The Open Collections website will be unavailable July 27 from 2100-2200 PST ahead of planned usability and performance enhancements on July 28. More information here.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The geomorphology and permafrost conditions of Garry Island, N.W.T. Kerfoot, Denis Edward 1969

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1970_A1 K47.pdf [ 29.51MB ]
Metadata
JSON: 831-1.0102152.json
JSON-LD: 831-1.0102152-ld.json
RDF/XML (Pretty): 831-1.0102152-rdf.xml
RDF/JSON: 831-1.0102152-rdf.json
Turtle: 831-1.0102152-turtle.txt
N-Triples: 831-1.0102152-rdf-ntriples.txt
Original Record: 831-1.0102152-source.json
Full Text
831-1.0102152-fulltext.txt
Citation
831-1.0102152.ris

Full Text

THE GEOMORPHOLOGY AND PERMAFROST CONDITIONS OF GARRY ISLAND, N.W.T. by DENIS EDWARD KERFOOT B.Sc. , King's College, University of London, 1961 M,A», University of B r i t i s h Columbia, 1964 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In the Department of Geography We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1969 In present ing th is thesis in pa r t i a l f u l f i lmen t of the requirements for an advanced degree at the Un ivers i t y of B r i t i s h Columbia, I agree that the L ibrary sha l l make i t f r ee l y ava i l ab le for reference and Study. I fur ther agree that permission for extensive copying of th is thes is for s cho la r l y purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or pub l i ca t i on of th is thesis for f i nanc i a l gain sha l l not be allowed without my wr i t ten permiss ion. Department of CSgo&fefifiV^V-The Un ivers i ty of B r i t i s h Columbia Vancouver 8, Canada Date Ho>i«.rrkV)ejp ^ . A^kT. i i ABSTRACT Garry Island, approximately 11 kilometres (7 miles) long and 0.8 to 3.2 kilometres (0.5 to 3.2 miles) wide, i s located at about l a t i t -ude 69° 28'N and longitude 135° 42'W in the southern part of the Beaufort Sea. The stratigraphy consists mainly of unconsolidated sands, s i l t s , clays and stony clays which have been intensively deformed by the thrusting action of glacier-ice moving from the south. The deformed sed-iments are l o c a l l y overlain by undisturbed sands and gravels containing marine f o s s i l s dated at>42,000 years. The absence of any evidence of g l a c i a l t i l l on top of the sands suggests that Garry Island lay beyond the northwestern l i m i t s of the Laurentide ice sheet during the late-Wisconsin glaciation. Elevated strand-lines, which may be of great antiquity and occur at approximately 7.5 metre (25 feet) intervals to an altitude of almost 46 metres (150 f e e t ) , indicate the extent of Pleistocene fluctua-tions of sea level and the drowning of a pre-existing topography. The development of tundra polygons, in small f l a t s behind sandspits or bars b u i l t across the drowned valleys i n association with the former sea l e v e l s , has imparted a d i s t i n c t i v e , stepped longitudinal p r o f i l e to the stream courses. The tundra vegetation of Garry Island is c l a s s i f i e d into ten major habitats which are primarily related to drainage conditions and type of geomorphic a c t i v i t y . The island is underlain by permafrost and the thickness of the active layer i s greatest, and ground temperatures i n this layer are highest, beneath unvegetated surfaces and where the substrate is composed predominantly of mineral s o i l . i i i S t ratigraphic, geomorphic and h i s t o r i c evidence indicates considerable recession of the coastline i n recent times. Current rates of retreat, reaching maxima of 10.5 metres (35 feet) per annum, are primarily related to the composition of the permafrost, being greatest i n areas of fine-grained sediments, containing high ice contents, with a southerly exposure. Thermal erosion of the permafrost i s the dominant process i n -fluencing c l i f f retreat and the primary role of wave action, on a short term basis, i s i n the removal of thawed debris from the base of the c l i f f s . Observations of three highly active mudslumps, created by the exposure of segregated ground i c e , show that the rate of headwall recession i s strongly correlated with ambient a i r temperatures. Maximum recession occurs where the ice content i s high and the slumped debris i s frequently removed from the base of the scarp. The c y c l i c development of a gully system on the ice face i s described. The longevity of mudslump a c t i v i t y i s prolonged where strong mudflows carry the thawed material away from the foot of the headwall, thus preventing the progressive b u r i a l of the scarp face. Mudflow v e l o c i t i e s reveal a rhythmic pulsation related to periodic blocking of their channels. Mud: levees, bordering the mudflows, r e s u l t from the progressive bleeding of moisture from, and subsequent stagnation of, the mud rather than as residual features pushed aside by the advancing mudflow. Patterned ground on Garry Island i s primarily r e s t r i c t e d to non-sorted types. Angular intersections of thermal contraction cracks, representing the inc i p i e n t stages of tundra polygons, exhibit a preferred tendency toward s l i g h t l y - o r i e n t e d , orthogonal systems. The i n i t i a l micro-r e l i e f of earth hummocks i s believed to originate through the accentuation of a miniature desiccation/frost- crack pattern. Following the establishment of a vegetation cover, t h e i r subsequent growth involves further d i f f e r e n t i a l f r o s t a c t i o n and s o l i f l u c t i o n . S t a t i s t i c a l tests show that the height, si z e and shape of earth hummocks are c l o s e l y r e l a t e d to t h e i r p o s i t i o n on the slope p r o f i l e . V ACKNOWLEDGMENTS It is impossible to give f u l l credit to a l l the people who gave assistance in the preparation of this thesis. Singled out for particular thanks are:- Dr. J. Ross Mackay, my advisor, whose teaching stimulated ray interests in the Canadian Arctic and who has been a con-stant source of inspiration in the classroom and in the f i e l d ; Dr. J. K. Stager for his assistance in the f i e l d and encouragement and advice in the preparation of the text; Dr. H. 0. Slaymaker and Dr. W. H. Mathews for their advice and comments on the text; Dr. M. A. Melton and my fellow graduate students for seminar discussions which enabled me to test and cla r i f y my thoughts; Dr. J. D. Ives and the staff of the former Geographical Branch, Ottawa, for the generous support of the f i e l d programme through the provision of funds, equipment, laboratory f a c i l i t i e s and a base map of Garry Island; Dr. J„ G. Fyles, of the Geological Survey of Canada, for advice in the f i e l d , radiocarbon dates and, with his colleague Dr. F. J. E. Wagner, identification of the f o s s i l specimens; Dr. E„ Hulten and Dr. He Personn of the Naturhistoriska Riksmusset, Stockholm, for identifications of the plant species; Mr. R. M. H i l l and the staff of the Inuvik Research Laboratory