Anhydrite-Carbonate Cycles of the Ordovician

Home , Bartlett Bay, Baumann Fiord, Castile Formation, Eurekan orogeny, Irene Bay, Norwegian Bay

ANHYDRITE-CARBONATE CYCLES OF THE ORDOVICIAN

BAUMANN FIORD FORMATION, ELLESMERE ISLAND, ARCTIC CANADA:

A GEOLOGICAL HISTORY

Grant Dilworth Mossop,

B.Sc.(Hons.), M.Sc.

Thesis submitted for the degree of

Doctor of Philosophy of the

University of London

Geology Department, Imperial College of Science and Technology,

London, England. November, 1973.

A 11

ABSTRACT

The Baumann Fiord Formation of central Elle'.-lmere Island is made up of between 700 and 1,500 feet of concordantly compounded carbonate — anhydrite cycles. The cycles, which average 12 feet in thickness, characteristically comprse a basal shallow—marine lime mudstone facies, followed upwards by an intertidal facies with algal stromatolites, in turn overlain by a supratidal anhydrite facies, the ~•mole bounded top and bottom by erosion surfaces. Each of these regressive sequences is representative of a discrete sabkha sedimentation cycle and the formation as a whole was built up through repetitous superimposition of successive sabkha cycles.

Much of the Baumann Fiord anhydrite underwent early—stage compactional flow, producing layered and laminar anhydrite. Deep burial, accompanied by pervasive recrystallization, further modified the textural character of some of the anhydrite, notably that of the seaward portion of the sedimentary wedge. The Ellesmerian Orogeny (Middle Devonian to Early Mississippian) brought folding and, where deformation was excessive, it caused remobilization and flow of the anhydrite. Subsequent thrust— faulting (Tertiary Eurekan Orogeny) produced shear zones at various levels within the formation. During exhumation the basal reaches of the formation locally underwent hydraulic fracturing. Water from the underlying stratamoved up into the evaporites and hydrated the anhydrite to form secondary gypsum. The excess calcium sulphate liberated during this volume—for—volume replacement process precipitated in the fracture system, forming satin— spar veins.

The latest stages of exhumation have taken place during the permafrost regime of Pleistocene and Holocene times. Hydration by present—day meteoric water promotes gypsification of only the outermost voneer of exposures, and in consequence much anhydrite is preserved essentially unaltered at outcrop. iii

Ari<21.0 .1F,DGETETTTS

This study was rupervi.sed by Dr. D.J. Shearman of the Imperial College, London. His guidance and encouragement in all phases of the research is most gratefully acknowledged. The author feels enriched for having had the opportunity to work so closely with a man of such infectious scientific enthusiasm and indomitable spirit.

Sponsorship of the project was by the Geological Survey of Canada (Dr. D.J. McLaren, Director; Institute of. Sedimentary and Petroleum Geology; Calgary, Alberta, Canada). The G.S.C. under- wrote all expenses incurred in the field component of the study and provided much in the way of scientific and logistical guidance. Direct financial support, in the form of an Overseas Scholarship, was from the Royal Commission for the Exhibition of 1851, London. To both these institutions, the author expresses his sincere gratitude.

Thanks are extended to the following persons for their help with the field phase of the study : Dr. R.L. Christie, who administered the project; Drs. H.P. Trettin, J.Wm. Kerr, R. Thorsteinsson, G. Davies and W.U. Nassichuk, who advised the author on matters of geology and logistics; Mr. D.S. Turner, who proved an able and companionable field assistant; and the pilots of Bradley Air Services and Dominion Helicopter Company, whose skill and willingness to try almost anything allowed the author access to very difficult terrain.

Of the teaching staff at Imperial College, special thanks is extended to Drs. P.R. Bush and G. Evans for much helpful discussion pertaining to numerous facets of the research. Dr. Bush also critically reviewed parts of the manuscript and furnished advice on thesis compilation. For uncounted kindnesses, the author extends particular gratitude to Miss Mary Pugh, without whom the Sedimentology Department would surely cease to function.

Technical advice and assistance from the following Imperial College personnel was greatly appreciated : Si wart and Martin iv

Gill (Sedimentology); john Blount (Rock Cutting)! an Bruce Hougl' (Sedimentary Geochemistry). Mr. J. Gee and his staff did much immaculate photographic work for the author and provided tuition in aspects of photoprinting. Typing was done quickly and efficient- ly by Mrs. Janice Tipping. Brian Moss drafted a number of the figures.

For contributing to productive discussion and for establishing a congenial atmosphere in which to work, appreciation is extended to fellow—students in the Geology Department of. Imperial College, and particularly to Hugh Dunsmore, Tony Corrigan, Stuart Smith and Steven Van der Haar.

Finally, the author wishes to express his warmest appreciation to his wife, Ruth, for her encouragement during times of setback, and for her unfailing grace and devotion. TV'cLE OF CONTENTS Page INTRODUCTION 1 General Statement Purposes 2

CHAPTER 1 GEOGRAPHY AND GEOLOGY OF ELLESMERE ISLAND: A REVIEW 5 1.1 Introduction 6 1.2 Geography of Ellesmere Island 7 Location, Size and Population 7 Climate 7 Physiography 10 1.3 Field Methods 14 Access and Travel 14 Exposure Conditions 15 1.4 History of Geological Investigation 16 Early Work : Pre 1955 16 Recent Work : Post 1955 19 1.5 General Geology 19 Precambrian Shield 19 Franklinian Geosyncline 22 Sverdrup Basin 24 1.6 Stratigraphic Setting of the Baumann Fiord Formation 26 General Stratigraphy 26 Structural Trend and Thickness Variation 29 Ordovician Stratigraphic Nomenclature 30 Baumann Fiord Formation 31 Summary 33

CHAPTER 2 CARBONATE—ANEYDRITE CYCLES OF THE BAUMANN FIORD FORMATION 36 2.1 Introduction 37 2.2 The Sabkha Setting 39 The Recent Sabkha 39 Ancient Sabkha Cycles 46 vi

Page 2.3 The Baumann Fiord Cycle 48 Erosion Surface 48 Description 48 Interpretation 48 Flat Pebble Conglomerate 51 Description 51 Interpretation 54 Lime Mudstone 56 Description 56 Interpretation 56 Algal Stromatolites 63 Descriptions 63 Interpretation 69 Dolomite 73 Description 73 Interpretation 73 Anhydrite 76 2.4 Summary and Conclusions 82

CHAPTER 3 COMPOSITE CYCLES FACIES VARIATIONS AND PALEOGEOGRAPHIC RECONSTRUCTION 86 3.1 Introduction 87 3.2 Section Descriptions 87 3.3 Vertical Variation in Sequence — Compound Cycles 104 3.4 Lateral Variation in Sequence — Facies Intertongu.ing 116 3.5 Paleoenvironmental Evolution of the Study Area 129 3.6 Paleogeography 130

CHAPTER 4 ANHYDRITE — ASPECTS OF DIAGENESIS TECTONISM AND METAMORPHISM 140 4.1 Introduction 141 4.2 Laminar Anhydrite 142 Displacive Growth of Nodular Anhydrite 142 Compaction of Nodular Anhydrite 142 Baumann riord Laminar Anhydrite 145 vii

Page Aspects of Petrography 153 4.3 Massive Anhydrite 159 4.4 Influence of Tectonism on Anhydrite 165 4.5 Other Aspects of the Baumann Fiord Formation's Secondary History 173 Replacement Anhydrite 173 Anhydrite Fracture—Fill 174 Chert Concretions 178

CHAPTER 5 SECONDARY GYPSUM 179 5.1 Introduction 180 5.2 Background Considerations 181 Anhydrite—Gypsum Thermo—chemical Stability Relationships 181 Volume Increase on Hydration 182 Water Availability and Access Modes 184 5.3 Secondary Gypsum of the Basal Baumann Fiord Formation 185 5.4 Surficial Weathering—Product Gypsum 192 5.5 Summary and Conclusions 207

REFERENCES 209

APPENDICES 222 Appendix A Published Reference Maps of the Study Area 223

Appendix B Section Lithologic Logs 225 Appendix C Sulphur Isotope Data for Baumann Fiord

Formation Sulphate Rocks 226 Appendix D New Paleontological Information on the

Baumann Fiord Formation 227 Appendix E Franklinian Overburden Above the Baumann

Fiord Formation — Trold Fiord Region 229 Appendix F Strontium Contents of Baumann Fiord

Formation Anhydrite and Gypsum Rocks 230 viii

LIST OF FIGURES Page CHAPTER 1 Figure 1 Index Map 8 Figure 2 Climatic Statistics - Ellesmere Island. Table. 9 Figure 3 Physiography - Ellesmere Island. Map. 12 Figure 4a Valley and Ridge Topography. Field Photograph. 13 b Dissected Plateau Physiography. Field Photograph 13 Figure 5a Frost-wedged blocks. Field Photograph. 17 Talus fans. Field Photograph. 18 Figure 6 Tectonic Subdivisions of Ellesmere Island. Map. 21 Figure 7 Paleo-physiography of the Frahklinian Geosyncline. Block Diagram. 23 Figure 8 Relationships Amongst the Geological Provinces - Ellesmere Island. Cross-Section. 25 Figure 9 Franklinian Stratigraphic Nomenclature. Correlation Chart. 28 Figure 10a Copes Bay - Baumann Fiord Contact. Field Photograph. 32 b Eleanor River Formation. Field Photograph. 32 Figure 11 Carbonate-Anhydrite Cycles. Field Photograph. 35

CHAPTER 2 Figure 12 Trucial Coast Sabkha. Cross Section. 42 Figure 13 Baumann Fiord Carbonate-Anhydrite Cycle. Schematic Drawing. 49 Figure 14 Purbeckian Sabkha Cycle - Warlingham Borehole. Slab Photograph. 50 Figure 15 Flat-Pebble Conglomerate. Field Photograph. 52 Figure 16a Flat-Pebble Conglomerate - Micrite Clasts. Photomicrograph. 53 b Flat-Pebble Conglomerate - Exotic Clasts. Photomicrograph. 53 Figure 17 Cross-Laraination - Lime Mudstone Facies. Field Photograph. 57 Figure 18 Massive Limestone. Field Photograph 58 Figure 19a Laminated Lime Mudstone. Photomicrograph. 59 ix

Page Figure 19b Laminated Lime Mudstone. Photomicrograph. 59 Figure 20 Digitate Stromatolites. Field Photograph. 65 Figure 21 Polygonal Stromatolites. Field Photograph, 66 Figure 22 Domal Stromatolites. Field Photograph. 67 Figure 23 Coalesced Algal Heads. Field Photograph. 68 Figure 24 Recent Algal Mats — Trucial Coast. Field Photograph. 70 Figure 25 Dolomite. Photomicrograph. 74 Figure 26 Mudcracked Lime Mudstone. Field Photograph. 75 Figure 27 Nodular Mosaic Anhydrite. Field Photograph. 77 Figure 28 Nodular Anhydrite. Slab Photograph. 78 Figure 29 Recent Anhydrite Crystal Fabric. Photomicro- graph. 79 Figure 30 Baumann Fiord Anhydrite Crystal Fabric. Photomicrograph. 80 Figure 31 Winnipegosis (M. Dev.) Nodular Anhydrite. Photomicrograph. 83

CHAPTER 3 Figure 32 Baumann Fiord Sections 1-1a-2-3. In Rear Pocket Figure 33 Baumann Fiord Section 4-7-5-6. In Rear Pocket Figure 34 Thickness and Exposure Data. Chart. 89 Figure 35 Overfolds in B Member. Field Photograph. 92 .Figure 36 Fracture System in Folded Carbonates. Field Photograph. 93 Figure 37 Section 1. =Field Photograph. 95 Figure 38 Section la. Field Photograph. 96 Figure 39 Section 2. Field Photograph. 97 Figure 40 Section 3. Field Photograph. 98 Figure 41 Section 4. Field Photograph. 100 Figure 42 Section 7. Field Photograph. 101 Figure 43 Section 5. Field Photograph. 102 Figure 44 Section 6. Field Photograph. 103 Figure 45 Thick Stromatolites. Field Photograph. 109 Figure 46 Thick Flat—Pebble Conglomerate. Field Photograph. 110 Figure 47 Algal Mat Breccia. Field. Photograph. 111 Page Figure 48 Vertical Jacking by Displacive Anhydrite Growth. Line Drawing. 114 Figure 49 Facies Pattern in Baumann Fiord Formation. Schematic Cross-Section. 117 Figure 50 Correlations Amongst Section 1a-2-3. Cross- Section. 120 Figure 51- C Member'Facies Transition from Anhydrite to Carbonate. Panoramic Field Photograph. 123 Figure 52a Redbed Quartz Sandstone and Marl. Sample Photograph. 126 b Redbed Quartz Sandstone. Photomicrograph. 126 Figure 53 B Member Limestones. Field Photograph 128 Figure 54 Biopelmicrite. Photomicrograph. 129 Figure 55 A Member Paleogeography. Map. 132 Figure 56 B Member Paleogeography. Map. 134 Figure 57 C Member Paleogeography. Map. 136 Figure 58 Extent of Baumann Fiord Sabkha Complex - Arctic Islands. Map. 138 Figure 59 Trucial Coast Sabkha. Field Photograph. 139

CHAPTER 4 Figure 60 Anhydrite Compaction Mechanisms. Line Diagram 144 Figure 61 Laminar Anhydrite. Field Photograph. 146 Figure 62 Aligned Fabric - Laminar Anhydrite. Photomicrograph. 148 Figure 63 Transition Continuum from Nodular Anhydrite to Laminar Anhydrite. a Mosaic Anhydrite. Field Photograph. 149 b Distorted Mosaic Anhydrite. Field Photograph. 150 Laminar Anhydrite. Field Photograph. 150 Figure 64 Elongate Anhydrite Nodules. Block Sample Photograph. 154 Figure 65a Anhydrite 7abric in nodule elongation-normal section, Photomicrograph. 156 Anhydrite fabric in nodule elongation-parallel section. Photomicrograph. 156 xi

Page Figure 66 Crystallographic and Optical Orientation of Anhydrite. Line Diagram. 157 Figure 67 Massive Anhydrite. Field Photograph. 160 Figure 68 Massive Anhydrite. Photomicrograph. 161 Figure 69 Massive Anhydrite with Proto-Lamination. Photomicrograph. 162 Figure 70 Chevron Folds - Locality 8. Field Photograph. 167 Figure 71 Minor Folds in Limbs of Chevron Folds. Field Photograph. 167 Figure 72 Boudinage in Limb of Chevron Fold. Field Photograph. 168 Figure 73 Aligned Fabric in Folded Anhydrites. Photomicrograph. 170 Figure 74 Microfolds in Tectonized Anhydrite near Shear Zone. Field Photograph. 172 Figure 75 Replacement Anhydrite. Photomicrograph. 175 Figure 76 Fracture-Fill Anhydrite. Photomicrograph. 176 Figure 77 Chert Concretions. Field Photograph. 177

CHAPTER 5 Figure 78 Anhydrite-Gypsum Volume Relationship. Chart. 183 Figure 79 Basal Baumann Fiord Gypsum. Line Diagram. 186 Figure 80 Secondary Gypsum with Satin-Spar. Field 187 Photograph. 5s- Figure 81 Porphyroblastic Secondary Gypsum. Photomicro- graph. 189 Figure 82 Massive Anhydrite Grading to Alabastrine Gypsum. Photomicrograph. 194 Figure 83 Surficial Gypsum Crust. Field Photograph. 197 Figure 84a Alabastrine Gypsum After Felted-Aligned Anhydrite. Photomicrograph. 199 b Preferentially Gypsified Zones. Photomicro- graph in 45° Position. 200 c Preferentially Gypsified Zones. Photomicro- graph with sensitive tint. 200 xii

Page Figure 85 Gypsum Optical Indicatrix, Line Diagram. 202 Figure 86 Alabastrine Gypsum with Preferred Optical Fabric. Photomicrographs. a Extinction Position. 205 b 45° Position. 206 c Sensitive Tint. 206 —Ma 'PaoTj utrawn7 uo Lutl@q. 114.T1P, uu-eutres: .1013T/1 u;--aldre3 'opaTd2TTuoJd 1

INTRODUCTION

6" 2

General Statement

At depth, anhydrite (CaS0 ) is the stable form of calcium 4 sulphate. But as anhydrite rocks approach the surface in the course of uplift and erosion, steadily decreasing temperature eventuallly renders anhydrite unstable as a mineral phase, and gypsum (CaS0 .2H 0) assumes thermodynamic preference. Once 4 2 calcium sulphate rocks have been exhumed above the anhydrite— gypsum transition depth (which commonly lies at 3,000-4,000 feet), hydration of anhydrite to secondary gypsum is dependant solely on the availability of water. In most parts of the world, hydro- logic conditions are such that water is made available at some time during the final stages of exhumation and it is thus relatively rare to find anhydrite rocks preserved at outcrop. One exception to this rule occurs in the permafrost zone of the arctic region, where all the water that would normally be available for the hydration of the near—surface anhydrite is perennially locked up as ice. In the Canadian Arctic Archipelago, there are a number of evaporite formations in which anhydrite is exposed at outcrop, with little or no surficial gypsum to obscure the rock's primordial aspect. The Baumann Fiord Formation of Ellesmere Island is one such instance.

To date most studies of ancient anhydrite sequences have been based, out of necessity, on material from borehole cores. In an outcrop study such as this, the principal advantage is one of being able to observe both vertical and lagieral variation in anhydrite rocks, a luxury that is not afforded to workers utilizing borehole material. The prospect of examining ancient anhydrite rocks at outcrop was thus one of the foremost induce- ments to the undertaking of this study.

Purposes

In that this study deals with rocks that have not been

There are notable exceptions; eg. Holliday (1966, 1968) made studies of Carboniferous anhydrite rocks exposed at outcrop in Spitsbere;en. 3

previously examined in detail, its first purpose is that of description. In this regard, the text encompasses aspects of the macroscopic characteristics of the formation, as exemplified in its field appearance, and microscopic features, as revealed in thin—section. The other purpose of the study is that of interpretation. Included is an interpretive analysis relating to all phases of the formation's geological history, from the time of its deposition through to the present—day.

The principal themes for discussion and documentation are as follows: (i)A review of the regional stratigraphic and structural setting in which the Baumann Fiord Formation resides (Chapter 1). (ii)A comprehensive description of the rocks' structures and textures with particular emphasis upon the nature of the formation's basic genetic unit, the carbonate—anhydrite cycle. The foremost objective of this endeavour is to effect an environmental reconstruction of the Baumann Fiord rocks, as based on direct comparison of the Ordovician evaporites with those of the Recent sabkha environment, with reference to all directly analogous features (Chapters 2 and 3). (iii)A discussion of the inherent differences between the Recent sabkha anhydrites and those of the Baumann Fiord Formation with accompanying documentation of the diagenetic, tectonic and metamorphic modification of the latter (Chapter 4). (iv)An examination of the nature and origin of those secondary gypsum rocks that are developed in the Baumann Fiord Formation, including that gypsum which stems from pre—permafrost hydration and that which is the result of present—day hydration by meteoric water (Chapter 5).

It is hoped that this document on the Baumann Fiord Formation will constitute a contribution to geological knowledge in a number of spheres: first, as a guide to the field occurence, lithology, facies variation and diagnostic features of the — formation as a whole; second, as an example of ancient sabkha facies evaporite genesis with its accompanying palcogeographical - and paleoclimatical implicetions; third, as an exemplifjcation of the profound textural transformation teat may result from 4

• flow and recrystallization of sedimentary anhydrite:; and fourthly, as a documentation of the secondary gypsum fabrics that result from various hydration processes. 5

CHAPTER 1

GEOGRAPHY AN]) GEOLOGY OF ELLESMERE ISLAND: A REVIEW 6

1.1 Introduction

This chapter deals primarily with background material. Essentials of the geography of the study area are outlined in order to familiarize the reader with the physical environment in which the Baumann Fiord Formation crops out. Fundamental aspects of the geology of the area are introduced in order that the reader may have an idea of the regional stratigraphic and structural setting in which the Baumann Fiord Formation resides. Much of the discussion contained in subsequent chapters relates back to material introduced in this first chapter.

The chapter begins with a short summary of the physiography and climate of the study area, followed by a brief review of field conditions and field methods in general. Aspects of access and travel are also discussed. Some of the history of discovery and exploration of thearea is reviewed insofar as it relates to early geological investigations.

The latter part of the chapter deals with the geology of the Ellesmere Island area, starting with a summary of the regional geology and becoming more and more specific to the Baumann Fiord Formation towards the end. First is an overall look at the major geological provinces of the northern Arctic Islands, their genesis and tectonic evolution. Next is a summary account of the Upper Proterozoic to Upper Devonian sequence in the area, followed by a more detailed examination of the Ordovician sequence. Finally, there is a general ccount of the Baumann Fiord Formation itself, its stratigraphic position, thickness variation, field appearance and lithologic character. Detailed descriptive and interpretive analysis of the Baumann Fiord Formation is confined to subsequent chapters. 7

1.2 GeorTaphy of Ellesmere Island

Location, Size and Population

Ellesmere Island_ is the most northerly S.sland of the Canadian Arctic Archipelago (Figure 1). Its east coast lies immediately adjacent to the northwest coast of Greenland. To the south and west of Ellesmere lie the remaining islands of the archipelago. To the north is the Arctic Ocean. The northern tip of Ellesmere Island lies less than 500 miles from the North Pole at lattitude 83°N. The total land area of Ellesmere Island is about 85,000 square miles, roughly equivalent to that of Britain. Its year- round population is about 200, all of which is contained in three widely separated communities; a small Eskimo settlement at Grise Fiord on the south coast (the most northerly native habitation in Canada), a weather station at Eureka on the west coast, and a Canadian Forces base and weather station at Alert on the north coast (Figure 3).

Climate Being over 1,000 miles inside the Arctic Circle, Ellesmere Island is subject to extremes of seasonal daylight bias. For a four-month period in winter (late-October to late-February), the island is without sunlight. In summer (late-April to late- August) the sun is above the horizon 24 hours a day. With continuous sunshine in summer, there is very little daily fluctuation in temperature, commonly only 4-5°F. In the warmest month (July), the mean temperature on most parts of Ellesmere is about 40°F (Figure 2). Annual precipitation is very small (Figure 2), so small in fact that Ellesmere may be legitimately classified as an "arctic dessert° (Fortier et al., 1963; p. 6). Most of the rainfall is in August and the bulk of the snowfall may be expected in September-October or in late spring. The small amount of snow that does accumulate is normally melted away by mid-June. For the geologist, Ellesmere Island affords a workable 8

Figure 1. Index map locating the Ellesmere Island study area. 9

Table of Climatic Statistics for Ellesmere Island

(adapted from Fortier et al., 1963; p.6)

Temperature (°F)

Coldest Month Warmest Month

Mean —15 to —39 38 to 44 Maximum —9 to —33 42 to 51 Minimum —26 to —46 33 to 57

Extremes 75 to —63

Annual Precipitation (inches)

Rainfall 1.74 to 3.93 Snowfall 15 to 64.9

* Total Precipitation (inches rain) 2.89 to 8.91

* Conversion factor : 1 in. rain E 13 in. snow

Figure 2. 10

field season of up to three months, from mid—June to early— September. The first part of any one season may be delayed due to late snow—melt runoff. Similarly, work may be curtailed before September as a result of early snowfall. The middle part of the season, July in particular, s normally very pleasant: continuous sunlight, very little rain and a complete lack of insect pests due to the relatively cold conditions. The only working hazards are those contingent upon summer blizzards, which may occur at unexpected times, and those imposed by fog, which may settle quickly and hug the ground for periods of up to two weeks.

Physiography

The physiography of the study area on southern and central Ellesmere Island is directly related to the bedrock geology, as evidence by the fact that the trend of the major physiographic provinces is essentially parallel to the regional structural trend (compare Figure 3 with Figure 6). Outcrops of the Baumann Fiord Formation are restricted to the valley and ridge province in the west and north of the island and to the dissected plateau region in the east (Figure 3). The valley and ridge areas consist of parallel or sub—parallel ridges separated by broad U—shaped glacial valleys (Figure 4a). In the area north of Baumann Fiord and adjacent to Trold Fiord, the ridges reach elevations of 3,000 feet or slightly more, the valleys being only a few hundred feet above sea level at most. Characteristically, the valleys are cut along the crests of anticlines and the ridges are comprised of the upturned limbs of adjacent synclines. Bedrock strata normally strike parallel to the valley sides and dip into the slope at angles up to 50 degrees. In the other region, the dissected plateau province in the east, the strata are essentially flat—lying and are best exposed along fiord—sides or in the steep canyon walls of deeply incised streams (Figure 4b). Oh Bache Peninsula, the plateau averages about 1,500 feet in elevation. Reliable physiographic maps of the study area on the scale of 250,000:1 (4 miles to the inch), with a contour interval of 500 feet are available from the Canada Surveys and 11

Figure 3. Physiographic subdivisions within the Baumann Fiord Formation outcrop belt on Ellesmere Island (Adapted from Roots in Fortier et al., 1963; p.267).

61110111•Ik Highlands Norwegian Bay Lowlands Eureka Sound Uplands Dissected Plateau Valley and Ridge

Baumann Fiord Formation Outcrop

The great majority of the Baumann Fiord outcrop falls into either the Valley and Ridge Area (Trold Fiord region) or the Dissected Plateau Province (Bache Peninsula region). 12

9s° 415.

es°

1.7)c) issIC

az°

Bache Peninsula

79°

78.

NORWEGIAN BAFFIN

BAY BAY

77 .

Miles 0 10 2:3 Sr. 40 SO W qO rise Fiord

1 95* 190° F35. Figure 4a. Valley and Ridge topography in the Troia Fiord region. View is northeast from Section 1. Kountain occupying the left centre of the photograph is some 5 miles distant and rises 3,100 feet above the valley floor.

Figure db. PiccectE,d Plateau physiography in the 7:ache Penimmla area, viewed frcm the air. The plateau averages 1,500 feet in elevation. 14

Napping Branch (Appendix A). Comprehensive vertical aerial photographs with complete stereoscopic overlap are also avai7able.

Ellesmere Island is almost devoid of vegetation. There are no trees or bushes, but locally, there are patches of creeping saxifrage and liather. In the low—lying areas that become water—logged in summer, small fields of Arctic cotton—grass may take hold. Sparce though the vegetation may seem, it is sufficient to support a modest population or caribou, muskox, lemming and hare. The remainder of the mammals. including the polar bear, arctic wolf and arctic fox are dominantly carnivorous.

For a more complete review of the biology, zoology and geography of the Arctic Islands, the reader is referred to Fortier et al. (1963; pp. 1-14).

1.3 Field Methods

The logistics of organizing and carrying out geological field work in the Canadian Arctic Islands are outlined in a short guide report by Kerr (1973). For those unfamiliar with the problems involved, it is a uaeful introduction.

Access and Travel

Access to the Arctic Islands is greatly fascilitated by the twice—weekly scheduled airline service that operates from Montreal, Quebeb and Edmonton, Alberta to the Canadian Ministry of Transport base at Resolute Bay on Cornwallis Island (Figure 1). Resolute is the main distribution centre for the northern islands and a number of air charter companies use it as their base. By chartering aircraft, it is relatively easy to transport equipment and provisions to any of the three settlemen:son Ellesmere Island (Figure 3), all of which have well—maintainEZ gravel airstrips.

For the present study, the station at Eureka was chosen as the most convenient base from which to operate. From there, a series of fear successive field camps were established in the summer of 1971 and a further four set up in the summer of 1972 15

(Mossop, 1972a, 1973a). The localities visited each year were those corresponding to sections 1 through 4 and 5 through 8 respectively (Figure 6). Transport was provided by single- engined Otter aircraft chartered from Eureka. Each camp was occupied for a period of one to two weeks, at the end of which the Otter would come from Eureka by previous arrangement to carry out the move to the next: locality. The aircraft used for each camp—move was also utilized to re—stock camp provisions, run the mail, and transport rock samples back to Eureka.

The Otter is admirably suited to Arctic conditions. It has short landing and take—off capabilities, a return—trip range of up to 350 miles and a payload capacity that is easily sufficient to accommodate all the food and equipment required by a two—man party. Most pilots are willing to attempt a landing on almost any approachable terrain but in general they prefer to land on alluvial terraces, gravel deltas or raised beaches, places where the oversized tyres of their aircraft can cope With the inevitable surface frost cracks and angular rocks. For localities 1 through 7 inclusive, the Otter was able to land within a few miles of the desired exposures. Access to locality 8 was by Bell 206 Ranger helicopter.

At each locality, examination of the Baumann Fiord Formation and the associated units was carried out in a series of foot traverses. Where exposure warranted, it was possible to follow the formation for distances of up to eight miles from camp in one direction then do the same in another direction. Considerable areas of ground were examined in this way.

Exposure Conditions

The whole of Ellesmere Island is well within the Arctic permafrost zone, meaning that the groundwater is frozen all year round, in some places to depths of 1,200-1,500 feet. In summer however, a surficial thaw zone develops. Though usually no more than one or two feet thick, it can give rise to patches of wet and muddy ground in the low—lying areas of the valleys. On the mountain sides, repeated summer thaws may also promote a 16

measure of in situ groundete.r hydration of anhydrite to form a white crust of powdery cyI:eum, normally of veneer thickness but, where drainage 13 impeCed, up to 2 or 3 feet thick. The freshest anhydrite was found in places where erosion rate exceeded hydration rate, notably in stream cuts and small landslips.

Scree cover is in fact the major handicap to the working geologist. Many slopes consist of nothing but a series of coalesced talus fans generated by active frost—wedging of the bedrock (Figure 5). The Baumann Fiord Formation, though essentially recessive in its weathering pattern, is usually spared from complete talus obliteration by the fact that the carbonate bands within the anhydrite are normally resistive enough to hold certain ridges proud from the talus slope (Figure 5b). The best and most continuous exposures are those occupying these buttresses between the talus—choked gullys.

1.4 History of Geological Investigation

Early Work : Pre 1955

The first geological observations made on Ellesmere Island were those by the Norwegian geologist Per Schei, a member of Otto Sverdup's 1898-1902 expedition aboard the'vessel '}Frain" (Schei, 1903; Holtedahl, 1917). Sverdrup's party were responsible for the initial discovery, exploration and mapping of most of western Ellesmere Island and Axel Heil:erg Island. Many of the prOminant physiographic features of the region were named by members of the expedition and many others subsequently named in their honour. Baumann Fiord, for instance, takes its name from the expedition's second in command, CEptain Victor Baumann (Sverdrup, 1904).

Over the next 50 years, scattered geological information was collected in the region by a number of explorers, adventurers and Canadian government officers. These are successively chron- icled in the works of Low (1906), McMillan (1910), Young (1926), Kindle (1939), Armstrong (1947) and Troeleen (1950). Figure 5a. Frost—wedged blocks of Baumann Fiord anhydrite strata. Section la. 18

Figure 5b. Coalesced talus fans between buttress of Baumann Fiord Format; on strata. Dark—toned talus comprises mainly limestone derived from the upper cliff—forming Eleanor River Formation. Light—toned talus comprises powdery 'weathering—product' gypsum. Ice—capped peak in the centre of the photograph is some 11 miles distant. Vertical cliff at right rises about 450 feet above the talus slope. Locality 5. 19

Recent Work,: Post 1955

The Geological Survey of Canada undertook the first large— scale geological reconnaissance of the Canadian Arctic Islands in the year 1955 (Operation Franklin, Fortier et al., 1963). Many parts of Ellesmere Island were mapped during the course of the operation and the foundations of the stratigraphic nomenclature for the region were established. Yumerous maps and reports on the geology of specific areas of Ellesmel.e Island have since reached publication, as have the results of many regional stratigraphic studies and some detailed paleontological and sedimentological studies. A complete bibliography of the accumulated references has been compiled by Christie (1971). Works that have a direct bearing on the present study are referred to in appropriate sections of subsequent text.

The comprehensive mapping programme of the Geological Survey of Canada is now vitually complete and reliable geological maps for the whole of the study are available (Appendix A). Most of the maps, compiled using air photo interpretation in conjunction with considerable surface section control, are on the scale of 250,000:1 (4 miles to the inch).

1.5 General Geolocz

Precambrian Shield

Precambrian crystalline rocks crop out over a large area of southeastern Ellesmere Island (Figure 6). K—Ar age determinations on suitable material from the region yield dates that clus- ter at 1,700 m.y., an indication that the rocks were deformed during the Hudsonian Orogeny (Winless, 1969). This terrain is thus considered an extention of the Churchill Province of the Canadian Shield (McGlynn in Douglas, 1970; pp. 100-101). It is thought that the Hudsonian rocks extend beneath the Franklinian Geosyncline to the north and west where they are considered to have acted as basement (Trettin et al., 1972). For the pertinent tectonic and isotopic age maps, refer to Appendix A. 20

Figure 6. Tectonic subdivisions in the Ellesmere Island part of the Innuition Province.

