. ‘13; '0? . u * am:- ice-.'>e~:tz.m ‘ mm. a. uni J. m Iii-L mu m 3—- L n‘ .v '2‘- E g t * ‘5‘" -.._ M4: Ir. k I a. u‘..' v 3.1: ‘1‘ w , i; q .. ‘— v,f ' mag: Jr! ', Jig";- Qua .‘a-w ”ray :» w -A—é ¢ w w 133.9% ' ,. .~ aw. .. "46:3;257' ‘ v .745". J.“ .93: .3." i W III‘I‘I'II'II‘II IIIIIIIIIII‘II‘IIIII'I’IIIIIIII'IIII " I 49434 University K ~——_ This is to certify that the thesis entitled Late Wisconsinan Drift Stratigraphy of the Saginaw and Lake Michigan Ice Lobes in Southwestern Michigan presented by George William Monaghan III has been accepted towards fulfillment of the requirements for ' Masters Geology degree in T) //7 J ,//l hd r professor Grahame J. Larson Dateié‘am /v//. /7X$’ 0.7639 MSU is an Affirmative Action/Equal Opportunity Institution MSU RETURNING MATERIALS: Place in book drop to remove this checkout from FINES will returned after the date fir LIBRARIES your record. be charged if book is stamped below. 35i1177§m4' 5‘1; 2 :1 . - . M15, '1'l\ 'x ~H! ’ K161 _ 168'< és 1143" ’" i 7 SE? 2 8 'I 1 .le hié ,SEP 2 , “15.3% we ll 9E _ LATE WISCONSINAN DRIFT STRATIGRAPHY OF THE SAGINAW AND LAKE MICHIGAN ICE LOBES IN SOUTHWESTERN MICHIGAN By George William Monaghan III A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the Degree of MASTER OF SCIENCE Department of Geological Sciences 198A ABSTRACT Late Wisconsinan Drift Stratigraphy of the Saginaw and Lake Michigan Ice Lobes in Southwestern Michigan By George William Monaghan III Three Lake Michigan Ice Lobe till units are exposed along the Lake Michigan shoreline in southwestern Michigan (Glenn Shores section). The lowermost (Glenn Shores), the middle (Ganges). and uppermost (Saugatuck) tills are each overlain by lacustrine deposits. All three tills are easily differentiated by.7A/lOA X-ray diffraction peak height ratios of their clay-size fraction. X-ray diffraction analyses performed on Lake Michigan Lobe till sampled from the Lake Border, Valparaiso, and Sturgis-Kalamazoo morainic systems suggest till from each correlates with the Saugatuck till at Glenn Shores, while till sampled from the Tekonsha Moraine and subsurface exposures correlates with the Ganges till at Glenn Shores. The occurrence of lake deposits between these tills at Glenn Shores implies a significant retreat of the Lake Michigan Lobe between the formation of the Tekonsha and the Sturgis-Kalamazoo moraines. Clay mineral analyses of Saginaw Lobe tills similarly suggest a significant retreat of the Saginaw Lobe between the Tekonsha and Kalamazoo moraines. DEDICATION To Heidi, for her patience with muddy boots and long drives. ACKNOWLEDGMENTS I am indebted to the members of my committee, Drs. John T. Wilband and David T. Long, for all their help and guidance given during this research, although special thanks must go to my chairperson, Dr. Grahame J. Larson. for his assistance and advice. both personal and professional, given throughout my graduate tenure. I am grateful to the following organizations for providing the funding to support this research: U.S. Environmental Protection Agency: Tri-county Planning Commission; Michigan Department of Natural Resources. Ground Water Quality Division: and the Department of Geology. Michigan State University. I am also indebted to the United States Geological Survey, Coal Division for stipend support over the past few years. I would like to thank Greg Gephart for his discussions, criticism, and help (as well as for stumbling over the Glenn Shores logs) and Steve Dworkin (for stumbling into the Bloomingdale pit). Thanks also go to Gordon Thompson for his dedication and countless hours of sample preparation and x-ray analyses. Although the ideas generated in this paper were derived from uncountable (sometimes painful) hours of discussion with Greg Gephart and Dr. Grahame Larson, I take the blame for any errors. TABLE OF CONTENTS LIST OF FIGURES AND TABLES INTRODUCTION CHAPTER I LAKE MICHIGAN ICE LOBE DRIFT STRATIGRAPHY Introduction Coastal Tills Inland surface tills Inland sub-surface tills . . . . . . . . . . . . Summary of till correlations in southwestern Michigan Regional Lake Michigan Lobe correlations CHAPTER 2 SAGINAW ICE LOBE DRIFT STRATIGRAPHY -Introduction . Till sheets of the Saginaw Lobe End moraines of the Saginaw Lobe . . . . . Summary of till correlations in south- central Michigan Saginaw and Huron-Erie Lobe correlations Saginaw and Lake Michigan Lobe Correlations Summary of correlations Discussion . CHAPTER 3 THE GLACIAL GEOLOGY 0F KALAMAZOO COUNTY, MICHIGAN INTRODUCTION . . . . . Pleistocene Geology . . . . . Fluvial and Lacustrine Deposits Climax- Scotts Outwash Deposits Galesburg- Vicksburg Outwash Deposits Augusta Creek Outwash Deposits . . Glacial Kalamazoo- Gun Plain Outwash Deposits Glacial Lake Deposits . . . Sand and Gravel, Undifferentiated Late Glacial Eolian Deposits Till and Moraine Deposits Tekonsha Moraine . Sturgis Moraine Kalamazoo Moraine Ti... Undifferentiated POST—PLEISTOCENE GEOLOGY . . Marsh and Swamp Deposits . . . . . . . . . . . . . Recent Floodplain Deposits . . . . . . . . . . . . II ll 12 22 28 3h 3h 37 37 AA AB 5A 5A 55 56 57 TABLE or CONTENTS (Cont.l PLEISTOCENE HISTORY Introduction Deglaciation of Kalamazoo County . Formation of the Leonidas drumlin field and Tekonsha moraines. Formation of the Climax- Scotts outwash plain General retreat of the Saginaw and Lake Michigan lobes. Formation of the Galesburg- Vicksburg outwash plain. Development of the Kalamazoo River Valley SUMMARY AND CONCLUSIONS APPENDIX A ANALYTICAL TECHNIQUES . . Semi-quantitative X-ray Analysis Qualitative X-ray Analysis APPENDIX B STATISTICS . . . Population Statistics and Data Tabulation Lake Michigan Lobe Tills Saginaw Lobe Tills T- test Statistics for Selected Populations Lake Michigan Lobe Tills Saginaw Lobe Tills APPENDIX C SECTION DESCRIPTIONS Comstock Section . . . . . . . Ravine Road Section Cooper Section Bloomingdale Section REFERENCES 80 80 BI 8I 82 8h 87 93 97 98 99 IOI lOl lOl 10h IO7 IO7 108 llO lIO llO lll ll3 llS FIGURE FIGURE TABLE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE 2-2. LIST OF FIGURES AND TABLES Map of Late Wisconsinan moraines and location of till sampled from the Lake Michigan Ice Lobe in southwestern Michigan The Glenn Shores Section showing the clay mineral variability of the till units along two vertical sampling transects Radio-carbon dates from organics at the Glenn Shores section Frequency histograms showing clay mineralogy of the three tills exposed at the Glenn Shores section compared with till sampled from Lake Michigan Lobe moraines Bloomingdale section showing the clay mineral variability of the three till units along two vertical sampling transects . . . . . Frequency histograms showing clay mineralogy of till from the Bloomingdale and Kalamazoo County sections compared with the clay mineralogy of till from the Glenn Shores section . Generalized section showing stratigraphic relationships of the Lake Michigan Lobe tili units in southwestern Michigan . . . . Correlation chart of the proposed Lake Michigan Lobe till units in southwestern Michigan with the Lake Michigan Lobe till stratigraphy from northern Illinois . . . . . . . . Map showing moraines in southern Michigan Map snowing Late Wisconsinan moraines of the Lake Michigan and Saginaw lobes and till sample point locations in southwestern Michigan 2h 27 3i 33 39 ‘AI FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE 2-3. Z-A. 2-5. 2-6. 3-2. 3-3- C-2. List of Figures and Tables (cont.) Cross-section showing lithology and clay mineralogy of the Bedford and Fulton tills sampled from the Eaton County borings ‘Frequency histograms showing clay mineralogy of the Bedford and Fulton tills from Eaton County bore-holes compared with histograms of till sampled from Saginaw Lobe moraines in south-central Michigan Generalized northeast to southwest stratigraphic cross-section through Saginaw Lobe deposits in south-central Michigan Correlation chart showing time stratigraphic relationships of tills of Lake Michigan, Saginaw, and Erie lobes in the Great Lakes region Map showing location of Kalamazoo County and the general extent of the Saginaw and Lake Michigan lobes . . . . . . . . . Map showing reconstructed position of the Lake Michigan and Saginaw lobes at about 16,500 BP during the construction of the Tekonsha moraines and the Climax-Scotts outwash plain Map showing reconstructed position of the Lake Michigan and Saginaw lobes at about l5,000 BP during the construction of the Kalamazoo moraines and the Galesburg-Vicksburg outwash plain . . . . . . . . . . . . . . . . . . Till fabric from Cooper exposure Bloomingdale measured section “3 “7 El 53 El 83 88 llZ ll2 INTRODUCTION Although in regions bordering Michigan recent glacial geologic research has been directed towards a Pleistocene, particularly [Wisconsinan, history based on drift stratigraphy, the Wisconsinan glacial history in Michigan remains based primarily on the morphology of end moraines first defined by Frank Leverett (Leverett and Taylor, I915) during the early part of this century. The focus of this research is to establish lithologic criteria which can be used to differentiate and correlate till deposits in southwestern and south-central Michigan. The goal of such research is to set up a stratigraphic system and to relate these units to major flucuations of the Laurentide Ice Sheet. In order to establish such a stratigraphic system in Michigan, two ice lobes, the Lake Michigan and Saginaw lobes, were studied. This involved: I) detailed mapping of geologic deposits, 2) locating and sampling tills in section, borings, and road-cuts, and, 3) analysis of the lithology of the till samples. Laboratory analysis focused primarily on differences in relative amounts of clay minerals found within the till samples. {/T’ This research has shown that the drift in southwestern and I I I I south-central Michigan can be differentiated based on unique clay mineralogies of each till sheet. Once the lithe-stratigraphic un ts were established for the Lake Michigan and Saginaw lobes, broad correlations between stratigraphic systems established for the Lake Michigan, Saginaw, and Huron-Erie ice lobes in other parts of the Great Lakes Region could be proposed. These correlations are based on morphologic relationships of drift deposits within areas where two adjacent lobes actually abutt. Kalamazoo County (Monaghan, et al., I983) was chosen for detailed mapping since at least two time equivalent ice-edge positions associated with both the Lake Michigan and Saginaw lobes (the Kalamazoo moraines and Tekonsha moraines) occur ‘ with in the are: These relationships are critical because by tracing till sheets established for each lobe to these moraines, the stratigraphic systems of both lobes can be chronologically correlated with each other. (Ra The paper is divided into three chapters. Chapter 1 outlines a proposed Late Wisconsinan drift stratigraphy for the Lake Michigan Lobe. Chapter 2, on the other hand, outlines a proposed drift stratigraphy of deposits associated with the Saginaw Lobe and also suggests correlations with other stratigraphic systems in