geological Survey Professional Paper, v. 504.

Hansel, A., and Johnson, W., 1986, Late quaternary record of the Chicago outlet area: Proceedings Quaternary Records of Northeastern Illinois and Northwestern Indiana: American Quaternary Association, Ninth Biennial Meeting, Field Guide Trip, v., 5.

Hemmings, S.R., 2004, Heinrich events: massive Late Pleistocence detritus layers of the North Atlantic and their global impreint: Reviews of Geophysics, v. 42, p. 1–43

Hough, J. L., 1958, Geology of the Great lakes, University of Illinois Press.

Hough, J. L., 1963, The prehistoric Great Lakes of North America: American Scientist, p. 84-109.

Hough, J.L., 1966, Correlation of glacial lake stages in the Huron-Erie and Michigan basins: The Journal of Geology, p. 62-77.

Howard, J. L., 2010, Late Pleistocene glaciolacustrine sedimentation and paleogeography of southeastern Michigan, USA: Sedimentary Geology, v. 223, no. 1, p. 126–142.

Howard, J. L., Clawson, C. R., and Daniels, W. L., 2012, A comparison of mineralogical techniques and potassium adsorption isotherm analysis for relative dating and correlation of late Quaternary soil chronosequences: Geoderma, v. 179, p. 81–95.

Johnson, W. H., 1986, Stratigraphy and correlation of the glacial deposits of the Lake Michigan Lobe prior to 14 ka BP: Quaternary Science Reviews, v. 5, p. 17–22.

Jones, G.A., Henry, G.H.R., 2003, Primary plant succession on recently deglaciated terrain in the Canadian High Arctic, Journal of Biogeography, v. 30, p. 277–296.

Kaiser, K. F., 1994, Two Creeks Interstade dated through dendrochronology and AMS: Quaternary Research, v. 42, no. 3, p. 288-298.

Karlen, I., Olsson, I.U., Kallburg, P. and Kilici, S., 1964, Absolute determination of the activity of two 14C dating standards: Arkiv Geofysik, v. 4, p. 465-471.

Karrow, P., 1989, Quaternary geology of the Great Lakes subregion: Quaternary geology of Canada and Greenland, no. 1, p. 326-350.

Karrow, P. F., Dreimanis, A., and Barnett, P. J., 2000, A proposed diachronic revision of late Quaternary time-stratigraphic classification in the eastern and northern Great Lakes area: Quaternary Research, v. 54, no. 1, p. 1-12.

Kehew, A.E., Kozlowski, A.L., 2007, Tunnel channels of the Saginaw lobe, Michigan, USA, Special Paper. In: Johannsson, P., Sarala, P. (Eds.), Applied Quaternary Research in the Central Part of Glaciated Terrain, Geological Survey of Finland, v. 46, p. 67-77.

Kehew, A. E., Piotrowski, J. A., and Jørgensen, F., 2012, Tunnel valleys: Concepts and controversies—A review: Earth-Science Reviews, v. 113, no. 1, p. 33-58.

Kilibarda, Z., Blockland, J., 2011, Morphology and origin of the Fair Oaks Dunes in NW Indiana, USA: Geomorphology, v. 125, p. 305–318.

Krist Jr, F. J., Lusch, D. P., Ehlers, J., and Gibbard, P., 2004, Glacial history of Michigan, USA: A regional perspective: Developments in Quaternary Sciences, v. 2, p. 111- 117.

Larson, G., and Schaetzl, R., 2001, Origin and evolution of the Great Lakes: Journal of Great Lakes Research, v. 27, no. 4, p. 518-546.

Larson, G. J., Lowell, T. V., and Ostrom, N. E., 1994, Evidence for the Two Creeks interstade in the Lake Huron basin: Canadian Journal of Earth Sciences, v. 31, no. 5, p. 793-797.

Lepper, K., Larsen N, A.,, McKeever, S.W., 2000, Equivalent dose distribution analysis of Holocene eolian and fluvial quartz sands from Central Oklahoma: Radiation Measurements, v. 32, p. 603-608.

Lepper, K., Fisher, T.G., Hajdas, I., and Lowell, T.V., 2007, Ages for the Big Stone moraine and the oldest beaches of glacial Lake Agassiz, implications for deglaciation chronology: Geology, v. 35, p. 667-670.

Lewis, C.F.M., 1969, Late Quaternary history of lake levels in the Huron and Erie basins: Proceedings of the 12th conference on Great Lakes Research, International Association of Great Lakes Research, p. 250-270.

Leverett, F., and Taylor, F. B., 1915, The Pleistocene of Indiana and Michigan and the history of the Great Lakes: US Government Printing Office United States Geological Survey Monograph, v. 53

Lowden, J.A., Blake, W. Jr., 1968, Geological Survey of Canada Radiocarbon Dates VII: Radiocarbon, v. 10, p. 207-245.