Sverdrup Basin (C-T)

~ Eugeosyncline Franklinian (-e-n) ~ l.~i 0 geosynclinc----Folded " Stable

Precambrian Shield (Hudsonian)

Baumann Fiord Formation Section Localities .1

The folded Franklinian I,Iioseosyncline is normally referred to as the Central Ellesmere IsICl.nd Fold Belt. The Stable Franklinian Miogeosyncline (unfold~d) is termed the Arctic Platform. 21

Bache Peninsula

NORWEGIAN BAY

95° \135. Aso 22

Franklinian Geosyncline

Franklinian rocks, which range in age from late Proterozoic to Late Devonian, may be meaningfully partitioned into two parallel belts. One crops out extensively on the north coast of Ellesmere Island and is of eugeosynclinal aspect, and the other extends in a broad zone almost the full length of the island from southwest to northeast and is of miogeosynclinal aspect (Figure 6). The miogeosynclinal belt passes transversely into the eugeosynclinal belt but the transition zone is exposed only in certain parts of northeast Ellesmere Island, younger sediments of the Sverdrup Basin having been superimposed elsewhere (Figure 6).

The eugeosynclinal region is dominantly comprised of metasedimentary and metavolcanic rocks along with some plutonic rocks (Trettin, 1969a, 1969b, 1971a, 1971b). The sediments were derived from a periodically uplifted erogenic arc in the present offshore region to the northwest, referred to as the Pearya Geanticline (Trettin et al., 1972). Some of the sediments debouched into the adjacent submarine depression (the Hazen Trough) where they constituted a flysch facies (Trettin et al., 1972, p. 98; block diagram reproduced here as Figure 7). Volcanism Was apparently confined to the southeastern edge of the geanti- clinal region, along the northwestern periphery of the Hazen Trough (Trettin et al., 1972). Metamorphism and plutonism in the region was perpetrated as a result of the tectonic activity of the Pearya Geanticline. This tectonic activity culminated in the Middle Devonian to Early Mississippian Ellesmerian Orogeny, which began in the northwest and spread southeastwards, terminating the life of the entire Franklinian Geosyncline.

The miogeosynclinal part of the Franklinian belt consists primarily of carbonate rocks, but carries some terrigenous rocks and evaporites. In general, the sequence is interpreted as being of shallow—marine shelf or bank origin (Figure 7). Along the western edge, the sequence is about 40,000 feet thick (Thorsteinsson and Tozer, 1970), but all individual rock units thin towards the east where they onlap onto the Precambrian Shield (Figures 6 and 7). The present erosional limit of the 23

STABLE MOBILE BELT REGION Pea rya Arctic TECTONIC DIVISIONS Geanti- Franklinian Geosyncline cline Platform Hazen Trough SEDIMENTARY PROVINCES clastic province carbonate province SUBPROVINCE 1 Hazen Trough VOLCANISM (approximate extent) Taconite Zebra Imno Cape Allen Boy -Read Bay FORMATION River Cbt s Phil- lips Member A B C ao 33 EE - EE SEDIMENTS TO O fi 3 To" 18A "67., rn qe, assn oa 12EZE.

PHYSIOGRAPHIC DIVISIONS Uplands .g Shelf Hazen Trough Shelf a. coalescing fans co canyon fon-volley % 8a sea-leap( principal Current; // directions / / j ,/ / / // / KM / / 2 / /

intermilfentUPLIFT. -4 LUTONISM, METAMORPHISM NW 6 SE 0 :DO 200KM

Figure 7. Block diagram of the paleo—physiographic and paleo- tectonic relationships between the Pearya geanticline and the Franklinian Geosyncline. The Baumann Fiord Formation is confined to the 'Shelf' area in the southeast. Diagram by Trettin et al., 1972; p.98. 24

sedimentary rocks along the edge of the shield is thought to be only slightly west of the original line of 'pinch—outl, the greater part of the present shield area having 1.:een positive during much of Early Paleozoic time.

The Franklinian Iliogeosyncline is traditionally partitioned into two parallel belts (Thorsteinsson and Tozer, 1970): the folded region in the west and north (termed the Central Ellesmere Fold Belt), and the stable belt along the margin of the shield (termed the Arctic Platform) (Figure 6). The line of deoar- cation between the two (termed the ( hinge°, Figure 7) marks the limit of Ellesmerian tectonic influence. In the folded belt, broad open folds predominate, commonly having amplitudes of 2,000-3,000 feet and wave—lengths of 2-3 miles. The deformation, 'caused by upslope overriding from the north and northwest, (was) of the decollement type, but did not involve significant low—angle thrustfaulting'1 (Trettin et al., 1972, p. 163). Post—tectonic normal faults are common, however. In the stable belt, the strata, though unfolded are normal—faulted in a similar manner. Exposures of Baumann Fiord Formation examined during the course of the present study fall in both the folded and stable regions (Figure 6).

Sverdrup Basin

After the Ellesmerian Orogeny, a crustal depression developed over much of the area formerly occupied by the Franklinian Geosyncline. This depression, termed the Sverdurp Basin, was the site of almost continuous sedimentary deposition from Pennsylvanian to Late Cretaceous time (Figure 6). In the axial region, some 35,000 feet of predominantly elastic sediment accumulated, most of it derived from Franklinian and Precambrian terrain to the south and east (Trettin et al., 1972). The Sverdrup Basin originally covered an area considerably greater than at present, though its exact bounds are not known. On central Ellesmere Island, there are Mesozoic outliers some 50 miles east of the main erosional edge. Much of the Franklinian terrain immediately adjacent to the present edge of the Sverdrup Basin must have been covered during most of the Mesozoic. SVERDRUP BASIN (C—T)

FOLDED — _ _ FRANKLINIAN MIOGEOSYNCLINE (£— D) ----•STABLE _ _ _ _

< V PRECAMBRIAN SHIELD

Figure 8. Schematic cross—section showing the relationships amongst the geological provinces in south—central Ellesmere Island (line of section is east—west through the Trold Fiord region, from the ice—capped shield in the east to Eureka Sound in the west — Figure 6). The Baumann Fiord Formation crops out in both the 'stable' and the 'folded' parts of the Franklinian MiOgeosyncline. 26

The Eurekan Orogeny of middle Tertiary time terminated the life of the Sverdrup Basin. Uplift was accompanied by open folding and thrust faulting in the axial region (Trettin et al., 1972). The degree and extent of Eurekan remobiliation of Franklinian rock has not been fully evaluated as yet. Thorstein- sson and Tozer (1970) suggested, however, that some low—angle thrust faults that cut Franklinian strata on west central Ellesmere island may have stemmed from Eurekan earth movements. These thrust faults, which often follow Lower Paleozoic evaporite horizons, are discussed in some detail in Chapter 4. The schematic cross—section shown in Figure 8 summarizes the relationships amongst the various geological provinces in the area of present interest. The Baumann Fiord Formation crops out in the folded Franklinian terrain in the westl 'immediately adjacent to the edge of the Sverdrup Basin. From there it can be followed eastward into the unfolded region, right to the edge of the shield.

For a more comprehensive summary account of the geology of the Canadian Arctic Islands the reader is referred to Thorstein- sson and Tozer (1970). The tectonic history of the region has been comprehensively reviewed by Trettin et al. (1972).

1.6 Stratigraphic Setting of the Baumann Fiord Formation

General Stratigraphy

A generalized stratigraphic chart of Franklinian formations in the study area on central Ellesmere Island is reproduced in Figure 9.. The illustrated 'Western Composite Section' is applicable over an extended area of the folded belt in the region between Baumann Fiord in the south and Carion Fiord in the north (see Figure 6). The 'Eastern Composite Section' is applicable in the region of Bache Peninsula in the east and throughout the stable Arctic Platform belt that borders the shield (Figure 6). Where equivalent time—stratigraphic units differ in formational name, the distinction between the nomenclature of eastern and 27

Figure 9. Stratigraphy and Correlation of Franklinian Formations, Central Ellesmere Island (adapted from Thorsteinsson and Tozer (1970a), Kerr(1968) and Christie (1967)). Both illustrated sections are somewhat.gcneralized in order that they may be applicable over a large area.

Conformable Contact Relationships Unknown

Anhy. Anhydrite Dol. Dolomite Ls. Limestone Lithologic Abbreviations Sand. Sandstone Sltst. Siltstone Sh. Shale

The 'Western Composite Section' refers primarily to the region of Trold Fiord. The 'Eastern Composite Section' refers primarily to the Bache Peninsula area.

d COMPOSITE SECTION FOR COMPOSITE SECTION FOR ies io WESTERN ELLESMERE IS. EASTERN ELLESMERE IS. Per Ser

,,,,,,...„,-...... ".....",,,,,.....„.

n Cape Rawson ss. ia n

Devo ...... ~Pe..~4~nrew~~~~....prow•ekrsi.e..

s h. Cape Phillips sltst. n

ia Allen Bay — Is. r

lu Read Bay dol. Si

Irene Bay Is. sh. Irene Bay Is. sh. Gp Gp

Thumb Mtn. Is. l Thumb Mtn. Is. l llis •

nwa nwa dol. dol.

ian M Bay Fiord Is. Bay Fiord Is.s. Cor Cor ic anhy• anti

dov Eleanor River Is. Eleanor River Is.

Or anhy. anhy. Baumann Fiord Baumann Fiord Is. Is. L ulinam uerod Is. d(g: Bay Is. gape Is. Cass h 1 1. U

ian M Parish Glacier Is. Parish Glacier Is. br

m Scoresby Bay dol. _ _ web ,■• ■•• •••■■• •••••■• r••■• •••Now

Ca L Rawlings Bay ss.

coc Ella Bay dol. ,.. -c, co ....— 29

western schemes reflects a fundamental lithological difference between the sediment deposited in close proximity to the shield (Eastern Composite Section) and that deposited further out (Western Composite Section).

Upper Hadrynian and Lower Cambrian formations are confined to the east—central coast of Ellesmere Island (Kerr, 1967a). These were deposited in a small basin that fore—ran the Frank- linian Miogeosyncline proper. By Middle Cambrian time, however, the Franklinian Miogeosyncline was fully established and sediments of the Parish Glacier Formation were deposited over a large area. Following the Upper Cambrian hiatus (Figure 9), Franklinian deposition was apparently continuous throughout the study area until the onset of the Ellesmerian Orogeny. Ordovician carbonates and evaporites accumulated over the whole of the region, followed In the southeast by the Upper Ordovician to Lower Devonian Allen Bay—Read Bay carbonates and in the northwest by the time— equivalent Cape Phillips shales.

The foregoing summary of Franklinian stratigraphy is intended to provide the reader with a rudimentary knowledge of the overall stratigraphic framework of which the Baumann Fiord Formation is a part. For the present work, a more detailed stratigraphic knowledge is not necessary. Readers interested in a comprehensive account of the Franklinian stratigraphy of Ellesmere Island are referred to Kerr (1967a, 1968 and in preparation) and to Thorsteinsson and Tozer (1970).

Structural Trend and Thickness Variation

The structural trend of the Franklinian Miogeosyncline in the southern portion of the study area is essentially north - northeast. Along the strike, component formations are generally of uniform thickness. But perpendicular to the strike, they thin markedly towards the east. In the eastern part of the study area,the sequence trends east — northeast and thins towards the south, again roughly perpendicular to the strike. As a rule, then, the structural trend of the Franklinian Miogeosyncline in the study area parallels the depositional 30

strike of its component formations (Figure 6).

Ordovician Stratigraphje Nomenclature

•Ordovician stratigraphic nomenclature for the Ellesmere Island sequence became somewhat confused in the early stages of mapping, partly due to a lack of reliable paleontological. control and pat y through a failure to recognize that there are I li not one but two major evaporite units, the Baumann Fiord Formation and the Bay Fiord Formation (Figure 9). The discrepancies were resolved by Kerr (1967b) and the nomenclature used in this work (Figure 9) is based upon his designations.

The Copes Bay Formation was erected by Thorsteinsson (1963) and has its type section at Copes Bay on eastern Ellesmere Island, where it measures about 4,800 feet. Throughout the study area it comprises shallow—water marine carbonates (Kerr, 1967b, 1968), mainly limestone and argillaceous limestone, but thin anhydrite beds occur in the upper part. Equivalent units on Bache Peninsula, designated by Christie (1967), are of similar lithology. The Copes Bay Formation has been dated as early Canadian (Kerr, 1967b, 1968).

The Baumann Fiord Formation lies conformably on the Copes Bay Formation and its equivalents. Aspects of its stratigraphic relationship to both the underlying and overlying formations are dealt with in the next section of this chapter.

The Eleanor River Formation comprises a sequence of slightly argillaceous, fine—grained limestones which have been interpreted as being of shallow—marine origin (Kerr, 1968). The formation is exposed throughout the study area and is directly correlative with the type section on Cornwallis Island (Thorsteinason, 1968). A reference section on Ellesmere Island (Kerr, 1967b), is 2,300 feet thick. Fossil collections over a wide area (reviewed in Kerr, 1967b) indicate that the Eleanor River For- mation is of late Canadian (Arenigian) to Chazyan age.

The Cornwallis Group on Ellesmere Island is comprised of three formations: the basal Day Fiord Formation, a unit of mixed 31

lithologies dominated by evaporites and carbonates; the Thumb Mountain Formation, an argillaceous and dolomitic limestone; and the Irene Bay Formation, a distinctive and highly fossiliferous argillaceous limestone. The Cornwallis Group has its type section on Cornwallis Island (Thorsteinsson, 1958). Separate type sections for the three component formations are on Ellesmere Island (Kerr, 1967b).

Baumann Fiord Formation

The type section of the Baumann Fiord Formation, as erected by Kerr (1967b), is located on the west side of Trold Fiord, Ellesmere Island (Locality 10 of Kerr, 1967b; Locality 1 in the present work, Figure 6). At the type locality and throughout the study area as a whole, the formation is conformably underlain by the resistive limestones of the Copes Bay Formation and its equivalents (Figure 9). A field photograph of a typical exposure of the contact is shown in Figure 10a. The Eleanor River Formation lies conformably above the Baumann Fiord Forma- tion throughout the study area. Because of erosional under- cutting of the recessive Baumann Fiord Formation, the resistant limestones of the Eleanor River Formation characteristically stand out as a prominent bluff or cliff (Figure 10b).

In the western region, the Baumann Fiord Formation averages about 1,500 feet (Sections 1, la, 2 and 3; Figure 6). It thins gradually towards the craton and in the vicinity of the shield averages 800 feet (Sections 4, 5, 6 and 7; Figure 6). For mapping purposes, the formation is divided into three members (Kerr, 1968). The basal A Member is comprised of rhythmically alternating bands of carbonate and anhydrite (Mossop, 1972a). Though there is considerable variability, each carbonate—anhydrite couple averages about 12 feet in thickness with anhydrite generally predominating over carbonate by a ratio of three to one. The

Logs of the sections measured in the course of the present study are figured in Appendix B and these are discussed in detail in Chapter 3. 32

Figure 10a. Uppermost beds of the Copes Bay Formation limestones with the recessive—weathering Baumann Fiord Formation above. Locality la.

Figure '21E.enor 7iver Formation limeton,7: fcrmin7 cliff—flankcd r.lateaus above recessive—weathering 7aumann Fiore Formation. View is from the air. Locality 4. 33

A Member is about 1,200 feet thick at the type section but thins to about 600 feet in the area adjacent to the craton (Mossop, 1973a). Member B is comprised of a resistive limestone that averages between 150 and 200 feet. It is a useful marker horizon throughout most of the study area. The upper C Member is essentially identical to Member A in that it is characterized by a series of carbonate—anhydrite couples. As a rule, the C Member does not exceed 200 feet in thickness. Its upper contact with the Eleanor River Formation is conformable and in places gradational. In some parts of the study area, the C Member is not developed and the B Member, which is lithologically similar to the overlying Eleanor River Formation, is mapped as part of the latter.

Dating of the Baumann Fiord Formation is hindered by its characteristic paucity of fossils. It is considered by Kerr (1967b, 1968) to be confined to the Lower Ordovician in that it is underlain by known Lower Ordovician strata (eg. Cape Clay Formation) and overlain by the Lower to Middle Ordovician Eleanor River Formation. On Bache Peninsula the B Member of the Baumann Fiord Formation contains Hystricursus sp. (GSC Localities 47271 and 47275) a trilobite genus which is thought to range from Tremadocian to early Arenigian (Twenhofel et al., 1954). The formation is thus tentatively considered to be entirely Early Ordovician in age, as indicated in Figure 9.

Faunas collected by the present writer from the. B Member at Locality 7 (GSC Locality C-19559) and identified by B.S. Norford yield a probable late Early Ordovician age. Conodonts from the same locality, examined by C.R. Barnes, are thought to be characteristic of late Early Ordovician assemblages. All new paleontological information gleaned in the course of the present study is tabulated in Appendix D.

Summary

The remaining chapters of this thesis are primarily devoted

Sulphur isotope determinations or the Baumann Fiord evaporites are broadly indicative of the Ordovician (see Appendix C). 34

to detailed description and interpretation of the Baumann Fiord Formation. The foregoing discussion, in that it relates to the most rudimentary aspects of the geology of the formation, is of a purely introductory nature and is intended to provide the reader with background knowledge of the field appearance, lithologic character and stratigraphic setting of the formation; i.e. a broad framework upon which subsequent discussion may be meaningfully founded.

Listed below are some points that, by way of summary, warrant iterative emphasis: — The Baumann Fiord Formation is part of an Upper Proterozoic to Lower Devonian sequence of dominantly carbonate rocks which crop out in a broad belt that borders the stable Hudsonian shield. — In the belt immediately adjacent to the shield, the formation is relatively thin ( 800 feet) but it thickens towards the west and north (to an observed maximum of 1,500 feet). — In the inner belt, the rocks are essentially undisturbed but further out in the geosyncline, they are gently folded (Elles- merian Orogeij and locally thrust—faulted (EunkanOrogeny). — The formation is divided into three members: the basal A Member, comprised of a complex series of carbonate—anhydrite couples; the middle B Member, a prominent carbonate unit; and the upper C Member, again a succession of carbonate—anhydrite couples.

The carbonate—anhydrite couple is in fact the basic building block of the entire formation (see Figure 11) and it is on this fundamental unit that attention is focused in the next Chapter. 35

Figure 11. Baumann Fiord Formation carbonate—anhydrite cycles. each of the prominant resistive bands is the carbonate component of a single cycle and the whole of the formation is mace up of rhythmically compounded carbonate—anhydrite couples. Locality 5. 36

CHAPTER 2

CARP.ONA11,24-ANHYDRITE CYCLES OF TIE BAUMANTT FIORD FORMATION 37

2.1 Introduction

Throughout the study area on Ellesmere Island, the Baumann Fiord Formation comprises a sequence of carbonate—anhydrite couples. Each couple consists of a basal limestone—dolomite component and an overlying anhydrite unit; the contact between the two being gradational. The top of each anhydrite unit is marked by a sharp erosion surface upon which rests the carbonate comronent of the succeeding couple. At the type section of the Baumann Fiord Formation (Locality 1, Figure 6), the 1,500 foot sequence is made up of some 125 such couples, each of which comprises about 12 feet of section on average. Over the remainder of the study area, the average thickness of the couples varies to some extent but the general character of the couples is everywhere similar.

Clearly, each carbonate—anhydrite couple represents a distinct genetic unit, self—contained but at the same time closely related to its enclosing couples, as evidenced by their mutual lithologic identity. In sedimentological terms, such a sequence of couples is best explained in terms of cyclical sedimentation, each couple having originated in response to environmental conditions that initially favoured carbonate deposition, followed naturally by a period in which anhydrite deposition was favoured. For the Baumann Fiord couples, each sedimentation cycle must have been self—terminating. The erosion surface at the top marked the cessation of evaporite development and at the same time recorded the re—establishment of environmental conditions conclusive to deposition of carbonate. Sedimentologcally, then, the Baumann Fiord Formation may be viewed in terms of a series of essentially similar sedimentation cycles, each of which in its self—termination gave rise to the next cycle.

Cyclical sedimentation is the principal theme of both this and the next chapter. The present chapter concentrates on establishing the internal nature and development of the single sedimentation cycle. Chapter 3 is concerned with the interrela- tionships between successive cycles and with the many manifest 38

variations on the cyclical theme; leading eventually to a comprehensive reconstruction of the paleoenVironmental and paleogeographical evolution of the study area for the whole of Baumann Fiord time.

What follows in this chapter, then, is a close examination of the single cycle. The carbonate component of the cycle may be partitioned into a sequence of distinctive lithofacies and these are discussed first in ascending order from the base to the top. The overlying anhydrite unit is then discussed as a single entity. Each lithofacies of the cycle is treated in a two—fold manner: a detailed description of the megascopic and microscopic character of the facies is given first, and this is followed by an interpretive analysis of its origin and its environment of deposition.

It has been postulated (Mossop, 1972b) that the Baumann Fiord cycle is directly analogous to the so—called 'sabkha cycle' of sedimentation, in the sense that each of the lithofacies of the Baumann Fiord cycle is explicable in terms of the sabkha model of cycle generation. In subsequent text, this argument is pursued and expanded by way of comparative analyses. The Recent environmental model to which the Baumann Fiord cycle is compared is that of the 'type area' of sabkha sedimentation, the Trucial Coast of the Persian Gulf.

The Trucial Coast sabkha has been studied by numerous workers over the past decade. Its nature and origin are well known, but for review purposes, the essential features are briefly discussed in the first part of this chapter. Reference is also made to some ancient sequences that have been interpreted as being of sabkha origin. The remainder of the chapter, the description and interpretation of the Baumann Fiord cycle, continually refers back to the environmental model outlined in the first part. 39

2.2 The s212-.-E..._ffElInE

The Recent Sabkha A review of the present—day sabkha environment is appropriate here for two reasons: first, to establish the character of the vertical seouence of lithofacies that developsin the sabkha setting, i.e. a reference sequence to which the Baumann Fiord cycle may be subsequently compared; and second, to outline the known environmental parameters that control the genesis of the sequence, i.e. a reference model upon which the environmental reconstruction of the Baumann Fiord cycle may be based. This review encompasses only the most basic aspects of the Recent sabkha setting. For a more comprehensive analysis, the reader is referred to a number of detailed papers on the subject, the most important of which are listed below by way of historical summary: 1963 Curtis, Evans, Kinsman and Shearman (1963) — the first reported occurence of Recent anhydrite. Shearman (1963) — a brief account of the sabkha environment. 1966 Shearman (1966), Kinsman (1966), Evans (1966), Evans et al. (1969), to Kinsman (1969), Kendall and Skipwith (1969a, 1969b) and date Butler (1970) — general accounts of the sabkha setting and views of its possible significance in interpreting ancient evaporite sequences.

A number of theses on the Trucial Coast setting have been completed. These include: Kinsman (1964) — primarily on the carbonate sediments of the area adjacent to the sabkha flat. Butler (1966) — on aspects of the sabkha setting, both the anhydrite facies and the associated carbonate facies. Skipwith (1966) and Kendall (1966) — mainly on the carbonate sediments adjacent to the sabkha. Other papers on the genesis and diagenesis of the carbonate sediments of the sabha and with particular reference to the algal mats include: Tiling et al. (1965), Kendall and Skipwith (1963) and Park (1973). 40

Brine geochemistry is referred to in a number of the works above but is discussed in detail by: Butler (1969) and Bush (1970, 1973)

For further information on the Trucial Coast, and for a comprehensive bibliography of papers pertaining to its carbonate and evaporite sediments, the reader is referred to Purser (1973).

The Trucial Coast of the Persian Gulf is an area of dominantly carbonate sediMentation. In the offshore region there is a complex of shoals, made up primarily of skeletal sands and oolites (see insert map Figure 12). Emergent parts of the shoals form a series of islands which parallel the coast. Between the islands and the mainland is a complex of shallow lagoons in which aragonite muds and pelleted muds are being generated. The broad coastal flat immediately adjacent to the lagoons is thesdpkha, proper. It extends for some 200 miles along the coast and in places is as much as 20 miles wide. The hinterland behind the sabkha is an area of arid continental sedimentation, bordered in the south and east by the Oman Mountains.

'Sabkha' is an Arabic word meaning 'flat salt—crusted desert'. The sabkha along the Trucial Coast (the stippled area,Figure 12) is a 'coastal sahkha' and its groundwaters are largely derived from the sea. 'Continental sabkhasl, such as those that are locally developed in the desert region inland, have groundwaters that are largely of terrestrial origin. It is with the coastal sabkha thPA this review is concerned. For simplicity, the term 'sabkha' is used throughout this text to mean 'coastal sabkha', i.e. a marine—derived sabkha.

The schematic cross—section shown in Figure 12 serves to illustrate most of the important features of the sabkha environment. The section transects only the zone of immediate interest in this review. Off the section to the left lie the offshore islands and shoals. Off the section to the right lie the dune— fields of the hinterland desert.

Pits dug in the sabkha surface consistantly reveal the

SABKHA km. SHOAL AREAS 0 20 affrazs=1 - A ALINE OF STUDY

Ras Gha nada

km. 0 200 L Abu Dhabi island. f % .•"" " --$ •N`.-- - 0 s. , 0 ...... "--'-d-, 4 - , e'''' f .. - - ; , i.,.ri .... 0 /.. / ...--,....43:C , ., ,_._.--•-•S--;) Jabal `, -° " .. X, ---0,i-s- , ,. ich or - , ... - • _. , -. _ - -...... a/ &az•-• Dha n na.-- --,..-

Trucial Coast of the Persian Gulf. Map courtesy of Dr. P. R. Bush PROGRADATION OF 3 MILES IN --3000 YEARS

SUBTIDAL INTERTIDAL SABKHA SURFACE LAGOON ZONE ALGAL MAT 3ft. MSL

Subtidal Carbonate

Nodular Anhydrite + Carbonate Gypsum /Anhydrite '==-0.- Algal Mat -0:4.0 Mat+Gypsurn •• : +Gypsum . Subtidal Subtidal ...... Carbonate Carbonate . . . .

Figure 12. Schematic cross—section of the Trucial Coast sabkha complex illustrating the form and distribution of early tevaporite minerals in the supratidal sabkha facies. Line of section equivalent to line A—A. in insert map above. 43

following vertical succession: a basal marine facies, the subtidal carbonate muds of the lagoon; a thin intertidal algal mat facies (six to eight inches); and a three—foot supratidal facies, the sabkha facies itself. The sequence is clearly the record of a regressive cycle of sedimentation, each facies being diachronous in its development. Buried algal mats some three miles in from the present shoreline have been dated by carbon-14 methods and are known to have been laid down 3,000 years B.F. (Evans et al., 1969). The system as a whole is thus undergoing seaward progradation at a rate of about one mile every thousand years.

The indigenous sediments of the supratidal facies consist predominantly of carbonates, part wind—blown carbonate sand, derived principally from the emergent offshore shoals, and part aragonite mud derived from the lagoons. The latter is washed over the sabkha plain in the course of storms which periodically inundate the flats with marineflood waters. All the other mineral species present in the supratidal facies are of early diagenetic origin. These include replacive dolomite and magnesite, and displacive gypsum and anhydrite. The mode of emplacement of these early diagenetic minerals is best considered by examining the sabkha's groundwater system. Throughout the sabkha, the stable groundwater level is about coincident with the old buried algal mat, some three feet below the sabkha surface (Figure 12). The groundwaters, which are being continually concentrated by capillarity and evaporation in this arid setting, are replenished in two ways: first, by gradual infra—sediment flaw, fluxing inland from the lagoon; and second by downward seepage of the floodwaters that occasionally cover the sabkha surface. The former appears to be continually operative (Bush, 1973), the marine—derived waters becoming increasingly concentrated as they progress further and further inland. This progressive concentration of the sea—water gives rise to a series of diagenetic minerals, some of which are direct precipitates end others the products of reactions between the brine and the earlier deposited sediment. 4.i

-The first diagentic mineral to form is gypsum. This grows within the algal mats in the upper part of the present intertidal zone (Shearman, 1966). The crystals are of the lenticular habit in which the c—crystal axis is roughly perpendicular to the. plane of flattening (see Masson, 1955). Intertidal zone gypsum crystals of this type are commonly about one centimeter in diameter. Accessory celestite (SrS0 ) is also found in the intertidal algal mat zone (Evans and Shearman, 1964).

tornticular gypsum crystals also occur in the supratidal sediment immediately inland froth the high—tide mark. These grow mainly by displacement but they often incorporate some carbonate sand or mud. 'Gypsum sand' crystals of this type often reach 20 cms. in diameter and are locally very abundant (Figure 12). The fluxing brines that give rise to the gypsum crystals are concentrated relative to sea—water by a factor of five or more. This prodigious concentration, whichcan only result through evaporative loss of 80 percent or more of the original sea—water, is a testiment to the intenselywaporitic conditions that characterize the sabkha setting. By the time the sea—water has been processed through the initial lagoonal, intertidal and near—shore supratidal environment, it is not only sufficiently concentrated to_precipitate gypsum, but is also markedly modified in its proportional concentration of dissolved substances. In particular, it is depleted in calcium, having precipitated two major calcium compounds in these early stages; viz. aragonite from the lagoonal waters, and aragonite and gypsum (CaCO3) 0) from the groundwaters of the intertidal and near— (CaS04 .2H2 shore supratidal sediments. With the consequent increase in the Kg/Ca ratio of the brine, a considerable measure of dolomitization is promoted in theMpratidal aragonitic and calcitic sediments (Figure 12), and magnesite (KgCO3) is also present in the region proximal to high—water mark (Bush, 1973). In a general way, then, environmental conditions in the intertidal and near—shore supratidal zones arc conducive to displacive development of gypsum with contemporaneous replacive development of dolomite and magnesite. 45

Anhydrite melees its first appearance in the supratidal sediments about -7' mile inland from the normal high water mark. It takes the form of either discrete nodules or bands of coalesced nodules. Groath of the nodules is by displacement of host carbonate, the calcium and sulphate being supplied by capillary draw from the underlying groundwater brines. In some parts of the sabkha, anhydrite constitutes more than 50 percent of the supratidal facies. Though many of the nodules appear to have originated as discrete entities in themselves, others are clearly derived from the alteration of pre—existing gypsum crystals (Holliday, 1968, 1971, 1973; Butler, 1970; Bush, 1973). This process results in the progressive loss of many of the earlier—formed 'algal mat' and 'gypsum mush' crystals (Figure 12). In a general way, the predominance of anhydrite in the sabkha facies increases progressively towards the hinterland and in parts of the sabkha far removed from the shoreline, it may be the sole evaporitic mineral present.

Detailed analysis of certain aspects of the Recent sabkha setting is contained in subsequent discussion, in a context more immediately related to aspects of the Baumann Fiord cycle. But the foregoing discussion of sabkha genesis and diagenesis, .intended as it is as an introductory statement, serves the present purpose ir that it establishes two important points. Firstly, the vertical sequence resulting from a complete sabkha cycle of sedimentation is characteristically comprised of a shallowing—upwards subtidal marine carbonate facies that grades up into an intertidal algal mat facies. This in turn is overlain by a supratidal facies in which carbonate and anhydrite are foremost constituents. Secondly, the cycle is generated in response to both depositional and diagenetic processes. The depositional phase, manifest by progradational advance of the algal mats and the supratidal carbonates, is followed by a series of early diagenetic fronts, which lag behind the depositional front but which eventually give rise to a widespread and essentially uniform carbonate—anhydrite assemblage in the supra— tidal facie. 46

Ancient Sabkha Cycles

Since the Pleistocene low stand of sea level, there has been sufficient time to develop only a single sabkha cycle on the Trucial Coast. The system is at present in a relatively stable state, in that the delicate balance between sedimentary build–up and regional subsidence is being maintained. But the system cannot continue to prograde indefinitely. There come a point at which the sedimentary processes that-are operating are simply not sufficient to equilibrate with the continuous subsidence. It seems reasonable to argue that at that critical point, a major marine incursion should ensue. Confirmation of the validity of this argument stems from knowledge of ancient sabkha sequences and it is on these sequences that attention is focused below.

In the decade since the Trucial Coast sabkha evaporite setting was first discovered, there have been a number of studies on ancient evaporite sequences in which workers have been able to show that sabkha sedimentation was responsible for their genesis. Shearman (1966) made a direct comparison between the Recent Trucial Coast sabkha sequence and that developed in the Purbeck evaporites (Upper Jurassic) of southern England. The latter is exposed in Dorset and Sussex (Howitt, 1964; West, 1964, 1965) and has been cored at a depth of about 2,100 feet in the Institute of Geological Sciences borehole at Warlingham in south London. The borehole section consists of a total of four distinct sabkha cycles comprising 17 feet of section. Sabkha cycles have also been reported in the Tournasian (Carbon- iferous) Hathern Series of Leicestershire (41-eige+itm-N–i4o4mIs**4. R444–sta494944491-1-94g1 and Llewellyn and Stabbins, 1968, 1970) and in the Visean (Carboniferous) Series in Ireland (West, Brandon and Smith, 1968). In Canada, a number of ancient evaporite sequences have been interpreted as being of sabkha origin. These include: Mississippian Madison Limestone evaporites of Saskatchewan, containing a total of seven cycles (see Fuller and Porter, 1969a; as well as Illing, Wood and Fuller, 1967 as based on Fuller, 1956); Upper Devonian evaporites in the 47

Stettler. Formation of Alberta and Saskatchewan, including 13 sabkha cycles in 50 feet of Crossfield Member strata (Fuller and Porter, 1969a); Upper Devonian Souris River Formation sabkha cycles in Saskatchewan (MacQueen and Price, 1969); and Middle Carboniferous (Windsorian) cycles in Nova Scotia. Other known instances of ancient sabkha facies development include: Middle Carboniferous supratidal anhydrite sequences in Spitsbergen (Holliday, 1965, 1968); carbonate—anhydrite cycles in the Arab/ Darb Formation of the Persian Gulf area (Uood and Wolfe, 1969); sabkha anhydrite in Lower Eocene rocks in Jamaica (Holliday, 1971); sabkha cycles in the Gachsaran Formation of Iran (Gill and Ala, 1972); supratidal sabkha evaporites in the shelf areas surrounding the Permian Basin (Guadalupian) of the southwest United States (Jacka, 1973); and cyclical carbonate—anhydrite development in Permian Bellerophon Formation of northern Italy (Bosellini and Hardie, 1973).