for their unfailing support in the f i e l d area; Mr. R. Reynolds and the staff of the Print Shop, Brock University, for their assistance in the printing of the thesis; Mr. P. J. Tighe and Mr. W. B. Windjack of the Audio Visual and Photographic Department, Brock University, for their assistance with the photographic illustrations; and last, but by no means least, to my wife, Helen, for the innumerable occasions that I was able to depend on her in the roles of f i e l d assistant, geographer, cartographer and secretary. August, 1969. Denis E. Kerfoot. v i i TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION 1 The Scope of the Study 4 Current Geomorphic Processes 4 The Role of Permafrost 7 Geomorphological History 8 The Study Area 10 H i s t o r i c a l Significance 12, Climate 14 I I . STRATIGRAPHY 22 Areas of Deformed Sediments 24 Sand Headland Areas 29 The Extent of the Marine Submergence 35 Longitudinal Stream P r o f i l e s 50 Summary 61 I I I . VEGETATION 63 Vegetation Types 65 I. Dryas-Hummock Type 65 I I . Alnus crispa Type 70 I I I . Cassiope Snowpatch Type 71 IV. Eriophorum Tussock Type 71 V. Hummock-Tussock Transition Type 73 VI. Stream Course Willow Thickets 73 v i i i CHAPTER PAGE VII. Sedge-Moss F l a t s 75 VIII. Mudslump Communities 75 IX. Strand Communities 78 X. Polygonal Ground 83 Summary 84 IV. PERMAFROST CONDITIONS 88 Depth of Thaw Measurements 90 Earth Hummock Experiments 100 Ground Temperature Measurements 108 Freeze-Back i n the Active Layer, 1964 118 Summary 122 V. GEOMORPHOLOGICAL PROCESSES 125 Coastal Recession 125 Evidence of Coastal Recession 126 Coastal Stake Measurements 131 Relevance of the P r o f i l e Studies to other Coastal Areas 144 Processes at work i n Coastal Recession 145 Muds lumps 149 Rates of Retreat 154 Ablation Studies 163 The Recession Process and Evolution of the Gully System 171 The Mudslump Cycle 178 i x CHAPTER PAGE Mudflows 184 Rates, of Movement . 191 Mud Levees 201 Patterned Ground 203 Tundra Polygons 204 Incipient Frost Crack Patterns 206 Low-Centred Polygons 217 High-Centred Polygons 225 Thermokarst Features 228 Earth Hummocks 232 The Structure of Earth Hummocks 234 The Size and Form of Earth Hummocks 254 The Origin and Development of Earth Hummocks 272 Summary 277 VI. SOME OBSERVATIONS ON THE GE0M0RPH0L0GICAL EVOLUTION OF GARRY ISLAND 282 Pleistocene Deposits and Glaciation 282 The Extent of the Marine Transgression 289 Current Geomorphological Processes 293 The Concept of a P e r i g l a c i a l Morphogenetic Region 296 BIBLIOGRAPHY 299 X APPENDIX I 305 A. L i s t of vascular species found on Garry Island 305 B. L i s t of Bryophytes found on Garry Island 308 x i LIST OF TABLES TABLE PAGE I. Mean Daily Temperatures for Aklavik, Inuvik, Tuktoyaktuk, Shingle Point and Nicholson Peninsula (Degrees Centigrade). 16 I I . Mean Monthly P r e c i p i t a t i o n Totals for Aklavik, Inuvik, Tuk-toyaktuk, Shingle Point and Nicholson Peninsula (Cms.) . . 18 I I I . Comparisons of Temperature and P r e c i p i t a t i o n Data i n the Delta area for the months of July-August, 1964-1966 . . . . 20 IV. Elevations of the breaks of slope on the longitudinal p r o f i l e s of Garry Island streams 55 V. Depths of Thaw beneath the major vegetation types, Garry Island, September 1, 1964 92 VI. Depths of Thaw beneath Earth Hummocks on Garry Island, July 19 - August 30, 1965 102 VII. Earth Hummock watering experiment, Garry Island, July 26 - August 30, 1965 106 VIII. Ground Temperatures recorded at different depths i n Mud, B o i l , Earth Hummock and Inter-Hummock Depressions (°C) 112 IX. Sand Headland P r o f i l e Changes 141 X. Mudslumps - Rates of Retreat (Metres) 156 XI. Rates of Recession at selected stake positions i n Slump B (Metres) 177 XII. Ice Contents (Weight of Ice to Dry S o i l ) of samples taken from exposure of ground ice i n Slump B 183 XII I . Coefficient of Viscosity of a Garry Island Mudflow 199 x i i TABLE PAGE XIV. Analysis of Garry Island Earth Hummock Data -Heights (Cms.) 263-264 XV. Analysis of Garry Island Earth Hummock Data -Sizes (Cms.) 266-267 x i i i LIST OF FIGURES FIGURE PAGE 1. Garry Island: Topography 11 2. Garry Island: Stratigraphy 23 3. Logarithmic Grain. Size D i s t r i b u t i o n Diagrams 27 4. Grain Size D i s t r i b u t i o n for s o i l samples taken from the Sand Headlands and areas of Deformed Sediments 32 5. Garry Island: Raised Shoreline Features 43 6. Stratigraphic section exposed i n coastal b l u f f s along northwest coast of Garry Island 47 7. Longitudinal Stream P r o f i l e s (1) 52 8. Longitudinal Stream P r o f i l e s (2) 53 9. Cross-Valley P r o f i l e s - Stream 'G' 59 10. Garry Island: Vegetation Types . 66 11. Vegetation Plot: Surface Contours above datum and Vegetation Types 95 12. Vegetation P l o t : Position of the Frost Table August 1 and September 1, 1964 97 13. Vegetation Plot: Depth of Thaw P r o f i l e s 98 14. Mean Ground Temperatures and Ground Temperature Fluctuations 111 15. Ground Temperature patterns i n an. Earth Hummock and adjacent Depression (°C) . . . . . 115 16. Ground Temperature patterns i n a Mud B o i l 116 17. Coastal Recession Features: Northwest coast of Garry Island 130 18. Sand Headland P r o f i l e s (1) 138 x i v F I G U R E P A G E 1 9 . S a n d H e a d l a n d P r o f i l e s ( 2 ) 1 3 9 2 0 . G a r r y I s l a n d : D i s t r i b u t i o n o f M u d s l u m p s 1 5 1 2 1 . M u d s l u m p : S l u m p B 157 2 2 . R e l a t i o n s h i p b e t w e e n r a t e o f r e t r e a t a n d n u m b e r o f t h a w i n g d e g r e e - d a y s f o r t h r e e G a r r y I s l a n d M u d s l u m p s 159 2 3 . A b l a t i o n S t u d i e s i n S l u m p B - P r o f i l e I 1 6 6 2 4 . A b l a t i o n S t u d i e s i n S l u m p B - P r o f i l e I I 167 2 5 . R e l a t i o n s h i p b e t w e e n a b l a t i o n r a t e s a n d n u m b e r o f t h a w i n g d e g r e e - h o u r s i n S l u m p B 1 7 0 2 6 . G a r r y I s l a n d M u d s l u m p P r o f i l e , 1 9 6 4 - 1 9 6 6 1 7 9 2 7 . R e c e s s i o n D i a g r a m o f a G a r r y I s l a n d M u d s l u m p , 1 9 6 4 - 1 9 6 6 . 1 8 0 2 8 . A G a r r y I s l a n d M u d f l o w 1 8 9 2 9 . M u d f l o w - R a t e s o f M o v e m e n t 1 9 3 3 0 . I n c i p i e n t F r o s t C r a c k P a t t e r n , K e n d a l l I s l a n d , N . W . T . . . 