the Great Lakes Region. Chapter 3 presents a detailed description of the drift deposits within Kalamazoo County as well as a summary of the deglaciation history of both the Lake Michigan and Saginaw ice lobes in southwestern and south-central Michigan. Although the techniques employed for clay mineral analysis of till samples is briefly outlined in chapters l and 2, a more detailed description is pre nted in Appendix A. Appendix B contains lists of the raw clay mineral data for each sample analyzed. It also has compilations of the statistics used in this paper. Detailed descriptions of four sections containing buried or multiple till units discussed in chapter I are shown in Appendix C. l0 CHAPTER I LAKE MICHIGAN ICE LOBE DRIFT STRATIGRAPHY INTRODUCTION Over the past two decades much research has been directed towards the establishment of an objective stratigraphy for Wisconsinan drift in the Lake Michigan basin (Frye and Willman, I960; Willman and Frye, I970; Johnson, I976; Lineback, et al., l97h, I979; Acomb, I978; Acomb et al., I982; McCartney and Mickelson, I982). While most of these investigations have relied on unique lithologic criteria for correlating and differentiating till sheets, the subdivision of the Wisconsinan in western Michigan (Melhorn, I95A; Bretz, l95ha, l95hb; Eschman and Farrand, I970; Evenson, I973; Farrand and Eschman, l97h; and Dreimanis, I977) is still based essentially on morpho-stratigraphic correlations that have not changed significantly since Frank Leverett originally mapped the state around the turn of the century (Leverett and Taylor, l9l5). In recent years, some of these correlations have been supplemented by radio-carbon dates of organic material buried between till units (Farrand, et al., I969; Zumberge and Benninghoff, I969; Miller and Benninghoff, I969; Eschman, 1968). The purpose of this chapter is: l) to establish litholcg.c criteria for recognizing Lake Michigan Lobe till units exposed along the Lake Michigan shoreline in southwestern Michigan, and 2) using these lithologic criteria, to correlate the till units along the II lakeshore with Lake Michigan Lobe tills exposed 2 to 50 kilometers further inland. In addition, the tills in Michigan will be correlated with Lake Michigan Lobe till units already defined by Willman and Frye (I970) in northern Illinois. COASTAL TILLS One of the most complete multi-till exposure in southwestern Michigan, informally named the Glenn Shores section (Gephart and Larson, in prep.), occurs in a wave-cut cliff along the Lake Michigan shoreline just west of the town of Glenn in Allegan County (Figure l-l). Here, three tills separated by lacustrine sediments are exposed in a 20 to 25 meter high bluff. The section (Figure l-Z) begins at the present Lake Michigan beach level (I77 m) with a dense, silty-clay till which is informally named the Glenn Shores till. The total thickness of this till unit is not known and only the uppermost meter is exposed. This till sheet is locally intruded by clay diapirs and dips beneath the present Lake Michigan beach both to the north and south of the section. The till has not been observed at any other locality. The Glenn Shores till is overlain by a 5 to ID centimeter thick discontinuous deposit of beach gravel containing detrital spruce wood and peat balls. Two meters of lacustrine sand and laminated silt, also containing detrital organic material, overlies this gravel deposit. Organics from both the beach gravel and sand/silt deposits have been radio-carbon dated from 37,l50 + 5h0 (BETA-33ll) to >h8,000 (ISGS-9h8) years BP (table I). No organic material has l2 FIGURE l-l. Map of Late Wisconsinan moraines and location of till sampled from the Lake Michigan Ice Lobe in southwestern Michigan. Location of sections shown by closed circles. Location of morainal samples shown by open circles with 7A/l0A ratios values labeled beside each point. Moraines of the Lake Michigan and Saginaw Lobes in southwestern Michigan: Lake Michigan Lobe Moraines: Lake Border System: LBM -- Inner and Outer Lake Border moraines Valparaiso System: IVM -- Inner Valparaiso Moraine OVM -- Outer Valparaiso Moraine Sturgis-Kalamazoo System: IKM -- Inner Kalamazoo Moraine OKM -- Outer Kalamazoo Moraine STM -- Sturgis Moraine Saginaw Lobe Moraines: TkM -- Tekonsha Moraine KM -- Kalamazoo Moraine LaM -- Lansing Moraine l3 £22.32; WIIIIIIIIJ mp O .muzaaoz III _>¢h2300 2...!on 6. ouc:too_ III FIGURE l-2. The Glenn Shores Section showing the clay mineral variability of the till units along two vertical sampling transects. l5 Lacuntrlno Sand Saugatuck ' 00 I TI” 0 :° I O .. O Lacustrine Send ELEVATION '(moure) .— I U "0 O Tunuct A I Transact 8 ‘WS —I90 —"5 I 0 IO ‘ O .0 . 3 Lamlnod Silt : . Gravel ° 0 - «,v u._—GlonnShoru 5 o P -- ill ' \ 3 I l T , l been observed above the sand and silt deposits. TABLE I. , Radio-Carbon dates form organics at the Glenn Shores Section : SAMPLE INC.) RADIOfiEIgéoN AG LAB. NUMBER Wood (GSOOI) ‘ 37,l50*540 BETA-33H Wood (68002) 38.30 i 740 BETA-saw >37,ooo Gx-8l93 ' i. PooI (68003) r>43,ooo Sl-5l83 Wood (68005) > 48,000 Ises-94e Above the lacustrine sand and silt is a 5 to 6 meter thick blue-grey, clay-rich till informally named the Ganges till. A 5 to 6 meter thick lacustrine sand deposit overlies the Ganges till; the contact between these two deposits is sharp. The lacustrine deposit above the Ganges till is overlain by a third till that is 2.5 meters thick and is informally named the Saugatuck till. This till is blue-grey in color and is clay-rich; the upper 0.3 meters consists of a brownish-red silty clay. The section is capped by 3 to A meters of Lake Chicago (Glenwood Level) sediments. The contact between this deposit and the underlying Saugatuck till is sharp. l7 FIGURE l-3. Frequency histograms showing clay mineralogy of the three tills exposed at the Glenn Shores section (above dashed line) compared with till sampled from Lake Michigan Lobe moraines (below dashed line). l8 . on". o.us .I on”. “0.0 .>uo.ba no.o zauu o mszmo: mrwzoxwh CO—\CP on." on... on... r IF P b I F d ESE hr o. Ibo - QUIN— b L b Du.u .>uu.ha no.9 zcu: On 1 zmpw>w OON¢E¢J¢XImHomDHw Ii .- D I. P b a cIc . £E=tE nN.o .>uo.hu no.0 zcu: v— z :mwm»m owH¢m¢m4¢> onu— EE%.H van—a . 3... $8.; : E. EVE—.1: w 8% zcuu . n .85. zmhwrw mmomom mxmq m a. :E = a as s. .. AJHH AKMZOJH wwmorw ZZMJO R sage a 8mm acuu .n JJHF HMJDDHZH mmozmo H I I I . . r . . . .._.1L . L . J .n n" .82 2... 4.8.5 6 on... :5: an .. no. .3: Hmmmmsg xuzpcazmw W CLAY MINERALOGY: In order to further define the lithologic characteristics of each of the three tills exposed in the Glenn Shores section, semi-quantitative X-ray diffraction analyses were performed on the clay-size fraction of 85 till samples. These samples were collected at set intervals along both vertical and horizontal transects, as well as randomly throughout the section. Each sample, weighing 25 to 50 grams, was first pre-treated to removed soluble salts, free calcium carbonates, and free iron oxides (Jackson, 1956; Carroll, 1972). The clay-size (<2 mu) fraction was then separated out by gravity fractionation, deposited on glass slides, and air-dried. Prior to X-ray analysis, the slides were placed in a desiccator containing ethylene glycol to expand the vermiculite to IAA and the smectite to 17A - 18A (Jackson, 1956; Carroll, 1972). The prepared slides were then scanned from 2 to 15 degrees 26-on a General Electric X-ray diffractometer (model no. XRD-S), rotated 180 degrees, and scanned again. The diffraction patterns were analyzed by comparing the intensities of the 7A and 10A peaks. These peak heights were measured to the nearest .025 cm from a predetermined base line (Grim, 1968; Carroll, 1972) and the 7A/10A ratio reported as an average of the two rotations. The results of the clay mineral analyses are presented in Figures 1-2 and 1-3. The clay mineral data displayed in Figure 1-2 are from two of the vertical transects. The elevation of each sample point is shown along the vertical axis, while its 7A/IOA ratio value is shown along the horizontal axis. Figure l-3, on the other hand, shows composite histograms of all the clay mineral data from each of the 20 three tills exposed at Glenn Shores. The horizontal axis of each histogram represents the calculated 7A/lOA peak height ratio values and the vertical axis indicates the number of samples that have a particular 7A/IOA value. The data in Figures 1-2 and 1-3 suggest that the three tills exposed in the Glenn Shores section have significantly different clay _ mineral compositions. Figure 1-2 shows that there is no regular vertical trend in clay mineralogy within a till unit, while Figure 1-3 suggests the occurrence of a significant step-wise increase in the amount of 10A clay, relative to 7A clays, from the lowermost till unit to the uppermost till. Based on the data shown in Figure 1-3, the mean 7A/10A ratio for the lowermost till unit is 1.22; the mean ratio value for the middle till is 0.85; and the mean ratio for the uppermost till is 0.58. Application of the T-test to this data shows that the clay mineralogy of each of the three tills is significantly different to the 99.9% confidence level. Since they have significantly different clay mineralologies, the three till units are unlikely to represent tiTT facies of the same ice advance. In addition to semi-quantitative analysis, qualitative clay mineral determinations were also performed on several samples. Sample pre-treatments were identical to that .used for semi-quantitative analysis. However, after gravity separation the clay-size fraction was deposited on porous ceramic plates, instead of glass slides. Each sample was then X-rayed after the following treatments: 1) Mg saturated.and glycerol solvated; 2) K-saturated and air-dried; . 3) K-saturated and heated to 300° C; and A) K-saturated and heated to 21 550° C. Following techniques outlined by Jackson (1956) and Carrol (1972), it was determined that although minor amounts of the minerals chlorite and vermiculite were present in the samples, kaolinite was the dominate mineral occurring in the 7A peak while illite was the dominate mineral in the IDA peak. INLAND SURFACE TILLS Inland from the Glenn Shores section, consecutively younger Late Wisconsinan (late Woodfordian) ice-edge positions of the Lake Michigan Lobe are represented by a series of end moraines (Figure 1-1) that were first identified by Frank Leverett (Leverett and Taylor, 1915) in the early 1900's. From oldest to youngest, they are: the Tekonsha, Sturgis*, Outer and Inner Kalamazoo, Outer and Inner Valparaiso, and Lake Border moraines. These moraines trend in a northeast-southwest direction and roughly parallel the present shoreline of Lake Michigan. To simplify discussion, several of the morainal ridges have been grouped into the following morainal systems: The Sturgis-Kalamazoo, Valparaiso, and Lake Border morainic systems (see Figure l-l). * Although Leverett (Leverett and Taylor, 1915) suggested that the Sturgis Moraine in Kalamazoo and western St. Joseph counties, Michigan, was formed by the Lake Michigan Lobe, he considered it time equivalent to the Sturgis Moraine of the Saginaw Lobe and, thus, stratigraphically older than the Tekonsha Moraine. Recent detailed mapping of the surficial deposits in southwestern Michigan by the author and others (Monaghan, et. al., 1983). however, shows that the Lake Michigan Lobe portion of the Sturgis Moraine more likely represents the outermost ridge of the Kalamazoo Morainic System and is therefore younger than the Tekonsha Moraine. Because of this relationship, the author and others have preferred to call Leverett's Kalamazoo Morainic System (Leverett and Taylor, 1915) the Sturgis-Kalamazoo System. 