Lowell, T. V., 1995, The application of radiocarbon age estimates to the dating of glacial sequences: an example from the Miami sublobe, Ohio, USA: Quaternary Science Reviews, v. 14, no. 1, p. 85-99.

Lowell, T. V., Larson, G. J., Hughes, J. D., and Denton, G. H., 1999, Age verification of the Lake Gribben forest bed and the Younger Dryas advance of the Laurentide Ice Sheet: Canadian Journal of Earth Sciences, v. 36, no. 3, p. 383-393.

MacDonald, G.M., Beukens, R.P., Kieser, W.E., 1991, Radiocarbon dating of limnic sediments: a comparative analysis and discussion, Ecology, v. 72, p. 1150–1155.

Malvern Instruments, Ltds., 1999, Operators Guide, Manual 0247,Worcestershire, U.K., v. 2 .

Malvern Instruments, Ltd., 2007, User Manual, Manual 0387, Worcestershire, U.K., v. 2.

Matsch, C. L., and Schneider, A. F., 1986, Stratigraphy and correlation of the glacial deposits of the glacial lobe complex in Minnesota and northwestern Wisconsin: Quaternary Science Reviews, v. 5, p. 59-64.

McCallum, K.J., Wittenberg, J., 1965, University of Saskatchewan radiocarbon dates IV: Radiocarbon, v. 7, p. 229-235

McCallum, K.J., Dyck, W., 1960, University of Saskatchewan radiocarbon dates II: American Journal of Science Radiocarbon Supplement, v. 2, p. 73-81.

McCartney, M., and Mickelson, D., 1982, Late Woodfordian and Greatlakean history of the Green Bay Lobe, Wisconsin: Geological Society of America Bulletin, v. 93, no. 4, p. 297-302.

Mickelson, D., Clayton, L., Fullerton, D., and Borns Jr, H., 1983, The Late Wisconsin glacial record of the Laurentide ice sheet in the United States: Late-Quaternary Environments of the United States, v. 1, p. 3-37.

Mörner, N.-A., and Dreimanis, A., 1973, The erie interstade: Geological Society of America Memoirs, v. 136, p. 107-134.

Moreau, M., Mercier, D., Laffly, D., Roussel, E., 2008, Impacts of recent paraglacial dynamics on plant colonization: a case study on Midtre Love´nbreen foreland, Spitsbergen (79N): Geomorphology, v. 95, p. 48–60.

Muller, E.H., Prest, V.K., 1985, Glacial Lakes in the Ontario Basin, in Calkin, P.E., ed., Quaternary evolution of the great lakes: Geological Association of Canada Special Paper, v. 30.

Odgen III, J. G., and Hay, R. J., 1973, Department of Botany and Bacteriology: Radiocarbon, v. 15, n. 2, p. 350-366.

Olsson, I.U., 1970, The use of oxalic acid as a standard: I.U. Olsson, ed., Radiocarbon Variations and Absolute Chronology, v. 12, p. 17.

Parrenin, F., Loulergue, L., Wolff, E., 2007, EPICA Dome C ice core timescles EDC3, IGBP pages : World Data Center for Paleoclimatology Data Contribution, v.083.

Rasmussen, S.O., Andersen, K.K., Svensson, A.M., Steffensen, J.P., Vinther, B.M., Clausen, H.B., SIggaard-Andersen, M.L., Johnsen, S.J., Larsen, L.B., Dahl- Jensen, D., Bigler, M., Rothlisberger, R., Fisher, H., Goto-Azuma, K., Hansson, M.E., Ruth, U., 2006, A new Greenland ice core chronology for the last glacial termination: Jourval of Geophysical Research, v. 111

Rech, J.A., Pigati, J.S., Lechmann, S.B., McGimpsey, C.N., Nekola, J.C., 2011, Assessing open system behavior of carbon-14 in terrestrail gastropod shells: Radiocarbon, v. 53, p. 325-336.

Ridge, J. C., 1997, Shed brook discontinuity and little falls gravel: evidence for the Erie interstade in central New York: Geological Society of America Bulletin, v. 109, no. 6, p. 652-665.

Rubin, M., and Suess, H.E., 1955, U.S. Geological Survey Radiocarbon Dates III: Science, v. 123, no. 3194, p. 442-448.

Rubin, M., and Alexander, C., 1958, U.S. Geological Survey Radiocarbon Dates IV: Science, v. 127, p. 1476-1487.

Ruhe, R. V., 1952, Topographic discontinuities of the Des Moines lobe: American Journal of Science, v. 250, p. 46-56.

Schneider, A. F., and Follmer, L. R., 1983, A buried Sangamon soil in southeastern Wisconsin: Geoscience Wisconsin, v. 7, p. 86-97.

Schneider, A.F., and Hansel J.K., 1990, Evidence for post-two creek age of the type Calumet shoreline: Geological Society of American Special Paper, Late Quaternary History of the Lake Michigan Basin, v. 251, p. 1-8.