2 1 1 3 1 . D i a g r a m m a t i c s k e t c h e s o f F r o s t C r a c k i n t e r s e c t i o n p a t t e r n s 2 1 4 3 2 . S u r f a c e C o n t o u r s a n d C r o s s - S e c t i o n a l P r o f i l e o f a t y p i c a l L o w - C e n t r e d P o l y g o n 2 1 9 3 3 . L o w - C e n t r e d P o l y g o n N e t w o r k 2 2 1 3 4 . L o w - C e n t r e d P o l y g o n N e t w o r k o n S a n d H e a d l a n d s 2 2 3 3 5 . S u r f a c e C o n t o u r s a n d C r o s s - S e c t i o n a l P r o f i l e o f a t y p i c a l H i g h - C e n t r e d P o l y g o n 2 2 7 3 6 . E a r t h H u m m o c k l o c a t e d o n u p p e r p a r t o f s l o p e p r o f i l e : S u r f a c e C o n t o u r s a n d S t r u c t u r a l P r o f i l e s 2 3 6 3 7 . E a r t h H u m m o c k l o c a t e d o n u p p e r p a r t o f s l o p e p r o f i l e : V e g e t a t i o n T y p e s 2 3 7 X V FIGURE PAGE 38. Earth Hummock located on middle part of slope p r o f i l e : Surface Contours and Structural P r o f i l e s 241 39. Earth Hummock located on middle part of slope p r o f i l e : Vegetation Types 242 40. Earth Hummock located on lower part of slope p r o f i l e : Surface Contours and Structural P r o f i l e s 244 41. Earth Hummock located on lower part of slope p r o f i l e : Vegetation Types 245 42. Structural P r o f i l e of an Earth Hummock located at the foot of a slope 246 43. Mud B o i l : Surface Contours 246 44. Mud B o i l : Vegetation Types 248 45. Earth Hummock Stripes: Surface Contours and Structural P r o f i l e s 250 46. Earth Hummock Stripes: Vegetation Types 251 47. Slope p r o f i l e s showing locations of Earth Hummock sampling stations 256 48. Histograms of Hummock Heights - P r o f i l e I 258 49. Histograms of Hummock Heights - P r o f i l e I I 259 50. Histograms of Hummock Heights - P r o f i l e I I I 260 51. Histograms of Hummock Heights - P r o f i l e IV 261 52. Histograms of Hummock Heights - P r o f i l e V 262 x v i LIST OF PLATES PLATE PAGE I. Stratigraphy 25 . I I . Raised Shoreline.Features 40 I I I . Vegetation Types 67 IV. Vegetation Types 74 V. Vegetation Types 81 VI. Coastal Recession 128 VII. Coastal Recession 133 VIII. Sand Headland P r o f i l e Studies 136 IX. Mudslumps 152 X. Mudslumps 164 XI. A Garry Island Mudflow 190 XII. Mudflow: Surge Phenomena 195 XI I I . Incipient Frost Crack Patterns. 208 XIV. Low-Centred Polygons 220 XV. High-Centred Polygon andThermokarst Features 226 XVI. Earth Hummocks 238 XVII. Earth Hummock Stripes 252 CHAPTER I INTRODUCTION The a r c t i c and s u b a r c t i c r e g i o n s of Canada have i n c r e a s i n g l y a t t r a c t e d the a t t e n t i o n o f s c i e n t i s t s i n the i n t e r v a l s i n c e the end of the Second W o r l d War. P r i o r t o 1945, r e l a t i v e l y l i t t l e a t t e n t i o n was p a i d t o these remote n o r t h e r n l a n d s , but the post-war r e a l i z a t i o n of t h e i r s t r a t e g i c s i g n i f i c a n c e and attempts t o e s t a b l i s h and d e v e l o p t h e i r economic r e s o u r c e p o t e n t i a l , have c o n t r i b u t e d g r e a t l y t o s t i m u l a t e a l l a s p e c t s of r e s e a r c h i n these h i g h l a t i t u d e s . Much v a l u a b l e d a t a has been o b t a i n e d as a 'by-product' o f the c o n s t r u c t i o n and o p e r a t i o n of the m i l i t a r y i n s t a l l a t i o n s and the e x p l o r a t o r y s u r v e y s s e e k i n g t o l o c a t e p o t e n t i a l m i n e r a l d e p o s i t s . One consequence o f the a v a i l a b i l i t y o f t h i s d a t a i s t h a t c o n s i d e r a b l e p r o g r e s s has been made i n the f i e l d of a r c t i c geomorphology d u r i n g the pa s t few decades. T h i s same time i n t e r v a l , s i n c e 1945, has a l s o w i t n e s s e d the growth and c o n s o l i d a t i o n of a number of new approaches t o the s u b j e c t o f geomorphology. T r a d i t i o n a l l y , g e o m o r p h o l o g i c a l s t u d i e s of land s c a p e development have r e l i e d h e a v i l y on q u a l i t a t i v e d e s c r i p t i o n . Such d e s c r i p t i o n s formed the b a s i s o f geomorphology d u r i n g the 19th c e n t u r y , c u l m i n a t i n g i n the c y c l i c c o n cepts of W i l l i a m M o r r i s D a v i s , and most of the s t u d i e s u n d e r t a k e n d u r i n g the f i r s t h a l f o f t h i s c e n t u r y . I n the pa s t two t o t h r e e decades however , attempts have been made t o e s t a b l i s h the s u b j e c t on a more p r e c i s e , o b j e c t i v e f o u n d a t i o n by s u b s t i t u t i n g q u a n t i t a t i v e measurements f o r v e r b a l d e s c r i p t i o n s . A l t h o u g h the a d o p t i o n 2 of t h i s quantitative approach to Landform studies i s s t i l l i n i t s infancy, a number of p r i n c i p a l avenues of research can be i d e n t i f i e d . These include the compilation and tabulation of data r e l a t i n g to the scale and shape of landforms; investigations of the mode of operation and i n t e r -relationships between degradational and aggradational forces, including their expression i n the forms of models, graphs or formulae; and measurements of the actual rates at which geomorphic processes are oper-ating on various parts of the earth's surface today. To date, most, though not a l l , of the success achieved by the employment of these methods has been i n the humid temperate regions of the world, and i n the f i e l d of f l u v i a l geomorphology where the drainage basin forms a convenient unit study area. The adoption of the quantitative approach to geomorphological studies r e f l e c t e d , i n part, a growing d i s s a t i s f a c t i o n with the subjective, genetically-oriented Davisian model of landscape development and the scant attention paid to process studies i n t h i s model. Additional reactions against the Davisian approach to the subject have also resulted i n increasing attention being paid to alternative theories of landscape development. Thus, objections to the long period of c r u s t a l s t a b i l i t y required for the production of a peneplain, have led to a r e v i v a l of the ideas of Penck and an emphasis on the significance of c r u s t a l mobility i n the formation of landforms. Other geomorphologists have concentrated on the role of climate, and the belief that the rate of operation of geo-morphological processes varies considerably from one climatic region to another. A considerable amount of interest i n a r c t i c geomorphology has been generated by the development and elaboration of t h i s concept of 3 climatic geomorphology.1 In this concept i t i s postulated that there i s a very close relationship between climate and geomorphology, to the extent that under a given set of climatic conditions certain geomorphic processes w i l l predominate, and these i n turn w i l l lead to the development of a charac t e r i s t i c assemblage of landforms. On the basis of these postulates, some proponents of the concept have further suggested that the influence of climate i s such that a series of morphogenetic regions can be i d e n t i f i e d i n which the topographic characteristics of an area can be differentiated from those of other areas developed under contrasting climatic regimes. In North America, the concept of climatic geomorphology was developed by P e l t i e r who tentatively i d e n t i f i e d nine morphogenetic 2 regions. For one of these regions he adopted the term p e r i g l a c i a l , to describe those parts of the earth's surface which have an annual temper-ature range of 5-30°F and an average annual r a i n f a l l range of 5-55 inches, and i n which the geomorphic processes are characterized by strong mass movement, moderate to strong wind action and a weak effect of running 3 water. Although the p e r i g l a c i a l environment i s currently r e s t r i c t e d to polar latitudes and high a l t i t u d e s , i t has attracted wide attention because many features, which have been described from more temperate l a t -itudes, have tentatively been interpreted as evidence of a s i m i l a r , more widespread environment during the Pleistocene period. B i r d , J . Brian (1967) The Physiography of A r c t i c Canada, The Johns Hopkins Press, Baltimore, Maryland, p. 158. 