Distal Record of Three Volcanic Centers During the Middle Campanian in the Western Interior Seaway

Home , Claggett Shale, Clidastes, Globidens, Platecarpus, Plioplatecarpus, Prognathodon

HIGH RESOLUTION EVENT STRATIGRAPHIC AND SEQUENCE STRATIGRAPHIC INTERPRETATION OF THE LOWER PIERRE SHALE (CAMPANIAN) WITH THE DESCRIPTION OF THE NEW WALHALLA AND CHAMBERLAIN MEMBERS

A dissertation submitted to the Division of Research and Advanced Studies of the University of Cincinnati in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY (Ph.D.) in the Department of Geology of the College of Arts and Sciences 2002 by Janet L. Bertog B. S. South Dakota School of Mines, 1995 M. S. South Dakota School of Mines, 1997

Committee Chair: Warren. D. Huff ABSTRACT The lower Pierre Shale represents a time of significant changes in the Cretaceous Western Interior Seaway, resulting from a complex interplay of tectonic events and eustatic sea level changes. Volcanic activity along the western margin of the seaway produced large volumes of ash that are recorded in the sediments of the basin in the form of bentonites. The bentonites provide information useful in interpreting volcanic activity that was active as well as providing event horizons in the strata that are useful for high-resolution stratigraphic analysis.

Bentonites of the lower Pierre Shale indicate that at least three major sources were active during this period, corresponding to forearc, island arc and backarc magmatism, as indicated by trace element whole rock geochemical signatures. Mineralogical suites of these bentonites indicate that both andesitic and rhyolitic sources were present within these major regions. Biotite phenocrysts further segregated magmatic sources of the bentonites.

Based on the composition of the bentonites, several individual layers of bentonite can be identified in the lower Pierre Shale. These bentonite horizons, combined with biostratigraphic correlation have provided high-resolution detailed stratigraphic evaluation of the lower Pierre Shale across the basin. The newly defined Walhalla Member represents a tectonically influenced sequence that precedes the eustatic sea level rise recognized as the Claggett Cycle in the Western Interior. The Walhalla Member is restricted to the northern part of the basin, where tectonic activity resulted in increased subsidence of the axial basin and the Williston Basin prior to the eustatic sea level rise. Following the Walhalla Member, eustatic sea level rise resulted in the deposition of the Sharon Springs and the Chamberlain members on the eastern stable platform while the Mitten Black Shale Member was deposited in the axial basin. Recognition of these distal sequences is necessary in order to understand the complete basin dynamics. Copyright (c) 2002 by: Janet L. Bertog Acknowledgements

I would like to thank Dr. Warren D. Huff for his time and effort in supporting this research and providing advice. I would also like to thank Dr. Tom Algeo, Dr. Gorden Bell, Dr. Carl Brett, Dr. James Martin, and Dr. Barry Maynard for their advice and support of this research. In addition, Dr. Craig Dietsch, Dr. Peter Harries, Dr. Attila Kilinc, Dr. Tom Lowell, Rick Bullard, Tresa Conrad, Sean Cornell, Dean Richmond, Jane Stormer, Jeff Von Loh, and David Ray provided active informal discussions of my research. Many landowners were very supportive of this research and provided me access to their lands to conduct my research. These people include Chuck Bonner, Ken Brown, Pete Bussen, Tom Conger, Ann Pfister, Vincent Wasserburger, the U. S. Forest Service, and the U. S. Corps of Engineers. The fieldwork for this project could not have been accomplished alone. Many people donated their time and effort to helping me collect samples in the field and provided discussion in the field for interpretation of the outcrops. These people include Dr. Gorden Bell, Dr. Carl Brett, Dr. Peter Harries, Dr. John Hoganson, Dr. Warren Huff, Dr. Gary Johnson, Dr. James Martin, Dr. Bruce Schumacher, Dr. Glenn Storrs, Pete Bussen, John Campbell, Sean Cornell, Mike and Pat Everhart, Diana Hecking, Darrin Pagnac, Jerry Riha, Marcus Ross, Bill Schurmann, Frank Varriale, Jeff VonLoh, Aric Wilisch, and John Zancanella. Joe Williams of the University of Cincinnati Advanced Materials Characterization Center supported my research through technical maintenance, wonderful ideas and assistance for sample preparation, support of the microprobe and SEM and moral support for many long months. Joe deserves special recognition in particular for helping me through the rough parts in my research and providing excellent suggestions at just the right time. Mike Menard always provided technical support for many different ideas that came up in sample preparation and analysis Dean Smith of Petrographic International prepared my thin sections for petrographic and microprobe analysis. Dr. Warren Huff and Dr. Barry Maynard provided training and support for the equipment available at the University of Cincinnati, Dpeartment of Geology for sample preparation. My parents Stan and Barbara Bertog have been very supportive and patient with my research and long years of school. Many organizations provided funding for the fieldwork and lab analysis required for this research. These associations include the American Association of Petroleum Geologists, Charlie Landis at American Colloid Company, the Clay Minerals Society, the Colorado Scientific Society, the Department of Geology at the University of Cincinnati, the Isabel and Mary Neff Fellowship at the University of Cincinnati, the University Research Council at the University of Cincinnati and the Fenneman Fellowship from the Department of Geology at the University of Cincinnati. The University of Cincinnati Department of Geology also provided funding through the summer research stipend, the geology research fund, the sedimentology fund and the Walter H. Bucher fund for travel. The Geological Society of America and the Rocky Mountain and North Central Sections of the Geological Society of America provided support for travel to annual meetings. i Table of Contents Table of contents ...... i

Table of figures ...... iv

INTRODUCTION ...... 1

Structural setting of the Western Interior Seaway ...... 1

Tectonic and eustatic controls on deposition of the lower Pierre Shale ...... 6

Volcanism of the western United States during the Campanian 9

Paleoecology of the Middle Campanian Western Interior Seaway 11

Campanian faunal pattern ...... 13

HISTORICAL DESCRIPTIONS OF THE STRATIGRAPHIC OF THE LOWER PIERRE SHALE ...... 19

The Pierre Shale ...... 19

The Sharon Springs Member ...... 22

The Pembina Member...... 26

The Gammon Ferruginous Member ...... 27

The Mitten Black Shale Member ...... 29

The Redbird Silty Member ...... 30

The Gregory Member ...... 30

Revised stratigraphy of the lower Pierre Shale ...... 31

Bibliography ...... 32 ii GEOCHEMICAL AND MINERALOGICAL RECOGNITION OF THE ARDMORE BENTONITE SUCCESSION AND ITS USE IN REGIONAL STRATIGRAPHIC CORRELATION ...... 41

Abstract ...... 41

Introduction ...... 42

Stratigraphy of the Ardmore succession ...... 47

Volcanism of the Middle Campanian ...... 51

Methods ...... 53

Results...... 55

Phenocryst mineralogy ...... 55

Biotite geochemistry ...... 58

Discussion ...... 61

Magma types ...... 61

Volcanic sources ...... 63

Sedimentation patterns and basin dynamics ...... 64

Conclusions ...... 72

Bibliography ...... 73

DISTAL RECORD OF THREE VOLCANIC CENTERS DURING THE MIDDLE CAMPANIAN IN THE WESTERN INTERIOR SEAWAY ...... 76

Abstract ...... 76

Introduction ...... 76

Stratigraphy of the Middle Campanian ...... 79

Stratigraphy of the “Sharon Springs Member” in the Black Hills 80

Methods ...... 82 iii Results...... 86

Clay mineralogy ...... 86

Whole rock geochemistry ...... 87

Phenocryst mineralogy ...... 89

Biotite geochemistry ...... 92

Discussion ...... 94

Conclusion ...... 97

References ...... 98

TECTONIC AND EUSTATIC CONTROLS ON FACIES DISTRIBUTIONS IN THE LOWER PIERRE SHALE (CAMPANIAN) WITH A DESCRIPTION OF THE NEW WALHALLA AND CHAMBERLAIN MEMBERS ...... 105

Abstract ...... 105

Introduction ...... 106

Histrorical descriptions of the stratigraphy of the lower Pierre Shale 109

Methods ...... 111

Results...... 112

Gammon Ferruginous Member ...... 113

Regional Unconformity ...... 115

Walhalla Member ...... 116

Sharon Springs Member ...... 122

Chamberlain Member ...... 126

Unassigned unit ...... 128

Other members of the Pierre Shale ...... 131 iv Discussion ...... 131

The Niobrara Cycle ...... 132

The Walhalla Sequence ...... 134

The Claggett Cycle ...... 137

Conclusion ...... 139

References ...... 140

RECORD OF TECTONIC ACTIVITY OF THE SEVIER OROGENIC BELT IN THE DISTAL MARINE SEDIMENTS OF THE CRETACEOUS WESTERN INTERIOR SEAWAY ...... 145

Abstract ...... 145

Introduction ...... 146

Foreland basin models ...... 148

Intracratonic basins ...... 150

Stratigraphy of the Walhalla Member...... 151

Correlation within the Walhalla Member ...... 153

Unconformities in the lower Pierre Shale ...... 154

Peripheral bulge development in the lower Pierre Shale ...... 156

Conclusion ...... 157

References ...... 158 v Table of Figures INTRODUCTION ...... 1

Figure 1.1: Generalized stratigraphy of the Cretaceous western interior of North America ...... 2

Figure 1.2: 3-dimensional reconstruction of the western interior seaway during the Baculites obtusus ammonite range zone ...... 3

Figure 1.3: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway ...... 6

Figure 1.4: Sea-level curves proposed for the western interior seaway during the Late Cretaceous ...... 8

Figure 1.5: Paleogeographic reconstructions of the basin during times of quiescence (A) and active tectonism (B) 9

Figure 1.6: Bentonites of the lower Pierre Shale record three volcanic sources based on their chondrite- normalized rare earth element patterns ...... 11

Figure 1.7a: Invertebrate biostratigraphy of the Santonian and Campanian of the western interior ...... 14

Figure 1.7b: Vertebrate biostratigraphy of the Santonian and Campanian of the western interior ...... 15

Figure 1.8a: Generalized stratigraphy of the lower Pierre Shale, as it is currently recognized in the literature ...... 21

Figure 1.8b: Stratigraphy of the lower Pierre Shale, as modified in this paper ...... 21 vi HISTORICAL DESCRIPTIONS OF THE STRATIGRAPHIC OF THE LOWER PIERRE SHALE ...... 19

Figure 2.1: Generalized stratigraphy of the Middle Campanian in the western interior of the United States ...... 43

Figure 2.2: Generalized stratigraphy of the Ardmore succession 45

Figure 2.3: Map of the western interior of North America showing major geographic features for the region ...... 46

Figure 2.4: Ardmore succession at Wasserburger Ranch, South Dakota ...... 47

Figure 2.5: Ardmore succession at Redbird, Wyoming ...... 48

Figure 2.6: Active volcanism and tectonics of the Cretaceous Western Interior Seaway ...... 52

Figure 2.7: Quartz grains from the Ardmore succession ...... 55

Figure 2.8: Potassium feldspar from the Ardmore succession 55

Figure 2.9: Plagioclase from the Ardmore succession ...... 55

Figure 2.10: Biotite grains of the Ardmore succession ...... 56

Figure 2.11: Apatites of the Ardmore succession ...... 56

Figure 2.12. Most zircons of the Ardmore succession ...... 56

Figure 2.13: Zircons ...... 56

Figure 2.14: Ilmenites are a rare accessory mineral ...... 56

Figure 2.15: Phenocryst composition of the bentonites of the Ardmore succession ...... 57

Figure 2.16: Biotites of the Ardmore succession ternary plots 59

Figure 2.17: Bavarient plots of biotite in the Ardmore succession 60

Figure 2.18: Stratigraphic correlation of the Ardmore succession from Montana to eastern South Dakota indicating bentonites of differing compositions ...... 65 vii Figure 2.19: Stratigraphic correlation of the Ardmore succession from eastern North Dakota to western Kansas ...... 66

Figure 2.20: Sea level curves of the Western Interior Seaway (Kauffman and Caldwell, 1993) vs. the eustatic sea level curve (Haq et al., 1987) ...... 67

Figure 2.21: Generalized cross-section of the northern part of the basin from the Black Hills region to eastern South Dakota 68

Figure 2.22: The Ardmore succession near Chamberlain, South Dakota ...... 69

Table 2.1: Localities of the Ardmore succession ...... 46

DISTAL RECORD OF THREE VOLCANIC CENTERS DURING THE MIDDLE CAMPANIAN IN THE WESTERN INTERIOR SEAWAY ...... 76

Figure 3.1: During the Middle Campanian, several volcanic centers were active in western North America ...... 77

Figure 3.2: Generalized stratigraphy of the lower Pierre Shale in the Black Hills and western Kansas ...... 78

Figure 3.3: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway ...... 79

Figure 3.4: Generalized stratigraphy of the major bentonites of the lower Pierre Shale in the Black Hills ...... 81

Fig. 3.5: Map of South Dakota showing localities used in this study 83

Figure 3.6: General clay mineralogy of bentonites in the lower Middle Campanian are montmorillonite, as shown by x-ray diffraction ...... 86

Figure 3.7: Representative chondrite-normalized plots for bentonites of the Sharon Springs...... 87

Figure 3.8: Discriminant function analysis produced four distinct groups in the bentonites of the “Sharon Springs Member” 88 viii Figure 3.9: Characteristic minerals of the bentonites in the “Sharon Springs Member” ...... 90

Figure 3.10: Mineralogy of bentonites of the lower Middle Campanian strata of the western interior ...... 91

Figure 3.11: Biotite geochemistry of the bentonites in the “Sharon Springs Member” ...... 93

Figure 3.12: Bivarient plots of major element analysis of biotites from bentonite samples in the “Sharon Springs Member” 93

Figure 3.13: Subduction zone environment showing the zones of magma generation ...... 95

Table 3.1 Localities of the lower Pierre Shale used in this study 83

Table 3.2: Concentrations (in ppm) from bentonite samples using INAA and ICP ...... 101

Table 3.3: Light mineral composition of bentonites in the lower Middle Campanian ...... 102

Table 3.4: Microprobe analysis of bentonites of the lower Middle Campanian ...... 103

TECTONIC AND EUSTATIC CONTROLS ON FACIES DISTRIBUTIONS IN THE LOWER PIERRE SHALE (CAMPANIAN) WITH A DESCRIPTION OF THE NEW WALHALLA AND CHAMBERLAIN MEMBERS ...... 105

Figure 4.1: Generalized stratigraphy of the Late Cretaceous 107

Figure 4.2: During the Middle Campanian, several volcanic centers were active in western North America ...... 108

Figure 4.3: Generalized stratigraphy of the lower Pierre Shale 110

Figure 4.4: Representative meaured sections of the Gammon Ferruginous Member ...... 114

Figure 4.5: Representative measured sections of the Walhalla Member ...... 117 ix Figure 4.6: The Walhalla Member overlying the Niobrara Formation in the Pembina Gorge near Walhalla, North Dakota ...... 118

Figure 4.7: Stratigraphy of the lower Pierre Shale near Chamberlain, South Dakota ...... 118

Figure 4.8: The Ardmore bentonite succession of the Walhalla Member ...... 119

Figure 4.9: Representative measured sections of the Sharon Springs Member ...... 123

Figure 4.10: The Sharon Springs Member at the type locality, McAllaster Buttes, Kansas ...... 124

Figure 4.11: Representative measured sections of the Chamberlain Member ...... 127

Figure 4.12: The unassigned unit (possibly Mitten Black Shale Member at Wasserburger Ranch, South Dakota 129

Figure 4.13: Representative measured sections of the unassigned unit ...... 129

Figure 4.14: Chronostratigraphic correlation of third and fourth-order sequences in the Lower Pierre Shale ...... 133

Figure 4.15: Paleogeographic reconstructions of the basin ...... 134

RECORD OF TECTONIC ACTIVITY OF THE SEVIER OROGENIC BELT IN THE DISTAL MARINE SEDIMENTS OF THE CRETACEOUS WESTERN INTERIOR SEAWAY ...... 145

Figure 5.1: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway ...... 146

Fig. 5.2: Thrusting activity in western Wyoming and Utah resulted in a broad foreland basin in the Western Interior ...... 147

Fig. 5.3: Comparison of the elastic and viscoelastic model of deformation ...... 148

Fig. 5.4: Interaction between the foreland basin and an intracratonic basin depends on the distance between the two basins 150 x Fig. 5.5: Stratigraphy of the lower Pierre Shale based on biostratigraphy and bentonite correlations ...... 152

Fig. 5.6: Variations in the whole rock geochemistry of bentonites in the Lower Pierre Shale ...... 153

Fig. 5.7: Biotite composition of bentonites of the lower Pierre Shale have been used to discriminate individual layers ...... 154

Fig. 5.8: Model of the forebulge migration in the Black Hills region 156 1 INTRODUCTION The lower Pierre Shale (Campanian, Cretaceous) represents a time of multiple changes in the environment of the Western Interior Seaway, resulting from a combination of tectonic and eustatic changes in the seaway. The lower Pierre Shale indicates a shift from carbonate environments of the Niobrara Formation to clastic environments of the Pierre Shale resulting from a eustatic sea level fall (Fig. 1.1, Kauffman and Caldwell, 1993) coupled with increased tectonic activity, particularly along the Absoroka thrust belt (Fig. 1.2, Monger, 1993). These environmental changes also resulted in a faunal turnover in both the marine and terrestrial realms (Martin, 1996; Russell, 1993). Based on lithology and ammonite zonation, the “Sharon Springs Member” is the most widely recognized unit in the lower Pierre Shale as the deep-water facies of the transgressive and highstand systems tracts of the Claggett Cyclothem (Fig. 1.1, Gill and Cobban, 1966). Despite the superficially monotonous nature of the “Sharon Springs Member”, faunal patterns had previously indicated a much more complex sequence architecture (Cobban, 1993; Kauffman, 1984; Martin, 1996; Russell, 1993). Through detailed stratigraphic analysis including bentonite correlation and high-resolution stratigraphic analysis, however, a more complex facies distribution is now confirmed. Revision of the stratigraphy of the lower Pierre Shale helps to clarify the facies patterns and provides the detail necessary for an evaluation of the tectonic and eustatic influences on the changes in the basin. In addition, bentonite geochemistry provides information regarding active volcanic sources during this time period. STRUCTURAL SETTING OF THE WESTERN INTERIOR SEAWAY The broad tectonic setting of the Cretaceous Western Interior of North America was a retroarc foreland basin resulting from thrust loading during the Sevier Orogeny along the 2

Figure 1.1: Generalized stratigraphy of the Cretaceous western interior of North America including European stage names corresponding with absolute age dates and North American ammonite range zones. 3

Figure 1.2: 3-dimensional reconstruction of the western interior seaway during the Baculites obtusus ammonite range zone. Colors indicate approximate elevations during this time. Blue is below sea level, green is above sea level. The darkest blue indicates the areas of deepest water. Red lines are modern state boundaries for reference.

western coast of North America (Fig. 1.2, Kauffman, 1977). Precambrian and Paleozoic structural features that were reactivated by compressional forces from the west dissected this broad basin (Shurr and Rice, 1986). During the Late Cretaceous, the western coast of North America was experiencing subduction of the Farallon Plate under the North American plate (Fig. 1.2, Monger, 1993), resulting in the accretion of exotic terrains along the coast, the formation of a magmatic arc along the western margin of North America and thrusting of the Sevier Orogenic belt further to the east. Coincident with the deposition of the Sharon Springs Member, at about 80 Ma, 4 the Farallon and Kula plates were both moving north-northeast as they subducted under North America. Along the margin of the North America plate, orthogonal components were translated into contractional structures, while a northward component resulted in strike- slip motion (Monger, 1993). Thrusting in the Sevier Orogenic belt shaped the basin of the Western Interior Seaway. The western margin was the deepest part of the basin structurally, due to subsidence resulting from high clastic sedimentation off the adjacent thrust belt. This zone, called the western foredeep (Fig. 1.2), was 150-200 km wide and was the coastal plain of the Western Interior Seaway, represented by the Mesaverde Group and Two Medicine Formation (Rice and Shurr, 1983). To the east of the western foredeep, the compressional forces reactivated basement faults and created a western forebulge (Fig. 1.2), resulting in structural highs that sequestered the coarse-clastics of the coastal plains within the western foredeep and separated them from the finer grained sediments to the east (Kauffman and Caldwell, 1993). The finer grained sediments to the east were deposited in the axial basin or the west median trough (Fig. 1.2), also called a back-bulge basin (DeCelles, 1994; Kauffman and Caldwell, 1993). Due to lower sedimentation rates in this part of the basin, it was the bathymetrically deepest part of the basin. The deepest areas during this time interval were near the Wyoming – South Dakota border (Rice and Shurr, 1983). The axial basin is separated from the eastern platform by the east-median hinge (Fig. 1.2), which is a peripheral bulge resulting from bending of the crust (Kauffman and Caldwell, 1993). This deformation results from tectonics and sediment load along the western margin, producing a basin adjacent to the sediment load and uplift on the distal margin of the basin (Miall, 1993) and was partially controlled by basement faults (Rice and Shurr, 1983). Basement faults in the basin weaken the crust and provide a region of slippage for the rebound, and so these basement faults are thought to have controlled the position of the peripheral bulge. The eastern platform, particularly in the northern part of the basin, was dissected by several lineaments and reverse faults trending dominantly northeast – southwest, 5 parallel to the transcontinental arch, with a minor northwest – southeast component that separated broad anticlinal blocks and narrow synclines and strike-slip features (Anna, 1986). One of the most active fault blocks during the Campanian was the Hartville Uplift in western Nebraska and southeastern Wyoming (Fig. 1.2, Weimer, 1984). These uplifts signify initiation of Laramide-style deformation within the Western Interior Basin, where basement faults were reactivated and deep crustal deformation segmented the basin (Shurr and Rice, 1986). These faults resulted in high areas along the Peace River Arch and Sweetgrass-Battleford Arch in Canada (Stelck, 1975) and the transcontinental arch in South Dakota and Wyoming (Fig. 1.2, Shurr, 1984; Shurr et al., 1989; Weimer, 1984) while the Williston Basin in North Dakota remained a low basin between the eastern platform and the transcontinental arch (Fig. 1.2, Shurr et al., 1989). In the northern part of the basin, tectonism was very active throughout the Campanian and the deposition of the lower Pierre Shale was widely affected by these events. However, in the southern part of the basin, the seaway was broadly shaped by ancient tectonism and smaller-scale tectonism that was present in Arizona and Nevada during this time (Bilodeau and Lindberg, 1983). Six major thrusting episodes took place during the Late Cretaceous – Paleocene (Villien and Kligfield, 1986). Thrusting along the Canyon Ridge Thrust in central Utah (Villien and Kligfield, 1986) and the Absoroka Thrust in Wyoming (DeCelles, 1994) is coincident with the initial deposition of the lower Pierre Shale (Fig. 1.3, Obradovich, 1993). The timing of these fault events and the deposition of the lower Pierre Shale have been considered coincident with the eustatic sea level rise (Fig. 1.3, Haq et al., 1988; Kauffman and Caldwell, 1993). 6

Figure 1.3: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway. Orogenic thrusting events occur synchronously with transgressions in the basin (Villien and Kligfield, 1993). The major cycles recognized by Kauffman and Caldwell (1993) and generally accepted for the northern part of the seaway are shown, relating thrusting events to the transgression of these cycles. Some of the major stratigraphic units, particularly within the Campanian, are given for the western margin and the distal sediments. Tectonic material compiled from Villient and Kligfield (1993), Wiltschko and Dorr (1983) and DeCelles (1994). TECTONIC AND EUSTATIC CONTROLS ON DEPOSITION OF THE LOWER PIERRE SHALE

Deposition of the lower Pierre Shale marks a time of numerous changes in the western interior seaway. During the middle Campanian, the Zuni II second order eustatic sea level cycle reached its peak with deposition of the Niobrara Formation when overall sea-level highs and tectonic quiescence favored carbonate deposition (Kauffman and Caldwell, 1993). This highstand was followed by a general regression that continued throughout the end of 7 the Cretaceous, and carbonate deposition gave way to clastic deposition, which dominated the seaway throughout this time. The organic-rich shale of the “Sharon Springs Member” has been considered the transgression of the third-order Claggett Cycle, superimposed on an overall regression, throughout the Western Interior Seaway (Kauffman and Caldwell, 1993). The Claggett Cycle is one of nine third-order cycles recognized in the Late Cretaceous strata of the Western Interior Seaway (Fig. 1.4). They are thought to be tectono-eustatic, resulting from increased sea floor spreading rates, which increased tectonism along seaway margins and increased subsidence rates within continental interiors, while also raising sea level due to displacement of water along the mid-ocean ridges (Kauffman and Caldwell, 1993). Kauffman and Caldwell (1993) have shown that these cycles roughly correlate to the eustatic sea-level changes proposed by Haq et al. (1987), although they do not match perfectly (Fig. 1.4). Although the cycles of Kauffman have been widely accepted, some authors disagree with details of this interpretation, contending that the cycles were proposed based on incomplete data. Lillegraven and Ostrech (1990) suggested that the transgression and regression associated with the Claggett Cycle was restricted to the northern part of the basin in Montana and Canada, whereas in the southern part of the seaway a smaller-scale transgression and highstand occurred within the San Juan Basin. They suggested that the northern and southern parts of the basin were separate and responded to different tectonic conditions (Lillegraven and Ostresh, 1990). Lillegraven and Ostrech (1990) recorded an overall regression from the zone of Collinaceras choteauensis (Santonian) through Didymoceras nebrascensis (Maastrichtian), interpretted by Kauffman and Caldwell (1993) as the regression of the Zuni Megasequence. From there, the interpretations vary significantly. In the northern part of the basin, the sea-level fluctuations recorded by Lillegraven and Ostrech (1990) match those of Kauffman and Caldwell (1993; Fig. 1.4). The Eagle Sandstone represents a progradation of the seaway correlated with the end of the Niobrara Cycle. The Claggett transgression is recognized by sediments of the “Sharon Springs Member” in the 8

Figure 1.4: Sea-level curves proposed for the western interior seaway during the Late Cretaceous. Refer to table 1 for the ammonite zonations referred to on this diagram. northern part of the basin. This period of transgression also correlates with timing on the Absoroka Thrust (Witschko and Dorr Jr., 1983) and may be a result of synorogenic subsidence in the basin. Cobban et al. (1994) supported the notion that the transgression was restricted to the northern part of the basin. In the southern part of the basin, Lillegraven and Ostrech (1990) recorded a number of localized transgressions and regressions, none of which can be correlated throughout the entire basin. Rather, they demonstrated segregation between

the northern and southern parts of the basin. Lillegraven and Ostrech (1990) recorded a lengthy interval of transgressions, stillstands and regressions from the Baculites sp. (weak flanked ribs) – Baculites compressus zones. The initial transgression of this interval nearly coincides with the transgression of the Claggett Cycle at the Baculites obtusus range zone recognized by Kauffman (1984), although it may have occurred slightly earlier in the San Juan Basin (Fig. 1.4). These various interpretations indicate that the seaway was much more complex than previously thought and casts doubt on the interpretations presented due to lack of 9 continuity between any two sets of data. As would be expected in an active continental basin, eustatic sea-level fluctuations alone are not the only factor in the sedimentation patterns of the basin. The eustatic sea-level fluctuations are compounded by active tectonism along the western margin of the seaway. High-resolution interpretation of the seaway indicates that during periods of active tectonism in Utah and Wyoming, thrusting resulted in increased subsidence in the northern part of the basin, corresponding to the axial basin and the Williston Basin and resulting in a north to south dichotomy in sedimentation patterns. However, during periods of tectonic quiescence, sedimentation patterns reflected a retroarc foreland basin with north-south trending parallel facies belts (Fig. 1.5, Chapter 4).

Figure 1.5: During times of tectonic quiescence, sedimentation patterns reflected a retroarc foreland basin with north-south trending parallel facies belts (A), however during times of tectonic activity in Wyoming and Utah, the axial basin and the Williston Basin in the northern part of the basin subsided, resulting in a north to south dichotomy in sedimentation patterns (B). VOLCANISM OF THE WESTERN UNITED STATES DURING THE CAMPANIAN Volcanism reached a peak between 75-80 Ma, in two broad regions, one along the western margin of the United States and a second further inland, with a dominant center in western Montana (Armstrong and Ward, 1993). Along the western margin of North America, sodic to normal calc-alkaline batholiths occurred along a thin, continuous band and numerous, small calc-alkaline plutons occurred in British Columbia and Washington (Fig. 1.2, Armstrong and Ward, 1993). The last Sierran volcanism in Nevada occurred about 80 Ma and was distinct in containing abundant muscovite (Vikre and McKee, 1985). Inland, the 10 Boulder Batholith and Little Elkhorn Mountains in Montana were a prominent source of volcanics, particularly at 80 Ma, where interplay of thrusting, Laramide-style deformation and magmatism produced very high volumes of ash (Fig. 1.2, Hyndman, 1983; Hyndman et al., 1975; McMannis , 1965 #197; Robinson et al., 1968; Armstrong and Ward, 1993). In the south, a shift to a more shallowly subducting slab resulted in a shift of magmatism from California into Colorado (Fig. 1.2, Cross and Pilger Jr., 1978; Dickenson, 1981). This frequent volcanic activity shed enormous amounts of ash, which were carried by winds into the Western Interior Seaway and deposited within the sediments of the basin (Christensen et al., 1994). Where these ashes were deposited in marine sediments, they were altered to bentonites rich in smectite or illite-smectite (or occasionally kaolinite) (Wherry, 1917). Where they were deposited in coal swamps, the ashes have been altered to tonsteins rich in kaolinite (Bohor and Triplehorn, 1993). Therefore, bentonites that are deposited in the distal sediments of the Western Interior Seaway also provide a record of volcanic activity along the western margin. Bentonites of the lower Pierre Shale record three distinct volcanic sources corresponding to with fore-arc, volcanic island arc and backarc volcanism (Fig. 1.6, Chapter 3). Within these three settings, rhyolitic to dacitic volcanic events occurred. While the back-arc volcanic source for these bentonites is known to be the Little Elkhorn Mountains Volcanic Complex in western Montana (Defant and Drummond, 1993; Monger, 1993), it remains unclear what the exact sources are for the fore-arc and island arc volcanic material (Chapter 3). Volcanism that may have been present in a forearc or island arc setting may not be easily recognized due to later collisional events along the western margin of the United States. 11

Figure 1.6: Bentonites of the lower Pierre Shale record three volcanic sources based on their chondrite-normalized rare earth element patterns. Bentonites with light rare earth element concentrations around 100 times chondrite (A) indicate a backarc magma source. Bentonites with less than 100 times chondrite (B) indicate an island arc magma source. Bentonites with a positive europium anomoly (C) indicate a forearc magma source.