22 _ FIGURE l-h. Bloomingdale section showing the clay mineral variability of the three till units along two vertical sampling transects. 23 canL nnul oval 8.. 8. 3. on. o. m .oooe-Ch I < 3.2....— O 5:330 :23 .2... 2.30:: 2C. cans—:0 2.30:: .2... (aiueui) NOllVAa'IS 2h CLAY MINERALOGY: In an attempt to correlate till associated with the inland moraines to the tills exposed at the Glenn Shores section, X-ray diffraction analyses were performed on the clay-size fraction of 65 till samples collected from exposures within these moraines. The location of the sample points and associated 7A/10A ratio values is shown in Figure 1-1. Samples were taken generally 2 to 3 meters below the ground surface from fresh exposures. The till samples were analyzed using the same methods applied to the Glenn Shores samples. The results of the above analyses are presented in Figure 1-3 as a series of frequency histograms comparing the calculated 7A/lOA ratios of the moraines with the three tills exposed at Glenn Shores. These histograms show that the clay mineralogy of till associated with the Sturgis-Kalamazoo, the Valparaiso, and the Lake Border morainic systems is most similar to that of the Saugatuck till, while the clay mineralogy of till associated with the Tekonsha Moraine is most similar to that of the Ganges till. Application of the T-test indicates that these associations can be supported to the 99% confidence level. Hence, on the basis of the statistical similarity of clay mineralogy, the author proposes that the till comprising the Lake Border, Valparaiso, and Sturgis-Kalamazoo morainic systems correlates with the Saugatuck (upper) till at Glenn Shores, and that the till comprising the Tekonsha Moraine correlates with the Ganges (middle) till. 25 FIGURE 1-5. Frequency histograms showing clay mineralogy of till from the Bloomingdale (below the upper dashed line) and Kalamazoo County (below lower dashed line) sections compared with the clay mineralogy of till from the Glenn Shores section. 26 a... o.“— D P ané .>uo. an a... scan m .3: sezzou DONazauax some oust ouHmam m oust mmzou uuaoazszooum u a ....s.. . . I . , , 45:. 4. Just wages: mumooz_zooum " raw... .3“...qu : ace :c H“ .3: Jose mums: mucoozszooum m 8.9 .39. no :G = ELF: =I; iw ecu: . some lmmzooc mMmoxm zzmuo " o_.o .suo... —*H m em... a.“ sure rmuooszs mwozac " - . I . . L:H#.. I o m .u .auxt nu.m suuuwm n” Jose .mmoa:.c xuahcoscm L H" 27 INLAND SUB-SURFACE TILLS In addition to the surface tills, four inland sections that contain buried or multiple tills were sampled. The locations of these sections (Bloomingdale, Cooper, Ravine Road, and Comstock sections) are shown in Figure 1-1. The Bloomingdale section (Figure l-h) occurs within a gravel pit southeast of the Inner Valparaiso Moraine near the town of Bloomingdale in northeastern Van Buren County. It contains three till units that are best exposed in the south wall of the gravel pit. Here, the lowest unit is represented by a b to 6 meter thick outwash which is composed of very coarse sand, pebbles, and cobbles. A 2 to 3 meter thick, grey, clay-rich till overlies the outwash. This till unit dips towards the south and is completely absent in the north wall of the pit.. Directly overlying the grey till is a 2 to 3 meter thick brown, sandy-clay till. The contact between these two till units is sharp. Overlying the brown till is a 2 to h meter thick outwash composed of medium to coarse sand and pebbles. The section is capped by a third till that is 1 to 2 meters thick. This till is brown, sandy, and pinches-out in a few places; the contact between it and the underlying outwash is sharp. The texture and discontinuous nature of the surface till suggests that it may be a flow till. CLAY MINERALOGY: In order to correlate the till units exposed at the Bloomingdale section with those exposed at Glenn Shores, clay mineral analyses (identical to those described for the Glenn Shores tills) were performed on a total of 3A till samples. These samples were 28 collected along both vertical and horizontal sample transects, as well as randomly throughout the Bloomingdale section. The results of the analyses are presented in Figures l-h and 1-5. Figure l-A shows the clay mineral data from two of the vertical transects, while Figure 1-5 is a series of 7A/10A histograms of all samples taken from each of the tills. Also shown in Figure 1-5 are the histograms of the clay mineral data from the till units exposed at 'Glenn Shores. Comparison of the Bloomingdale and Glenn Shores clay data (Figure 1-5) suggests that the Bloomingdale lower till is most similar to the Ganges (middle) till at Glenn Shores, while the Bloomingdale middle and upper tills are most similar to the Saugatuck (upper) till. Application of the T-test indicates that these above associations can be supported to at least the 99% confidence level. Therefore, based on the clay mineral similarities, the author proposes that the Bloomingdale middle and upper tills correlate with the Saugatuck (Upper) till at Glenn Shores, while the Bloomingdale lower till correlates with the Ganges (middle) till. Three additional sections, each containing a single till unit buried beneath outwash, occur within gravel pits located east of the Bloomingdale section in Kalamazoo County (Figure l-l). The Ravine Road and Comstock sections contain a brown, sandy, and loose-textured till, while the till in the Cooper_ section is brown, dense, and clay-rich. The clay mineralogies c. samples taken from these exposures is shown in Figure 1-5 as a composite histogram and suggests that the till in each of these exposures correlates with the Ganges (middle) till at Glenn Shores. 29 FIGURE 1-6. Generalized section showing stratigraphic relationships of the Lake Michigan Lobe till units in southwestern Michigan. 30 WEST 15 10 kllomotore 0 I I c o -2‘ c 2 3 a: -K o 4 I I miles 31 FIGURE 1-7. Correlation chart of the proposed Lake Michigan Lobe till units in southwestern Michigan with the Lake Michigan Lobe till stratigraphy from northern Illinois. Notes: 1 - After Zumburge (1960) 2 - After Frye and Willman (1973) 32 moox mAijomZuI< m.<<. Z_OI_o>z v $9000 «3. 0.1.. _ 7 ————-'V Glenn Shores Saugatuck Mb < D 1? O «3. (I O .3 onoamsoufifl . 03.250 moox m3>jom>ox< zm. Fr... (233.3 WEDRON Yorkville V—_7 .3633 3 0 com Ax 7009 33 SUMMARY OF TILL CORRELATIONS IN SOUTHWESTERN MICHIGAN A generalized east-west cross-section through southwest Michigan is presented in Figure 1-6 and summarizes the correlations proposed in this paper for tills of the Lake Michigan Lobe. It shows that the Saugatuck till can be traced inland, along the surface, from the Lake Michigan shoreline to the Sturgis-Kalamazoo Morainic System, while the Ganges till can be traced in the subsurface from Lake Michigan to where it crops out at the Tekonsha Moraine. The presence of lacustrine sands between the Ganges and Saugatuck tills at Glenn Shores (Figure 1-2) suggests that a significant retreat - of the Lake Michigan Ice Lobe into the Lake Michigan basin must have occurred following the formation of the Tekonsha Moraine. This retreat also suggests that the distinct increase in 10A clay incorporated into the Saugatuck till resulted as the Lake Michigan Lobe ice subsequently readvanced to the Sturgis-Kalamazoo Morainic System. REGIONAL LAKE MICHIGAN LOBE CORRELATIONS In 1915, Leverett (Leverett and Taylor, 1915; plate V) suggested that the Kalamazoo Morainic System in Michigan is time correlative to the outermost ridges of the Valparaiso Morainic System in Illinois. More recent morpho-stratigraphic correlations proposed by Zumberge (1960), however, disagree with Leverett's original interpretat.on. Based on reconnaissance mapping of terrace deposits associated with the Kankakee Flood (Ekblaw and Athy, 1925), Zumburge suggests that the Tekonsha Moraine in Michigan and the West Chicago Moraine (outermost 3h ridge of the Valparaiso System) in Illinois are time equivalent (Figure 1-7). Fullerton (1980) has agreed with the correlation proposed by Zumburge (1960) and has further suggested that till associated with both the Tekonsha and West Chicago moraines represents the base of the Wadsworth till member of the Wedron Formation as defined by Willman and Frye (1970) in Illinois. If correct, the conclusions of Zumberge and Fullerton would not only imply that the Ganges and Saugatuck tills exposed in southwestern Michigan both correlate with the Wadsworth Till Member (Figure 1-7), but would also suggest that the retreat of the Lake Michigan Lobe, which occurred between the formation of the Tekonsha Moraine and the Sturgis-Kalamazoo Morainic System in southwestern Michigan, must have taken place in Illinois during the formation of the Valparaiso Morainic System. However, a significant Late Wisconsinan (late Woodfordian) retreat of the Lake Michigan Lobe, on the order of that recorded at Glenn Shores, has not been recognized within the lower part of the Wadsworth Till. Such a retreat,"though, has been documented between the top of the Yorkville and base of the Wadsworth tills (Willman and Frye, 1970). Therefore, the author proposes it is more reasonable to conclude that the Ganges till in southwestern Michigan correlates with the upper part of the Yorkville Till, and that the overlying Saugatuck till correlates with the Wadsworth Till (Figure 1-7). These correlations, which generally agree with the morainal correlations first proposed by Leverett (Leverett and Taylor, 1915), suggest that the Sturgis Moraine (outermost ridge of the Sturgis-Kalamazoo System) 35 in Michigan is time equivalent to the West Chicago Moraine in Illinois, while the Tekonsha Moraine in Michigan must be approximately time equivalent to the Minooka or Marselles moraines in northeastern Illinois. 36 CHAPTER 2 SAGINAW ICE LOBE DRIFT STRATIGRAPHY INTRODUCTION The establishment of an objective Wisconsinan drift stratigraphy for the Great Lakes Region has been the focus of much recent research, particularly, as was noted in Chapter 1 of this paper, with respect to the Lake Michigan Ice Lobe (Willman and Frye, 1970, 1975; Johnson, 1976; Acomb, 1978; Lineback, et al., l97h, 1979; Glass, 1981; Acomb et al., 1982) and Huron-Erie Ice Lobe (Gooding, 1961, 1963, I971; Dreimanis and Karrow, 1972; Dreimanis and Goldthwait, 1973; Dreimanis, 1977; Eschman, 1978). To date, however, and primarily because of the absence of any multiple till exposure, few studies have attempted to establish a stratigraphy for till deposits associated with the Saginaw Ice Lobe. In fact, the stratigraphy of the Saginaw Lobe is still based on the morphologic relationships of moraines that were first described by Frank Leverett around the turn of the century (Leverett and Taylor, 1915). Even in recent summaries by Dreimanis (1977), Fullerton (1980), and Mickelson et al. (1983) the stratigraphy of the Saginaw Ice Lobe has been described either as a series of 12 to 16 named moraines (Figure 2-1), or has been ignored. The focus of this -uapter is two fold. First, to establish Iithological criteria that can be used to define till sheets associated with the Saginaw Lobe in south-central Michigan, and second, to suggest probable correlation of these till sheets with the 37 FIGURE 2-1. Map showing moraines in southern Michigan (after Leverett and Taylor, 1915). 