2 P e l t i e r 9 L„C. (1950) "The geographical cycle i n p e r i g l a c i a l regions as i t i s related to climatic geomorphology", Ann. Assoc. Amer. Geog., Vol. 40, pp. 214-236. 3 I b i d . , p. 215. 4 One of the major shortcomings of P e l t i e r ' s paper was that i t was construed within the framework of the t r a d i t i o n a l Davisian approach to geomorphology. As such, and as P e l t i e r himself r e a l i z e d , i t was r e s t r i c t e d to qualitative description and was based on inadequate data. Further discussion of the v a l i d i t y and lim i t a t i o n s of the concept of morphogenetic regions need not be debated here, save to mention that before i t can be given additional credence i t w i l l have to be established on a quantitative basis through detailed geomorphic process studies. As yet, however, there has been only limited research into the p e r i g l a c i a l processes operating i n 4 northern Canada. The number of process studies i s steadily increasing, but many more are needed before a r c t i c geomorphology can be f u l l y i n t e r -preted, and only as a r e s u l t of such investigations can the proper r o l e of variations i n the climatic regime be evaluated. The Scope of the Study. The aims of this thesis are to make a threefold contribution to the general f i e l d of a r c t i c geomorphology: 1. To study some of the geomorphic processes currently operating i n an a r c t i c environment on Garry Island, N.W.T. 2. To examine the role of permafrost i n the operation of these geomorphic processes. 3. To attempt to decipher the complex geomorphological history of the outer Mackenzie Delta area i n g l a c i a l and post-glacial times. Current Geomorphic Processes. The s p e c i f i c geomorphic pro-cesses investigated were those involved i n coastal recession, mudslumps B i r d , J . Brian (1967) op. c i t . , p. 157. 5 and associated mudflows, and the genesis of certain types of patterned ground. This l i s t is by no means exhaustive of a l l the contemporary processes operating on Garry Island. The most obvious omission is the process of sol i f l u c t i o n which is undoubtedly one of the most familiar and widespread agencies moulding the landscape in arctic latitudes and has been more intensively investigated than most of the other geomorphic pro-cesses. For this reason, an examination of the comprehensive aspects of solifluction was excluded from the programme of studies on Garry Island. Observations of the rate of coastal recession were made on approximately 2.5 kilometres (1.5 miles) of coastline along which stakes were installed during the summer of 1964. Most of these stakes were located along the exposed northwest coast of the island, and measurements of the amount of recession that had taken place were made at the beginning and end of each of the f i e l d seasons. To supplement these observations, and provide additional data on the processes involved in c l i f f retreat, four profile stations were established on the prominent sand headlands along the north coast of the island. At each of these stations wooden stakes were driven into the c l i f f face, normal to the surface, and these were surveyed periodically throughout one of the field^seasons to detect profile changes. Mudslumps are created by the melting out of large bodies of ground ice, and the fine-grained sediments, containing substantial masses of segregated ice, which underlie much of Garry Island provide favourable sites for the development of these features. A series of stakes was installed around three highly active mudslumps, and measurements of the amount of retreat were made throughout each of the f i e l d seasons. A programme of ablation studies was also carried out on the ice face 6 exposed in the headwall of one of the mudslumps. At intervals of two to three weeks, a number of stakes was installed normal to the ice face and the amount of ablation was measured daily for five or six consecutive days each time. Attempts were made to correlate the data on the rates of headwall retreat and rates of ablation with meteorological observations recorded at a small weather station established on the island. Further observations on the manner in which a mudslump headwall retreats were obtained through investigations of the seasonal evolution of a distinctive gully system which, while i t exists, imparts a miniature 'badland' topog-raphy to the ice face. These process studies were combined with observations on mudflows to describe a model of the cyclic development of mudslumps and the factors which influence the longevity of this cycle. Mudflows are the most effective agents by which thawed debris is removed from the foot of an actively retreating mudslump headwall. An active mudflow was surveyed in detail, and markers were installed on i t s surface to determine the rate of movement and the nature of the flow pattern. Samples of mud were collected to try to correlate variations in the velocity of the mudflow with changes in the viscosity of the mud. Detailed measurements and excavations of the mud levees bordering both old and active mudflows were made to determine variations ln the height, slope, symmetry, composition and mode of origin of these features. The studies of patterned ground were restricted to investig-ations of some of the more controversial aspects of the development of tundra polygons and earth hummocks. Particular attention was paid in the tundra polygon studies to the incipient frost crack stage, and the nature of the angular intersections of the cracks which can only be measured with any reliable degree of accuracy at this stage. Several areas of polygonal 7 ground were also surveyed in detail to examine some of the factors i n -fluencing the size and spacing of the polygonal units and the transition from low- to high-centred forms. Most of the literature pertaining to earth hummocks consists of very generalized, descriptive statements concerning their shape and size, and a part of the f i e l d programme was aimed at replacing these qualitative statements with quantitative data. Preliminary observations of the hummocks on Garry Island suggested that their size and spacing were related to their position on the slope profile. Five slope profiles were surveyed and a series of six survey