PALEOECOLOGY OF THE MIDDLE CAMPANIAN WESTERN INTERIOR SEAWAY Paleoecological interpretations of the Western Interior Seaway have focused on the distribution of previously recognized members, especially the widely recognized “Sharon Springs Member”. By evaluating facies distributions based on the member terminology, erroneous correlations have been made over the region, and paleoecological interpretations have been based on these correlations. Because bentonites of the Lower Pierre Shale indicate multiple volcanic sources, these bentonites can be used in conjunction with biostratigraphic and event stratigraphic horizons for high-resolution regional correlations of the basin (Chapter 2-3). Three primary interpretations have previously been suggested for deposition during the lower Pierre Shale. According to one interpretation, during deposition of the Niobrara Formation, the seaway was relatively open-marine and had a maritime climate with warming temperatures, rising surface salinity and decreasing seasonality. However, as sea level began to fall during the regression of the Zuni Megasequence and the basin became more segregated, resulting in the ecology of the seaway during deposition of the lower Pierre Shale changing dramatically. At this time, the seaway was density stratified. Overall, the seaway was thought to have been dominantly brackish due to freshwater influx of the highlands to the 12 west. The lighter freshwater tended to form a lower salinity cap on the seaway, and the deeper waters were slightly higher in salinity (Kauffman, 1984). During sea-level highstands, warm southern waters encroached the seaway and the basin became thermally stratified over denser, cooler, northern water masses (Kauffman and Caldwell, 1993). According to a second hypothesis, the organic carbon accumulation of the “Sharon Springs Member” resulted from mixing of the northern and southern water masses during periods when the two water masses would produce a blended water of greater density that would sink, carrying with it abundant organic material, such as surface plankton. The abundance of organic matter in the lower water column would result in reduced levels of dissolved oxygen (Hay et al., 1993). The third hypothesis is one in which high productivity associated with upwelling produces the organic-rich sediments of the “Sharon Springs Member” (Parrish and Gautier, 1993). Features that they used to support an upwelling environment were: 1) abundant phosphate nodules, 2) lateral facies distribution of glauconitic sandstone, phosphate and organic-rich carbon, 3) abundant remains of marine organisms from high in the food chain, 4) sediments consisting almost entirely of organic-rich fecal pellets, dominantly composed of marine phytoplankton, 5) abundant pyrite, and 6) abundant vertebrate predators. They also point out that the organic-rich facies is widespread, particularly on the eastern platform, and is not confined to the deepest part of the basin. Their evaluation of these facies, however, included the entire “Sharon Springs Member” in bulk in several localities, and did not look at the stratigraphic distribution on a finer scale. Although multiple hypotheses have been proposed as alternatives for the organic- rich facies in the Western Interior Seaway, the anoxic bottom water, stratified water column model has been the most widely accepted and supported theory. This theory also appears to be supported by the high-resolution stratigraphic interpretation presented herein (Chapter 4). 13 CAMPANIAN FAUNAL PATTERNS Due to the many changes that took place within the seaway during deposition of the Sharon Springs Member, it is not surprising to see significant turnovers in the fauna at many levels. Foraminifera are abundant in the Late Cretaceous strata of the Western Interior and are dominantly controlled by paleoecology. As a result, the distribution of foraminifera in the Western Interior tends to parallel the cycles (Caldwell et al., 1993). During marine transgressions, particularly the Greenhorn and the Niobrara, Tethyan fauna were brought as far north as Alberta and Saskatchewan (Caldwell et al., 1993). Highstands and regressions, on the other hand, were dominated by arenaceous and benthic fauna (Caldwell et al., 1993). During the deposition of the “Sharon Springs Member”, however, the foraminifera were severely limited. As opposed to other transgressions, an arenaceous benthic fauna dominated the transgression of the Claggett Cycle (Caldwell et al., 1993; McNeil and Caldwell, 1981; Mello, 1971), contradictory to the expected higher ratio of planktonics to benthics during a rise in sea level. Radiolarians also tend to be more common than foraminifera in general during this transgression (Bertog, unpublished data). These trends may be due to the high volcanic activity and high preservation of bentonites within the transgression of the Claggett Cycle, which created a more acidic environment. The bentonites generally produced an acidic diagenetic environment, which would be expected to preferentially destroy calcareous skeletons. In addition, silica enrichment in the water column due to the ash would cause preferential production of diatoms over calcareous organisms. Although a benthic fauna of forams existed, the diversity is low, restricted to less than 10 species (Fig.1.7, Caldwell et al., 1993; McNeil and Caldwell, 1981; Mello, 1971) and the species represent low oxygen conditions, consistent with a dysoxic bottom condition. The primary forams present were Trochamina ribstonensis and Bathysiphon vitta. The Trochamina, a planispiral form, tended to be large and flattened, increasing the surface to volume ratio, while the Bathysiphon vitta were thin and elongate in order to accomplish the high surface area. Increased surface area allowed them to survive in lower oxygen conditions 14 bivalves have been used to establish biozones in the western interior. Lines represent known distribution of the fuana. Lines represent establish biozones in the western interior. Figure 1.7a: Invertebrate biostratigraphy of the Santonian and Campanian of the western interior. Foraminifera, ammonites and 1.7a: Invertebrate biostratigraphy of the Santonian and Campanian western interior. Figure 15 es and Marine Vertebrate Ages es and Marine Vertebrate that have been proposed are shown to the left. are that have been proposed Figure 1.7b: Vertebrate biostratigraphy of the Santonian and Campanian of the western interior. North American Land Mammal Ag biostratigraphy of the Santonian and Campanian western interior. 1.7b: Vertebrate Figure 16 by maximizing the amount of area capable of incorporating oxygen (Lipps et al., 1979). The molluscs also took a major hit during the deposition of the Sharon Springs Member. Mollusca are by far the most common invertebrates of the western interior seaway. Brackish water conditions (Kauffman and Caldwell, 1993) and soft, muddy substrates (Gill and Cobban, 1966) prohibited migration and diversification of typical marine fauna including sponges, bryozoans, brachiopods, corals and echinoderms. During times of transgression, warm tropical waters brought immigrant species into the western interior basin (Kauffman, 1984). As a result, peaks in diversity can be seen in conjunction with these transgressions. Among ammonites, major diversification peaks are seen in the middle late Cenomanian, middle Turonian, Coniacian and Santonian, and the late Campanian and early Maastrichtian. The transgression of the Claggett Cycle is once again an exception to the rule of higher diversity during transgression. During the zone of Baculites obtusus, only Baculites, Trachyscaphites and Placenticeras are known (Fig. 1.7, Cobban, 1993). Baculites and inoceramids are the most common molluscs in the lower Pierre Shale (Gill and Cobban, 1966). Most of the molluscs are found only in concretions where early diagentic conditions were favorable to preservation (Gill and Cobban, 1966). Inoceramids are adapted to low oxygen conditions and are opportunistic, growing on any hard substrate they could find, including vertebrate remains. This adaptability allowed these organisms to survive even in the low oxygen conditions that persisted through the transgressive phase of the Claggett Cycle. Unlike the reduced invertebrate fauna, vertebrate fossils are very common in the lower Pierre Shale. However, significant faunal changes are recognized in this interval, many of which merit further attention. A significant turnover is recognized in the marine fossils of this interval, from a fauna typical of the fauna of the Niobrara Cycle (“Niobraran Age”), to a fauna more typical of the upper Pierre Shale (“Navesinkian Age”) (Fig. 1.7, Russell, 1993). The end of the Niobraran “Age” is characterized by the loss of the fishes 17 Apsopelix, Xiphactinus, and the bird Ichthyornis (Russell, 1993). The fauna of the lower Pierre Shale also differs from the typical “Niobraran” fauna in that no Cretoxyrhina (shark) have been found, and Ichthyodectys (fish), toxychelids (turtles), hesperornids (birds) and polycotylids (short-necked plesiosaurs) are more abundant (Russell, 1993). It is interesting to note that the transition of many of the marine vertebrates occurred after the third order cyclic shift from the Niobrara Cycle to the Claggett Cycle recognized by (Kauffman and Caldwell, 1993). Brett (1998) have shown that most faunal shifts occur associated with maximum highstands rather than at sequence boundaries, and this appears to be the pattern followed here. A significant transition in mosasaurs is observed at this interval (Martin et al., 1996). Mosasaurs of the “Niobraran Age”, including Clidastes and Tylosaurus terminate within the Sharon Springs – Mitten Black Shale members and Mosasaurus and Prognathodon first occur within this interval (Fig. 1.7). In addition, the evolutionary change of Platecarpus to Plioplatecarpus occurred during this interval. Globidens makes a brief appearance during the Niobraran “Age” (Russell, 1988), and then is relatively unknown until much later in the Navesinkian (Martin, 1996). Rare specimens of Globidens jaw fragments are also present in the Upper Sharon Springs Member in Kansas (Everhart and Everhart, 1996). The faunal turnover was not restricted to the marine realm, terrestrial faunas were also strongly affected. Cretaceous Land Mammal Ages were defined by (Fig.1.7, Russell, 1975), who attempted to correlate these ages to the European stages and appears to have considered the existence of the organisms to be restricted to the physical limits of particular rock units (Lillegraven and Ostresh, 1990). Lillegraven and Ostresh (1990) modified and more clearly defined the Cretaceous North American land mammal ages, using the terms introduced by Russell (1975). More recent research shows that these land mammal “ages” may be restricted to only the northern part of the seaway (Weil, 1999). The Aquilan fauna is Early Campanian in age (Scaphites hippocrepis III to Baculites asperformis ammonite range zones (Fig.1.7, Lillegraven and McKenna, 1986)) and is the 18 earliest defined and accepted “age” term (Lillegraven and McKenna, 1986). This turnover appears to occur slightly later than the marine faunal turnover, equivalent to the top of the Baculites asperformis ammonite range zone (Lillegraven and Ostresh, 1990). The Aquilan represents the last occurrence of many of the more primitive animals and a radiation of the more advanced Late Cretaceous mammals. The triconodonts and symmetrodonts make their last appearance in the Aquilan (Russell, 1975). The multituberculates exhibit a radiation in the Aquilan, becoming more advanced and many of the multituberculates of the Aquilan persist into the Late Cretaceous (Russell, 1975). Although marsupialsappear prior to the Aquilan, they began their radiation during this time, with five genera (including the first didelphoids, Eodelphis cutleri (Lillegraven and McKenna, 1986)) recognized in the Aquilan, as opposed to one from older strata (Russell, 1975). Eutherians made their first appearance in the Early Cretaceous (two genera), but were much more diversified in the Aquilian (Russell, 1975). At the species level, all the known fauna of the Aquilan are unique, with the possible exception of one species that may continue into the Judithian (Lillegraven and McKenna, 1986). The Judithian fauna differs from the Aquilian in the loss of the triconodonts and symmetrodonts, it is intermediate in the multituberculates and marsupials, and the eutherians are more advanced. The Judithian has only first appearances at the species level and no fauna are restricted to this time (Russell, 1975). While the faunas exhibited an overall turnover during the deposition of the lower Pierre Shale, both marine faunas (unpublished data, Bell, Everhart, Bertog) as well as terrestrial faunas (Weil, 1999) suggest that a major geographic or ecologic boundary exists throughout the western interior seaway at approximately the position of the transcontinental arch. This boundary is also exhibited in the stratigraphy and the sea level fluctuations of this time (Lillegraven and Ostresh, 1990). 19 Historical descriptions of the stratigraphy of the lower Pierre Shale

THE PIERRE SHALE The Pierre Shale is one of the most widespread formations in the western interior. The Pierre Shale was first described near Fort Pierre, South Dakota along the Missouri River for exposures of dark gray shale (Hayden, 1862). The Pierre Shale exhibits vertical variation and has been divided into several members. The facies also vary laterally across the region and most members are local in extent. The Pierre Shale can be divided into two parts, particularly in the northern part of the basin, where two 3rd order cycles are clearly recorded (Kauffman, 1984). The division between the upper and lower Pierre Shale is marked by the regressive facies of the Redbird Silty Member and the Gregory Member, distal equivalents to the Judith River Formation. The lower Pierre Shale below these members is equivalent to the Claggett Shale in Montana and the upper part is equivalent to the Bearpaw Shale (Dyman et al., 1994). The first subdivision of the Pierre Shale was in northeastern Colorado (Mather et al., 1928). Two years later, in the northern Black Hills, the Pierre Shale was divided into the Gammon Ferruginous Member, including the Pedro Bentonite and the Groat Sandstone; the Mitten Black Shale Member, the Monument Hill Bentonitic Member; and an upper unnamed fissile shale and mudstone (Rubey, 1930). In the western Black Hills, the original members of Rubey (1930) were recognized, but a second unnamed member was added between the Mitten Black Shale and the Monument Hills bentonitic member and the Kara Bentonitic Member was named for sediments below the upper unnamed shale member (Robinson et al., 1959). The lower two unnamed shale members were later named the Redbird Silty Member (Gill and Cobban, 1962). In a review of the Pierre Shale, the Sharon Springs Member, defined in Kansas (Elias, 1931) was also recognized in the Black Hills. Currently, the accepted members of the Lower Pierre Shale in the Black Hills region include the Gammon Ferruginous Member, the “Sharon Springs Member”, the Mitten Black Shale 20 Member and the Redbird Silty Member (Figure 4.2). As shown in this research, the so- called “Sharon Springs Member” of the Black Hills is not equivalent to the Sharon Springs Member at the type locality in Kansas and is therefore referred to the new Walhalla Member (Figure 4.3). Along the Missouri River, the Pierre Shale was divided into five members (Searight, 1937), with the lower part of the Pierre Shale referred to the Gregory Member. The Gregory Member was later refined to include the Sharon Springs Member as the non-calcareous bituminous shale and the Gregory Member as the calcareous shale (Gries and Rothrock, 1941; Moxon et al., 1939; Moxon et al., 1938). The Sully Member, as described by Searight (1937) has been revised to include the Crow Creek Member and the DeGrey Member (Gries and Rothrock, 1941), while the Verendrye, Virgin Creek and the Elk Butte members have been retained from their origin description by Searight (1937). Currently, the accepted members of the Lower Pierre Shale along the Missouri River include the Sharon Springs Member and the Gregory Member (Figure 4.2). In this paper, the lowest part of the “Sharon Springs Member”, shown not to correlate with the type Sharon Springs Member of Kansas, is referred to the Walhalla Member (Figure 4.3). In Kansas, the Pierre Shale was divided into the Sharon Springs Member in the lower part of the Pierre Shale, the Weskan Shale Member, the Lake Creek Shale Member, and the Salt Grass Shale Member, an unnamed shale member and the Beecher Island Member (Elias, 1931). Gill et al. (1962) further described the Sharon Springs, but the terminology remains unchanged. Currently, the term Sharon Springs Member is used for the entire lower Pierre Shale in Kansas (Figure 4.2). However, as shown in this paper, the lower part of the Sharon Springs Member is equivalent to the Gammon Ferruginous Member, and the Sharon Springs Member is redefined to exculde this unit (Figure 4.3). In Colorado, the subdivisions of the Pierre Shale are highly variable due to the increased sandy sedimentation and proximity to the western margin of the seaway. In the northeastern part of the state, the members follow that described in the southern part of the 21 Black Hills (Gill et al., 1975). In northwestern Colorado, several sandstone members are interbedded with unnamed shale units. The only shale unit that was recognized in this area is the “Sharon Springs Member”, and it was restricted to the most-organic rich interval of shale with distinct buttress-weathering outcrops and as many as eight thin bentonites (Izett et al., 1971). In southeastern Colorado, again only the sandstones and the Sharon Springs Member have been recognized, with large intervals of unnamed shale between these units (Scott, 1969; Scott and Cobban, 1959).

Figure 1.8a: Generalized stratigraphy of the lower Pierre Shale, as it is currently recognized in the literature.

Figure 1.8b: Stratigraphy of the lower Pierre Shale, as modified in this paper. The Sharon Springs Member, Gammon Ferruginous Member and Pembina Member are redefined to include the Walhalla Member and teh phosphate nodule and gray shale facies of the Chamberlain Member, two new members. 22 THE SHARON SPRINGS MEMBER Kansas

The Sharon Springs Member was first described for exposures of organic-rich black shale in northwestern Kansas (Elias, 1931). Similar facies were recognized in Colorado and in the Black Hills of South Dakota and Wyoming (Elias, 1931). The member was divided into three units, a lower dark soft shale unit, a middle organic-rich shale unit and an upper phosphatic unit (Fig. 1.8, Gill et al., 1972). The lower dark soft shale unit consists of light to medium-gray, light-olive-gray, and black shale and contains some limestone concretions (Gill et al., 1972). Certain layers in the shale are altered to dark yellowish orange, causing the upper part of the exposure to appear banded. Gill and Cobban (1972) suggested that this was a result of pre-Tertiary oxidation. The shale is soft and weathers easily to a gentle slope, making outcrops uncommon. Thin bentonite beds are present near the base of the unit, but they are not laterally extensive in western Kansas. The top of the unit is marked by an iron-stained interval approximately 0.25m thick. The lower dark shale unit is within the Baculites sp. (weakly ribbed) ammonite range zone. Baculites obtusus are found at the top of the unit, within the dark yellowish orange and iron-stained intervals (Gill et al., 1972). The organic-rich unit is characterized by dark brownish-black organic-rich shale with abundant layers of closely spaced limestone concretions. The shale contains abundant fish fragments, particularly scales and teeth, but other bones are also recognized. The organic material may also contain marine and terrestrial plant material (Gill et al., 1972). This unit is harder than the lower unit and forms vertical cliffs in weathering. It contains on average 4.9% organic carbon, and up to 9.6% in some beds near the top (Gill et al., 1972). The unit contains abundant jarosite, limonite and gypsum and these minerals typically coat joints and fossils. The upper part of the bituminous unit contains numerous layers of limestone concretions, some of which are septarian. One layer of large septarian concretions contains abundant invertebrate fossils. Thin bentonite beds are present in the organic-rich 23 unit, but most are thin and not laterally persistent. One 3cm thick bentonite about 10m above the contact of the dark soft shale and the organic-rich shale is present at every locality. Fish and large marine vertebrates are common in the organic-rich shale but the nature of the shale prevents preservation of invertebrates except in the concretions. Although only Baculites asperformis has been recognized in the concretions near the top of the organic- rich shale unit, it is presumed that the deposition occurred during the Baculites mclearni and Baculites asperformis ammonite range zones (Gill et al., 1972). The phosphatic shale unit at the top of the material presently assigned to the Sharon Springs Member in Kansas consists of dark brownish black to grayish brown shale with numerous layers of phosphate nodules and several thin bentonites (Gill et al., 1972). The phosphate nodules are concentrically laminated internally and typically have a rind of gypsum filled with shale (Elias, 1931). The nodules are dark brown to brownish black when fresh and alter to bluish gray. Above the phosphatic unit are several layers of limestone lenses as much as 1-foot thick and 50 feet across. Numerous bentonite beds are present, one of which is at least 1- foot thick (Elias, 1931). This interval was included in the Sharon Springs Member by Elias (1930) but was moved to the Weskan Shale Member by Gill and Cobban (1972). In this paper, the lower dark soft shale unit is moved to the Gammon Member, restricting the Sharon Springs Member to the organic-rich unit and the phosphate unit. This is based on the similarity in lithology between the Gammon Member in the Black Hills and this unit in western Kansas. Missouri River, South Dakota

The Sharon Springs Member was correlated to the Missouri River in South Dakota, where the lower part of the Gregory Member (Searight, 1937) was recognized as having a similar lithology to the Sharon Springs Member of Kansas (Moxon et al., 1939; Moxon et al., 1938) and was divided into two units (Fig. 1.8). The “Sharon Springs Member” was identified as all beds above the marl of the Niobrara Formation and below the marl of the 24 Gregory Member. The lower part of the “Sharon Springs Member” along the Missouri River is called the Fish Scale Zone (Gries and Rothrock, 1941). This unit consists of dark gray, fissile, bituminous shale with an abundance of fish scales and other bones of fish. This unit overlies the Niobrara Formation unconformably and the unconformity is marked by a bed of rusty-colored selenite varying in thickness from less than 2 cm to greater than 0.25m (Gries and Rothrock, 1941). The top of the unit is marked by a layer of white concretions. Numerous bentonite beds are common in this unit and may be up to 0.6m thick. The bentonitic unit consists of much more organic-rich, “earthy” shale than the overlying more fissile shale of this unit (Gries and Rothrock, 1941). The upper shale zone is described by Gries and Rothrock (1941) as a thin bed of soft, bluish-gray shale devoid of fish remains with occasional specks of red hematite, the presence of numerous, very fine, tubelike holes and no concretions. This unit is less than 0.5 m thick. In this paper, the bentonitic interval is moved to the Walhalla Member. This newly recognized member is restricted to the northern part of the basin. The upper shale zone is referred to the Sharon Springs Member and a previously undescribed phosphate unit is referred to the Chamberlain Member. Black Hills, South Dakota, Wyoming and Montana

The “Sharon Springs Member” is recognized in the southern Black Hills (Fig. 1.8). When Elias (1931) described the Sharon Springs in Kansas, he recognized similar lithology in the southern Black Hills. The term Sharon Springs Member was formally assigned to the organic-rich interval of the Pierre Shale in the Black Hills in 1939 (Moxon et al., 1939). In this description, beds equivalent to the Gammon Ferruginous Member, the Mitten Black Shale Member (Rubey, 1930) and the Sharon Springs Member, as it is currently recognized (Gill and Cobban, 1966), were included in the Sharon Springs Member. In the revised description of the Sharon Springs Member (Gill and Cobban, 1966), the base of the Sharon Springs Member in the Black Hills is marked by the Ardmore bentonite succession, described near the town of Ardmore, South Dakota (Spivey, 1940). The lower part of the “Sharon 25 Springs Member” in the Black Hills as it has been recognized consists of highly organic- rich black, siliceous shale that is very hard and weathers in thick blocks. It is inter-bedded with the beds of the Ardmore bentonite succession. The upper part of the “Sharon Springs Member” consists of dark-gray to grayish-black hackly, soft shale with thin bentonites. Concretions are common in the Sharon Springs Member in the eastern Black Hills but are mostly absent at Redbird, Wyoming in the western Black Hills. Specimens of Baculites obtusus are preserved in concretions within the Ardmore succession. Baculites are present in the “Sharon Springs Member” outside the concretions but they are not preserved well enough to be identified. Vertebrate fossils are common in the “Sharon Springs Member” of the Black Hills. In this paper, the Sharon Springs Member in the Black Hills is moved to the Walhalla Member. The Gammon Member and the Mitten Black Shale members are retained. Colorado

In northeastern Colorado, the stratigraphy of the lower Pierre Shale is described as being similar to the Black Hills. The “Sharon Springs Member” as previously recognized is marked at the base by the Ardmore bentonite succession (Gill et al., 1975), herein referred to the Walhalla Member. However, throughout most of Colorado, the lower part of the Pierre Shale has several silty layers due to the proximity to the sediment source to the east. The Sharon Springs Member in this area is confined to only the most organic-rich, buttress- weathering shale, generally within the Baculites obtusus ammonite range zone (Fig. 1.8). This interval is recognized in eastern Colorado based on lithology, stratigraphic position and fossil content (Izett et al., 1971). In this area, the member has as many as 8 thin bentonite beds in the lower part of the member and more organic-rich shale than the surrounding shale. The member in the Kremmling area in eastern South Dakota is divided into two units. The lower part consists of the thin bentonite beds and resistant organic-rich shale. The upper unit is consistent with the shale of the type locality in Kansas, dark brownish- black buttress-weathering shale with abundant fish remains. 26 THE PEMBINA MEMBER The Pembina Member was first described in Manitoba as the uppermost unit of the Vermillion River Formation. This member was named for exposures of noncalcareous shale overlying the calcareous Boyne Member of the Vermillion River Formation on Pembina Mountain and along the Pembina River in southern Manitoba (Kirk, 1930). The type description on Pembina Mountain indicates no more than 27 meters of dark shale. North of the type locality, along the Pembina River, the lower 3.5 meters of the member was indicated as having at least 6 bentonite beds (Kirk, 1930). This lower shale was black, contrasting with the upper part of the member, which was black and chocolate brown fissile shale with abundant fish material. The same area was described in further detail in 1945 (Wickenden, 1945) and 1948 (Tovell, 1948). Wickenden (1945) measured slightly greater than 27 meters of the member, with the lower 6 meters consisting of black shale with 18 layers of bentonite. Tovell (1948) described the two-fold division of the member (Fig. 1.8). The lithology of the Pembina Member of the Vermilion River Formation is consistent with the temporally equivalent Sharon Springs Member of the Pierre Shale in the United States, as it has been previously described. In North Dakota, strata of lithology resembling the Pembina Member of Manitoba and the Sharon Springs Member of Kansas and South Dakota were assigned to the Pembina Member, but were placed in the Pierre Shale rather than the Vermillion River Formation (Gill and Cobban, 1965). The term Pembina Member was chosen over the Sharon Springs Member due to the proximity of the exposures in northeastern North Dakota to the type locality of the Pembina Member in Manitoba. In fact, the exposures near Walhalla in North Dakota are along the same escarpment as the type locality of the Pembina Member and are less than 30 miles from the type section. However, Gill and Cobban (1965) chose to include the Pembina Member of North Dakota in the Pierre Shale, the United States terminology, and retain the Niobrara Formation below the Pierre Shale, as opposed to using the formation names applied in Manitoba (Gill and Cobban, 1965). 27 The Pembina Member of North Dakota can be divided into three parts. The lower part is approximately 2-3 meters of black, carbonaceous clay-shale alternating with white- yellow bentonites (Leonard, 1904). The lower part of the member has as many as 18 bentonites. The lower unit overlies the Niobrara Formation unconformably. The top of the Niobrara Formation has extensive weathering, possibly indicating karst formation prior to the deposition of the Pierre Shale. A second unconformity also is present within the lower unit, indicated by a hard iron-stained brecciated shale. The lower unit is discontinuous, and in some areas the middle unit rests directly on the Niobrara Formation unconformably (Gill and Cobban, 1965). The middle part is 6-20 meters of black and highly carbonaceous shale cut by joints filled with jarosite and gypsum. Bentonites are rare in the upper part of the member, but thin bentonites are present. The middle unit unconformably overlies the lower unit, as indicated by an iron-stained shale breccia between the two units. Along the Sheyenne River in North Dakota, the lower unit is missing and the middle unit is only 6 meters thick (Gill and Cobban, 1965). At the top of the middle unit in this area, a few siderite concretions are present 5-6 meters above the base. Two thin bentonites are present at the base of the unit in this area. Poorly preserved specimens of Baculites perplexus were recorded within the middle unit. Overlying the middle unit is a third unit of grayish-brown to brownish black shale with gypsum encrusted phosphate nodules (Gill and Cobban, 1965). In this paper, the bentonitic unit is referred to the Walhalla Member, a newly recognized member that is present in the northern part of the basin. The middle organic- rich shale and the upper phosphatic unit are referred to the Sharon Springs Member as they exhibit a pattern identical to the Sharon Springs Member in Kansas. THE GAMMON FERRUGINOUS MEMBER The Gammon Ferruginous Member was named for exposures of dark-gray mudstone and shale containing abundant siderite concretions in the northern Black Hills of Wyoming one year before the Sharon Springs Member was defined in Kansas (Fig. 1.8, Rubey, 1930). In the northern Black Hills, the Gammon Ferruginous Member is 265 meters thick and the 28 top 17 to 50-meters consists of sandstone, sandy shale and siltstone named the Groat Sandstone Beds (Rubey, 1930). When the Sharon Springs Member was described, it was noted that equivalent strata were present in the lower part of the Pierre Shale in the Black Hills (Elias, 1931) and the Gammon Ferruginous Member was included in the “Sharon Springs” in the southern part of the Black Hills. At Redbird, Wyoming in the southern Black Hills, Gill and Cobban (1966) later recognized the Gammon Ferruginous Member at the base of the Pierre Shale and separated the Sharon Springs Member into the Gammon Ferruginous Member, the “Sharon Springs Member” and the Mitten Black Shale Member. The term “Sharon Springs Member” was then restricted to only the most organic-rich interval, herein referred to the Walhalla Member, while the Gammon Ferruginous Member is a black shale with abundant siderite concretions. The siderite concretions are not present in the Sharon Springs Member at Redbird, Wyoming. Based on this revised correlation of the strata, Gill and Cobban (1966) recognized that the member thinned from north to south, partially by inter-fingering with the Niobrara Formation below and partially due to truncation by the Sharon Springs Member above. In the Black Hills area, the Gammon Ferruginous Member conformably overlies the Niobrara formation and is recognized by a shift from calcareous shale to non- calcareous shale. The Gammon Ferruginous Member was recognized in northern Colorado and northwestern Nebraska where it is thin below the Sharon Springs Member (Gill et al., 1975). It was also recognized as a discontinuous member below the Pembina Member in North Dakota (Gill and Cobban, 1965). In this paper, the Gammon Ferruginous Member is shortened to the Gammon Member, removing the implication of an iron-rich member and the member is extended into Kansas, where a similar lithology is present at the base of what has traditionally been called the Sharon Springs Member. 29 THE MITTEN BLACK SHALE MEMBER The Mitten Black Shale Member was named along with the Gammon Ferruginous Member in the northern Black Hills of Wyoming along Driscoll Creek (Fig. 1.8, Rubey, 1930). The member is recognized only in the Black Hills. It has been fully described by Rubey (1930) in the northern Black Hills and by Gill and Cobban (1966) at Redbird, Wyoming. The member ranges from 280m at Redbird, Wyoming to 44m at the type locality in the northwestern Black Hills and 24m in the southeastern Black Hills. Four units can be recognized in the Mitten Black Shale Member. The thickness variations are partially due to thinning of the upper and lower units and pinching out of the middle units to the north and pinching out of all but the upper unit to the east. The lower unit of the black fissile shale facies consists of 24-64m of moderately hard dark-gray shale tht weathers brown. In the northern Black Hills, the base contains a layer of phosphatized pebbles and rounded bone fragments, resting unconformably on the Gammon Ferruginous Member (Gill and Cobban, 1966). In the southwestern Black Hills, at Redbird, Wyoming, the unit contains many layers of siderite and septarian limestone concretions. The unit is not present to the east. This unit is within the Baculites asperformis ammonite range zone. The second unit consists of 60m of dark-gray to black, soft, flaky shale, lacking concretions. 75cm of bentonitic shale is present 20m below the top of this unit. This unit is only recognized at Redbird, Wyoming. This unit is within the Baculites sp (smooth). Ammonite range zone (Gill and Cobban, 1966). The third unit is a brownish-gray to gray bentonitic shale that weathers to grayish- brown and lacks vegetation. The top of this unit is marked by a 75cm thick bentonite. Abundant siderite concretionary layers and a few thick limestone concretions are present in this unit. This unit is only recognized at Redbird, Wyoming. This unit is within the Baculites perplexus (early form) ammonite range zone (Gill and Cobban, 1966). The fourth unit consists of black soft flaky shale with numerous rusty limestone 30 concretions in the upper part and siderite concretionary horizons in the lower part. In the northern Black Hills, this unit is 20m thick and has a thin layer of phosphate nodules at the base (Rubey, 1930). At Redbird, the unit is 105m thick and in southeastern South Dakota the unit is 24m thick (Gill and Cobban, 1966). The unit is within the Baculites perplexus (early form) to Baculites gilberti ammonite range zone. THE REDBIRD SILTY MEMBER The Redbird Silty Member was described at Redbird, Wyoming in the southern Black Hills for light- to medium-gray soft silty shale approximately 200 meters thick (Fig. 1.8, Gill and Cobban, 1962). The member is considered to overly the Mitten Black Shale Member conformably in the southern Black Hills. The member is conformably overlain by dark-weathering unnamed shale. Silty limestone concretions are common in the member, sometimes forming almost a solid bed of limestone. The member is represented in a north-south trending band through east-central Colorado, eastern Wyoming, western South Dakota, eastern Montana and western North Dakota. Further to the west, the member is equivalent to the Hygiene Sandstone Member of the Pierre Shale in Colorado, the Parkman Sandstone Member of the Mesaverde Formation in Wyoming, and the Judith River Formation in Montana (Gill and Cobban, 1966), where increased sedimentation increased the sand content. To the east, the unit grades laterally into the calcareous Gregory Member of the Pierre Shale in eastern South Dakota and North Dakota. In Kansas, the time interval represented by the Redbird Silty Member is within the Weskan Shale Member of the Pierre Shale (Gill and Cobban, 1965). This interval is indicated by the Baculites perplexus through the Baculites scotti ammonite range zones (Gill and Cobban, 1966). THE GREGORY MEMBER The Gregory Member was described for outcrops along the Missouri River in Gregory County, South Dakota (Fig. 1.8, Searight, 1937). In the type description, the member included all beds from the top of the Niobrara Formation to the top of the chalky shale in the lower 31 part of the Pierre Shale. This description includes strata now recognized as the Sharon Springs Member and the Gregory Member (Gries and Rothrock, 1941). The member was divided into two units, the lower and upper Gregory in the type description. The Lower Gregory, which included the highly bituminous black shale with bentonites, was later considered the Sharon Springs Member (Moxon et al., 1938; Moxon et al., 1939; Gries and Rothrock, 1941). The upper Gregory is a thin succession of chalky shale or marl that is characteristically distinct from the lower Gregory and was not made a separate unit by Searight (1937) only because it was too thin. This classification was altered when the Sharon Springs terminology was applied to the bituminous shale that was the lower Sharon Springs, leaving the Gregory Member as only the marl unit (Moxon et al., 1939; Moxon et al., 1938). Gries and Rothrock (1941) modified the Gregory Member to include all beds between the base of the Gregory marl and the base of the upper calcareous zone. They noted that the Gregory Member described by Moxon et al. (1939) included a unit at the base of the Sully Member (now the Crow Creek Member) of sand and marl that are distinct from the lower part of the Upper Gregory Member. The Gregory Member, as described by Gries and Rothrock (1941), can be divided into two units. The lower unit is a chalk zone, observed only at the type locality. This interval consists of about 2.5 meters of impure, light gray chalk with small shale pebbles and sand grains. To the north, the lower part of the Gregory Member is marked by an intermittent zone of large limestone concretions, near the White River, and further north towards Chamberlain, this zone is absent. The upper unit is a shale zone and is varied in color from light buff to dark gray with a typical banding of the outcrop, characteristic of interbedded non-calcareous and calcareous shale with brown ironstone concretions and light gray calcareous concretions. REVISED STRATIGRAPHY OF THE LOWER PIERRE SHALE Based on high-resolution stratigraphic correlations, the sediments of the lower Pierre Shale have been revised to reflect the more complex facies distribution exhibited by the 32 strata (Fig. 1.8, Chapter 4). The Gammon Ferruginous Member, previously only recognized in the Black Hills region, has been extended to western Kansas, where sediments of the lower “Sharon Springs Member” exhibit similar facies patterns. The “Sharon Springs Member” of the lower Pierre Shale is now recognized as a composite member, and has been herein revised to reflect the different facies patterns recognized. The newly defined Walhalla Member includes strata previously referred to the “Sharon Springs Member” in the Black Hills and eastern South Dakota and the Pembina Member in eastern North Dakota. This member is restricted to the northern part of the basin and is absent in western Kansas. It reflects deposition resulting from a tectonic cycle associated with the Absoroka Thrust belt in Wyoming. Above the Walhalla Member, the strata indicate north-south trending facies belts. In the east, the restricted Sharon Springs Member and the newly defined Chamberlain Member, which include sediments previously referred to as the “Sharon Springs Member” in western Kansas and eastern South Dakota and the “Pembina Member” in North Dakota are recognized. The Mitten Black Shale Member is recgognized in the Black Hills. The term Pembina Member is herein discarded in North Dakota.