These ancient sequences exhibit all the essential characteristics of sabkha—type deposition, though all of them have aspects that are unique, eg. their component cycles may be extraordinarily thick, or their supratidal facies may be lacking in dolomite. There are as many variations on the basic sabkha- cycle theme as there are ancient examples. But without exception, these fossil sabkha cycles are similar in their essential aspect: a shallowing—upwards marine facies, commonly, but not universally, dominated by carbonates; grading upwards into an intertidal facies, normally characterized by algal stromatolites; overlain by a supratidal facies in which anhydrite is a principal component. Some cycles are further overlain by a continental redbed facies. The top of each of these regressive cycles is invariably marked by an erosion surface upon which rests the sub— tidal component of the next cycle. Thus it is apparent that the gradual regressive phase of sabkha—type sedimentation is intrinsically subject to termination by rapid marine transgression, the erosion surface at the top commonly being the only recognizable physical manifestation of its passing.

Throughout the analysis of the Baumann Fiord cycle that 48

follows, it is argued that these essential features of the sabkha model of cyclical evaporite genesis are consistent with the observed features of the Ordovician sequence.

2.3 The Baumann Fiord Cycle

Figure 13 schematically depicts the vertical succession exhibited in a single Baumann Fiord cycle. As a general rule, these cycles average 12 feet in thickness but there is considerable regional variation (± four feet). Taken in its essentials, the cycle is made up of a basal limestone facies, dominated by lime mudstone, an intermediate stromatolite facies, and an upper anhydrite facies. On closer inspection, these facies can be sub—divided into distinct components, each of which is the subject of detailed analysis below. It is argued that each of the cycle's components is justifiably explicable in terms of the sabkha model.

Erosion Surface

Description

In the field, the erosion surface at the base of the cycle is characteristically sharp, the contact between the top of the underlying anhydrite and the base of the carbonate sometimes being noticeably undulose but commonly being essentially concor- dant. The surface as such is in places marked by a thin film of dolomite residuum.

Interpretation

It is thought that the erosion surface marks the passing of a marine transgression that swept over the pre—existing supra— tidal flat, eroding and in part dissolving the uppermost reaches of the supratidal facies. That such a transgression would be accompanied by erosive action is demonstrated by the way in which storm—flooding of the Trucial Coast sabkha, a process that in many ways simulates that of rapid transgression; brings about 49

Erosion surface A A A A A A A A A A A A A A A Anhydrite A A A A A A A A A A A A A

(dolomitic ~12 Feet -occasional mud cracks)

Stromatolite (algal mat)

(occasional symmetrical ripples)

Laminated Lime Mudstone (occasional ripples)

Massive Lime Mudstone

Ripple Cross- laminated Lime Mudstone

Limestone Flat-Pebble Conglomerate Erosion surface

Fire 13. Schematic representation of the make—up of a typical Baumann Fi3rd rar'.7.onatanhydrite cycle. 50

-Erosion Surface

Sabkha Facies

Intertidal Facies

Lagoonal Facies

Figure 14. Core-slab photograph of sabkha cycle in the Purbeckian of southern England (Warlingham borehole). Top of slab delimits the contour of the upper erosion surface; i.e. the base of the superseding cycle. 51

erosion. Tn the wake of the Trucial Coast floods, the uppermost carbonate sediments of the supratidal flat are stripped away and much of the anhydrite that was previously some three or four inches below the surface is laid barer In places, anhydrite nodules are physically eroded and in others they are partially dissolved by the relatively dilute sea—water (Bush, 1973). But as stated previously, the Trucial Coast sequence does not in itself illustrate the ramifications of marine transgression, there being only a single cycle developed there to date. It is necessary to turn to ancient examples in order to find an analogue. In the Purbeck sabkha cycles at Warlingham (Shearman, 1966) for example, the supratidal carbonates and evaporites are clearly truncated at the top by sharp erosion surfaces which are immediately superseded by the subtidal component of the next cycle (Figure 14). A similar relationship is observable in the Arab/barb sabkha cycles. Thus it appears that erosion of the top of the one cycle is a common, if not essential, by—product of the marine transgression that marks the onset of the new sabkha cycle and in this regard, the Baumann Fiord cycle appears typical.

Flat—Pebble Conglomerate

Description

Immediately above the erosion surface there is a limestone flat—pebble conglomerate, normally two to three inches thick but occasionally thinner (Figure 13). The conglomerate consists of well rounded flat pebble7u7 to two inches across and with diameter/thickness ratios of about 4/1. The plane of flattening of the pebbles is usually parallel to bedding (Figure 15). The pebbles themselves are characteristically composed of essentially homogenous micrite, minor amounts of quartz silt (< 5 percent) being the only common impurity. ilore than 90 percent of the pebbles are of this type. The remaining pebbles exhibit a great variety of internal lithologies, with constituents comprising skeletal fragments, pellets, intraclasts, lumps and ooliths, all set in a micritic matrix. In thin—section, these exotic pebbles stand out in striking contrast to the normal micrite pebbles (Figure 16), The matrix in which pebbles reside consists dominantly V

Figure 15. Limestone flat—pebTAe conglomerate from the base of a Baumann Fiord cycle. Plan view onto bedding surface. 53

Figure 16a. Flat—pebble conglomerate composed of micrite clasts set in a matrix of lime mudstone. Matrix is extensively dolomitized in this example (light—toned crystals). Monochrome photomicrograph. Crossed—nicols. Field 1.8 cm. across.

Figure 16b. Flat—pebble conglomerate, here with an extraordinary preponderance of 'exotic' pebbles. Note the abundant allochems in the pebbles and the matrix. Pervasive sparry cement is detectable in the groundmaos. Monochrome photomicrograph. Crossed—nicols. Field 2.3 cm. across. 54

of lime mudstone, occasionally structureless but often having a pelleted appearance. Minor amounts of skeletal detritus, with accompanying intraclasts and other allochems, locally constitute a small fraction of the matrix. Dolomite is also a common accessory constituent of the matrix (up to 35 percent). Pervasive crystalline calcite is a characteristic matrix cement (Figure 16b).

Interpretation

Flat—pebble conglomerates of this type have been described in the Lower Ordovician Tribes Hill Formation of New York (Braun and Friedman, 1969). There, it is argued that the conglomerate originated as a result of reworking of desiccated mud flat sediments. Former mud polygons were ripped up in the course of a storm—induced flood, and the resulting mud ships transported and rounded, and eventually laid down again as a lag deposit. Braun and Friedman (1969) report that the Tribes Hill conglomerate can sometimes be traced laterally into an in situ lime mudstone in which desiccation cracks are a prominant feature.

It seems reasonable to suggest that the Baumann Fiord flat— pebble conglomerate originated in a manner similar to that outlined by Braun and Friedman (1969). Furthermore, it would seem that the deposition of a conglomerate of this type would be a natural by- product of marine transgression over a pre—existing sabkha flat; i.e. the flat—pebble conglomerate seems a logical complement to the erosion surface if the two are considered in conjunction with one another. What evidence is there to corroborate these suggestions?

Firstly, there are ancient sabkha analogues. For example, Wood and Wolfe (1969) report that flat—pebble conglomerate is a characteristic feature of the zone immediately above the erosion surface in the Arab/Darb sabkha cycles and they interpret this as a transgressive lag deposit. Similarly, some of the Purbeck sabkha cycles are characterized by carbonate conglomerate at the base and it is suggested (Shearman, personal communication) that these are the result of reworking of former supratidal carbonate sediments (see Figure 14). Secondly, there is a recent analogue. In this regard, it is interesting to note that the Trucial Coast sabkha does not in 55

itself seem to afford the potential for flat—pebble conglomerate generation pursuant upon a future marine transgression. This is a function of the fact that, though there is a twashover' mud component in the supratidal carbonates, the bulk of the sabkha carbonate sediment is wind—blown calcarenite and this sand—size material does not lend itself to desiccation cracking and/or cohesive pebble formation. The supratidal carbonates of the Baumann Fiord sequence, on the other hand, apparently consisted almost exclusive- ly of 'washover' lime muds and these would seem to be intrinsically susceptible to desiccation and subsequent reworking (this point is developed further in subsequent discussion). But there is an interesting occurence of desiccated lime mud in a supratidal flat setting on Sugar Loaf Key, Florida (Shinn, 1968). There, 'washover' lime mud has been subjected to two years of continuous desiccation and the mud polygons are undergoing gradual disaggredation to form flattened, pebble—size clasts (see illustration p. 615; Shinn, 1968). The matrix in which the mud pebbles reside is selectively dolomitized (up to 25 percent dolomite). The present writer would suggest that a high—energy reworking of these sediments, either by marine transgression or spring—tide storm flooding, would produce a flat—pebble conglomerate lag deposit essentially identical to that developed in the Baumann Fiord cycle.

Lastly, there is evidence from the Baumann Fiord conglomerate itself, the pebbles have clearly undergone some measure of transport, as evidenced by their well—rounded aspect. Those pebbles that consist of homogeneous micrite may well have been derived from reworking of dessicated 'washover' lime mud on the supra— tidal flat proper; i.e. reworked essentially in situ. However, it is apparent that at least some of the pebbles, notably those with exotic internal lithologies, were derived from some distance away, their original environment of deposition being more readily attributable to open marine conditions, i.e. toward the seaward edge of the sabkha complex, either in or immediately adjacent to the island shoals. Part of the pebble matrix, which includes

The internal lithology of the exotic pebbles is characteristic of near—shore, shallow marine deposition but in situ deposits of this facies are not expobed within the study area. Discussion of the possible location and character of the shoals is contained in Chapter 3. 56

skeletal detritus and other allochems, may stem from the same environment as that of the exotic pebbles. Taken as a whole, however, it is clear that the character of the flat-pebble conglomerate is consistent with the thesis that it originated by reworking of former supratidal lime mud and that the transgression that brought about erosion of the top of the previous cycle was also responsible for the deposition of the flat-pebble lag. As such, it is considered that the sabkha model of cycle genesis is consistent with these observed features of the Baumann Fiord cycle.

Lime Nudstone

Description

Above the flat-pebble conglomerate, there is a sequence of lime mud.stones that, although essentially uniform in composition, show a considerable range of sedimentary structures (Figure 13). This lime mudstone facies is characterized by a complete lack of recognizable fossils and the only accessory constituents are small amounts of quartz silt and clay (usually 4:5 percent) and disseminated organic remains. At the base, the lime mudstone is normally marked by asymmetrical ripple cross- lamination. The contour of the ripple marks is often highlighted by thin partings of silt or clay-rich lime mudstone (Figure 17). Above the rippled lime mudstone, there is a zone that is essentially structureless, composed of very pure micrite but containing as well some wisps of organic matter (Figure 18). The overlying laminated material is again composed of relatively pure micrite but some laminae are slightly enriched in quartz silt while others are enriched in organic matter (Figure 19). Near the top of the lime mudstone facies, immediately below the stromatolitic zone, there are occasional symmetrical ripple marks (Figure 13).

Interpretation

The lime mudstone facies is here interpreted as having been deposited in a shallow lagoonal environment comparable with the present-day Trucial Coast lagoon complex. Current knowledge of

..--..-- •,,, ,_._„,„,. _ _...,- .__ c ------"_,,e-,4 - ..>, -,---i- .....,,....._— - _ . _ 1.-::-.'- ,„1"._. • 44-', ----)- --1.- 1,-...-- ---".----"',. i'l - __, .,.... ‘,..r.,,,-_, • ,...... 7004 _f - '-:._ ------97-e--' - --,% ' - — ' - ' -■- ' '.4771!:464-Z1177 ' ' : ( , r 4 27i

Figure 17. Asymmetrical cross-lamination in lime mudstone facies, with the contour of some of the laminae highlighted by partings of silty shale. 58

Figure 18. Massive lime mudstone with scattered silt grains and wisps of dark organic matter. Monochrome photomicrograph. Ordinary light. Field .8 cm. top -to bottom. Figure 19a. Laminated lime mudstone from the carbonate component of a Baumann Fiord cycle.

Figure 19h. Laminated lime mudstone. Lamination here defined by contrasting proportions of Quarts silt (light—to7led grains) in successive laminr. Monochrome photomicrograph. Ordinary light. Field .4 cm. across. 6o

the sedimentary processes that operate in the Trucial Coast lagoons stems largely from a detailed sedimentological, minera- logical and chemical analysis carried out by Kinsman (1964). In essence, Kinsman was able to establish that the lagoons are areas of cuiet water sedimentation, characterized by active generation of carbonate mud in the form of aragonite needles. Using various lines of evidence, he was further able to show that °80-90 percent of the mid and inner lagoon sediments must have been chemically precipitated° (Kinsman, 1964; p. 178); i.e. there is a relatively small contribution of aragonite from organic sources. He reports lagoonal salinites ranging between 35o dissolved substances ( == normal sea—water) in the seaward portions of the lagoons up to 65%o in some of the inner parts of the lagoons. It is suggested that it is this evaporative concentration of the lagoonal waters that may make possible the direct precipitation of CaCO as aragonite. 3 The major difference between the present—day lagoonal lime muds and the Baumann Fiord lime mudstones is the characteristic lack of pellets in the latter. Kinsman (1964) attributed the ubiquitous pellets of the Trucial Coast lagoons to a fecal source and there is no doubt that the abundant infauna and epifauna of the lagoonal area is capable of bringing about the observed degree of pelleting. It is not known exactly what faunal group is responsible for the lagoon pellets but Kinsman (1964) suggests that the Annelid worm may be the principal contributor, as it is in the BaharAjn lagoons (Cloud, 1962). There appears to be at least one possible explanation for the lack of pellets in the Baumann Fiord lime mudstones and that is that in Early Ordovician time, the deposit feeding faunas were not evolved to the point that they were able to tolerate high—salinity conditions. If, as postulated, the Baumann Fiord lime mudstones were deposited as a direct precipitate from a very shallow and areally extensive lagoon, then it seems reasonable to suggest that evaporative concentration of lagoonal waters may well have been prodigious. In a relatively shallow embayment in the Persian Gulf (the Gulf of Salwa to the west of Qatar Peninsula), the salinity increases by a fator of 1.5 from its seaward tc land:Tard extremities 61

(Sugden, 1963). In an even shallower setting, the salinity might easily reach three or four times that of sea—water. This order of concentration is manifest even in the Trucial Coast lagoons where Kinsman (1964) reports salinities 111 excess of 100%o in restricted areas. Certainly there were pelleting organisms in the Early Ordovician (Annelids, for instance, are known in the Cambrian (Moore (ed.), 1962)) but these may have been restricted to the seaward extremity of the sabkha complex; i.e. the area from which the exotic pellet—carrying pebbles of the basal conglomerate may have been derived.

But apart from the lack of pellets, and by the same token the lack of fossils, the Baumann Fiord lime mudstones appear to be directly comparable to the carbonate muds of the Trucial Coast lagoons, this in the sense that the observed features of the Baumann Fiord lime mudstones are consistent with those that characterize the subtidal component of the model sabkha cycle. The basal lime mudstones, the first to be deposited after the postulated marine transgression, in that they carry asymmetrical ripple—marks, appear to be diagnostic of deposition in the seaward portion of a lagoon complex where current action may be expected to have had an influence on the sedimentary structures. Mud—size sediment does not, as a rule, lend itself to current rippling and in this regard the Baumann Fiord ripples are somewhat anomolous. It is possible, however, that the presence of small amounts of quartz silt altered the physical character of the lime mud to an extent that allowed current rippling to take place. Certainly the most prominent ripples are those in which the concentration of silt is highest (to 20 percent). The environmental significance of the massive lime mudstone above the rippled zone is more difficult to assess. It is possible, of course, that the direct precipitation of lime mud in a very shallow and quiet setting tends to produce a structureless deposit. Some of the lime mud of the Trucial Coast lagoons is massive and homogeneous but it is not known whether this is due precipitate (i.e. the result of to rapid settling of CaCO3 deposition from 'whitings') or due to extensive bioturbation of a formerl y laminated mud (Kinsman, 1964). For reasons outlined 62

above, it is unlikely that in the Baumann Fiord setting there was a faunal community capable of extensive bioturbation. The presence of considerable amounts of organic matter in the massive mudstone suggests that organic production or entrapment of CaCO3 may have played a role in its genesis but there is no conclusive evidence in support of this. With no documented present—day analogue upon which to base interpretive discussion, the origins of the massive lime mudstone component of the Baumann Fiord sequence must remain conjectural. Laminated lime mudstone, on the other hand, would appear to be a natural product of the shallow lagoonal setting envisaged for the 7_ Baumann Fiord lime mud facies. Each lamina is distinct from the next because of minor differences in their respective accessory compositions (one may have a slightly higher concentration of quartz silt than the other, for instance) and this distinction is clearly a result of minor temporal changes in salinity, temperature or other physico—chemical conditions. This type of temporal fluctuation in environmental conditions is surely what would be expected in a shallow lagoonal setting and a delicately laminar deposit should be the inherent result. In the uppermost reaches of the lime mudstone facies, symmetrical ripples are occasionally developed and these are interpreted as resulting from oscillatory water movement in the swash zone of the shallowest foreshore region.

Taken as a whole/ the lime mudstone facies of the Baumann Fiord cycle is seen as a shallowing—upwards lagoonal deposit, now composed mainly of micrite but probably originally laid down as aragonite mud, much of which may have been a direct chemical precipitate. The deposit bears many essential similarities to the Trucial Coast lagoonal sediments and direct comparison of the two does not reveal any irreconcilable points of diver- gence. FJore importantly, all the intrinsic characteristics of the Baumann Fiord lime mudstone facies are explicable in terms of shallow subtidal genesis and this, of course, is the single requisite for its interpretive identity with the general sabkha model. 63

Algal Stromatolites

As a rule, the stromatolitic component of the Baumann Fiord cycle is on the order of 8-12 inches in thickness and it invariably occupies an intermediate position between the lime mudstone facies and the anhydrite facies (Figure 13). The algal stromatolites themselves exhibit a number of distinctive forms (Mossop, 1973) and these are dealt with in detail below, But all the forms have one aspect in common: they are made up almost entirely of micrite, with trace amounts of organic matter defining a delicately laminar internal structure.

In describing the different forms, the author has chosen to employ the terminology defined by Aitken (1967), primarily because it accords most closely with historically established usage. The extremely precise and adaptable designations of the classification scheme set up by Logan, Rezak and Ginsburg (1964)* are parenthetically included as well.

Descriptions

There are three distinctive groups into which the Baumann Fiord stromatolites may be categorized, as based upon their external form: branching digitate stromatolites, polygonal stromatolites and domal stromatolites.

The digitate forms are the most common and these exhibit the following characteristics (Figure 20): (i) Individual 'fingers', which normally measure about one inch across, are comprised of stacked hemispheroids (SH—V), with relief to * in. or less. (ii) The hemispheroidal laminae are usually very closely spaced ( < 1 mm. apart). (iii) As a rule, the 'fingers' show a tendancy to branch upwards; i.e. single 'fingers' repeatedly bifurcate upwards. (iv) The matrix in which the 'fingers' are imbedded is normally composed of dense micrite. * Based on three basic geometric configurations: Vertically Stackr.d Hemispheroids (SM), Laterally Linked Hemispheroids (LLH) and Spheroidal Structures (SS). 64

(v) Viewed in plan, it is not possible to discern any alignent or pattern amongst the 'fingers'.

Polygonal Stromatolites (Figrxe.21):

(i) The polygons measure about 6 inches across and, though essentially linear along their edges they are often slightly rounded on the corners. (ii) In cross-section, the polygons, which average 2 inches in thickness, can be seen to consist of closely spacea algal laminae that terminate abruptly along the often slightly upturned edges of the polygon (SH-I; Kendal and Skipwith, 1969). (iii) Interstices between polygons are commonly filled with dense micrite.

Domal Stromatolites (Figure 22):

(i) Individual domes, up to 12 inches high, consist of stacked hemisperoids (SH-V) with low internal relief ( ,.,-, one inch), the lamination being delicate and closely spaced, (1 mm.). (ii) Discrete domes, usually about four inches across at their base, often increase in radius upwards and may coalesce with adjacent domes (SH-V-.LLH-C). (iii) In plan, the domes are circular to slightly eliptical and are often aligned in rows.

These three basic forms make up virtually all of the stromatolites of the Baumann Fiord Formation (Digitate 655, Polygonal 255 and Domal 105). In any one place, one form tends to dominate, but others are usually in evidence as well. The most characteristic associations are those in which digitate forms dominate in the basal two-thirds of the stromatolite horizon, giving v!ay to domal or polygonal forms in the upper third. In a small number of cases, no distinctive stromatolite horizon occurs in the cycle but in these, the carbonates immediately below the anhydrite show an 'algal--lamination' and appear to accord with the designation: 'cryptalgalaminite' (Aitken, 1967). 65

Figure 20. Digitate stromatolites viewed in cross-section. The matrix compriset light grey dense micrite. Figure 21. Polygonal stromatolites, viewed in plan. Mottled appearance is due to lichen cover. Field 3.5 feet across. Figure 22. Domal stromatolites, viewed in plan. Erosion of the tops of the domed algal heads has produced the raggedly concentric profiles of the domes/ internal lamination. ad. ^1,

- Oir es

• 4

Figure 23. Coalesced algal heads with knobby external surfaces, Locality 2. 69

This type of cryptalgalaminite is also found in assocition with the stromatolites (usually 'immediately below the stromatolite horizon) and may represent a transition between 'ron—algal' and true 'stromatolite' laminar development.

One further type of cryptalgal structure, observed only twice but perhaps worth noting because of its rather spectacular aspect, is that of huge coalesced algal heads (Figure 23) similar to some 'thrombolites,' described by Aitken (1967). These assume the aspect of small bioherms. Their original internal structure, now largely obscured by recrystallization, is thought to have been of the 'clotted' or 'spongy' type typical of thrombolites.

Interpretation

The study of ancient and recent cryptalgal structures has, in the last 40 years, produced a wealth of knowledge on the nature and mode of growth of these prolific and ubiquitous rock formers and sediment trappers. This mass of data and conjecture is admirably condensed and summarized in a review by Bathurst (1971; Chapter 5).

The most obvious recent analogue to the Baumann Fiord stromatolites, at least from the point of view of evaporite- association, is that of the Trucial Coast. There, the algal mats are restricted to the intertidal zone (Kendall and Skipwith, 1968). Because the slope of the shore—edge of the sabkha is so gentle, the intertidal zone is up to one mile wide in places. Along much of the coast, this intertidal area is almost completely covered by spreads of algal mat (Figure 24). Four mat zones were recognized by Kendall and Skipwith (1968). From the lagoon edge, these are: 1. the Cinder zone, made of rubbery, black algal mats resembling volcanic cinder deposits; 2. the Polygonal zone, comprised of regular polygons separated by sediment—filled cracks; 3. the Crinkle zone, with a 'blistered', 'leathery' algal skin; and 4. the Flat zone, in which a very thin (3 mm.) smooth algal skin is developed. The polygonal stromatolites of the Baumann Fiord cycle bear a direct resemblance to the 70

Figure 24. Band of Recent algal mats occupying the intertidal zone between the shallow lagoon (right) and the supratidal sabkha surface (left), as viewed from the air. Trucial Coast of the Persian Gulf. The algal mat spread is approximately 1' mile wide. Photograph courtesy of Dr. D.J. Shearman. 71

Trucial Coast polygonal zone algal mats and there seems little douht that the former were laid down in a comparably high intertidal zone setting; i.e. one subject to alternate wetting and desiccation hut with desiccation predominating. Similar polygonal algal mats have been described in the high intertidal zone in other parts of the Persian Gulf (Illing et al., 1965), the Bahamas (Black, 1933), and Florida (Ginsburg et al., 1954). Although algal stromatolites resemhling.those developed in the other three Trucial Coast zones were not recognized in the Baumann Fiord cycles, some of the internal lamination of these present—day algal mats is similar to the cryptalgalaminite described in the ancient sequence.

Recent domal stromatolites have been reported in periodically exposed tidal flat settings in the Bahamas (Black, 1933), Florida (Ginsburg et al., 1954), and Shark Bay, Western Australia (Logan, 1961; Logan et al., 1965; Davies, 1970; Hoffman, 1971). Many of these present—day forms are strikingly similar to the Baumann Fiord domal stromatolites and it follows that the ancient examples may have formed under conditions similar to those of the Recent.

The origin of digital forms is somewhat more problematical because recent analogues are scarce. Hoffman (pers. comm.) reports digitate forms in a very shallow subtidal setting at Hamelin Pool, Sh,=1: Bay. It is not known, however, whether these are actually growing in this quiet subtidal setting or whether they are relict from a previous period in which mean sea level was slightly lower. In this regard, Logan (1961) noted that when intertidal zone algal mat was subjected to prolonged wetting, in semipermanent tidal pools for example, active algal mat growth was terminated. Logan et al. (1964) deduced that some digitate forms must develop in the low inter—

Growing subtidal algal mats are known in the Bahamas (Neumann et al., 1970; Scoffin (1970) and Bathurst (1971) pp. 122-126) and Bermuda (Gebelein, 1969), but these do not exhibit stroma- tolic forms. Rather, they are characterized by indistinct cryptalgalanination of a type that is often completely obscured upon buria]. 72

tidal zone. Aitken (1967) also assigned digitate strornatolites to the low intertidal zone. The Baumann Fiord digitate stromatolites are thus considered to have originated in the low intertidal zone but at least come of them may have grown in a shallow subtidal setting.

It is concluded overall that the stromatolites of the Baumann Fiord Formation are representative of intertidal zone accretion. The polygonal and domal forms have presentday interidal analogues. The digitate forms may well have originated in either the low intertidal or very shallow subtidal setting. One provisory note warrants inclusion here, however. Garrett (1970) has demonstrated that if certain living intertidal and supratidal algal mats are transplanted into the low intertidal or shallow sublittoral zone and there fenced off from the browsing cerithid gastropods that would normally devour them, they are capable of survival and even accretion. If, as maintained in the previous section of this text, the Baumann Fiord Lagoons were practically devoid of browsers due to prohibitively high salinities, then it seems possible that the algal mats may have spread into the sublittoral zone. Some of the cryptal- galaminites of the Baumann Fiord sequence may be of sublittoral origin.

It is interesting to note that in the Baumann Fiord stromatolites, there is no definitive evidence of the former presence of interstitial gypsum, either in the form of observable gypsum pseudomorphs or otherwise. On the Trucial Coast, 'algal—mat' gypsum is restricted to the high intertidal zone, where evapori- tive concentration of the sea—water is sufficient to allow gypsum precipitation. In the Baumann Fiord setting, the lack of 'algal—mat' gypsum suggests that ether the landward fluxing brines did not become sufficiently concentrated within the near— shore zone or that high intertidal algal mats were not extensively developed. It is difficult to assess which is the more likely answer.

In summary, the following reconstruction is proposed: Algal mats were able to spread over the shallowest sediments of the 73

lime mudstone facies. These initial mats may have been either wholly or partially subtidal in nature and their propagation produced cryptalgalaminated fabrics and possibly as well some digitate stromatolite structures. Other digitate stromatolites along with polygonal and domal forms, structures that characteristically overly the cryptalgalaminite, grew in the intertidal zone and were founded upon the earlier formed algal mats.

Dolomite

Description

The anhydrite facies of the Baumann Fiord cycle very often rests directly upon the top of the stromatolitic horizon. But in some instances, there is preserved a thin zone of carbonate material between the stromatolites and the anhydrite (Figure 13). This zone, rarely more than two or three inches thick, consists of two components: a crudely laminated micrite and a microdolomite. The two are present in the ratio of 3 dolomite/1 calcite but they are rarely mixed in a uniform manner. Rather, the horizon is characteristically segregated into small lenses of relatively pure micrite residing in a mass of microdolomite. The dolomite is itself crudely laminated (Figure 25) and in some cases individual laminae can be followed from a micrite pocket into the adjacent dolomite. Desiccation cracks, unknown elsewhere in the cycle, are oc..c ionally preserved within this dolomite—micrite facies (Figure 26).

Interpretation

The original sediments of this facies are thought to have consisted of pure lime mud, washed over onto a supratidal flat by high spring tides or storms, and subsequently subject to desiccation—cracking due to prolonged exposure. This type of 'washover' lime mud is well known from the Trucial Coast sabkha and, as outlined previously, is highly susceptible to dolomitization by the marine brines that subsequently pass through it (Bush, 1973). It is not knoon why certain lenses in this washover facies were spared from dolomitization but perm ability 74

Figure 25. Dolomite with crude lamination inherited from its lime mud precursor. Monochrome photomicrograph. Crossed—nicols. Field .8 cm. top to bottom. Figure 26. Mudcracked lime mudstonc from the uppermost carbonate beds of a Baumann Fiord cycle, as viewed in plan. 76

differences may have been an important factor, the slightly 'tighter' pocicets having bCen by—passed by the fluxing brines.

Anhydrite

It would be misleading in the extreme to suggest that the anhydrites of the Baumann Fiord cycle are characteristically of the nodular form that typifies those of the Recent sabkha environment. In fact, most of the Baumann Fiord anhydrites have virtually none of the textural or morphological characteristics normally associated with anhydrites of supratidal origin. Rather, thinly layered or 'bedded' anhydrites are the norm; these are modes of occurence practically unknown in the Recent. It is nevertheless maintained, and much of Chapter 4 is devoted to the documentation of this assertion, that the Baumann Fiord anhydrites were originally of the typical nodular form but that they underwent extensive secondary modification: compactional and tectonic flow and in some cases metamorphic recrystallization; thereby eaten- sively transforming and even obliterating their primordial aspect and form.

In a few places, however, nodular structure in the anhydrite is preserved more or less intact and it is on these occurences that the following discussion is based. In the field, these occurences of anhydrite exhibit 'chicken—wire' structure (named as such by Forgotson, 1958), with closely spaced, almost inter- locking nodules, separated from one another by thin wisps of matrix material, either microdolomite, micrite or organic matter (Figure 27). Often, the wispy partitions between nodules (the actural 'chicken—wires-I) are offset or slightly stretched, indicating a degree of deformation in the nodular mass as a whole. In the classification scheme devised by Maiklem, Bebout and Glaister (1969), this 'chicken—wire' structure is termed mosaic or nodular mosaic grading, in the deformed parts, to distorted mosaic or distorted nodular mosaic. Close inspection of the nodules reveals that they are of the spherical to oblate form that characterizes many Recent sabkha anhydrite nodules (Figure 28). They are very densely packed and in many instance, fi

Figure 27. Nodular mosaic anhydrite in the upper part of a Baumann Fiord cycle, here preserved in a relatively undistorted form. Dark toned 'chicken—wire' between nodules comprises organic matter, micrite and microdolomite. 78

Figure 28. Nodular anhydrite viewed in slab surface normal to bedding. Field 3 inches across. 79

Figure 29. Anhydrite crystal fabric in halite—cemented nodule from the Trucial Coast sabkha; cleavage flakes arranged in a decussate manner. Monochrome photomicrograph. Ordinary light. Field 1.2 mm. across. Photograph courtesy of Dr. D.J. Shearman. 0

Figure 30. Anhydrite crystal fabric in nodule from the Baumann Fiord Formation. Note identity of fabric with that of the Recent nodule (Figure 29). Monochrome photomicrograph. Ordinary light. Field 1.8 mm. across. 81

matrix material constitutes less than 5 percent of the bulk volume of the rock.

Anhydrite nodules have been shown to grow in sediment by displacement (Kinsman, 1966; Shearman, 1966), the mechanical force of nodule growth being sufficient to push the host material aside (Shearman and Fuller, 1969). This mode of growth produces a distinctive internal fabric in the anhydrite. In thin—section, Recent anhydrite nodules from the Trucial Coast sabkha are seen to consist of lath—like cleavage flakes of anhydrite arranged in a decussate manner (Figure 29). This fabric arises as a natural function of nodule growth (Shearman and Fuller, 1969). The petrographic character of the Baumann Fiord nodular anhydrites (Figure 30) is, in all essential regards, identical to that of the Recent sabkha anhydrites (compare Figures 29 and 30). Because of this petrographic identity between the Recent and ancient anhydrites, and taking into account their mutual morphological similarities, it is concluded that the Baumann Fiord nodular anhydrites must have arisen in a supratidal environment; i.e. in a sabkha setting. This conclusion, of course, is a strictly uniformitarian one, for nodular anhydrite is unknown in the Recent except in the supratidal environment.