Schnurrenberger, D., Russel, J., and Kelts, K., 2003, Classification of lacustrine sediments based on sedimentary components, Journal of Paleolimnology, v. 29, p. 141-154.

Shane, L. C., 1987, Late‐glacial vegetational and climatic history of the Allegheny Plateau and the till plains of Ohio and Indiana, USA: Boreas, v. 16, no. 1, p. 1-20.

Shane, L. C., and Anderson, K. H., 1993, Intensity, gradients and reversals in late glacial environmental change in east-central North America: Quaternary Science Reviews, v. 12, no. 5, p. 307-320.

Stuiver, M., 1980, Workshop on 14C data reporting: Radiocarbon, v. 22, p 964-966.

Stuiver, M. and Polach, H.A., 1977, Discussion of reporting of 14C data: Radiocarbon, v.19, p. 355-363.

Stuiver, M., and Reimer, P.J., 1993, Extended 14C data base and revised CALIB 3.0 14C age calibration program, Radiocarbon, v. 35, p. 215-230.

Suess, H.E., 1954, U.S. Geological Survey Radiocarbon Dates I: Science, v. 120, no. 3117, p. 467-473.

Svensson, A., Andersen, K. K., Bigler, M., Clausen, H. B., Dahl-Jensen, D., Davies, S., Johnsen, S. J., Muscheler, R., Parrenin, F., and Rasmussen, S.O., 2008, A 60,000 year Greenland stratigraphic ice core chronology: Climate of the Past, v. 4, no. 1, p. 47-57.

Totten, S., 1982, Pleistocene beaches and strandlines bordering Lake Erie, Glacial Geology of Northeast Ohio, Ohio Geological Survey Bulletin, v. 86, p.52-60.

Appendix A

Loss-on-Ignition

LOI Fennel Lake Raw Data cr wet dry wt LOI5 LOI1 wt of wt loss at % loss wt loss at % loss wt samp 50 000 sampl LOI 550 for 550 LOI for le e 1000 1000 4.67 7.176 6.405 6.345 5.884 1.728 0.0603 3.4 0.5216 30.1 77 8 7 4 1 4.83 7.794 6.979 6.893 6.568 2.142 0.086 4.0 0.4112 19.1 73 8 9 9 7 6 5.38 6.785 6.590 6.558 6.479 1.208 0.0318 2.6 0.1116 9.2 19 1 7 9 1 8 5.46 7.906 7.263 7.224 6.799 1.801 0.0394 2.1 0.464 25.7 29 7 9 5 9 5.56 6.914 6.384 6.359 6.055 0.817 0.0244 2.9 0.329 40.2 66 6 3 9 3 7 4.49 5.843 5.341 5.316 5.009 0.850 0.0246 2.8 0.3318 39.0 07 8 3 7 5 6 5.08 6.258 5.822 5.799 5.536 0.741 0.0232 3.1 0.2863 38.6 1 7 5 4 7 4.67 6.365 5.725 5.692 5.314 1.050 0.0329 3.1 0.4112 39.1 52 5 6 7 4 4 5.35 6.931 6.349 6.320 5.962 0.994 0.0283 2.8 0.3862 38.8 48 6 7 8 2 5.46 7.745 6.906 6.863 6.354 1.443 0.0437 3.0 0.552 38.2 34 5 9 2 9 5 5.77 7.67 6.992 6.955 6.537 1.220 0.0376 3.0 0.4555 37.3 2 7 1 2 7 5.31 7.440 6.62 6.577 6.131 1.309 0.0421 3.2 0.4886 37.3 03 8 9 4 7 4.90 6.600 5.972 5.937 5.571 1.068 0.0349 3.2 0.4007 37.5 82

36 3 1 3 4 5.40 7.001 6.409 6.374 6.017 1.003 0.0345 3.4 0.3917 39.0 54 6 2 7 5 8 5.25 6.411 5.989 5.963 5.704 0.733 0.0264 3.5 0.2852 38.8 63 8 8 4 6 5 T5 5.70 7.428 6.425 6.338 6.119 0.723 0.0877 12.1 0.3062 42.3 2 5 7 5 7 5.53 8.091 6.577 6.456 6.103 1.044 0.1207 11.5 0.4736 45.3 31 8 3 6 7 2 5.13 7.643 6.076 5.947 5.629 0.945 0.1291 13.6 0.4468 47.2 07 8 5 4 7 8 5.41 8.433 6.695 6.560 6.138 1.277 0.1345 10.5 0.5574 43.6 83 4 4 9 1 5.74 8.380 6.877 6.758 6.387 1.127 0.1188 10.5 0.4894 43.4 95 5 1 3 7 6 4.92 5.536 5.795 5.668 5.381 0.873 0.1267 14.5 0.4136 47.3 23 1 5 8 9 2 4.74 8.732 6.064 5.894 5.430 1.321 0.17 12.8 0.6342 48.0 33 5 4 4 2 1 5.38 7.917 6.283 6.124 5.845 0.900 0.1593 17.6 0.4378 48.6 27 7 5 2 7 8 5.13 7.678 6.270 6.147 5.735 1.137 0.1233 10.8 0.5352 47.0 28 7 4 5 9