lines was established from each profile running parallel to the contours. Using a horizontal sight line, the height and spacing of each hummock and depression on each of the survey lines were recorded, and sufficiently large samples were obtained so that any differences in the calculated means could be tested s t a t i s t i c -a l l y . At the same time observations were made relating to the alignment and profiles of the hummocks, and these were combined with a series of excavations to determine the structure and mode of origin of these micro-r e l i e f features. The Role of Permafrost. The second contribution of the thesis is closely a l l i e d to the f i r s t and is related to the concept of climatic geomorphology and, in particular, the role of the permafrost conditions. The various geomorphic processes operating in arctic regions today may be peculiar to the northlands, or they may simply differ in degree from those shaping the landscape in more southerly latitudes. Any such differences, either in kind or degree, may possibly be related to the presence of permafrost, the role of which is imperfectly understood. The thickness and composition of the permafrost, and particularly the presence or 8 absence of large bodies of segregated ground i c e , have a d e f i n i t e influence on the occurrence or non-occurrence and the rates of operation of certain geomorphic processes. Of particular significance to the genesis of some of the microrelief features i n high latitudes are the conditions i n the active layer. This i s the layer, extending from the ground surface down to the permafrost table, which undergoes seasonal thawing during the summer months. A knowledge of the changes i n the thermal regime of this active layer i s relevant to a l l attempts to develop the economic potential of the northlands, as wel l as to a f u l l e r understanding of ar c t i c ; geomorphology. To date, however, r e l a t i v e l y few measurements of ground temperature patterns i n the active layer have been recorded i n northern Canada. Accordingly, the second aim of this thesis i s to examine the influence of the composition of the permafrost on the rates of operation of the geomorphic processes, and some of the factors which contribute to variations i n the ov e r a l l thickness and thermal regime of the active layer. Geomorphological History. The t h i r d contribution of the thesis i s an attempt to decipher the complex geomorphological history of the outer Mackenzie Delta area i n g l a c i a l and post-glacial times. This latest epoch of geological time witnessed immense expansions of glaciers i n middle and high l a t i t u d e s , the extent of which fluctuated i n response to clim a t i c changes. The waxing and waning of the ice sheets was accompanied by profound o s c i l l a t i o n s i n the l e v e l of the sea. Large tracts of the earth's crust were depressed by the weight of these ice masses, and the subsequent melting of the ice contributed to r i s i n g sea levels and an extensive sub-mergence of the coastal lowlands. At a later date these same lowlands emerged from beneath the sea as the earth's crust slowly responded to the 9 removal of the recently melted ice cover. The number of g l a c i a l and i n t e r g l a c i a l episodes, and their orderly arrangement into a chronological sequence, i s s t i l l imperfectly understood i n the a r c t i c regions of Canada. There i s no doubt that the area i n the v i c i n i t y of the Beaufort Sea was affected by glaciers advan-cing from the south, but there are differences of opinion as to the number of advances that affected the area and their relationship to the c l a s s i c a l Pleistocene sequence established for more southerly latitudes. Similar controversies exist i n the interpretation of evidence indica t i v e of changing patterns of land-sea relationships and the extent of the post-g l a c i a l marine transgression. The i d e n t i f i c a t i o n and interpretation of t h i s evidence, and an attempt to arrange the sequence of events into a chronological order, constitutes the t h i r d aim of t h i s thesis. The selection of Garry Island as a f i e l d study area was at the advice of Dr. J . Ross Mackay, my advisor, on the basis of his extensive knowledge of the Mackenzie Delta area. The f i e l d work was carried out during the periods June 25 - September 8, 1964, June 22 - September 13, 1965 and July 26 - August 29, 1966. I am deeply indebted to the former Geographical Branch, Department of Mines and Technical Surveys, Ottawa, for the generous provision of funds and equipment i n support of each of the f i e l d seasons. I also gratefully acknowledge support for the f i e l d programme which was received by Dr. J . Ross Mackay from the Department of Northern A f f a i r s and National Resources, by way of the Committee on A r c t i c and Alpine Research, the University of B r i t i s h Columbia, and from research funds of the University of B r i t i s h Columbia. 10 THE STUDY AREA Garry Island i s centred at latitude 69° 28' N and longitude 135° 42' W i n the southern part of the Beaufort Sea (Figure 1). I t l i e s at the d i s t a l end of the Mackenzie Delta; a low, f l a t area approximately 80 kilometres (50 miles) wide and 160 kilometres (100 miles) long."5 The modern delta i s characterized by a myriad of i n t e r l a c i n g channels and small lakes, and Garry Island forms part of an arcuate chain of islands which represent the seaward remnants of a formerly more extensive P l e i s t -ocene, or e a r l i e r , ancestor of the modern delta. The island i s approximately 11 kilometres (7 miles) long, oriented i n a northwest-southeast d i r e c t i o n , 0.8 to 3.2 kilometres (0.5 to 2 miles) wide, and reaches elevations exceeding 46 metres (150 feet) above sea l e v e l . These elevations, and those of adjacent islands, are consider-ably higher than the heights found i n the modern delta of the Mackenzie River, and they owe their altitude to the development of ice-thrust features and the growth of substantial bodies of segregated ground i c e . The major topographic features of Garry Island, which may be described as 6 gently r o l l i n g , are shown i n Figure 1. Extensive f l a t summit areas are lacking and the higher ground i s characterized by smooth slopes seldom The metric standard of measurement i s used i n this thesis. However, on some of the maps, i f the f i e l d surveys were done using B r i t i s h measures, the B r i t i s h units are used. The topographic map of Garry Island shown i n Figure 1 was compiled from a e r i a l photographs by the former Geographical Branch, Department of Mines and Technical Surveys, Ottawa. Although this map accurately portrays the general features of the topography, and has there-fore been reproduced with only minor modifications, the absence of any accurate ground height control implies that the positions of, and the numerical values assigned to, the form lines are only estimates. 