As a reconstruction of how the Baumann Fiord nodular anhydrites came into being, the following sequence of depositional and early diagenetic events is postulated: Above the intertidal zone algal stromatolites, there was deposited some thickness of supratidal 'washover' lime muds. Subsequent dolomitization of these muds, due to reaction between the sediments and the fluxing brines, yielded the 'dolomite' facies. This dolomitized washover horizon is only locally preserved in the Baumann Fiord sequence and it is suggested that its lack of persistence is due to the disruption that early diagenetic growth of nodular anhydrite affected within it. The nodules grew displacively and jacked up the host carbonate, incorporating only minor amounts of the carbonate (the wisps between nodules) as they grew; eventuall yielding a nodular mosaic anhydrite, deposit. The viability of this displacive 'jacking' mechanism is further investigated in Chapter 3. 82

It is concluded that the nodular anhydrite of the Baumann Fiord sequence grew within washover lime muds and dolomitzied washover lime muds. Support for this conclusion stems from the observation that the matrix 'chicken—wire' of the nodular facies is comprised of mud—size calcite and microdolomite in about the same proportions that they are found in the undisturbed 'washover' facies.

Some of the anhydrite must have been emplaced in pure organic matter for, in certain instances, the 'chicken—wire' comprises brown organic matter only. This mode of growth is not manifest in the present—day Trucial Coast setting but has been fully documented and illustrated in some of the Devonian T:Tinnipcgosis anhydrites of Saskatchewan (Shearman and Fuller, 1969). Growth of anhydrite nodules in organic matter appears to be as viable as in any other host material (Figure 31). Pre- sumably, the organic matter in which some Baumann Fiord nodules grew was the muscillaginous remains of former algal mats.

The foregoing review of the mode of growth of the Baumann Fiord anhydrites has been included here for the sake of complete- ness; i.e. to show that the anhydrites themselves are consistent with the sabkha model of cycle genesis, as advocated for the remaining components of the cycle. A more detailed analysis of the Baumann Fiord anhydrites is undertaken in Chapter 4.

24 Summary and Conclusions

The interpretive ideas presented in this chapter have been largely substanstiative in nature in that they have focused on establishing the intrinsic viability of the sabkha model in explaining the observed features of the Baumann Fiord cycle, rather than on attempting to discredit possible alternative interpretations. The sabkha model, in its most fundamental form, comprises the following- essential requisite elements: a basal shallow marine facies, followed upwards by an intertidal facies, in turn overlain by an evaporite facies in which anhydrite is a dominant constituent; the whole being bounded top and bottom 33

Figure 31. Nodular anhydrite set in a matrix of calcitized organic matter. Displacive nodule growth predates calcitization (Shearman and Fuller, 1969). Large calcite crystals are distinguished by the pseudopleochroism imparted by the included organic matter. Winnipegosis Formation (M. Dev.), Saskatchewan. Colour photomicrograph. Ordinary light. Field 2.7 cm. across. Photograph courtesy of Dr. D.J. Shearman. 84

by erosion surfaces. It is felt that the sequential components of the Baumann Fiord cycle accord completely with this basic sabkha model of cycle genesis. Furthermore, there do not appear to be any critical aspects of the Baumann Fiord cycle that mili- tate against the model.

Outlined below are the principal conclusions of this chapter, as they relate to the evolution of a single Baumann. Fiord cycle: (i) The cycle commences with rapid marine transgression over a pre—existing sabkha surface, producing erosion of the uppermost reaches of the anhydrite facies (erosion surface) and depositing a carbonate lag in its wake (flat—pebble conglomerate). Many of the pebbles of the lag deposit are thought to have been locally derived, by autobrecciation, transport and rounding of desiccated supratidal carbonate muds. Other pebbles are clearly exotic, derived from an area nearer to the open marine front. (ii) The shallow sublittoral environment established by the transgression was conclusive to deposition of lime muds, some of massive aspect, other delicately laminar. The salinity of this lagoonal environment may have been such that direct precipitation of aragonite was possible. Lack of fossils at this level may be attributable to high salinity conditions. The lime mudstone facies as a whole appears to 'shallow—upwards', with swash—zone symmetrical ripples being not uncommon at the uppermost levels. (iii) Algal mats, which constituted the leading edge of a regressive front, colonized the top of the lagoonal lime mud facies. Cryptalgalamini.te and some digitate stromatolites may have been partially of subtidal origin but overlying polygonal and domal stromatolites are thought to have been laid down in the intertidal zone. (iv) Supratidal lime muds were deposited above the algal mats, probably as a result of 'washover' from the lagoon. Early diagenetic dolomitization of these lime muds was extensive, due to the influence of landward—fluxing brines. 85

(v) The same landward—fluxing brines, undergoing continual concentration by capillarity and evaporation, gave rise to a trailing diagenetic front characterized by displacive growth of nodular anhydrite within the supratidal carbonate muds, leading eventually to the generation of a thick nodular mosaic anhydrite deposit above the stromatolites. (vi) Marine transgression over this supratidal flat terminated the cycle, and at the same time initiated the next cycle.

Throughout the study area, the Baumann Fiord Formation is made up of successive repetitions of this basic cyclical sequence. This is not to say, however, that there is no vertical or lateral variation in the way the cycles were compounded. Indeed, there must have been countless areal and temporal variations on the basic cyclical theme, resulting in extremely complex sequential records in different places. In the next chapter, an attempt is made to explain these complexities in terms of the environmental model erected in this chapter, thereby further justifying and reinforcing the conclusions reached here. 86

CHAPTER 3

COMPOSITE CYCLES, FACIES VARIATIONS AND

PALEOGEOGRAPHIC RECONSTRUCTION 87

3.1 Introduction

It is argued in this chapter that the whole of the Baumann Fiord Formation is explicable in terms of sabkha model ge,Iesis. In many places, the formation is made up of dozens of sequential repetitions of the basic cycle described in the previous chapter. But in most localities, at least part of the sequence is not so straight forward. There are instances of extraordinarily thick carbonate development and instances of thick anhydrite development. Lateral intertongaing of anhydrite and carbonate lithofacies is often extremely intricate. Indeed, the vertical-and lateral variations in the Baumann Fiord Formation of the study area are generally of a rather complex nature and, if the sabkha concept of its genesis is to retain credibility, these complexities must be explained.

This chapter deals initially with the nature and character of each of the sections examined, thereby establishing the scope of manifest complexity. The interpretive discussions that follow relate to vertical variation in the sequence, lateral facies variation across the depositional trend and parallel to it and the overall time and rock—stratigraphic relation of the Baumann Fiord Formation to its enclosing formations: The chapter ends with a reconstruction of the paleoenvironmental and paleogeographical evolution of the formation as a whole. In short, this chapter details the many sedimentalogical peculiarities of the formation and attempts to explain and assimilate them in terms of the sabkha model.

3.2 Section Description

The Baumann Fiord Formation was examined at eight seperate localities (Figure 6). Complete sections were measured at six of these (Sections lc, 2, 3, 5, 6 and 7). Near—complete sections were measured at Localities 1 and 4. At Locality 8, yree cover was too extensive to allow reasonable compilation of the section. Lithologic logs of all eight measured sections are depicted 88

in Figures 32 and 33 (inserted in pocket). These are grouped according to their mutual geographic proximity and their relative positions within the depositional belt. 'Figure 32 (Sections 1-1a-2-3) transects sections in the region of Troia Fiord, sections well within the folded belt and far removed from the onlap edge of the miogeosyncline to the east (Figure 6). The sections in Figure 32 are so arranged at to delineate a cross— section that is perpendicular to the regional trend in the area. Figure 33 (Sections 4-7-5-6) transects a line of section essentially parallel to the regional trend. All these sections lie within the undisturbed part of the Franklinian Miogeosynclinal belt and all are in close proximity to the onlap edge of the miogeosyncline. The horizontal datum on which the sections are placed is the base of the B Member. This datum was chosen because it is thought to be consistent 'time line'. Both the top and bottom of the formation are somewhat diachronous.

The primary purpose of the section logs is to depict gross lithology but where accessary lithologies assume significant proportions (between 10 percent and 40 percent) these are noted (right extremity of each column). Sedimentary structures have not been included in the logs because of the small scale on which these are normally developed. It should be remembered, however, that each of the thin carbonate bands is cbaracteris- - tically comprised of_the typical cycle's structures (flat— pebble conglomerate, lime muastone, stromatolites). Differen- tiation between anhydrite and gypsum is on inherent lithology and is not a reflection of degree of surficial weathering. Thicknesses shown reflect true stratigraphic thickness, not gross thickness, the latter often being considerably greater due to tectonic enhancement (overfolding and thrusting).

Thickness and exposure data for the measured sections is

It should be noted that the base of the B Member was everywhere ass -:ne( to the first resistive carbonate beds above the recessive --.ember strata. In certain instances, the uppermost A Member beds consist not of anhydrite but of anhydritic limestone (eg. Sections 1 and la, Figure 32). 89

Thickness and Ex osure Data for Measured :Bauman Fiord Sections

Total Covered Exposure Section Region Thickness

1 Trold Fiord 1357+ 18 99%

la Trold Fiord 1552 96 94%

2 Flat Pebble Bay 1193 65 95%

3 Star Fish Bay 1022 126 88%

4 Sverdrup Pass 670+ 33 98%

5 Sandola Creek 868 85 90%

6 Bartlett Bay 760 437 42%

7 Witch Mountain 912 101 89%

Figure 34 90

Tar. Mbr. Thickness Covered Exposure Carbonate

C 162 0 100% 54% B 144 0 l00% l00% A 1051+ 18 98% 31%

C 72 57% 41% B 0 100% i00% ,A ri 24 98% 28% n/a - - - - n/a - - - - A 193 65 95% 47% n/a - - - - n/a. - - - - A 1022 126 88% 35%

C 206 33 84% 28% B 127 0 l00% 100% A 337+ 0 l00% 18%

c 145 20 87% 10% B 107 0 100% i00% A 616 65 89% 11%

c 115 79 31% 16% B 87 0 i00% 100% A 558 358 36% 16%

C 187 48 74% 11% B 154 0 l00% l00% A 571 53 90% 18%

Figure 34 (contd.) 91

compiled in Figure 34. Exposure generally exceeds 90 percent with the exception of Section 6, Bartlett Bay (42 percent). Such exposure is not necessarily manifest in a single stream cut or scree buttress. Each illustrated section was compiled by com- bining the results of measurement in a number of laterally adjacent exposures, usually within about half a mile of one another.

Sections 1, la, '2 and 3 can be considered as a group (Figure 32). They have the following characteristics in common: (i) They all lie well out into the miogeosyncline. Section 1 is some 50 miles west of the onlap edge of the sequence and Section 3 some 35 miles (Figure 4). (ii) All the sections lie within the folded part of the Frank- linian Miogeosyncline (Figure 6). Folding in this region is generally of a subdued nature, amplitude 2,000-3,000 feet, wave—length 2-3 miles. Locally the deformation is prodigious, with overturned folds being not uncommon in certain confined horizons (Figure 35). (iii) Due to tectonism, most of the carbonate rocks carry calcite—filled fractures (Figure 36). The anhydrite rocks are not fractured but show some evidence of flow (see Chapter 4, Section 4.4).

Section 1 is the type section of the Baumann Fiord Formation (Kerr, 1967b). It is well exposed on the eastern slope of a mountain ridge that parallels Trold Fiord (Figure 37). The underlying Copes Day Formation is not exposed in the mountain side and there is a reverse fault near the base of the sequence (mapped as such by Thorsteinsson (1972) and evidenced as a 'repeat' sequence in the section log; Figure 32, levels 1,400- 1,500). Both these shortcomings render it somewhat deficient as a type section. Section la is some 6 miles east of Section 1 and comprises a complete succession through the Baumann Fiord Formation, including the contact with the underlying Copes Bay Formation

■■•■••■■■••••••.... thickness covered. ) percent Exposure is calculated as follows: 100 ( total thickness Figure 35. Overfolds in basal B Member carbonate beds; Locality 1. Folded rocks are etratigraphically confined to a 60 foot interval. Regionally, the deformation is very subdued (see text). 93

Figure 36. Fracture system in folded carbonate beds of the Baumann Fiord Formation; Trold Fiord/egion. Fractures are filled with white calcite. . 94

(Figure 38). It is well exposed (9e, Figure 34) and is not marred by any apparent overfolds or faults. In that it lies within the designated type area (Kerr, 1967b), it is herein recommended that this section serve in future as the principal reference section for the formation.

Section 2 lies some 25 miles south of Section la but relative to the regional strike, it is only some 6 or 8 miles further east; i.e. closer to the craton (Figure 32). A well developed monocline within the Baumann Fiord Formation at Locality 2 does not complicate the section, as it is exposed in a relatively undeformed setting along a prominent adjacent ridge (Figure 39). Both Sections 2 and 3 lack evaporite development in the zone which would normally be designated as the C Member. Consequently the B Member, in that it is lithologically indistinguishable from the Eleanor River Formation, is mapped as part of the latter.

Section 3 is still closer to the craton than Section 2 (Figure 32). The Baumann Fiord Formation is well exposed at Locality 3 at the end of a long mountain ridge where a deeply incised stream breaches the topography (Figure 40). Although there is some talus cover near the base of the section, the contact with the Copes Bay Formation is exposed.

Even though Figure 32 is not intended as a correlation section, a,number of interesting relationships may be delimited through comparative analysis of adjacent sections. Firstly, it is clear that the formation as a whole thins toward the east, i.e. towards the craton. Secondly, it is apparent that while in many instances correlative units can be traced laterally from section to section, there are significant portions of each section in which the sequence is distinct from that of its neighbours. The intertonguing of the carbonate and anhydrite lithofacies is complex and the lateral variation appears to be manifest both perpendicular to the depositional strike (compare Sections 1 and la) and. along the depositional strike (compare Sections la and 2)- Finally, it is clear that the C Member in the region of Trold Fiord dies out towards The south, apparently 95

Figure 37. Baumann Fiord Formation Section 1. (Type Section). Cliff—forming Eleanor River Formation at the top; thin recessive C Member at base of cliff; bluff—forming B Member limestone below; and long slope of recessive A Member. Copes Bay Formation not exposed. Peak in right centre is 3,200 feet (4 miles distant). 96

Figure 38. Typical exposure of Baumann Fiord Formation. Section la. Dark toned material is limestone (with hammer resting on it). Light toned material is anhydrite (with some surficial gypsum). 97

Figure 39. Baumann Fiord Formation A Member, Section 2. Note vertically dipping monocline limb in lower right of photograph (see text). Figure 40. Baumann Fiord Formation Section 3. C Member not developed in this region. Jointed Eleanor River strata at the top are here highly prone to landsliding. Mountain 2,600 feet in elevation. 99

along strike.

The other four sections (4, 7, 5 and 6, Figure 33) have the following characteristics in common: (i) They all lie in close proximity to the onlap edge of the Franklinian niogeosyncline (Figure 33). (ii) All the sections lie within the stable part of the Frank- linian Niogeosyncline (Figure 6) and the strata are flat— lying or very gently tilted but otherwise undisturbed. (iii) Unlike the Trold Fiord sections, the carbonates in this region are unfractured and the anhydrites do not show evidence of tectonic flow.

Section 4 is well exposed in the steep canyon walls of a deeply incised stream (Figure 41). Unfortunately, the base of the formation, and the contact with the underlying Copes Bay is buried beneath alluvium. The evaporites near the base of the pictured section (Figure 41) consist of secondary gypsum, not anhydrite (Figure 33); and these beds are ramified throughout with satin—spar gypsum veins.

Section 7 is some 5 miles east of Section 4 and within 2 miles of the outcrop edge of the Precambrian Shield (Figure 33). Baumann Fiord strata are there well exposed along the southern slope of Witch Mountain (Figure 42). In the 100 foot zone above the Copes Bay contact there is extensive development of secondary gypsum and satin—spar veins are again present in abundance.

Section 5 comprises a complete section of the Baumann Fiord Formation, ideally exposed along the steep slopes of the canyon cut by Sanddola Creek (Figure 43). This section contains some of the best exposed nodular anhydrite encountered in the study area (particularly at Level 550-580, Figure 33). Secondary gypsum and satin—spar veins are again in evidence in the basal parts of the formation.

Section 6 at Bartlett Bay is the most easterly section examined in the study Lrea (rigure 33), but it is unfortunate that this extremity of Baumann Fiord outcrop is not better exposed (Figures 33 and 44). Talus obscures large parts of the 100

Figure 41. Baumann Fiord Formation Section 4. Exposed thickness is 670 feet. Figure 42. Paumann Fiord_ Formation Section 7. Copes Bay Formation crops out along a line level with the lowermost snow patch in right—centre of photograph. Formation is here 912 feet thick. Tents and Canadian Flag in foreground for scale. Figure 43. Baumann Fiord Formation Section 5, as viewed from the air. Resistive Eleanor River Formation forms the plateau. The Copes Bay cliff—forming limestones are visible at bottom— centre of photograph. a.03

Figure 44. Badly weathered and completely gypsified beds of the Baumann Fiord Formation A Member. Locality 6. 104

sequence and exposed strata are altered to gypsum to a depth of up to 2 feet, making it difficult to procure fresh samples.

Comparative analysis of these sections (4-7-5-6, Figure 33) reveals some points of interest. Firstly, there seems to be a gradual thinn]ng of the formation towards the east, particularly in the B and C Members (Figure 33). It is possible that the Baumann Fiord Formation in the region of Section 6 is very close to the untimate 'along—strike' extremity of its development, as evaporites at this level are unknown in Greenland (Troelsen, 1950). Secondly, there is much more elastic material in these sections than in those further out in the miogeosyncline. Proximity to a elastic source area, the Precambrian Shield, presumably accounts for the shaly and occasionally sandy nature of the sequences in this region. Finally, it is apparent that lateral variation in facies development is pronounced at certain levels, for correlation between sections is not always obvious. In this regard, these !stable region' sections are not unlike the Trold Fiord sections.

Apart from purely documental considerations, the foregoing description of the eight measured sections has attempted to outline the scope of the formation's variability over the study area. net follows is a detailed discussion of the ways in which this variability may be explained in terms of the sabkha model. Firstly, an analysis is made of the vertical variation in the formation to explain how departures from strict rhythmic cyclicity may have resulted. Secondly, lateral variations in the formation are discussed to show how the complex intertonguing of facies may have developed.

3.3 Vertical Variation in Sequence — Compound Cycles

One of the remarkable features of the sabkha model is that it seems inherently conducive to the generation of cycles. This point was touched on in previous discussion but its further elaboration is appropriate here. 105

A single sabkha cycle is comprised of two distinct phases: a regressive phase and a transgressive phase. Sedimentary accumulation takes place during the regressive phase, not strictly y vertical build-up but primarily by gradual lateral accretion. This lateral accretion is brought about by means of a series of progradational fronts, each advancing seaward at approximately the same rate. First is the algal mat front, which progressively advances over the shallowest lagoonal sediments. Behind the algal mats is the front of supratidal carbonate accumulation. Following this is a series of diagenetic fronts; including among others the dolomitization front, the interstitial gypsum front and the front of displacive anhydrite growth. Once the anhydrite front has passed, the system apparently stabilizes, with no new mineral species arising, but with anhydrite continuing to grow within the supratidal facies as long as the brine supply is maintained. The system remains in this relatively stable state until the onset of the transgressive phase, when the sabkha flat is inundated and lagoonal conditions are established once again over the area. As was argued earlier, this transgressive phase tends to be rapid.

The question which arises is this:- why, in a system of such apparent tectonic stability, should the regressive phase so consistently produce sedimentary accumulation while the transgressive phase produces only erosion? The answer to this question seems to lie in the assessment of the delicate inter- play between sedimentary accumulation and regional subsidence. Consider a single point in a sabkha complex and assume that subsidence at that point proceeds at a constant rate. At the beginning of a given cycle, that point will be in the shallow lagoon. As the progradational front passes, vertical accummulation will be very rapid at the point, for. it will go from a few feet below mean sea-level to a few feet above in a relatively- short period of time. During the passing of the front, then, accumulation rate greatly exceeds subsidence rate. Subsecuently, vertical accretion at the point will tail off, for as the shoreline progrades further and further away, so the mechanisms of sediment supply become less and less efficient. Washover mud will be less 106

likely to reach the point and vertical jacking by displacive anhydrite growth will diminish because brine supply mechanisms will become less efficient. As the vertical accumulation rate wans,- subsidence will begin to dominate and the point will start to sink relative to mean sea level.

Expanding the corollary to encompass the whole of the sabkha complex, consider what will happen when progradation brings the front of the sabkha to the seaward extremity of the system, either to the marine barrier complex or, in an extreme case, conceivably as fax as the edge of the shelf. Iihen progradation can proceed no further, virtually the whole of the sabkha surface behind the shoreline will be in the state where sedimentation cannot keep up with the continuing subsidence. Due to various processes, sabkha surfaces tend to be perfectly flat and if the whole of the sabkha plain is sinking relative to sea level, the time will inevitably come when land—level and sea—level coincide and the sea will be free to immediately inundate the whole of the surface. For a time subsequent to the transgression, sedimentation in all but the landward portion of the newly established lagoon must remain subservient to subsidence, seaward parts gradually deepening until the passing of the progradational front of sedimentation boosts them once again above sea level.

According to this theoretical model, steady and uninterupted subsidence should provide for the generation of repeated cycles with little or no variation in their vertical character or lateral extent. For the Baumann Fiord Formation, the model would seem to apply in a number of instances; as exemplified, for example, in the near—perfect rhythmic cyclicity of the sequence in Section la, levels 800-1,200 (Figure 32), where there are 32 complete cycles in 400 feet of section. At least certain parts of the other sections also appear to accord with the model. Clearly, there must have been long periods during the build—up of the Baumann Fiord Formation when subsidence was perfectly constant and rhythmic compounding of cycles was possible.

Accepting the dictum that constant subsidence rate over a 107

very broad area is a requisite for the development of rhythmic sahkha cycles, it seems logical to postulate that slight local or temporal changes in subsidence rate could promote very consid- erablediversification in the way cycles are compounded. Indeed) it is argued below that fluctuation in subsidence rate is the key factor in the generation of 'non—rhythmic' or 'composite' cycles.

In order to elucidate what is meant by 'composite' cycles, it is perhaps best to. consider specific instances in which the ;aumann) Fiord sequence departs from conventional 'rhythmic' compounding of cycles. One of the most obvious departures from rhythmic cyclicity is the case where anomolously thick limestone units are developed (Figures 32 and 33). In the field, it is immediately plain that these are comprised of the same type of lime mudstone that normally makes up the carbonate component of the typical cycle. However, close examination of the sedimentary structures in a thick carbonate unit often reveals that it is made up not of a single Ishallowing—upward' sequence but rather of a number of incomplete lagoonal sequences. In a case such as this, it is clear that cyclical development was still operative within the system as a whole but that the progradational front repeatedly failed to reach that specific locality before it was swamped by the next transgression. Thus, that particular carbonate unit, even though it was built up through continuous deposition, is in fact representative of a number of cycles, each of which was prematurely terminated in that particular area. The best explanation of this is that subsidence rate increased for a time, throwing the system out of equilibrium and bringing about untimely swamping of the progradational front.

Most of the thick carbonate developments are of this 'composite' type, meaning that they are made up of two or more 'semi—cycles', each consisting of part or all of the carbonate component of the full cycle. The tops of these thick carbonate units characteristically exhibit normal transition into the overlying anhydrite: lime mudstone grading up into a stromatolite zone, immediately overlain by anhydrite. Clearly, subsidence ultimately slowed down to the extent that the progradational front was finlly aide to reach and pass byon0. that particular 108

locality.

Locally, however, thick carbonate developments are not of the 'composite' type. In certain places, for instance, the carbonate component of a cycle may be enhanced by inordinate stromatolite development; i.e. the cycle is perfectly normal save that the stromatolite zone, instead of being about one foot in thickness, is five or six feet thick (Figure 45). Accretion of such a thickness clearly records temporary halting of the progradational front: the intertidal zone having remained in one place for an inordinate period of time. It is thought that this resulted from an increase in subsidence rate, not so pronounced as to induce a transgression, but sufficient to balance the accretion rate of the intertidal algal stromatolites. With the front of progradation stalled, the stromatolites in that particular locality would be free to build up apparently anomolous thicknesses. Only when subsidence rate decreased could progradational advance of the intertidal zone once again proceed. Examples of thick stromatolite development of this type are relatively rare and are of restricted areal extent. This is certainly what would be expected. For the intertidal zone to remain stationary for an extended period of time requires that stromatolite accretion must be exactly balanced by regional subsidence.

Enhancement of the carbonate component of a cycle may also come about by means that are seemingly not controlled by subsidence variation. One example of this is the local occurence of thick flat—pebble conglomerate (Figure 46). Where such conglomerates were observed, they usually exhibited an imbricate configuration suggesting that they marked the strand line of a marine embayment. Another example of thick carbonate development is the local occurence of 'algal mat breccia', consisting of disordered arrays of cryptalgalaminated clasts, some of which appear to have been contorted while still in a very 'rubbery' state (Figure 47). It is thought that these accumulations of algal mat debris were the result of local 'rip—up' of mat, by severe storm action or by high—energy reworking during •• h • •

1 -"V )

. !. ")ilt4 :*'4 ..,.,

Figure 45. Stromatolites forming the upper part of an extraorinarily thick carbonate component of a cycle. Light toned material (on which pipe is resting) is laminar anhydrite. Note how digitate stromatolites grade upwards into domal forms. 110

Figure 46. Thick accumulation of limestone flat—pebble conglomerate. Note tendancy toward pebble imbrication — sloping downwards to the left. Figure 47. Algal mat breccia, here accumulated as a thick pile of ripped—up mat, contorted in an appartently rubbery fashion. Note overturned piece of cryptalgalaminated material below the pen at left centre of photograph. 112

transgression. Occurences of both beach—pebble conglomerate and the algal mat breccia are relatively rare and they do not account for any areally extensive enhancement of carbonate thickness.

In any one section of the Baumann Fiord Formation, there are numerous examples of departure from rhythmic cyclicity. Thick carbonate development, usually taking the form of composite carbonate cycles, is one example of this departure. The other important variation on the cyclical theme is the devalopment of anomolously thick anhydrite units. Like the composite carbonates, the thick anhydrites are thought to arise in response to fluctuation in subsidence rate. Various facets of accretion—subsidence balance and their possible effect on variations in anhydrite are discussed below.

The first ouestion that must be answered is: What maximum thickness of supratidal nodular anhydrite may arise in the course of a single sabkha cycle? Recent nodular anhydrite has been shown to grow displacively within the capillary zone above the groundwater table (Shearman, 1966; Kinsman, 1966). But while capillary concentration may be a requisite of nodule growth, there is abundant evidence that, once formed, anhydrite can persist in the zone below the groundwater table (Bush, 1973). Thus it is apparent that the potential thickness of the nodular anhydrite facies is not restricted to the two to three feet that resides above the stable groundwater level. Given a background of steady subsidence, anhydrite may continue to grow within the capillary zone while previously formed nodules subside below the groundwater reference datum. As long as displacive nodule growth produces enough vertical jacking of the sabkha surface to keep pace with subsidence, marine transgression will not occur and the anhydrite facies may continue to thicken.

The- Baumann Fiord cycles average twelve feet in thickness. The basal three to four feet of each cycle is the carbonate component and the anhydrite facies is characteristically limited to eight or nine feet. Clearly, there must have been environmental factors that tended to limit the anhydrite facies to this order of thickness. 113

As postulated earlier, there would appear to be two such limiting factors. The first relates to brince supply mechanisms and the efficiency thereof. Bush (1973) has shown that while the sonation of the diagenetic facies within the sa-bna arises in response to -the evaporation and selective depletion of fluxing intra—sediment brine, the continued growth of anhydrite in areas well removed from the leading progradational front is likely contingent upon flood—water replenishment of brine. The efficiency of the flood—water supply mechanism must decrease as a function of distance from the progradational front. Even though a supratidal surface may be perfectly flat, seawater driven onto the plain in the course of a storm will not necessarily penetrate to the back of the sabkha, for it will be undergoing absorption into the sabkha sediments all the while. Only the most severe of storms would appear to be cwpable of completely inundating a sabkha surface of the dimensions of the Baumann Fiord system (up to 70 miles wide and hundreds of miles long). Clearly, then, there are restrictions on the amount of calcium and sulphate that can be supplied through the flood—water mechanism. Once supply diminishes beyond a critical point, the delicate balance between subsidence and accretion is tipped in favour of subsidence, marine transgression becomes inevitable and the cycle will terminate.

The other factor that may tend to limit the thickness of the anhydrite component of a cycle is related to the amount of potential host sediment. If anhydrite nodules grow by displacement, then they must have a host in which to grow. Consider a system in which there is initially only one foot of supratidal carbonate sediment. If growth of nodular anhydrite within that sediment yields a nodular mosaic consisting of 95 percent anhydrite, 5 percent matriX, then the carbonate will be totally consumed only when the facies reaches 20 feet in thickness (Figure 43). There is, of course, a very broad scope for variation on this theme and the assessment of the limiting influence of the 'host consumption' parameters is spurious at best. There is no reason, cf course, why anhydrite could not act as host for some later_anhydrite, Suffice to say that if ~1~;:?;] Host Sediment - Matrix

~ r~odu'ar Mosaic Anhydrite (95·'.anhydrite,5·/.matrix)

\ 20ft. \

1ft.~r.fJ~ ___.-..J/ --_/ Continual Anhydrite Growth By Displacement Of Host Sediment ------> Vertical Jacking

Figure 48. Schematic a.epictioll of the ,-ray in vlhich one foot of supratidal sediment may be jacked I-' -to 20 feet of t chicken-Nire t anh;:rrlri te by c:)ntinued displaci. ve nodule groT,rth. I-' ~ 115

the amount of potential host material is llmited, then. the continued growth of displacive nodular anhydrite may be adversely affected.

To return to the original question of how thick the anhydrite component of a single cycle may become, it is clear that there is no single answer. The thicTmess that a supratidal anhydrite facies may attain is theoretically unlimited. All that is required is that accretion rate equals or exceeds subsidence rate. Figure 48 may serve to graphically clarify this point. If, during the time it takes to accrete 20 feet of nodular anhydrite, subsidence in the area is less than 20 feet, the top of the facies will remain positive. If, however, subsidence exceeds 20 feet, the system will undergo marine transgression and the next cycle will be initiated.

In summary, it is clear that although there may be no theoretical restrictions on the accumulation of sabkha facies anhydrite, there are practical restrictions. For the Baumann Fiord system, it would appear that these restrictions normally made their influence felt when the anhydrite facies reached eight to nine feet. What the foregoing discussion has attempted to elucidate is the physical nature of the restrictions. Two possible influences have been postulated: brine supply efficiency limits and host consumption limits. There may well be other influences.

Extending the subsidence—accretion argument to the problem of anomolously thick anhydrite units, the following corollary seems applicable. If local or temporal increase in subsidence rate fascilitates the compounding of thick composite carbonate units, then local or temporal decrease in subsidence rate may explain thick anhydrite units. Consider an area that for a time was subject to subsidence at a rate somewhat slower than that of the remainder of the sabkha plain. That area would remain positive when the next transgression took place. Yore importantly, accretion of anhydrite within the positive area should be regenerated in direct response to its newfound proximity, to the shoreline and the enhancement of .wine supply effic::;_ency that 116

the new situation afforded. In other words, conditions favourable to rar:kha facies accretion would be restored despite the fact that the transgression did not establish a new discrete cycle in that area. If subsidence were to lag ehind that of the remainder of the system for a considerable period of time and accretion of anhydrite was repeatedly regenerated, then there is no reason to believe that very thick anhydrite development could not take place.

This 'sabkha regeneration' argument is a theoretical one but it may have considerable applicability in the Baumann Fiord sequence. Its validity can only be evaluated in conjuction with a knowledge of how the Baumann Fiord cycles vary laterally; how they intertongue and how they pinch out. These relationships are considered in detail in the section that follows.

3.4 Lateral V7,riation in ")ecuence-Facies Intertonguing

In the foregoing discussion the author has attempted to elucidate some of the vertical variation in the Baumann Fiord sequence, as exemplified in the measured sections. But a full picture of the facies pattern within the formation as a whole is central to a complete understanding of the depositional history of the formation and this can be accomplished only if the lateral interrelations between sections is taken into account. -lhat follows is a documentation of how the sections correlate with one another and how the thickness and character of the cycles change laterally, both perpendicular to the regional trend and parallel to it. By thus assimilating the regional facies pattern, it becomes pesSible to assess the ultimate applicarcility of the sabkha model in explaining the origin of the formation as a whole.

The cross-sec-tin shown in Figure 49 schematically depicts the lithostratigraphic variation in the Baumenn Fiord Formstion. The left extremity of the section is representative of the sequence in the region of Troia Fiord, i.e. that part of the depositional belt furthest removed from the crato. The right extremity is representative of the sequence immediately Eleanor I SE River Fm

Copes Bay Fm

Anhydrite.. _ Limestone EA 20 Miles

Figure 49. ;7chematic representation of the facies pattern within the Baumann Fiord Formation. Carbonate units tend to die out towards the craton in the southeast and the formation as a whole thins in the same direction. 118

adjacent to the onlap edge of the formation. The cross—section thus transects the formation along a line normal to the depositional strike.