T4 5.35 7.909 5.978 5.746 5.621 0.621 0.2313 37.1 0.3569 57.3 62 1 8 2 9 5.12 6.915 5.430 5.242 0.309 0.1878 60.6 5.4301 1755.0 07 4 1 3 4 5.65 8.366 6.245 5.990 5.897 0.592 0.2545 42.9 0.3472 58.5 23 5 1 6 9 8

T6 5.70 8.297 7.202 7.163 6.591 1.492 0.0392 2.6 0.6103 40.8 94 2 2 9 8 4.49 6.988 6.116 6.053 5.588 1.618 0.0635 3.9 0.5287 32.6 79 1 7 2 8 5.40 8.022 6.996 6.879 6.324 1.593 0.1165 7.3 0.6723 42.2 34 1 4 9 1 5.18 8.133 6.935 6.875 6.233 1.747 0.0601 3.4 0.7016 40.1

75 6 2 1 6 7 4.90 6.881 6.166 6.130 5.679 1.261 0.0361 2.8 0.4873 38.6 52 7 5 4 2 3 4.67 7.351 6.402 6.357 5.709 1.728 0.045 2.6 0.6935 40.1 42 5 9 9 4 7 4.48 7.550 6.288 6.217 5.585 1.799 0.0711 3.9 0.7033 39.0 92 1 3 2 1 4.48 6.973 5.894 5.848 5.301 1.413 0.0463 3.2 0.5928 41.9 15 1 5 2 7 5.98 7.982 6.945 6.891 6.400 0.963 0.0536 5.5 0.5452 56.5 2 2 4 8 2 4

LOI Raw Data Sweet Lake cr wet dry LO LOI wt of wt loss at % loss wt loss at % loss wt samp wt I55 100 sample LOI for 550 LOI for 1000 les 0 0 550% 1000 4.8 6.897 6.6 6.6 6.56 1.8146 0.0306 1.6 0.0654 3.6 16 7 306 52 5.0 7.333 7.0 7.0 6.92 1.94699 0.0117 0.6 0.1022 5.2 780 1 25 133 28 1 5.4 7.229 6.3 6.2 6.20 0.8521 0.045 5.2 0.1093 12.8 631 9 152 702 59 5.4 7.583 6.4 6.3 6.31 1.0498 0.061 5.8 0.1473 14.0 077 575 965 02 5.1 6.684 5.9 5.9 5.90 0.8137 0.0295 3.6 0.0877 10.7 802 1 939 644 62 4.8 6.048 5.4 5.4 5.40 0.6464 0.0198 3.0 0.0691 10.6 302 766 568 75 5.3 7.022 6.2 6.1 6.13 0.879 0.0333 3.7 0.0988 11.2 536 9 326 993 38 4.4 5.865 5.1 5.1 5.10 0.7257 0.0371 5.1 0.0887 12.2 67 2 927 556 4 4.6 6.241 5.4 5.4 5.37 0.8223 0.0364 4.4 0.1 12.1 55 8 773 409 73 5.3 7.401 6.4 6.3 6.27 1.0233 0.0577 5.6 0.134 13.0 823 3 056 479 16 5.5 7.564 6.5 6.4 6.36 0.9551 0.0701 7.3 0.1451 15.1 574 8 125 424 74 5.5 6.867 6.1 6.0 6.03 0.6052 0.0557 9.2 0.1062 17.5 317 7 369 812 07 5.3 6.475 5.7 5.6 5.66 0.4134 0.0593 14.3057 0.0922 22.3 84