BIBLIOGRAPHY Anna, L.O., 1986. Structural influences on Cretaceous sedimentation, northern Great Plains. In: J.A. Peterson (Editor), Paleotectonics and sedimentation in the Rocky Mountain Region, United States. AAPG Memoir, pp. 173-193. Armstrong, R.L. and Ward, P., 1993. Late Triassic to Earliest Eocene Magmatism in the North American cordillera: Implications for the Western Interior basin. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 49-72. Bilodeau, W.L. and Lindberg, F.A., 1983. Early Cretaceous tectonics and sedimentation in southern Arizona, southwestern New Mexico, and northern Sonora, Mexico. In: M.W. Reynolds and E.D. Dolly (Editors), Mesozoic paleogeography of west-central 33 United States. Rocky Mountain Section SEPM paleogeography symposium, pp. 173- 188. Bohor, B.F. and Triplehorn, D.M., 1993. Tonsteins: Altered volcanic-ash layers in coal- bearing sequences. Geological Society of America Special Paper, 285: 44. Caldwell, W.G.E. et al., 1993. Foraminiferal biostratigraphy of Cretaceous marine cyclothems. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada, pp. 477-520. Christensen, E.H., Kowallis, B.J. and Barton, M.D., 1994. Temporal and spatial distribution of volcanic ash in Mesozoic sedimentary rocks of the Western Interior: An alternative record of Mesozoic magmatism. In: M.V. Caputo, J.A. Peterson and K.J. Franczyk (Editors), Mesozoic systems of the Rocky Mountain region, USA. Rocky Mountain section SEPM, Denver, pp. 73-94. Cobban, W.A., 1993. Diversity and distribution of Late Cretaceous ammonites, Western Interior, United States. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada, pp. 435-452. Cross, T.A. and Pilger Jr., R.H., 1978. Tectonic controls of late Cretaceous sedimentation, western interior, USA. Nature, 274: 653-657. DeCelles, P.G., 1994. Late Cretaceous-Paleocene synorogenic sedimentation and kinematic history of the Sevier thrust belt, northeast Utah and southwest Wyoming. Geological Society of America Bulletin: 32-56. Defant, M.J. and Drummond, M.S., 1993. Mt. St. Helens: Potential example of the partial melting of the subducted lithosphere in a volcanic arc. Geology, 21: 547-550. Dickenson, W.R., 1981. Plate tectonic evolution of the southern Cordillera. In: W.R. Dickenson and W.D. Payne (Editors), Relations of tectonics to ore deposits in the wouthern Cordillera. Arizona Geological Society Digest, pp. 113-135. Dyman, T.S. et al., 1994. Stratigraphic transects for Cretaceous rocks, Rocky Mountains and Great Plains regions. In: M.V. Caputo, J.A. Peterson and K.J. Franczyk (Editors), 34 Mesozoic systems of the Rocky Mountain region, USA. Rocky Mountain Section SEPM, Denver, pp. 365-392. Elias, M.K., 1931. The Geology of Wallace County Kansas. Kansas Geological Survey Bulletin 18: 254. Everhart, M.J. and Everhart, P.A., 1996. First report of the shell crushing mosasaur, Globidens sp., from the Sharon Springs Member of the Pierre Shale (upper Cretaceous) of western Kansas. Kansas Academy of Science Transactions, 15: 17. Gill, J.R. and Cobban, W.A., 1962. Red Bird Silty Member of the Pierre Shale, a new stratigraphic unit. U. S. Geological Survey Professional Paper, 450-B: 21-24. Gill, J.R. and Cobban, W.A., 1965. Stratigraphy of the Pierre Shale, Valley City and Pembina Mountain areas, North Dakota. U. S. Geological Survey Professional Paper 392-A: A1-20. Gill, J.R. and Cobban, W.A., 1966. The Redbird section of the Upper Cretaceous Pierre Shale in Wyoming. U. S. Geological Survey Professional Paper 393-A: A-1-A73. Gill, J.R., Cobban, W.A. and Schultz, L.G., 1972. Stratigraphy and composition of the Sharon Springs Member of the Pierre Shale in Western Kansas. U. S. Geological Survey Professional Paper, 728: 50. Gill, J.R., Cobban, W.A., Scott, G.R. and Burkholder, R.E., 1975. Unedited stratigraphic sections of the Pierre Shale near Roundbutte and Buckeye in Larimer County, Northern Colorado. U. S. Geological Survey Open-File Report, 75-129, Denver, 12 pp. Gries, J.P. and Rothrock, E.P., 1941. Manganese deposits of the Lower Missouri Valley in South Dakota. U. s. Geological Survey Report of Investigations, 38, Vermillion, 96 pp. Haq, B.U., Hardenbol, J. and Vail, P.R., 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level change. In: C.K. Wilgus et al. (Editors), Sea-level changes: An integrated approach. SEPM Special Publication, pp. 71-108. 35 Hay, W.W., Eicher, D.L. and Diner, R., 1993. Physical oceanography and water masses in the Cretaceous Western Interior Seaway. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada Special Paper, pp. 297-318. Hayden, F.V., 1862. On the geology and natural history of the Upper Missouri. American Philosophical Society Transactions, 12(1): 218. Hyndman, D.W., 1983. The Idaho Batholith and associated plutons, Idaho and western Montana. Geological Society of America Memoir, 159: 213-240. Hyndman, D.W., Talbot, J.L. and Chase, R.B., 1975. Boulder batholith: A result of emplacement of a block detached from the Idaho batholith infrastructure? Geology, 3: 401-404. Izett, G.A., Cobban, W.A. and Gill, J.R., 1971. The Pierre Shale near Kremmling, Colorado and its correlation to the east and the west. U. S. Geological Survey Professional Paper 684-A: A1-A19. Kauffman, E.G., 1977. Geologic and biologic overview: Western Interior Cretaceous Basin. Mountain Geologist, 14: 75-99. Kauffman, E.G., 1984. Paleobiogeography and evolutionary response dynamic in the Cretaceous Western Interior seaway of North America. In: G.E.G. Westerman (Editor), Jurassic-Cretaceous Biochronology and paleogeography of North America. Geological Association of Canada Special Paper, pp. 273-306. Kauffman, E.G. and Caldwell, W.G.E., 1993. The Western Interior basin in space and time. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 1-30. Kauffman, E.G. et al., 1993. Molluscan biostratigraphy of the Cretaceous western interior basin, North America. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada, pp. 397-434. Kirk, S.R., 1930. Cretaceous stratigraphy of the Manitoba escarpment. Canadian Geological 36 Survey Summary Report 1929B: 112B-135B. Leonard, A.G., 1904. Topographic features and geological formations of North Dakota. North Dakota Geological Survey, 3d Bienn. Rept: 127-177. Lillegraven, J.A. and McKenna, M.C., 1986. Fossil mammals from the “Mesaverde” Formation (Late Cretaceous, Judithian) of the Bighorn and Wind River Basins, Wyoming, with definitions of Late Cretaceous North American Land-Mammal “Ages”. American Museum Novitates, 2840: -. Lillegraven, J.A. and Ostresh, L.M., Jr., 1990. Late Cretaceous (earliest Campanian, Maastrichtian) evolution of western shorelines of the North American Western Interior seaway in relation to known mammalian faunas. Geological society of America special paper, 243: 30. Lipps, J.H., Berger, W.H., Buzas, M.A., Douglas, R.G. and Ross, C.A., 1979. Foraminiferal ecology and paleoecology. SEPM Short Course, 6. Society of Economic Paleontologists and Minerologists, 198 pp. Martin, J.E., 1996. Biostratigraphic ranges and biozonation of mosasaurs (Reptilia) within the Late Cretaceous of the North American epicontinental seaway. Geological Society of America Rocky Mountain Section meeting abstracts with programs, 28(4): 16. Mather, K.F., Gilluly, J. and Lusk, R.G., 1928. Geology and oil and gas prospects of northeastern Colorado. U. S. Geological Survey Bulletin, 796-B: 86-92. McNeil, D.H. and Caldwell, W.G.E., 1981. Cretaceous rocks and their foraminifera inthe Manitoba escarpment. The Geological Association of Canada Special Paper, 21: 439. Mello, J.F., 1971. Foraminifera from the Pierre Shale (Upper Cretaceous) at Red Bird, Wyoming. U. S. Geological Survey Professional Paper 393-C: C1-C54. Miall, A.D., 1993. Collision-related foreland basins. In: L.E.A.S. Frostick, R. J. (Editor), Tectonic controls and signatures in sedimentary successions. Blackwell Scientific Publications, pp. 393-424. 37 Monger, J.W.H., 1993. Cretaceous tectonics of the North American cordillera. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 31-48. Moxon, A.L., Olson, O.E. and Searight, W.V., 1939. Selenium in rocks, soils, and plants. South Dakota Agr. Expt. Sta. Tech. Bull 2: 94. Moxon, A.L., Olson, O.E., Searight, W.V. and Sandals, K.M., 1938. The stratigraphic distribution of selenium in the Cretaceous Formation of South Dakota and the selenium content of some associated vegetation. American journal of botany, 25(10): 794-809. Obradovich, J.D., 1993. A Cretaceous Time Scale. In: W.G.E. Caldwell and E.G. Kaufman (Editors), Evolution of the Western Interior Basin. Geological Association of Canada, pp. 379-396. Parrish, J.T. and Gautier, D.L., 1993. Sharon Springs Member of the Pierre Shale: Upwelling in the Western Interior seaway? In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, Tulsa, pp. 319-332. Rice, D.D. and Shurr, G.W., 1983. Patterns of sedimentation and paleogeography across the western interior seaway during time of deposition of Upper Cretaceous Eagle Sandstone and equivalent rocks, northern Great Plains. In: M.W. Reynolds and E.D. Dolly (Editors), Mesozoic paleogeography of west-central United States. Rocky Mountain Section SEPM Paleogeography symposium, pp. 337-358. Robinson, C.S., Mapel, W.J. and Cobban, W.A., 1959. Pierre Shale along western and northern flanks of Black Hills, Wyoming and Montana. American Association of Petroleum Geologists Bulletin, 43(1): 101. Robinson, G.D., Klpeer, M.R. and Obradovich, J.D., 1968. Overlapping plutonism, volcanism and tectonism in the Boulder Batholith region, western Montana. Geological Society of America Memoir, 116: 557-576. 38 Rubey, W.W., 1930. Lithologic studies of fine-grained Upper Cretaceous sedimentary rocks of the Black Hills region. U. S. Geological Survey Professional Paper, 165-A: 1-54. Russell, D.A., 1988. A checklist of North American marine Cretaceous vertebrates including fresh water fishes. Occasional papers of the Tyrrell Museum of Palaeontology, 4. Russell, D.A., 1993. Vertebrates in the Cretaceous Western Interior seaway. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada, pp. 665-680. Russell, L.S., 1975. Mammalian faunal succession in the Cretaceous System of western North America. In: W.G.E. Caldwell (Editor), The Cretaceous system in the Western Interior of North America. Geological Association of Canada, Waterloo, Ontario, pp. 137-162. Scott, G.R., 1969. General and engineering geology of the northern part of Pueblo, Colorado. US Geological Survey Bulletin, 1262: 131. Scott, G.R. and Cobban, W.A., 1959. So-called Hygiene group of northeastern Colorado [and Wyoming], Rocky Mountain Association of Geologists, Symposium on Cretaceous rocks of Colorado and adjacent areas, 11th field conference, Washakie, Sand Wash and Piceance Basins, pp. 124-131. Searight, W.V., 1937. Lithologic stratigraphy of the Pierre Formation of the Missouri Valley in South Dakota. South Dakota State Geological Survey, Report of Investigations 27: 63. Shurr, G.W., 1984. Regional setting of Niobrara Formation in Northern Great Plains. American Association of Petroleum Geologists Bulletin, 68(5): 598-609. Shurr, G.W., Larson, B.S. and Watkins, I.W., 1989. Zuni sequence in Williston basin - Evidence for Mesozoic paleotectonism. American Association of Petroleum Geologists Bulletin, 73: 68-87. Shurr, G.W. and Rice, D.D., 1986. Paleotectonic controls on deposition of the Niobrara Formation, Eagle Sandstone, and equivalent rocks (Upper Cretaceous), Montana 39 and South Dakota. In: J.A. Peterson (Editor), Paleotectonics and sedimentation in the Rocky Mountain Region, United States. AAPG Memoir, pp. 193-212. Spivey, R.C., 1940. Bentonite in southwestern South Dakota. South Dakota Geological Survey Report of Investigations: 56. Stelck, C.R., 1975. Basement control of Cretaceous sand sequences in western Canada. In: W.G.E. Caldwell (Editor), The Cretaceous System in the western interior of North America. Geological Association of Canada Special Paper, pp. 427-440. Tovell, W.M., 1948. Geology of the Pembian Valley-Deadhorse Creek area, Winnipeg Division, Manitoba. Manitoba Department of Mines and Natural Resources, Mines Br. Preliminary Report 47-7: 7. Vikre, P.G. and McKee, E.H., 1985. Zoning and chronology of hydrothermal events in the Humboldt Range, Pershing County, Nevada. Isochron/West, 44: 17-24. Villien, A. and Kligfield, R.M., 1986. Thrusting and synorogenic sedimentation in central Utah. In: J.A. Peterson (Editor), Paleotectonics and sedimentation in the Rocky Mountain Region, United States. AAPG Memoir, pp. 281-307. Weil, A.I., 1999. Multituberculate phylogeny and mammalian biogeography in the Late Cretaceous and earliest Paleocene Western Interior of North America. PhD Thesis, University of California, Berkeley, 243 pp. Weimer, R.J., 1984. Relation of unconformities, tectonics, and sea-level changes, Cretaceous of Western Interior, USA. In: J.S. Schlee (Editor), Interregional unconformities and hydrocarbon accumulation. American Association of Petroleum Geologists Memoir. AAPG, pp. 7-35. Wherry, E.T., 1917. Clay from volcanic dust. Journal of the Washington Academy of Sciences: 567-583. Wickenden, R.T.D., 1945. Mesozoic stratigraphy of the eastern plains, Manitiboa and Saskatchewan. Canadian Geological Survey Memoir 239: 87. Witschko, D.V. and Dorr Jr., J.A., 1983. Timing and deformation in overthrust belt and 40 foreland of Idaho, Wyoming and Utah. American Association of Petroleum Geologists Bulletin, 67(8): 1304-1322. 41 Geochemical and mineralogical characterization of the Ardmore bentonite succession, Late Cretaceous of the Western Interior, and its use in regional stratigraphic correlation

ABSTRACT

The Ardmore bentonite succession is a thick succession of bentonites that has been extensively used based on stratigraphic position for regional correlations. Although the bentonites exhibit similar characteristics to each other in the outcrop, they differ significantly in their internal composition, indicating at least 15 volcanic eruptions and four different magmatic sources, two rhyolites and two andesites. Both andesites and one of the rhyolites contain abundant biotite phenocrysts that can be used to interpret the magmatic composition. The more aluminous andesite source is the Little Elkhorn Mountains volcanic complex in western Montana, based on correlation to bentonites preserved in the Two Medicine Formation in western Montana.

Although the Ardmore bentonite succession in the Sharon Springs Member has been reported as the initial transgression of the Claggett Cycle, high resolution correlation of the bentonites indicates that the initial transgression in the northern part of the basin, coincident with deposition of the Ardmore bentonite succession, occurred earlier than the eustatic sea level rise, and was a response to thrusting along the Absoroka Thrust belt. Subsidence in the axial basin and the Williston Basin resulted in initial transgression into these areas. A complete sequence is preserved in the lower part of the Ardmore bentonite succession, and an unconformity is present within the Ardmore succession marking a sequence boundary. The transgression of a second sequence is preserved in the upper part of the Ardmore bentonite succession. 42 Geochemical and mineralogical recognition of the Ardmore bentonite succession and its use in regional stratigraphic correlation

INTRODUCTION Thick bentonites in the Late Cretaceous western interior are commonly used for stratigraphic correlation. The criteria traditionally used for correlation include lithostratigraphy and biostratigraphic zonation, and radiometric dating. The value and reliability of these criteria varies as a function of both their degree of preservation and the regional variation in facies relationships. The use of stratigraphic position, for example, depends on the recognition of other regionally persistent lithologic units or event beds and surfaces whose relationships to key bentonite beds remain consistent. Biostratigraphic criteria are most useful when diagnostic taxa are abundant and persistent on a regional scale. In siliciclastic-dominated basins, however, lateral persistence of taxa or event beds is rare. Bentonites, on the other hand, offer internal criteria that can be used for correlations. Internal criteria of bentonites such as whole rock chemistry, composition characteristics of discrete phenocrysts, or population studies of primary minerals are all predicated on the assumptions of lateral homogeneity of these features. The validity of this assumption is supported by studies of contemporary explosively erupted pyroclastic flows and fallout ash beds. To the extent that bentonite correlation can be based on internal criteria, these beds serve as powerful tools for independently testing the presumed synchroneity of sequence boundaries and other separately established age equivalent units. The Middle Campanian Ardmore bentonite succession is a thick succession, characterized by a 1-meter thick bentonite at the base that has been extensively used for regional correlation (Fig. 2.1). On the basis of stratigraphic position, the Ardmore succession has been correlated from western the Claggett Shale in Montana to the Pierre Shale in eastern South Dakota (Dyman et al., 1994). Within the Pierre Shale, the Ardmore succession 43 nized stratigraphic position of the Ardmore succession and the ammonite range zones used for biostratigraphic correlation. the Ardmore Figure 2.1: Generalized stratigraphy of the Middle Campanian in western interior United States indicating recog Figure 44 is recognized as the base of the “Sharon Springs Member” in South Dakota (Gill and Cobban, 1966), although the Ardmore succession is not present at the type locality of the Sharon Springs Member in Kansas. However, outside the Pierre Shale, the correlation of the Ardmore succession has been more subjective. For example, the Ardmore succession was first correlated with the Claggett Shale in Wyoming (Fig. 2.1), based partly on biostratigraphic position combined with the presence of the thick bentonite succession. Below the Ardmore succession in the Pierre Shale, the Gammon Ferruginous Member contains Baculites sp. (weakly ribbed). This baculite is also present in the lower Claggett Shale. In the Claggett Shale, Baculites obtusus is present with the thick bentonites that have been correlated to the Ardmore succession. However, no baculites are present in the Ardmore succession in the Pierre Shale. The correlation to the Claggett Shale, then, is based on the occurrence of Baculites sp (weakly ribbed) below the thick bentonites in both the Claggett Shale and the Pierre Shale. The Ardmore succession has also been correlated with the terrestrial Two Medicine Formation in Montana (Fig. 2.1, Roberts, 2000). This correlation was based on radiometric dating of feldspars in the bentonites in Montana and Wyoming (Rogers et al., 1993). The succession of bentonites in Montana does not have the same stratigraphic appearance as the bentonites of the Ardmore succession in other parts of the basin, however. For example, no 1-meter thick bentonite is present in Montana and the bentonites are more widely spaced due to higher sedimentation rates. The Little Elkhorn Mountains in Montana was previously considered the source of the ash that composes the Ardmore bentonite succession (Gill and Cobban, 1966). If this were the case, one would expect to see thicker and more abundant layers in Montana than in the more distal parts of the basin. While the variations in bentonite preservation may be due in part to a terrestrial source with a higher clastic influx and more reworking, it also suggests that not all the bentonites of the Ardmore succession are actually present in Montana and therefore Montana is not the only source. This suggestion is supported through geochemical and mineralogical analysis of the bentonites, as shown below. Although all the bentonites of the Ardmore succession are 45 similar in physical appearance, their internal compositions vary significantly, indicating at least 15 volcanic eruptions from four different magmatic sources. Stratigraphically, in the Black Hills the Ardmore succession consists of a 1-meter thick bentonite near the base and as many as eight additional bentonites that are stratigraphically close to this thick bentonite (Fig. 2.2). This configuration changes significantly to the east and west where the bentonites

Figure 2.2: Generalized stratigraphy of the Ardmore succession. The Ardmore bentonite of Spivey (1940) was restricted to the 1-meter thick bentonite at the base of the succession. Gill and Cobban (1966) included all the thick bentonites at the base of the Sharon Springs with the Ardmore bentonite. 46

Figure 2.3: Map of the western interior of North America showing major geographic features for the region. Open circles represent localities of the Ardmore succession that were included in this study. 47 are split into multiple layers due to higher sedimentation rates. Due to the complex nature of the Ardmore succession, groups of bentonites within the Middle Cretaceous may be misidentified as the Ardmore succession and in some localities the Ardmore succession is incomplete, making it difficult to recognize. STRATIGRAPHY OF THE ARDMORE SUCCESSION Spivey (1940) first described the Ardmore bentonite as the approximately 1-meter thick bentonite at the base of the “Sharon Springs Member” of the Pierre Shale (Fig. 2.2) in southwestern South Dakota at Wasserburger Ranch near the town of Ardmore (Fig. 2.3). The 1-meter thick bentonite was mined in the southern Black Hills and this bed was the primary bentonite of interest. Spivey (1940) recognized this bentonite around the southern Black Hills (Fig. 2.4), from Wallace Ranch to Buffalo Gap (Fig. 2.3). However, other workers subsequently pointed out that several additional bentonites are present in close stratigraphic proximity to the Ardmore bentonite (Gill and Cobban, 1966). Five to eight

Figure 2.4: Ardmore succession at Wasserburger Ranch, South Dakota. The Ardmore Bentonite was originally described for the 1-meter thick bentonite at this locality. Bentonites are indicated by the white lines in the photograph. The Ardmore succession marks the base of the Sharon Springs Member in the Black Hills. 48 Figure 2.5: Ardmore succession at Redbird, Wyoming. When Gill and Cobban (1966) extended the Ardmore Bentonite to Redbird, they included all the thick bentonites above the 1-meter thick bentonite at the base of the Sharon Springs Member. These bentonites are now termed the Ardmore succession. Bentonites are grayish blue when fresh, but alter to grayish yellow, standing out in contrast to the black shale.

bentonites, averaging 15 cm thick, are present above the 1-meter thick basal bentonite (Fig. 2.2). Two thin bentonites, 3 cm thick each, are also present less than 5 cm above the highest of the 15cm thick bentonites (Fig. 2.2). A similar distribution of bentonites was described in the lower part of the Pierre Shale in the northern Black Hills of South Dakota, although the Sharon Springs Member is not recognized in that area (Wing, 1940). Gill and Cobban (1966) expanded the term Ardmore bentonite to Redbird, Wyoming (Fig. 2.3, 5), where they included all the thick bentonites in close proximity to a 1-meter thick bentonite (Fig. 2.2, 5). The Ardmore bentonit succession is now used extensively for stratigraphic correlation, based on its stratigraphic position. It is the only thick bentonite succession recognized in the Middle Campanian strata, and has been correlated from western Wyoming to eastern South Dakota (Dyman et al., 1994). To the west, the bentonites within the equivalent part of the succession become thicker and more widely spaced stratigraphically. In the Claggett Shale of Montana, three 49 thick bentonites, each about 1-meter thick are present at approximately the same position as the Ardmore bentonite (Fig. 2.1). These beds were called beds S, T and U (Knechtlel and Patterson, 1956), but were not correlated with the Ardmore succession when they were first described. They are now considered to be equivalent to the Ardmore succession, based on stratigraphic position (Dyman et al., 1994). The Ardmore bentonite (sensu stricto) was recognized in the subsurface of the Powder River Basin using spontaneous potential and resistivity wire-line logs (Fox, 1993a; 1993b; 1993c; 1993d). At the scale of most published wire-line logs, only the thick basal bentonite is recognizable in the subsurface. The Ardmore succession is also recognized in the terrestrial Two Medicine Formation in western Montana (Fig. 2.1, Roberts, 2000). At this locality, eight bentonites, ranging from 5 cm to 100 cm thick, are present in the Two Medicine Formation floodplain deposits. These bentonites have a mineral suite similar to that of the Ardmore succession (Roberts, 2000). To the east, several thin bentonites are present in the Pembina Member of the Pierre Shale in North Dakota that have been recognized as the Ardmore bentonite succession (Fig. 2.1, Gill and Cobban, 1965). The bentonites in North Dakota are all very thin compared to the bentonites of the Ardmore succession in South Dakota, with the thickest bentonite only 17cm thick. The Pembina Member of the Pierre Shale in North Dakota is equivalent to the Sharon Springs Member in South Dakota. The name Pembina Member has been used in North Dakota due to its proximity to the type locality in Manitoba. However, use of this terminology is confusing because the Pembina Member in Manitoba is part of the Vermillion Creek Formation, a formation that includes strata equivalent to the upper Niobrara Formation and lower Pierre Shale in the United States. In this paper, the “Sharon Springs Member” will be used in place of the Pembina Member for the strata in North Dakota that are identical to the “Sharon Springs Member” in South Dakota. It is recognized, based on correlation of the Ardmore succession and lithology, that the “Sharon Springs Member” in the northern part of the basin is not present at the type locality, and this member will be redefined in a later paper (Chapter 4). 50 To the south, the Ardmore succession is recognized in some parts of northern Colorado (Fig. 2.1, Gill et al., 1975), where the bentonites appear to have a stratigraphic pattern similar to that of the Black Hills. The Ardmore succession is only recognized in Colorado in areas where the other members of the Black Hills’ stratigraphic section are also recognized (Gill et al., 1975) and in the subsurface in Nebraska, using wire-line logs (DeGraw, 1975). The Ardmore succession is reported at only a single locality in western Kansas (Gill et al., 1972); however, this locality could not be relocated in this study. Ar:Ar radiometric dating of sanidine crystals from the thick bentonite of the Ardmore succession (the “Ardmore bentonite”, sensu stricto) at Redbird, Wyoming yielded an age of 80.04 +/- 0.4Ma (Hicks et al., 1999). The basal thick Ardmore bentonite was recognized in the Claggett Shale of the Elk Basin in Wyoming where it is present within the Baculites obtusus ammonite range zone. Here the bentonite yielded a date of 80.54 +/- 0.55Ma (Obradovich, 1993) using Ar:Ar radiometric dating on sanidine crystals. This date is used for the Baculites obtusus ammonite range zone, and the bentonite and ammonite range zone are considered to be coeval, although the ammonites are not found in association with the bentonite at the type locality in South Dakota. Previous accounts of the Ardmore bentonite succession have not provided detailed stratigraphic information about the individual bentonites in the succession, nor have they provided independent variables for identifying the Ardmore bentonite succession. Based on detailed stratigraphic correlation, distinct patterns can be recognized in the Ardmore succession. Around the southern Black Hills, the bentonites of the Ardmore succession exhibit nearly identical stratigraphic patterns to the type locality, with a 1-meter thick bentonite and eight additional bentonites, each approximately 15 cm thick. Below the Ardmore succession, the Gammon Ferruginous Member in the Black Hills also has several thick bentonites. At Redbird, Wyoming, the Gammon Ferruginous Member has five bentonites ranging from 3cm to 17cm thick. Although none of the bentonites are as thick as the basal bentonite of the Ardmore succession, they are closely spaced and superficially 51 resemble the Ardmore succession. In the Black Hills of South Dakota, the Gammon Ferruginous Member is much thicker and these bentonites are more widely spaced, so they do not look like the Ardmore succession. Moreover, they are chemically and mineralogically distinct, as described below. Outside the Black Hills region, the Ardmore bentonites vary significantly in thickness, distribution and number. To the east, along the Missouri River in South Dakota, the bentonite succession is discontinuous owing to unconformities in the lower Pierre Shale that cut out the succession in some places (described below). Where bentonites are present, only 3-4 beds are present and the thickest is only 10 cm thick. These bentonites superficially appear to be the Ardmore succession because they are present at the base of the Sharon Springs Member and they are closely spaced stratigraphically. Their correlation to the Ardmore succession is confirmed by their mineralogy and geochemistry. To the northeast, in North Dakota, 16 closely spaced bentonites occur at the base of the Pierre Shale. The bentonites range in thickness from 1 cm to 17 cm thick. The 17 cm thick bentonite is not at the base of the succession, but near the top. The entire succession is correlative with the Ardmore succession of the Black Hills on the basis of mineralogy and geochemistry, although the bentonite pattern is significantly different. VOLCANISM OF THE MIDDLE CAMPANIAN During the Late Cretaceous, the western interior was covered by a shallow inland seaway (Fig. 2.6) created by compressional stresses resulting from the subduction of the Farallon Plate under the North American plate and the resulting Sevier Orogeny (Kauffman, 1977). The Sevier Orogeny created the foreland basin and provided siliciclastic sediments to the basin. In addition to the tectonic activity, the subduction of the Farallon Plate created widespread volcanism along the western margin and several volcanic centers were present. At around 80 Ma, the western part of North America experienced a culmination in volcanic activity. Magmatism shifted from Sierran volcanism to Laramide style volcanism during this time and an overlap of the two styles produced extensive volcanic activity (Armstrong 52

Figure 2.6: During the Middle Campanian, several volcanic centers were active in western North America (Gray areas), resulting from the subduction of the Kula and Farallon Plate (inset). The Boulder Batholith and Little Elkhorn Mountains were very active during this time. The subduction also created the Sevier Orogeny, with the Absoroka Thrust being active at about 80Ma. The compressive stresses resulted in the back-arc foreland basin with four distinct facies belts. The Williston Basin in eastern North Dakota resulted from a response to compressive stresses coupled with structural weaknesses in the subsurface. 53 and Ward, 1993). The subducting plate is thought to have changed to a shallower angle of subduction, shifting a great deal of the volcanism inland to Colorado, New Mexico and Montana (Armstrong and Ward, 1993). The largest of these new volcanic centers was the Little Elkhorn Mountain volcanic complex in western Montana. These volcanoes produced extensive ash that was transported into the western interior seaway and developed as a result of the complex interplay between the thickening of the crust due to the Sevier Orogeny, faulting associated with the orogeny, and melting of the Farallon Plate in the back-arc setting (Robinson et al., 1968). METHODS Detailed stratigraphic sections of the Ardmore succession and associated bentonites were measured in the “Sharon Springs Member” of the Pierre Shale in the southern Black Hills (Table 2.1) and samples of the bentonites were collected for analysis. The sections of the Ardmore succession in the southeastern Black Hills at and around the type locality at Wasserburger Ranch were used to define the Ardmore succession stratigraphically and were used as the basis for definition based on mineralogical and geochemical characteristics. Stratigraphic sections from equivalent units in Montana, Wyoming, South Dakota, North Dakota, Colorado and Kansas (Table 2.1, Fig. 2.3) were also measured (Appendix A) and bentonites were collected to determine the lateral extent of the Ardmore succession. Bentonite samples were disaggregated by soaking in water and then agitated in a blender. Samples were then wet-sieved and material between 62.5 and 250-micrometer diameter was retained for further analysis. This material was soaked in dilute nitric acid to break up aggregates and remove carbonates, sieved again, and dried in an oven. Bromoform, with a specific gravity of 2.84 g/cc, was used to separate light grains from heavy grains. Biotite can occur in both fractions because it is close to the density of bromoform and the slightest alteration will reduce the specific gravity to less than 2.84. The heavy mineral fraction was further separated using a Frantz magnetic separator at 0.2, 0.4, 0.6, 0.8 and 1.0 amps. The non-magnetic fraction was separated using methylene 54 iodide, with a specific gravity of 3.1g/cc. This separated apatite, in the light fraction, from zircon, in the heavy fraction. Isolated grains were identified using binocular microscope and petrographic microscope characteristics and verified using a Hitachi S-4000 scanning electron microscope equipped with Oxford energy-dispersive x-ray capabilities at the University of Cincinnati Advanced Materials Characterization Center. Relative abundances of minerals were described for each bentonite. For the light mineral fraction, 300 grains were counted using a petrographic microscope for all samples to determine the relative abundance of quartz and feldspars. The 0.4-amp magnetic separation fraction contained the most biotites and this fraction was used for microprobe analysis. The grains of the 0.4-amp fraction were mounted in epoxy and polished using 0.5-micron diamond paste. The samples were put in brass holders and analyzed using an ARL SEMQ microprobe equipped with wavelength dispersive LiF, PET, ADP and TAP crystal spectrometers and energy dispersive spectrometers at the University of Cincinnati Advanced Materials Characterization Center. For each sample, ten grains were analyzed with 5 points on each grain. This provided statistical data for analysis of variance within a sample and between samples. The microprobe was set to 15kV at 100 amps. Samples were analyzed with count times of 10 seconds on peak and 4 seconds off-peak. The standards used were magnesium olivine for magnesium on the TAP crystal, magnetite for iron on the LiF crystal, albite for silica on the PET crystal and calcium on the ADP crystal, anorthite for aluminum on the ADP crystal and sodium on the TAP crystal, orthoclase for potassium on the PET crystal and TiO for titanium on the LiF crystal, from the University of Cincinnati, Department of 2 Geology standards database. 55 RESULTS Phenocryst Mineralogy

Phenocrysts of the Ardmore succession include quartz, plagioclase, sanidine and biotite with trace amounts of zircon, apatite and ilmenite (fig. 2.7-2.14). Minerals were generally euhedral or fractured, although some rounded zircon and quartz grains were present. Zircon and apatite are found as isolated grains, but also as inclusions in biotite grains (Fig. 2.13). Minerals were identified by petrographic

Figure 2.7: Quartz grains from the Ardmore microscopy and verified using a scanning electron succession. A: Fractured quartz grains and B: Beta-form quartz grains microscope with EDAX capabilities. Quartz is recognized by the first-order yellow birefringence and is typically beta-form or fractured (Fig. 2.7). Potassium feldspar is recognized by first-order gray birefringence with 90° cleavages (Fig. 2.8). Most potassium feldspar grains exhibited dissolution features. Plagioclase is recognized by Figure 2.8: Potassium feldspar from the Ardmore succession. first order gray birefringence with polysynthetic twinning and 90° cleavages (Fig. 2.9). Most plagioclase grains were fractured. Biotite grains are euhedral and range from dark brown (Fig. 2.10) to green, depending upon the degree of alteration. Apatite crystals were short and prismatic with

Figure 2.9: Plagioclase from the Ardmore first-order gray birefringence (Fig. 2.11). Zircon succession. 56

Figure 2.10: Biotite grains of the Ardmore succession are typically euhedral. When Figure 2.12. Most zircons of the Ardmore fresh they are dark brown, but alter to green succession are elongate and prismatic, or light brown. although fracturing and rounding are present in some specimens.