12 exceeding 5-10 degrees. The high ground i s broken by a series of shallow valleys and depressions, the sides of which may have slopes of 25-35 degrees. On the north side of the island, three prominent aprons of coarse sand produce f l a t to gently-sloping surfaces ranging from 7.5-15.0 metres (25-50 feet) above sea l e v e l . Drainage conditions over much of the island are poor and r e f l e c t the presence of permafrost at shallow depths beneath the ground surface. Despite the presence of a number of f a i r l y well-defined stream courses, integrated drainage patterns are poorly developed. The present channels serve primarily as conduits for surface runoff derived from melting snow and the thawing of the active layer during the spring and early summer. Throughout the rest of the summer these channels are kept moist by seepage from the thawing ground, but surface runoff i s generally lacking except for a short time following exceptionally prolonged periods of heavy r a i n f a l l . The fl o o r s of the depressions, p a r t i c u l a r l y at lower elevations, and parts of the stream courses are further characterized by the development of polygonal ground with associated pond and marsh areas. H i s t o r i c a l Significance. Despite i t s small s i z e , Garry Island has played a controver-s i a l r o l e i n the history of the exploration and mapping of t h i s north-western section of the Canadian A r c t i c . In 1789, Alexander Mackenzie completed his epic voyage down the Grand River,^ subsequently renamed i n his honour, at a small island which he named 'Whale Island' after the numerous beluga, or white whales, which he observed i n the surrounding Stager, J.K. (1965) "Alexander Mackenzie's exploration of the Grand River", Geographical B u l l e t i n , Vol. 7, pp. 213-241. 8 waters. In 1825, S i r John Franklin journeyed down the same r i v e r and named the s i t e of his most northerly camp i n the delta Garry Island i n 9 honour of his friend the Deputy Governor of the Hudson's Bay Company. Franklin included a map of the delta i n the account of his voyage and although he did not see 'Whale Island', he located i t on his map using Mackenzie's l a t i t u d i n a l and longitudinal observations. 1^ Consequently, for more than a century, Garry Island and 'Whale Island' appeared adjacent to one another on Canadian topographic maps of the Mackenzie Delta area u n t i l the f i r s t accurate maps were produced from photographs taken during a e r i a l reconnaissances of the delta flown during the Second World War. 1 1 Since these maps f a i l e d to reveal any presence of land at the position described by Mackenzie, the name 'Whale Island' was rescinded by the Can-12 adian Board on Geographical Names i n 1960. There are s t r i k i n g s i m i l a r i t i e s i n the respective explorers' descriptions of 'Whale Island' and Garry Island with respect to their s i z e , elevation above sea l e v e l and panoramic vistas of adjacent islands 8 Mackenzie, A. (1801) Voyages from Montreal on the River St. Laurence through the continent of North America to the Frozen and P a c i f i c  Oceans i n the years 1789 and 1793, T. Cadell, Jun. and W. Davies, Strand; Cobbett and Morgan, P a l l M a l l ; and W. Creech at Edinburgh. Reprinted editi o n (1966) i n March of America Facsimile Series, Number 52, University Microfilms Inc., Ann Arbor, Michigan, pp. 64-65. 9 Frank l i n , S i r John (1828) Narrative of a^  second expedition to the shores of the Polar Sea i n the years 1825, 1826 and 1827, Carey, Lea and Carey, Philadelphia, p. 49. 10 I b i d . , Map Frontispiece. 11 Bredin, T.F. (1962) "'Whale Island' and the Mackenzie Delta: charted errors and unmapped discoveries, 1789 to 1850", A r c t i c , Vol. 15, p. 52. 12 Anon. (1960) "Geographical Names i n the Canadian North", A r c t i c , Vol. 13, p. 143. 14 i n the delta and the front of the Richardson Mountains. The major dis-crepancies are i n the l a t i t u d i n a l positions of the two islands, and the compass directions of the topographic features i n the panoramas. In attempting to reconstruct the o r i g i n a l route of Mackenzie's voyage through the delta to the coast, Bredin discovered a number of errors i n the dis-tances and directions as reported by Mackenzie. The most s i g n i f i c a n t of these errors are a consistent recording of the latitudes south of their actual positions, and recorded compass directions of t r a v e l always more westerly than his true directions, even allowing for magnetic v a r i a t i o n s . 1 I f due consideration i s taken of these factors, the revised location of 'Whale Island' coincides, almost i d e n t i c a l l y , with that of Garry Island. A similar conclusion, using somewhat d i f f e r i n g c r i t e r i a , was also reached by Mackay, by combining Mackenzie's descriptions of natural features with 14 his own detailed knowledge of the conditions i n the delta. Thus, the enigma of 'Whale Island' has been solved, for there i s l i t t l e doubt that the 'Whale Island' of S i r Alexander Mackenzie and the Garry Island of S i r John Franklin are one and the same island. Climate. There are f i v e operating meteorological stations i n the v i c i n i t y of the Mackenzie Delta, but few of these have records which ex-tend back over a large number of years. F a i r l y continuous records are available for the town of Aklavik, from 1926 onwards, but, with the estab-lishment of Inuvik, they were terminated i n 1959. Data for the Inuvik Bredin, T.F. (1962) op. c i t . , pp. 52-53. 14 Mackay, J , Ross (1963) "The Mackenzie Delta area, N.W.T.", Geographical Branch Memoir, No. 8, p. 6. 15 airport are available from 1958. The most continuous records of the e x i s t i n g meteorological stations, dating from 1948, are those of Tuktoyak-tuk. With the establishment of the Distant Early Warning (DEW line) system, additional weather-recording sit e s were provided at Shingle Point, i n the Yukon Territory to the west; Tununuk, at the southern end of Richards Island; Atkinson Point and Nicholson Peninsula, approximately 80 and 160 kilometres (50 and 100 miles) northeast of Tuktoyaktuk respect-i v e l y . 1^ The sit e s at Tununuk and Atkinson Point have subsequently been abandoned, and the records for Shingle Point and Nicholson Peninsula date back to the summer of 1957. With the lone exception of Inuvik, the published observations at each of the meteorological stations are r e s t r i c t e d to p r e c i p i t a t i o n t o t a l s and temperature extremes. The lengths of the records are, i n most cases, too short to provide t r u l y r e l i a b l e means and, consequently, the values presented i n the following tables should be interpreted accordingly. Furthermore, there i s the problem of the variable length of the records at each of the stations: 31 years at Aklavik, 21 years at Tuktoyaktuk, and 11 years at Inuvik, Shingle Point and Nicholson Peninsula. These two factors, lack of r e l i a b i l i t y and non-comparability of the means, are fundamental to any discussion of the regional climate. The mean daily temperatures for each of the meteorological stations i n the Mackenzie Delta area are shown i n Table I. Mean daily temperatures are below freezing for eight months of the year. January i s the coldest month for most of the stations, except for Tuktoyaktuk and ->For locations of the place-names see the World Aeronautical Chart, I.C.A.O., 1:1,000,000. Sheet 2062, F i r t h River (1967) included at the back of the thesis. 