423 9 557 964 35 5.4 6.872 5.8 5.7 5.75 0.4746 0.0913 19.2 0.1246 26.2 038 9 784 871 38 5.1 6.703 5.6 5.5 5.47 0.4678 0.1015 21.6 0.1325 28.3 347 3 025 01 5.2 6.805 5.6 5.5 5.56 0.4423 0.0994 22.4 0.1297 29.3 566 1 989 995 92 5.3 7.103 5.8 5.6 5.66 0.5191 0.1251 24.0 0.1572 30.2 025 4 216 965 44 5.4 6.781 5.8 5.7 5.73 0.3383 0.0841 24.8 0.1059 31.3 987 4 37 529 11 4.4 5.968 4.8 4.7 4.76 0.3786 0.0848 22.3 0.1064 28.1 897 9 683 835 19 5.6 7.438 6.1 6.0 5.99 0.5158 0.0976 18.9 0.1354 26.2 112 2 27 294 16 4.4 5.816 4.8 4.7 4.72 0.3446 0.0773 22.4 0.1009 29.2 8 8 246 473 37 5.3 7.127 5.9 5.8 5.81 0.566 0.09 15.9 0.1358 23.9 84 6 5 6 42 6.1 7.830 6.5 6.4 6.43 0.4134 0.0981 23.7 0.1225 29.6 397 2 531 55 06 5.5 8.418 6.3 6.1 6.07 0.7521 0.1957 26.0 0.239 31.7 639 8 16 203 7 5.7 7.030 6.0 5.9 5.89 0.3043 0.0958 31.4 0.1107 36.3 011 8 054 096 47 4.4 6.171 5.1 5.1 5.04 0.7582 0.0824 10.86784 0.1522 20.0738 391 4 973 149 51 49 5914 4.9 6.341 5.2 5.1 5.17 0.3597 0.0917 25.4 0.1111 30.8 214 9 811 894 4.3 5.646 4.9 4.8 4.77 0.5234 0.0833 15.9 0.1228 23.4 781 3 015 182 87 4.9 6.552 5.3 5.2 5.21 0.3762 0.1202 31.9 0.1358 36.0 704 3 466 264 08 4.9 6.602 5.3 5.1 5.17 0.4128 0.1259 30.4 0.1433 34.7 072 2 2 941 67 5.7 6.793 6.0 5.9 5.91 0.2726 0.0778 28.5 0.0902 33.0 327 1 053 275 51 5.4 6.936 5.8 5.7 5.71 0.3685 0.1023 27.7 0.1187 32.2 612 2 297 274 1 4.6 6.877 5.2 5.0 5.04 0.5675 0.1579 27.8 0.1804 31.7 612 2 287 708 83 5.1 6.866 5.5 5.3 5.35 0.3906 0.1339 34.2 0.1498 38.3 187 5 093 754 95 5.7 7.663 6.2 6.1 6.09 0.4793 0.1385 28.8 0.1615 33.6 85

733 8 526 141 11 4.9 7.367 5.5 5.3 5.30 0.6011 0.179 29.7 0.2039 33.9 097 2 108 318 69 4.7 6.395 5.2 5.1 5.13 0.4665 0.1056 22.6 0.1283 27.5 976 641 585 58 5.6 7.362 6.0 5.8 5.88 0.3292 0.1268 38.5 0.1452 44.1 97 1 262 994 1 4.7 7.062 5.4 5.2 5.24 0.6476 0.1493 23.0 0.1832 28.2 84 8 316 823 84 5.2 7.123 5.7 5.6 5.58 0.4841 0.1281 26.4 0.1487 30.7 53 3 371 09 84 4.7 6.496 5.2 5.0 5.06 0.4084 0.1232 30.1 0.141 34.5 975 2 059 827 49 5.7 7.761 6.2 6.0 6.05 0.5092 0.1417 27.8 0.1631 32.0 097 8 189 772 58 4.6 7.125 5.2 5.0 5.00 0.5406 0.1886 34.8 0.2102 38.8 792 7 198 312 96 5.5 7.078 5.8 5.7 5.75 0.3142 0.1141 36.3 0.1265 40.2 639 1 781 64 16 5.6 7.716 6.2 6.1 6.12 0.6083 0.1237 20.3 0.1555 25.5 708 6 791 554 36 5.8 7.999 6.4 6.3 6.30 0.647 0.1227 18.9 0.1552 23.9 142 5 612 385 6 5.0 6.428 5.3 5.2 5.26 0.2985 0.0835 27.9 0.0979 32.7 613 598 763 19 5.6 7.008 5.9 5.8 5.84 0.3376 0.0836 24.7 0.1011 29.9 042 4 418 582 07 4.9 7.576 5.2 5.4 5.41 0.3634 -0.1859 -51.1 -0.15 -41.2 047 681 54 81 4.9 6.319 5.6 5.1 5.16 0.6384 0.4184 65.5 0.4318 67.6 627 4 011 827 93