38 FIGURE 2-2. Map showing Late Wisconsinan moraines of the Lake Michigan and Saginaw lobes and till sample point locations in southwestern Michigan. NOTE: 1 - After Leverett and Taylor (1915) Moraines of the Lake Michigan and Saginaw Lobes in southwestern Michigan: Lake Michigan Lobe Moraines: Lake Border System: LBM -- Inner and Outer Lake Border moraines Valparaiso System: IVM -- Inner Valparaiso Moraine OVM -- Outer Valparaiso Moraine Sturgis-Kalamazoo System: IKM -- Inner Kalamazoo Moraine OKM -- Outer Kalamazoo Moraine STM -- Sturgis Moraine Saginaw Lobe Moraines: TkM -- Tekonsha Moraine KM -- Kalamazoo Moraine LaM -- Lansing Moraine hO nee—0.5:... .mmzqmo: IIIIIII F5.32:3 upamoucmhz..il.|l mmann >FZDOU 2...!on .0. 101 FIGURE 2-3. Cross-section showing lithology and clay mineralogy of the Bedford and Fulton tills sampled from the Eaton County borings. Contacts shown between tills sheets are based on an intervening sand unit between tills, and/or on a significant difference in clay mineralogy. The insert 7A/10A frequency histogram shows the distinct difference in clay mineralogy between the Bedford and Fulton tilIs in BORE-2 AND BORE-6. AZ I-“I NDIlUAB'IB 02? 08C 0.! co“); m._ N.— 0.0 v.0 rprbrh»>.pr.._ 9 ea m4." efi . mm . u." I mo H \Ei o "fl, .4 «; 9° to ... z Llrltlrrrrurtt 0 WW / o my .00.! nu. i . m . N... do 0 mfi mw m-&numw b. wmw N o ; S I // E n . u... m m wmom v N v M ‘C: 3.. O... I 00%. K nWC-o . # g ...I.. id c g a.“ ... Q . m" 44_— zapaau . 2 ...x a a a u 44.. exocoum ” 5:): _m>o_o cco ocom hwy ‘DIIIID'I :: omEocca E .... :23... g .2 Eoeumm fl x..u N.— ..o v.0 flirt-... .. -..... ¢o~\¢b ./ ... ~._ ..o ..o I Ellrlrtlplrtll. Ill 0 \ ... ,// z/ \\o\\ ” \\\ \ii 0 /. . _ C O m mmom DC H9 NC Pb’UIRL—P ¢O~\¢h ... ~.. ..o ..o IFFDPDDFFPDRDIL- ¢O~\¢b . o. a.“ 0.0 0.0 DPhbbbb 0 'IM” '3: . 3E . v“ 1L1 (I: CD 03 3'5 $3 2 .N S H 1|. 9 MS NE SE 1 .OCD IIII NOIIUABTB ’43 established drift stratigraphy of the Huron-Erie and Lake Michigan ice lobes. TILL SHEETS OF THE SAGINAW LOBE In an attempt to establish a till stratigraphy for drift of the Saginaw Lobe in south-central Michigan, six 9 meter deep borings were drilled between the cities of Charlotte and Grand Ledge in Eaton County, Michigan. The locations of these borings are shown on Figure 2-2 (8-1 to 8-6). For each boring, detailed lithologic descriptions were logged; in addition, split-spoon samples of the drift were taken at approximately 1.5 meter intervals. The lithologies and sampling intervals for each bore-hole is shown on the cross-se:tion presented in Figure 2-3. Based on the lithologic logs, at least two till units, separated by sand and gravel, were identified in BORE-2 AND BORE-6 (Figure 2-3). The upper unit, informally named the Bedford till, consists of a red to brown, sandy to sandy clay till. The lower unit is composed chiefly of a grey to tan, dense, clay-rich till, and is informally named the Fulton till. CLAY MINERALOGY: To further define their lithologic characteristics, semi-quantitative X-ray diffraction analyses were performed on the clay-size fraction of samples taken from the two tills. Preparation of the till samples for analysis included pre-treatments that removed soluble salts and free calcium carbonates for each 25 to 50 gram sample (Jackson, 1956; Carroll. 1972). The clay-size (<2 mu) fraction was then separated out by gravity fractionation, deposited on glass hh slides, and air dried. Prior to X-ray analysis, the slides were placed in a desiccator with ethylene glycol to expand the vermiculite to ILA and the smectite to 17A - 18A (Jackson, 1956; Carroll, 1972). The prepared slides were then scanned from 2 to 15 degrees 29- on a General Electric X-ray diffractometer (model XRD-S), rotated 180 degrees, and scanned again. The diffraction patterns were analyzed by comparing the intensities of the IDA and 7A peaks. These peak heights were measured to the nearest .025 cm from a predetermined base line (Grim, 1968: Carroll, 1972). The 7A/IOA peak height ratio was calculated and reported as an average of the two rotations. The results of the clay mineral analyses are presented in the cross-section shown on Figure 2-3. The lithology of the logs is indicated on the Figure Z-and the 7A/lOA ratio at each sampling interval is shown to the right of each log. These data clearly show that where they are separated by sand deposits in BORE-2 and BORE-6, the Bedford (upper) and Fulton (lower) tills have distinct clay mineralogies. The clay mineralogy of till from these two borings has been compiled in Figure 2-3 as 7A/10A frequency histograms. The horizontal axis of each histogram represents the calculated 7A/10A ratio, while the vertical axis represents the number of samples with a particular ratio value. It shows that in BORE-2 and BORE-6, the Bedford till has ratio values less than 0.90 while the Fulton till has ratio values greater than 0.90. On the basis of the distinct difference at the 0.90 ratio value of till sampled from these two borings, contacts between the till units in adjacent borings -were drawn. The samples from the 10 meter (30 foot) interval in BORE-2, “5 FIGURE Z-h. Frequency histograms showing clay mineralogy of the Bedford and Fulton tills from Eaton County bore-holes (above dashed line) compared with histograms of till sampled from Saginaw Lobe moraines in south-central Michigan (below dotted line). A6 .ouxu .ouzu r. oom— . ww.IIL. oo oww_ pllcow— . oowo . ow.o . 0.. 2.6 $3.5. C G E C G . L 8..» :5" .~ oumLL zLuzamo wcoLzoud m N. a :9: r t . dIEwaE q: - L . . t L 2 » zcuu .~ mzLamo: czmzoxwe u mmwm .Bufim I I . . . . -C a :2: w tdq u. mzLamo: oo~¢z¢u¢x H 8.. £8... ( i . I CE I r . r L L . . L av.” zcmu TN moszom some uuLe Lmezoue .omzazzp. a an... imam . _ I . . LEE: doc. : . i _ I ,M . . moszom some uuLL LMJQQLZL ZOLJDL a 8.5 .593 I t . . . . r I LC CU Pd E: . L 3% zcuu .N moszom some JJLL Lawmmsu omomomm m 1.7 BORE-3, and BORE-h show ratio values that are greater than l.h5. Their stratigraphic position within the borings, as well as the distinct difference in their clay mineralogies, suggests that these samples may represent the top of an even older till sheet. The clay mineral data from all samples associated with the Bedford and Fulton tills have been compiled as 7A/10A frequency histograms (Figure Z-h). These histograms show the probable range of clay mineralogy of the two till units. The mean ratio value for the Bedford till is 0.69, while the mean ratio value for the Fulton till is 1.06. Application of the T-test to the clay data indicates that the Bedford and Fulton tills are different to the 99% confidence level. END MORAINES OF THE SAGINAW LOBE Two ice edge positions of the Saginaw Lobe, the Tekonsha and Kalamazoo moraines, occur approximately ho to 60 kilometers southwest of the bore-holes. These moraines were first identified by Leverett (Leverett and Taylor, l9l5) and are shown on Figure 2-2. Also shown is the location of the Lansing Moraine, which occurs a few kilometers south of boring 8-6. A drumlinized till plain (informally named the Leonidas drumlin field) occurs southwest of the Tekonsha Moraine. The southwest-northeast orientation of these drumlins indicates that they were formed by Saginaw Lobe ice (Leverett and Taylor, 1915; Monaghan, et. al., 1983). CLAY MINERALOGY: In an attempt to correlate the Bedford and Fulton .tills with till occurring south of the bore-holes, clay mineralogical k8 analyses (identical to those described for the till sampled 'from the bore-holes) were performed on 28 till samples taken from exposures within the Kalamazoo and Tekonsha moraines as well as the Leonidas drumlin field. Samples were generally taken 2 to 3 meters deep from fresh, unweathered exposures. The locations and associated 7A/IOA values of these sample points are shown on Figure 2-2. The results of the analyses are presented on Figure Z-h in the form of 7A/10A frequency histograms. These data show that the clay mineralogy of the till associated with both the Tekonsha Moraine and the Leonidas drumlin field is significantly different from the till associated with the Kalamazoo Moraine. The mean ratio value for the Leonidas drumlin field is 1.28, the mean ratio value for the Tekonsha Moraine is 1.13 and the mean ratio value for the Kalamazoo Moraine is 0.78. Comparison of the frequency histograms for the clay data from moraine samples with those associated with the tiIls defined in the bore-holes (Figure Z-h) suggests that the till from the Kalamazoo Moraine is most similar to the Bedford till, while till from both the Tekonsha Moraine and the Leonidas drumlin field is most similar to the Fulton till. Application of the T-test shows that these associations can be supported to at least the 99% confidence level. Based on the statistical similarity of their clay mineralogy, the author proposes that the till associated with the Kalamazoo Moraine correlates with the Bedford till, while till associated with the Tekonsha Moraine and the drumlin field correlates with the Fulton till. ‘19 FIGURE 2-5. Generalized northeast to southwest stratigraphic cross-section through Saginaw Lobe deposits in south-central Michigan. 50 ...—E _|||Inll._ o- 23052... .I|i|||_ n— o 51 FIGURE 2-6. Correlation chart showing time stratigraphic relationships of tills of Lake Michigan, Saginaw, and Erie lobes in the Great Lakes region. Notes: 1 - After Frye and Willman (1973) 2 - After Wayne (1963, 1965) and by Dreimanis and Goldthwait (1973) 3 - After Dreimanis and Goldthwait (1973) 52 O. . § 1‘52 “l- <<~ I z m 82 3 in 5 [fl 3 t :85 '1. ‘° 1, 5 2 mm m . I; ~J a 8 g o 015 ... O U — 2 "E I; I - I: I J‘ l . o l I o- 2?— $ é «- ) 13' I a\ 7 ICE MARGIN _. FIGURE 3-2. Map showing reconstructed position of the Lake Michigan and Saginaw lobes at about 16,500 BP during the construction of the Tekonsha moraines and the Climax-Scotts outwash plain. 