Figure 2.13: Zircons are sometimes present as inclusions within biotites.

Figure 2.11: Apatites of the Ardmore succession were generally euhedral, although Figure 2.14: Ilmenites are a rare accessory some also exhibited linear dissolution mineral in the bentonites of the Ardmore features. succession. 57 Figure 2.15: Composition of the bentonites of the Ardmore succession indicate 4 magma 2.15: Composition of the bentonites Ardmore Figure indicated. are groups rhyolite succession. Two types. A: Rhyolites of the Ardmore is indicated by 1 has very few phenocrysts and no biotites. This group Rhyolite group is indicated by circles 2 has abundant phenocrysts. This group plussed. Rhyolite group 1 in the 1-meter thick bentonite of Ardmore B: Part of andesite group and squares. in circles succession. Bentonites of the lower part 1-meter thick bentonite are 1 plusses. C: Andesite group and in the upper part of 1-meter thick bentonite are 2 within the thick succession and Andesite group other bentonites of the Ardmore from 1 can be succession, indicated by a diamond. Andesite group bentonite of the Ardmore 5 (crossed 4 (plusses), subgroup 3 (stars), subgroup Subgroup further segregated. 7 (circles). subgroup 6 (squares), subgroup squares), 58 crystals are long and prismatic with third-order birefringence (Fig. 2.12), and some were elongate but rounded. Zircon and apatite crystals are frequently found as inclusions in biotites (Fig. 2.13). Ilmenite is present but rare in the Ardmore succession (Fig. 2.14). Mineral suites of the bentonites in the Ardmore succession can be divided into three groups (Table 2.2, Fig. 2.15). In the first group, potassium feldspar was more abundant than quartz or plagioclase (average quartz = 29%, average potassium feldspar = 39%, average plagioclase = 32%) and biotite was abundant. In the second group potassium feldspar and plagioclase were approximately equal (average quartz = 21%, average potassium feldspar = 40%, average plagioclase = 40%), but overall very few phenocrysts were present. No biotite was present. In the third group, plagioclase was greater than quartz or potassium feldspar (average quartz = 19%, average potassium feldspar = 36%, average plagioclase = 46%) and biotite was very abundant. Zircon and apatite were common accessory minerals.

Biotite Geochemistry

Biotite was analyzed for SiO , TiO Al O , FeO*, MgO, CaO, Na O, and K O (Table 2 2, 2 3 2 2 2.3). Concentrations of Fe2+, Fe3+, Mg2+ and the magnesium number were calculated for all samples (Table 2.3). The general formula for biotite is KZ (X,Si) O (O,OH,F) . Biotite 2-3 4 10 2 crystals are arranged in layers of tetrahedra-octahedra-tetrahedra, separated by inter-layer cations, in this case potassium. The tetrahedral consist of (X,Si) O , where X is usually 2 5 aluminum, but can also include Be, B, and Fe3+ and can replace silica by up to 50%. Most substitution occurs in the octahedral layer, Z, where Al, Fe, Ti and Mg are the primary elements that can occur. The presence of these cations in the biotite structure indicates their availability in the magma (Adbel-Rahman, 1996). Magnesium and iron values in biotite and amphibole are only slightly higher than the coexisting melt (Hess, 1989), so the biotite composition is a reflection of the melt composition. The magnesium number is a useful indicator of the degree of crystal fractionation, assuming a mantle-derived primary magma. 59 The magnesium number is calculated with the equation Mg2+/(Mg2++Fe2+)*100 (Tatsumi and Eggins, 1995). Initial magma generated from the mantle has a magnesium number of 90. In early stages of crystallization, magnesium-rich minerals such as olivine and pyroxene crytallize out, so that differentiated magmas have a lower magnesium number. Typical subduction zone magmas have a magnesium number around 60-70% (Hess, 1989). Samples within the Ardmore succession have magnesium numbers between 39-54%, indicating a highly fractionated magma. Total oxide weight percents are between 73 at 99% of the total for biotite samples. Most of the variation accounting for the differences in weight percents between samples was in the SiO content. A small portion of the biotite crystal consists of volatiles such as 2 water and fluorine, which accounts for some of the missing weight in the microprobe totals. In unaltered biotites, the K O concentration is approximately 9%. K O values for the biotite 2 2

Figure 2.16: Biotites of the Ardmore succession plot in two groups. Andesite 1 and rhyolite 2 plot together in the calcalkaline magma suit, plotted as circles in the diagrams. Andesite group 2 plots in the alkaline magma suite, plotted as diamonds. (After Abdel-Rahman, 1996) 60

Figure 2.17: Bavarient plots of biotite in the Ardmore succession and surrounding bentonites. A: Plot shows the differentiation trend of the bentonites in the Ardmore succession. Rhyolite 1 and andesite 2 exhibit the least differentiation while bentonites outside the Ardmore succession exhibit the highest differentiation. B: Plot of Al2O3 compared to magnesium number. Three groups can be discerned, andesite group 2, non- Ardmore bentonites and other bentonites of the Ardmore succession. C: Plot of TiO2 compared to magnesium number. Overall, the bentonites show a trend in titanium from low concentrations to high concentrations from the base of the section to the top. samples were between 4.6-10.4%. Even in grains with low K O values, concentrations of 2 other elements, particularly TiO , Al O , FeO* and MgO remained consistent within a sample, 2 2 3 between altered and unaltered biotites.

Biotite data were plotted on bivarient and ternary plots (Fig. 2.16, 17). Fig. 2.16 shows ternary plots of Al O , FeO and MgO. This plot is based on data of known volcanic 2 3 compositions from 26 separate volcanoes and 329 biotite samples representing alkaline, calc-alkaline and peraluminous magma sources (Adbel-Rahman, 1996). Al O , FeO and 2 3 MgO were used because they are interchangeable in the octahedral cation site of biotite. Alkaline magmas are iron-rich, due to crystal fractionation and the fact that iron oxides and iron-titanium oxides form late in the fractionation sequence (Adbel-Rahman, 1996). Calcalkaline magmas are relatively magnesium-rich as a result of increased water content 61 that allows iron oxides and iron-rich amphiboles to crystallize early, removing iron from the system. Peraluminous magmas are enriched in aluminum due to partial melting of the continental crust, with abundant aluminum-rich minerals. Most biotite samples of the Ardmore bentonite succession plot within the calc-alkaline magma source based on this plot (Fig. 2.16). One bentonite within the succession plots within the alkaline magma source. Bivarient plots show variations of cation concentrations within the biotite structure. Factors useful in segregating bentonites of the Ardmore succession included Al O , Fe3+, TiO and Mg#. Al O , Fe3+, and TiO were plotted against 2 3 2 2 3 2 magnesium number to show variations in these elements with differentation (Fig. 2.16, Tatsumi and Eggins, 1995). DISCUSSION Magma Types

A combination of biotite geochemistry and bulk mineralogy was used to distinguish four major magma types for the bentonites of the Ardmore succession. Magma composition was determined by the mineral suite composition (after Streckeisen, 1976) and further characterized by the biotite geochemistry (after Abdel-Rahman, 1996). Concentrations of quartz, potassium feldspar and plagioclase are used to determine the magma source based on the International Union of Geological Sciences classification (Streckeisen, 1976, Fig. 2.15). This characterization is further refined by biotite compositions (Fig. 2.16, 2.17). Four magma compositions were identified in the bentonites of the Ardmore succession, two rhyolites and two andesites. General magma composition is based on light mineral concentrations, as shown in Table 2.2 and Fig. 2.15. The andesites had abundant phenocrysts, especially biotite and plagioclase. Accessory minerals, such as zircon and apatite were present in the andesites. Ilmenite was present but rare. One of the rhyolites had abundant phenocrysts, including biotite and zircon and apatite, as accessory minerals, but the other rhyolite had very few phenocrysts and no biotite. Rhyolite group 1 is distinguished from other bentonites of the section due to the 62 small number of phenocrysts and the lack of biotite (Table 2.2). Rhyolite group 2 has abundant phenocrysts including biotite. Biotite of rhyolite group 2 plot with the andesites on geochemical plots (Fig. 2.16, 2.17). Biotite geochemical composition discriminates the two groups within the bentonites of the Ardmore succession (Fig 2.16). Biotite from andesite group 1 is similar in composition to rhyolite group 2 and is consistent with a calcalkaline magma derived from subduction zone activity, based on data presented by Abdel-Rahman (1996). Biotite in andesite group 2 is distinguished based on lower Al O composition (Fig. 2.16, 2.17), indicating an alkaline 2 3 anorogenic magma source (Adbel-Rahman, 1996). Bivarient plots of the magnesium number compared to Fe3+ show a differentiation trend. All biotite samples fall along a single line of magmatic differentiation. As magma differentiates, the magnesium number decreases and Fe3+ increases (Tatsumi and Eggins, 1995). Magnesium numbers are between 39-54% for the Ardmore bentonite succession and 28-36% for bentonites outside the Ardmore succession. Andesite 2 and rhyolite 1 plot near the less differentiated end of the spectrum, with magnesium numbers between 47-54% (Fig. 2.17). Biotite in the Ardmore succession indicate a differentiation trend from the base of the succession to the top and bentonites higher in the succession indicate a high degree of differentiation compared to the Ardmore succession (Fig. 2.17). Al O is also useful in segregating biotite within the Ardmore succession (Fig. 2.17). 2 3 Al O can replace MgO and FeO* in the biotite structure. Al O is plotted against the 2 3 2 3 magnesium number to show the concentration of Al O with degree of crystal fractionation. 2 3 Three groups can be distinguished in this plot. Biotite of andesite group 2 plots separately from the other bentonites in the Ardmore succession and these are both separate from other bentonites outside the Ardmore succession. Higher Al O concentrations in most bentonites 2 3 of the Ardmore succession are consistent with increased incorporation of the continental crust through partial melting (Hess, 1989). TiO can also substitute for other cations in the biotite structure. However, in the 2 63 bentonites of the Ardmore succession, TiO cannot be used for segregation. A general 2 trend is seen, however, in the TiO concentration of biotite. The least segregated bentonites, 2 rhyolite 1 and andesite 2, have higher TiO concentrations while the most segregated 2 bentonites, those outside the Ardmore succession, have lower TiO concentrations. This 2 pattern is consistent with increased crystal fractionation, where iron-titanium oxides such as ilmenite crystallize out of the melt (Hess, 1989). Overall, andesite 2 and rhyolite 1 are less differentiated than other bentonites of the Ardmore succession. The Ardmore succession and overlying bentonites outside this succession indicate a progression of crystal fractionation overall from the lowest bentonites to the highest bentonites. Volcanic Sources

Based on a combination of mineral concentration, biotite geochemistry, and stratigraphic position, at least 15 individual volcanic event horizons representing four magma sources can be identified in the Ardmore succession from Montana to eastern North Dakota (Fig. 2.18). Correlation of these horizons across the basin can aid in interpretation of magma sources. Traditionally, the Little Elkhorn Mountains Volcanics Complex has been considered the source for the Ardmore bentonite succession (Fig. 2.6). The Little Elkhorn Mountains were a very active volcanic source around 80Ma, producing pyroclastic flows covering 25,000 km2 (Robinson et al., 1968; Smedes, 1966). Bentonites associated with the Two Medicine Formation in Montana are all within andesite group 1 magma source. Bentonites in the Two Medicine Formation vary considerably in thickness, even across an outcrop, due to the terrestrial sedimentation and frequent local erosion or non-deposition on floodplains. Within the Two Medicine Formation, a single bentonite is present, dated at 80.54 Ma, which ranges in thickness from 0.5-2m in the area of Choteau, Montana (Roberts, 2000). Associated with this bentonite is a crystal- lithic tuff that buried a forest on the floodplains of the Two Medicine Formation. The 64 source of this tuff and bentonite has been traced to the Little Elkhorn Mountains just to the west (Roberts, 2000). This bentonite correlates with the andesitic part of the 1-meter thick bentonite in the Black Hills region and confirms the Little Elkhorn Mountains as the source for andesite group 1 in the Ardmore bentonite succession (Fig. 2.18). In addition to this thick bentonite, seven additional bentonites are present in the Two Medicine Formation of Montana that are also within andesite group 1 and correlate to the other andesitic volcanics of the Ardmore succession in South Dakota and North Dakota. Only bentonites within andesite group 1 are present in Montana, indicating that the other three magma sources are not from the Little Elkhorn Mountains Volcanic Complex. It remains unclear where the source for these bentonites would be. Several volcanic centers were active during the Late Campanian that could have contributed to these bentonites (Fig. 2.6). Calc-alkaline volcanism occurred along the western margin of North America, particularly in Washington and British Columbia, as well as in Nevada, Colorado and New Mexico. The last Sierran volcanism in Nevada occurred about 80Ma and was distinct in containing abundant muscovite (Viker and McKee, 1985). None of the bentonites in the Ardmore succession contained muscovite, however, and so this is an unlikely source. In the south, volcanism in Colorado resulted from a more shallowly subducting oceanic slab under North America (Cross and Pilger Jr., 1978; Dickenson, 1981). A southern source for the bentonites of the Ardmore succession seems unlikely because bentonites of this succession do not appear to be present in the southern part of the basin. While this may be due in part, to erosion of this interval, prevailing winds may have prevented deposition of the ash in the southern part of the basin from a more northerly source such as Washington or Canada. Sedimentation patterns and basin dynamics

Bentonites are geologically instantaneous event horizons that can be used for high- resolution correlation across the basin. Fifteen individual volcanic eruptions can be identified in the Ardmore succession (Fig. 2.18 and 2.19). These bentonites can be used to interpret sedimentation patterns and dynamics of the western interior basin. In the northern part of 65

Figure 2.18: Stratigraphic correlation of the Ardmore succession from Montana to eastern South Dakota indicating bentonites of differing compositions. 66

Figure 2.19: Stratigraphic correlation of the Ardmore succession from eastern North Dakota to western Kansas. 67 Figure 2.20: Sea level curves of the Western Interior Seaway (Kauffman and Caldwell, 1993) vs. the eustatic sea level curve (Haq et al., 1987). Generalized stratigraphy of the Western Interior Seaway is also shown, with the Ardmore bentonite succession marked. Note that the Ardmore bentonite succession is deposited during inital transgression of the Claggett Cycle, and slightly earlier than the eustatic sea level rise.

the basin, active tectonism plays are role in the basin dynamics, resulting in subsidence that responded to thrusting along the Absoroka fault in Wyoming. In the southern part of the basin, dynamics were different due to less active thrusting in the west. Northern basin: The base of the Ardmore bentonite succession is considered to mark the initial transgression of the Claggett Cycle (Fig. 2.20). The Claggett Cycle has been thought to represent a eustatic sea level rise, although tectonism was also active in North America during the time. The Claggett Cycle is best recognized in the northern part of the basin and more poorly identified in the southern part of the basin. This suggests that a northern tectonic influence may have had more of a control on this cycle than eustatic controls. The lowest bentonite of the Ardmore bentonite succession has been dated at 80.5 +/- 0.55 Ma at Redbird, Wyoming (Obradovich, 1993) and in the Two Medicine Formation in Montana (Roberts, 2000). This is half a million years earlier than the eustatic sea level rise recognized by Haq et al. (1987). Thrusting of the Absoroka fault (Fig. 2.6) occurred at 68 80Ma (DeCelles, 1994), coincident with the initial transgression of the Claggett Cycle in the northern part of the basin, which resulted in subsidence of the axial basin. Uncertainty in the calculation of the dates for the thrusting, eustatic sea level rise and deposition of the Ardmore succession makes it difficult to evaluate the exact cause of the sea level changes at this interval, and the changes may be a combination of eustatic and tectonic processes. Early thrusting along the Absoroka Thrust caused rapid subsidence in the axial basin, but may not have broken the ocean surface and therefore, no major sediment source was available (cf. model of Ettensohn, 1998). In this sediment-starved basin, multiple volcanic ash layers were deposited with little or no deposition of clastic sediments between volcanic events. The deepest part of the axial basin during this time was in the Black Hills region. A single 1-meter thick bentonite represents as many as 12 volcanic event horizons in the Black Hills region. Further to the east, these 12 volcanic event horizons are represented by discrete bentonites, separated by organic-rich shale, resulting from increased sedimentation

Figure 2.21: Generalized cross-section of the northern part of the basin from the Black Hills region to eastern South Dakota. During deposition of the Ardmore succession, eastern North Dakota has a similar depositional pattern to the Black Hills. Deposition is less continuous in the eastern part of the basin, resulting from more frequent erosion on the eastern platform. 69 from the eastern margin of the basin (Fig. 2.18). While primary subsidence associated with thrusting of the Absoroka Thrust Fault occurred in the axial basin, subsidence also occurred in the Williston Basin, in North Dakota. In the Williston Basin and the axial basin, bentonite layers Rhyolite 1A and Andesite 1A were deposited (Figure 2.21). These layers were not deposited in eastern South Dakota. The deposition of these volcanic layers indicates that sedimentation occurred earlier in the axial basin and the Williston Basin, where the basin was deeper. In eastern South Dakota, deposition did not occur in the initial transgression. The occurrence of early sedimentation in the Williston Basin, coincident with early sedimentation in the axial basin indicates that the intracratonic Williston Basin became joined with the axial basin during early subsidence (Fig. 2.6). The Williston Basin subsided resulting from flexural stresses associated with the Absoroka Thrust Fault coupled with ancient weaknesses in the continental crust. Migration of peripheral bulges away from the axial basin and the Williston Basin resulted in overlap of the bulges and yoking of the two basins so that the Williston Basin was directly connected to the foreland basin (Ettensohn, 1998). Deepening of the seaway continued in the northern part of the basin as sea level rose until it was deep enough for sedimentation to occur on the eastern platform in eastern South Dakota. Sedimentation on the eastern platform began with depsoition of Rhyolite 2A and

Figure 2.22: The Ardmore succession near Chamberlain, South Dakota. The Ardmore succession is bracketed by solid white lines. Below the Ardmore succession is the Niobrara Formation in this area. Within the Ardmore succession, a thin iron- stained channel is present, deliniated by the dashed lines. 70 continued through accumulation of Andesite 1D (Fig. 2.21). Following Andesite 1D, an unconformity formed in eastern South Dakota and North Dakota (Fig. 2.21), representing a fall in sea level and sequence boundary. This unconformity is not evident in the Black Hills. In the Williston Basin, erosion associated with the unconformity removed Andesite 1B-1D (Fig. 2.21). In eastern South Dakota, the unconformity has a great deal of relief and removed anywhere from just the shale above Andesite 1D to all of the bentonites below the unconformity. Broad thin channels are preserved in outcrop in eastern South Dakota (Fig. 2.22). The erosion associated with this unconformity in North Dakota indicates a significant amount of time in which the area was exposed to erosion. This erosion in the Williston Basin appears to have taken place while sedimentation continued in the axial basin. This could have resulted from the Williston Basin being shallower than the axial basin, so that during sea level lowstand, the Williston Basin was not deep enough for sedimentation. However, the deposition of sediments during the earliest transgression indicates that the Williston Basin was relatively deep. A second possibility is that the Williston Basin subsided associated with tectonic thrusting and as the stresses relaxed, the Williston Basin also relaxed, resulting in rebound of the basin. Alternatively, the Williston Basin subsided associated with tectonic thrusting and following initial subsidence, high sedimentation rates resulted in the Williston Basin becoming shallower after the initial subsidence. Following erosion associated with the unconformity, a similar pattern to the basal part of the Ardmore succession is represented. Andesite 1E at Redbird, Wyoming is a composite bentonite resulting from starved sedimentation in the axial basin (Fig. 2.20). Sediment starvation was not as widespread in this instance, and in the eastern Black Hills, this bentonite is correlative to three individual bentonites. Two to four bentonites are present in eastern South Dakota and North Dakota that are correlative to Andesite 1E. Following Andesite 1E, sedimentation across the basin was approximately equal during deposition of Rhyolite 1B and Andesite 1F (Fig. 2.20). At the top of the Ardmore succession, septarian 71 concretions are present representing maximum flooding at the top of the Ardmore succession. In eastern South Dakota, later erosion removed part of the upper Ardmore succession, from Rhyolite 1B to Andesite 1E. At Yankton, all bentonites above Andesite 1B has been removed by multiple erosion events (Fig. 2.19). The second transgression is represented across the entire basin, although sedimentation remained higher on the eastern side of the basin (Fig. 2.20). This transgression occurred in the Baculites mclearni ammonite range zone, which is dated at approximately 80.5 Ma (Obradovich, 1993). This transgression would be coincident with the eustatic sea level transgression (Haq et al., 1987), indicating that the initial transgression at the base of the Ardmore succession was tectonically controlled and preceded the eustatic sea level rise. Southern basin: In Kansas, no thick bentonites occur in the Sharon Springs Member (Fig. 2.19). The Sharon Springs Member in Kansas is equivalent to the Gammon Ferruginous Member, the “Sharon Springs Member” and the Mitten Black Shale Member in the Black Hills. Only a few thin bentonites are present in the type Sharon Springs Member in Kasnas, and they are near the base of the member. One single persistent bentonite is present near the middle of the member. Within the middle of the member, a significant unconformity is present. An iron-stained brecciated shale layer 25cm thick is present with Baculites obtusus specimens. Baculites obtusus is only found within this horizon in western Kansas. Mineralogical analysis of this iron-stained shale unit revealed euhedral biotite phenocrysts.

Although these biotites were heavily altered to kaolinite, it is reasonable that these biotites are from the Ardmore bentonite succession. Within this interval, only one other bentonite has biotites. This indicates that the Ardmore bentonite succession was at least partially deposited in Kansas and then removed by subsequent erosion. Since the entire interval was removed in Kansas (Fig. 2.19), this would indicate that the area was significantly shallower than the northern basin, where sedimentation was preserved at least in part even in the eastern part of the basin. In the southern part of the basin, no tectonics were active, and the basin was more stable, so littlesubsidence occurred. Although the basin must have been 72 deep enough to have Baculites obtusus present, the basin probably remained very shallow thoughout the entire interval. CONCLUSIONS The Ardmore succession is an important unit in the Middle Campanian that has been used for regional correlation based on stratigraphic position. The mineralogy and geochemistry of the Ardmore succession indicates deposition from multiple volcanic sources and that the succession is incomplete in some areas. In addition, bentonites that have been incorrectly identified as the Ardmore succession in some areas. The underlying Gammon Ferruginous Member contains several bentonites and in condensed sections, these bentonites superficially resemble the Ardmore succession. However, they do not share the internal characteristics of these bentonites. It is important to identify these characteristics to ensure proper correlation. Recognition of the phenocryst compositions has allowed correlation into areas where the Ardmore succession is not traditionally recognized because the succession is incomplete. The phenocrysts have also been identified in unconformities indicating that the Ardmore succession was deposited and then removed. Four magma sources are identified in the bentonites of the Ardmore succession, two andesitic and two rhyolitic sources. The more aluminous andesite source is the Little Elkhorn Mountains Volcanic Complex in Montana. This is supported by stratigraphic correlation of the andesitic bentonites into the Two Medicine Formation, where an associated ash flow tuff preserved buried trees. Paleocurrent reconstructions point to the Little Elkhorn Mountains Volcanic Complex (Roberts, 2000). The other three volcanic sources are uncertain. Several volcanic centers were active during this time, particularly along the western margin of Washington and British Columbia, as well as in Nevada, New Mexico and Colorado. High-resolution correlation of individual volcanic events in the Ardmore bentonite succession provides detail useful in interpreting the basin dynamics. In the northern part of the basin, initial transgression during deposition of the Ardmore succession occurred earlier than eustatic sea level rise and was a response to subsidence associated with thrusting along 73 the Absoroka Thurst belt. Subsidence in the axial basin and Williston Basin resulted in initial transgression into these areas. Continued sea level rise eventually covered the eastern stable platform deeply enough preserve ash and sediments. A complete sequence is recognized in the lower part of the Ardmore bentonite succession, and an unconformity is present within the Ardmore succession marking a sequence boundary. Above the sequence boundary, transgression of a second sequence is preserved within the top of the Ardmore bentonite succession. This second sequence transgresses across the entire basin and is coincident with the eustatic sea level rise. In Kansas, the basin was very shallow and erosion dominated during this interval. Although the Ardmore bentonite succession was at least partially deposited in this area, the entire interval was subsequently removed due to erosion. BIBLIOGRAPHY Adbel-Rahman, A.-F.M., 1996. Discussion on the comment on nature of biotites in alkaline, calk-alkaline and peraluminous magmas. Journal of petrology, 37(5): 1031-1035. Armstrong, R.L. and Ward, P., 1993. Late Triassic to Earliest Eocene Magmatism in the North American cordillera: Implications for the Western Interior basin. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 49-72. DeGraw, H.M., 1975. The Pierre-Niobrara unconformity in western Nebraska. In: W.G.E. Caldwell (Editor), The Cretaceous System in the Western Interior of North America. The Geological Association of Canada Special Paper, pp. 589-607. Dyman, T.S. et al., 1994. Stratigraphic transects for Cretaceous rocks, Rocky Mountains and Great Plains regions. In: M.V. Caputo, J.A. Peterson and K.J. Franczyk (Editors), Mesozoic systems of the Rocky Mountain region, USA. Rocky Mountain Section SEPM, Denver, pp. 365-392. Fox, J.E., 1993a. Stratigraphic cross sections A-A’ through F-F’, showing electric logs of Upper Cretaceous and older rocks, Powder River Basin, Montana and Wyoming. U. 74 S. Geological Survey Oil and Gas Investigations Chart, 0135: 3. Fox, J.E., 1993b. Stratigraphic cross sections G-G’ through L-L’, showing electric logs of Upper Cretaceous and older rocks, Powder River Basin, Montana and Wyoming. U. S. Geological Survey Oil and Gas Investigations Chart, 0136: 3. Fox, J.E., 1993c. Stratigraphic cross sections m-M’ through R-R’, showing electric logs of Upper Cretaceous and older rocks, Powder River Basin, Montana and Wyoming. U. S. Geological Survey Oil and Gas Investigations Chart, 0137: 3. Fox, J.E., 1993d. Stratigraphic cross sections S-S’ through V-V’, showing electric logs of Upper Cretaceous and older rocks, Powder River Basin, Montana and Wyoming. U. S. Geological Survey Oil and Gas Investigations Chart, 0138: 3. Gill, J.R. and Cobban, W.A., 1965. Stratigraphy of the Pierre Shale, Valley City and Pembina Mountain areas, North Dakota. U. S. Geological Survey Professional Paper 392-A: A1-20. Gill, J.R. and Cobban, W.A., 1966. The Redbird section of the Upper Cretaceous Pierre Shale in Wyoming. U. S. Geological Survey Professional Paper 393-A: A-1-A73. Gill, J.R., Cobban, W.A. and Schultz, L.G., 1972. Stratigraphy and composition of the Sharon Springs Member of the Pierre Shale in Western Kansas. U. S. Geological Survey Professional Paper, 728: 50. Gill, J.R., Cobban, W.A., Scott, G.R. and Burkholder, R.E., 1975. Unedited stratigraphic sections of the Pierre Shale near Roundbutte and Buckeye in Larimer County, Northern Colorado. U. S. Geological Survey Open-File Report, 75-129, Denver, 12 pp. Hess, P.C., 1989. Origin of Igneous Rocks. Harvard University Press, Cambridge, MA, 336 pp. Hicks, J.F., Obradovich, J.D. and Tauxe, L., 1999. Magnetostratigraphy, isotope age calibration and intercontinental correlation of the Red Bird section of the Pierre Shale, Niobrara County, Wyoming, USA. Cretaceous Research, 20: 1-27. 75 Kauffman, E.G., 1977. Geologic and biologic overview: Western Interior Cretaceous Basin. Mountain Geologist, 14: 75-99. Knechtlel, M.M. and Patterson, S.H., 1956. bentonite deposits in marine Cretaceous formations, Hardin district, Montana and Wyoming. U. S. Geological Survey Bulletin, 1023: 116. Obradovich, J.D., 1993. A Cretaceous Time Scale. In: W.G.E. Caldwell and E.G. Kaufman (Editors), Evolution of the Western Interior Basin. Geological Association of Canada, pp. 379-396. Roberts, E.C., 2000. MS Thesis, University of Montana, Missoula. Robinson, G.D., Klpeer, M.R. and Obradovich, J.D., 1968. Overlapping plutonism, volcanism and tectonism in the Boulder Batholith region, western Montana. Geological Society of America Memoir, 116: 557-576. Rogers, R.R., Swisher, I., C. C. and Horner, J.R., 1993. 40Ar/39Ar age correlation of the nonmarine Two Medicine Formation (Upper Cretaceous), northwestern Montana,

U. S. A. Canadian Journal of Earth Science, 30: 1066-1075. Streckeisen, A., 1976. To each plutonic rock its proper name. Earth science reviews, 12(1): 1-33. Tatsumi, Y. and Eggins, S., 1995. Subduction zone magmatism. Blackwell Science, Cambridge, 211 pp. Wing, M.E., 1940. Bentonites of the Belle Fourche district. South Dakota Geological Survey Report of Investigations, 35: 29. 76 Distal record of three volcanic centers during the middle Campanian in the Cretaceous Western Interior Seaway ABSTRACT Bentonites are common in the lower Pierre Shale of the Cretaceous Western Interior. These bentonites provide a distal record of volcanic activity during this time frame and form event-stratigraphic horizons useful for regional correlations. Whole rock geochemical analysis indicates that volcanism was active in a forearc, island arc and backarc setting during the deposition of the Lower Pierre Shale. Bentonites can be further distinguished based on the mineral compositions and biotite geochemistry, when available.

INTRODUCTION During the Late Cretaceous, the Western Interior of North America was covered by a shallow retro-arc seaway created by compressional stresses in the west associated with the subduction of the Kula and Farallon Plate under the North American Plate under the North American Plate (Fig. 3.1, Kauffman, 1977). The Sevier Orogenic belt was active in Montana, Wyoming, Idaho and Utah (Monger, 1993) and provided siliciclastic sediments to the Western Interior Seaway. In addition to the tectonism associated with the Sevier Orogenic Belt, several areas of volcanism are also active. The Little Elkhorn Mountains volcanic complex was very active during the Middle Campanian and produced widespread volcanic ash as a result of the interplay between the volcanism and tectonism of the Sevier Orogeny (Robinson et al., 1968). Sodic to normal calc-alkaline volcanic activity was present along the western margin in Washington and British Columbia (Armstrong and Ward, 1993). Volcanism was also present in the Sierra Nevadas during the early part of the Middle Campanian (Vikre and McKee, 1985), but shifted to Colorado during the latter part of the Middle Campanian (Cross and Pilger Jr., 1978; Dickenson, 1981). Although general trends in magmatic activity are known (Armstrong and Ward, 1993; Christensen et al., 1994), little information is available on the specific timing of different volcanic centers. However, these volcanoes produced ash that has transported eastward into the Western Interior Seaway, where it is preserved in the marine sediments as bentonites. These bentonites provide a distal record of multiple volcanic events and provide information concerning the relative 77

Figure 3.1: During the Middle Campanian, several volcanic centers were active in western North America (Gray areas), resulting from the subduction of the Kula and Farallon Plate (inset). The Boulder Batholith and Little Elkhorn Mountains were very active during this time. The subduction also created the Sevier Orogeny, with the Absaroka Thrust being active at about 80Ma. The compressive stresses resulted in the back-arc foreland basin with four distinct facies belts. The Williston Basin in eastern North Dakota resulted from a response to compressive stresses coupled with structural weaknesses in the subsurface. 78 timing of these volcanic events. The Middle Campanian was a period of extensive volcanic activity. The distal strata of the Middle Campanian (Fig. 3.2) contain bentonites representing this volcanic activity. Cretaceous bentonites of the Western Interior were first identified as volcanic ash layers in the Big Horn Basin of Wyoming by Hewett (1917). The presence of angular or terminated crystals of plagioclase, orthoclase and biotite, which are not present in the surrounding siliciclastic rocks, compared favorably with volcanic ash from several localities. The volcanic glass is recognized to have altered to clay with abundant silica and aluminum by the hydration of glass soon after deposition. Bentonites are dominantly formed by the alteration of silica-rich volcanic ash from highly explosive volcanic eruptions (Fisher and Schmincke, 1984). The explosive nature of these volcanic events ejects the ash into the stratosphere where it is carried long distances by prevailing winds and fall out over large areas.