Wall Lake LOI Raw Data cr wt wet dry LIO LOI wt of wt loss % wt % sample wt 550 1000 sample at LOI loss loss loss 550 for for for 550 LOI 1000 1000 5.44315 6.0385 5.886 5.88 5.835 0.44344 0.0032 0.7 0.051 11.5 2 3 8 4 5.39905 5.9944 5.854 5.85 5.807 0.45584 0.0023 0.5 0.047 10.4 2 2 8 5 6.30285 6.8982 6.700 6.69 6.630 0.39744 0.0056 1.4 0.070 17.6 2 4 8 1 6.6444 7.4944 7.142 7.12 6.959 0.498 0.0223 4.4 0.182 36.6 0 5 6.1334 6.9834 6.749 6.73 6.626 0.6158 0.0182 2.9 0.123 19.9 1 5.9857 6.8357 6.512 6.49 6.357 0.5272 0.0204 3.8 0.155 29.5 2 7 4.9547 5.8047 5.499 5.48 5.343 0.5449 0.0178 3.2 0.155 28.5 1 8 5.3957 6.2457 5.908 5.87 5.742 0.513 0.0332 6.4 0.165 32.3 5 9 6.5432 7.3932 6.969 6.93 6.796 0.4261 0.0297 6.9 0.172 40.4 9 4 6.3867 7.2367 6.639 6.59 6.490 0.2526 0.0405 16.0 0.149 59.0 8 2 5.2456 6.0956 5.664 5.63 5.520 0.4184 0.0271 6.4 0.143 34.2 6 3 6.0404 6.8904 6.417 6.39 6.274 0.3768 0.0244 6.4 0.142 37.8 2 6 5.1954 6.0454 5.616 5.58 5.509 0.4208 0.027 6.4 0.106 25.3 9 8 5.9189 6.7689 6.279 6.24 6.156 0.3607 0.0334 9.2 0.123 34.2 6 6 4.7519 5.6019 5.133 5.10 5.018 0.3812 0.0313 8.2 0.114 30.0 1 6 5.9719 6.8219 6.149 6.09 6.012 0.1774 0.0527 29.7 0.136 76.9 6 5 6.2072 7.0572 6.652 6.63 6.567 0.4456 0.0211 4.7 0.085 19.1 1 2 6.1202 6.9702 6.467 6.43 6.360 0.3472 0.03 8.6 0.106 30.7 7 9 5.484 6.334 5.846 5.81 5.735 0.3629 0.0314 8.6 0.111 30.6

5 1 5.9878 6.8378 6.154 6.11 5.998 0.1663 0.0381 22.9 0.156 93.8 6 1 6.078 6.928 6.374 6.33 6.254 0.2962 0.0376 12.6 0.119 40.3 6 5 5.72 6.57 6.073 6.03 5.983 0.3538 0.042 11.8 0.090 25.4 1 1 6.1482 6.9982 6.282 6.22 6.178 0.134 0.0552 41.1 0.103 77.3 7 6 5.324 6.174 5.589 5.54 5.519 0.2655 0.0481 18.1 0.069 26.2 1 7 5.5694 6.4194 5.633 5.57 5.523 0.0644 0.0561 87.1 0.110 171. 7 7 8 5.002 5.852 5.139 5.08 5.039 0.1376 0.0547 39.7 0.100 72.9 4 4 4.8737 5.7237 5.057 5.00 4.976 0.184 0.054 29.3 0.081 44.1 3 3 4.4796 5.3296 4.733 4.68 4.669 0.2534 0.0498 19.6 0.064 25.2 3

Appendix B

Magnetic Susceptibility

Fennel Lake MS reading Depth MS Reading 0 18.7 0 8 0 2 0 18.6 2 18.7 2 5 2 8 2 13.6 4 18.8 4 5 4 11 4 16.5 6 15.8 6 8 6 4 6 10.5 8 15.8 8 9 8 2 8 14.5 10 14.8 10 11 10 0 10 9.5 12 8.8 12 6 12 -1 12 15.4 14 7.8 14 8 14 1 14 15.4 16 9.8 16 17 16 0 16 17.4 18 10.8 18 17 18 0 18 13.4 20 6.8 20 22 20 -2 20 5.3 22 5.8 22 26 22 -3 22 6.3 24 7.8 24 25 24 -3 24 5.3 26 7.8 26 25 26 0 26 6.3 28 7.8 28 26 28 1 28 3.2 30 9.8 30 23 30 1 30 2.2 32 9.8 32 19 32 4 32 0.2 34 8.8 34 20 34 7 34 -0.9 36 8.9 36 24 36 13 36 -1.9 38 8.9 38 22 38 5 38 -1.9 40 2.9 40 26 40 -2 40 -1.9 42 2.9 42 22 42 -5 42 -8 44 3.9 44 15 44 -1 44 -9 46 2.9 46 14 46 -2 46 -9 48 1.9 48 19 48 3 89

48 -8.1 50 3.9 50 26 50 -2 50 -9.1 52 4.9 52 11 52 -3 52 -9.1 54 3.9 54 21 54 -3 54 -9.1 56 3.9 56 25 56 4 56 -9.1 58 3.9 58 27 58 3 58 -9.2 60 2.9 60 11 60 8 60 -5.2 62 8.9 62 11 62 6 62 -4.2 64 5 64 4 64 -5.2 66 20 66 6 66 -5.2 68 18 68 6 68 -3.3 70 22 70 8 70 -3.3 72 21 72 14 72 -3.3 74 25 74 9 74 -3.3 76 28 76 9 76 -2.3 78 33 78 5 78 -2.3 80 28 80 7 80 -4.4 82 18 82 4 82 -5.4 84 7 84 2.6 86 6