A number of fundamental relationships are illustrated in the section. Firstly, it is clear that limestone comprises a greater proportion of the formation in the west than it does in the east. In the Trold Fiord region (Figure 32), the A Member is made up of 34 percent limestone, 66 percent anhydrite on average (Figure 34). These figures closely mimic the proportions in which the two components are represented in each of the cycles that make up the section: 3 to 4 feet limestone (24-335) and 8 to 9 feet anhydrite (66,40 — 75%). But the proportion of carbonate in the seouence progressively diminishes in the direction of the craton. In the 'near—craton' sections (Figure 33), the percentage of carbonate in the A Member averages 15 percent (Figure 34). There is a similar trend in the C Member, carbonate being considerably more abundant in the west than in the east (Figure 34). Clearly, the development of the lagoonal component of the cycle was more prevalent and more regular in the distal portion of the depositional belt tha9 it was in the portion proximal to the craton.

This observation leads directly to the second fundamental relationship that Figure 49 attempts to bring out. The limestone stringers in the Baumann Fiord sequence tend to die out from west to east, towards the craton. This relationship is evident in the field. In places it is possible to follow individual beds for distances of up to 8 miles. ',There this was done along exposures that paralleled the regional trend, the . limestone stringers were observed to have remarkable continuity and uniformity. But where individual sets of strata were traced laterally from west to east along exposures that tran- sected the regional trend, it was observed that while the sequence still showed a considerable degree of continuity, some of the limestone stringers pinched out. In one such transverse section in the region of Locality 1, two limestone beds were observed to terminate within one and a half miles while a further seven 119

persisted virtually unchanged over the same distance, west to east. Attempts to correlate lithologic logs of the measured sections are greatly facilitated by the remarkable lateral continuity that individual beds often exhibit. Indeed, it is commonly possible to make reasonable correlations between sections that are some 10 or 20 miles apart, so extensive are many individual beds. Figure 50 illustrates some correlations amongst Sections la, 2 and 3. Acknowledging that there is some risk in correlating over such considerable distances, it is thought that the correlations shown in Figure 50 are justified and that, as a rule, the Baumann Fiord sequence tends to be uniform and continuous over broad areas. Of course, it is the exceptions to the rule that are of immediate concern here. In this regard, the sections in Figure 50 appear to bear out the relationship that was locally observable in the field; i.e. that where carbonate units do pinch out, they do so in an easterly direction, toward the craton.

Figure 49 provides conceptual reconstruction of how the Baumann Fiord Formation was built up. Each limestone bed, although regressive, marks a marine transgression. Diany of the transgressions penetrated right to the back of the sallh:ha system. Others only partially inundated. the previous sablcha surface. The explanation for this would appear to lie in the application of altuments developed earlier based on subsidence versus accretion. If subsidence in the distal portions of the depositional belt exceeded that in the proximal portions to an extent that more than compensated for any accretionary advantage that the distal region may have had, then transgression could proceed only as far as that subsidence contrast allowed. As suggested earlier, a new and distinct cycle could thus be initiated in the distal region with commensurate regneration of sabkha facies accretion in the proximal region. Numerous transgressions, each contingent upon subsidence balance and each followed by progradational advance of the sabhha, could produce the variability in sequence illustrated in Figure 49. It is thought that this reconstruction is both compatible with the observed character of the formation as a whole and intrinsically consistent with the Figure 50. Correlations within the basal Baumann Fiord Formation; Trold Fiord region. Line of section is west to east, towards the craton (see Figure 32, pocket insert).

Section la Section 2 Section 3 030

V V V V V

4•■••■■••■■•■ 0021 0 3 0 Ogl I 1 I I I I I I 1 I I I I 1 I I I I

✓ V V ✓ V ✓ V V 00p1 ✓ V ✓ V V ✓ V ✓ V

✓ V ✓ V V ✓ V 030 ✓ V V v ✓ V V ✓ V ✓ V V V V ✓ V V 0001 ✓ ✓ V V ✓ V V ✓ V OOCL ✓ V V ✓ V

✓ V ✓ V V

✓ V V ✓ ✓ V V ✓ V V ✓ V ✓ V 006 006 ✓ V V ✓ V V ✓ V V V ✓ V V 00Z1 ✓ ✓ V ✓ V ✓ V V ✓ V V ✓ V ✓ V V V V ✓ V ✓ V V ✓ - V V V V ✓ V V ✓ V V V V V ✓ V V V ✓ V ✓ y V V

✓ V ✓ V V ✓ V V ✓ V ✓ V V V ✓ V V ✓ V ✓ V ✓ ✓ V V ✓ V V 008 008 ✓ V V ✓ V V ✓ V 0011 ✓ V V ✓ V V ✓ V ✓ V V ✓ V V ✓ V V V ✓ V V ✓ V V ✓ v ✓ V V v ✓ V V V ✓ v ✓ V V v ✓ ✓ V ✓ V V ✓ V V ✓ v V V V V v V V ✓ ✓ V v ✓ ✓ V V ✓ V V ✓ ✓ V V v ✓ ✓ v ✓ ✓ V V ✓ V v ✓ V ✓ V V V 1 OOL V v V OOL ✓ V v ✓ ✓ V ✓ V V V ✓ V V ✓ 0001 I ✓ V ✓ V ✓ V ✓ V V ✓ V V ✓ V V v

V V V ✓ V V ✓ V ✓ V ✓ V V ✓ V V ✓ V 1 I I ✓ V ✓ V V Vi ✓ V V ✓ V v V V ✓ ✓ V v v V V ✓ V V ✓ V I I V V V V ✓ ✓ V ✓ V V ✓ V V V ✓ V ✓ V V ✓ ✓ V V I I 1 I I V ✓ V ✓ I I ✓ v ✓ v V 009 ✓ V V 009 1 1 Y V v I I ✓ V 006 I I • V V ✓ V ✓ V V v E VI V I ✓ 1 • v ✓ V I 1 • • V ✓ V V 122

sabkha model.

Application of the differential subsidence model is also thought to explain the lateral variation along strike. In Section 2 for example, there is a preponderance of carbonate between levels 100 and 500 (Figure 32), much of which is clearly time—equivalent to anhydrite in the sections to the north. This relationship is readily explained by postulating that for the time interval in question, subsidence in the vicinity of Section 2 tended to exceed that in the area to the north. That being the case, sa'okha development would tend to be favoured in the north while lagoonal conditions would persist in the south. In other words, it appears that there was a semi—permanent embayment in the Section 2 region for that time interval.

Another example of 'along—strike' pinch out of sag'_ ha facies anhydrite is the loss of the C Member in the Trold Fiord area (Figure 32). It is present in the north (Sections 1 and la) but it is not developed in the south (Sections 2 and 3). Fortunately there are some exposures in the intervening area. Running north from Section 3, for instance, there is a linear ridge that affords continuous exposure of the Baumann Fiord Formation (Figure 51). At the northern end of the ridge, the three members of the formation are all distinctly evident. Towards the south, the recessive C Member becomes more and more ill—defined, anhydrite becomes less and less prevalent until at the southern extremity of the ridge (Section 3 itself), there is virtually no evaporite development at the C Member horizon; all the anhydrite having graded into pure limestone. It is thought that during C Member time there was a permanent lagoon in the southern region, induced and propagated as such by higher subsidence rates in that area. Noteworthy in this regard is that the Trold Fiord region may mark the ultimate southern terminus of C ?'ember evaporite development, for in the area further south, the C Member is absent, both in the Fllesmere Island extension of the belt and on Devon and Cornwallis Islands to the south and west (Kerr at al,, 1973 and pers. comm.). Figure 51. Panoramic view of linear ridge north of Section 3, encompassing a 180 degree arc north to south (The ridge, which is 14 miles long, here appears distorted due to the effect of 'fish-eye' parallactic perspctive imbalance). The Baumann Fiord C Member, clearly evident in the north (left), dies out toward the south by gradual facies transition from anhydrite to carbonate. 124

In summary, the Baumann Fiord cycles tend to he very widespread and persistent. If the carbonate components do pinch out laterally, it is either in an easterly direction towards the craton, or, in a north—south direction roughly parallel to the depositional strike. These relationships are thought to be explicable in terms of differential subsidence. It is thus concluded that the sabkha model is a viable one to account for the origin of the formation as a whole.

3.5 Paleoenvironmental Evolution of the Study Area

What follows is a reconstruction of how the Baumann Fiord evaporite complex evolved through time. Incorporated in the reconstruction are many of the ideas already set forth regarding the way the sabkha cycles were compounded and how differential subsidence may have. influenced the system. All of these relate ultimately to the environmental model outlined in Chapter 2. Also included is discussion of some of the features of the formation that have not as yet received specific comment.

The Copes Bay Formation, which regionally underlies the Baumann Fiord Formation, has been described by Kerr (1968) and is interpreted as ioeing of very shallow marine origin. The upper reaches of the formation have been reported to contain abundant limestone flat—pebble conglomerate and locally anhydrite beds are developed (Kerr, 1968). Clearly, near the end of Copes Bay time, depositional conditions in the study area were very similar to those which later developed in the Baumann Fiord lagoons. Certainly it is the author's experience that the uppermost carbonates of the Copes Bay Formation are virtually indistinguishable from many of the Baumann Fiord limestones.

The boundary between the Copes Bay Formation and the Baumann Fiord Formation is diachronous in the study area. Some of the lowermost Baumann Fiord anhydrite units in the region proximal to the craton demonstrally time—eouivalent to the uppermost limestone 'reds of the Copes BaL, Formation in the distal region. This relationship, schematically depicted in Figure 49, is 125

evident in the correlation section shown in 'fl6ure 50. The transition from Copes Say to Baumann Fiord development thus apparently proceeded as a series of regressive sabkha cycles progressively prograded further and further out into the depositional belt. The transition was completed when prograditional advance of sabkha system penetrated to the most distal portions of the region.

Accumulation of the Baumann Fiord A Member is thought to have taken place by compounding of successive sabkha cycles; each marine transgression marking the initiation of a new regressive cycle. Many of the transgressions penetrated to the proximal extremity of the sabkha, system. Others inundated only the distal reaches. Over the time interval involved, subsidence in the distal region clearly exceeded that in the proximal area for the A Member thins significantly from west to east (Figure 49). This subsidence differential is thought to have been the principal control on the extent of successive transgressions.

At Locality 7 there is a redbed sandstone horizon. at the top of the A Member (Figure 33). Trough crossbedding and cut—and—fill structures .are recognizable in the unit and locally, small channels can be distinguished. The sandstone comprises medium—grained cuartz with moderate sorting, well rounded grains of moderate to high spherioity, cemented by. calcite spar (Figure 52). In that the sandstone unit is laterally confined and is in close proximity to the stable Precambrian paleo—landmass, it is thought to be representative of foreland elastic fan deposition. The rounding and sorting suggest however that it may be of a secondary nature, perhaps originating by alluvial reworking of former dune sands. Presence of a non—marine elastic influence was also noted in the vicinity of Section 4, levels 420 to 465 (Figure 32).

The onset of 7 Member deposition is thought to have resulted as a function of increased subsidence over the whole of the stud-, area. Althougb the Member wid espread and relatively thic' it is clear that it was not deposited ul,der open marine conditions. Quite the contrary, the litholoio character of Figure 52a. Redbed quartz sandstone, viewed in plan (right) and cross—section (upper left); and red marl (lower left). Redbed facies in Baumann Fiord B Member and uppermost A Member beds, Section 7. Sandstone hand sample at right is 5 inches across.

Figure 52b. (quartz sands tone from sanple shown in Tigure 52a; Baumann Fiord. A Member, Locality 7. Cement is calcite. Monochrome photomicrograph. Crossed—nicols. Field .8 cm. across. 127

B Member strata is llormally virtually identical to tbi.t of the carbonate units in the A and C Members: ripple cross—laminated limo mudstone, massive lime mudstone, occasional flat—pebble conglomerates and, in the upper reaches, cryptalgalaminated lime mudstone and stromatolites (Figure 53). It is therefore concluded that the B Member is representative of long—standing lagoonal deposition, perpetuated as such by subsidence rates that did not allow sabkha development to gain a foothold, even in the proximal region. The fact that the B Member is characteristically devoid of fossils is seen as being supportive of this 'shallow—lagoonal' interpretation; elevated salinites induced by rapid evaporation under arid zone climatic conditions having had an inhibitory effect on faunal development. Noteworthy in connection with this cuestion of the influence of salinity on restricting faunas is the occurence at Locality 7 of numerous fossil—rich horizons in the B Member (brachipods, gastropods and arthropods — see Appendix C). At these horizons the limestones consist dominantly of biopelmicrite with intraclasts, pellets and skeletal grains (Figure 54). Invariably associated with the fossiliferous horizons are intercalations of red marl (Figare 52a) which are thought to be of non—marine origin. The conclusion drawn from this association is that during B Member time there was a fresh—water influence in the Section 7 region, periodically producing sufficient dilution of the concentrated lagoonal waters to permit development of a faunal community.

Transition from B Member limestone deposition to cyclical carbonate—anhydrite deposition in the C Member appears to have resulted when a series of progradational sabkha advances eventually brought the supratidal front to the most distal regions. Many of the sabkha cycles in the basal C Member in the east are thought to be time—equivalent to uppermost B Member limestones. in the west (Figure 49). Accumulation of the C Member is thought to have proceeded in a manner very similar to that which characterized the A Member: successive compounding of sabkha cycles.

The basal p`a-t of the 71eanor River Thrmation is made up of lime muconer, of a character very similar to the 7alimarIll Tliord carbonates. It is therefore concluded that the transition Figure 53. Baumann Fiord B Member limestone. Alternating massive and laminated—rippled beds of micrite. Locality 6. 129

Figure 54. Biopelmicrite from fossiliferous B Member beds at Locality 7; comprising brachiopod, gastropod and echinoderm skeletal debris, pellets and other allochems. Monochrome photomicrogaphy. Crossed—nicols. Field .8 cm. across. 130

to Eleanor River depositi,,n marks the -widespread restoration of lagoonal conditions in the study area. This restoration may have been somewhat diachronous as basal Eleanor River strain the distal region are apparently equivalent to upper C Kember sabkha cycles in the proximal part of the depositional belt (Figure 49).

In summary, it is evident that from Copes Bay to Eleanor River time, paleoenvironmental conditions were such that deposition was temporally and spatially partitioned- into two major categories: lime mud facies development and anhydrite facies development. This duality is thought to be reconcilable if the system is viewed in terms of the sabkha model, taking into account all the variability that the model allows and at the same time adhering to its prescribed restrictions.

3.6 Paleogeography

The foregoing reconstruction of the paleoenvironmental evolution of the Baumann Fiord system contains numerous implications regarding the paleogeography of the study area. These implications may be synthesized into a meaningful whole through the construction of paleogeographic maps. On the basis of all the control that is available and utilizing all the paleoenvironmental information that has been deduced, the following paleogeographic reconstructions are tendered.

During A ilember time, sabhhas were prevalent over the whole of the study area on Ellesmere Island. Figure 55 depicts an instance in.which the sabkha front has prograded to the distal reaches of the belt. Periods of maximum sabkha development such as this were interspersed with periods in which, pursuant upon marine transgression, lagoonal conditions prevailed over part or all of the area between the landmass and the open marine front. Some embay ents along the front of the sabhha are shown in an attempt to depict how some of the 'along strike' facies varielion may have arisen. At Locality 2, for instance, there were long periods in which lagoonal conditions prevailed despite there being major sr7,:ha facies anhydrite development in more distal 131

Figure 55. Paleogeography during Baumann Fiord A Member time. Sabkha plain here shown in position of near—maximum progradational advance. 132 133

Figure 56. Paleogeography during Baumann Fiord B Member time. Lagoonal conditions prevailed over the whole of the depositional belt. 3.34 135

Figure 57. Paleogeography during Baumann Fiord C Member time. Sabkha development less prevalent than during A Member time. 136 137

areas further north (Compare Section 2 levels 200 to 500 with equivalent levels in Sections 1, la and 3; Figure 32). The open marine front marked on the map delineates the inferred limit of the sabkha, system. The exact location of this front is not known but it is thought to parallel the regional depositional trend, as shown. Whether or not the open marine front was marked by a barrier complex similar to that developed in the Trucial Coast remains a matter of conjecture. The relevant rocks are not exposed on Ellesmere Island-..'

During B Member time (Figure 56), the essentials of the physiography remained similar to those developed previously save for the fact that lagoonal conditions prevailed over the whole of the platform. There may have been, as well, a number of foreland alluvial fan systems developed along the edge of the craton; systems similar to that delimited in the Section 7 region during latest A Member and B Member times.

C Member time saw the return of major sabkha development (Figure 57). In many regards, the paleogeography during this period was similar to that of A Member time. The sabkhas were of a more patchy nature, however, as evidenced by the fact that the C Member is not developed in certain areas. Clearly, after B Member time, conditions in some parts of the depositional belt were not conducive to sabkha development and these remained as shallow lagoonal regions. The area south of Section 1 and la appears to be in this category (Figure 57).

One of the remarkable features of the Baumann Fiord sabkha complex is its prodigious areal extent. On Ellesmere Island, the sabkha system, which is thought to have been some 70 miles wide, can be traced for well over 200 miles along strike. The reported occurences of Baumann Fiord evaporite development extend well beyond_the confines of the present study area. Kerr et al. (1973) and Kerr (pers. comm.) report Baumann Fiord anhydrite with associated limestone flat—pebble conglomerates and stromatolites in parts of Grinnell Peninsula, Devon Island (Figure 58). Thors- teinsson and Kerr (1968) describe thick Baumann Fiord anti drites with associated ]imestones in the central dome of Cornwallis 138

r.sri -Ns..„,.. ELLESMERE ISLAND

Miles 0 1.1 1001 O 160 Kilometres

C,3 Grinnell Pen.

OCornwallis Dome•

Figure 58. Projected regional extent of Baumann Fiord sabkha complex. Apart from the study area, evaporites of the Baumann Fiord Form- ation are known at: Sydkap Fiord, Ellesmere Island; Grinnell Peninsula, Devon Island; and Cornwallis Dome, Cornwallis Island (see text). 139

Island (Figure 54). If these occurences are in fact parts of one continuous belt, then the Baumann Fiord sabkh system extended for at least 700 miles along the edge of the Franklinian Geosyncline (Figure 58).

A very extensive sabkha evaporite complex along the southeast margin of the Franklinian sea is thus indicated for much of Early Ordovician time. The whole of the belt must have been characterized by stable tectonic conditions, allowing, with only very minor fluctuation, the steady and gradual subsidence that sabkha development requires. Climatically, the region must have lain within an arid torrid zone and in this regard it is interesting to note that a recent paleomagnetic reconstruction of the relative positions of the continents during the Ordovician places the Arctic Islands region in the centre of the equatorial zone (Smith et al., 1973). ':Tith regard to topography, the area must have been remarkable for its complete lack of relief. Figure 59 perhaps conveys an appropriate impression of what it looked like.

Figure 59. Trucial Coast sabkha plain, as viewed from the ground. Photograph courtesy of Dr. D.J. Shearman. 140

CHAPTER 4

ANHYDRITE - ASPECTS OF DIAOENESIS,

TECTONISM AND METAMORPHISM 141

4.1 Introduction

The anhydrite rocks of the Baumann Fiord Formation are of three distinct types. Firstly, there ara those that exhibit nodular mosaic structure. As mentioned previously, these are very rare. Secondly, there are those that are of bedded or laminar appearance. Anhydrite rocks of this type are peculiar to the proximal reaches of the Baumann Fiord outcrop belt on Ellesmere Island. Finally, there are those that exhibit massive structure. These are confined to the distal parts of the outcrop belt — Localities 1, la, 2 and 3. Of the three, the nodular mosaic form is thought to represent the primordial aspect of the anhydrite, the other two being secondary modifica- tions of it.

The present chapter is devoted principally to an examination of the nature and origin of each of the Baumann Fiord anhydrite rock types. Detailed descriptions of both the megascopic and petrographic character of each rock type are presented and each description is followed by interpretive discussion on how that particular form originated. Genetic affinities amongst the three basic classes are postulated and a model that is thought to account for the progressive secondary modification of the anhydrite is erected. In its essentials, the model comprises the following elements: early ddagenetic displacive growth of nodular anhydrite; compactional flow of anhydrite during initial burial stages, yielding laminar anhydrite; and dynamic burial metamorphism of the anhydrite inducing pervasive recrystallization with consequent transformation to massive anhydrite. Early stage compactional flow is thought to have affected the anhydrites of the whole of the study area. Subsequent metamorphic recrystallization of the distal anhydrites is thought to be related to their having undergone a minimum of 20,000 feet of burial. Proximal areas were never so deeply buried and thus the anhydrites of that region retained their lsminar character. In effect, then; documentation of the mode of origin of the various Baumann Fiord anhydrite rocks involves tracing their histories from the time of implacement in the sabkha right 142

through to The time at which they reached their maximum depth of

Subseouent to the discussion of the mode of origin of the various Baumann Fiord anhydrite rocks, there is an examination of the tectonic behaviour of the rocks in diverse structural settings. Also included are some petrofabric observations on anhydrites subjected to intense folding and on anhydrites exposed to fault-zone shear stress.

The latter part of the chapter deals with other noteworthy aspects of the secondary history of the Baumann Fiord anhydrite rocks: the origin of some replacement anhydrites is discussed; anhydrite fracture-fill -textures are noted; and chert concretions with length-slow chalcedony are examined.

4.2 Laminar Anhydrite

Displacive Growth of Nodular Anhydrite

The manner in which anhydrite nodules grow and displace their host has been fully documented by Shearman and Fuller (1969). It would be redundant to undertake herein a comprehensive detailing of the processes involved. Suffice to say that the fabric resultant upon displacive growth is normally that of decussate aggregates of platy anhydrite cleavage flakes, as pic- * Lured in Figure 29. This open 'felted' texture, in that it is thought to be representative of the primordial Baumann Fiord anhydrites (Figure 30) serves below as the starting point for discussion on the secondary textural evolution of the anhydrites.

Compaction of Nodular Anhydrite

Probably the first phenomenon to affect any nodular anhydrite after its formation is that of compaction. Because the internal framework of anhydrite nodules is so delicate and open and because their constituent cleavage laths are so readily subject

Divergent radiating fabrics are also known but these are less common (see Shearman, 1966). 143

to further breakage along any or all of their three cleavage planes, it is likely that compactional disruption of nodule fabrics takes place very early in the burial history of anhydrite sediments. Recent anhydrite nodules from the Trucial Coast are very moist and putty-like, easily deformed between one's fingers (Shearman, 1966). Barring very early cementation, it appears likely that the lath framework of nodules would. break down under the load imposed by deposition of a few feet of superimcum- bent sediment. nany ancient nodular anhydrites show evidence of having been affected by such load-induced crushing, breakage of the laths and collapse of the structure producing 'pile-of-bricks' fabric (equivalent to 'subfelted' or 'microcrystalline' texture, naiklem et al., 1969). But there are many ancient anhydrites in which 'pile-of-bricks' fabrics are not in evidence. In these examples, it is common to find evidence that liquefication and flow played a role in the early compaction history (Shearman and Fuller, 1969).

Whether anhydrite nodules compact by breakage or laths, producing 'pile-of-bricks' fabric, or by liquefication and flow, producing 'felted-aligned' fabric, appears to be controlled by the drainage characteristics of the host sediment (Shearman and Fuller, 1969). If the host is free-draining and there is nothing to support the open framework of the anhydrite during loading, the constituent laths fracture one another along cleavage planes and 'pile-of-bricks' fabric is generated (Figure 60a). If, on the other hand, drainage in the host is somehow impeded during loading, pore-water pressure builds up within the lath framework, thereby easing the frictional contact between laths. Slippage at the contact points induces temporary liquefication of the system and by virtue of their platy habit, the laths flow easily past one another to assume an aligned fabric (Figure 60b). Shearman and Fuller (1969) compare the process to the collapse of a house of cards.

On the basis of fabrics alone, the Baumann Piord anhydrites appear to have been affected by 'flow' compaction only. Felted- aligned texture Is c.'_eveloped in virtually all of the unrecrys- tallized anhydrite rocks examined. Pile-of-bricks fabric, A B HOST SEDIMENT FREE-DRAINING HOST SEDIMENT IMPERVIOUS

•

INCREASING P7*.#".- :\-----. COMPACTION 12 I

IN.WISMIOWSNO• Ming/NS • AP. .01•• flENNMISR.11ZINACOM astxxosmeesroo de=2/2MIMIIIIMMUlle 110610101M.M aseamagerszaresro• exersessmees .semonmos•remososa ••■•• wade.. Ma..44. 'MUM •411, 1440 kr. V I. al41.4 ■•••• • .00 ••••• ms•M IMINetiaNSIMNNO 4111•M ••■•■•••..... OI1WeqeaWz.otsles.sPA01 1OMNOM•■M•'""*""".10MMP • • • • • • . • . • • •

PILE OF BRICKS FABRIC ALIGNED FABRIC

Figure 60. Schematic representation of the effect of drainage parameters on early—stage compaction modification of anhydrite fabrics. With host free—draining, laths are unsupported and original fa'cric Breaks down into 'pile—of—Bricks' (A). With host sediment impervious, pore—water pre-:sure dcvlops in the interstices Between laths, inducing temporary quickening and flow (B). Diagram after -hcarman and Fuller, 1969; p.506. 145

diagnostic of 'crush' compaction is practically unknown in the Baumann Fiord rocks. A review of all the evidence pertaining to the compaction history of the Baumann Fiord anhydrites is undertaken below.

Baumann Fiord Laminar Anhydrites

It has been shown that compactional flow of nodular anhydrite can give rise to the sluggy and discoidal nodules that are so common in ancient anhydrite rocks (Shearman and Fuller, 1969). That extreme compactional flow yields delicately laminar anhydrite was considered a theoretical necessity by Shearman and Fuller (1969), a natural end—member to the compaction spectrum. But this hypothesis has not been_fully tested to date. The laminar anhydrites of the Baumann Fiord Formation afford an opportunity to so examine this supposition.

In the field, the laminar Baumann Fiord anhydrites can be seen to consist of delicate and uniform layers. The layering is in fact defined by minor differences in the proportion of accessory carbonate or organic matter in the anhydrite of individual laminae. White to cream coloured laminae are relatively free of impurities. Tan, brown and dark—brown laminae respectively reflect the presence of increased amounts of organic matter. Grey and grey—brown laminae commonly contain significant micrite or micro—dolomite impurities (to 30 percent in extreme cases). There is considerable variation in the thickness of individual layers, some being as much as two or three inches. Most laminae are on the order of 1/10 to 1/2 inch, however. One's first impression in the field is that this laminated material might be a sedimented deposit, perhaps deposited as a direct precipitate from a standing body of water, so unvarying and persistent are the laminae (Figure 61). Close inspection of individual laminae soon gives cause to seriously question this thought, however. Small pinch—and—swell structures can be observed (Figure 61); some laminae are wavy or show irregular undulations (Figure 61); scmsi laminae are extremely contorted, with small scale recumbent foNs anfflame structures (Figure 61). All of these features are suggestive of flow and it is difficult to 14r)

74 e • t Nutt;_ .0 7 •

_ # • ,...-r•- \ . •=1 q-)..;'- .1. -:- •_

Figure 61. Baumann Fiord laminar anhydrite (light tones) with intervening bed of limestone flat-pebble conglomerate (dark tones). Note pinch-and-swell structures (eg. left-centre, pen- clip level); wavy laminations (eg. left-centre above limestone)T and small scale overfolds and flame structures (eg. within darkish band at lower right). 14.7

envisage how any of them could have been produced_ by simple settling of anhydrite precipitate.

Petrographic examination of the laminar material gives the final lie to any thoughts that it may be indicative of a sedimented. origin (Figure 62). Firstly, the constituent grains of anhydrite, though generally exhibiting parallel to sub—horizontal alignment, do have many irregular undulations, a characteristic that is normally interpreted as being indicative of flow (Shearman, 1966), not settling of anhydrite precipitate. More decisive is the fact that the fabric is defined by aligned cleavage flakes of anhydrite, not discrete anhydrite crystals. One would expect anhydrite precipitate to crystallize in its lanceolate habit, elongate in the 0—crystal direction and flat parallel to 010 (Cuff, 1969). No crystals of this type have heen.found in any of the Baumann Fiord anhydrite rocks. The megascopic and microscopic evidence pertaining to the laminar. Baumann Fiord anhydrites thus indicates that: (a) it is unlikely that the rock was deposited as a direct precipitate from a standing:body - of brine, and (b) structures indicative of flow are common. Examination of the laminar material itself does not provide conclusive evidence of any direct genetic link between the laminar anhydrite and the presumed nodular precursor; save for the fact that the cleavage flake constituents of the laminar anhydrite are known to be generated as a natural function of displacive nodule growth and there is no other known mechanism to produce them. In this regard, the author was fortunate in coming upon some field exposures in which the genetic relationship between nodular and laminar anhydrite is strikingly evident. At Sand8la Creek, for instance, there is .a seven—foot interval (levels 558 to 565, Section 5) in which nodular anhydrite is beautifully exposed in the walls of a prominent buttress, proud from the scree slope. One end of the exposure is comprised of mosaic anhydrite, preserved in an almost completely undistorted state (Figure 63a). Tracing the horizon laterally, one finds that the nodules become elongated, flattened and otherwise distorted (Figure 63b). Further on, the nodules become so distorted that Figure 62. Aligned fabric of flowed anhydrite cleavage flakes, here shown in 45 position of maximum illumination. Colour photomicrograph. Crossed—nicols. Field .7 mm. across.

lA9

Figure 63. Transition continuum between nodular anhydrite and laminar anhydrite, as manifest by lateral change within a single level at Section 5. Mosaic Anhydrite --•• Distorted Mosaic Anhydrite.. Laminar Anhydrite.

Figure 63a, Mosaic Anhydrite (with slight distortion). 150

-1014, m*.1040, 44

a v4

1016le 7 kilt Ilk

4111"121099.h...twale±A

aditorik

Figure 63b. Distorted Mosaic anhydrite. White material is surficial gypsum.

Figure 63c. Laminar Anhydrite. Some surficial gypsum present but dark band at hammer base is anhydrite. 151

they lose their individual outline a,,d at the far end of the exposure, the anhydrite is of a bedded or laminar character (Figure 63c). In the space of some 35 feet, the anhydrite grades laterally from distinctly nodular to distinctly laminar structure (Mosaic ----•- Distorted Mosaic Bedded Massive in the classification scheme of Mailem et al., 1969). Thus, there can be little doubt that, as a rule, the laminar Baumann Fiord anhydrites originated through flow of pre—existing nodular anhydrite.

Compactional flow to produce laminar anhydrite requires fulfilment to two essential requisites: (a) the host must be itself susceptible to extreme compaction, and (b) the host must be relative impervious (Shearman and Fuller, 1969). With regard to the former; the matrix in which the Baumann Fiord anhydrites grew, in that it consisted dominantly of mud—size carbonate and organic matter, would appear to be highly compactible in its own right. Furthermore, because matrix material often constituted less than 10 percent of the bulk volume ofthe sediment, it is to be expected that the matrix would be easily carried along with the anhydrite during flow. Regarding host perlability; the Baumann Fiord system would appear to offer ideal 'self— sealing' properties. Both mud—size carbonate sediments and organic matter have high entry—pressure characteristics and for the small pore—water pressure enhansement required, they would seem to be capable of providing a suitable seal.

A number of questions regarding the mechanics of flow warrant further examination. Why is it that, in the laminar rocks, the impurities are so neatly segregated, some laminae being relatively rich in carbonate or organic matter, others relatively poor? Two answers seem possible but the author has been unable to find decisive evidence in support of either. Firstly, it is plausible that distinct laminae arise through flow—merging of a pre—existing band of like nodules, a row of, * Mosaic anhydrite preserved as such characteristically contains < 10% matrix; and the flowed anhydrite although sometimes containing as much as 30% matrix minerals, also normally comprises < 10% non—sulphate. 152

say, organic—rich nodules undergoing flow to produce an organic— rich lamina. Alternately, it is conceivable that, given a more heterogeneous matrix to start with, there could be a density segregation of the non—sulphate components during mass flow of that selected zone. Anhydrite is slightly more dense than dolomite and calcite and is very much denser• than organic matter.