Figure 3.2a: Generalized stratigraphy of the lower Pierre Shale in the Black Hills and western Kansas (the type locality of the Sharon Springs Member), as it is currently recognized in the literature. b: Stratigraphy of the lower Pierre Shale in the Black Hills and western Kansas, as modified in this paper. 79 STRATIGRAPHY OF THE MIDDLE CAMPANIAN The strata of the Western Interior Seaway record several sea-level fluctuations (Fig. 3.3). Ten third-order sea-level fluctuations were recognized by Kauffman and Caldwell (1993). These local ossilations may correlate with global sea-level fluctuations (Haq et al., 1988) attributed to tectono-eustatic processes as a result of increased sea floor spreading rates (Kauffman and Caldwell, 1993). Alternatively, sea-level fluctuations of the Western Interior Seaway could correlate with tectonic pulses in the Sevier Orogeny. Pulses of thrusting in the Sevier Orogenic Belt can be correlated with timing of the third-order sea-level fluctuations (DeCelles, 1994; Villien and Kligfield, 1986; Witschko and Dorr Jr., 1983). Lillegraven and Ostrech (1986) suggested that these relative sea-level fluctuations are restricted to the northern part of the basin, where tectonic activity was most active.

Figure 3.3: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway. Orogenic thrusting events occur synchronously with transgressions in the basin (Villien and Kligfield, 1993). The major cycles recognized by Kauffman and Caldwell (1993) and generally accepted for the northern part of the seaway are related to eustatic sea level changes (Haq, 1988) and thrusting events of the Sevier Orogeny Some of the major stratigraphic units, particularly within the Campanian, are given for the western margin and the distal sediments. Tectonic material compiled from Villien and Kligfield (1993), Wiltschko and Dorr (1983) and DeCelles (1994). 80 The Claggett Cyclothem was deposited during the Middle Campanian (Kauffman and Caldwell, 1993; Obradovich, 1993) and this time frame is represented by Baculites obtusus through Baculites scotti ammonite range zones (Fig. 3.2, Obradovich, 1993). Transgression of the Claggett Cycle in the Western Interior Seaway occurred in the Baculites obtusus ammonite range zone and is represented by extensive black shales associated with the Ardmore succession (Kauffman, 1977). These extensive black shales are referred to as the Sharon Springs Member of the Pierre Shale and extend from Kansas to South Dakota. In North Dakota, the equivalent strata are called the Pembina Member of the Pierre Shale [Gill, 1965 #155]. The Pembina Member designation was used in North Dakota due to its proximity to the type locality in Manitoba. However, the use of this term is confusing because in Manitoba, the Pembina Member is in the Vermillion Creek Formation, not in the Pierre Shale. Therefore, in this paper, the Pembina Member of North Dakota will be referred to as the “Sharon Springs Member”. In the Black Hills, strata equivalent to the Sharon Springs Member in Kansas are divided into the Gammon Ferruginous Member, the “Sharon Springs Member” and the Mitten Black Shale Member (Gill and Cobban, 1966; Rubey, 1930). The “Sharon Springs Member” is restricted to only the most organic rich interval in this area. It is recognized that the “Sharon Springs Member” in the northern part of the basin is not present in the type locality in Kansas, based partly on data presented in this paper, and the stratigraphy will be revised in a later paper (Chapter 4). STRATIGRAPHY OF THE “SHARON SPRINGS MEMBER” IN THE BLACK HILLS The “Sharon Springs Member” in the Black Hills can be subdivided into three intervals on the basis of unconformities (Fig. 3.4, Martin, 1996; Martin et al., 1996). In the Black Hills region, the stratigraphy of the lower Pierre Shale has been extensively described and distinct bentonites have been recognized in the field. The base of the Sharon Springs Member is marked by the Ardmore bentonite succession in the Black Hills region (Gill and Cobban, 1966; Spivey, 1940). 81 Although bentonites are common throughout the Lower Pierre Shale, only the Ardmore bentonite succession has been regionally correlated based on stratigraphic position (Dyman et al., 1994). The Ardmore bentonite succession is a thick succession of bentonites. In the Black Hills the lowest bentonite is 1-meter thick and was originally described by Spivey (Fig. 3.4, 1940). Geochemical analysis of the bentonites of the Ardmore succession indicate that the 1- meter thick bentonite is a composite bentonite and that at least two volcanic sources are responsible for the 13 volcanic events represented by the bentonites of the Ardmore succession (Chapter 2). Interval II includes strata between

Figure 3.4: Generalized stratigraphy of the major the top of the Ardmore succession and below bentonites of the lower Pierre Shale in the Black Hills. The Ardmore bentonite of Spivey (1940) was restricted the intra-member unconformity of the to the 1-meter thick bentonite at the base of the succession. Gill and Cobban (1966) included all the “Sharon Springs Member” (Fig. 3.4, Martin, thick bentonites at the base of the Sharon Springs with the Ardmore bentonite. 1996). In this interval, two additional thick bentonites are present. These bentonites are 25cm thick each and in the eastern Black Hills these bentonites are separated by less than 5cm of shale. These bentonites make a distinct couplet in the field that is easily recognized in the Black Hills and has been 82 termed the Bentonite couplet (Martin et al., 1996). Above the bentonite couplet are several thin bentonites, less than 1cm thick that are not distinct in the field. Interval III extends from the intra-member unconformity to the unconformity that marks the top of the “Sharon Springs Member” in the Black Hills (Fig. 3.4, Martin, 1996). The three thin bentonites in this interval are distinct in the field. Two thin bentonites make a distinct couplet. The two bentonites are each 2cm thick and separated by less than 5cm of shale. These bentonites are termed the “first stringer” (Martin et al., 1996). Above the first stringer is a single bentonite that is 3cm thick, this bentonite is termed the “second stringer” (Martin et al., 1996). These bentonites are only recognized in the Black Hills based on field observations. METHODS Bentonites of the lower Pierre Shale were sampled from 13 localities from the Pierre Shale in South Dakota, North Dakota, Wyoming, Colorado and Kansas, the Steele Shale in Wyoming and the Claggett Shale in Montana (Fig. 3.2, Table 3.1). Detailed stratigraphic sections were measured at each locality and samples were taken of each bentonite layer. For bentonite greater than 5cm thick, multiple samples were taken to determine vertical variability. A combination of whole rock and phenocryst analysis can be used to identify magma sources, which can further be used for regional correlation. Mineralogical analysis confirms that clay layers are bentonites and thus are event stratigraphic deposits, potentially useful in regional correlation. Whole rock rare earth element patterns can also be used for segregating magma sources. As altered volcanic ash layers, bentonites have two distinct grain sizes. The clay size fraction of bentonites is consistent with alteration from volcanic glass (Hewett, 1917; Slaughter and Early, 1965). The non-clay fraction consists of sand sized euhedral or fractured minerals including quartz and feldspar. Whole rock analysis of immobile elements in bentonites can provide information regarding magmatic source. Explosively erupted ash beds commonly show only minor 83

Fig. 3.5: Map of South Dakota showing localities used in this study. 84 variation in composition with distance from the source (Heiken and Wohletz, 1992; Izett, 1981). One hundred five samples were analyzed for 34 elements using instrumental neutron activation analysis. Results were then plotted on chondrite-normalized graphs (after Floyd and Winchester, 1978). Chondrite values are assumed to represent primitive Earth compositions and deviations from this composition indicate differentation of the magma. Chondrite-normalized rare earth element distribution of the bentonites reveals three magma sources for the bentonites of the lower Pierre Shale, as discussed below. Multivariate discriminant function analysis separates statistical groups of bentonites that can be used to identify discrete horizons across the basin. Discriminant function analysis was performed using Systat Version 10 for Windows by SPSS. In addition to whole rock analysis, phenocryst mineralogy and geochemistry was performed. Samples were wet sieved using 62.5 and 250-micrometer sieve sizes. The material that was less than 62.5 micrometers was allowed to settle for two hours, so that only the clay-sized fraction, less than 2 micrometers, remained in suspension. The suspended material was analyzed for the clay composition using oriented clay slides by concentrating clay-sized material suspended in water on a slide by evaporation. Samples were analyzed using powder x-ray diffraction of the samples under air-dried, glycolated and heated conditions. Samples were run from 2θ to 32θ with a step size of 0.05θ and a count time of 2 seconds (Moore and Reynolds, 1997). Material that was between 62.5 and 250-mircometers was soaked in dilute nitric acid to break up aggregates and remove carbonates. The samples were then sieved again to remove any additional clay-sized particles and rinse the acid. Phenocrysts from 62.5 to 250-micrometers were dried in an oven and kept for mineralogical and geochemical analysis. Bromoform, with a specific gravity of 2.84 g/cc was used to separate light grains from heavy grains. Biotite has a tendency to be found in both the heavy and the light mineral fraction as a result of its density being close to that of bromoform, any slight amount of weathering will reduce its specific gravity. Analysis of biotite in the light fraction indicated 85 that the biotites in the light fraction were slightly altered, so only the heavy biotites were used in analysis. The heavy mineral fraction was further separated magnetically using the Frantz magnetic separator. Magnetic separation was done at 0.2, 0.4, 0.6, 0.8 and 1.0 amps. The non-magnetic fraction was further separated using methylene iodide with a specific gravity of 3.1 g/cc. Selected grain mount samples were thin sectioned for point count analysis and phenocrysts from other samples were mounted for SEM and EDAX analysis to verify identification of grains. The 0.4 amp magnetic separation contained most of the biotites and this fraction was mounted in epoxy and polished using 0.5-micron diamond paste for microprobe analysis. Single grains were chosen by visual inspection during microprobe analysis. The samples were put in a brass sample holder and the grains were analyzed using an ARL SEMQ microprobe equipped with wavelength dispersive LiF, PET, ADP and TAP crystal spectrometers and energy dispersive spectrometers at the University of Cincinnati Advanced Materials Characterization Center. For each sample, 10 biotite grains were analyzed and five points were analyzed on each grain. This provided for analysis of both within the grain and within the sample variation so that between beds variation statistics could be performed. The microprobe was set to 15kV at 100 amps. Samples were analyzed with count times of 10 seconds on peak and 4 seconds off-peak. The standards used were magnesium olivine for magnesium on the TAP crystal, magnetite for iron on the LiF crystal, albite for silica on the PET crystal and calcium on the ADP crystal, anorthite for aluminum on the ADP crystal and sodium on the TAP crystal, orthoclase for potassium on the PET crystal and TiO for titanium on the LiF crystal, from the University of Cincinnati, Department of 2 Geology standards database. 86 RESULTS Clay Mineralogy

In most samples, the clay mineral present was montmorillonite. Air-dried samples have a dominant, broad peak at 14-14.8A d-spacing with a secondary peak at 4.4-4.5A d- spacing. Upon glycolation, these broad peaks changed to a sharp strong peak at 16.8-16.9A with a secondary peak at 4.5A (Fig. 3.3). This is consistent with previous reports of bentonite composition in the northern Black Hills (Wing, 1940) and in parts of Montana, Wyoming and South Dakota (Schultz et al., 1980). Some samples are reported to have kaolinite and beidellite (Schultz et al., 1980). This is particularly true of the bentonites in the Sharon Springs Member of Kansas (Gill et al., 1972). In areas that have undergone extensive secondary burial and/or uplift, mixed-layer clays with a high proportion of illite are present (Schultz et al., 1980). Montmorillonite is the most common mineral that forms from the alteration of volcanic glass in marine environments. It is considered to have formed during early diagenesis, probably before significant burial and possibly while still in the water column, as a result of interaction with the seawater (Hewett, 1917; Tomita et al., 1993). Clay mineralogy is not a particularly diagnostic feature of the bentonites, as the clay composition is dependent on the depositional and diagenetic conditions under which the beds are formed.

Figure 3.6: General clay mineralogy of bentonites in the lower Middle Campanian are montmorillonite, as shown by x-ray diffraction. 87 However, it does confirm that the clay layers are bentonites derived from a volcanic source, making them useful for regional correlations (Hewett, 1917). Whole Rock Geochemistry

Immobile elements of whole rock samples (Table 3.2) indicate three distinct groups for bentonites of the lower Pierre Shale, corresponding to a forearc, island arc and backarc setting. The first group includes bentonites of the Ardmore bentonite succession. In this group, the light rare earth elements are enriched, with values around 100 times chondrite values. The heavy rare earth elements have concentrations around 10-20 times chondrite. The bentonites have a slight europium anomaly (Fig. 3.4). The second group had a pattern similar to the Ardmore succession except that the rare earth element concentrations overall were lower. The light rare earth elements were around 10 times chondrite values and the heavy rare earth elements were around 5-8 times chondrite values. These bentonites exhibited a slight europium anomaly (Fig. 3.4). The third group of bentonites had similar concentrations to the second group, except for a few notable differences. In this group, a positive europium

Figure 3.7: Representative chondrite-normalized plots for bentonites of the Sharon Springs and Mitten Black Shale Members showing variations in plots resulting from separate magma sources. A: Bentonites representing a backarc magma source, with a source in the Little Elkhorn Mountains volcanics complex. The high concentrations of light rare earth elements indicate partial melting of the continental crust. C: Range of values for bentonites with a volcanic arc volcanic source, with a comparison to Mt. Adams data shown as stars(Defant and Drummond, 1993). D: Range of values for bentonites with a forearc volcanic source with a comparison to Mt. St. Helens data shown as stars (Defant and Drummond, 1993). 88 anomaly is present and the heavy rare earth element concentrations are very low, between 1-5 times chondrite values (Fig. 3.4). Discriminate function analysis segregates four distinct groups within the bentonites (Fig. 3.5), distinguishing two separate groups within the backarc group above. Samples from bentonites that were easily recognized in the field or distinct based on chondrite- normalized patterns were used to determine statistical groups for discriminant function analysis. Five groups were established, the Ardmore succession, the lower and upper bentonites of the couplet, the first stringer and the bentonites with positive europium anomalies. Two functions were calculated based on these groups: Function 1 = 13.799 + 0.109*La – 0.265*Ce – 0.002*Nd + 0.632*Sm + 0.925*Eu – 1.336*Tb – 0.505*Yb – 0.194*Lu +3.747*Eu/Eu* Function 2 = 21.399 + 0.069*La – 0.193*Ce + 0.192*Nd – 0.322*Sm + 2.58*Eu – 0.773*Tb + 1.131*Yb – 0.734*Lu + 19.857*Eu/Eu*

Although five groups were used in the discriminant analysis, only four distinct groups can be identified. The upper bentonite of the bentonite couplet grouped with the first stringer based on these functions.

Figure 3.8: Discriminant function analysis produced four distinct groups in the bentonites of the “Sharon Springs Member”. The Ardmore succession (group 1) and the lower bentonite of the couplet (group 2) are both from a backarc setting, but statistical analysis indicates that they are sufficiently different from each other. The lower bentonite of the couplet and the first stringer plot together (group 3). Both are from an island arc setting. Other bentonites of the “Sharon Springs Member” also plot with the first stringer. Group 4 are bentonites with a forearc volcanic signature. 89 Phenocryst Mineralogy

Phenocrysts are very abundant in most bentonites of the lower Pierre Shale. These phenocrysts are sand sized, dominantly between 62.5 and 250-micrometers (Fig. 3.9), although biotite grains greater than 250-micrometers were noted in some samples. Phenocrysts are typically concentrated at the base of the bentonite bed and fine upward. Minerals were separated based on specific gravity and magnetism and then identified using petrographic microscopy and verified using a scanning electron microscope with EDAX capabilities. Minerals separated in the light fraction of bromoform specific gravity separations included quartz, potassium feldspar, and plagioclase (Fig. 3.9) as well as secondary minerals, such as gypsum and barite, and some altered biotites. Minerals separated in the heavy fraction of bromoform density separations were further separated using magnetic separation. Minerals in the 0.2 amp separation were dominantly ilmenite and some biotite (Fig. 3.9). Minerals in the 0.4 amp separation were dominantly biotite and the secondary jarosite minerals. Altered biotite was found in 0.6-1.0 amp seprations. Apatite and zircon (Fig. 3.9) were in the non-magnetic fraction. Quartz minerals were identified with a petrographic microscope and verified with the scanning electron microscope with EDAX. They are characterized by first-order yellow birefringence and typically were beta-form or had angular fractures and exhibited 90°-cleavage; although most potassium feldspar showed some degree of dissolution. Plagioclase feldspar is characterized by first-order gray birefringence with polysynthetic twinning, although thicker samples had yellow birefringence and typically exhibited 90°-cleavage although they also were partially dissolved. Ilmenite was opaque with well-defined crystals. Fresh biotite was dark brown and the color of the mineral masked any birefringence. Altered biotite was various shades of brown or green, depending on the degree of alteration. Most biotite had inclusions of apatite or zircon. Zircon and apatite generally had well-defined, prismatic crystals. Apatite had first-order gray birefringence, whereas zircon had fourth-order birefringence. Zircon crystals were typically elongate 90

Figure 3.9: Characteristic minerals of the bentonites in the “Sharon Springs Member” include a) quartz, b) potassium feldspar, c) plagioclase, d) ilmenite, e) biotite, f) apatite, g) zircons. 91 while apatite was shorter. Some zircon crystals were rounded. Some apatite showed linear corrosion, but this could only be seen with a scanning electron microscope. Presence and abundance of all minerals characterized mineral suites (Table 3.3). Point counts of light mineral fraction (quart, plagioclase and potassium feldspar) characterize the magma source when plotted using the International Union of Geological Sciences classification (Fig. 3.10). Most bentonites plot within the rhyolite field, with the exception of some bentonites of the Ardmore succession, which plot in the andesite field. Although the majority of the bentonites are within the rhyolite field, individual bentonite layers show a variation in mineralogic composition that can be used in conjunction with geochemistry to distinguish them.

Figure 3.10: Mineralogy of bentonites of the lower Middle Campanian strata of the western interior. A: Bentonites of the Gammon Ferruginous member. B: Representative bentonites of the Ardmore succession. C: Bentonites of the “Bentonite Couplet”. D: Bentonites of the “First Stringer”. 92 Biotite Geochemisty

Biotite is abundant in the bentonites of the Ardmore succession and in the two bentonite beds above the Ardmore succession in the “Sharon Springs Member”. Biotite phenocrysts are useful because their chemistry can aid in interpreting volcanic source magma. Eight major oxides were analyzed for the biotite (Table 3.4). The magnesium number is a useful indicator of the degree of crystal fractionation, assuming a mantle-derived primary magma. Initial magma generated from the mantle has a magnesium number of about 90. In early stages of crystallization, magnesium-rich minerals such as olivine and pyroxene crystallize out, so that differentiated magmas have a lower magnesium number (Tatsumi and Eggins, 1995). Concentrations of Fe2+, Fe3+, Mg2+ and the magnesium number were calculated for all samples (Table 3.4). The magnesium number was calculated with the equation Mg2+/(Mg2++Fe2+)*100 (Tatsumi and Eggins, 1995). Total oxide weight percents are between 73 and 99% of the total for the biotites. Most of the variation accounting for the differences in weight percents between samples was in the SiO content. A small portion of the biotite crystal consists of volatiles such as 2 water and fluorine, which account for some of the missing weight in the micoprobe totals. In unaltered biotites, the K O concentration is approximately 9%. K O values for the 2 2 bentonite samples were between 4.6-10.4%. Even in samples with low K O values, 2 concentrations of other elements, particularly TiO , Al O , FeO* and MgO remained 2 2 3 consistent within a sample. These components are useful in segregating biotite composition because these four elements occupy the same position in the octahedral cation site in the biotite crystal. Biotite chemical concentrations were plotted on bivariate and ternary plots. A ternary plot of Al O , FeO and MgO has been used to distinguish alkaline, calc-alkaline and 2 3 peraluminous magma sources (Fig. 3.11). The position of fields on this plot is based on data of known volcanic compositions from 26 volcanoes and 329 biotite samples representing three different volcanic sources (Adbel-Rahman, 1996). Alkaline magmas are iron-rich, 93 Figure 3.11: Biotite geochemistry of the bentonites in the “Sharon Springs Member” indicate three groups. Two of the groups, are calcalkaline and one is alkaline. The calcalkaline bentonites include most of the Ardmore succession, the lower bentonite of the bentonite couplet and the first stringer.

due to crystal fractionation and the fact that iron oxides and iron-titanium oxides form late in the fractionation sequence (Adbel-Rahman, 1996). Calcalkaline magmas are relatively magnesium-rich as a result of increased water content that allows iron oxides and iron-rich amphiboles to crystallize early, removing iron from the system. Peraluminous magmas are enriched in aluminum due to partial melting of the continental crust, with abundant aluminum- rich minerals. Most of the bentonites plot within the calc-alkaline field. Bivariate plots of Al O , TiO , and FeO compared to magnesium number indicate variations in the bentonites 2 3 2 based on degree of fractionation (Fig. 3.12).

Figure 3.12: Bivarient plots of major element analysis of biotites from bentonite samples in the “Sharon Springs Member”. Plot A of Mg# vs. Fe3+ indicates the degree of fractionation in the magma. Lower Mg# indicates a more fractionated magma. Plot B indicates three groups based on Al2O3 vs. Mg#. Higher Al2O3 indicates a higher degree of crustal contamination. Plot C of TiO2 vs. Mg# groups the Ardmore bentonites with the first stringer, separate from the bentonite couplet. 94 DISCUSSION The combination of mineral suites and whole rock and biotite geochemistry of the bentonites establishes multiple magmatic sources for the bentonites of the lower Pierre Shale. Three primary magma sources are indicated, a backarc, island arc and forearc volcanic center, and within these major centers, rhyolite to andesite magma compositions are represented. Trends shown in the chondrite-normalized rare earth element distributions indicate at least three source areas for the bentonites, coinciding with fore-arc, island arc and backarc volcanism in a subduction zone setting (Fig. 3.7, Tatsumi and Eggins, 1995). In all of these sources, the light rare earth elements are enriched relative to the heavy rare earth elements, consistent with calcalkaline magmas in subduction related environments (Taylor and McLennan, 1988). However, variations in the three plot types are distinctive for the three different magmatic settings. In typical subduction zones, the subducting oceanic crust undergoes a series of decompositions resulting from increased pressure and temperature, causing a release of fluids from the oceanic slab (Tatsumi, 1989). The first breakdown of the oceanic crust occurs between 75-85km where temperatures are high enough to breakdown the amphibolite of the oceanic slab, producing eclogite and the release of fluids. In typical subduction zones, the release of water from the oceanic slab into the overlying mantle wedge does not lower the melting temperature of the mantle wedge enough to cause melting (Fig. 3.13). However, in cases where the oceanic slab is young and still relatively hot, the oceanic crust itself will melt, resulting in forearc volcanism with a trace element signature reflecting altered mid-ocean ridge basalts. These magma sources are characterized by a positive europium anomaly, resulting from the incorporation of plagioclase from the subducting oceanic slab into the melt. A modern example of this type of volcanism is Mt. St. Helens, produced as a result of subduction of the young Juan de Fuca Plate under the North American Plate (Defant and Drummond, 1993). In the lower Pierre Shale, two bentonites have trace 95

Figure 3.13: Subduction zone environment showing the zones of magma generation corresponding to forearc, island arc and back arc magmatism. element signatures consistent with this forearc volcanic setting (Fig. 3.7). With continued subduction, the eclogite facies of the subducting oceanic slab breaks down between 120-150km depth, again releasing hydrous fluids and altering the oceanic crust to phlogopite (Tatsumi, 1989). The release of hydrous fluids at this depth results in lowering the melting temperature of the mantle wedge (Defant and Drummond, 1993). The trace element signature of volcanoes in this volcanic arc setting is consistent with mantle chemistry, with higher light rare earth element concentrations and a negative europium anomaly (Fig. 3.7). Several bentonites in the lower Pierre Shale have a trace element signature consistent with an island arc setting (Fig. 3.7). Within this island arc group, the upper bentonite of the “Bentonite couplet” and the “first stringer” are diagnostic bentonites that can be identified based on their phenocryst composition, geochemical signature and stratigraphic position. The “bentonite couplet” is a pair of thick bentonites that are closely spaced together in the Black Hills region. The upper bentonite of the couplet has very few phenocrysts, but the phenocrysts that are present indicate a rhyolitic source (Fig. 3.10). 96 The first stringer has light mineral compositions that indicate a rhyolite source (Fig. 3.10), but also includes biotite phenocrysts that can be used to further characterize the source magma. Biotite of the “first stringer” has magnesium numbers of 44-48%, indicating a moderately fractionated melt (Fig. 3.12). TiO and Fe3+ concentrations of the “first stringer” 2 are slightly lower than other bentonites of the lower Pierre Shale, and are useful in segregating the bentonites (Fig. 3.12). A third episode of volcanism is consistent with continued subduction of the oceanic plate and breakdown of the eclogite facies at about 180km depth, again producing hydrous fluids that are introduced into the mantle wedge. Mantle magma is introduced into a thickened continental crust where crustal contamination occurs, resulting in a trace element signature that is enriched in light rare earth elements (Tatsumi, 1989). Within the lower Pierre Shale, andesites of the Ardmore bentonite succession are consistent with a backarc volcanic source (Fig. 3.7). The Ardmore bentonite succession is a unit of thick layers of bentonite representing as many as 15 individual volcanic events (Chapter 2). The bentonites are separated into rhyolites and andesites based on their light mineral fraction (Fig. 3.10). The andesites have abundant bioite in them, which can be used to further define the bentonites. Most of these andesites plot within the calcalkaline field based on biotite compositions (Fig. 3.11, Adbel- Rahman, 1996). The first bentonite of the “bentonite couplet” shares features similar to the andesites of the Ardmore bentonite succession except that the light mineral fraction indicates a rhyolite magma source. The bentonites show an overall fractionation trend from the base of the Ardmore succession to the bentonite couplet based on biotite geochemistry. Bentonites of the Ardmore succession have a magnesium number between 45-53 and the lower bentonite of the “bentonite couplet” has a magnesium number between 26-49, indicating that the lower bentonite of the couplet is more fractionated than the bentonites of the Ardmore succesion (Fig. 3.12). Bentonites of the lower Pierre Shale indicate three primary volcanic centers, a forearc, backarc and island arc magma source (Fig. 3.7). Known magmatism was active in Arizona, 97 Nevada, Colorado, Washington, Idaho and Montana (Fig. 3.2, Armstrong and Ward, 1993). Prevailing wind directions from isopach maps of the Ardmore succession (Gill and Cobban, 1973) indicate that a southern source is not a likely source of the Ardmore succession. It is likely that prevailing wind direction remained relatively constant through the duration of this interval, and it can be assumed that the other bentonites in the sections are also not from Arizona. The Little Elkhorn Mountains volcanic complex in Montana has long been considered the source of the Ardmore bentonite succession. This volcanic source formed in a backarc setting (Fig. 3.13, Tatsumi, 1989) and is consistent with the andesites of the Ardmore bentonite succession. For the other bentonites of the lower Pierre Shale, however, the magma source is less clear. Island arc and forearc volcanism was common during this time frame, and the timing of these volcanic centers is less certain. CONCLUSION Bentonites of the lower Pierre Shale in the Black Hills indicate three primary volcanic sources. Fore-arc, island arc and back-arc volcanism were all active during this time frame. Fore-arc volcanic ash can be identified based on chondrite-normalized rare earth element patterns with low light rare earth element concentrations and a positive europium anomaly. Two bentonites in the lower Pierre Shale have rare earth element patterns consistent with forearc volcanism. Fore-arc volcanism occurs when the subducting oceanic plate is hot enough to melt, inducing an oceanic crust signature on the melt composition. Island arc volcanic ash is enriched in light rare earth elements with a slightly negative europium anomaly. Back arc volcanic ash is highly enriched in light rare earth elements, with concentrations around 100 times chondrite values due to partial melting of the continental crust. Identification of individual bentonite layers based on their magmatic signature allows for high-resolution correlation of the lower Pierre Shale and provides a means for clear identification of unconformities and condensed sections in distal marine shales where unconformities are not always easily recognized. 98 REFERENCES Adbel-Rahman, A.-F.M., 1996. Discussion on the comment on nature of biotites in alkaline, calk-alkaline and peraluminous magmas. Journal of petrology, 37(5): 1031-1035. Armstrong, R.L. and Ward, P., 1993. Late Triassic to Earliest Eocene Magmatism in the North American cordillera: Implications for the Western Interior basin. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 49-72. Christensen, E.H., Kowallis, B.J. and Barton, M.D., 1994. Temporal and spatial distribution of volcanic ash in Mesozoic sedimentary rocks of the Western Interior: An alternative record of Mesozoic magmatism. In: M.V. Caputo, J.A. Peterson and K.J. Franczyk (Editors), Mesozoic systems of the Rocky Mountain region, USA. Rocky Mountain section SEPM, Denver, pp. 73-94. Cross, T.A. and Pilger Jr., R.H., 1978. Tectonic controls of late Cretaceous sedimentation, western interior, USA. Nature, 274: 653-657. Defant, M.J. and Drummond, M.S., 1993. Mt. St. Helens: Potential example of the partial melting of the subducted lithosphere in a volcanic arc. Geology, 21: 547-550. Dickenson, W.R., 1981. Plate tectonic evolution of the southern Cordillera. In: W.R. Dickenson and W.D. Payne (Editors), Relations of tectonics to ore deposits in the wouthern Cordillera. Arizona Geological Society Digest, pp. 113-135. Fisher, R.V. and Schmincke, H.U., 1984. Pyroclastic Rocks. Springer-Verlag, New York. Floyd, P.A. and Winchester, J.A., 1978. Identification and discrimination of altered and metamorphosed volcanic rocks using immobile elements. Chemical Geology, 21(3- 4): 291-306. Gill, J.R. and Cobban, W.A., 1973. Stratigraphy and geologic history of the Montana Group and equivalent rocks, Montana, Wyoming, and North and South Dakota. U. S. Geological Survey Professional Paper, 776: 37. Gill, J.R., Cobban, W.A. and Schultz, L.G., 1972. Stratigraphy and composition of the Sharon 99 Springs Member of the Pierre Shale in Western Kansas. U. S. Geological Survey Professional Paper, 728: 50. Heiken, G. and Wohletz, K., 1992. Volcanic Ash. University of California Press, Berkeley, 246 pp. Hewett, D.F.D., 1917. The origin of bentonite and the geologic range of related materials in Big Horn Basin, Wyoming. Journal of the Washington Academy of Sciences: p. 196-198. Izett, G.A., 1981. Volcanic ash beds: Recorders of upper Cenozoic silicic pyroclastic volcanism in the western United States. Journal of Geophysical Research, 86: 10200- 22. Kauffman, E.G., 1977. Geologic and biologic overview: Western Interior Cretaceous Basin. Mountain Geologist, 14: 75-99. Monger, J.W.H., 1993. Cretaceous tectonics of the North American cordillera. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 31-48. Moore, D.M. and Reynolds, R.C., Jr., 1997. X-ray diffraction and the identification and analysis of clay minerals. Oxford University Press, 378 pp. Robinson, G.D., Klpeer, M.R. and Obradovich, J.D., 1968. Overlapping plutonism, volcanism and tectonism in the Boulder Batholith region, western Montana. Geological Society of America Memoir, 116: 557-576. Schultz, L.G., Tourtelot, H.A., Gill, J.R. and Boerngen, J.G., 1980. Composition and properties of the Pierre Shale and equivalent rocks, northern Great Plains Ragion. U. S. Geological Survey Professional Paper 1064-B: 114. Slaughter, M. and Early, J.W., 1965. Mineralogy and geological significance of the Mowry Bentonites, Wyoming. Geological Society of America Special Paper, 83: 95. Tatsumi, Y., 1989. Migration of fluid phases and genesis of basalt magmas in subduction zones. Journal of Geophysical Research, 94(B4): 4697-4707. 100 Tatsumi, Y. and Eggins, S., 1995. Subduction zone magmatism. Blackwell Science, Cambridge, 211 pp. Taylor, S.R. and McLennan, S.M., 1988. The significance of the rare earths in geochemistry and cosmochemistry. In: K.A. Gschneider and L. Eyring (Editors), Handbook on the physics and chemistry of rare earths. Elsevier, New York, pp. 485-578. Tomita, K., Yamane, H. and Kawano, M., 1993. Synthesis of smectite from volcanic glass at low temperature. Clays and Clay Minerals, 41(6): 655-661. Vikre, P.G. and McKee, E.H., 1985. Zoning and chronology of hydrothermal events in the Humboldt Range, Pershing County, Nevada. Isochron/West, 44: 17-24. Wing, M.E., 1940. Bentonites of the Belle Fourche district. South Dakota Geological Survey Report of Investigations, 35: 29. 101 Table 3.2: Concentrations (in ppm) from bentonite samples using INAA and ICP. Sample abbreviations are: BG – Buffalo Gap; BR – Brown Ranch; WsR – Wasserburger Ranch (Ardmore); WR – Wallace Ranch; C – Chamberlain.