83 traced no further than just west of the village of Scotts. The meltwater that produced the Climax-Scotts outwash plain probably drained south, along the distal margin of the Lake Michigan Lobe, and then into the glacial St. Joseph River (Leverett and Taylor, 1915). From there, drainage may have continued south and west into Illinois down the Kankakee River Valley (Zumberge, 1960). At least some of the drainage associated with the outwash plain was also channeled south down narrow valleys in the Leonidas drumlin field. When traced into St. Joseph County, these valleys terminate at the upper terraces of the St. Joseph River. The surface gradient for these channels is the same as that of the Climax-Scotts outwash plain. This suggests that the channels and the plain were formed at approximately the same time and were likely graded to the same drainage threshold. Deposition of outwash sediments on the Climax-Scotts outwash plain ended when the ice from both lobes retreated from the Tekonsha Moraine. General retreat of the Saginaw and Lake Michigan lobes Subsequent to the formation of the Tekonsha Moraine and the Climax-Scotts outwash plain, both the Saginaw and Lake Michigan lobes began to progressive retreat across Kalamazoo County. This retreat, which probably began about 16,500 to 17,000 years ago (Chapters 1 and 2 of this paper), may have been characterized by minor still-stands of the margin of both lobes. Leverett and Taylor (1915, 192A) has cited as evidence for one such still-stand a 15 sq km (6 sq 'mi) undifferentiatated, high gravel plain located north of the Kalamazoo 8A River Valley, just east of the town of Augusta. He associated it with the Saginaw Lobe and suggests that it was formed by a “later moraine” (probably the Battle Creek Moraine) after the ice margin had retreated from the Tekonsha Moraine. Though too small to define with certainty as an outwash sequence, this remnant has an apparent southwestward gradient of about 6.5 m/km (12 ft/mi) and may represent an outwash deposited during the northward retreat of the Saginaw Lobe. The magnitude of the retreat of the Lake Michigan and Saginaw lobes following the formation of the Tekonsha Moraine is not clear from drift deposits within Kalamazoo County alone. Recent stratigraphic research, however, does suggest that it was significant. For example, Gephart, et.al. (1983) and Monaghan (Chapters 1 and 2 of this paper) have traced the till within the Tekonsha Moraine of the Lake Michigan Lobe west to an exposure of buried till along the Lake Michigan shoreline in Allegan County. They further suggest that following the formation of the Tekonsha Moraine the margin of the Lake Michigan Lobe retreated. at least as far as the Lake Michigan shoreline, a distance of 60 kilometers (A0 mi). In Chapters 1 and 2 of this paper, the author also has suggested that during the same time interval the Saginaw Lobe retreated as far north as the city of Lansing, a distance of 60 kilometers (A0 mi). He bases his interpretation on correlation of till from the Leonidas Drumlin field and till occurring in the subsurface in northeastern Eaton County. Stratigraphic evidence for a retreat of the Lake Michigan Lobe has also been observed at several exposures in Kalamazoo County. These show outwash both overlying and underlying till. The author and 85 others (Chapters 1 and 2 of this paper and Gephart, et al., 1983) suggest that the till from these exposures is stratigraphicly equivalent to till of the Tekonsha Moraine. This relationship is particularly apparent within exposures along the western wall of the Kalamazoo River Valley north of the city of Kalamazoo. In one such exposure, located just north of and behind the outermost ridge of the Kalamazoo Moraine (Cooper section, sw of sw, sect. 15, T15 Rllw), dense basal till is overlain by south and southwestward dipping, cross-bedded outwash that fines upward. The till is underlain by medium to fine sand and silt. A pebble fabric of the till from this exposure also suggests that it was deposited by the Lake Michigan Lobe. In Chapters 1 and 2 of this paper, the author has proposed that, following the retreat of the Saginaw Lobe north of the city of Lansing, it readvanced' to the Kalamazoo Moraine in Barry County (Figure 3-3). This readvance is based on shallow boring in Eaton County which’show that the till comprising the Kalamazoo Moraine stratigraphically overlies till associated with the Tekonsha Moraine. A similar readvance of the Lake Michigan Lobe has been proposed by the author and others (Chapters 1 and 2 of this paper and Gephart, et al., 1383). They suggest that, following the retreat of the Lake Michigan Lobe west of the Lake Michigan shoreline, it readvanced to the position of the Sturgis Moraine (in Kalamazoo COLHLY)*. This is based on correlation of till within the Sturgis Moraine to that associated with the Kalamazoo, Valaraiso, and Lake Border morainic systems. Exposures between Kalamazoo County and the Lake Michigan 86 shoreline also show that this till stratigraphicly overlies till correlated with the Tekonsha Moraine. Formation of the Galesburg-Vicksburg outwash plain The formation of the Galesburg-Vicksburg outwash plain began during the early phases of the westward retreat of the Lake Michigan Lobe from the Sturgis Moraine and culminated when the ice margin of the Lake Michigan and Saginaw lobes occupied a position defined by the Kalamazoo moraines (Figure 3-3). The author suspects that the retreat of Lake Michigan Lobe ice across Kalamazoo County from the position of the Sturgis Moraine probably occurred by progressive back-wasting of the ice margin, and that the Galesburg-Vicksburg outwash plain was constructed in stages. The “outer“ Kalamazoo Moraine most likely represents a final forward “push” of the ice during this construction. During the initial construction of the Galesburg-Vicksburg outwash plain, meltwater. was channeled between the retreating ice front and the tOpographically higher Leonidas drumlin field and the Climax-Scotts outwash plain. Erosion associated with this meltwater formed the prominent scarp that defines the western boundary of the * In 1915, Leverett (Leverett and Taylor, 1915) originally proposed that the Sturgis Moraine of the Lake Michigan Lobe in Kalamazoo County and the Sturgis Moraine of the Saginaw Lobe in St. Joseph County were formed at the same time. However, based on the above correlations, the author feels that this moraine more likely represents the outermost ridge of the Kalamazoo Morainic System. Although the name (Sturgis) given by Leverett to the morainal ridge has been kept, the author and others have elsewhere (Gephart, et al., 1983) informally refered to Leverett's Kalamazoo Morainic System as - the Sturgis-Kalamazoo Morainic System. 37 Lahd } o 9.“ / “.‘fl| Loo. _./ . L". \, a' 0 smw“ pr” / 35 o g w3- ”1 o l" m (<1-b (no 5 o: N AD _, >' .33 E '- 3 1” ¢<> ¢ . 3 z “”3 ¢ . Z I g .5 g 01... ; ° 0 - F s s . 8 l I I I I 04 z<———— J:- 2* i \ \ __ ’— \. ‘ \R r \. hx \\ "(K \ .(‘d-\\ .../I. 4?"? 1% -%D Z 5 K /V7£W g1; ”917/14 J *5 .7 FIGURE 3'3. Map showing reconstructed position of the Lake Michigan and Saginaw lobes at about 15,000 BP during the construction of the Kalamazoo moraines and the Galesburg-Vicksburg outwash plain. Climax-Scotts outwash plain. Meltwater drainage probably also eroded or partly buried the northern extension of the Sturgis Moraine. In addition, remnant ice masses became detached from the ice margin as it retreated towards the west. These masses were subsequently buried by outwash and are represented today by depressions and kettle lakes, such as Gull, Indian, and Austin lakes. While the ice retreated from the Kalamazoo Moraine to the Valparaiso Moraine in Allegan County (Gephart and Larson, 1983), southeastward drainage along the Galesburg-Vicksburg outwash plain ended. Meltwater was then channeled southward along an irregular back-wasting ice-front. That this margin was irregular is indicated by the extensive kame field which makes-up the undiffereniated sand and gravel deposits between the Galesburg-Vicksburg outwash plain and the Gun Plain outwash. Development of the Kalamazoo River Valley As the Lake Michigan Lobe retreated from the Kalamazoo Moraine, the Saginaw Lobe also retreated progressively northward. During the. initial phase of this retreat, meltwater was channeled southward along a chain of kettle depressions in the Galesburg-Vicksburg outwash plain and deposited the Augusta Creek outwash. The upper terraces of this valley train system have been traced from north of the Kalamazoo Moraine, in Barry County, to a 290 meter (900 ft) threshold along the Kalamazoo River Valley at the town of Augusta in Kalamazoo County. Originally, this drainage system may have extended further south, along the western margin of the Tekonsha Moraine and Climax-Scotts 89 outwash plain and then down the valley of the Portage River. The lack of terraces along Portage River, however, suggests that this drainage must have been short lived. Leverett (Leverett and Taylor, 1915) has suggested that a rather large remnant ice mass must have persisted in the Kalamazoo River Valley during the construction of the Galesburg-Vicksburg outwash plain. As evidence for such a mass, he cites a broad 90 sq km (35 sq mi) depression located between the cities of Galesburg and Kalamazoo. The author proposes that as the Lake Michigan Lobe retreated from the Sturgis Moraine this ice mass became detached from the retreating ice margin. The mass was partly buried by sediments associated with the Galesburg-Vicksburg outwash and by sediments deposited during the last push of the Lake Michigan Lobe as it formed the outer ridge of the Kalamazoo Moraine. .Subsequently, as the Cake Michigan Lobe. retreated westward out of the county, the mass began to melt and left a large depression in the surface of the Galesburg-Vicksburg outwash plain. Today, the occurence of lake sediments within this depression suggests that y, perhaps associated with the Augusta Creek outwash, collected in this basin and formed a shallow lake. Initially the lake probably drained to the south. However, continued sagging in the northern part of the depression eventually allowed the lake to drain northward through a breach in the Kalamazoo Moraine. Water flowing through rbls breach, which is located just north of the city of Kalamazoo, then merge with the drainage that formed the Gun Plain outwash. When the Kalamazoo Moraine was breached, it allowed unrestricted westward flow of meltwater down the 90 valley of the Kalamazoo. River. As this precussor to the Glacial Kalamazoo River downcut rapidly through the lake plain, it captured the drainage flowing down Augusta Creek and Portage River. That the downcutting was rapid is indicated by the lack of terrace levels from 25A meter (860 ft) to 235 meters (830 ft) elevation along the Kalamazoo River between Kalamazoo and the Allegan County border. Complete westward flow of meltwater along the the valley of the Kalamazoo River initiated the Glacial Kalamazoo-Gun Plain outwash sequence. The Glacial Kalamazoo-Gun Plain system served as the major meltwater drainageway for the eastern part of the Saginaw Lobe until the ice front retreated north of the city of Lansing (Leverett and Taylor, 1915). This is suggested by recent mapping in Barry County which shows that the Gun Plain continues north to the Thornapple River. Meltwater that formed the upper terraces along the Thornapple River between the towns of Hastings in Barry County and Vermontville in Eaton County probably flowed southward down Gun Plain. Terraces along the glacial Kalamazoo River, on the other hand, can be traced to terraces in the valley of the Battle Creek in Calhoun County. The lower terraces along the Battle Creek have also been traced to- the Charlotte Moraine near the town of Charlotte in Eaton County (Monaghan and Larson, 1980). After both the Lake Michigan and Saginaw lobes hac completely retreated from Kalamazoo County, remnant ice masses buried by outwash began to melt and formed kettle depressions. These kettles sometimes coalesced into long, deep elongate depressions, like that between Gull 91 Lake and the city of Kalamazoo. When the Glacial Kalamazoo-Gun Plain outwash eroded through the Galesburg-Vicksburg outwash plain, it probably lowered the water table in the surrounding area. In response, sapping occured along the margin of the Kalamazoo River Valley and produced deep furrows. Many of these furrows are quite long, such as along Ravine Road between the Kalamazoo Moraine and the City of Kalamazoo. Most of the major landforms occurring in Kalamazoo County were developed during the retreat of Late Wisconsinan ice. Those materials deposited since deglaciation generally represent minor modification of glacial landscape. They include floodplain deposits along the Kalamazoo River and muck deposits infilling kettle depressions. Significant deposits of wind blown sand also occur in some areas. These deposits were probably formed soon after deglaciation and before significant amounts of vegetation grew on the land surface. 