Another important question is: At what stage in the burial history does flow of anhydrite characteristically occur? Some arguments favouring flow under only a few feet of overburden were presented earlier. Evidence from the Trucial Coast anhydrites tends to support this contention. Numerous workers have reported the occurence of soft and pliable anhydrite layers within the supratidal facies (eg. Shearman, 1966; Kinsman, 1966, 1969), the so—called scream—cheese' anhydrite (Park, 1969). Although it has not proved possible to sample this anhydrite without distur- bing the fabrics, it is possible to occasionally discern flow— lines within the anhydrite mass and locally, small diapirs are developed (Shearman, 1966; and pers. comm.). If flow can take place only one or two feet below the surface of the Trucial Coast sabkha, with its relatively high proportion of permeable calcarenite host, then flow under a few feet of overburden in the Baumann Fiord setting would appear to be intrinsically viable, considering the relatively impermeable nature of its host materials. To summarize the evidence concerning the origin of the laminar Baumann Fiord anhydrites, the following reconstruction is proposed: Nodular anhydrite grew within the supratidal sediments of the Baumann Fiord sabkha plain. These nodules were comprised of decussate aggregates of platy cleavage flakes (felted texture). The nodules probably persisted in their primordial form for only a short period of time, just until a few feet of superincumbant sediment had accumulated. By that time, the seal provided by the matrix material had impeded drainage to the extent that anomoloua pore—water pressures were generated within the framework of laths. Slippage of laths and collapse of the framework induced quickening of the sediment and extensive flow ensued, yielding laminar anhydrite with felted—aligned texture. 153

Compactonal flow of pre-existing nodular snhys.rite is thought to have affected virtually all of the Baumann Floret sabkha-facies deposits. Survival af nodular anhydrite was certainly the exception rather than the rule. Drainage conditions in this ancient sabkha system must have been inherently conducive to early-stage flow. Of course, the drainage characteristics of the Baumann Fiord system were themselves a function of the rather specialized high-mud, high-organic nature of the depositional environment. The causal relationship between genesis and early stage flow diagenesis is thus seen as a direct one.

Aspects of Petrography

During the compactional flow process, there is a short period of time when overburden. is supported by the water only, the framework of laths having given way. This condition persists only until the water.finds an escape, at which time there is physical movement of water in the escape direction. It is to be expected that laths caught up in this flow will align themselves according to the hydrodynamic dictates of the system. In the field, it is often difficult to assess the direction of escape flow, especially in the laminar rocks where reliable vector structures are often rare or absent. But where nodules have been distorted only to the extent where they are flattened and 'pulled-out', as have those in Figure 63b, it is clear that escape flow paralleled the axis of nodule elongation. Thin- sections cut at various attitudes relative to this elongation vector reveal a number of interesting things about how felted- aligned texture is generated and how compactional flow operates.

Pictured in Figure 64 is a. hand sample of 'pulled-out' anhydrite nodules from the Baumann Fiord Formation (Section 5). The sample has been slabbed such that one face lies perpendicular to the axis of nodule elongation, the other face parallel to it. Besides being elongated, the nodules are of course somewhat flattened in the plane of the sample. Thin sections cut from this sample reveal that the nodules were deformed by flow within certain wispy zones only, the laths in the other parts of the rock having been spared from flew and having thereby retained Figure 64. Photograph of a block of Baumann Fiord nodular anhydrite. The slabbed surface on the left is parallel to the nodule elongation axis. The surface on the right is normal to the axis of elongation. Block is lz inches thick.

Nodules Oblate

Anhydrite Nodules Elongate

Pi.eferred Orientation of Cleavage Flakes 155

their felted fabric. That flow was localized, not all-pervading, indicates that the paleo-drainage in this particular rock was relatively free, small pore-pressure anomalies giving rise to lath-slippage at scattered points only.

The zones that have 'undergone flow (i.e. those with felted- aligned texture) exhibit a number of striking optical features when viewed in thin-section under crossed-nicols. In order to understand the significance of these features it is necessary to first review the crystallographic and optical properties of anhydrite.

For an orthorhombic mineral such as anhydrite, there are six possible ways of assigning the crystal axes, three of which have been widely employed in the literature on anhydrite. That of Deer, Howie and Zussman (1962, pp. 219-225) is employed here (Figure 66) because their choice of unit cell orientation permits direct comparison between the anhydrite structure and those of the hydrates of calcium sulphate (hemihydrate, gypsum etc.). In dealing with anhydrite of nodular origin, it is the cleavage characteristics of the mineral that are of critical importance. According to the determined cleavage fissility ratings of anhydrite in its three pinacoidal planes, cleavage flakes of the mineral should tend to assume the habit illustrated in Figure 66: platy in the plane of the best cleavage (010), and elongate in the direction of intersection of the best cleavage (010) and the next best cleavage (100), i.e. elongate in the c-crystal direction. Many of the optical peculiarities of flowed anhydrite rocks can be explained in terms of the shape and orientation of the constituent cleavage flakes, as outlined below: - Aligned anhydrite cleavage flakes, such as those pictured in Figure 65, immediately exhibit an obvious optical fabric when viewed under crossed-nicols. Rotating the stage, one finds that all the cleavage flakes go into extinction simultaneously. This is simply a function of the fact that anhydrite, being orthorhombic

*eg. a = Y b.X c.Z(Deer, Howie and Zussman, 1962) a .X b.Y c.Z(Shearman and Fuller, 1969; Cuff, 1969) a .Z c.X(the morphological designation - Dana, 1932) Figure 65a. Anhydrite fabric in thin-section normal to nodule elongation axis. Note anomolously low birefringence of most of the aligned anhydrite laths indicating acute bisectrix- centred preference (c-crystal axis normal to plane of thin- 0 section). Colour photomicrograph. Crossed-nicols. 45 position. Field .9 mm. across.

4' re 65b. Anhydril;e fabric in thin-section parallel to nodule elongF,ticn axis (still normal to bedding). Note normal birefringence of aligned anpydrite laths Colour photomicrograph. Crossed-nicols. 450 position. Field .6 mm. across. ANHYDRITE Orthorhombic (+) c = Z

MORPHOLOGICAL SETTING CLEAVAGE FLAKE

4)( =1.571(X) 010 Perfect (3 =1.576(Y) 100 Very Good 7 =1.614(Z) 001 Good

1Pigure (6. Grys'Gallographic and optical orientation of anhydrite. Given the illustrated elcr,vage fissi_lity ratings, cleavage flakes of the mineral tend to be platy in the A—C plane and clon6ate 7.:arallel to the c—crystal direction. 158

and having pinacoidal cleavages, characteristically exhibits straight extinction. If the cleavages of all the flakes are parallel, so they will all extinguish together. — Use of a quartz wedge reveals that the cleavage flakes, within the aligned zone all show positive elongation (length slow). This observation simply confirms that the flakes are platy in the 010 plane, normal to the X vibration at axis (assuming, of course, that the thin—section is cut normal to the plane of alignment, as are those in Figure 65). No matter how such a plate is cut, the X direction will be normal to the elongation of the flake. — The ultimate key to the orientation of the cleavage in the aligned zones stems from the following observations: In sections cut parallel to the elongation axis (but still normal to the alignment plane: Figure 65b) the flakes characteristically show near—normal birefringence (.030 to .045). But in sections cut perpendicular to the nodule elongation axis, there is a high proportion of flakes that show anomolously low birefringence (.005 to .014) — Figure 65a. Point counts on the flowed zones in this elongation—normal section (Figure 65a) ih fact reveal that 71 percent of the flakes exhibit birefringence of less than .014 (first—order yellow). Clearly, these flakes are preferentially oriented such that the Z vibration axis lies normal to the plane of the thin—section, parallel to the axis of nodule elongation. As such, it is apparent that the flakes in Figure 65a are bounded top and bottom by 010, sides by 100, with the 001 plane facing. This unique solution to the optical orientation of the cleavage flakes allows two important conclusions to be drawn. First, it appears that cleavage flakes generated by growth of nodular anhydrite do in fact tend to be platy in the 010 plane and elongate in the c—crystal direction, as one would expect from

In Z—centred view, birefringence is (3-0‹ = .005 (first—order grey). Because of the small size of the flakes, acute bisectrix- centred figures proved difficult to obtain and confirmation of the orientation solution was thereby precluded. 159

the cleavage ratings (Figure 66). Second, it is apparent that during flow, anhydrite cleavage flakes tend to align themselves such that the long axes of the plates (c—crystal direction) parallel the escape—flow vector, as depicted in Figure 64.

4.3 Massive Anhydrite

In the proximal part of the Baumann Fiord depositional belt (sections 4 through 7 inclusive), laminar anhydrites with felted—aligned texture have been preserved as such, their having been spared from any form of major post—compactional textural transformation. However, in the distal region (Sections 1 through 3 inclusive), there is abundant evidence that the anhydrites did undergo further textural modification, over and above that pursuant upon early compactional flow. This is immediately evident in the field, for the distal anhydrites are of massive aspect (Figure 67); medium to finely crystalline, light to dark bluish grey in hand speciment. Close inspection reveals, hoWever, that the anhydrite is not absolutely homogeneous but is in fact thinly bedded or laminated, the layering being defined by subtle colour differences between adjacent bands.

In thin—section (Figure 68), the anhydrite can be seen to consist of blocky subhedral crystals, characteristically elongate in outline with diameter/thickness ratios of about 2.5/1. The size of individual crystals generally varies between 40 and 100 microns but some crystals are as large as 300 microns. Twinning (sometimes repeated) is observable in certain crystals (twinned on 011). The crystals are aligned parallel to bedding and, on the whole, these anhydrites can be said to have a weak to moderate dimensional fabric. This dimensional preference is reflected in a moderate optical fabric, with the majority of crystals extin- guishing parallel to the dimensional foliation (62 percent of the crystals have extinction parallel to bedding t 10 degrees). The layering in these massive anhydrites is due to differences in the accessory compositions of individual bands. Some laminae are rich in calcite or dolomite (to 30 percent), others poor.

1,- 4 .410* tc --,..,. -,...... - ---.%...... - -■■••''' .?1' - • .Aishi:r. ..-00-- _:..,,... • ltIP----- _— - - -

•

. _...... „ --.■.....t --- - als&i. vs,.:sidassist . ii!kr

Figure 67 . Massive anhydrite (dark tones). Irregular. chert concretions visible to the left of the pen. White surface in lower right quadrant of photograph is gypsum crust. Hairline alteration gypsum evident elsewhere. Figure 68. Massive anhydrite, showing blocky crystal habit and moderate to weak foliation. Note occasional twinned crystals. Colour photomicrograph. Crossed—nicols. Field 1.8 mm. across. 162

Figure 69. Massive anhydrite with foliation. The proto—lamination, here defined by sloping bands of carbonate—rich anhydrite, lies oblique to the gravity—normal foliation. Colour photomicrograph. Crossed—nicols. Field .8 cm. across. 163

This compositional differt:Intiation is petrographically reinforced by distinct textural differentials amongst the laminae (Figur& 69). In bands that are relatively rich in accessory carbonate, the anhydrite is finely crystalline; bands with lesser amounts of matrix material are more coarsely crystalline in their anhydrite component; pure anhydrite, such as that pictured in Figure 68, is the most coarsely crystalline end-member.

Thus, in thin-section, as well as on polished slabs, it is possible to trace individual laminae, because each band is distinctive with regard to accessory composition and mean anhydrite crystal size. These laminae are usually co-planar and the preferred orientation of the component crystals is parallel to bedding. In a number of instances, however, the laminae are contorted, with small-scale overfolds and flame structures. Pinch-and-swell structures are also occasionally evident. In fact, the internal structures of the massive anhydrites are comparable in every essential regard to the 'compactional-flow' structures described earlier. In instances where the proto- lamination is thus contorted, as is the case for instance in the rock pictured in Figure 69, the plane of flattening of the blocky anhydrite is oblique to the disturbed laminae, parallel to the bedding of the unit as a whole. In other words, the foliation is pervasive and through-going, always parallel to overall bedding, notwithstanding small-scale contortions.

The foregoing observations allow some conclusions to be drawn regarding the origin of the massive Baumann Fioidanhydrites. Firstly, there can be little doubt that the massive anhydrites stem from a laminar precursor (flowed anhydrite), for the pre- existing laminae are immaculately preserved in relict form (Figure 69). The only major difference between the laminar 'flow' anhydrites of the proximal region and the massive anhydrites of the distal region is that the textural character of the latter has been transmogrified the extent that the previous 'platy' habit of the constituent anhydrite has been totally .convertee, to a blocky habit. Secondly, it is clear that the foliation in the massive anhydrites was not directly inherited 164

from the pre—existing laminar anhydrites, for in the cases where the proto—laminae were contorted the present foliation cuts the inferred direction of 'flow' preference. Thus it is concluded that, although the present foliation characteristically parallels bedding (as did that of its flow—laminar precursor) it was not strictly a case of 'fabric inheritance'.

What was the nature of the textural transformation from platy to blocky anhydrite; what was the physio—chemical environment in which the transformation occured; and when did the transformation take place? Precedents are lacking. Of the metamorphic anhydrite textures described by Borchert and Muir (1964), none is comparable to that of the massive Baumann Fiord anhydrite. Hoen (1964), and Schwerdtner and Clark (1967), described textural changes in anhydrites involved in major diapiric activity but none of the resultant fabrics resembles that under consideration here. Goldman (1952) set forth a great deal of textural detail relating to the anhydrites of salt dome cap rock. He showed how pervasive recrystallization can bring about dramatic textural changes in sedimentary anhydrites. The recrystallization processes he outlined were shear—induced and the fabrics thereby generated revealed a history of graulation, crushing and mylonitization. Evidence of any form of shearing or other stress—related textural transformation is lacking in the massive Baumann Fiord anhydrites.

It is nevertheless considered worthwhile to undertake a reconstruction of how the textural modification of the Baumann Fiord anhydrites took place. This may be done through the application of some basic principles of metamorphic transformation. In monominerallic aggregates, thermally activated changes in atomic arrangement are well known. These give rise to what is normally referred to in metamorphic petrology as simply 'recrystallization' (Spry, 1969). Recrystallization during the early (pre—tectonic) stages of metamorphism often produces

*'Recrystallization' is here used in the defined by Spry (1969): reconstitution of an existing mineral phase. Recrystallization may proceed by either 'grain growth' or 'coalescence'. 165

preferred orientction in monominerallic rocks because -the stress field is uniaxial ( er vertical :gravity) and the rate of grain growth in the plane normal to CT' exceeds the rate in the OF direction (Spry, 1969, pp. 212-225). It is this type of thermally activated recrystallization that is envisaged for the Baumann Fiord anhydrites. The rationale behind the postuation of such a mechanism of textural transformation is as follows:— The foliation invariably parallels bedding thus suggesting (a) that the stress indicatrix during recrystallization was uniaxial, with Cri5.1 gravity, and (b) that the recrystallization was pre—tectonic, induced before the beds were tilted relative to the gravity— induced principal stress. The mean crystal size of the blocky anhydrites is considerably greater than that of its laminar precursor suggesting that grain growth was the dominant recrystallization agent. Development of primary twins commonly accompanies recrystallization (Spry, 1969; pp. 67-72) and the twinned anhydrite crystals of the Baumann Fiorirocks are thus to be expected.

The massive Baumann Fiord anhydrites are restricted to the most distal reaches of the miogeosyncline (Troll Fiord region). Conservative estimates of the thickness of Pranklinian overburden in the area run to 20,000 feet or more (Appendix Assuming 'normal' geothermal gradients, the temperature at 20,000 feet should be on the order of 24000 (Levorsen, 1954; P. 415). That thermally activated recrystallization can take place in evaporites at such comparatively low temperatures has been established (Borchert and Muir, 1964). Thus it is thought that recrystallization to produce massive anhydrite was a matter of pre—Ellesmerian burial metamorphism. More proximal anhydrites were not recrystallized because they were never sufficiently deeply buried.

4.4 Influence of Tectonism on Anhydrite

Further evidence that anhydrite recrystallization in the Troll Fiord region was thermally activated rather than stress induced stems from the observation that in more proximal parts 166

of the deformed belt (areas wherT, the rocks were not as deeply buried), the anhydrite is not recrystallized to any signigicanL degree, despite being deformed under what, at certain localities, must have been very high tectonic stress conditions. At Locality 8, for example, a series of tight chevron folds are exposed (Figure 70). These structures die out laterally and it is clear that the prodigious deformation in this zone is Strictly local- * ized, for the deformation in the region as a whole is much more subdued. Even though the anhydrites have been subjected to intense folding in this locality they have not recrystallized. The discussion that follows relates to some of the textural changes that have in fact takenyplace in response to folding and it attempts to throw some light on how the deformation was achieved.

The folds are exposed in a vertical rock face that is presently being actively undercut by a stream (Figure 70). With continuous rapid erosion, the exposed anhydrite rocks are very fresh, erosion rate greatly exceeding rate of gypsification. The folds are of amplitude 100 feet, and wave—length 40 feet, on average. In the field, it is immediately evident that the competence contrast between the anhydrites and the interbedded limestones was great indeed. The anhydrites show every sign of having behaved in an extremely ductile manner during folding. Minor s—folds and z—folds are very common in the limbs of the structures (Figure 71). In one place, a syncline hinge has been breached and anhydrite injected into the core, forming a small— scale fan fold with minor m—folds (Figures 70 and 71); a striking indication of the plasticity of the anhydrite. The limestones, on the other hand, behaved in a considerably more competent manner, deforming principally by brittle failure rather than by plastic flow. Radial tension fractures are common in the hinge zones of the folds. In the limbs of the folds, boudinage structure is often developed, anhydrite bridging the necks

The deformation is thought to be Ellesmerian in age but, with detailed knowledge of the tectonic history of the region only just now emerging (Trettin et al., 1972), it is not possible to confirm this as yet. Figure 70. Chevron folds in Baumann Fiord anhydrite and carbonate rocks; Locality 8. Note the small fan fold in the hinge of the central syncline. Cliff exposure is 80 feet high.

;;;iiit. “

Figure 71. Detail of minor folds in limb of Locality 8 chevron fold. Left half of photograph constitutes fan fold injection structure in the hinge of the central syncline (Figure 70). Figure 72. Limb of chevron fold; Locality 8. Doadinage stromatolitic limestone in centre (the lowermost boudin lying right of the hammer) with laminar anhydrite either side and necking the boudins. 169

(Figure 72).

That the anhydrites deformed by plastic flow is further evidenced by their petrographic character (Figure 73). The anhydrites are made up almost entirely of platy cleavage flakes, immaculately aligned parallel to the external lamination. Where the laminae are contorted in small scale overfolds or flexures, the cleavage flakes are tangentially aligned along the contours of the structures, giving sweeping extinction under crossed- nicols. Very small, regular undulations (on the order of a few millimeters) give the impression of kink bands (Figure 73). There is no evidence of recrystallization in any of the material examined.

In many regards, these deformed anhydrite rocks closely resemble the laminar 'flow' anhydrites, in their megascopic as well as their microscopic characteristics. The cleavage flakes are identical in size and habit to those of the laminar 'flow' anhydrites. Individual laminae are defined according to their accessory composition just as is the case in the undeformed 'flow' anhydrites (see section '4.2). In general aspect, the flow textures in the Locality 8 anhydrites are not at all dissimilar to the flow textures of the laminar anhydrites. It is thought therefore, that the Locality 8 anhydrites deformed under conditions of enhansed pore—water pressure, triggered tectonically. The form and configuration of the flow laminae would have been controlled by compressional rather than gravi- tational stresses but otherwise, the physical conditions under which the flow took place would have been similar to those that operated in the compactional flow regime. If compactional flow is viable mechanically, then the evidence at Locality 8 indicates that remobilization of the same rocks by tectonically induced pore—pressure build up is viable as well.

As a rule, the structural style in the deformed part of the Franklinian Miogeosyncline is very subdued. Broad open folds with wave—lengths of 2-3 miles, amplitudes of 2,000-3,000 feet Figure 73. Aligned fabric in anhydrite from Locality 8 chevron folds. 'Kink banding' here highlighted by virtue of having the cleavage flakes or one element of each kink aligned parallel to the upper nicol (i.e. extinction position). Colour photomicrograph. Crossed—nicols. Field 4 mm. across. 171

are the norm. This order of folding is not consistent with the development of any form of stress—related foliation, even in anhydrite rocks (N.J. Price, pers. comm.). Indeed, the evidence from the Locality 8 anhydrites suggests that much higher orders of folding do not necessarily give rise to stress—related foliation (i.e. the equivalent of slaty cleavage). Thus it is not surprising that the Ellesmerian Orogeny passed without leaving any regionally significant imprint on the anhydrites. In the Trold Fiord region, the massive anhydrites underwent Ellesmerian folding without detectable textural modification or recrystallization. In other, more proximal parts of the deformed belt, the laminar anhydrites of the Baumann Fiord Formation survived the folding with little or no change in gross aspect of petrographic character (restricted zones such as that at Locality 8, are exceptions to this general rule). In the stable platform region, there was, of course, no Ellesmerian influence and the laminar anhydrites have persisted unchanged to the present day.

Tectonic overprinting on the Franklinian rocks by the Tertiary Eurekan Orogeny was limited to low—angle thrust faulting (Thorsteinsson and Tozer, 1970; Trettin et al., 1972). These thrust faults, which normally strike parallel to the Ellesmerian structures, preferentially follow the Ordovician evaporites (both the Bay Fiord and Baumann Fiord Formations). But in the Baumann Fiord sections examined in this study, the influence of thrusting on the anhydrites proved very difficult to assess, primarily because, under Arctic weathering conditions, intensely sheared anhydrite alters to gypsum much more readily than does undeformed anhydrite. Where mappable thrust faults (or associated splay faults) were encountered, as for instance at level 1360 near the base of Section 1 (Figure 32), the sheared zone .was found to be so extensively gypsified as to preclude the procuring of fresh anhydrite samples.

The thrust zones tend to be strictly confined. Rarely does the affected level exceed 10 feet in thickness and the vertically adjaoent anhydrite strata invariably appear virtually unaffected by the fault movement. Near the edge of some shear zones it is 172

Figure 74. Tectonized Baumann Fiord anhydrite rock at the edge of a major shear zone; Locality 1. The anhydrite is in fact invariably altered to secondary gypsum in such instances, meteoric water having brought about hydration of at least the outermost 1-2 feet of the exposures. 173

possible to find highly .)ontorted. material (Figure 74), not itself sheared in the strict sense but obviously affected by the associated shear translation. Such contorted strata was also found to be completely gypsified. Even in instances where shearing was on the smallest of scales (one or two inches of horizontal translation), the affected rock is totally gypsified. Clearly, the assessment of anhydrite textural changes pursuant upon shearing must await the time when borehole cores of the relevant material are made available. In the meantime, it is nevertheless apparent that the Baumann Fiord anhydrites have a very low shear yield threshold and it is for this reason that the formation served as a major decollment during deformation.

4.5 Other Aspects of the Baumann Fiord Formation's Secondar History

This chapter has dealt with a significant portion of the secondary history of the Baumann Fiord Formation: early—stage compactional flow of the anhydrites (manifest throughout the study area); burial metamorphism—recrystallization of the anhydrites (restricted to the most deeply buried parts of the belt — notably the Trold Fiord region); and the effects of tectonic deformation (reflowage of anhydrites caught up in high— order folding and shear of anhydrites in thrust zones). But other more specialized secondary changes in the Baumann Fiord Formation warrant documentation as well and the present context is deemed an appropriate one in which to discuss them.

Replacement Anhydrite Replacement anhydrite (anhydrite after calcite or aragonite) is occasionally found in limestones associated with ancient x evaporite sequences (see for example Fuller, 1956; Murray, 1964).

Figure 77 shows a slightly rotated chert concretion, offset as a result of inferred shearing in the immediately adjacent rocks. The photo was taken in rocks only a feY feet above a significant. shear zone, and this small—scale shear girdle is thought to be a splay arising from it. 174

Its occiArence in some of the Baumann Fiord lime mud tones is reported here as a matter of documentary interest, not because the Ellesmere Island examples throw any new light on how the replacement takes place.

The Baumann Fiord replacement anhydrites are observable in only about 5 percent of the thin—sections of lime mudstone examined in the course of this study. The replacement crystals are usually very small ( 1 mm.) and are normally clustered along zones of preferred permeability (eg. small fractures or silt—rich laminae). In a few instances, the crystals reach inch in size (Figure 75). That they are replacive in origin is evidenced by the fact that they are invariably riddled with floating carbonate relics (Figure 75). The replacement crystals tend to be crystalloblastic but their boundaries are often castillated. In many of the Baumann Fiord examples, the rims of the crystals have been gypsified (late—stage hydration by meteoric water — Figure 75). Replacement of calcite by anhydrite involves a volume increase of about 25 percent (35 percent increase for aragonite --.anhydrite). Because fractures are often associated with replacement anhydrites (Murray, 1964; Shearman, pers. comm.), its is thought that the replacement normally takes place in a very early diagenetic setting, the mechanical force of crystallization producing fissures in the partially indurated host carbonate (Shearman, in prep.).

Anhydrite Fracture—Fill

Many of the limestone bands in the Baumann Fiord Formation have responded to deformation by undergoing brittle failure. The resultant fractures are normally filled with sparry calcite 36). But in one such fracture, the vein—fill mate ial was found to consist of anhydrite, with individual crysta up to 1 inches across. In thin—section (Figure 76), the fracture—fill anhydrite crystals can be seen to have well developed twins. In many crystals there is only one set of twin lamellae (011), but others have conjugate twin sets (the second set, rarely 175

Figure 75. Replacement AnhyLrite (anhydrite after calcite). Note floating relics of unreplaced dolomite and calcite. Halo is that due to alabastrine gypsification by meteoric water. Much of the halo gypsum was removed in the manufacture of the thin— section. Colour photomicrograph. Crossed—nicols. Field 4 mm. across. Figure 76. Fracture-fill anhydrite. Large crystals of twinned anhydrite with alabastrine gypsum along cleavage planz?s. (where meteoric water was able to find access). Note the displacement of part of one crystal occasioned by the mechanical force of crystallization of the alabastrine gypsum. Colour photomicrograph. Croosed-nicols. Field 1 cm. across. Figure 77. Chert concretions in massive anhydrite. Note clockwise rotation of concretion at left. Mesh of light—toned cracks is contained within surficial veneer of alabastrine gypsum. 178

developed, is on 120). This type of conjugate twinning is normally attributed to deformation (Spry, 1964; pp. 67-85) and there is no reason 1..o believe that this is not the case here.

Chert Concretions

Chert concretions are common in both the anhydrites and the limestones of the Baumann Fiord Formation (Figure 77). They occur isolated, in clusters (Figure 67), or in discrete bands (particularly prominent bands are logged as such in Figures 32 and 33). As a rule, the concretions are spherical to oblate in shape, and in the latter cases, they are normally oriented with their maximum dimension parallel to bedding. In size, they vary from less than an inch in diameter to more than six inches. The chert is characteristically dark grey to black.

In thin—section, individual chalcedony crystals are not easily resolved because of their fine crystal size, but in the instances where it did prove possible to perform optical tests, it was found that the chalcedony needles invariably show positive elongation. Chalcedony, of course, is normally length—fast but in cherts associated with evaporites, positive elongation appears to be the norm (Michel—Levy and Munier—Chalmas, 1892; Cayeux, 1916; Folk and Pitman, 1971). The occasional clusters of radially arranged leutecite that are in evidence in the anhydrites, are also length—slow (radially slow).

Spherules of leutecite in the Baumann Fiord rocks are of the type originally described by Cayvax (1916, 19'29). 179

CHAPTER 5

SECONDARY GYPSUM 1F.X.)

5.1 Introduction

The previous chapter has dealt with aspects of the burial history of the Baumann Fiord Formation, with consideration also given to tectonic effects. In this chapter, the chronicalling of the secondary history of the formation is continued by discussion relating to aspects of the exhumation history.

The Baumann Fiord Formation has had a rather specialized exhumation history, specialized in the sense that, since the onset of the Pleistocene—Holocene permafrost regime, water that would normally have been available for the gypsification of the anhydrite has been locked up as ice. In consequence, much of the anhydrite has escaped pervasive secondary alteration to gypsum and is preserved at outcrop in a largely unaltered state. However, there is a significant amount of secondary gypsum present in the Baumann Fiord Formation and it is with the nature and origin of this gypsum that the present chapter is primarily concerned.

There are two distinct categories into which. the Baumann Fiord secondary gypsum rocks may be divided. The first consists principally of replacive porphyroblastic gypsum and is found in association with satin—spar veins. This type of secondary gypsum is developed only in certain parts of the Baumann Fiord outcrop, and where it is in evidence, it is invariably confined to the basal parts of the formation (Figure 33). The second type is that which consists largely of replacive alabastrine gypsum. This development is confined to the outermost surfaces of the present—day outcrops and it is in evidence in all of the sections examined, at virtually every level in the formation (save, of course, in the carbonate bands). The first type is thought to have arisen in response to pre—Pleistocene hydraulic fracturing of the basal parts of the Baumann Fiord Formation, by water moving up into the evaporites from the underlying aquifer. This promoted in situ gypsification of the anhyOrite with concomitant precipitation of gypsum in the fractures, forming satin—spar veins. The second type appears to arise solely as a functicn of present—day weathering, by meteoric waters penetrating the 181

summer thaw zone and perpetrating a measure of r,urficial hydration. Much of the discussion that follows is devoted to the documentation of these postulates for the origin of the two different secondary gypsum types.

5.2 Background Considerations

Before delving into descriptive and interpretive detail regarding the Baumann Fiord secondary gypsum rocks, it is useful to review some of the basic principles of secondary gypsum genesis. Secondary gypsum rock may be defined as "any gypsum . rock that was derived either directly or indirectly by hydration of pre-existing anhydrite rock" (Mossop and Shearman, 1973). Thus, in considering the mode of occurence or origin of any secondary gypsum deposit, it is important to keep in mind that there are inescapable physical controls on the anhydrite-gypsum system. Of these, there are three which are considered of critical importance: (a) thermo-chemical stability relationships in the anhydrite-gypsum system; (b) volume increase on hydration, anhydrite —4-gypsum; and (c) water availability and access modes.

Anhydrite-Gypsum Thermo-chemical Stability Relationships

Theoretical and experimental work on the phase relationships in the anhydrite-gypsum system has been comprehensively reviewed by Hardie (1967). Some general principles pertaining to the geological implications of these phase relationships have been set forth by Mossop and Shearman (1973).

At depth, anhydrite is the stable form of calcium sulphate. But on exhumation, steadily decreasing temperature eventually renders anhydrite unstable and gypsum immediately assumes thermodynamic preference. Posnjak (1938) determined the solubility curves of gypsum and anhydrite in distilled water and concluded that because anhydrite is less soluble than gypsum above 42°C, anhydrite is favoured in that field. Posnjak's figure of 42°C has since gained widespread acceptance, having been confirmed by subsequent workers (see MacDonald, 1953). The presence of 182

other ions in the host solution (particularly sodium chloride) has a profound effect, increasing concentrations of dissolved salts tending to proportionately depress the transition temperature. MacDonald (1953) considered that if the amount of sodium chloride in normal sea—water were present in a calcium sulphate rock system, the anhydrite—gypsum transition temperature would be reduced to 35°C. Dickson (in Hsu et al., 1973) argued that in solutions saturated with sodium chloride, the transition temperature could be as low as 20°C.

Though the anhydrite—gypsum transition temperature may thus vary between 42°C and 20°C, in any one rock system the transition temperature is fixed (given set host solution concentrations). The depth at which this fixed temperature'is realized is, of course, a direct function of the prevailing geothermal gradient. In practice, it is found that the maximum depth at which gypsum may be encountered is about 3,500 feet (Murray, 1964).

In general terms, then, the situation is clear. On exhumation, anhydrite rocks eventually enter the temperature realm in which gypsum is favoured and are thereby rendered potentially susceptible to hydration, dependent only on the availability of water.

Volume Increase on Hydration

The anhydrite—gypsum transformation involves a prodigious volume increase. Its magnitude may be precisely determined either by direct' comparison of the volumes of the respective unit cells or by calculations based on the respective molecular weights and densities of the two minerals. Both the anhydrite and gypsum unit cells contain equal numbers of molecules (z = 4) and can be directly compared. Allen (1971) thereby determined the increase to be 63 percent (100 volumes of anhydrite —1163 volumes of gypsum). Calculations based on the molecular weight and density also yield a figure of 63 percent, with possible fluctuation of ± 2-3% depending on what densities are chosen (Figure 78). In dealing with any occcurence of secondary gypsum after anhydrie, it is important to bear in mind that excess calcium sulphate is

183

ANHYDRITE + WATER ----..- GYPSUM

CaSO + 2H2O CaS0 .2H 0 4 4 2

MOLECULAR NT. 136 + 36 172

DENSITY 2.899 to 2.985 + 1.000 2.314 to 2.328

Mass TRT MASS cr

MASS 136 t 136 36 172 t 172 DENSITY 2.899 ° 2.985 1.000 2.314 ° 2.328

46.9 to 45.8 36 74.4 to 73.9 Volumes Volumes Volumes

ANHYDRITE WATER GYPSUM 100 volumes 78 volumes 163 volumes

Figure 78. Volume changes involved in hydration of anhydrite 184

an essential by—product of hydration and must be satisfactorily accommodated.

Water Availability and Access Modes

When exhumation brings anhydrite strata into the thermo- chemical stability field of gypsum, hydration becomes possible. But anhydrite rocks are notoriously tight and water may find access only in rather specialized circumstances. Thus it is possible for anhydrite to persist in a metastable state within the gypsum stability field even if temperatures are well below that of strict anhydrite—gypsum equilibrium.