La Ce Nd Sm Eu Tb Yb Lu La Ce Nd Sm Eu Tb Yb Lu BG1a 67 128 50 10.5 2.2 1.4 3.6 0.5 WR8 4 11 0 2.1 0.8 0 0.9 0.17 BG1a 72 137 50 8.6 2.1 1 3.4 0.48 WR7 12 21 0 1.3 1.2 0 1.1 0.17 BG1c 64 128 50 10.3 1.6 1.1 3.8 0.52 C1 22 38 20 4.5 0.8 0.5 1.3 0.2 BG2 27 55 30 6.7 1.5 1.1 3.7 5.1 C2 30 63 30 9.1 2 1.4 3 0.47 BG3a 58 108 40 8.1 1.9 0.9 3 0.46 C3a 67 124 50 8 1.3 0.8 2.5 0.36 BG3b 59 109 50 8.2 1.5 1 3 0.43 C3b 84 149 60 9.7 1.7 0.9 1.5 0.24 BG4 53 99 40 7.4 1.3 0.9 3 0.41 C4 30 51 20 2.9 0.3 0 0.8 0.15 BG5 47 92 40 8 1.9 1 4.2 0.49 C5 34 49 20 3.9 0.7 0 1 0.17 BG6 62 114 50 8.2 1.6 0.9 3.5 0.5 C6 8 11 0 0.8 0.4 0 0.2 0.05 BG7a 53 101 50 9.2 1.8 1.1 3.8 0.56 C7 8 12 0 0.7 0.3 0 0.2 0.06 BG7b 63 112 40 8.2 1.5 0.9 3.6 0.5 C8 29 42 10 1.2 0.07 0 0 0 BG8 61 114 50 9.2 1.8 0.9 3.2 0.44 C9 4 12 20 5 1.2 0 0.4 0.07 BG9 7 14 0 1.9 0.5 0 1.3 0.17 WSR1a 62 110 40 7.7 1.5 0.9 2.9 0.38 BG10 11 22 10 1.7 0.6 0 0.7 0.16 WSR1b 66 116 40 7.9 1.7 1 3.7 0.55 BG11 51 92 40 7 1.1 0.7 2.3 0.32 WSR1c 60 108 40 7.7 1.3 0.9 3.3 0.49 BG12 17 32 10 2.7 0.7 0 1.2 0.18 WSR1d 68 123 50 9.3 1.8 1.2 3.2 0.5 BG16 68 45 10 1.3 0.6 0 0.8 0.15 WSR1e 71 131 50 9.8 2.2 1.2 3.3 0.48 BG16 10 18 0 1.6 0.5 0 1 0.2 WSR1f 79 148 60 10.1 2.2 1.2 3.3 0.46 BG17 14 25 10 2.4 0.5 0 1.6 0.23 WSR1g 47 90 40 7.6 1.7 1 3 0.48 BG18 23 44 20 3.4 0.5 0 1.2 0.2 WSR1h 71 129 50 9.1 1.4 1 3.1 0.49 BG19 192 236 60 6 1.8 0 0.8 0.1 WSR1i 39 73 30 6.4 1.3 0.9 2.9 0.4 BG17 3 6 0 0.9 0.5 0 0.3 0.09 WSR1j 27 50 20 4.9 1 0.7 3.5 0.55 BG19 42 79 30 5.8 1.1 0.7 2.1 0.29 WSR1k 75 140 60 10.3 1.9 1.3 3.9 0.61 BR1a 39 72 30 5.6 1.1 0.7 3.5 0.53 WSR2 56 104 40 7.7 1.7 0.9 3.1 0.5 BR1b 44 90 30 6.6 1.5 0.8 3.3 0.46 WSR3 18 37 20 4.2 1 0.8 2 0.3 BR1c 31 56 20 4.6 0.9 0.7 2.8 0.41 WSR4 63 115 40 7.8 1.4 0.8 3 0.48 BR1d 53 96 40 7 1.4 0.8 2.8 0.41 WSR5 61 109 40 7.7 1.6 0.9 3 0.49 BR1e 72 120 50 7.9 1.7 0.8 3.3 0.47 WSR6 71 135 60 10.7 2.3 1.3 4.6 0.63 BR2 51 94 40 6.8 1.1 0.7 2.9 0.42 WSR7 47 88 30 6.4 1.3 0.9 2.6 0.36 BR3 29 51 20 4.1 1 0.5 2.8 0.43 WSR8 8 17 10 2.3 0.4 0 1 0.14 BR4 60 103 40 7 1.4 0.7 3.1 0.43 WSR9 8 19 10 2.5 0.4 0 1.2 0.15 BR5 48 85 30 6.4 1.6 0.8 3 0.46 WSR10 54 89 40 6.4 1.1 0.7 2.5 0.34 BR6a 32 93 40 10.1 2 1.1 4.8 0.51 WSR11 51 95 40 6.7 1.9 0.7 2.5 0.35 BR6b 45 83 30 5.7 1 0.7 2.3 0.32 WSR12 19 28 10 1.6 0.8 0 0.8 0.16 BR6c 61 108 40 6.8 1.4 0.6 2.3 0.31 WSR13 35 63 20 4.3 0.9 0.5 1.4 0.21 BR7 56 98 30 5.6 0.8 0.6 2.6 0.41 WSR14 7 13 0 1.8 0.5 0 1.3 0.19 BR8 23 41 10 2.7 0.8 0.5 2.5 0.46 WSR15 11 15 0 0.6 0.5 0 0.2 0.06 BR9 9 16 10 1.6 0.4 0 0.9 0.17 WSR16 7 13 0 1.6 0.6 0 0.8 0.12 BR10 3 5 0 0.8 0.4 0 0.5 0.13 WSR19 3 6 0 0.8 0.3 0 0.7 0.1 BR11 15 23 10 1.6 0.4 0 1 0.14 WSR17 12 20 0 1.2 0.6 0 0.6 0.13 BR12 7 12 0 1.3 0.6 0 0.7 0.13 WSR18 3 6 0 0.9 0.5 0 0.3 0.09 BR13 18 22 10 1.6 0.4 0 0.7 0.17 WSR17 15 25 10 1.5 0.5 0 1.1 0.16 BR11 16 28 10 2.1 0.8 0 1.4 0.23 WSR18 4 7 0 1 0.2 0 0.3 0.09 BR12 11 20 10 1.8 0.6 0 1.1 0.17 BR13 10 18 10 1.9 0.7 0 0.8 0.15 BR14 8 12 0 1.3 0.4 0 0.6 0.09 BR15 126 146 30 3.5 1 0 1.1 0.16 BR16 5 10 0 1.2 0.5 0 0.4 0.07 BR17 15 24 10 1.5 0.4 0 0.7 0.11 WR1 56 30 5.6 0.8 0.6 2.6 0.41 WR2 87 122 90 19 5 2.4 3.3 0.38 WR3 46 80 23.7 5.1 3.8 3.7 0.45 WR4 67 50 10.2 2.2 1.3 3.6 0.53 WR5 64 112 50 8.4 1.7 1.2 3.7 0.51 WR6 59 107 40 8.5 2.4 1.1 3.8 0.56 WR11 21 32 10 2.3 0.6 0 0.8 0.1 WR10 5 9 0 1.7 1.3 0 0.7 0.08 WR9 9 19 10 3 0.7 0 1.1 0.15 WR8 3 8 0 1.6 0.5 0 0.6 0.1 WR7 12 19 0 1.2 0.9 0 1 0.16 102 Table 3.3: Light mineral composition of bentonites in the lower Middle Campanian. Averages of samples for bentonites of the Ardmore succession are provided.

Quartz K-spar Plag Quartz K-spar Plag Ardmore Rhyolite 1 (8) Gammon Bentonites Average 25.33 42.67 32.00 RED1-3 very few phenocrysts StDev 3.50 2.42 2.83 RED1-4 very few phenocrysts RED1-7 very few phenocrysts Ardmore Rhyolite 2 (10) RED1-8 very few phenocrysts Average 31.75 37.5 30.75 RED1-9 48 40 12 StDev 8.73 3.70 5.32 RED1-10 40 46 14 WAL3-B2 very few phenocrysts Ardmore Andesite 1 WR24 46 30 24 subgroup 1 (10) WR23 44 40 16 Average 17.70 41.50 40.80 WR22 very few phenocrysts StDev 2.11 2.88 2.15 WR21 very few phenocrysts WR20 50 30 20 Subgroup2 (2) Average 23 36 41 First stringer StDev 4.243 0 4.243 BG18 31 42 27 BR30 6 14 5 Subgroup 3 (4) WsR32 29 49 22 Average 18.25 26.75 55 BG17 31 49 13 StDev 0.957 0.957 0.816 BG18 53 90 48 WsR31 18 57 25 Subgroup 4 (3) Average 26.33 25 48.67 Couplet StDev 2.309 0 2.309 BG15 50 32 18 BG16 39 41 20 Subgroup 5 (4) WsR24 13 9 4 Average 16 39 45 WsR26 38 45 17 StDev 1.732 1 2.646 Red3-99-8 56 33 10 Red3-99-9 39 42 18 Subgroup 6 (2) Pem1-20 34 56 18 Average 16.5 34 49.5 StDev 2.121 2.828 0.707

Subgroup 7 (2) Average 17.5 24 58.5 StDev 2.121 1.414 3.536

Andesite Ardmore 2 RED2-98-08 8 48 44 103 Table 3.4: Microprobe analysis of bentonites of the lower Middle Campanian. The Ardmore succession, the bentonite couplet and the first stringer have biotites. Representative samples of the Ardmore succession are provided.

Sample ID SiO2 TiO2 Al2O3 FeO MgO CaO Na2O K2O Fe2+ Fe3+ Mg Mg# MgO Representative Ardmore Samples RED2-98-08 38.374 5.481 6.119 10.788 15.071 0.015 0.592 8.726 8.091 2.997 9.090 0.529 15.071 RED2-98-08 37.150 5.319 5.438 11.101 15.083 0.015 0.666 8.955 8.326 3.084 9.097 0.522 15.083 RED2-98-08 35.658 4.866 5.229 11.148 13.734 0.004 0.534 8.483 8.361 3.097 8.283 0.498 13.734 RED2-98-08 32.404 4.825 6.100 11.526 14.860 0.007 0.530 8.685 8.645 3.202 8.963 0.509 14.860 RED2-98-08 34.404 5.399 4.090 11.532 14.483 0.000 0.538 8.769 8.649 3.204 8.735 0.502 14.483 RED2-98-08 34.210 5.592 4.207 11.602 15.613 0.015 0.664 9.057 8.701 3.223 9.417 0.520 15.613 RED2-98-08 35.379 5.460 4.997 11.914 14.842 0.008 0.577 9.102 8.935 3.310 8.952 0.500 14.842 RED2-98-08 33.436 5.125 5.031 12.492 13.754 0.000 0.555 8.579 9.369 3.471 8.296 0.470 13.754 BS28 35.665 5.059 12.535 12.328 14.934 0.029 0.775 8.079 9.246 3.425 9.007 0.493 14.934 BS28 34.690 5.296 12.065 11.111 14.974 0.019 0.670 7.709 8.333 3.087 9.031 0.520 14.974 BS28 35.922 5.346 12.977 12.091 15.261 0.028 0.665 8.285 9.068 3.359 9.204 0.504 15.261 BS28 33.756 5.042 11.909 11.307 14.416 0.036 0.610 7.673 8.480 3.141 8.695 0.506 14.416 BS28 32.947 4.772 11.447 11.568 14.229 0.063 0.597 7.684 8.676 3.214 8.582 0.497 14.229 BS28 34.336 5.264 12.315 11.292 14.950 0.000 0.666 7.974 8.469 3.137 9.017 0.516 14.950 BS28 35.564 4.723 11.822 14.053 14.248 0.013 0.701 8.191 10.540 3.904 8.593 0.449 14.248 BS28 35.656 5.244 12.157 12.291 15.134 0.013 0.625 8.313 9.218 3.415 9.128 0.498 15.134 WsR01 39.226 5.603 13.228 11.222 13.597 0.088 0.481 7.937 8.416 3.118 8.201 0.494 13.597 WsR01 35.920 5.550 11.429 11.285 14.189 0.088 0.518 7.643 8.464 3.135 8.558 0.503 14.189 WsR01 35.762 5.589 12.486 11.897 14.320 0.046 0.509 8.272 8.923 3.305 8.637 0.492 14.320 WsR01 39.630 5.917 12.364 11.918 14.681 0.035 0.510 7.870 8.938 3.311 8.855 0.498 14.681 WsR01 38.763 5.814 12.391 12.086 14.476 0.067 0.466 7.863 9.065 3.358 8.731 0.491 14.476 WsR01 36.485 5.729 12.195 12.229 14.394 0.119 0.515 7.759 9.172 3.398 8.682 0.486 14.394 WsR01 39.608 5.639 12.970 12.241 14.508 0.074 0.526 8.450 9.181 3.401 8.750 0.488 14.508 WsR02 38.242 4.391 11.663 10.643 12.044 0.148 0.539 7.602 7.982 2.957 7.264 0.476 12.044 WsR02 40.002 4.826 15.157 11.223 12.072 0.151 0.466 7.269 8.417 3.118 7.281 0.464 12.072 WsR02 36.321 5.348 12.961 11.500 14.086 0.104 0.566 8.271 8.625 3.195 8.496 0.496 14.086 WsR02 35.655 5.696 12.490 11.826 14.655 1.361 0.693 8.157 8.870 3.286 8.839 0.499 14.655 WsR02 37.673 5.966 12.851 11.958 14.666 0.051 0.607 8.406 8.969 3.322 8.846 0.497 14.666 WsR02 36.863 5.313 11.937 12.020 14.050 0.087 0.503 8.525 9.015 3.339 8.474 0.485 14.050 WsR02 38.191 5.572 12.229 12.746 14.596 0.009 0.516 8.617 9.559 3.541 8.803 0.479 14.596 WsR03 39.580 5.562 12.972 11.575 13.969 1.110 0.551 8.108 8.681 3.216 8.425 0.493 13.969 WsR03 39.096 5.903 12.032 12.062 15.245 2.125 0.561 9.026 9.047 3.351 9.195 0.504 15.245 WsR03 39.066 5.705 12.916 12.338 14.489 0.037 0.464 8.731 9.253 3.428 8.739 0.486 14.489 WsR03 39.042 5.684 12.349 12.338 14.699 0.011 0.578 8.657 9.253 3.428 8.866 0.489 14.699 WsR03 38.751 5.713 13.684 12.396 13.837 0.043 0.541 8.298 9.297 3.444 8.346 0.473 13.837 WsR03 37.908 5.954 12.492 12.420 14.382 0.049 0.532 8.683 9.315 3.451 8.674 0.482 14.382 WsR03 39.174 5.633 13.442 12.840 13.925 0.032 0.480 8.564 9.630 3.567 8.399 0.466 13.925 WsR03 37.741 5.453 13.021 13.760 14.361 0.014 0.516 8.983 10.320 3.823 8.662 0.456 14.361 WsR13 39.462 5.311 13.071 11.382 13.680 0.161 0.544 8.183 8.537 3.162 8.251 0.491 13.680 WsR13 41.843 5.490 11.453 11.444 13.924 0.404 0.508 8.119 8.583 3.179 8.398 0.495 13.924 WsR13 36.273 5.324 11.857 11.532 13.783 0.108 0.564 7.852 8.649 3.204 8.313 0.490 13.783 WsR13 39.973 5.599 12.387 11.560 14.245 0.064 0.564 8.389 8.670 3.212 8.592 0.498 14.245 WsR13 38.548 5.361 10.925 11.619 12.101 0.153 0.509 7.544 8.714 3.228 7.299 0.456 12.101 WsR13 40.773 5.579 12.192 11.732 14.731 0.064 0.659 8.442 8.799 3.259 8.885 0.502 14.731 WsR13 40.286 5.670 12.478 12.123 15.002 0.037 0.581 8.960 9.092 3.368 9.048 0.499 15.002 104 Sample ID SiO2 TiO2 Al2O3 FeO MgO CaO Na2O K2O Fe2+ Fe3+ Mg Mg# MgO Bentonite Couplet RED2-98-26 33.859 4.806 4.817 18.127 11.418 0.011 0.480 9.200 13.595 5.036 6.887 0.336 11.418 RED2-98-26 34.166 5.331 11.598 20.916 9.152 0.039 0.519 7.697 15.687 5.811 5.520 0.260 9.152 RED2-98-26 36.118 5.279 12.893 12.521 14.820 0.055 0.668 7.806 9.391 3.479 8.938 0.488 14.820 RED2-98-26 34.773 4.511 12.369 18.241 11.606 0.064 0.513 8.310 13.681 5.068 7.000 0.338 11.606 Pem1-20 39.289 4.811 13.759 12.051 14.573 0.055 0.613 8.401 9.038 3.348 8.790 0.493 14.573 Pem1-20 44.472 4.734 10.516 16.802 10.242 0.088 0.500 8.381 12.602 4.668 6.177 0.329 10.242 Pem1-20 38.228 4.595 10.706 17.370 8.514 0.005 0.411 7.623 13.028 4.826 5.135 0.283 8.514 Pem1-20 38.250 4.699 11.960 17.500 12.180 0.020 0.442 9.002 13.125 4.862 7.346 0.359 12.180 Pem1-20 38.249 4.699 11.964 17.501 12.181 0.020 0.442 9.002 13.126 4.862 7.347 0.359 12.181 Pem1-20 39.120 4.905 11.862 17.761 12.169 0.019 0.479 9.018 13.321 4.934 7.340 0.355 12.169 Pem1-20 36.487 4.924 11.100 18.094 9.061 0.042 0.416 7.862 13.571 5.027 5.465 0.287 9.061 Pem1-20 39.154 5.015 12.297 18.752 11.355 0.011 0.470 9.055 14.064 5.210 6.849 0.327 11.355 Pem1-20 37.848 4.941 12.238 18.837 11.219 0.018 0.475 8.847 14.128 5.233 6.767 0.324 11.219 WsR26 38.300 5.862 11.364 17.046 10.874 0.040 0.440 7.770 12.785 4.736 6.559 0.339 10.874 WsR26 37.210 6.243 12.033 17.498 11.907 0.050 0.490 8.445 13.124 4.861 7.182 0.354 11.907 WsR26 39.960 5.434 11.070 17.388 11.169 0.090 0.490 7.954 13.041 4.831 6.736 0.341 11.169 WsR26 39.960 5.434 11.070 17.388 11.169 0.090 0.490 7.954 13.041 4.831 6.736 0.341 11.169 WsR26 37.720 5.653 9.832 17.334 11.300 0.050 0.420 8.088 13.000 4.816 6.815 0.344 11.300 WsR26 40.000 5.817 12.221 11.028 15.326 0.020 0.780 7.470 8.271 3.064 9.244 0.528 15.326 WsR26 39.280 5.872 12.841 18.135 11.283 0.020 0.450 8.378 13.601 5.038 6.805 0.333 11.283 WsR26 39.200 5.094 9.642 16.149 10.327 0.080 0.420 7.444 12.112 4.487 6.229 0.340 10.327

First Stringer BG19 39.167 3.887 13.644 9.639 10.695 0.023 0.323 7.600 7.229 2.678 6.451 0.472 10.695 BG19 35.817 4.505 10.996 11.394 11.397 0.033 0.443 8.624 8.546 3.166 6.874 0.446 11.397 105 Revised stratigraphy of the Lower Pierre Shale (Campanian): Implications for the tectonic and eustatic controls on facies distributions with description of the new Walhalla and Chamberlain members

ABSTRACT The lower Pierre Shale represents a time of significant changes in the Cretaceous Western Interior Seaway, resulting from complex interactions of tectonism and eustatic sea level changes. The complex nature of this interval has been masked by the recognition of a single stratigraphic unit, the “Sharon Springs Member”, which has been recognized from Kansas to South Dakota, suggesting widespread stability of a single facies. Detailed stratigraphic analysis of this interval indicates a much more complex distribution of facies. The recognition of two new members, the Walhalla Member and the Chamberlain Member, and the revision of existing member terminology clarifies the facies distribution and allows for evaluation of tectonic and eustatic processes on the changes in the basin. The newly defined Walhalla Member is restricted to the northern part of the basin and represents tectonically influences sequences. These sequences are a response to rapid subsidence of the axial basin and the Williston Basin corresponding to tectonic activity along the Absoroka Thrust in Wyoming. Overlying the Walhalla Member, the facies exhibit north to south trending facies belts represented by the Mitten Black Shale Member in the west and the Sharon Springs and Chamberlain members to the east. These members represent fourth-order eustatic sea level flucatuations within the third-order Claggett Cycle. Recognition of these complex facies patterns and distal sequences is necessary to understand complete basin dynamics. 106 Revised stratigraphy of the Lower Pierre Shale (Campanian): Implications for the tectonic and eustatic controls on facies distributions with description of the new Walhalla and Chamberlain members

INTRODUCTION The lower Pierre Shale (Campanian, Cretaceous) represents an interval of significant changes in the Cretaceous Western Interior Seaway, corresponding to the Claggett Cycle of Kauffman (1984) and dated at 81-75Ma (Fig. 4.1, Obradovich, 1993). The interaction of tectonic events in the Sevier Orogeny and eustatic sea level changes resulted in changes in sedimentation and ecology that ultimately resulted in the demise and turnover of several groups of animals (Russell, 1993). The complex nature of this interval is masked by the present recognition of a supposedly single stratigraphic unit, the “Sharon Springs Member”, which has been extended from Kansas to South Dakota, with a duration of approximately 5 million years. The correlation of this member over such a large area appears to suggest widespread stability of a single facies. However, detailed stratigraphic analysis of the “Sharon Springs Member” throughout the basin indicates a much more complex distribution of facies. The recognition of two new members and the revision of existing member terminology clarifies this facies distribution and provides the detail necessary for an evaluation of the impact of tectonic and eustatic processes on the changes in the basin. The lower Pierre Shale represents distal sedimentation during multiple sea level

fluctuations within the overall 3rd order cycles, and results from both tectonic and eustatic controls. The lower Pierre Shale was deposited within the eastern part of the axial basin, across the east median hingeline and onto the eastern stable platform. Broad structural zones were created in response to the tectonic forces to the west (Fig. 4.2, Kauffman and Caldwell, 1993). Superimposed on these zones, structural weaknesses in the crust caused local faulting, complicating sedimentation patterns (Rice and Shurr, 1983). Within the eastern part of the basin, structural weaknesses allowed for movement along the eastern 107

Figure 4.1: Generalized stratigraphy of the Late Cretaceous including primary formations of the Santonian- Maastrichtian. The study interval is the Lower Pierre Shale, Baculites sp. (smooth) through Baculites scotti. 108

Figure 4.2: During the Middle Campanian, several volcanic centers were active in western North America (Gray areas), resulting from the subduction of the Kula and Farallon Plate (inset). The Boulder Batholith and Little Elkhorn Mountains were very active during this time. The subduction also created the Sevier Orogeny, with the Absoroka Thrust being active at about 80Ma. The compressive stresses resulted in the back-arc foreland basin with four distinct facies belts. The Williston Basin in eastern North Dakota resulted from a response to compressive stresses coupled with structural weaknesses in the subsurface. 109 stable platform, particularly in North Dakota, where ancient weaknesses of the Williston Basin allowed for increased subsidence associated with thrusting stresses (Fig. 4.2, Russell, 1993). In the southern part of the basin, the broad structural zones are not as well defined. This is partly a result of lower tectonic activity in the southern part of the basin and partly due to the basin opening to the south into the open ocean. The northern and southern parts of the basin are at least partially isolated from each other by a structural high associated with the Transcontinental Arch (Fig. 4.2). Sedimentation patterns indicate that the Transcontinental Arch was a prominent positive feature throughout the Late Cretaceous, and may have been subaerially exposed during periods of sea level lowstand.

HISTORICAL DESCRIPTIONS OF THE STRATIGRAPHY OF THE LOWER PIERRE SHALE The Pierre Shale, one of the most widespread formations in the Western Interior, was first described near Fort Pierre, South Dakota along the Missouri River for exposures of dark gray shale (Hayden, 1862). The Pierre Shale exhibits vertical variation and has been divided into several members (Fig. 4.3a), which vary across the region and tend to be individually local in extent. The Pierre Shale can be divided into two parts, particularly in the northern part of the basin (Kauffman, 1984). The division between the upper and lower Pierre Shale is marked by the contact between the regressive facies of the Redbird Silty Member and the equivalent but more distal Gregory and Crow Creek members below and the unnamed shale member and the more distal DeGray Member above. The lower Pierre Shale is equivalent to the Claggett Shale and Judith River Formation in Montana (Fig. 4.3a, Dyman et al., 1994). Within the lower Pierre Shale, the Sharon Springs Member, the dominant member in this interval, was first defined in Logan County, Kansas, as the bituminous shale above the calcareous Niobrara Formation at McAllaster Buttes (Elias, 1931),and represents the Baculites sp. (smooth form) to Baculites scotti ammonite range zone. The Sharon Springs Member was correlated on the basis of lithology to South Dakota and Colorado (Fig. 4.3a). 110

Figure 4.3a: Generalized stratigraphy of the lower Pierre Shale, as it is currently recognized in the literature.

Figure 4.3b: Stratigraphy of the lower Pierre Shale, as modified in this paper. The Sharon Springs Member, Gammon Ferruginous Member and Pembina Member are redefined to include the Walhalla Member and teh phosphate nodule and gray shale facies of the Chamberlain Member, two new members.

In eastern South Dakota the so-called “Sharon Springs Member” is the bituminous shale between the calcareous Niobrara Formation and below the calcareous Gregory Member and represents the Baculites obtustus to Baculites asperformis ammonite range zones. In the Black Hills, the “Sharon Springs Member” was originally described as the bituminous shale above the Niobrara Formation and below the Redbird Silty Member and represented the Baculites sp. (smooth form) to Baculites asperformis ammonite range zones. However, the “Sharon Springs Member” was subsequently restricted to only the most highly bituminous shale, within the Baculites obtusus to Baculites mclearni ammonite range zones (Gill and 111 Cobban, 1966). Below the “Sharon Springs Member” is the Gammon Ferruginous Member and above is the Mitten Black Shale Member (Fig. 4.3a). The Gammon Ferruginous Member and the Mitten Black Shale Member were distinguished by the presence of abundant siderite concretions. All three members in the Black Hills have been considered correlative to the Sharon Springs Member in Kansas (Gill and Cobban, 1966). In North Dakota, the Pembina Member has also been considered equivalent to the Sharon Springs Member (Gill and Cobban, 1965). The “Sharon Springs Member” in South Dakota and the Pembina Member in North Dakota actually differ significantly from the Sharon Springs Member in Kansas in that several thick bentonite layers are present in South Dakota and North Dakota that are not present in Kansas. The Ardmore bentonite succession marks the base of the “Sharon Springs Member” in South Dakota and the Pembina Member in North Dakota (Spivey, 1940). Above the Ardmore bentonite succession are as many as twelve additional bentonites, ranging from 1cm to 25cm thick. These bentonites have unique compositions that can be used to fingerprint and correlate them across the basin (Chapter 3). Stratigraphic analysis can be combined with bentonite stratigraphy and ammonite biostratigraphy for regional chronostratigraphic correlations. METHODS Detailed stratigraphic sections were measured and described for outcrops of the “Sharon Springs Member” in Kansas, Colorado, South Dakota and Wyoming, and for the Pembina Member in North Dakota and the Gammon and Mitten Black Shale members in South Dakota and Wyoming. Published sections and subsurface information were used to supplement the field investigations for Kansas, Colorado, Nebraska, South Dakota and North Dakota. The stratigraphic characteristics of the members were compared to the type locality of the Sharon Springs Member and individual facies of the lower Pierre Shale were defined based on lithology. These facies were then used to redeine the Sharon Springs Member, the Pembina Member and the Gammon Ferruginous Member and to define two new members, 112 the Chamberlain and Walhalla members. Correlations were based on bentonite chemostratigraphy, as described in chapters 2 and 3 and ammonite biostratigraphy where the information is available. Ammonites, however, are relatively rare in the organic-rich shales assigned to the “Sharon Springs Member”, and are frequently confined to concretionary horizons. Outcrops of the lower Pierre Shale were examined in from eastern North Dakota to western Kansas at 13 outcrops (Fig. 4.2). Additional information was obtained from published measured sections and subsurface data: eastern South Dakota localities (Gill and Cobban, 1966); Nebraska subsurface (DeGraw, 1975); Redbird, Wyoming (Gill and Cobban, 1966); northern Black Hills (Wing, 1940); southern Black Hills (Spivey, 1940); Kansas (Elias, 1931; Gill et al., 1972); eastern and central South Dakota(Gries and Rothrock, 1941; Searight, 1937); and Colorado (Izett et al., 1971; Landis, 1959). RESULTS A re-evaluation of the member distribution in the lower Pierre Shale provides a much more complete understanding of the dynamics of the Western Interior Basin during the Late Campanian. The term “Sharon Springs Member” has been used in different localities as a catch-all phrase for all organic-rich shales that did not fit into one of the other member definitions. The strata referred to as the type Sharon Springs Member actually correlate, in part, to strata equivalent to the underlying Gammon Ferruginous Member and the overlying Mitten Black Shale Member (Fig. 4.3). Ironically, the strata referred to as the “Sharon Springs Member” in the Black Hills are not present at the type locality of the Sharon Springs Member in Kansas, due to an unconformity within the member. For this reason, as described below, the term “Sharon Springs Member” herein is restricted to the upper part of the strata originally described as the Sharon Springs Member in western Kansas, and a new member name, the Walhalla Member, is defined for the black shale unit in South Dakota, which had formerly been called the “Sharon Springs Member” (Fig. 4.3). The Walhalla Member is present throughout South Dakota and eastern North Dakota but had been mapped as the 113 lower part of what has been called the Pembina Member. The Pembina Member of North Dakota is split into the Walhalla Member, the Sharon Springs Member and another new unit the Chamberlain Member (Fig. 4.3b). The Chamberlain Member is a unit of organic- rich shale with abundant phosphate nodules that is present above the Sharon Springs Member (sensu stricto) throughout Kansas, eastern South Dakota and North Dakota. Gammon Ferruginous Member

The term Gammon Ferruginous Member was applied to the approximately 300m thick interval of dark, friable, soft olive-black shale with abundant tabular ferruginous concretions above the Niobrara Formation and at the base of the Pierre Shale along the Gammon Prong in the northern Black Hills (Rubey, 1930). This member has been recognized around the Black Hills (Gill and Cobban, 1966) and detailed representative measured sections are shown in Figure 4.4. The Gammon Ferruginous Member was deposited during the Baculites sp. ammonite range zone (Gill and Cobban, 1966). The Gammon Ferruginous Member thins significantly from eastern South Dakota to western Wyoming over a distance of approximately 40 miles. At Redbird, the member is 5.5 meters thick (reported as 10m by Gill and Cobban, 1966) and at Slurp Flats the member is approximately 10m thick. Bentonites within this interval are present at both Redbird and Slurp Flats (Fig. 4.4) and indicate that the unit thins to the west due to condensation and not due to erosion and truncation (Chapter 3). While the ferruginous concretions are restricted to the Black Hills region, the shale facies is present in Kansas, where it has been previously referred to the “dark soft shale facies” of the Sharon Springs Member (Gill et al., 1972). The lower Sharon Springs Member, as it was originally described, thus includes differentiated strata equivalent to the Gammon Ferruginous Member (Elias, 1931), and is re-assigned to the Gammon Ferruginous Member in this report (Fig. 4.4). Although ironstone concretions are not present, several layers of shale are iron-stained yellowish-orange and some scattered fossiliferous limestone and siderite concretions are present. Gill and Cobban (1972) suggested that this was pre-Tertiary 114

Figure 4.4: Representative meaured sections of the Gammon Ferruginous Member in the southern Black Hills and western Kansas. Section localities are marked in Fig. 4.2. 115 oxidation. The most complete section of the Gammon Ferruginous member in Kansas is along Burris Draw southwest of Elkader in Wallace County, Kansas. Two meters of the facies are exposed at McAllaster Buttes, where it is unconformable with the overlying buttress-weathering shale facies of the Sharon Springs Member, as described below. No complete section of the Gammon Ferruginous Member is present in outcrop in Kansas. As described in the literature, this unit is 30 meters thick in western Kansas at Burris Draw (Elias, 1931; Gill et al., 1972); however, only 15 meters was measured during this study. The top of the member was not present at Burris Draw, where a thick iron-stained shale unit marking an unconformity is present. The majority of those 15 meters is exposed along Burris Draw. Near the tops of the bluffs, increased vegetation made recognition of shale units impossible. The Gammon Ferruginous Member is not present in eastern South Dakota and North Dakota, where a prominent unconformity at the top of the Niobrara Formation indicates a period of erosion during the time of deposition of this facies to the west. At Redbird, Wyoming this unconformity is less evident, and two tabular siderite layers mark the top of the member. At Igloo, South Dakota, the unconformity is well developed and overlain by a channel sandstone. Regional Unconformity

A well-developed unconformity is present across the region at the top of the Gammon Ferruginous Member or, where the Gammon Ferruginous Member is not present, at the top of the Niobrara Formation. The unconformity can be recognized at all localities from North Dakota to Kansas, and appears to represent subaerial exposure in some parts of the basin and subaqueous erosion in other parts of the basin, as described below. In northwestern Nebraska, the unconformity is well developed and subaerial exposure is suggested by a pattern of fluvial channels apparently from rivers flowing both north and south off the transcontinental arch (DeGraw, 1975). The unconformity is also well developed at Igloo, South Dakota, some localities along the Missouri River north and south of 116 Chamberlain, South Dakota and at Walhalla, North Dakota. The unconformity is heavily iron-stained with jarosite. At Walhalla, North Dakota, the unconformity is represented by karsting at the top of the Niobrara marl as well as a 7cm iron-stained, selenite-rich brecciated hash layer. At Redbird, Wyoming, the unconformity is not well developed. However, two tabular siderite concretionary layers are present near the top of the Gammon Ferruginous Member at this locality, which may correlate with this unconformity. Walhalla Member

The Walhalla Member is herein defined as the bituminous shale facies of the lower Pierre Shale and includes strata that have been traditionally assigned to the lower part of the Pembina Member in North Dakota (Gill and Cobban, 1965), the “Sharon Springs Member” in the Black Hills (Moxon et al., 1939) and northwestern Nebraska (DeGraw, 1975), and the “lower Sharon Springs Member” in eastern South Dakota (Gries and Rothrock, 1941). Although the strata have been referred to the “Sharon Springs Member” in South Dakota and the Pembina Member in North Dakota, strata of equivalent age and lithology are not present in Kansas, where the Sharon Springs Member was originally defined. The Walhalla Member is well exposed in the Pembina Gorge near Walhalla, North Dakota, where it has been traditionally referred to the lower part of the Pembina Member (Fig. 4.5, 4.6). This interval is approximately 2-3m of black, bituminous clay-shale alternating with as many as 18 white-yellow bentonites (Leonard, 1904). The Walhalla Member can be correlated to eastern South Dakota and the Black Hills, where equivalent strata previously have been assigned to the Sharon Springs Member (Fig. 4.7, 8). The thickness of the member is highly variable, ranging from less than 1m to 24m and the unit is locally missing, particularly in eastern South Dakota. The variation in thickness is primarily due to several unconformities within or at the top of the unit that remove several meters of section at some localities. This member is the most organic-rich facies of the lower Pierre Shale. This shale is black, weathering to dark brown, appearing “pulpy” and very hard, making it more resistant 117

Figure 4.5: Representative measured sections of the Walhalla Member. Bentonites of the Ardmore succession are present at all localities at the base of the Walhalla Member. Wasserburger Ranch is the type locality of the Ardmore bentonite succession. 118

Figure 4.6: The Walhalla Member overlying the Niobrara Formation in the Pembina Gorge near Walhalla, North Dakota. The yellow bands in the Walhalla Member are bentonites of the Ardmore bentonite succession. Unit two of the Walhalla Member is partly covered by vegetation.