16 TABLE I MEAN DAILY TEMPERATURES FOR AKLAVIK, INUVIK, TUKTOYAKTUK, SHINGLE POINT AND NICHOLSON PENINSULA (Degrees Centigrade). Aklavik Inuvik Tuktoyaktuk Shingle Point Nicholson Peninsula Jan. -27.7 -30.9 -28.1 -26.1 -25.9*(3) Feb. -27.1 -27.5 -29.2 -24.3*(3) -27.9*(2) Mar. -22.7 -23.9 -26.4 -26.1 -25.8*(4) Apr. -13.0 -13.3 -18.6 -17.5 -.18.8 May - 0.5 - 0.8 - 5.2 - 4.1 - 6.0 June 9.4 10.2 5.0 5.1 4.2 July 13.6 13.9 10.1 10.3 7.1 Aug. 10.1 10.4' 9.2 8.8 7.4 Sept. 3.4 3.1 1.8 1.7 0.4 Oct. - 6.8 - 7.6 - 7.2 - 7.7 - 7.8 Nov. -19.4 -20.8 -21.3 -20.3 -21.1 Dec. -26.9 -28.2 -25.6 -23.8 -25.9 * Unfortunately the mean daily temperatures are only intermittently reported for these months, and the means are of correspondingly less value. The figures i n brackets indicate the number of years records on which these means are based. Source: Canada, Department of Transport, Meteorological Branch, Monthly Record(s): Meteorological Observations i n Canada, Queen's P r i n t e r , Ottawa. 17 possibly Nicholson Peninsula, where the average February temperatures are colder. The t r a n s i t i o n from winter to summer temperatures i s quite rapid. As Table I shows, mean daily temperatures increase by 26-28°C between A p r i l and J u l y , which i s usually the warmest month. Using a climatic d e f i n i t i o n of the a r c t i c , the mean July temperatures of approximately 10°C (50°F) for the warmest month at Tuktoyaktuk and Shingle Point i n d i c -ate a location on the boundary between a r c t i c and subarctic climates despite their locations w e l l to the north of the t r e e - l i n e . The trans-i t i o n from summer to winter temperatures i s as rapid, with the onset of sub-freezing temperatures again i n late-September or early-October. The mean monthly p r e c i p i t a t i o n t o t a l s for these same stations are shown i n Table I I . Annual t o t a l p r e c i p i t a t i o n i s low, averaging 14-19 cms. (5-7 inches) at the coastal locations and increasing to 25-28 cms. (10-11 inches) further inland. Each of the stations records a summer maximum, i n the form of r a i n , during July and August,, and t h i s maximum i s more pronounced at the coast with 40-50 per cent of the annual t o t a l occurring i n these two months, compared to less than 30 per cent further inland. Approximately one-half of the p r e c i p i t a t i o n i s i n the form of snow. Garry Island, occupying an intermediate position between Tuktoyaktuk and Shingle Point, probably has an annual temperature and p r e c i p i t a t i o n pattern similar to these stations, with the coldest month being January or February and below-freezing temperatures for eight months of the year. Light snowfalls were encountered during August and September i n 1964 and 1965, but the snow did not persist for any length of time. Total snowfall i s probably quite l i g h t , and the thickness of the cover i s related to the action of the wind. Much of the snow i s swept from the 18 TABLE I I MEAN MONTHLY PRECIPITATION TOTALS FOR AKLAVIK, INUVIK, TUKTOYAKTUK, SHINGLE POINT AND NICHOLSON PENINSULA (Cms.)-Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Total Aklavik 1.68 1.50 1.12 1.52 1.40 2.13 3.53 3.66 2.39 2.46 2.21 1.22 24.82 Inuvik 2.39 1.42 1.35 2.16 1.30 2.18 4.60 2.97 1.78 3.71 1.96 1.80 27.62 Tuktoyaktuk 3.45 1.02 0.23 0.91 0.84 1.30 3.33 4.09 1.30 0.99 0.61 0.56 18.63 Shingle Point 0.64 0.18 0.15 0.66 0.51 2.49 4.65 3.25 1.37 3.81 0.71 0.18 18.60 Nicholson Peninsula 0.18 0.25 0.25 0.43 0.48 1.88 2.64 4.14 1.45 1.45 0.41 0.25 13.81 Source; Canada, Department of Transport, Meteorological Branch, Monthly Record(s): Meteorological Observations i n Canada, Queen's P r i n t e r , Ottawa. 19 exposed slopes and summit areas and i t i s pi l e d into thick d r i f t s i n the depressions and against the coastal b l u f f s . A climate station was established on Garry Island, for the duration of each f i e l d season,, i n conjunction with observations on var-iations i n the rate and depth of thaw of the active layer at selected s i t e s on the island. Meteorological records from t h i s station permit the only v a l i d comparisons with the published records of the other stations i n the delta area. The results of these comparisons are shown i n Table I I I . The temperature data contained i n Table I I I demonstrate the modifying effect of proximity to the Beaufort Sea. Mean temperature values for each of the summer months on Garry Island were consistently 2-3°C cooler than those experienced at Inuvik. The fact that the Garry Island temperatures also averaged 1-2°C cooler than those at Tuktoyaktuk and Shingle Point probably r e f l e c t s i t s insular character. A comparison of the temperature extremes recorded during the same time period exhibits the same features. Although there was l i t t l e or no discernible pattern i n the recorded minima, the maximum temperatures for each month at Inuvik were consistently 2-4°C warmer than those recorded on Garry Island. The values for the monthly p r e c i p i t a t i o n t o t a l s do not reveal much of a pattern though, with the exception of Shingle Point, the amounts recorded at the coastal stations were less than further inland at Inuvik. The number of hours of bright sunshine was also recorded on Garry Island during the 1964^  and 1965 f i e l d seasons. An average of 321 hours of bright sunshine was recorded i n the month of July and 197 hours i n August. These t o t a l s were 7 and 22 hours less than Inuvik respectively. Although each station averaged four days without sunshine i n the month of July , the corresponding figures for August for Garry Island and Inuvik 20 TABLE I I I COMPARISONS OF TEMPERATURE. AND PRECIPITATION DATA IN THE DELTA AREA FOR THE MONTHS OF JULY - AUGUST, 1964 - 1966. Mean Daily Temperatures (Degrees Centigrade) 1964 1965 1966 1964-66 July Aug. July Aug. July Aug. July Aug. Garry Island 9.0 7.3 9.3 7.3 9.7 5.9 9.3 6.8 Tuktoyaktuk 8.3 8.2 11.6 9.9 11.1 7.1 10.3 8.4 Nicholson Pen. 5.7 5.8 9.5 8.2 8.5 5.3 7.9 6.4 Shingle Point 8.8 8.2 9.8 9.2 11.7 7.1 10.1 8.2 Inuvik 11.6 10.2 13.9 10.7 14.3 9.6 13.3 10.2 Daily Temperature Extremes (Degrees Centigrade) 1964 1965 1966 July Aug. July Aug. July Aug. Garry Island Max. 26.7 22.5 24.6 20.0 26.7 20.0 Min. -3.3 -1.1 1.9 -2.8 -0.6 -0.6 Tuktoyaktuk Max. 26.7 22,8 23.3 21.1 25.0 . 