There are three possible sources from which water for hydration may be derived (Shearman et al., 1973). Firstly, from within; i.e. water present in whatever pore spaces may exist in an anhydrite body. Such inherent water would in theory promote hydration as soon as exhumation had brought the strata up to the anhydrite—gypsum transition level for that particular rock system. The change from anhydrite to gypsum would thus take place under essentially equilibrium conditions (Mossop and Shearman11973). Secondly, water may be introduced from below, from underlying water—bearing formations. If pore— water pressure is allowed to build up in porous strata beneath an anhydrite unit, hydraulic fracturing of the basal part of the overlying anhydrite may ensue. Such fractures would be characteristically closely spaced and parallel to bedding plane fissility. Hydraulic jacking would hold the fractures open while volume—for- volume replacement of the host sulphate proceeded. Excess Calcium sulphate from the hydration would precipitate as fracture—fill. gypsum. When all the water was consumed the overburden wouldbe transferred back onto the vein—fill gypsum, thereby imposing the fibrous satin—spar aspect (Shearman et al., 1972). If hydraulic fracturing occured early in exhumation, at a level where the rocks had only just moved up out of anhydrite stability field, the resultant hydration would be expected to take place under near—equilibrium conditions. If, however, water did not move up into the evaporites until late in exhumation, at levels where 185

temperaure was well below that of anhydrite-gypsum transition, the reaction would take place under conditions of strong disequilibrium (Mossop and Shearman, 1973). Thirdly, water for hydration may be introduced from above,by downward percolation of meteoric water. The depth to which surface waters may penetrate varies as a function of inherent or induced permeability in the rock system and is also controlled in part by climatic factors (see Mossop and Shearman, 1973). Such near—surface hydration of anhydrite must.normally take place under conditions of strong disequilibrium, at temperatures well below that of anhydrite—gypsum equilibrium (Mossop and Shearman, 1973). The fact that water may be made available to anhydrite rocks in various ways and at various levels in exhumation means that secondary gypsum rocks often show evidence of having undergone more than one episode of hydration. As a general rule, it appears that hydration perpetrated at or near to equilibrium conditions gives rise to porphyroblastic gypsum and that hydration under conditions of disequilibrium favours the development of alabastrine fabrics (Mossop and Shearman, 1973). The applicability of this argument for the Baumann Fiord secondary gypsum rocks is assessed in some detail in the discussions that follow.

5.3 Secondary Gv sum of the Basal Baumann Fiord Formation

At Localities 4, 5, 7 (and probably 6 as wellijthe basal parts of the Baumann Fiord Formation are extensively gypsified (Figure 79). The affected zone, which may vary in thickness from 60 to 200 feet, is invariably ramified throughout by gypsum Veins (Figure 80). The great majority of the veins are horizontal to sub—horizontal and consist of fibrous satin—spar, with the fibres aligned vertically, irrespective of the attitude of the vein. Thus, in veins that ar6 slightly inclined, the fibres are slightly oblique to the vein wall. This verticality of fibre orientation appears to be a characterstic feature of satin— spar veins wherever they are encountered (Richardson, 1920). The Baumann Fiord satin—spar veins vary in width from less than BASAL BAUMANN FIORD GYPSUM

.) BAUMANN FIORD FORMATION

Secondary Porphyroblastic Gypsum -----_.. _-, ,-----"" + Satin - Spar Veins

Ficure 79. SchemCl.tic representation of ~ydration by "later in-troduced from the und.erl~ying Copes Bay Formation. Gypsified strata restricted to the basal rea.ches of the Baumann Fiord I-' Form:-,-tion. co 0'\ Figure 80. Secondary gypsum rock in the basal part of Section 4, with multitudinous horizontal veins of fibrous satin--spar gypsum. Vein gypsum represents the excess volume of CaSO liberated by volume—for—volume gypsum replacement of 4 the former anhydrite. 188

I inch to about inches. They are clearly displacive in origin as irregularities on the one wall of a vein can be invariably matched with counterparts on the other wall.

Where fresh, the rock between the veins can be seen to consist largely of porphyroblastic secondary gypsum, with individual crystals ranging in size from two to three hundred microns up to one centimetre. In some places, unaltered anhydrite is also in evidence. Where this is the case, the porphyroblastic secondary gypsum usually lies immediately adjacent to the veins while the anhydrite occupies what is usually a rather thin zone midway between successive veins. In other instances, the pre- servation of anhydrite is more patchy.

In thin—section (Figure 81), these secondary gypsum rocks can be seen to comprise aggregates of porphyroblasts, each clearly defined and showing crisp extinction. They tend to meet along linear compromise boundaries. Virtually all porphyroblasts are poikiloblastic in nature, with corroded anhydrite relics and inclusions of former matrix calcite and dolomite. The presence of these floating relics leaves no doubt that the gypsum was replacive in origin. Furthermore, there is no evidence of disruption of the original anhydrite textures around the porphyroblasts and where the parent anhydrite showed felted— aligned texture, the fabric is continued undisturbed through the porphyroblasts, as marked by the corroded relics therein preserved. Thus it is clear that the replacement took place on a volume—for- volume basis, with no in situ expansion. This mode of replacement is characteristic of porphyroblast growth, as has been firmly established by the petrographic studies of Holliday (1970).

Most workers agree that where satin—spar veins are present, the gypsum of the veins represents the excess volume of gypsum generated as a by—product of volume—for—volume replacement (eg. Bundy, 1956; West, 1965; Holliday, 1967, 1970; Shearman et al., 1972). This certainly seems to be the case in the basal Baumann Fiord setting. A model explaining the generation of satin—spar veins and the gypsification associated therewith has been erected by Shearman et al. (1972) — outlined briefly in the preceding 1F, 9

Figure 81. Porphyroblastic secondary gypsum from the bazal Baumann Fiord Formation; relics of unreplaced anhydrite and former matrix dolomite and calcite. Wedge of unreplaced anhydrite laths in upper centre of photograph. Colour photomicrograph. Crossed—nicols. Field 7 mm. across. 190

section of this text. With the aid of this model it is possible to reconstruct the events that gave rise to the basal Baumann Fiord secondary gypsum rocks:

First of all, it must be argued that this gypsification took place before the Pleistocene epoch. However, it is difficult to assess precisely when hydration did occur. It is possible only to say that the formation must have been exhumed to the extent that the basal part lay within the gypsum stability field (assuming 'normal' pre—Pleistocene geothermal gradients,-.the base of the formation must have been at about 3,500 feet or shallower). The top of the formation could not have reached the meteoric zone before the Pleistocene as there is no evidence of pervasive hydration in the upper reaches of the formation (the pre Pleistocene meteoric zone may be reasonably assumed to have been no more than 500-600 feet — see Mossop and Shearman, 1973). Thus, it seems logical to postulate that hydration of the basal part of the formation occured while the formation lay between 3,500 feet and 500 feet from the surface. The only possible exception could be that the formation was somewhat deeper before the Pleistocene and that during the onset of permafrost, the critical anhydrite—gypsum equilibrium isogeotherm deepened beyond the level of the formation, rendering it potentially susceptible to hydration.

In the areas where basal gypsification took place (see Figure 33), the Baumann Fiord Formation lies on porous Copes Bay Formation equivalents (limestones and dolomites of the Cape Clay, Cass Fiord, and an unnamed formation — Figure 9). Following the hydraulic fracture model of Shearman et al. (1972), it would be in these underlying strata that elevated pore—water pressures must have been generated during exhumation. When hydrostatic pressure had reached a critical level ( X = 1.0 or 1.1 ( X hydrostatic pressure/lithostatic pressure); Shearman et al., 1972),

Permafrost presently extends to about 1,5C0 feet and may have been even deeper during the height of the Pleistocene. During tho onset of permafrost, then, the isogeotherms must have deepened by a minimum of 1500 feet. 191

hydraulic fracturing of the overlying Baumann Fiord strata could proceed. It is unlikely that the entire fracture network was initiated in a single episode of water injection. Rather, the fracturing is thought to have proceeded by degrees, each episode of water injection seeking a higher level in the formation. Differences in the amount of vertical stopping that took place at different localities (Figures 33 and 79) probably reflects differences in the intensity or duration of the hydraulic fracture : regime. Once a fracture had opened and overburden was supported by the water, hydration apparently proceeded by piecemeal volume—for- volume replacement of the anhydrite on either side of the vein, resulting in the generation of secondary gypsum porphyroblasts in the zone adjacent to the fracture. The excess calcium sulphate thus liberated diffused back into the fracture to precipitate as pellucid vein—fill. When all the water in the fracture had been consumed by the hydration process, the overburden load would be transferred back onto the vein—fill gypsum and it is at this stage that the satin—spar fibres are thought to have arisen, imposed upon the vein—fill gypsum as a function of gravity—induced stress (fibres always vertical). The fibres are not in fact discrete crystals in themselves. Rather, several hundred adjacent fibres are all part of one and the same crystal (a former pellucid vein—fill crystal), individual fibres being domains within that crystal (Shearman et al., 1972). That anhydrite is occasionally preserved in the zone midway between two vertically adjacent veins is likely an indication that the total amount of water made available through the two fractures was simply not sufficient to hydrate the total mass of anhydrite. Either succeeding episodes of water injection by- passed the already established fractures in favour of penetrating to higher levels or the whole of the hydraulic fracture regime terminated before a succeeding episode of injection occured.

The fact that gypsification by hydraulic fracture gave rise 192

to porphyroblastic rather than alabastrine fabrics, suggests that the hydration took place under near equilibrium conditions (Mossop and Shearman, 1973). This being the case, it is perhaps valid to suggest that hydraulic fracturing of the basal Baumann Fiord Formation took place when the rocks were still quite deep ( 3,500 feet?), rather than when the top of the formation was within 500-600 feet of the surface.

5.4 Surficial ',leathering—Product Gypsum

As noted previously, Ellesmere Island is an area of perennially frozen ground, the permafrost zone extending to depths between 1,200 and 1,500 feet. From September through June, all interstital water in the rocks is completely frozen. But in summer, a surficial thaw zone develops. By the end of August, the thickness of thawed ground normally reaches two to three feet and it is only in this relatively thin zone that hydration of anhydrite rock is even potentially possible. The availability of water is a problem, however, for the snowmen runoff takes place within scarcely a few days during June and at that time of year, the thaw zone is usually two to three inches at most. Hydration by meltwater is thus severely restricted. During the latter parts of the summer, when the thaw zone is thicker, the only water likely to be available is that derived from rainfall. However, Ellesmere Island normally receives less than 2 inches of rain per summer (Figure 2) and, because the fiord—sides and mountain slopes have no soil horizon and no vegetation to retain the water, the runoff from rainfall is usually extremely rapid.

Erosion is characteristically rapid as well. During the summer season, weather conditions are sometimes such that repeated freezing and thawing occurs and frost wedging becomes an extremely powerful erosive tool. In areas of moderate to high relief, it is unlikely that any rock surface is exposed for more than a few years before it is spalled off by frost action. It is therefore not surprising that prodigious talus fans mantle the mountain sides, creek valley slopes and fiord—sides (see 193

Figures 5b and 42). What is surprising at first is that, although anhydrite rocks tend to spill off in blocks (see Figure 5a), very few pieces of anhydrite can be found in the scree material. Normally, the talus comprises gypsum powder and limestone blocks and it is clear that, once exposed to weathering on all sides, the anhydrite blocks are very quickly hydrated to gypsum. This gypsum is in turn open to leaching and the rock is soon broken down into gypsum flour. In low—lying areas, where drainage is poor, the Baumann Fiord Formation is usually completely leached away and the ground surface ctovered with more resistive carbonate debris. Where outcrops do occur, they are invariably so rotted that large pieces of gypsum, the size of muskoxen, can be picked up and crumpled into powder in one's hands. In other instances, the anhydrite weathers into thin plates of gypsum, extremely friable and powdery in aspect (Figure 44). In the arctic, virtually all the surface water is only just above freezing, being everywhere in intimate contact with ice. The weathering characteristics of the Baumann Fiord anhydrite rocks testify to the extreme vulnerability of anhydrite at these low temperatures. Once liberated from the permafrost, it is clear that hydration of anhydrite proceeds very rapidly indeed. Close inspection of partially hydrated anhydrite blocks reveals a number of interesting points about how surficial hydration proceeds. In the massive Baumann Fiord anhydrites, the gypsified zone in any one block tends to be of uniform thickness (Figure 82), indicating that the water penetrated into the rock from the exposed surface (or surfaces); i.e. the water for hydration was meteoric. This mode of weathering is very similar to that reported by Holliday (1967) in the anhydrite rocks of Spitsbergen. Hydration by interstitial water (recently melted) is not common, for the rocks have very low porosity. Occasionally, small randomly—spaced patches of secondary gypsum may be detected in the anhydrite, likely promoted by interstitial water hydration, but these are not of volumetric significance.

The secondary replacement gypsum occasioned by penetratioli of meteoric water is invariably of the alabastrine type (Figure 82). 194

Figure 82. Massive anydrite grading to alabastrine gypsum toward the exposed surface of the rock. Hydration is by penetration of meteoric water from the upper surface. Colour photomicrograph. Crossed—nicols. Field 1.2 cm. top to bottom. 195

The petrologic altributas of alabastrine secondary gypsum have been comprehensively reviewed by Holliday (1970) and as he points out, the term 'alabastrine' actually refers to a range of textures which form a continuum amongst a series of end members. All of the Baumann Fiord 'alabastrine' gypsum is of one particular type (Type 1 Hydration Texture; Holliday, 1970): In thin— section (Figure 82), the gypsum appears at first sight to consist of uniform aggregates of tiny, apparently discrete, crystals (to 1 mm.). However, under crossed nicols, distinct grain boundaries are not discernible and, on rotating the stage, the whole appears as a mass of erratically anastomosing extinction shadows. In fact, no discrete crystals are normally apparent, each small sub—crystal having lattice linkage to each and every one of its neighbours, as evidenced by the sweeping brush extinction that characterized the contacts between adjacent sub—individuals. The optical and crystallographic disorder inherent in this fabric is thought to be directly related to its having originated under conditions of strong disequilibrium (very low temperature), with hydration proceeding at what is known to be a very rapid rate. Once water is available to penetrate this 'thaw zone' anhydrite rock, multitudinous nuclea- tion centres are initiated and growth of the tiny crystals proceeds very quickly. Mutual interference amongst the competitively growing crystals is thought to cause the considerable lattice distortion that characterizes this fabric (Mossop and Shearman, 1973).

In some secondary alabastrine gypsum rocks, there is evidence that a degree of in situ expansion accompanied the hydration (eg. in the Permian Castile Formation of Texas and New Mexico (David and Kirkland, 1970) — see Mossop and Shearman, 1973). This disturbance can only be attributed to mechanical forces exerted by growth of the gypsum. In the Baumann Fiord surficial gypsum rocks, it is sometimes possible to illustrate that a measure of in situ displacement did in fact accompany the replacement. For example, in some coarsely crystalline vein— fill anhydrite (Figure 76), where meteoric water has worked its way along prominant cleavage planes promoting alabastrine 196

gypsification either side, the residual pieces of anhydrite can be seen to be slightly displaced relative to one another. However, the extent of disturbance is usually much less than that which would be expected to be produced by in situ emplacement of the full 63 percent additional volume of gypsum. Thus it is not at all uncommon to find that a certain amount of gypsum is present in the cracks between adjacent weathering blocks (Figure 83), calcium sulphate that has clearly been sweated out of the parent rock and precipitated as an efflores- cerpeon the outer surface. In some instances the efflorescent gypsum has been redistributed over an outcrop surface by the action of trickling rainwater (Figure 83) and, of course, in many cases the excess gypsum has been dissolved and washed away.

The powdery efflorescent gypsum is interesting from the point of view of its trace strontium content. On the basis of a limited amount of data (Appendix F), it appears that efflorescent gypsum invariably carries higher strontium than does the anhydrite rock from which it is derived. Furthermore, the alabastrine replacement gypsum in these cases always carries less strontium than does the parent anhydrite. From these associations, one can only conclude that, during surficial alabastrine gypsification, strontium is preferentially leached from the anhydrite, a disproportionate amount of strontium diffusing out to the rock surface and being incorporated in the efflorescent gypsum. Preferential leaching of strontium during hydration (sometimes with consequent crystallization of celestite) has been noted in a number of secondary gypsum rocks from diverse settings (Goodman, 1952; Ham, 1962; Stewart, 1963; and Holliday, 1967, 1970).

In the laminar anhydrite rocks, surficial gypsification proceeds in much the same manner as in the massive anhydrite rocks: meteoric water penetrates into the exposed anhydrite, yielding

Celestite (SrS0A ) does not occur as a separate mineral phase in the efflorescence; nor does it occur in either the parent anhydrite or the replacement alabastrine gypsum. Figure 83. Surficial crust of white gypsum covering the surface of an exposed block. On the sunlit portion, the gypsum has been washed over the surface by trickling rainwater. Along the shaded surfaces, the efflorescent gypsum is present in its natural form. 198

alabastrine secondary replacement gypsum. Some in situ expansion accompanies the gypsification but most of the excess calcium sulphate diffuses back to the surface and precipatates as efflorescent gypsum. In these laminar anhydrites, however, the water does not always penetrate uniformly into the exposed rock. Rather, it appears that certain laminae may provide preferred access for the water and gypsification along these zones is thereby favoured. In thin-section, a partially hydrated laminar anhydrite rock often appears as a series of wispy alabastrine bands intermingled with zones of unreplaced anhydrite (Figure 84b). It is immediately clear that the preferred water access was along zones of flow fabric perfection in the parent anhydrite (Figure 84a). Either the zones of perfect lath alignment are more naturally permeable than are the slightly less ordered zones, or, the tiny cracks that open up as a result of relief of confining pressure tend to follow the zones of greatest alignment fissility, thereby providing for the preferred access of meteoric water (see Mossop and Shearman, 1973). The remarkable feature of this secondary alabastrine gypsum is that it exhibits a strong preferred fabric, quite clearly inherited from the felted-aligned fabric of the parent anhydrite. Details of the fabric are as follows: Firstly, there is a recognizable dimensional foliation (Figure 84a). The sub-crystals of the alabastrine gypsum tend to be flattened in the plane of bedding, parallel to the plane of flattening of the parent anhydrite laths. This dimensional preference is uniaxial, however, as the foliation is equally marked in sections cut normal to the axis of parent lath elongation as it is in sections parallel to it; i.e. the alabastrine sub-crystals are flattened parallel to the 010 plane of the parent laths but are essentially equidimensional within that plane (may be thought of as discoidal in shape). Secondly, the alabastrine sub-crystals invariably show parallel extinction; i.e. they extinguish parallel to their

••■••■■■1.01.11.41. Holliday (1970; Fig. 5) illustrated foliation in secondary alabastrine gypsum from Spitsbergen and noted that the foliation was inherited from the parent anhydrite. 199

Figure 84a. Anhydrite (upper right) grading to alabastrine gypsum (towards lower left ) along the edge of a preferentially gypsified zone. The alabastrine gypsum follows zones of well— aligned anhydrites laths, relics of which are visible within the rather large alabastrine subcrystals. Monochrome Photomicro- graph. Crossed—nicols. Field 1.2 mm. across. Figure 84b.

Figure 84c.

Figures 84b and 84c. Anhydrite preferentially altered to secondary alabastrine gypsum along zones of latil alignment fissility. The alabastrine gypsum has a preferred dimensional foliation (parallel to the fabric of parent ani7drile: 84b) and a preferred optical fabric (apparent when snsiive tint is inserted; 84c). Colour photomicrographs. 45 position. Field 2 cm. across. Crossed—nicols: 84b. nth quartz sensitive tint: 84c. 201

Fi gure 85. n-44-,1 incli catrix of cr7rn,lim lritawPri in the a—c Y. P. 1.5g3., (I) normal to plane of projection). Angle (a A c) = 113 50 (Be Jong and Bouman cell). GYPSUM OPliCS

010 SECTION

NB. TRACE OF INDICATRIX NOT TO SCALE 203

plane of flattening. In this respect, of course, they mimic their anhydrite lath precursors. Finally, the use of a sensitive tint reveals that the alabastrine sub-crystals are dominantly length-slow (Figure 846), with about 70 percent reinforcing the plate when the elongation direction of the sub-crystals coin- cides with the N direction of the plate. The remaining 30 percent are length-fast.

In order to explain this preferred optical fabric, it is necessary to take into account the shape and orientation of the biaxial indicatrix of gypsum (Figure 85). The crystal axes shown in Figure 85 are based on the A_ cell of gypsum (De Jong and Douman, 1939) and accord with the axis orientation employed by Deer, Howie and Zussman (1962, pp. 202-203). The refractive indices auoted are those that apply when 211.1.- = 58°, the standard optic axial angle for gypsum at the temperatures in which it is normally encountered.

The X and Z vibration directions of gypsum are oblique to both the crystal axes and to the rational crystal faces of the mineral. Thus it is clear that parallel extinction can be manifest only when the mineral is viewed in the X-Z plane, with 010 lying normal to the plane of the thin-section. This is a demonstration of the structural affinity between anhydrite and gypsum for it means that, during alabastrine gypsification, the gypsum nucleates with its 010 plane co-parallel with the 010 plane of the parent anhydrite (the platy anhydrite cleavage flakes lie with 010 normal to the plane of the thin-section in bedding-normal sections). The extent of structural identity between the anhydrite laths and the alabastrine sub-crystals derived therefrom appears, however, to be limited to the inheritance of 010, for the preferred length-slow fabric of the alabastrine gypsum can be explained in terms of random orientation of the X and Z vibration axes (Y. being fixed). In sections normal to bedding, with Y standing vertically in the plane of the thin-section, any random cut exposes a 180° arc in the optic axial plane (Figure 85). Statistically, 58/180 (about 30 of the sub-crystals) will approach acute bisectrix-centred orientation (length-fast) 204

while the remaining 122/180 (about 70) will show obtuse bisectrix orientation (length—slow).

If '010 inheritance' can be taken as a univerally applicable manifestation of alabastrine gypsification, then this attribute may prove to be a useful tool in diagnosing the pre—hydration textures of other ancient anhydrite rocks now gypsified. Certainly it is of use in the Baumann Fiord alabastrine gypsum rocks. For example, in rocks that have been totally gypsified, it is possible to determine with relative accuracy whether the anhydrite precursor was of massive texture, felted texture or felted—aligned texture. Figure 86 shows an alabastrine gypsum rock that undoubtedly stemmed from an anhydrite precursor with very well developed felted—aligned fabric. With one of the nicols parallel to bedding, there is a near—total extinction (Figure 86a). In the maximum illumination position (Figure 86b), the sub—crystals show birefringence between .003 and .007 (retardation between 80 and 220 mu in this 300,A section) but with most tending toward the upper limit (approaching obtuse bisectrix—centred; Figure 85). With the sensitive tint inserted (Figure 86c), 70 percent of the subcrystals are length—slow (reinforcing the plate to second—order blue) and 30 percent are length—fast (compensating the plate to first—order orange). This series of optical peculiarites can beamlained only if the parent anhydrite consisted of well aligned cleavage flakes.

Where the parent anhydrite comprised a nodular mosaic.with randomly oriented cleavage flakes, the alabastrine gypsum derived thereform shows neither a dimensional nor an optical fabric. In those massive anhydrites where there is a preferred fabric (010 horizontal), the alabastrine replacement gypsum often shows an optical preference similar to that in the laminar rocks, but this is not normally so well defined. For the laminar anhydrites, of course, the replacement gypsum fabric is always well defined, as outlined above. 205

Figure 86. Alabastrine secondary gypsum after felted aligned anhydrite. Identical field of view in all three photographs:

86a. Extinction position. The great majority of the gypsum is here extinguiShed;small points of light stem from strained boundaries of sub—individuals.

86b. 45° position of maximum illumination. Dimensional foliation well developed.

86c. Sensitive tint inserted (N from upper left to lower right). 70-`• reinforcement to second—order blue. 30% compensation to first—order orange.

Colour photomicrographs. Crossed—nicols. Sensitive tint in 84c. Field 3.2 mm. across.

Figure 86a. Figure 86b.

Figure 86c. 207

5,5 Summary and Conclusions

The secondary gypsum rocks of the Baumann Fiord Formation are of two distinct types, developed respectively in two distinct geological settings during two temporally distinct hydration regimes. First; in the comparatively early stages of exhumation, when the formation resided at depths in the vicinity of 3,000 4,000 feet, the basal reaches of the formation underwent hydraulic fracturing. Water from the underlying strata was forced up into the evaporites, producing a network of horizontal to sub-horizontal fractures that were jacked open by the water. Hydration to porphyroblastic secondary gypsum proceeded on a volume-for- volume basis and the excess calcium sulphate thereby liberated reprecipitated in the fractures. With reloading of the overburden onto the veins, the fracture-fill gypsum underwent stress- induced lattice reordering and was transformed to satin-spar. Second; under the present-day weathering regime, 'thaw-zone' gypsification by meteoric water takes place. The hydration proceeds very rapidly, under conditions of strong thermal disequilibrium, and invariably yields the alabastrine secondary gypsum. Some in situ expansion accompanies the hydration but most of the excess calcium sulphate precipitates as an efflorest- eence on the surfaces of the exposed blocks. Where the parent anhydrite possessed a preferred fabric, so the alabastrine gypsum derived therefrom possesses a distinctive fabric, the genetic link between the two being one of crystallographic inheritance of 010.

Because of the intervention of perMafrost during the latest stages of exhumation, fresh anhydrite rock is exposed at the surface in the Baumann Fiord Formation. Furthermore, of the secondary gypsum rocks developed in the formation, there is a clear and demonstrable dicotomy between those that are pre- Pleistocene in origin and those that result from present-day 'thaw zone' hydration. For the study of both anhydrite and secondary gypsum rocks, then, these peculiarities of the Baumann 208

Fiord setting afford a rare and challenging opportunity.

What is now exposed at the surface on Ellesmere Island would normally be observable only in borehole cores, for, with the onset of permafrost, the whole of the Baumann Fiord Formation was literally 'frozen' in the state in which it existed at 3,000 — 4,000 feet (i.e. pre—Pleistocene depth). Because of the erosive grinding of the Ice—Age glaciers, the cover rock has been differentially excavated and the formation is exposed, in its pre—Pleistocene form, along the resultant fiord—sides and valley walls. It is perhaps not inappropriate, therefore, to end this thesis with an imagerative tribute to these retreating but unbowed glaciers. 209

REFERENC7T,S 210

Aitken, J.D. 1967. Classification and environmental significance of cryptalgal limestones and dolomites, with illustrations from the Cambrian and Ordovician of southwestern Alberta. J. Sedim. Petrol., 37, pp. 1163-1178.

Allen, B.G. 1971. The origin of sheet fractures in the Galore Creek copper deposits, British Columbia. Can. J. Earth Sci., 8, pp. 704-711. Armstrong, J.E. 1947. The Arctic Archipelago; geology and economic minerals of Canada. Geol. Surv. Canada Econ. Geol. Series No. 1, 3rd ed., pp. 311-324.

Bathurst, R.G.C. 1971. Carbonate sediments and their diagenesis. Amsterdam London New York, Elsevier, 620 pp.

Black, M. 1933. The algal sediments of Andros Island, Bahamas. Phil. Trans. Roy. Soc. London, Serv. B., 222, pp. 165-192.

Borchert, H. and Muir, R.D. 1964. Salt deposits. The origin, metamorphism and deformation of evaporites. London, Van Nostrand, 338 pp. Bosellini, A. and Hardie, L.A. 1973. Depositional theme of a marginal marine evaporite. Sedimentology, 20, pp. 5-27.

Braun, N. and Friedman, G.M., 1969. Carbonate lithofacies and environments of the Tribes Hill Formation (Lower Ordovician) of the Mohawk Valley, New York. J. Sedim. Petrol., 39, pp. 113-135. Bundy, W.M. 1956. Petrology of gypsum-anhydrite deposits of southwestern Indians. J. Sedim. Petrol., 26, pp. 240-252.

Bush, P.R. 1970. Chloride rich brines from sabkha sediments and their possible role in ore geneSis. Trans. Instn. Min. Metall., B79, pp. 137-144. Bush, P.R. 1973. Sedimentology and groundwater chemistry of some Recent coastal plain sediments, Abu Dhabi, Trucial Coast, Persian Gulf. Ph.D. Thesis, Univerity of London, 356 pp. Butler, G.P. 1966. Early diagenesis in Recent sediments of the Trucial Coast of the Persian Gulf. M.Sc. Thesis, University of London, 192 pp. 211

Butler, G.P. 1969. Modern evaporite deposition and geochemistry of co—existing brines, the sabkha, Trucial Coast, Arabian Gulf. J. Sedim. Petrol., 39, pp. 70-89. Butler, G.P. 1970. Holocene gypsum and anhydrite of the Abu Dhabi sabkha, Trucial Coast: An alternative explanation of origin. Proc. 3rd. Salt Symp., Vol. 1. Cleveland, N. Ohio Geol. Soc., pp. 120-152. Cayeux, L. 1916. Iltude p4trographique des roches skimentaires: Texte. Paris, Imprimerie Nationale, 524 pp. Cayeux, L. 1929. Les roches sedimentaires de France — roches siliceuses. Paris, Imprimerie Nationale, 774 pp.

Christie, R.L. 1967. Bache Peninsula, Ellesmere Island, Arctic Archipelago. Geol. Surv. Canada Memoir 347, 63 pp. Christie, R.L. 1971. Publications on the geology of the Arctic Islands by the Geological Survey of Canada. Geol. Surv. Canada Paper 71-10, 21 pp. Cloud, P.E. 1962. Environment of calcium carbonate deposition west of Andros Island, Bahamas. U.S. Geol. Surv. Prof. Paper 350, 138 pp. Cuff, C. 1969. Lattice disorder in Recent anhydrite and its geological implications.(demonstration). Free. Geol. Soc. Lond., No. 1659, pp. 326-330. Curtis, R., Evans, G., Kinsman, D.J.J. and Shearman, D.J. 1963. Association of dolomite and anhydrite in the Recent sediments of the Persian Gulf. Nature, Lond., 197 (4868), pp. 679-680. Dana, E.S. 1932. Textbook of mineralogy (4th ed.). London, Chapman and Hall, 851 pp. Davies, G.R. 1970. Algal—laminated sediments, Gladstone Embayment, Shark Bay, Western Australia. AniAssoc. Petrol. Geol. Hem. 13; pp. 169-205.

Davis, J.B. and Kirkland, D.W. 1970. Native sulfur deposition in the Castile Formation, Culberson County, Texas. Econ. Geol., 65, pp. 107-121. 212

Deer, W.A., Howie, R.A. and Zussman, J. 1962. Rock forming ilrinerals, Vol. 5, Non—silicates. London, Longhams, 371 pp.

De Jong, W.F. and Bauman, J. 1939. Des reziproke and das Bravais- sche Gitter von Gips. Zeit. Krist., 100, p. 275.

Douglas, R.J.W. (ed.) 1970. Geology and economic minerals of Canada. Geol. Surv. Canada Econ. Geol. Rept. No. 1, 5th ed., 838 pp.

Evans, G. 1966. The sedimentary facies of the Persian Gulf region. Phil. Trans. Royal Soc., A259, pp. 291-298.

Evans, G. and Shearman, D.J. 1964. Recent celestite from the sediments of the Trucial Coast of the Persian Gulf. Nature Lond., 2o2, No. 4930, pp. 385-386.

Evans, G., Schmidt, V., Bush, P. and Nelson, H. 1969. Strati- graphy and geologic history of the sabkha, Abu Dhabi, Persian Gulf. Sedimentology,12, pp. 145-159.

Folk, R.L. and Pittman, J.S. 1971. Length—slow chalcedony: A new testament for vanished evaporites. J. Sedim. Petrol., 41, pp. 1045-1058. Forgotson, J.M. 1958. A correlation and regional stratigraphic analysis of the formations of the Trinity Group of the Comanchean Cretaceous of the Gulf Coast Plain. Trans. Gulf Coast Assoc. Geol. Socs., 6, pp. 91-108.

Fortier, Y.O. Blackadar, R.G., Glenister, B.F., Greiner, H.R., McLaren, D.J., McMillan, N.J., Norris, A.W., Roots, E.F., Souther J.G., Thorsteinsson R. and Tozer E.T. 1963. Geology of the north—central part of the Arctic Archipelago, Northwest Territories (Operation Franklin). Geol. Surv. Canada Memoir 320, 671 pp. Fuller, J.G.C.M. 1956. Madison limestone in southeastern Saskatchewan. Oil in Canada, viii, No. 40, pp. 34-58. Fuller, J.G.C.M. and Porter, J.W. 1969a. Evaporite formations with petroleum reservoirs in Devonian ard Mississippian of Alberta, Saskatchewan and North Dakota. Bull. Am. Assoc. Petrol. Geol., 53, pp. 902-926. 213

Fuller, J.G.C.M. aild Porter, J.W. 1969b. Evaporites and carbonates; two Devonian basins of western Canada. Bull. Can. Petrol. Geol., 17, pp. 182-193.