92 SUMMARY AND CONCLUSIONS Three till sheets associated with advances of the Lake Michigan Ice Lobe are exposed in a 20 to 25 meter high section (Glenn Shores section) along the Lake Michigan shoreline in Allegan County, Michigan. The lowermost till (Glenn Shores till) is directly overlain by gravel and lacustrine silt and sand containing organic material dated at 37,150 to >A8,000 years B.P. The middle (Ganges till) and uppermost (Saugatuck till) tills are also each overlain by lacustrine deposits. All three tills in the section are easily differentiated by the 7A/10A X-ray diffraction peak height ratio of their clay-size fraction. In an attempt to define litho-stratigraphy for tills associated with the Saginaw and Lake Michigan lobes, X-ray diffraction analyses were performed on over 150 samples collected from moraines, gravel pits, and borings in southwestern Michigan. The results of these analyses show that Lake Michigan Lobe till associated with the Lake Border, Valparaiso, and Sturgis-Kalamazoo morainic systems can be correlated with the Saugatuck (upper) till exposed at the Glenn Shores section, while the Ganges till can be traced in the subsurface from crops out within the Tekonsha Moraine to the Glenn Shores section. The occurrence of lake deposits between the Ganges and Saugatuck tills at Glenn Shores implies a significant retreat of the Lake Michigan 93 Lobe (greater than 50 kilometers) between the formation of the Tekonsha Moraine and the Sturgis-Kalamazoo Morainic System. Based on clay mineral analysis of Saginaw Lobe till sampled from six 10 meter borings in Eaton County, Michigan, at least two till units can be defined. The upper till (Bedford till) from the borings can be traced along the ground surface to as far south as the Kalamazoo Moraine, while the lower lower till (Fulton till) can be correlated with till sampled from and just south of the Tekonsha Moraine. The clay mineralogies of these till units suggest a significant retreat of the Saginaw Lobe between the formation of the Tekonsha and Kalamazoo moraines. Morpho-stratigraphic correlations between moraines of the Lake Michigan, Saginaw, and Erie lobes suggest that the West Chicago and Sturgis-Kalamazoo moraines of the Lake Michigan Lobe, the Kalamazoo Moraine of the Saginaw Lobe, and Powell-Union City Moraine of the Erie Lobe are all the same age and were formed as the margin of all three lobes readvanced following the Erie Interstade. These correlations imply that the interval between the deposition of the Ganges and Saugatuck tills of the Lake Michigan Lobe as well as the interval between deposition of the Fulton and Bedford tills of the Saginaw Lobe are both time equivalent to the Erie Interstade of the Erie Lobe. If correct, these correlations then imply that during the Late Wisconsinan major retreats and re advances of the ice front of the Lake Michigan, Saginaw, and Erie lobes were probably approximately time synchronous. 9A The clay mineralogy of the stratigraphic units defined for tills in southwestern Michigan suggests that a relative increase in 10A clay occurs in progressively younger tills sheets. The reason for this progressive change in clay mineralogy has not yet been defined, however, it is apparently related to slgnificant retreats of ice. The fact that this progressive change in clay mineralogy occurs within drift of both the Lake Michigan and Saginaw lobes suggests that the processes responsible for this clay fractionation may relate to a more general erosional/deposition system occurring throughout the Great Lakes Region. Although the processes of clay fractionation in tills in southwestern Michigan are not yet known with certainty, the author suggests that one of the following three models may explain them. 1) The relative increase in 7A clays in consecutively older till units resulted from greater diagenesis of the older tills. 2) The fraction of 7A and 10A clays is related to differential settling velocities of specific clay minerals within the proglacial lake which formed during ice retreat. As the ice front subsequently readvanced, these fractionated lake clays would then be incorporated within the ice and re-deposited as till with increased amounts of 10A clay. 3) The stratigraphy of the Glenn Shores section suggests that the three exposed tills span the entire Wisconsinan. If this is correct, the oldest till unit (Glenn Shores till) should include relatively high amounts of authigenic kaolinite and chlorite 95 weathered from the Canadian Shield during Sangamon time, as well as illite derived primarily from erosion of Paleozoic shales as the ice advanced. Tills associated with subsequent re-advances of the ice would include mostly illite derived from Paleozoic shales and only small amounts of kaolinite eroded from older tills and from small remnants of the Sangamon surface on the shield. Hence, successive fluctuations of the ice margin, and subsequent dilution of 7A clays, should result in a progressive, step-wise increase of the IDA clay in consecutively younger till units. 96 APPENDIX A APPENDIX A ANALYTICAL TECHNIQUES Samples used in this study have all undergone some pre-treatments to simplify and standardize the analysis. These pre-treatments included removal of soluble salts and carbonates, and in some cases, free iron oxides. While the use of pre-treatments does involve the risk of altering or destroying some important mineral components, the treatments selected are designed to have minimal effects on important clay mineral constituents. Removal of soluble salts and carbonates. Soluble salts and carbonates are removed from the sample because they can act as cementing agents and promote flocculation. They also may interfere with X-ray analysis. All samples used in this study were_ pre-treated by adding 50 mls of 1N sodium acetate solution (buffer to pH 5 with acetic acid) to each 5 grams of sample. This causes the following reaction: CaC03 + 2H+NaAc r) ta2+ + H2C03 --- :02? + H20 The sample solution was heated to 50 to 60 degrees centigrade for a 30 minute period, centrifuged, and washed 3 times with distilled water. Removal of free iron oxide. Iron oxides should be removed from samples both because they promote flocculation and can act as dilutant which reduces‘the X-ray diffraction intensity. Removal of free iron oxides was accomplished 97 as follows: 1) To each sample A0 mls of 0.3N Na-citrate and 5 mls of 1N NaHCO; were added. 2) The suspension was heated to 75 to 80 degrees centigrade and 1 gram of NaZSZO‘ added and digested for 15 minutes. 3) The sample was centrifuged and washed with distilled water. These steps remove the free iron oxides by the following reactions: Na citrate Fe203 -—9 Fe3 + chelates Addition of bicarbonate maintains neutrality and iron is reduced by: OH‘ + 520;: -+ zso3= + 2H20 + 2e0 2Fe3+ + 2e” —-> 2Fe2+ Not all samples were treated to remove free iron oxides. Experiments performed on samples from the Bloomingdale section (Appendix C) indicated that the treatment had little regular effect of the calculated 7A/10A ratios. Semi-quantitative X-ray analysis. The clay-size fraction (<2 mu) of each treated sample was separated out by gravity settling, deposited on a glass slide, and air dried, which produced a slide of oriented clay mineral grains. These slides were then placed in a desiccator with ethylene glycol for at least A8 hours to expand the vermiculite to IAA and the smectite to l7A-18A. The prepared clay slides were scanned for 3 degree to 15 degree 2f} on a General Electric X-ray diffractometer using 1 degree medium 98 range beam slit, a medium range soller and a 0.2 detector slit Samples were run at the 1,000 to 2,000 cps range (depending on the intensity of the peaks). Each sample was rotated 180 degrees and scanned again. Peak heights were measured to the nearest 0.01 inch on a Numonics digitizer from an established baseline and the 7A/10A ratios reported as the average of the 2 rotations. Qualitative X-ray analysis. Methods used to determine which clay minerals are present in Lake Michigan and Saginaw Lobe tills in southwestern Michigan are similar to those described by Jackson (1956), Carroll (1972), and Gephart (1982). Four (two each from tills associated with the Saginaw and Lake Michigan ice lobes) pre-treated, fractionated till samples were deposited on a porous porcelain plate by vacuum suction to produce an oriented sample slide ready for X-ray analysis. Each of these samples was then X-rayed once after each of the following treatments: 1) Mg saturated and glycerol solvated; 2) K saturated and air dried; 3) K saturated and heated to 300 degrees C for a two hour period: and A) K saturated and heated to 550 degree C for a two hour period. lllite is the only common clay mineral that produces a 10A peak which is not affected by either the chemical or heat treatments (Carroll, 1972: Grim, I968). Kaolinite will produce a 7A X-ray peak: however, after heating the sample to 550 degree C the structure of kaolinite is destroyed and the 7A peak associated with kaolinite will disappear (Carroll, 1972; Grim, 1968). Vermiculite and chlorite both produce 1AA peaks after Mg saturation and glycerol solvation, and will 99 produce a smaller second order (002) 7A peak (Carroll, 1972; Grim, 1968). Smectite will produce a 17A to 18A peak after Mg saturation and glycerol solvation. Both vermiculite and smectite will collapse to 10A after heating to 300 degrees C (Carroll, 1972; Grim, 1968). The results of these analyses show that the dominant clay minerals present in Saginaw Lobe tills are illite, kaolinite, and to a lesser extent, vermiculite. Chlorite is generally quite rare in Saginaw Lobe drift. The