Figure 4.7: Stratigraphy of the lower Pierre Shale near Chamberlain, South Dakota. The Whalla Member is condensed at this locality. The Sharon Springs Member is approximately 5m thick and the Chamberlain Member, including the phosphatic unit and grya shale unit, is also approximately 5m thick. 119

Figure 4.8: The Ardmore bentonite succession of the Walhalla Member. Yellow bands on the outcrop are the bentonites. A: The Ardmore bentonite succession at Wasserburger Ranch, the type locality of the Ardmore bentonite succession. B: The Ardmore bentonite succession at Redbird, Wyoming. 120 to weathering than the other facies. Fragments of plants can be discerned in the shale and fish scales are very abundant. Weathered exposures of the bituminous shale do not support vegetation due to toxic materials such as selenite in the shale. The Walhalla Member can be divided into two units, a lower unit that is hard and highly bituminous and includes the Ardmore bentonite succession (Chapter 2) and an upper unit of bituminous, hackly shale with several thin bentonites (Chapter 3). The Walhalla Member is bounded by unconformities. The regional unconformity, described above, marks the base of the member. At the top of the member, another unconformity is locally well developed. The member overlies the Gammon Ferruginous Member in the Black Hills and the Niobrara Formation in North Dakota and eastern South Dakota. Well-preserved specimens of Baculites are rare, being found only in concretions. Within the concretions, Baculites obtusus are present in association with the Ardmore bentonite succession in Unit 1 and Baculites mclearni is found in Unit 2. Other invertebrates are rare in this facies. Vertebrates, including mosasaurs, plesiosaurs, turtles, sharks and fish, are common in the bituminous shale facies, particularly in association with the Ardmore succession. Vertebrate specimens almost inevitably have selenite throughout the bone, often making the specimens nearly unidentifiable. Unit 1: Above the basal unconformity of the Walhalla Member is the Ardmore bentonite succession. This thick succession of bentonites was first described near the town of Ardmore, South Dakota for the 1-meter thick bentonite at the base of the section that was mined commercially (Spivey, 1940). Overlying the 1-meter thick bentonite are as many as eight additional bentonites, each approximately 25cm thick (Fig. 4.5). The bentonites are bluish gray when fresh and alter to grayish yellow, standing out in stark contrast against the black shale. Around the southern Black Hills, the Ardmore succession has a very similar pattern at all outcrops. Exposures of this unit can be seen around the southern Black Hills from Redbird, Wyoming to Buffalo Gap, South Dakota (Fig. 4.5). Further to the east, this 121 unit is exposed at several outcrops along the Missouri River between Chamberlain and Yankton, South Dakota and at limited outcrops in eastern North Dakota (Fig. 4.5). Unit 1 is also present in northwestern Nebraska and northeastern Colorado, where it is primarily recognized in the subsurface and in limited outcrops. In the eastern sections, the unit is discontinuous, and locally it is completely removed by erosion associated with an overlying unconformity. At the eastern sections, the bentonites of the Ardmore succession are separated into individual beds that range in thickness from less than 1cm to 17cm thick. Shale that is not present around the southern Black Hills separates these thin bentonites, making this interval thicker where it is complete at the eastern sections (Chapter 2). The thickness of the interval ranges from 0 to 300cm. The base of Unit 1 is marked by the regional unconformity, described above. Deposition of Unit 1 began earlier in the Black Hills and in eastern North Dakota than in eastern South Dakota as evidence by the presence of bentonites from five volcanic events that are preserved at the base of Unit 1in the Black Hills and North Dakota and are not present in eastern South Dakota. Higher volcanic event layers are distributed across the entire basin. In addition to the basal unconformity at the eastern localities in South Dakota and North Dakota, two other unconformities are well developed in Unit 1. The first unconformity is in the middle of the unit and locally removes as much as 1m of section and three bentonites (Fig. 4.4). The second unconformity is at the top of the section and also locally removes as much as 1m of section and three bentonites (Fig. 4.4). At Yankton, South Dakota, the eastern-most locality, these two unconformities are superimposed, removing 2m of section. In the Black Hills, the first unconformity is not well developed and the section may be conformable. At Buffalo Gap, a thin layer of limestone concretions, present at this level, may mark the unconformity. These unconformities are locally indicated by broad, shallow sandy channels or by iron-stained brecciated shale. At the top of Unit 1 in the Black Hills, septarian concretions are locally present, suggesting sediment starvation, but no evidence 122 of erosion is present. The unconformity at the top of Unit 1 at the eastern localities is correlative with the unconformity at the top of Unit 2 in the Black Hills, and unit 2 is missing to the east (Fig. 4.4). Unit 2: Unit 2 is present in the southern Black Hills and in eastern North Dakota (Fig. 4.5). This interval is locally absent, where it is removed by erosion marked by an unconformity at the top of the Ardmore succession, such as at Wallace Ranch, but ranges up to approximately 17m at Redbird, Wyoming and Walhalla, North Dakota. This unit differs from the Ardmore succession in that the shale is slightly more hackly than in the Ardmore succession and only two thick bentonites are present, although as many as nine thin bentonites (0.25-3cm) occur within this interval (Chapter 3). The two thick bentonites make a closely spaced couplet in the eastern part of the Black Hills, but this couplet is more widely spaced and not as easily recognized stratigraphically in the western Black Hills or outside this region. These two bentonites have a distinct mineralogy and chemistry that make them recognizable (Chapter 3). Most of this interval does not have any concretions. At Buffalo Gap, a thin layer of siderite concretions and a layer of cone-in-cone concretions are present near the top of the unit. The unit is incomplete at this locality. An iron-stained brecciated layer with abundant fish remains marks an unconformity at the top of the unit at other localities. Sharon Springs Member

The term Sharon Springs Member is restricted to the highly organic-rich, buttress- weathering shale unit of the traditional Sharon Springs Member (Fig. 4.9), as described by Elias (1931) and Gill et al. (1972) in western Kansas. The member consists of inter-bedded layers of dark brown and brownish black shale when fresh, but it weathers to bluish gray or grayish black (Fig. 4.10). The shale is hard but hackly and weathers to a nearly vertical, strongly jointed, “buttress-forming” slope (Gill et al., 1972). Thin bentonites are present in this unit, but they are difficult to trace from one area to the next. Limestone concretions are common, particularly near the top of the unit in Kansas, and several of these concretions are 123

Figure 4.9: Representative measured sections of the Sharon Springs Member. *Pembina Mtn section from Gill and Cobban, 1965, ** Chamberlain section is a composite section, *** McAllaster Buttes is the type locality of the Sharon Springs Member. 124

Figure 4.10: The Sharon Springs Member at the type locality, McAllaster Buttes, Kansas. At the top of the outcrop is a layer of large septarian concretions. septarian with abundant invertebrate fossils. Fish and large marine vertebrates are common in the member, but the nature of the shale prevents preservation of invertebrates, except in concretions. The Sharon Springs Member was first described in Logan County, Kansas at McAllaster Buttes, and expanded to include several outcrops in western Kansas (Elias, 1931) and finally to South Dakota, where the member was incorrectly identified as the “organic-rich unit” (Gill and Cobban, 1966). The Sharon Springs Member in Kansas was originally divided into three units, a lower and upper shale unit and an upper phosphatic unit (Elias, 1931). Gill et al. (1972) named the lower unit the dark soft shale unit, the upper shale unit the buttress-weathering shale unit and separated the phosphatic unit out from the Sharon Springs, but did not assign it a formal name. Strata originally referred to the lower part of the Sharon Springs in western Kansas have been herein referred to the Gammon Ferruginous Member, because the facies closely resembles the facies of the Gammon Ferruginous Member in the Black Hills. The phosphatic unit is herein referred to the Chamberlain Member, described below. This restricts the Sharon Springs Member to the buttress-weathering shale unit in Kansas, where it is 18m thick. This member is completely exposed at McAllaster Buttes, Logan County Kansas. In Kansas, the member overlies the Gammon Ferruginous Member unconformably, and a clear iron-stained unconformity is present at the type locality. In eastern South Dakota, the term “Sharon Springs Member” was applied by Searight 125 (1937) who divided the interval into two units, the fish scale zone and an upper buttress weathering shale (Moxon et al., 1939). The fish scale zone has herein been referred to the Walhalla Member and the Sharon Springs Member is restricted to the buttress-weathering shale. A similar pattern is also seen in the Pembina Member of eastern North Dakota. In the Pembina Member, three units were previously recognized (Gill and Cobban, 1965), a lower bentonitic unit, a middle organic-rich shale and an upper phosphatic unit. The lowest part of the Pembina is herein referred to the Walhalla Member, the middle unit is referred to the Sharon Springs Member and the phosphate unit is referred to the Chamberlain Member. In eastern South Dakota, the Sharon Springs Member is locally missing, removed by erosion indicated by an unconformity at the top of the member, but may be as thick as 5m. In North Dakota, the member is as much as 15m thick. The Sharon Springs Member had been correlated to the Black Hills (Gill et al., 1972), where it was restricted to the most organic-rich interval within the Baculites obtusus ammonite range zone, and a time-transgressive facies shift was proposed, so that deposition occurred earlier in the west than in the east (Parrish and Gautier, 1993). Careful examination of these units, however, indicates that the two facies are distinct lithologically and the Black Hills sections are referred to the Walhalla Member. The Walhalla Member is black and hard, whereas the Sharon Springs Member is interbedded dark brown and brownish-gray hackly shale. In the Black Hills, the temporal equivalent to the Sharon Springs Member is the Mitten Black Shale Member, which also has a lithology distinct from the Sharon Springs Member. In Colorado, lithology equivalent to the Sharon Springs Member could not be verified. The Mitten Black Shale Member is reported from the northeastern part of Colorado (Gill et al., 1975), but this could not be verified. Baculites asperformis is reported from a layer of large septarian concretions in Kansas (Gill et al., 1972). Although no Baculites mclearni have been found in the member, it is presumed that deposition occurred through the Baculites mclearni to the B. asperformis 126 ammonite range zones (Gill et al., 1972). Most specimens of Baculites are too poorly preserved and encrusted with selenite to be identifiable. Vertebrates from the member include abundant fish remains, some mosasaurs and plesiousaurs. Vertebrates are typically encrusted with selenite, which often penetrates the bone making identification difficult. Chamberlain Member

The Chamberlain Member is herein defined for a phosphate-rich facies and overlying gray shale facies above the Sharon Springs Member in the eastern part of the basin (Fig. 4.11) and includes strata previously assigned to the Sharon Springs Member in Kansas (Fig. 4.3a, Elias, 1931) and South Dakota and to the Pembina Member in North Dakota (Gill and Cobban, 1965). Gill et al. at removed the phosphate-rich facies from the Sharon Springs Member in Kansas, but did not assign it a formal name. The facies has not be previously described in South Dakota. The member can be divided into two units, the lower phosphate-rich facies, which includes several layers of selenite-encrusted phophate nodules, and an overlying gray shale facies lacking nodules and concretions. The member is well exposed at several outcrops along the Missouri River near Chamberlain, South Dakota, where it is as much as 10m thick (Fig. 4.7). The phosphate facies and the gray shale facies are each up to 5m thick at the type locality. At some outcrops along the Missouri River the member is represented by only a 50cm thick phosphate nodule lag overlying the Sharon Springs Member. The member can be correlated to Kansas and North Dakota. In Kansas, it is approximately 250cm thick and in North Dakota the member is up to 7m thick. The gray shale facies does not appear to be present in North Dakota. This member is assigned to the Baculites sp. (smooth) through Baculites perplexus ammonite range zone in Kansas (Gill et al., 1972). Baculites perplexus is found in ironstone concretions just below the lowest phophate nodules in North Dakota (Gill and Cobban, 1965). Phosphate facies: The first unit of the Chamberlain Member is the phosphate nodule 127

Figure 4.11: Representative measured sections of the Chamberlain Member. *Pembina Mtn section from Gill and Cobban, 1965 128 facies. While this unit remains very organic-rich, similar to the Walhalla Member and the Sharon Springs Member, this facies is characterized by very abundant selenite-encrusted phosphate nodules. These nodules usually occur in discrete layers, but may also be disseminated throughout the shale. The facies is present from western Kansas to eastern North Dakota. In Kansas the interval is 2m thick and it thickens to the north into North Dakota where it is 7m thick. The facies is discontinuous between outcrops along the Missouri River, but has an average thickness of approximately 5m where it is present. The facies in not present in the Black Hills, it is equivalent in part to the Redbird Silty Member. In Kansas, one bentonite 18cm thick and five bentonites less than 1cm thick are present within the phosphate nodule facies. In North Dakota, 4-5 layers of bentonite ranging from less than 1cm to 4cm thick are present. No bentonites were recognized in this interval in South Dakota. In Kansas, directly above the phosphate layer, an 8cm thick layer of shale is present that is very organic-rich and harder than the shale above or below. Gray shale facies: The gray shale facies is best developed in eastern South Dakota, where it is 5m thick and exposed in cliff faces above the phosphate nodule facies. The facies also appears to be present in Kansas, where 46cm of gray-black, fissile shale overlies the last layer of phosphate nodules and a hard heavily organic-rich layer. The facies does not appear to be present in North Dakota.

The gray shale facies is characterized by gray, soft fissile shale that weathers to a gentle slope at most outcrops. In eastern South Dakota, the gray shale facies has two layers of bentonite, each approximately 25cm thick, that are often cemented with calcite forming resistant ledges. Two thin bentonites are present within the interval in Kansas. Unassigned Unit – Mitten Black Shale Member?

A unit of organic-rich shale is present at the top of the Walhalla Member in South Dakota (Fig. 4.12). This unit is restricted to the Black Hills region and overlies the second unit of the Walhalla Member unconformably. This unit ranges from less than 2m at Buffalo 129

Figure 4.12: The unassigned unit (possibly Mitten Black Shale Member at Wasserburger Ranch, South Dakota. The two white lines indicate the position of the first stringer (bottom) and second stringer (top) bentonite layers.

Figure 4.13: Representative measured sections of the unassigned unit (possibly part o the Mitten Black Shale). The first and second bentonite stringers are prsent at Wasserburger Ranch and Wallace Ranch, but the second bentonite stringer is missing at Buffalo Gap. 130 Gap to approximately 5m at Wallace Ranch and is present around the southeastern Black Hills (Fig. 4.13). This unit is similar to the second unit of the Walhalla Member except that the shale is slightly grayer than that of the Walhalla Member. It is separated from the Walhalla Member by a disconformity and has three thin bentonites (Fig. 4.12). In the southeastern Black Hills, the lowest two bentonites of this unit make a thin couplet, and these two bentonites are referred to as the first stringer (Chapter 3). Although they are thin, they make a distinct continuous band across the outcrop. The third bentonite is also thin and is referred to as the second stringer (Chapter 3). The second stringer is locally removed by erosion, such as at Buffalo Gap. An unconformity marks the top of this unit, indicated by an iron-stained unit in most localities. At Buffalo Gap, three thin brecciated shell layers are present near the top of the unit, indicating multiple disconformities. It is unclear what part of the section this unit is equivalent to in other parts of the basin. Overlying this unit is the Mitten Black Shale Member. The Mitten Black Shale Member is divided into four units at its most complete locality at Redbird, Wyoming (Gill and Cobban, 1966). At Redbird, this unassigned unit is conformable with the first unit of the Mitten Black Shale Member and unconformable with the Walhalla Member. An iron- stained layer of brecciated shale overlies the unconformity at the based of this unassigned unit. The first occurrence of siderite concretions marks the top of this unit. In the eastern Black Hills, this unit is unconformable at the top and bottom, where it overlies the Walhalla Member and underlies the fourth unit of the Mitten Black Shale Member. The first, second and third units of the Mitten Black Shale Member are missing in the eastern Black Hills. Further work is needed on the biostratigraphy and bentonite stratigraphy of this unit in order to determine its correlation. It is most reasonable to assume that this unit, which is conformable with the base of the Mitten Black Shale Member at the most complete sections, could be included as part of the Mitten Black Shale Member. 131 Other members of the Pierre Shale

Other members of the lower Pierre Shale that were not evaluated in this study include the Mitten Black Shale, the Redbird Silty and the Gregory members. The Mitten Black Shale Member is temporally equivalent to the Sharon Springs Member and the Chamberlain Member and overlies the Walhalla Member in the Black Hills of South Dakota and Wyoming. The Mitten Black Shale Member is restricted to the Black Hills region, where increased detrital sedimentation occurred during deposition of the organic-rich Sharon Springs Member elsewhere in the basin. The Mitten Black Shale Member can be divided into four units at the most complete locality. If the unassigned unit described above is included in the Mitten Black Shale, then the member would include 5 units at Redbird, Wyoming. Siderite concretions are common in the first unit, third unit and fourth unit. The second unit is devoid of concretions. In the northern Black Hills, only units 1 and 4 are present and clear unconformities marked by lags of phosphatized pebbles are present at the base of each unit. In the eastern Black Hills, only the fourth unit is present, and unconformably overlies the unassigned unit. The Redbird Silty Member is also restricted to the Black Hills (Gill and Cobban, 1962), and is temporally equivalent to the calcareous Gregory Member to the east. The Gregory Member is calcareous marl overlying the Chamberlain Member in eastern South Dakota and North Dakota (Gries and Rothrock, 1941). DISCUSSION The redefinition of members in the lower Pierre Shale allows for a much clearer understanding of the dynamics of the basin during this time. During deposition of the Gammon Ferruginous Member, sedimentation was nearly was restricted to the axial basin, with only minor deposition along the western margin of the eastern stable platform. During deposition of the Walhalla Member, sedimentation was restricted to the northern part of the basin and in particular to the axial basin and the Williston Basin. During this time, the axial 132 basin and Williston Basin rapidly subsided, controlled by tectonic forces from the Absoroka Thrust fault in Wyoming (Witschko and Dorr Jr., 1983). Rapid subsidence resulted in sediment stravation, preserving high organic content in the Walhalla Member. Following deposition of the Walhalla Member, sedimentation patterns shifted so that there were north- south trending facies belts and facies differed from the axial basin to the eastern platform. Sedimentation was not differentiated in the Williston Basin compared to the rest of the eastern platform during the deposition of the Sharon Springs Member through the Gregory Member. These facies belts appear to extend from North Dakota to Kansas with little variation. In addition to these broad, basinal patterns, a shift in sedimentation of distal deltaic sediments in the Black Hills, corresponding to sedimentation associated with the Eagle delta, indicates a shift of the delta from terminating near the northern Black Hills during the Gammon Ferruginous Member to terminating near the southern Black Hills during the Mitten Black Shale through the Redbird Silty Member. The Niobrara Cycle

The Gammon Ferruginous Member represents the regressive phase of the Niobrara Cycle (Fig. 4.14, Kauffman and Caldwell, 1993). The Gammon Ferruginous Member was deposited around the Black Hills in South Dakota and Wyoming and in western Kansas (Fig. 4.15). Insufficient information is present for Colorado. However, previous studies suggest that sedimentation in Colorado may be similar to that seen in the Black Hills (Gill et al., 1975). If this is the case, it would suggest a west to east dichotomy of deposition of the Gammon Ferruginous Member. The Black Hills lie along the eastern margin of the axial basin and is at the distal edge of a large delta that has formed in Wyoming and Montana, as represented by the Eagle Sandstone (Rice and Shurr, 1983). The Gammon Ferruginous Member in the Black Hills represents the distal edge of this delta. Similar deltas also may have been present in Colorado, as represented by the Mesaverde Formation, which would cause similar sedimentation in the Pierre Shale in Colorado (Roehler, 1990). The Gammon Ferruginous Member represents the late highstand systems tract of 133

Figure 4.14: Chronostratigraphic correlation of third and fourth-order sequences in the Lower Pierre Shale. The Niobrara Cycle terminates in the Gammon Ferruginous Member in the Black Hills of South Dakota and Wyoming and western Kansas. The Walhalla sequences are tectonically controlled sequences restricted to the northern part of the basin. The Claggett Cycle is a second eustatic with at least three sequences. The first sequence is represented by the unassinged unit and Unit 1 of the Mitten Black Shale Member in the west and the Sharon Springs Member in the east. Sequence II is represented by the Mitten Black Shale Member units 2-4 in the west and the Chamberlain Member in the east. Sequence III was not evaluated in this study but is represented by the Redbird Silty Member in the west and the Gregory and Crow Creek members in the east. 134 the third-order eustatic sequence of the Niobrara seaway. The Gammon Ferruginous Member is thickest in the northern Black Hills, where it is 300m thick (Rubey, 1930), and thins to the south at Redbird, where it is only 5.5m thick. Parasequences are also best developed in the northern part of the basin, their tops marked by siderite concretionary horizons. The Eagle delta terminated near the northern Black Hills during deposition of the Gammon Ferruginous Member as indicated by this sedimentation pattern and the presence of the Groat Sandstone Member at the top of the Gammon Ferruginous Member in the northern Black Hills. The Walhalla Sequences

The Walhalla Member represents two sequences that appear to be controlled by a single tectonic pulse associated with the Absoroka Thrust in Wyoming and flexure of the Western Interior Basin resulting in subsidence of the axial basin and the Williston Basin. Subsidence of the Williston Basin in conjunction with the axial basin was a result of structural weaknesses in the basement rocks of the Williston Basin, which allowed the basin to subside in response to the thrust forces. This subsidence was primarily restricted to these basins, and the eastern stable platform remained shallow and may have even bulged upward slightly resulting in shallowing of this region (Ettensohn, 1998). These sequences are preserved in the northern part of the basin (Fig. 4.15), supporting a tectonic control that resulted in a more regional distribution of the sequences. The thick

Figure 4.15: During times of tectonic quiescence, sedimentation patterns reflected a retroarc foreland basin with north-south trending parallel facies belts (A), however during times of tectonic activity in Wyoming and Utah, the axial basin and the Williston Basin in the northern part of the basin subsided, resulting in a north to south dichotomy in sedimentation patterns (B). 135 bentonite of the Ardmore bentonite succession near the base of the Walhalla Member has been dated at 80.5Ma +/- 0.55Ma (Hicks et al., 1999; Obradovich, 1993; Rogers et al., 1993). While this is within the range of potential error to coincide with the eustatic sea level change, it appears that the eustatic sea level rise, which is represented across the entire basin, actually occurred about 0.5Ma later than the sea level rise associated with the Walhalla Member. The restriction of this member to the northern part of the basin supports a tectonic control on the sequence, as opposed to a eustatic control. In Kansas, a lag bed at the unconformity that caps the Gammon Ferruginous Member contains phenocrysts from the bentonites of the Ardmore bentonite succession, indicating that deposition did occur in Kansas, and the sediments were subsequently eroded (Chapter 2). This indicates that the area was shallow, compared to the eastern stable platform, throughout this time frame. Two sequences are clearly preserved in the Walhalla Member (Fig. 4.14). In the first sequence, highly bituminous shale is interbedded with bentonites of the Ardmore bentonite succession. A locally well-developed transgressive lag marks the base of this sequence (Fig. 4.15). Deposition of this sequence began earlier in the axial basin and the Williston Basin and during maximum flooding, progressed onto the eastern stable platform. In the axial basin, very low sedimentation rates resulted in deposition of multiple bentonites with no shale between them, resulting in a composite bentonite that correlates to multiple bentonites in the Williston Basin and the eastern platform (Chapter 2). An unconformity within the Ardmore bentonite succession marks the top of the first sequence. This unconformity has considerable relief on it, in some places removing all of the sediments between this unconformity and the base of the Walhalla Member. This unconformity is best developed on the eastern stable platform. In the axial basin, no evidence of this unconformity is present. Above this unconformity, and its correlative conformity in the axial basin, a second phase of transgression occurred. In the axial basin, this transgression is again indicated by 136 sediment starvation and deposition of a composite bentonite that correlates to multiple bentonites to the east. The top of the Ardmore bentonite succession was deposited during the transgressive phase of this sequence. Unit 2 of the Ardmore bentonite succession represents the highstand systems tract of this sequence. Increased sedimentation during this period resulted in lower organic content in the shale, more hackly shale, and thinner bentonites. This unit was only deposited in the axial basin and in the Williston Basin, indicating that the eastern platform became shallow and by the end of the second sequence, erosion occurred on the eastern platform. Following this second sequence, erosion extended across the basin and an unconformity with considerable relief, locally removed all of unit 2, even in the axial basin. The two sequences appear to represent distal sedimentation during a single tectonic phase, as described by Ettensohn (1998). The first sequence records sedimentation resulting from rapid subsidence of the basin responding to deformational loading in the thrust belt that occurs in the subsurface. Because the deformation occurred in the subsurface, no clastic source was available, resulting in the accumulation of organic matter and sediment starvation in the axial basin. Preservation of organic matter was facilitated by dysoxic bottom conditions, created by a stratified water column typically associated with the initial transgression, resulting from circulatory restriction in a partly enclosed basin (Ettensohn, 1998). Low oxygen conditions are indicated by abundant pyrite and selenite in the shale, as well as the abundance of planktic and nektonic fauna (particularly vertebrate fauna) and the absence of benthic fauna. The axial basin was deepest during this initial phase, when tectonic forces were strongest. The regression of the first sequence resulted from relaxation of this structural load. This regression was relatively minor, and in the axial basin sedimentation was continuous while in the shallower eastern platform erosion occurs. The second sequence resulted from the formation of an anti-peripheral bulge, which results in a second transgression. This transgression was not as extensive as the initial subsidence and transgression, but it does cover the eastern platform and allow sedimentation across the 137 basin. Finally, relaxation resulting from unloading resulted in regression. Ettensohn (1998) indicates that during this phase marginal marine sediments may extend across the basin and redbeds may be deposited, as displayed in the Appalachians. In this case, the second sequence of the Walhalla Member was entirely marine but does indicate increased clastic sedimentation and reduction of organic matter. This was followed by a second sequence boundary as the tectonic pulse ends and the seaway relaxes. The Claggett Cycle