88 8 0 5 0 9 0 22 0 28 2 5 2 27 2 23 2 13 4 3 4 35 4 23 4 18 6 3 6 33 6 26 6 21 8 1 8 59 8 30 8 17 10 1 10 22 10 26 10 23 12 1 12 17 12 30 12 19 14 2 14 18 14 25 14 16 16 3 16 22 16 26 16 22 18 3 18 11 18 27 18 21 20 3 20 13 20 26 20 22 22 3 22 12 22 25 22 22 24 2 24 13 24 31 24 22 90

26 2 26 16 26 28 26 17 28 2 28 8 28 31 28 19 30 2 30 8 30 31 30 19 32 3 32 5 32 13 32 25 34 7 34 8 34 16 34 18 36 12 36 7 36 27 36 15 38 12 38 10 38 14 38 12 40 10 40 10 40 14 40 16 42 21 42 14 42 31 42 18 44 20 44 13 44 32 44 18 46 23 46 8 46 33 46 20 48 19 48 3 48 32 48 24 50 14 50 2 50 25 50 24 52 11 52 10 52 37 52 24 54 12 54 22 54 28 54 17 56 14 56 26 56 27 56 20 58 18 58 34 58 30 58 19 60 18 60 42 60 30 60 24 62 17 62 15 62 29 62 21 64 16 64 28 64 26 66 16 66 34 66 29 68 21 68 33 68 26 70 12 70 28 70 21 72 12 72 34 72 21 74 16 74 35 74 33 76 17 76 34 76 28 78 14 78 37 78 22 80 14 82 14 84 16 86 19 88 19 90 22 Sweet Lake MS Reading Depth MS MS Reading Depth Reading 0 12 0 7 2 9 25 8 4 5 50 5 6 3 75 5

8 3 100 5 10 3 125 5 12 3 150 5 14 3 175 3 16 4 200 4 18 5 225 4 20 6 250 10 22 5 275 9 24 6 300 6 26 6 325 3 28 7 350 2 30 9 375 1 32 9 400 2 34 13 425 -1 36 16 450 -1 38 16 475 1 40 21 500 6 42 19 525 1 44 25 550 0 46 38 575 -1 48 33 600 -1 50 43 625 -1 650 -2 675 -3 700 -2 725 -2 750 -2 775 -2 800 0 825 -1 850 0

Wall Lake MS Reading Depth MS MS Reading Depth Reading 0 50 0 14 2 51 25 14 4 7 50 19 6 7 75 17 8 2 100 26

10 58 125 27 12 54 150 27 14 0 175 32 16 0 200 36 18 0 225 38 20 0 250 23 22 0 275 18 24 0 300 27 26 0 325 20 28 0 350 11 30 0 375 9 32 0 400 5 34 0 425 6 36 0 450 4 38 0 475 2 40 0 500 1 42 0 525 4 44 0 550 -1 46 0 575 0 48 0 600 -2 50 0 625 2 52 0 650 -2 54 0 675 -1 56 0 700 -2 58 0 725 -1 60 0 750 -1 62 0 64 0 66 7

Appendix C

Grain Size Analysis

Fennel Lake F1 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0% Coarse Sand (500.00—1000.00 um): 0% Medium Sand (250—500 um): 1.07% Fine Sand (125—250 um): 7.33% Very Fine Sand (62.50—125 um): 14.16%

Coarse Silt (31.0—62.5 um): 17.23% Medium Silt (16.60—31.0 um): 22.67% Fine Silt (7.80—15.60 um): 22.05% Very Fine Silt (3.00—7.80 um): 13.29%

Clay (0.06—3.00 um): 2.20%

Skewness: 2.299 Kurtosis: 6.071

Fennel Lake F2 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0% Coarse Sand (500.00—1000.00 um): 0% Medium Sand (250—500 um): 0.31% Fine Sand (125—250 um): 5.60% Very Fine Sand (62.50—125 um): 11.54% 94

Coarse Silt (31.0—62.5 um): 21.47% Medium Silt (16.60—31.0 um): 24.14% Fine Silt (7.80—15.60 um): 19.33% Very Fine Silt (3.00—7.80 um): 14.77%

Clay (0.06—3.00 um): 2.95%

Skewness: 2.359% Kurtosis: 6.295

Fennel Lake F3 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0% Coarse Sand (500.00—1000.00 um): 0.61% Medium Sand (250—500 um): 4.67% Fine Sand (125—250 um): 6.97% Very Fine Sand (62.50—125 um): 10.99%

Coarse Silt (31.0—62.5 um): 18.23% Medium Silt (16.60—31.0 um): 20.91% Fine Silt (7.80—15.60 um): 17.63% Very Fine Silt (3.00—7.80 um): 15.56%

Clay (0.06—3.00 um): 4.43%

Skewness: 3.101 Kurtosis: 11.114

Sweet Lake S1 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.04% Coarse Sand (500.00—1000.00 um): 4.75% Medium Sand (250—500 um): 24.91% Fine Sand (125—250 um): 23.23% Very Fine Sand (62.50—125 um): 10.57%