dominant clay minerals present in Lake Michigan Lobe tills, on the other hand, are illite, kaolinite, and to a lesser extent, chlorite. Vermiculite is rarely present in any significant amounts within Lake Michigan Lobe drift. IOO APPENDIX B APPENDIX B STATISTICS Population Statistics_and Data The following listing is a compilation of the population statistics for each of the till units and is grouped by moraine, bore holes, or reference section. LAKE MICHIGAN LOBE MORAINES AND REFERENCE SECTIONS Kalamazoo (KzM) - Sturgis (StM) Moraine Sample size ------------------------- 1A Population mean --------------------- 0.61 Population standard deviation ------- 0.25 DATA Sample 7A/10A Sample 7A/10A KzM-l 0.63 KzM-Z 0.60 KzM-3 0.A5 KzM-A 0.A9 KzM-S 0.51 KzM-6 0.50 KzM-7 0.60 KzM-8 0.60 KzM-9 0.55 KzM-lO 0.83 KzM-ll 1.08 StM-l 0.57 StM-Z 0.7A StM-3 0 A3 Tekonsha Moraine Sample size ------------------------- 6 Population mean --------------------- 0.83 Population standard deviation ------- 0.Al DATA Sample 7A/10A Sample 7A/10A TkM-1 0.67 TkM-2 0.69 TkM-3 0.88 TkM-A 0.87 TkM-5 l.ll TkM-6 0.88 101 Bloomingdale Upper till Sample size ----------* -------------- 10 Population mean --------------------- 0.52 Population standard deviation ------- 0.21 DATA Sample 7A/10A Sample 7A/10A Bdut-l O 58 Bdut-Z 0 38 Bdut-3 0.75 Bdut-A 0 5A Bdut-S 0 51 Bdut-6 0 A3 Bdut-7 0.50 Bdut-8 0 A9 Bdut-9 0 6A Bdut-lO 0 A2 Bloomingdale Middle Till Sample size ---------------------- r-- 6 Population mean --------------------- 0.A6 Population standard deviation ------- 0.21 DATA Sample 7A/10A Sample 7A/10A Bdmt-l 0 A3 Bdmt-Z n 0 A3 Bdmt-3 0 A9‘ Bdmt-A 0 52 Bdmt-5 0.AA Bdmt-6 0 A5 Bloomingdale Lower Till Sample size ------------------------- 18 Population mean --------------------- 0.83 Population standard deviation ------- 0.22 DATA Sample 7A/10A Sample 7A/10A Bdlt-l 0.90 Bdlt-Z 0.88 Bdlt-3 0.8A BdIt-A 0.83 Bolt-5 0.92 Bdlt-6 0.77 Bdlt-7 0.73 Bdlt-8 0.8A Edit-9 0.78 Bdlt-lO 0.75 BdIt-11 0.83 Bdlt-lZ 0.8A Bdlt-13 0.87 BdIt-IA 0.7A BdIt-IS 0.83 Bdlt-lé 0.70 Bdlt-17 1.11 BdIt-18 0.78 102 Buried Tills in Kalamazoo County Sample size ------------------------- 11 Population mean --------------------- 0.97 Population standard deviation ------- 0.35 Comstock Section DATA Sample 7A/10A Sample 7A/10A Cooper Section DATA Sample 7A/10A Sample 7A/10A Cop-l 1 31 Cop-2 l 10 Cop-3 1.03 Cop-A l 07 Ravine Road Section DATA Sample 7A/10A Sample 7A/10A er-l 0.81 er-2 0.8A er-3 0.83 er-A 0.95 103 Kalamazoo Moraine SAGINAW LOBE MORAINES AND BORINGS (and surface samples north of moraine) Sample size ------------------------- 17 Population mean --------------------- 0.75 Population standard deviation ------- 0.31 DATA Sample 7A/10A Sample 7A/10A KzS-l 1.3A KzS-Z 0.78 KzS-3 0.35 KzS-A 0.50 KzS-5 0.76 KzS-6 0.82 KzS-7 0.83 KzS-8 0.75 KzS-9 0.7A KzS-IO 0.5A KzS-ll 0.38 KzS-12 0.96 KzS-l3 0.77 KzS-IA 0.79 KzS-15 0.3A KzS-16 0.90 KzS-17 0.69 Leonidas Drumlin Field Sample size ------------------------- 5 Population mean --------------------- 1.28 Population standard deviation ------- 0.70 DATA Sample 7A/10A Sample 7A/10A Lde-l 1.15 Lde-2 1 0 Lde-3 0.99 Lde-A 1 6 Lde-5 1.61 Tekonsha Moraine Sample size ------------------------- 9 Population mean --------------------- 1.12 Population standard deviation ------- 0.A2 DATA Sample 7A/10A Sample 7A/10A TkS-l 1.13 TkS-Z 1.A2 TkS-3 0.95 TkS-A 1 23 TkS-5 1.07 TkS-6 1.10 TkS-7 0.98 TkS-8 1.05 TkS-9 1.15 10A Upper Till From Borings Sample size ------------------------- 10 Population mean --------------------- 0.69 Population standard deviation ------- 0.26 Middle Till From Borings Sample size ------------------------- 16 Population mean --------------------- 1.06 Population standard deviation ------- 0.29 Lower Till from Borings Sample size ------------------------- 3 Population mean --------------------- 1.A9 Population standard deviation ------- 1.05 Clay Mineral Data from Borings Boringfi Depth Till Unit** Sample 7A 8-1 5' ut 8-1-5 0.5A 8-1 10' ut 8-1-10 0.82 8-1 15' mt 8-1-15 1.00 B-l 20' mt 8-1-20 1.09 8-2 5' ut 8-2-5 0.69 8-2 10' ut 8-2-10 0.58 8-2 15' mt 8-2-15 0.97 8-2 20’ mt 8-2-20 1.11 8-2 30' 1t 8-2-30 1.52 8-3 5' ut 8-3-5 0.60 8-3 10' mt 8-3-10 1.19 8-3 15' mt 8-3-15 1.00 8-3 20' mt 8-3-20 0.85 8-3 25' mt 8-3-25 0.93 8-3 30' 1t 8-3-30 1.A7 B-A 5' mt 8-A-5 .00 B-A 30' It 8-A-3O .A8 8-5 5' ut 8-5-5 0.78 8-5 10' ut 8-5-10 0.8A 8-5 15' mt 8-5-15 1.15 8-5 20' mt 8-5-20 1.09 8-5 25' mt 8-5-25 0.98 8-5 30' mt 8-5-30 1.1A 105 Clay Mineral Data from Borings (cont.) 8oring* Depth Till Unit** Sample 7A/10A 8-6 5' ut 8-6-5 0.72 8-6 10' ut 8-6-10 0.71 8-6 l5' ut 8-6-15 0.8A 8-6 20' mt 8-6-20 1.10 8-6 25' mt 8-6-25 1.08 8-6 30' mt 8-6-30 1.12 * Boring number refers to location on figure 6 *8 Till Unit refers to named units on figure 9 ut -- upper Till (Bedford till) mt -- middle till (Fulton till) It -- lower till (unnamed till) 106 T-test Statistics for Selected Populations. follows: The following is a tabulation of T-test statistics for selected till populations. Till populations are abbreviated as Glenn Shores upper (Saugatuck) till 8 -- GS-UT Glenn Shores middle (Ganges) till * -- GS-MT Bloomingdale upper till -- BD-UT Bloomingdale middle till -- BD-MT Bloomingdale lower till -- BD-LT Kalamazoo-Sturgus Moraine (Lake Michigan Lobe) -- KzM-M Tekonsha Moraine (Lake Michigan Lobe) -- TkM-M Tekonsha Moraine (Saginaw Lobe) -- TkM-S Kalamazoo Moraine (Saginaw Lobe) -- KzM-S Upper till from Eaton County borings -- Br-UT Middle till from Eaton County borings -- Br-MT * Data used in T-test for these samples from Gephart (1982) LAKE MICHIGAN LOBE TILLS GS-UT * KzS-M Number ------- 25 -- 1A Mean --------- 0.58 -- 0.61 Variance ------ 0.02 -- 0.06 St. Dev. ------ 0.13 -- 0.25 T-statistic ----------- 0.881 Degrees of Freedom ---- 37 GS-MT * Tk-M Number ------- 39 -- 6 Mean --------- 0.85 -- 0.83 Variance ---- . 0.03 -- 0.17 St. Dev. ------ 0.18 -- 0.Al T-statistic ----------- 0.203 Degrees of Freedom ---- A3 107 Number ------- 25 -- 10 Mean --------- 0.58 -- 0.52 Variance ------ 0.02 -- 0.0A St. Dev. ------ 0.13 -- 0.21 T-statistic ----------- 0.986 Degrees of Freedom ---- 33 GS-UT * BD-MT Number ------- 25 -- 6 Mean --------- 0.58 -- 0.A6 Variance ------ 0.02 -- 0.0A St. Dev. ------ 0.13 -- 0.21 T-statistic ----------- 1.811 Degrees of Freedom ---- 29 GS-MT * BD‘LT Number ------- 39 -- 18 Mean --------- 0.85 -- 0.83 Variance ------ 0.03 -- 0.05 St. Dev. ------ 0.18 -- 0.22 T-statistic ----------- 0.A50 Degrees of Freedom ---- 55 SAGINAW LOBE TILLS KzM-S TkM-S Number ------- 17 -- 9 Mean --------- 0.75 -- 1.12 Variance ------ 0.10 -- 0.18 St. Dev. ------ 0.31 -- 0.A2 T-statistic ----------- 2.52 Degrees of Freedom ---- 2A Br-UT Br-MT Number ------- 10 -- 16 Mean --------- 0.69 -- 1.06 Variance ------ 0.07 -- 0.08 St. Dev. ------ 0.26 -- 0.29 T-statistic ----------- -3.260 Degrees of Freedom ---- 2A KzM-S Br-UT Number ------- 17 -- 16 Mean --------- 0.75 -- 1.06 Variance ------ 0.10 -- 0.08 St. Dev. ------ 0.31 -- 0.29 T-statistic ----------- 0.589 Degrees of Freedom ---- 31 TkM-S Br-MT Number ------- 9 -- 16 Mean --------- 1.12 -- 1.06 Variance ------ 0.18 -- 0.08 St. Dev. ------ 0.A2 " 0.29 T-statistic ----------- 1.020 Degrees of Freedom ---- 23 109 APPENDIX C APPENDIX C SECTION DESCRIPTIONS The following Appendix discribes three of the sections refered to in this paper. Two of these sections were discovered by the author and the third, Ravine Road, was described by Straw (1978). Comstock Section The Comstock section is exposed in a gravel pit in the north wall of the Kalamazoo River valley located just east of the town of Comstock (SW of SW of SW, sect. 16, t25, rIOw). It contains about 1 meter (2 to 3 feet) of a brown, coarse, sandy to cobbley till. The till'unit pinches out to the north and east which suggests that it was probably deposited as a flow till from the retreating margin of the Lake Michigan Lobe. This is also suggested by deltaic deposits composed of course sand and pebbles which grade into the till where it pinches out. The till and deltaic sediments are underlain by fine to very fine horizontally bedded and laminated sand and silt. The till in overlain by cross-bedded course sand and gravel and the contact between these two units is sharp. Ravine Road Section The Ravine Road section occurs in an abandoned gravel pit located just north of the City of Kalamazoo (sw,sw,sw, sec. 9, t 2 s, r 12 w). The base of the exposure is represented by a brown, sandy, 110 loose-textured till unit. The total thickness of this till is not known and only the upper 2 to 3 meters (6 to 8 feet) is exposed. Vertical samples derived from this till do not show any trends. The till is overlain by medium to fine sand with some pebbles. Bedding structures within this sand unit indicate that it was deposited within a braided stream environment and its stratigraphic position, therefore, suggests that it was probably formed during the final stages of the Galesburg-Vicksburg outwash plain (Chapter 3). Cooper Section The Cooper section is exposed in a gravel pit within the west wall of the Kalamazoo River Valley (SW of SW of SW, sect. 15, tls, rllw) located about 8 miles north of the city of Kalamazoo. A l to 3 meter (3 to 10 feet) thick brown, dense clay-rich till is directly overlain by 6 to 10 meters (20 to 30 feet) of cross-bedded, coarse sand and gravel, which includes some elongated till balls. Samples taken from a vertical transect within the till .do not indicate any significant trends in clay mineralogy. The till is underlain by 3 to 5 meters (9 to 15 feet) of fine sand and silt which becomes coarser at the base of the exposure. In an attempt to verify that the till exposed within this section was deposited by the Lake Michigan Lobe, the orientation of A0 pebbles occurring within the till was measured (Figure C-l). The orientation of these pebbles suggests that the flow direction of the ice which deposited this till was WSW (azimuth of approx. 100-110 degrees) and implies a Lake Michigan Lobe source.. 111 FIGURE C-l. Till fabric from Cooper exposure. lull, blue". sandy nut-uh, ammo- to (Ina an“ 7111, blown, “no, (1., Honey “H. [[01, (lay nu. alunay 1 Own-ash, u-aru sand and 'f ‘1 II J "a 7"710: o Transecl A a Transecl 8 FIGURE C-2. Bloomingdale measured section. 