The third order sequence: The Claggett Cycle is considered a eustatic cycle within the Western Interior Seaway (Kauffman and Caldwell, 1993). Traditionally, the base of the Claggett Cycle has been place at the base of the Ardmore bentonite succession. As discussed above, the Ardmore bentonite succession within the Walhalla Member was deposited as a result of tectonic compression and preceded the eustatic sea level rise. Therefore, the eustatic sea level rise began within the Sharon Springs Member in the east and the Mitten Black Shale Member to the west (Fig. 4.14). The end of the third-order sequence is marked by the deposition of the deltaic Judith River Formation in Montana, the Redbird Silty Member in the Black Hills region and the phosphatic Chamberlain Member, the calcareous Gregory Member and the silty Crow Creek Member in the east. The eustatic nature of this sequence is supported by deposition of the sediments across the entire basin, from Kansas to North Dakota, indicating that regional tectonic processes did not play a role. Higher order sequences: Within the Claggett Cycle, at least four higher-order sequences are present (Fig. 4.14). The unassigned unit and Unit 1 of the Mitten Black Shale Member in the Black Hills and the Sharon Springs Member to the east represent the first sequence. Units 2-4 of the Mitten Black Shale in the Black Hills and the Chamberlain Member represent the second sequence. The Redbird Silty Member in the Black Hills and the Gregory Member to the east represent the third sequence. North to south trending facies belts are recognized in these sequences and deposition in North Dakota was similar to the rest of the eastern platform as opposed to the axial basin, 138 contrasting with deposition during the Walhalla sequence (Fig. 4.18). In both sequences 1 and 2, deposition began earlier in the Black Hills than on the eastern platform. On the eastern platform, only the early highstand and beginning of the late highstand systems tract are preserved in both sequences. This is a result of lack of deposition during the transgression and erosion of the top of the sequence associated with the next sequence boundary. Erosion/ transgression surfaces mark the sequence boundaries on the eastern platform. In the Black Hills, full sequences are more clearly developed. The transgressive systems tracts of both sequences 1 and 2 are preserved as gray-black shale devoid of concretions. The unassigned unit described above and unit 2 of the Mitten Black Shale Member represent the two transgressive systems tracts. The bases of these units are marked by sequence boundaries with considerable relief on them. At some localities, particularly in the eastern Black Hills, as much as 6m of relief occurs on the first unconformity, sometimes completely removing the entire highstand systems tract of the Walhalla sequence. Erosion associated with the second unconformity also completely removed the highstand systems tract of the first sequence in the eastern Black Hills. The highstand systems tracts are preserved as gray-black shale with layers of siderite concretions marking parasequence boundaries. Units 1, 3 and 4 of the Mitten Black Shale Members represent the highstand systems tracts. The sequences are most completely developed in the western Black Hills, especially around Redbird, where the axial basin was deepest. Erosion on the first sequence boundary was most prevalent in the most south-west corner of South Dakota, and is evident in outcrops that are trending in a north-south line just south of Edgemont, South Dakota. Erosion on the second sequence boundary is more prevalent further to the east, near Buffalo Gap, South Dakota, where part of the transgressive systems tract of unit 1 was removed along with the highstand systems tract. This increased erosion may have resulted from a local high associated with a peripheral bulge. During the Niobrara Cycle, the Eagle delta terminated in the northern Black Hills (Rice and Shurr, 1983). During the Claggett Cycle, however, sedimentation patterns shifted 139 and the Mitten Black Shale and Redbird Silty members are thicker and more complete near the southern Black Hills. CONCLUSION Recognizing members based on distinct, subtle lithologic changes can allow for a clearer interpretation of sequences and facies changes. Two new members are recognized in the lower Pierre Shale, the Walhalla Member and the Chamberlain Member. The Mitten Black Shale, Pembina and Sharon Springs members are redefined. Distribution of these members indicates that north-south trending facies belts were present throughout deposition of most of the lower Pierre Shale. However, during deposition of the Walhalla Member, facies patterns indicate that a north-south partitioning of the basin was present. Recognition of the Walhalla Member in the northern Black Hills indicates the presence of a sequence created as a result of tectonic processes that were regionally restricted to the northern part of the basin and were a response to thrusting on the Absoroka Thrust in Wyoming. In the axial basin and the Williston Basin subsidence resulted, while the eastern platform in eastern South Dakota remained shallow. Prior to the Walhalla sequence, the Gammon Ferruginous Member was deposited, representing the end of the Niobrara eustatic cycle. Following the deposition of the Walhalla Member, the Claggett Cycle represents a second eustatic sea level rise. Patterns of sequences in North Dakota no longer indicate a deeper basin and the eastern platform continues from Kansas to North Dakota. North-south trending facies belts coincide with structural facies belts of the foreland basin. The axial basin, east median hingeline and eastern platform are represented in the sediments of the lower Pierre Shale. In the axial basin, the Mitten Black Shale and Redbird Silty Members were deposited. These are the distal equivalents of the Parkman Sandstone and Judith River Formation in Wyoming and Montana. To the east, the Sharon Springs, Chamberlain and Gregory Members were deposited. Within this third- order eustatic sea level change, three fourth order cycles are also represented. The lower Mitten Black Shale Member and the equivalent Sharon Springs member represent the first 140 sequence. The upper Mitten Black Shale and the Chamberlain members represent the second sequence. The Redbird Silty and the Gregory Members represent the third sequence. During the eustatic sea level cycles, the transgressive systems tract is generally restricted to a thin transgressive lag while the early highstand is indicated by gray-black shale devoid of concretions. The late highstand systems tracts typically have numerous siderite concretions marking parasequence boundaries and periods of sediment starvation combined with higher oxygen conditions in the seawater. This is especially true in the Black Hills, where distal delta sedimentation was taking place. Sequence boundaries have considerable relief on them, locally removing as much as 10m of section. In all the sequences of the lower Pierre Shale, the transgressive systems tract is best developed in the axial basin, with only a transgressive lag preserved on the eastern platform. The early highstand systems tract is typically present across the entire basin and the late highstand systems tract is usually missing in the eastern platform, where erosion associated with the overlying sequence boundary has removed sediments. Erosion associated with sequence boundaries was extensive across the basin, often extending into the axial basin as well. Recognition of sequences in distal sediments requires the recognition of subtle facies changes and sequence boundaries separating two shale units, which may not be easily recognizable. Traditionally the focus on sequence stratigraphy has been in more proximal environments where typical sequence patterns can be recognized. The recognition of these sequences in the distal parts of the basin is critical to understand the complete basin dynamics. REFERENCES DeGraw, H.M., 1975. The Pierre-Niobrara unconformity in western Nebraska. In: W.G.E. Caldwell (Editor), The Cretaceous System in the Western Interior of North America. The Geological Association of Canada Special Paper, pp. 589-607. Dyman, T.S. et al., 1994. Cretaceous rocks from southwestern Montana to southwestern Minnesota, Northern Rocky Mountain, and Great Plain region. In: G.W. Shurr, G.A. 141 Ludvigson and R.A. Hammond (Editors), Perspectives on the eastern margin of the Cretaceous Western Interior Basin. Geological Society of America Special Paper, Boulder, pp. 5-26. Elias, M.K., 1931. The Geology of Wallace County Kansas. Kansas Geological Survey Bulletin 18: 254. Ettensohn, F.R., 1998. Compressional tectonic controls on epicontinetal black-shale deposition: Devonian-Mississippian examples from North America. In: J. Schieber, W. Zimmerle and P.S. Sethi (Editors), Shale and Mudstones I: Basin studies, sedimentology and paleontology. E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart, pp. 109-128. Gill, J.R. and Cobban, W.A., 1962. Red Bird Silty Member of the Pierre Shale, a new stratigraphic unit. U. S. Geological Survey Professional Paper, 450-B: 21-24. Gill, J.R. and Cobban, W.A., 1965. Stratigraphy of the Pierre Shale, Valley City and Pembina Mountain areas, North Dakota. U. S. Geological Survey Professional Paper 392-A: A1-20. Gill, J.R. and Cobban, W.A., 1966. The Redbird section of the Upper Cretaceous Pierre Shale in Wyoming. U. S. Geological Survey Professional Paper 393-A: A-1-A73. Gill, J.R., Cobban, W.A. and Schultz, L.G., 1972. Stratigraphy and composition of the Sharon Springs Member of the Pierre Shale in Western Kansas. U. S. Geological Survey Professional Paper, 728: 50. Gill, J.R., Cobban, W.A., Scott, G.R. and Burkholder, R.E., 1975. Unedited stratigraphic sections of the Pierre Shale near Roundbutte and Buckeye in Larimer County, Northern Colorado. U. S. Geological Survey Open-File Report, 75-129, Denver, 12 pp. Gries, J.P. and Rothrock, E.P., 1941. Manganese deposits of the Lower Missouri Valley in South Dakota. U. s. Geological Survey Report of Investigations, 38, Vermillion, 96 pp. 142 Hayden, F.V., 1862. On the geology and natural history of the Upper Missouri. American Philosophical Society Transactions, 12(1): 218. Hicks, J.F., Obradovich, J.D. and Tauxe, L., 1999. Magnetostratigraphy, isotope age calibration and intercontinental correlation of the Red Bird section of the Pierre Shale, Niobrara County, Wyoming, USA. Cretaceous Research, 20: 1-27. Izett, G.A., Cobban, W.A. and Gill, J.R., 1971. The Pierre Shale near Kremmling, Colorado and its correlation to the east and the west. U. S. Geological Survey Professional Paper 684-A: A1-A19. Kauffman, E.G., 1984. Paleobiogeography and evolutionary response dynamic in the Cretaceous Western Interior seaway of North America. In: G.E.G. Westerman (Editor), Jurassic-Cretaceous Biochronology and paleogeography of North America. Geological Association of Canada Special Paper, pp. 273-306. Kauffman, E.G. and Caldwell, W.G.E., 1993. The Western Interior basin in space and time. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 1-30. Landis, E.R., 1959. Radioactivity and uranium content, Sharon Springs Member of the Pierre Shale Kansas and Colorado. U. S. Geological Survey Bulletin, 1046-L: 299- 318. Leonard, A.G., 1904. Topographic features and geological formations of North Dakota. North Dakota Geological Survey, 3d Bienn. Rept: 127-177. Moxon, A.L., Olson, O.E. and Searight, W.V., 1939. Selenium in rocks, soils, and plants. South Dakota Agr. Expt. Sta. Tech. Bull 2: 94. Obradovich, J.D., 1993. A Cretaceous Time Scale. In: W.G.E. Caldwell and E.G. Kaufman (Editors), Evolution of the Western Interior Basin. Geological Association of Canada, pp. 379-396. Parrish, J.T. and Gautier, D.L., 1993. Sharon Springs Member of the Pierre Shale: Upwelling in the Western Interior seaway? In: W.G.E. Caldwell and e.G. Kauffman (Editors), 143 Evolution of the Western Interior basin. Geological Association of Canada, Tulsa, pp. 319-332. Rice, D.D. and Shurr, G.W., 1983. Patterns of sedimentation and paleogeography across the western interior seaway during time of deposition of Upper Cretaceous Eagle Sandstone and equivalent rocks, northern Great Plains. In: M.W. Reynolds and E.D. Dolly (Editors), Mesozoic paleogeography of west-central United States. Rocky Mountain Section SEPM Paleogeography symposium, pp. 337-358. Roehler, H.W., 1990. Stratigraphy of the Mesaverde Group in central and eastern greater Green River Basin, Wyoming, Colorado and Utah. U. S. Geological Survey Professional Paper, 1508: 52. Rogers, R.R., Swisher, I., C. C. and Horner, J.R., 1993. 40Ar/39Ar age correlation of the nonmarine Two Medicine Formation (Upper Cretaceous), northwestern Montana,

U. S. A. Canadian Journal of Earth Science, 30: 1066-1075. Rubey, W.W., 1930. Lithologic studies of fine-grained Upper Cretaceous sedimentary rocks of the Black Hills region. U. S. Geological Survey Professional Paper, 165-A: 1-54. Russell, D.A., 1993. Vertebrates in the Cretaceous Western Interior seaway. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the western interior basin. Geological Association of Canada, pp. 665-680. Searight, W.V., 1937. Lithologic stratigraphy of the Pierre Formation of the Missouri Valley in South Dakota. South Dakota State Geological Survey, Report of Investigations 27: 63. Spivey, R.C., 1940. Bentonite in southwestern South Dakota. South Dakota Geological Survey Report of Investigations: 56. Wing, M.E., 1940. Bentonites of the Belle Fourche district. South Dakota Geological Survey Report of Investigations, 35: 29. 144 Witschko, D.V. and Dorr Jr., J.A., 1983. Timing and deformation in overthrust belt and foreland of Idaho, Wyoming and Utah. Amerian Association of Petroleum Geologists Bulletin, 67(8): 1304-1322. 145 Record of tectonic activity of the Sevier Orogenic belt in the distal marine sediments of the Cretaceous Western Interior Seaway

ABSTRACT Interpretation of thrust activity in a tectonically active foreland basin is generally complicated by later fault activity that dissects the thrust belt and the proximal sediments that are frequently used in the interpretation. In the Cretaceous Western Interior seaway, movement along the Absoroka Thrust is relatively well constrained based on proximal synorogenic sediments. In addition, the Walhalla Member records movement of the fault in the distal marine environment. Recognition of tectonically controlled distal marine sedimentation provides a unique opportunity for a more thorough evaluation of basin dynamics in response to thrust movement. Unconformities associated with the Walhalla Member record a migrating peripheral bulge in the Black Hills region corresponding to a single tectonic pulse on the Absoroka Thrust. Initial thrusting resulted in cratonward migration of the foreland basin and associated peripheral bulge. Continued cratonward migration resulted in overlap of the foreland peripheral bulge with the margin of the Williston Basin, resulting in yoking of the basin and deposition of the first sequence of the Walhalla Member. Following cessation of thrust movement, deposition of clastic sediments into the foreland basin resulted in thrustward migration of the peripheral bulge in the Black Hills. Finally, as clastic sedimentation filled the foreland basin and prograded eastward, the peripheral bulge migrated cratonward, and is reflected by thinning of the Mitten Black Shale Member to the east in the Black Hills region. 146 Record of tectonic activity of the Sevier Orogenic belt in the distal marine sediments of the Cretaceous Western Interior Seaway

INTRODUCTION Timing of thrust activity in tectonically active thrust belts, such as the Cretaceous Sevier Orogenic belt in the western United States, is generally estimated by the recognition of synorogenic clastic sedimentation responding to the thrust activity (Fig. 5.1; see for example Cross and Pilger Jr., 1978; Jordan, 1981; Wiltschko and Dorr Jr., 1983; DeCelles, 1994). Dependence on proximal synorogenic sedimentation can be complicated by two factors: 1) Initial thrusting events may not break the surface and uplift may remain below sea level. As a result no coarse clastic deposition will occur (Ettensohn, 1998). 2) Proximal sediments are often deformed by later thrust activity (Jordan, 1981). Careful examination of distal sediments can reveal a record of tectonic activity that is not complicated by the issues that affect proximal settings, providing an independent method of evaluating tectonic activity.

Figure 5.1: Thrusting events of Utah and Wyoming compared to stratigraphic units of the western margin of the western interior seaway. Orogenic thrusting events occur synchronously with transgressions in the basin (Villien and Kligfield, 1993). The major cycles recognized by Kauffman and Caldwell (1993) and generally accepted for the northern part of the seaway are shown, relating thrusting events to the transgression of these cycles. Some of the major stratigraphic units, particularly within the Campanian, are given for the western margin and the distal sediments. Tectonic material compiled from Villient and Kligfield (1993), Wiltschko and Dorr (1983) and DeCelles (1994). 147 In the Cretaceous Western Interior, a broad foreland basin resulted as a response to the Sevier Orogenic Thrust Belt to the west (Fig. 5.2; Kauffman, 1977). The western interior seaway is significantly wider than typical foreland basins produced by thrust-related subsidence (Jordan, 1995), resulting from a composite of a western foreland basin and a very broad axial basin. While it remains clear why subsidence occurred over such a large area, it is suggested that an underlying shallowly dipping subducting plate may be the cause. The forebulge associated with thrusting activity in the western interior is poorly understood. It has been debated whether the multiple uplifts within the basin were forebulge uplifts responding to Sevier thrusting (Quinlan and Beaumont, 1984) or Laramide-style basement faulted uplifts (Schwarz and DeCelles, 1988). The Absoroka Thrust belt is one of the more precisely dated faults in the Idaho- Wyoming belt (Fig. 5.2; Jordan, 1981). The Absoroka Thrust was active from the Santonian to Maastrichtian, with multiple discrete tectonic pulses (Schwartz and DeCelles, 1988). The earliest event is recorded by a conglomerate containing clasts of the Lower Cretaceous Aspen Shale (Wiltschko and Dorr Jr., 1983). Shaly interbeds within the conglomerate contain mid-Santonian microfossils. Assuming synchronous deposition of the conglomerate with fault activity, the fault was active in the Santonian (Royse et al., 1975; Vietti, 1977). This early conglomeratic unit was subsequently folded in the hanging wall of the Absoroka Thrust. The fault offsets strata as young as latest Cretaceous. This places motion on the

Fig. 5.2: Thrusting activity in western Wyoming and Utah resulted in a broad foreland basin in the Western Interior. Cratonward of the foreland basin, a forebulge developed due to the stresses in the basin. A: Cross- section of the Cretaceous Western Interior Seaway (Modified from Kauffman, 1977). B: Map of the Cretaceous Western Interior Seaway in the northern United States showing the position of active faults in the Idaho- Wyoming thrust belt. 1 = Paris Fault, 2 = Meade-Crawford Fault, 3 = Absoroka Fault, 4 = Darby Fault, 5 = Hogsback Fault. (Modified from Jordan, 1981). 148 Absoroka Thrust between Santonian and Maastrichtian, with primary movement in the Maastrichtian (Armstrong and Oriel, 1965; Royse et al., 1975). Tectonic activity on the Absoroka Thrust is also recorded in the distal sediments of the Walhalla Member (Fig. 5.1; Chapter 4). Deposition of the Ardmore bentonite succession coincident with initial thrusting along the Absoroka Thrust places a date on this thrusting event at 80.5 Ma (Hicks et al., 1999). Recognition of a tectonic influence on the Walhalla Member provides a unique opportunity to evaluate distal basin response to tectonic thrusting. FORELAND BASIN MODELS Two basic models have been proposed for the deformation of continental crust in response to a thrust load. In the first model, the crust responds to the load through elastic deformation (Fig. 5.3; Flemings and Jordan, 1990; Jordan, 1995). According to the elastic deformation model, lithospheric flexure occurs simultaneously with thrusting and the crust maintains its deformation until the stress is changed (Jordan, 1995). In viscoelastic deformation, the lithosphere weakens in response to increased thrust loading (Fig. 5.3; Quinlan and Beaumont, 1984). The occurrence of synorogenic conglomerates proximal to the thrust belts, recording multiple transgressive-regressive cycles, appears to support the viscoelastic model (Schedl and Wiltschko, 1982). However, modelling of the Late Cretaceous A B

Fig. 5.3: Comparison of the elastic and viscoelastic model of deformation. A: In the elastic model, initial thrusting results in thrustward migration of the peripheral bulge (1). With continued thrusting and relaxation, the peripheral bulge migrates cratonward (2, 3). B: In the viscoelastic model, thrusting results in deepening of the axial basin and basinward migration of the peripheral bulge. Triangles indicate initial position of the peripheral bulge. 149 Western Interior is complicated by the initiation of Laramide-style basement faulting beginning in the Campanian (Jordan, 1981; Monger, 1993; Wiltschko and Dorr Jr., 1983). As a result of this basement faulting, it is often difficult to discern whether a topographic high is a peripheral bulge responding to flexure of the basin or a response to basement faulting (Jordan, 1981). Flemings and Jordan (1990) predicted changes in basin wavelength corresponding to basinal response on an elastic lithosphere (Fig. 5.3). In their model, initial thrusting results in narrowing of the basin and migration of the forebulge towards the thrust belt. Continued thrusting and relaxation results in migration of the forebulge cratonward. According to their model, flexure responding to tectonic loading takes place on a scale of less than five million years. This is consistent with an elastic model in which the lithosphere is responding almost instantaneously to thrusting. In contrast, Quinlan and Beaumont (1984) proposed a viscoelastic model for the Appalachian foreland basin and the development of peripheral bulges such as the Cincinnati and Findlay Arches. In a viscoelastic model, initial response to a thrust load is identical to that of an elastic model, with the formation of a broad shallow foreland basin and a low forebulge. With a viscoelastic model, however, continued presence of the thrust load results in weakening of the lithosphere, resulting in flow of the lithosphere. With a viscoelastic model, continued presence of a thrust load results in deepening of the foreland basin and migration of the forebulge towards the thrust belt (Fig. 5.3). Quinlan and Beaumont (1984) also proposed a modified viscoelastic model, the temperature-dependent viscoelastic model. Under this model, the timing of the deformation and ultimate equilibrium is dependent on the temperature of the lithosphere, which ultimately controls the viscosity. The temperature- dependent viscoelastic model suggests that the lower part of the lithosphere flows on time scales of less than 1 Ma, while the upper part of the lithosphere is unable to flow on time scales within the age of the Earth. Because of the rigidity of the upper lithosphere, the crust in the viscoelastic model behaves identically to that of an elastic model. Therefore, the 150 general consensus is that the shallow lithosphere behaves ultimately in an elastic fashion, responding directly to thrust loading and remaining deformed as long as the load is in place. INTRACRATONIC BASINS The Cretaceous Western Interior is similar to the Paleozoic Eastern Interior in that both regions experienced development of intracratonic basins in addition to the foreland basin (Monger, 1993; Quinlan and Beaumont, 1984). In the Cretaceous Western Interior, the Williston Basin in North Dakota was a prominent intracratonic basin during the Cretaceous (Fig. 5.2; Monger, 1993). Quinlan and Beaumont (1984) describe the interaction of the Paleozoic foreland basin with the Michigan and Illinois intracratonic basins. Interaction between the basins depends in large part on the position of the two basins and the amount of space between them. In the case of widely spaced basins, no interaction will occur between the two (Fig. 5.4). However, with basins that are more closely spaced, the two peripheral bulges will interact with each other, producing arching when the peripheral bulges overlap completely, or yoking of the basins when the peripheral bulges work to cancel each other (Fig. 5.4). In the case of yoking, the two peripheral bulges coincide so that the overall

Fig. 5.4: Interaction between the foreland basin and an intracratonic basin depends on the distance between the two basins. If the basins are widely spaced, the basins will be decoupled and no interaction will occur (A). With closer spacing, the peripheral bulge of the intercratonic basin and the foreland basin will overlap producing arch (B). With even closer spacing, the peripheral bulge of the foreland basin and the intracratonic basin will overlap creating a lowered peripheral bulge and resulting in yoking of the basins (C, D) (From Quinlan and Beaumont, 1984). 151 amplitude of the two basins is lower than individual peripheral bulges, and the intracratonic basin becomes connected with the foreland basin. The Williston Basin is a structurally controlled basin, resulting from weaknesses in the crust due to Paleozoic activity (Monger, 1993). Therefore, the peripheral bulge associated with the Williston Basin was likely also in a fixed position. The foreland basin, on the other hand, was a dynamic basin and the peripheral bulge was free to migrate in response to stresses associated with thrust belt loading (Quinlan and Beaumont, 1984). Therefore, the interaction of the foreland peripheral bulge with the Williston Basin peripheral bulge was a dynamic interaction resulting in a range from a decoupled basin to a yoked basin depending on the position of the foreland peripheral bulge. STRATIGRAPHY OF THE WALHALLA MEMBER The strata of the Western Interior Seaway recorded ten third-order sea level fluctuations that are considered to represent tectono-eustatic sea level fluctuations (Fig. 5.1; Kauffman and Caldwell, 1993). Alternatively, some of the sea level fluctuations could correlate with tectonic pulses in the Sevier Orogeny (DeCelles, 1994; Villien and Kligfield, 1986; Wiltschko and Dorr Jr., 1983). Lillegraven (1986) suggested that these relative sea- level fluctuations were restricted to the northern part of the basin where tectonic activity was most active. The Echo Canyon Conglomerate in Utah and Wyoming is considered a synorogenic conglomerate associated with thrusting on the Meade-Crawford and Absoroka Thrust faults (Wiltschko and Dorr Jr., 1983). Distal marine sediments in the foreland basin include the Claggett Shale and the Walhalla Member of the Pierre Shale (Fig. 5.1; Dyman et al., 1994; Chapter 4). Two unconformity-bounded sequences are recorded within the Walhalla Member (Fig. 5.5; Chapter 4), which are interpreted as being controlled by a single tectonic pulse associated with the Absoroka Thrust Fault in Wyoming, corresponding to thrusting and then relaxation of the thrust (Ettensohn, 1998). The Walhalla Member is preserved in the 152

Fig. 5.5: Stratigraphy of the lower Pierre Shale based on biostratigraphy and bentonite correlations. Sequence boundaries and thickness variations record a migrating peripheral bulge in the region around the Black Hills. 153 northern part of the basin in South Dakota and North Dakota, within the axial basin and the Williston Basin. Below the Walhalla Member, the Gammon Ferruginous Member thins to the west across the Black Hills region (Fig. 5.5; Chapter 4). Above the Walhalla Member, the Mitten Black Shale Member thins to the east across the Black Hills (Fig. 5.5; Gill and Cobban, 1966) due to erosion associated with unconformities in the middle of the member. CORRELATION WITHIN THE WALHALLA MEMBER High-resolution correlation of the Walhalla Member in the Black Hills is facilitated by the presence of laterally continuous bentonites (Fig. 5.5). These bentonites have geochemical and mineralogical signatures that have been used to identify multiple magmatic source terrains (Fig. 5.6, 5.7; Chapter 3). Because these bentonites have variable signatures, individual bentonite layers can be identified that can also be useful in regional correlations. The Ardmore bentonite succession is a composite interval of bentonites representing at least 15 volcanic episodes. In the Black Hills, the Ardmore bentonite succession consists of a 1-meter thick bentonite at the base of the Walhalla Member and as many as six additional bentonites ranging in thickness from 12 to 26cm (fig. 5.5). Bentonites of the Ardmore succession represent at least four distinct magma sources, two rhyolites and two andesites

Fig. 5.6: Variations in the whole rock geochemistry of bentonites in the Lower Pierre Shale has been used to identify bentonite layers that are useful in regional correlation. The lines indicate the ranges of values for each bentonite. 154

Fig. 5.7: Biotite composition of bentonites of the lower Pierre Shale have been used to discriminate individual layers useful in regional correlation (Chapter 3).

(Chapter 2). Whole rock signatures of most bentonites of the Ardmore succession indicate a backarc volcanic source (Fig. 5.6). The rhyolites indicate an island arc volcanic source (Fig. 5.6; Chapter 3). Above the Ardmore bentonite succession, the “bentonite couplet” and the “first stringer” are two pairs of bentonites that can also be distinguished based on geochemical and mineralogical signatures (Chapter 3). The “bentonite couplet” consists of two bentonites, each about 25cm thick, separated by less than 5cm of shale in the Black Hills. Although these bentonites make a distinct couplet in the field, their geochemical signatures are very distinct. The lower bentonite is similar to andesites of the Ardmore succession with abundant biotite. The biotite geochemistry of the lower bentonite indicates that the source magma for this bentonite was more fractionated than the bentonites of the Ardmore succession (Fig. 5.7; Chapter 3). The upper bentonite has very few phenocrysts and the whole rock signatures indicate an island arc magmatic source. The “first stringer” has a whole rock geochemical signature that indicates an island arc magmatic source (Fig. 5.6). It can be distinguished from other bentonites in the Walhalla Member by the presence of biotite (Chapter 3). UNCONFORMITIES IN THE LOWER PIERRE SHALE Correlation of the bentonites in the Walhalla Member can be used to identify unconformities in the distal marine black shale that otherwise would be difficult to identify. Some of these unconformities were previously identified by Martin (1996) based on field 155 correlations and are confirmed and clarified by the correlation of the bentonites. The unconformity at the base of the Walhalla Member is clearly defined at most localities, and locally a sandy channel lag is present. The Gammon Ferruginous Member thickens to the east below this unconformity. At Redbird, Wyoming, the member is only 5m thick. On the eastern side of the Black Hills, the member is 30m thick (Fig. 5.5). Above the basal unconformity, deposition occurs earlier in the axial basin and the Williston Basin than on the eastern platform in eastern South Dakota, as indicated by the presence of five bentonite layers in the Black Hills and Williston Basin that are not present in eastern South Dakota (Fig. 5.6). Within the Ardmore succession, an unconformity marks the top of the first sequence (Fig. 5.6; Chapter 2). This unconformity is difficult to identify in the field at localities in the Black Hills, but it is well developed at eastern localities. In North Dakota, erosion associated with this unconformity removes four bentonites that are present in the Black Hills and eastern South Dakota Fig. 5.6). The intra-member unconformity (Martin, 1996) marks the top of the second sequence (Chapter 4) and has considerable relief on it (Fig. 5.6). At Redbird, Wyoming 12 bentonites are present below the intra-member unconformity, including the “bentonite couplet”. In the eastern Black Hills, the second sequence is half as thick as at Redbird, Wyoming, and at least two bentonites are missing at the top of the sequence that are present at Redbird. At several localities in the south-central Black Hills, including Wallace Ranch, Igloo, Slurp Flats and Whitley’s Wash, the second sequence is considerably thinner (Fig. 5.6). Only three bentonites at the top of the Ardmore succession are present at these localities. The remainder of the sequence is missing due to erosion associated with the overlying unconformity. Above the Walhalla Member, an unassigned unit (possibly the base of the Mitten Black Shale Member) and the Mitten Black Shale Member also contain several unconformities. Thinning of the Mitten Black Shale to the east is a result of the lower and 156 middle parts of the Mitten Black Shale being removed due to erosion and thinning of the upper part of the Mitten Black Shale (Fig. 5.5; Gill and Cobban, 1966). PERIPHERAL BULGE DEVELOPMENT DURING DEPOSITION OF THE LOWER PIERRE SHALE Within the Walhalla Member and the Mitten Black Shale Member, the multiple unconformities indicate the presence and migration of a peripheral bulge responding to the stresses of the Absoroka Thrust belt. At the base of the Walhalla Member is a regional unconformity that is present throughout the basin. This unconformity represents more time in the western Black Hills than in the eastern Black Hills, indicating that erosion occurred for a longer period of time to the west. This is consistent with the rapid migration of a peripheral bulge responding to initial thrusting (Quinlan and Beaumont, 1984). As the thrust advances toward the basin, the basin and the associated foreland bulge migrate cratonward (Fig. 5.8). In the Paleozoic, this resulted in the regional Knox Unconformity (Quinlan and Beaumont, 1984). Following this regional erosion, deposition began earliest in the Black Hills region and in eastern North Dakota, within the Williston Basin (Fig. 5.5). This suggests that the peripheral bulge continued to migrate eastward, overlapping with the margin of the Williston

Fig. 5.8: Model of the forebulge migration in the Black Hills region. 1) With initial thrusting, the forebulge migrates cratonward, producing a regional unconformity (the top of the Gammon Ferruginous Member, Fig. 5.5). 2) Continued thrusting and cratonward migration of the forebulge results in overlap of the forebulge of the foreland basin and the peripheral bulge of the Williston Basin, causing yoking of the two basins. 3) Relaxation of thrusting results in deposition of coarse clastics in the basin, causing the basin to deepen and narrow, with thrustward migration of the forebulge. 4) Continued deposition of coarse clastics into the basin results in cratonward migration of the foreland basin and forebulge. 157 Basin and resulting in yoking of the two basins (Fig. 5.8). During this period, the resultant uplift had low amplitude and sedimentation occurred across the uplift. As thrusting on the Absoroka Thrust ceased, the foreland basin subsided as a result of the deposition of coarse clastics to the west. Initially, this subsidence resulted in the peripheral bulge migrating westward, responding to subsidence of the foreland basin (Fig. 5.8). Initial subsidence of the foreland basin and retrograde movelment of the forebulge results in the initiation of the second sequence within the Walhalla Member. This sequence is thicker in the western Black Hills than the eastern Black Hills, consistent with a transition from the axial basin in the west to the backbulge basin to the east (Fig. 5.8). Clastic sedimentation continued to prograde into the foreland basin, distributing the stresses across the basin and resulting in cratonward migration of the forebulge during deposition of the Mitten Black Shale Member (Fig. 5.8). CONCLUSION Forebulge development in the Black Hills region during deposition of the Walhalla Member is consistent with an elastic strain model for foreland basin development in response to thrust loading, coupled with the margin of an intracratonic basin. A temperature dependent viscoelastic model behaves identical to the elastic model in the shallow lithosphere and is an equally likely possibility (Quinlan and Beaumont, 1984). Corresponding to the advance of the thrust belt, the foreland basin and the forebulge migrate cratonward, responding to increased stresses in the thrust belt. As the forebulge advanced cratonward, it overlapped with the margin of the intracratonic Williston Basin, resulting in yoking of the basins and deposition was continuous across the basins during early deposition of the Walhalla Member. As thrusting halted, the foreland basin subsided responding to increased load from clastic sediments shed off the thrust belt. This resulted in deepening and narrowing of the foreland basin and thrust-ward migration of the peripheral bulge. Finally, with continued tectonic quiescence, clastic material was shed further eastward into the basin, distributing the load more uniformly across the foreland basin, resulting again in cratonward migration of the 158 forebulge. Recognition of a forebulge in the Black Hills region corresponding to tectonic thrusting on the Absoroka Thrust provides for a more detailed analysis of the basin dynamics within this critical part of the basin. REFERENCES Cross, T.A. and Pilger Jr., R.H., 1978. Tectonic controls of late Cretaceous sedimentation, western interior, USA. Nature, 274: 653-657. DeCelles, P.G., 1994. Late Cretaceous-Paleocene synorogenic sedimentation and kinematic history of the Sevier thrust belt, northeast Utah and southwest Wyoming. Geological Society of America Bulletin: 32-56. Dyman, T.S. et al., 1994. Stratigraphic transects for Cretaceous rocks, Rocky Mountains and Great Plains regions. In: M.V. Caputo, J.A. Peterson and K.J. Franczyk (Editors), Mesozoic systems of the Rocky Mountain region, USA. Rocky Mountain Section SEPM, Denver, pp. 365-392. Ettensohn, F.R., 1998. Compressional tectonic controls on epicontinetal black-shale deposition: Devonian-Mississippian examples from North America. In: J. Schieber, W. Zimmerle and P.S. Sethi (Editors), Shale and Mudstones I: Basin studies, sedimentology and paleontology. E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart, pp. 109-128. Flemings, P.B. and Jordan, T.E., 1990. Stratigraphic modeling of foreland basins: Interpreting thrust deformation and lithosphere rheology. Geology, 18: 430-434. Gill, J.R. and Cobban, W.A., 1966. The Redbird section of the Upper Cretaceous Pierre Shale in Wyoming. U. S. Geological Survey Professional Paper 393-A: A-1-A73. Hicks, J.F., Obradovich, J.D. and Tauxe, L., 1999. Magnetostratigraphy, isotope age calibration and intercontinental correlation of the Red Bird section of the Pierre Shale, Niobrara County, Wyoming, USA. Cretaceous Research, 20: 1-27. Jordan, T.E., 1981. Thrust loads and foreland basin evolution, Cretaceous, Western United States. The American Association of Petroleum Geologists, 65(12): 2506-2520. 159 Jordan, T.E., 1995. Retroarc foreland and related basins. In: C.J. Busby and R.V. Ingersoll (Editors), Tectonics of sedimentary basins. Blackwell Science, pp. 331-362. Kauffman, E.G., 1977. Geologic and biologic overview: Western Interior Cretaceous Basin. Mountain Geologist, 14: 75-99. Kauffman, E.G. and Caldwell, W.G.E., 1993. The Western Interior basin in space and time. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 1-30. Martin, J.E., 1996. Disconformities of the lower Pierre Shale (Cretaceous) of South Dakota. Geological Society of America Rocky Mountain Section meeting abstracts with programs, 28(4): 16. Monger, J.W.H., 1993. Cretaceous tectonics of the North American cordillera. In: W.G.E. Caldwell and E.G. Kauffman (Editors), Evolution of the Western Interior basin. Geological Association of Canada, St. John’s, Newfoundland, pp. 31-48. Quinlan, G.M. and Beaumont, C., 1984. Appalachian thrusting, lithospheric flexure, and the Paleozoic stratigraphy of the Eastern Interior of North America. Canadian Journal of Earth Sciences, 21: 973-996. Schedl, A. and Wiltschko, D.V., 1982. Sedimentological effects of a moving terrain. Journal of Geology, 92: 273-287. Schwartz, R.K. and DeCelles, P.G., 1988. Cordilleran foreland basin evolution in response to interactive Cretaceous thrusting and foreland partitioning, southwestern Montana. In: C.J. Schmidt and J. Perry, W. J. (Editors), Interaction of the Rocky Mountain Foreland and the Cordilleran thrust belt. Geological Society of America Memoir, Denver, pp. 489-513. Villien, A. and Kligfield, R.M., 1986. Thrusting and synorogenic sedimentation in central Utah. In: J.A. Peterson (Editor), Paleotectonics and sedimentation in the Rocky Mountain Region, United States. AAPG Memoir, pp. 281-307. Wiltschko, D.V. and Dorr Jr., J.A., 1983. Timing and deformation in overthrust belt and 160 foreland of Idaho, Wyoming and Utah. Amerian Association of Petroleum Geologists Bulletin, 67(8): 1304-1322.