Garrett, P. 1970. Phanerozoic stromatolites: noncompetitive ecologic restriction by grazing and burrowing animals. Science, 169, pp. 171-173. Gebelein, C.D. 1969. Distribution, morphology and accretion rate of Recent subtidal algal stromatolites, Bermuda. J. Sedim. Petrol., 39, pp. 49-69. Gill, W.D. and Ala, A. 1972. Sedimentology of the Gachsaran Formation (Lower Fars series), southwest Iran. Bull. Am. Assoc. Petrol, Geol., 56, pp. 1965-1974. Ginsburg, R.N., Isham, L.G., Bein, S.J. and Kuperburg, G. 1954. Laminated algal sediments of South Florida and their recognition in the fossil record. Coral Gables Marine Laboratory Rept. 54-21, 337 pp. Goldman, M.I. 1952. Deformation, metamorphism, and mineraliza- tion in gypsum-anhydrite cap-rock, Sulphur salt dome, Louisiana. Geol. Soc. Am. Mem. 50, 169 pp. Goodman, N.R. 1952. Gypsum and anhydrite in Nova Scotia. Nova Scotia Dept. Mines Nem. 1, 75 pp. Ham, W.E. 1962. Economic geology and petrology of gypsum and anhydrite in Blaine County. Oklahoma Surv. Bull., 89, pp. 100-151. Hardie, L.A. 1967. The gypsum-anhydrite eauilibrium at one atmosphere pressure. Am. Mineral., 52, pp. 171-200.

Hoen, E.W. 1964. The anhydrite diapirs.of central western Axel Heiberg Island. Axel Heiberg Island Research Reports, Geology; No. 2, McGill Univ. 120 pp.., Hoffman, P. 1971. Recent stromatolites and loferites, Bay, Western Australia (abs.). .VIII Internat. Sediment. Cong., pp. 1,2.

Holliday, D.W. 1966. Nodular gypsum and anhydrite rocks in the Billefjorden region, Spitzbergen. Norsk Polarinstitutt Arbok, 1965, pp. 65-73. 214

Holliday, D.W. 1961. Secondary gypsum in Middle Carboniferous rocks of Spitsbergen. Geol, Mag.. 104, pp. 171-177. Holliday, D.W. 1968. Early diagenesis in Middle Carboniferous nodular anhydrite of Spitsbergen. Proc. Yorks. Geol. Soc.; 36, pp. 277-292. Holliday, D.W. 1970. The petrology of secondary gypsum rocks: a review. J. Sedim. Petrol., 40, pp. 734-744. Holliday, D.W. 1971. Origin of Lower Eocene gypsum—anhydrite rocks, southeast St. Andrew, Jamaica. Trans. Instn. M. Metall., B80, pp. 305-315. Holliday, D.W. 1973. Early diagenesis in nodular anhydrite rocks. Trans. Instn. Min. Metall., B82, pp. 81-84. Holtedahl, O. 1917. Summary of geological results in Report of the second Norwegian expedition in the Pram, 1898-1902. Vidensk.—Selsk. I. Kristiania, 4, 27 pp. Howitt, F. 1964. Stratigraphy and structure of the Purbeck inliers of Sussex (England). Q.J. Geol. Soc. Lond., 120, pp. 77-113. Hstt, K.J., Ryan, W.B.F. and Cita, M.B. 1973. Late Miocene dessication of the Mediterranean. Nature, Land., 242, PP. 240-244. Illing., L.V., Wells, A.J. and Taylor, J.C.M. 1965. Penecon- temporary dolomite in the Persian Gulf. Spec. Pub. Soc. Econ. Palaeot. Miner., 13, pp. 89-111. Illing, L.V., Wood, G.U. and Fuller, J.G.C.M. 1967. Reservoir rocks and stratigraphic traps in non—reef carbonates. 7th World Petroleum Congress, 2, pp; 487-499. Jackal A.D. 1973'. Deposition and diagenesis of Permian evaporites and associated carbonates and elastics on shelf areas of the Permian Basin (abs.). Proc. 4th Salt Symposium, Cleveland, N. Ohio Geol. Soc. pp. 16. Kendall, C.G. St.C. 1966. Recent carbonate sediments of the Western Thor al Bazam, Abu Dhabi, Trucial Coast. Ph4D. Thesis, University of London. 286 pp. 215

Kendall, C.G. St.C. and Skipwith, Sir Patrick, A.d'E. 1968. Recent algal mats of a Persian Gulf lagoon. J. Sedim. Petrol., 38, pp. 1040-1058.

Kendall, C.G. St.C. and Skipwith, Sir Patrick A.d'E. 1969a. Holocene shallow—water carbonate and evaporite sediments of Khor al Bazam, Abu Dhabi, Southwest Persian Gulf. Bull. Am. Assoc. Petrol. Geol., 53, pp. 841-869.

Kendall, C.G. St.C. and Skipwith, Sir Patrick A.D'E. 1969b. Geomorphology of a Recent shallow Water carbonate province; Khor al Bazam, Trucial Coast, Southwest Persian Gulf. Bull. Geol. Soc. Am., 8o, pp. 865-892. Kerr, J.Wm. 1967a. Stratigraphy of central and 'eastern Ellesmere Island, Arctic Canada: Part I, Proterozoic and Cambrian. Geol. Surv. Canada Paper 67-27(I), 63 pp. Kerr, J.Wm. 1967b. New nomenclature for Ordovician rock units of the eastern and southenlQueen Elizabeth Islands, Arctic Canada. Bull. Canadian Petrol. Geol., 15, pp. 91-113.

Kerr, J.Wm. 1968. Stratigraphy of central and eastern Ellesmere Island, Arctic Canada: Part II, Ordovician. Geol. Surv. Canada Paper 67-27(11)1 92 pp. Kerr, J.Um. 1973. Tips on organizing Arctic geological field work. Geol. Surv. Canada Open File rept. 151. Calgary, Orhans Ltd., 32 pp.

Kerr, J.Wm. in preparation. Stratigraphy of central and eastern Ellesmere Island, Arctic Canada: Part III, Upper Ordovician (Ashgillian), Silurian and Devonian. Geol. Surv. Canada Paper 67-27(111). Kerr, J.Wm., Morrow, D.W., and Savigny, K.W. 1973. Grinnell Peninsula, Devon Island, District of Franklin in Report of Activities, Part A: April to October 1972. Geol. Surv. Canada Paper 73-1A, pp. 262-263. Kindle, E.M. 1939. Geology of the Arctic Archipelago and 7 Interior Plains of Canada in Geologie der Erde. E. Krekel (ed.), Berlin, Gebrvden Borntraeger. 216

Kinsman, D.,J.J. 1964. Recent carbonate sedimentation near Abu Dhabi, Trucial Coast, Persian Gulf. Ph.D. Thesis. Unive,:sity of London, 302 pp.

Kinsman, D.J.J. 1966. Gypsum and anhydrite of Recent age, Trucial Coast, Persian Gulf. Proc. 2nd. Salt Symp., Vol. 1. Cleveland, N. Ohio Geol. Soc., pp. 302-326.

Kinsman, D.J.J. 1969. Modes of formation, sedimentary associations, and diagnostic features of shallow—water and supra— tidal evaporites. Bull. Am. Assoc. Petrol. Geol., 53, pp. 830-840. Levorsen, A.I. 1954. Geology of Petroleum. London, Freeman and Co., 724 pp.

Llewellyn, P.G., Mahmoud, S.A., and Stabbins, R. 1968. Nodular anhydrite in Carboniferous limestone, West Cumberland. Trans. Instn. Min. Metal., B77, pp. 18-25.

Llewellyn, P.G. and Stabbins, R. 1968. Core material from the anhydrite series, Carboniferous limestone, Hathern borehole, Leicestershire. Proc. Geol. Soc. Lond., No. 1650, pp. 171-180.

Llewellyn, P.O. and Stabbins, R. 1970. The Hathern anhydrite series, Lower Carboniferous, Leicestershire, England. Trans. Instn. Min. Metall., B79, pp. 1-15.

Logan, B. 1961. Cryptozoon and associate stromatolites from the Recent, Shark Bay, Western Australia. J. Geol., 69, PP. 517-533. Logan, B.W., Rezak, R. and Ginsburg, R.N. 1964. Classification and environmental significance of algal stromatolites. J. Geol., 72, pp. 68-83. Low, A.P. 1906. Cruise of the Neptune. Ottawa, Queen's Printer, pp. 183-247. MacDonald, G.J.F. 1953. Anhydrite—gypsum equilibrium relations. Am. J. Sci., 251, pp. 884-898. MacQueen, R.W. and Price, L.L. 1968. Carbonate—evaporite cycles, Souris River Formation, Saskatchewan in Report of Activities, Part A: April to October, 1967. Geol. Surv. Canada Paper 68-1, pp. 202-203. 217

Maiklem, W.R., Debout, D.G. and Glaister, R.P. 1969. Classifica- tion of anhydrite — a practical approach. Bull. Can. Petrol. Geol., 17, pp. 194-233.

Masson, P.M. 1955. An occurence of gypsum in southwest Texas. J. Sedim. Petrol., 25, pp. 72-77. McMillan, J.G. 1910. Geographical report in Report of the Dominion of Canada Government Expedition to the Arctic Islands and the Hudson Strait on board the D.G.S. Arctic (1908-09). J.E. Bernier (ed.), Ottawa, queen's Printer.

Michel—Levy and Munier—Chalmas. 1892. Memoires sur les diverses formes affectees par le r4seau 41ementaire du quartz. Bull. Soc. Fr. Min., XV, pp. 159-190. Moore, R.C. (ed.) 1962. Treatise on invertebrate paleontology, Part W., miscellanea. University of Kansan Press, pp. 164.

Mossop, G.D. 1972a. Baumann Fiord Formation evaporites of central Ellesmere Island in Report of Activities, Part A: April to October, 1971. Geol. Surv. Canada Paper 72-1A, pp. 216-217. Mossop, G.D. 1972b. Baumann Fiord Formation evaporites of central Ellesmere Island, Arctic Canada in Report of Activities, Part B: November 1971 to March 1972. Geol. Surv. Canada Paper 72-1B, pp. 86-90. Mossop, G.D. 1973. Lower Ordovician evaporites of the Baumann Fiord formation, Ellesmere Island in Report of Activities, Part April to October 1972. Geol. Surv. Canada Paper 73-1A, pp. 264-267. Mossop, G.D. and Shearman, D.J. 1973. Origins of secondary gypsum rocks. Trans. Instn. Min. Metall., B82, pp. 147-154. Murray, R.C. 1964. Origin and diagenesis of gypsum and anhydrite. J. Sedim. Petrol., 34, PP. 512-523. Neumann, A.G., Gebelein, C.D. and Scoffin, T.P. i970. The composition, structure and erodability of subtidal mats, Abaco, Bahamas. J. Sedim. Petrol., 40, pp. 274-r297. 218

Park, R.K. 1969. Written contribution in Lattice disorder in Recent anhydrite and its geological implications: C. Cuff. Proc. Geol, Soc. Lond., No. 1659, pp. 331-332. Park, R.K. 1973. Unpublished Ph.D. Thesis, University of Reading.

Posnjak, E. 1938. The system CaSO4—H20. Am. J. Sci., 35A, PP.247-272. Purser, B.H. (ed.) 1973. The Persian Gulf. Berlin Heidelberg New York, Springer—Verlag, 471 pp. Richardson, W.A. 1920. The fibrous gypsum of Nottinghamshire. Miner. Mag., 19, pp. 77-95. Schei, Per 1903. Summary of geological results; second Norwegian north polar expedition in the Pram, 1693-1962. Geol. J., 22, pp. 56-65.

Schwerdtner, W.M. and Clark, A.R. 1967. Structural analysis of Mokka Fiord and South Fiord domes, Axel Heiberg Island, Canadian Arctic. Can. Journ. Earth Sci., 4, pp. 1229-1245. Schenk, P. 1969. Carbonate—sulphate—redbed facies and cyclic sedimentation of the Windsorian Stage (M. Carboniferous), Maritime Provinces. Canadian Journ. Earth Sci., 6, pp. 1019-1066.

Scoffin, T.P. 1970. The trapping and binding of subtidal carbonate sediments by marine vegetation in Bimini Lagoon, Bahamas. J. Sedim. Petrol., 40, pp. 249-273. Shearman, D.J. 1963. Recent anhydrite, gypsum, dolomite and halite from the coastal flats of the Arabian shore of the Persian Gulf. Proc. Geol. Soc., No. 1607, pp. 63-64. Shearman, D.J. 1966. Origin of marine evaporite by diagenesis. Trans. Instn. Min. Metall., B75, pp. 208-215.

Shearman, D.J. in preparation. Replacement anhydrites in the Madison formation of southeast Saskatchewan. Bull. Can. Petroleum Geol. 219

Shearman, D.J. and Fuller, J.G.C.M. 1969. Anhydrite diagenesis, calcitization and organic laminites, Winnipegosis Formation, piddle Devonian, Saskatchewan. Bull. Can. Petrol. Geol., 17, pp. 496-525. Shearman, D.J., Mossup, G.D., Dunsmore, H. and Martin, M. 1972. Origin of gypsum veins by hydraulic fracture. Trans. Instn. Min. Metall., B81, pp. 149-155. Shearman, D.J., Mossop, G.D., Dunsmore, H. and Martin, M. 1973. Discussion and contributions: Origin of gypsum veins by hydraulic fracture. Trans. Instn. Min. Metall., B82, pp. 66-67. Shinn, E.A., 1968. Selective dolomitization of Recent sedimentary structures. J. Sedim. Petrol., 38, pp. 612-616. Skipwith, Sir Patrick A.d'E. 1966. Recent carbonate sediments of Eastern Khor al Bazam, Abu Dhabi, Trucial Coast. Ph.D. Thesis, University of London, 271 pp. Smith, A.G., Briden1 J.C. and Drewry, G.E. 1973. Phanerozoic world maps in Organisms and continents through time. N.F. Hughes (ed.); Special Paper in paleontology No. 12, London, The Paleontological Association, pp. 1-42. Spry, A. 1969. Metamorphic textures. Oxford, Pergamon Press, 350 pp. Stewart, F.H. 1963. Marine evaporites. U.S. Geol. Surv. Prof. Paper, 440—Y, 52 PP. Sugden, W. 1963. The hydrology of the Persian Gulf and its significance with respect to evaporite deposition. Am. J. Sci., 261, pp. 741-755. Sverdrup, O. 1904. New land: Four years in the Arctic regions. Transla. by E.H. Hearn. London, Longmans Green, 2 vols., 496 PP., 504 PP. Thode, H.G. and Monster, J. 1964. Sulphur isotope geochemistry of petroleum, evaporites, and ancient seas. Am. Assoc. Petrol. Geol. Memoir 4, pp. 367-377. 220

Thorsteinsson, R. 1958. Cornwallis and Little Cornwallis Islands, District of Franklin, Northwest Territories. Geol. Surv, Canada Memoir 294, 134 pp.

Thonteinsson, R. 1963. Copes Bay in Operation Franklin. Geol. Surv. Canada Memoir 320, pp. 386-394.

Thorsteinsson, R. 1972. Geology, Eureka Sound South, District of Franklin; Scale 1:250,000. Geol. Surv. Canada Map 1300.

Thorsteinsson, R. and Kerr, J.14m. 1968. Cornwallis Island and adjacent smaller islands, Canadian Arctic Archipelago. Geol. Surv. Canada Paper 67-64, 16 pp.

Thorsteinsson, R. and Tozer, E.T. 1970. Geology of the Arctic Archipelago in Geology and economic minerals of Canada. R.J.W. Douglas (ed.). Geol. Surv. Canada Econ. Geol. Rept. No. 1, 5th ed., pp. 548-590.

Trettin, H.P. 1969a. Pre—Mississippian geology of northern Axel Heiberg and northwestern Ellesmere Islands, Arctic Archipelago. Geol. Surv. Canada Bull. 171, 82 pp. Trettin, H.P. 1969b. Geology of Ordovician to Pennsylvanian rocks, M'Clintock Inlet, north coast of Ellesmere Island, Canadian Arctic Archipelago. Geol. Surv. Canada Bull. 183, 93 pp. Trettin, H.P. 1971a. Geology of Lower Paleozoic formations, Hazen Plateau and southern Grant Land Mountains, Ellesmere Island, Arctic Canada. Geol. Surv. Canada Bull. 203, 134 pp. Trettin, H.P. 1971b. Reconnaissance of Lower Paleozoic geology, Phillips Inlet region, north coast of Ellesmere Island, District of Franklin. Geol. Surv. Canada Paper 71-12, 29 pp.

Trettin, H.P., Frisch, T.O., Sobczak, L.H., Weber, J.R., Niblett, E.R., Law, L.K., De Laurier, J.E. and Whitlam, K. 1972. The Innuitian Province in Variations in tectonic style in Canada. R.A. Price and R.J.W. Douglas (eds.). Geol. Assoc. Canada Spec. Paper 11, 688 pp. 221

Troelsen, J.C. 1950. Contributions to the geology of northwest Greenland, Ellesmere Island and Axel Heiberg Island. Medd. om Gr$nland, Ed. 149, Mr. 7, 85 pp. Twenhofel, N.H. 1954. Correlation of the Ordovician formations of North America. Bull. Geol. Soc. Am., 65, pp. 247-298.

Wanless, R.K. 1969 (comp.). Isotopic age map of Canada, scale 1:5,000,000. Geol. Surv. Canada Map 1256A.

West, I.M. 1964. Evaporite diagenesis in the Lower Purbeck Beds of Dorset. Proc. Yorks. Geol. Soc., 34, pp. 315-330. West, I.M. 1965. Macrocell structure and entrolithic in Britich Purbeck gypsum and anhydrite. Proc. Yorks. Geol. Soc., 35, PP. 47-58. West, I.M., Brandon, A. and Smith, M. 1968. A tidal flat evaporitic facies in the Visean of Ireland. J. Sedim. Petrol., 38, pp. 1079-1093. Wood, G.V. and Wolfe, M.J. 1969. Sabkha cycles in the Arab/ Darb formation off the Trucial Coast of Arabia. Sedimen- tology, 12, pp. 165-191. Young, G.A. 1926. Arctic Archipelago and Hudson Bay Lonland; geology and economic minerals of Canada. Geol. Surv. Canada Econ. Geol. Series No. 1, 2nd ed., pp. 122-127. 212

Deer, W.A., Howie, R.A. and Zussman, J. 1962. Rock forming minerals, Vol. 5, Non—silicates. London, Longhams, 371 pp.

De Jonc;, W.F. and Bcuman, J. 1939. Das reziproke and das Bravais- sche Gitter von Gips. Zeit. Krist., 100, p. 275.

Douglas, R.J.W. (ed.) 1970. Geology and economic minerals of Canada. Geol. Surv. Canada Econ. Geol. Rept. No. 1, 5th ed., 838 pp.

Evans, G. 1966. The sedimentary facies of the Persian Gulf region. Phil. Trans. Royal Soc., A259, pp. 291-298. Evans, G. and Shearman, D.J. 1964. Recent celestite from the sediments of the Trucial Coast of the Persian Gulf. Nature Lond., 2o2, No. 4930, pp. 385-386.

Evans, G., Schmidt, V., Bush, P. and Nelson; H. 1969. Strati- graphy and geologic history of the sabkha, Abu Dhabi, Persian Gulf. Sedimentology, 12, pp. 145-159. Folk, R.L. and Pittman, J.S. 1971. Length—slow chalcedony: A new testament for vanished evaporites. J. Sedim. Petrol., 41, pp. 1045-1058.

Forgotson, J.M. 1958. A correlation and regional stratigraphic analysis of the formations of the Trinity Group of the Comanchean Cretaceous of.the Gulf Coast Plain. Trans. Gulf Coast Assoc. Geol. Socs., 6, pp. 91-108.

Fortier, Y.O. Blackadar, R.G., Glenister, B.]?., Greiner, H.R., McLaren, D.J., McMillan, N.J., Norris, A.W., Roots, E.F., Souther J.G., Thorsteinsson R. and Tozer E.T. 1963. Geology of the north—central part of the Arctic Archipelago, Northwest Territories (Operation Franklin). Geol. Surv. Canada Memoir 320, 671 pp. Fuller, J.G.C.M. 1956. Madison limestone in southeastern Saskatchewan. Oil in Canada, viii, No. 40, pp. 34-58. Fuller, J.G.C.M. and Porter, J.W. 1969a. Evaporite formations with petroleum reservoirs in Devonian and Mississippian of Alberta, Saskatchewan and North Dakota. Bull. Am. Assoc. Petrol. Geol., 53, pp. 902-926. 223

APFETTDIX A

PUBLISHED REFERENCE DAPS OF THE STUDY AREA

Listed below are the principal reference maps of the study area on Ellesmere Island, as denoted by their index designations in the GSC and NTS referencing systems.

Geological Maps:

Canada. Geological Survey of Canada Map 1250A. Scale 1:5,000,000. Compiled by R.J.W. Douglas, 1967.

Detailed Geology Maps of Study Area Encompasses Section Localities Baumann Fiord. Geological Survey of Canada Map 1312A. Scale 1:250,000. Compiled by J.W. Kerr and R. Thorsteinsson, 1971.

Eureka Sound South. Geological Survey of Canada 1, la, 2, 3 Map 1300A. Scale 1:250,000. Compiled by R. Thorsteinsson, 1971.

Strathcona Fiord. Geological Survey of Canada Map 1307A. Scale 1:250,000. Compiled by R. Thorsteinsson, 1971.

Eureka Sound North. Geological Survey of Canada Map 1302A. Scale 1:250,000. Compiled by R. Thorsteinsson, 1970. Carion Fiord. Geological Survey of Canada Map 4, 7,8 1308A. Scale 1:250,000. Compiled by R. Thorsteinsson, 1971.

Maps that do not encompass sections measured in this study are included in order to provide for a complete geological picture of the area. All such listed maps do encompass Baumann Fiord Formation outcrop. 224

Sawyer Bay. Geological Survey of Canada Open 5 File 111. Scale 1:125,000

Bobbin Bay. Geological Survey of Canada Open 6 File 111. Scale 1: 125,000

Maps:

Each of the seven geological maps noted above has a counterpart of the physiography, all with scale — 1:250,000, countour interval — 500 feet. They encompass the whole of the study area and are vital for access to the area and navigation within it. Numberical designation is that of the National Topographic System.

Baumann Fiord NTS 49C

Eureka Sound South NTS 49F

Strathcona Fiord NTS 49E

Eureka Sound North NTS 49G

Carion Fiord NTS 49H

Sawyer Bay NTS

Bobbin Bay NTS 39H and 29G

Other vans.

Tectonic Map. Geological Survey of Canada Map 1251A. Scale 1:5,000,000. Compiled by the Tectonic Map of Canada Committee (C.H. Stockwell, chairman).

Isototopic Age Map. Geological Survey of Canada Map 1256A. Scale 1:5,000,000. Compiled by R.K. Wanless, 1969.

All of- the maps listed above are available through the Geological Survey of Canada (Publications):

Geological Survey of Canada, Institute of Sedimentary and Petroleum Geology, 601 Booth Street, 3303-33ra Street, N.W., Ottawa, Ontario, Canada. Calgary, Alberta, Canada. 225

APPENDIX B

SECTION LITHOLOGIO LOGS

Contained in the rear insert pocket of this volume are lithologic logs of the eight sections of the Baumann Fiord Formation measured in the course of this study. Logs of Sections l—la-2-3 (denoted Figure 32) transect a line of section perpen dicular to the region strike in the Trold Fiord area. Logs of Sections 4-7-5-6 (Figure 33) transect a line of section roughly parallel to the onlap erosional edge of the formation in the stable Arctic Platform region. All sections are related to what is thought to be the most reliable time—line datum in the formation, the base of the B Member.

Section Localities

1, 84°401 W 78°22'N la 84°25'14 78°22111 2 84°24IW 78°061N 3 84°05'w 78:13'N 4 80°40'w 79°15'N 5 78°08'W 79°15'N 6 74°50'w 79°08'N 7 80°16'w 79°09'N 8 82°1111i 79°14'N 226

APPENDIX C

SULPHUR ISOTOPE DATA FOR BAUMANN FIORD FORMATION SULPHATE ROCKS

The following sulphur isotope determinations were carried out by the United States Atomic Energy Commission (Dr. H.R. Adler, Washington D.C.). Analyses were based upon 20 gram hand samples of Baumann Fiord Formation gypsum and anhydrite rocks.

s34 o/00 Sample Number

GM-7-2 +28.2 GM-7-4 +27.6 GM-7-6 +28.7 GM-7-8 +30.9 GM-7-10 +28.3 GM-7-12 +25.4 GM-7-14 +24.0

GM-8-1 +28.0 GM-8-3 +26.0 GM-8-12 +27.0 GM-8-13 +28.2 GM-8-14 +28.2 GM-8-15 +30.4 GM-8-16 +29.2 GM-8-17 +27.6

32 Within theS /IS34 spectrum for marine sulphate deposits, the above determinations "agree very well with the published data on Ordovician sulphate, particularly with the data provided by Thode and Monster (1964) on the sulphate of the Red River Formation, Saskatchewan" (H. Adler, pers. comm.)

Samples were derived from localities 7 and 8, as denoted in the middle digit of each sample number 227

APPENDIX D ral PALEONTOLOGICAL IITT1'07.1,IATION ON THE BAIMANN FIORD FORMATION

Macrofauna

Field No. & StrtIzElEla Locality. FPuna & Age o o Baumann Fiord Fm., 79 09 N, 80 16 W, Witch Mountain, 125-142 ft. above Sverdrup Pass. base of B Member indeterminate gastropods Section 7 fragments of trilobites GSC Locality No. C-19559 condonents Polytoechia sp. age: late Early Ordovician, Late Canadian or early Middle Ordovician (Whiterock)

Comments

The genus Polytoechia is not well known, most of the described species are late Early Ordovician in age, but an undescribed species similar to the present material has been collected by Christie (C-3307) from rocks on Devon Island that probably are Whiterock in age.

Determinations and comments by B.S. Norford (Institute of Sedimentary and Petroleum Geology, Calgary).

Condononts

Total number of specimens: 257 Locality as above; GSC C-19559

Condonents were sorted and identified as follows (the number of specimens noted in parentheses). All are form taxa.

Acodus oneotensis Furnish (9) Acontiodus staufferi Furnish (2) Drepanodus arcuatus Pander (25) D. sp. cf. D. homocurvatus Lindstr8m (16) D. subcuatus 228

D. sp. cf. D. suberectus Branson & Mehl (2) D. n. sp. (4) Oistodus sp. cf. O. inclinatus Branson & Mehl (4) O. sp.- aff. O. lanceolatus Lindstrom (22) O. linlaatus Lindstr8m (4) Scandodus pipa Lindstr8m (9) S. n. sp, (2) Scolopodus cornutiformis Branson & Mehl (56) S. gracilis Ethington & Clark (48) S. quIL1,2111a1L15 Branson & Mehl (27) Stolodus. (= Distacodus) stola Lindstr8m (5) Ulrichodina sp. (1)

Taxonomic note: Acodus oneotensis, Oistodus sp. aff. 0. lanceolatus, Scolopodus Llacilis, and. Stolodus stola are interpreted widely for each represents a transition series for which additional form species have been applied in the past. In terms of multi—element species, D. sp. cf. D. homocurvatus,, D. sp. cf. D. suberectus, and 0. sp. cf. O. inclinatus belong within the single natural species of D. sp. cf. D. homocurvatus.

Comments

The large :conodont fauna is characterized by scolopodid and drepanodid elements characteristic of platform facies of the Lower Ordovician in North America. A few elements, notably Oistodus linguatus and Stolodus stola, are typical components of Lower Ordovician faunas of the North Atlantic province.

Conedents were picked and mounted by T.T. Uyend (Institute of Sedimentary and Petroleum Geology, Calgary). Identifi- cations and comments are by C.R. Barnes (University of Waterloo, Ontario). 229

APPENDIX E

FRANKLINIAN OVERBURDEN ABOVE THE BAUMANN FIORD FORMATION — TROLD FIORD REGION

In the Trold Fiord region, the Baumann Fiord Formation was buried by more than 20,000 feet of superincumbent Franklinian rocks. Some of the overburden is preserved uneroded at the present day. That load which is thought to have been removed from the area is here projected as being equivalent in thickness to strata preserved some 50 miles south along strike, on Bjorne Peninsula. Thickness data is that recorded by Thorteinsson and Tozer (1970) — Sections 24 (Trold Fiord) and 25 (Bjorne Penin- sula):

Awe Formation Thickness (ft.)

Ellesmerian Orogeny Okse Bay Fm. Equivalents 9,950 Bird Fiord Fm. Equivalents 2,400 Devonian 12,350 TOTAL THICKNESS OF BJORNE PENINSULA EQUIVALENTS ______12,350 Cape Rawson Fm. 500+ Silurian Cape Phillips Fm. 3,100 Irene Bay Fm. 1,000 Thumb Mountain Fm. 1,300 Ordovician Bay Fiord Fm. 2,100 Eleanor River Fm. 1,500

Baumann Fiord 9,500+ TOTAL THICKNESS OF EXISTING OVERBURDEN 9,500+ TOTAL PROJECTIO THICKNESS OF FRE—ELLESMERIAN OVERBURDEN ABOVE 21;850+ TIE', BAUMANN FIORD FORMATION - TROLD FTORDAEGTON.

230

APPENDIX F

STROI:TIM CONT=TS IN BAUMANN FIORD FORMATION ANHYDRITE AND GYPSUM ROCKS

Trace strontium analyses were carried out by atomic absorption methods. Material analysed included anhydrite rock, alabastrine gypsum rock (latin situ hydration), surfacial efflorescent gypsum, porphyroblastic secondary gypsum (early in situ hydration), and some satin—spar gypsum. All values are expressed in parts—per—million strontium:

Sample Weathering Product Surficial Efflor- /section No.) Alabastrine Gypsum escent Gypsum 26b(1) 2752 2558 29f(la) 2262 2740 30a(la) 2794 3000 33d(1) 1749 * 37b(1) 2687 2380 3091 39e(la) 2721 2535 * 43c(la) 2762 1965 2991 52a(2) 2741 2082 * 57a(2). 2209 1831 3265 * 65a(3) 2650 2208 2941 72b(3) 2481 2887 79c(4) 2362 2793 * 102a(5) 2644 1931 2986 105b(5) 2763 105c(5) 2692. 2226 111a(6) 2408 2977 * 115c(7) 2641 1994 2760 117d(7) 2740 2322 231

Porphyroblastic Secondary 122212_ipection No. Gypsum Satin—Spar

76b(4) 2256 760(4) 3143 107b(5) 1994 118c(7) 2006 4016 119a(7) 2331 2976

* Denotes analyses that include original anhydrite, its alabastrine gypsum hydration product, and the surficial efflorescent gypsum (see text section 5.4).

+ Denotes instances where porphyroblastic gypsum and satin—spar determinations were obtainable from the same hand specimen., LITHOLOG IC LOGS OF BAUMANN FIORD FORMATION SECTIONS 4-7-5-6

PRINCIPAL LITHOLOGY ACCESSORY L ITHOLOGY (10 % TO 400f0) ':1 LIMEST ONE SANDY " , " , " r , , C ALCAREOUS ANHYDR I TE " , , " , , -4 r r r DOLOMITIC r r GYP SUM r r r r r [- .. r... _. PROMINENT CHERT CONC RET IONS SANDSTON E iii ARGILLACEOUS Bache , FOLOEO ~~:c~ Peninsula ANHYDRI TIC , " " " , MEMBER DESIGNATION , 400- -c " , " LEVEL BELOW TOP OF FORMATION " -1 ( FE ET ) Bay Fiord PRECAMBRIAN ~~ COVERED SHiElD [>(

Oe -ELEANOR RIVER FORMAT ION STABLE BElT OeD - CO PES BAY FORMATION (OR E OUI VALENTS )

SECTION 4 SECTION 7

Oe Oe .t:; 00 SECTION 5

X1\ f. "

S ECTIO N 6 O e , " c c , " , '" 100

Oe --...,.....--..,,/ 00 " " " >< , , , / , , , ~ lr _· 10 0 c , " " c

~~ 100 T~"" ~~~ : B B B B t8~~ ~ h-1l~ = ~ ~~ 200

" "...... " . .." ;..,. ." " ,,'"" 300 , , ' , , , ~ ..." .. . ': . .. ~ ... :-... " " , , ~ " '1.= ••, ::' • ,• • !' •••, h--+ ,'- 400 I 400 r; -;- "l - " " '" ", .. .': .• ~• .• !'. ~ J 300 , ..!. , 1 . f:=Cr

, " A , , A .,. ~ . ,.. t ..., A A , , 400 ;..' " :'" . ~ . 500 A A , , .., ...... , , . 500 , , , ~ . ~' " '

- - - - .

A 600 A A A

r - ~

500

A " • • , .~• • • t'• . A , , •• • :'• •• t' •• 7 ? 7 " " , , , " ...... 600 " , , 700 " r , :.., .. ';.' .. '.,: •• • ':' ••• t'.• \ i ~ " "._-: " , " A ,

" , '\ \ " , ',, " ',.: ' ' ;'" ...., ...,. .. , , , \ 600 ...... , , .. , , " , , " r r r r r r r r rr I ~ 70 0 800 ~ /

, " , ~ - r " r r r r r 700 r r r r r r r r r r r r r r -

r r r r r - r - r 800 r 90 0 r r r - O e D r OC O r