112 Bloomingdale Section The Bloomingdale section (Figure C-2) occurs in a gravel pit in north-central Van Buren County (sw, sect. 28, t l s, r 1A w). The base of the exposure is represented by a very coarse sand and pebble unit. Large scale cross-beds and channel scours suggest that this unit was deposited as outwash very near the ice margin. In several places in the gravel pit, this outwash shows normal faulting and slump features which suggest that blocks of ice detached from a retreating ice margin were present in the unit and melted, forming small kettle depressions in the surface. Some very fine organic particles occur in one of these kettle depressions. Directly overlying this outwash unit is a l to 2 meter (A to 6 feet) thick grey, clay-rich till. This till unit dips towards the south and is completely absent in the north wall of the gravel pit. Directly overlying the lowest grey till is a 2 to A meter (6 to 12 feet) brown sandy-clay till. The contact between these two till units is sharp and is represented both by a distinct color change and the occurrence of deformation structures such as shears and I'flame" structures. Large (>10 cm) cobbles commonly occur in both of these till units. A 2 to A meter (6 to 12 feet) thick outwash composed of ,medium to coarse sand and pebbles overlies the brown till. The section is capped by a brown, sandy and loose-textured till that pinches out in a few places. This till is 1 to 2 meters (2 to 5 feet) thick and the contact between it and the underlying outwash is sharp. Because of the presence of three distinct till units, detailed 113 sampling was penformed in this exposure. Two vertical transects were sampled at 1 foot intervals in each till unit. The clay mineralogy associated with these samples (Figure C-2) does not suggest any significant trends within a till unit, but does show a significant change in 7A/10A ratios between the upper two brown tills and the lower grey till. 11A REFERENCES REFERENCES Acomb, L.J, 1978, Stratigraphic relations and extent of Wisconsin's Lake Michigan Lobe red tills: M.S. Thesis. Univ. Wisc., Madison,- Wisc. Acomb,L.J., D.M. Mickelson, E.B. Evenson, 1982, Till stratigraphy and late glacial events in the Lake Michigan Lobe of easten Wisconsin, GSA Bull. v.93, p1289-296. Benninghoff, W.S. and C.W. Hibbard, 1961, Fossil pollen associated with a late glacial woodland musk-ox in Michigan: Papers of Mich. Acad. Sci., v. A6, pp. 155-159. 81ack,R.F., 1978, Comment on ”Greatlakean Substage: a replacement for Valderan Substage in the Lake Michigan basin” by E.8. Evenson, W.R. Farrand, D.F. Eschman, E.M. Mickelson, and L.J. Maher: Quat. Research, V.9, No. l, p.119-123. Bretz, J.H., 19513, Causes of the glacial lake stages in Saginaw basin, Michigan: Jour. Geology, v.59, p.2AA-258. ------- 1951b, The stages of Lake Chicago -- Their causes and correlations: Am.Jour.Sci., v.2A9, p.A01-A29. Carroll, 1972, Clay Minerals: A Guide to Their Identification: Geol. Soc. America Spec. Paper 126, 88 pp. Dreimanis, A., 1977. Late Wisconsinan glacial retreat in the Great lakes region, North America: Annals of the New York Acad. Sci., v.288a Pp-70'89. Dreimanis,A and R.Goldthwait, 1973, Wisconsin glaciations in the Huron, Erie, and Ontario lobes, in R.F. Black, R.P. Goldthwait, and H.8. Willman, eds., The Wisconsinan Stage: Geol. Soc. America Mem. 137, p.71-106. Dreimanis,A. and P. Karrow, 1972, Glacial history of the Great Lakes-St. Lawrence region, The classification of the Wisconsin(an) Stage, and its carrelatives: Internat. Gelo. Cong., 2Ath, Montereal, 1972, Quaternary Geology, Sec. 12, p. 5-15. Ekblaw,G.E. and L.F. Athy, 1925, Glacial Kankakee torrent. in northeastern Illinois: Geol. Soc. America Bull. v. 36, p. A17-A27. 115 Eschman,D.F., 1978, Pleistocene geology of the Thumb area of Michigan. in: Kesler, R.V., ed., Field Excursions form the University of Michigan, 1978 Mtg. of North-Central Section, Geol. Soc. America, Ann Arbor, p. 35-62. Eschman, D.F. and W.R. Farrand, 1970, Glacial history of the glacial Grand Valley, in Guidebook for field trips: Michigan Basin Geol. Soc., East Lansing, Mich, 1970, p.131-157. Eschman, D.F., 1981, Some evidence of Mid-Wisconsinan events in Michigan: Mich. Acad., v. 12, n. A, pp. A23-A35. Evenson, E.8., 1973. Late Pleistocene shorelines and stratigraphic relations in the Lake Michigan basin: G.S.A. Bull., V.8A,p.2281-2297. Evenson, E.8., D.F. Eschman, D.M. Mickelson, and L.J. Maher, 1976, Greatlakean Substage -- a replacement for Valderan Substage in the Lake Michigan Basin; Quaternary Res., V.6, p.All-AZA. Farrand, W.R., R. Zahner, and W.S. Benninghoff, 1969, Cary-Port Huron interstade -- Evidence for a buried bryophyte bed, Cheboygan County, Michigan: G.S.A. Spec. Paper 123, p.2A9-262. Farrand, W.R. and D.F. Eschman, 197A, Glaciation of the Southern Peninsula of Michigan -- a review: Michigan Academician, v.7, p.32-56 Forstat. D., 1982, Drift Thickness of Kalamazoo County, in: Monaghan, G.W., G.J. Larson, D.W. Forstat. and H.0. Sorensen, 1983, Selected Geologic Maps of Kalamazoo County, Michigan: County Geologic Map Series -- Kalamazoo. Mich. Geol. Surv. Div., Lansing, Michigan. Forstat. D. and H.O. Sorensen, 1982, Bedrock Geology of Kalamazoo County, in: Monaghan, G.W., G.J. Larson, D.W. Forstat. and H.O. Sorensen, 1983, Selected Geologic Maps of Kalamazoo County, Michigan: County Geologic Map Series -- Kalamazoo. Mich. Geol. Surv. Div., Lansing, Michigan. Frye, J.C. and H.8. Willman, 1960, Classification of the Wisconsinan stage in the Lake Michigan glacial lobe: Ill. Geol. Surv. Circ. 285, 16 p. Frye, J.C. and Willman, H.8., 1973, Wisconsinan climatic history interpreted from Lake Michigan Lobe deposits and soils, in Black, R.F., Goldthwait, R.R., and Willman. H.8., eds., The Wisconsinan Stage: Geol. Soc. America Memior 136. PP. 135-152. Frye, J.C., et al., 197A, The late Woodfordian Jules Soil. and associated moluscan faunas: Ill. Geol. Survey Circ. A86, 11p. Fullerton, 0.5., 1980, Preliminary correlation of Post-Erie Interstadial events (16,000 - 10,000 radio carbon years B.P.), 116 Central and Eastern Great Lakes region, and Hudson, Champlain, and St. Lawrence Lowlands, United States and Canada: U.S. Geol. Surv. Prof. Paper 1089, 51p. Gephart, G.D., 1982. Differentiation and correlation of till sheets using 7A/10A X-ray diffraction peak height ratios, Allegan County, Michigan: M.S. Thesis, Michigan State Univ., East Lansing, Mich. Gephart, G.D., G.W. Monaghan, G.J. Larson, 1982, A Mid-Wisconsinan event in the Lake Michigan Basin, Programs with Abstracts, Geol. Soc. Am., v.15, no. 7. ------- 1983. Stratigraphic evidence for a significant readvance of the Michigan ice lobe to the Kalamazoo Moraine -- southwestern Michigan: Programs with Abstracts, Geol. Soc. Am., v.16, n.A. Gephart, G.D. and G.J. Larson, 198A, Surficial Geology of Allegan County, in: Gephart, G.D., G.J. Larson, D.W. Forstat. and H.O. Rieck, 1983, Selected Geologic Maps of Allegan County, Michigan: County Geologic Map Series -- Allegan. Mich. Geol. Surv. Div., Lansing, Michigan. ------- , 198A, Glacial Geology of Allegan County, Michigan. in press with Mich. Dept. Nat. Resoc., Geol. Surv. Div., Lansing, Michigan. Glass, H.D., 1981, Clay mineral stratigraphy of the Pleistocene tills of Lake Michigan: (abst.) Symposium on the Valders Problem, N.A.G.T. East-Central Region, p.7. Gooding, A.M., 1961, Illinoian and Wisconsin history in southeastern Indiana, in Guidebook for field trips, Cincinnati meeting: Geol. Soc. America, pp. 99-130. ----- , 1963, Illinoian and Wisconsin glaciations in the Whitewater Basin, southeastern Indiana, and adjacent areas: Jour. Geology, V. 71, p. 665-682. ----- , 1971, Character of late Wisconsin tills in eastern Indiana: Indiana Geol. Survey. Grim, 1969, Clay Mineralogy. 2nd ed., McGraw-Hill book Co., New York, 569 pp- Jackson, M.L., 1956, Soil Chemical analysis: Advanced course: Univ. of Wisc., Madison, 89A p. Johnson, 1976, Quaternary stratigraphy in Illinois -- Status and current problems, in W.C. Mahaney, ed., Quaternary stratigraphy of North America: Stroudsburg, Pa., Dowden, Hutchenson, and Ross, p.16l-l96. 117 Jahns, R.H., 19A1, Outwash chronology in northeastern Massachusetts, Geol. Soc. Amer. Bull., v.52, n.12, pt.2, p.1910 ----- , 1953. Surficial geology of the Ayer quadrangle, Mass: U.S. geol. Survey Geol. Quad. Map GQ-Zl. Koteff, C., 1971, The morphologic sequence concept and deglaciation of southern New England. Leverett, F. and F. Taylor, 1915, The Pleistocene of Indiana and Michigan and the History of the Great Lakes, U.S. Geol. Surv. Monograph, v. 53, 527p. Leverett, F., 192A, Map of the Glacial Deposits in the Southern Peninsula of Michigan, Mich. Dept. of Conserv. pub. 2A. Lineback, J.A., D.L. Gross and R.P. Meyer, 197A, Glacial tills under Lake Michigan: 111. Geol. Survey, Environmental Geol. Note 69, A8p. ------- 1979, Glacial and post-glacial sediments in Lakes Superior and Michigan: G.S.A. Bull., part 1, v.90, p.781-791. Martin, H, Map of the Bedrock Formations of Michigan, Mich. Dept. of Conserv. pub. ??. McCartney, M.C. and D.M. Mickelson, 1982, Late Woodfordian and Greatlakean history of the Green Bay Lobe, Wisconsin: G.S.A. Bull., v.93. Melhorn, W.N., 195A, Valders glaciation of the Southern Peninsula of Michigan: Univ. Michigan Ph.D. dissert.,178p. Miller, N.G. and W.S. Benninghoff, 1969, Plant fossils from a Cary-Port Huron interstade deposit and their paleoecological interpretation: Geol. Soc. America Spec. Paper 123, p. 225-2A8. Monaghan, G.W. and G.J. Larson, 1980, Surface Geology in Eaton County, Michigan. Map on file with Tri-County Planning Commission, Lansing, Michigan. ------- , 1983, Map of Surficial Geology of Kalamazoo County. in: Monaghan, ~G.W., G.J. Larson, D.W. Forstat. and H.0. Sorensen, 1983, Selected Geological Maps of Kalamazoo County: County Geological Map Series -- Kalamazoo, Michigan Department of Natural Resources, Geol. Surv. Div., Lansing, Michigan. Monaghan, G.W., G.J. Larson, D.W. Forstat. and H.O. Sorensen, 1983, Selected Geological Maps of Kalamazoo County: County Geological Map Series -- Kalamazoo, Michigan Department of Natural Resources, Geol. Surv. Div., Lansing, Michigan. Morner, N.A. and A. Dreimanis, 1973, The Erie Interstade: Geol. Soc. America Mem. 136, p. 107-13A. 118 Ogden, J.G. and Hay, R.J, 1965, Ohio Wesleyn University natural radiocarbon measurements 11. Radiocarbon 7, pp. 166-73. Rieck, R.L., H.A. Winters, D.L. Mokma, and M.M. Mortland, 1979, Differentiation of surficial glacial drift in southeastern Michigan from 7A/10A X-ray diffraction rations of clays: G.S.S. Bull., Part 1, v.90,p.216-220. Rieck, R.L. and H.A. Winters, 1980, Distribution and significance of glacially buried organic matter in Michigan's southern Peninsula: Physical Geography, v. 1, n. 1, pp 7A-89. Shaw, B.P., 1969, Evaluation of Natural Aggregates in Kalamazoo County and Vicinity, Michigan. M.S. Thesis, Mich. State Univ. 191 pp. Straw, T.S., 1972, Some Aspects of Glacial Geology in the Kalamazoo Area. Guidebook, Geol. Soc. America Northcentral Sect. Springer, J.R. and R.C. Flemmal, 1981, Paleontolgical and geological results from two fossil Proboscidean finds in northern Illinois. Transactions of the 111. State Acad. Sci., v. 7A, no. 3-A, pp. 87-99. Wayne, W.J, 1963, Pleistocene formations of Indiana: Indiana Geol. Survey Bull. 25, 85 p. ----- , 1965, The Crawfordsville and Knightstown Moraines in Indiana: Indiana Geol. Survey Rept. Prog. 28, 15 p. Willman, H.8. and J.C. Frye, 1970, Pleistocene stratigraphy of Illinois: 111. Geol. Surv. Bull. 9A, 20Ap. Zumberge, J.H., 1960, Correlation of Wisconsin drifts in Illinois, Indiana, Michigan, and Ohio: G.S.A. Bull., v.71, p.1177-1188. Zumberge, J.H. and W.S. Benninghoff, 1969, A mid-Wisconsinan peat in Michigan, U.S.A.: Pollen et Spores, v. 11, pp. 585-601. Zumberge, J.H., and J.E. Potzger, 1956, Late Wisconsin chronology of the Lake Michigan Basin correlated with pollen studies: Geol. Soc. America Bull., V. 67, p.271-288. 119 "Illlillllllllllllllll“