Coarse Silt (31.0—62.5 um): 12.23%

Medium Silt (16.60—31.0 um): 12.14% Fine Silt (7.80—15.60 um): 7.42% Very Fine Silt (3.00—7.80 um): 4.08%

Clay (0.06—3.00 um): 0.63%

Skewness: 1.099 Kurtosis: 1.064

Sweet Lake S2 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 0.00% Medium Sand (250—500 um): 1.70% Fine Sand (125—250 um): 12.64% Very Fine Sand (62.50—125 um): 19.47%

Coarse Silt (31.0—62.5 um): 16.04% Medium Silt (16.60—31.0 um): 16.53% Fine Silt (7.80—15.60 um): 16.64% Very Fine Silt (3.00—7.80 um): 14.17%

Clay (0.06—3.00 um): 2.80%

Skewness: 1.1688 Kurtosis: 3.011

Sweet Lake S3 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 0.52% Medium Sand (250—500 um): 13.53% Fine Sand (125—250 um): 26.44% Very Fine Sand (62.50—125 um): 21.25%

Coarse Silt (31.0—62.5 um): 12.95% Medium Silt (16.60—31.0 um): 9.96% Fine Silt (7.80—15.60 um): 7.86%

Very Fine Silt (3.00—7.80 um): 6.11%

Clay (0.06—3.00 um): 1.38%

Skewness: 1.158 Kurtosis: 1.021

Sweet Lake S4 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 0.00% Medium Sand (250—500 um): 0.00% Fine Sand (125—250 um): 7.14% Very Fine Sand (62.50—125 um): 25.32%

Coarse Silt (31.0—62.5 um): 27.45% Medium Silt (16.60—31.0 um): 18.57% Fine Silt (7.80—15.60 um): 10.97% Very Fine Silt (3.00—7.80 um): 8.21%

Clay (0.06—3.00 um): 2.25%

Skewness: 1.124 Kurtosis: 0.915

Wall Lake S1 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 0.06% Medium Sand (250—500 um): 22.87% Fine Sand (125—250 um): 43.70% Very Fine Sand (62.50—125 um): 21.54%

Coarse Silt (31.0—62.5 um): 3.98% Medium Silt (16.60—31.0 um): 2.79% Fine Silt (7.80—15.60 um): 2.50% Very Fine Silt (3.00—7.80 um): 2.15%

Clay (0.06—3.00 um): 0.41%

Skewness: 0.551 Kurtosis: 0.031

Wentworth Size Class Wall Lake W3

Very coarse sand (1000.00—2000.00 um): 0.01% Coarse Sand (500.00—1000.00 um): 8.10% Medium Sand (250—500 um): 27.62% Fine Sand (125—250 um): 24.68% Very Fine Sand (62.50—125 um): 12.15%

Coarse Silt (31.0—62.5 um): 9.15% Medium Silt (16.60—31.0 um): 7.80% Fine Silt (7.80—15.60 um): 5.49% Very Fine Silt (3.00—7.80 um): 3.98%

Clay (0.06—3.00 um): 0.93%

Skewness: 0.0552 Kurtosis: 0.045

Wall Lake W3 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 2.016% Medium Sand (250—500 um): 3.58% Fine Sand (125—250 um): 13.39% Very Fine Sand (62.50—125 um): 19.45%

Coarse Silt (31.0—62.5 um): 16.43% Medium Silt (16.60—31.0 um): 16.07% Fine Silt (7.80—15.60 um): 14.84% Very Fine Silt (3.00—7.80 um): 12.59%

Clay (0.06—3.00 um): 3.48%

Skewness: 1.50 Kurtosis: 1.01

Wall Lake W4 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0.00% Coarse Sand (500.00—1000.00 um): 0.00% Medium Sand (250—500 um): 0.00% Fine Sand (125—250 um): 7.14% Very Fine Sand (62.50—125 um): 25.32%

Coarse Silt (31.0—62.5 um): 27.54% Medium Silt (16.60—31.0 um): 18.57% Fine Silt (7.80—15.60 um): 10.97% Very Fine Silt (3.00—7.80 um): 8.21%

Clay (0.06—3.00 um): 2.25%

Skewness: 1.124 Kurtosis: 0.915

Wall Lake W6 Wentworth Size Class

Very coarse sand (1000.00—2000.00 um): 0% Coarse Sand (500.00—1000.00 um): 0% Medium Sand (250—500 um): 1.07% Fine Sand (125—250 um): 12.65% Very Fine Sand (62.50—125 um): 19.47%

Coarse Silt (31.0—62.5 um): 16.04% Medium Silt (16.60—31.0 um): 16.53% Fine Silt (7.80—15.60 um): 16.64% Very Fine Silt (3.00—7.80 um): 14.17%

Clay (0.06—3.00 um): 2.80%

Skewness: 1.688 Kurtosis: 3.011

100