20.6 Min. -1.7 -0.6 1.7 0.6 1.1 -0.6 Nicholson Pen, Max. 27.2 22.2 25.0 19.4 25.6 18.9 Min. -7.8 -4.4 0.0 0.0 -1.1 -3.3 Shingle Point Max. 27.8 21.7 26.1 23.9 27.8 21.1 Min. -3.9 0.0 2.2 -4.4 1.1 -1.7 Inuvik Max. 28.9 23.9 27.2 24.4 29.4 23.9 Min. -2.2 3.9 -1.1 -1.7 -3.3 -6.1 Monthly P r e c i p i t a t i o n Totals (Cms.) 1964 1965 1966 1964-66 July Aug. July Aug. July Aug. July Aug. Garry Island 0.6 1.3 _ _ 0.7 0.2 . _ _ Tuktoyaktuk 2.9 1.7 3.1 4.4 0.6 1.1 2.2 2.4 Nicholson Pen. 0.4 •2.5 T 2.4 0.7 1.6 0.4 2.2 Shingle Point 7.3 1.5 3.5 5.9 2.2 2.2 4.3 3.2 Inuvik 6.2 1.9 4.6 9.0 3.3 1.4 4.7 4.1 Source: Data for Tuktoyaktuk, Nicholson Peninsula, Shingle Point and Inuvik taken from Monthly Record(s): Meteorological Observations  i n Canada, Queen's P r i n t e r , Ottawa,, 21 were nine and three days respectively. Comparisons of wind speeds and directions showed that the dominant winds during each of the summer months on Garry Island were from the northwest and east, the same as Inuvik. Periods of calm were r e l a t -i v e l y rare on the island however, and the mean wind v e l o c i t i e s for each month, 8-10 m.p.h., were consistently 2-3 m.p.h. higher than those recorded at the inland station. CHAPTER I I STRATIGRAPHY The purpose of this chapter i s to describe the major s t r a t i -graphic units occurring on Garry Island and the evidence for changes i n the former r e l a t i v e positions of land and sea. This material w i l l then be used, i n the f i n a l chapter, i n an attempt to decipher the geomorpho-l o g i c a l history of the island i n late- and pos t - g l a c i a l times. Garry Island i s the westernmost member of an arcuate chain of islands, which includes adjacent P e l l y , Kendall, Hooper and Pullen Islands, located i n the outer part of the Mackenzie Delta. These islands, together with most of Richards Island, the Tuktoyaktuk Peninsula, a coastal fringe along the south side of the Eskimo Lakes, and an area stretching north and northeast of the Caribou H i l l s , represent the discontinuous remnants of one or more deltas constructed by a Pleistocene, or e a r l i e r , ancestor 1 of the modern Mackenzie River. The stratigraphy consists e n t i r e l y of a sequence of unconsolidated sands, gravels, s i l t s , clays and stony-clays, cemented by i c e , i n which many of the beds have been deformed from their o r i g i n a l position. Figure 2 i s a map showing the major stratigraphic features of Garry Island. This map i s based on a limited number of clean exposures, mainly i n wave-cut b l u f f s , which have not been obscured by the combined effects of s o l i f l u c t i o n processes and slumping, or mantled beneath a thin veneer of g l a c i a l t i l l . 1 Mackay, J . Ross (1956) "Mackenzie Deltas - A Progress Report", The Canadian Geographer, No. 7, pp. 3-7. Figure 2 GARRY ISLAND - STRAT IGRAPHY 24 Areas of Deformed Sediments. The oldest sediments found on Garry Island are those represented i n a series of discontinuous exposures i n a number of mud-slumps and coastal bluffs along the south and west coasts of the island. The following description outlines the major stratigraphic and structural features observed i n a series of transects along the l a t t e r coastline where active slumping and coastal recession have resulted i n the greatest number of clean exposures. The f i r s t transect A-B (see inset, Figure 2), includes approximately 900 metres (3,000 feet) of coastline i n which a sequence of sands, s i l t s and clays has been uncovered i n a number of active mudslumps. The clays contain large quantities of segregated ground ice which has a d i s t i n c t i v e banded appearance caused by an alternation of bands of frozen ground, with a high s i l t and clay content, and clear ice* These sediments, which dominate the stratigraphic sequence, are intercalated with beds of fine sand, 4.5-9.0 metres (15-30 feet) thick, containing an abundance of twigs, washed wood, bone fragments and s h e l l s . Most of the shells have been severely crushed during deformation of the s t r a t a , but two complete specimens, indicating a marine depositional environment, were i d e n t i f i e d as species of Portlandia a r c t i c a (Gr ay) by Dr. F.J.E. Wagner of the Geological Survey of Canada. A l l the sediments i n this transect have been deformed (Plate I.-A) , and measurements taken from both the segregated ice bands and the sand beds show a remarkable degree of consistency in their attitude, being t i l t e d to the southwest at angles ranging from 30-75 degrees. The transect B-C traverses a series of high bluffs reaching maximum heights of approximately 29 metres (95 feet) above sea l e v e l . Samples were taken systematically from each of the major stratigraphic 25 Plate I S T R A T I G R A P H Y A. Deformed body of segregated ground i c e . Banded structures dip to the southwest at angles of 60-65 degrees. B. Shear plane developed i n s i l t y - c l a y . C. G r a n i t i c g l a c i a l e r r a t i c at an e l e v a t i o n of approximately 85 feet above sea l e v e l on the south side of Garry Island. D. S t r u c t u r a l (?) r i d g e at an e l e v a t i o n of 60-65 feet above sea l e v e l on the south side of Garry Island. 26 units exposed i n these b l u f f s , with the exception of those occurring i n the mudslumps near C, and were analysed to determine their granulometric composition. The results are i l l u s t r a t e d graphically i n Figure 3A. The sediments are primarily s i l t s and s i l t y - c l a y s , with occasional beds of sandy-silt and, l i k e the sediments i n the previous transect, they exhibit many signs of intensive deformation. Bedding planes, where preserved, are t i l t e d and occasionally contorted into a series of gentle folds. The most salient feature of the deformation however is the occurrence of large shear planes with well-preserved slickensided surfaces which cut obliquely across, and l o c a l l y o f f s e t , many of the o r i g i n a l bedding structures (Plate I-B). These shear planes, spaced at intervals ranging from a few centimetres to several metres, are frequently concave up and in places extend almost to the top of the b l u f f s . Their presence produces a much more complex pattern of deformation than that described in the previous transect, and the attitude of the individual beds often changes rapidly over very short distances. Further evidence of the complexity of the stratigraphic relationships i n this section was found in the sequential examinations of constantly changing exposures at the same location as a result of continued marine erosion and slumping. Seldom, i f ever, did these new exposures repeat i d e n t i c a l l y the pattern which had been observed at an e a r l i e r date. Nevertheless, the general pattern of deformation i s similar to, albeit less consistent than, that found i n the previously described transect, with the strata maintaining dips approximately to the southwest. Where the shear planes are concave, the orientation of the concavity i s also towards the southwest. The deformed sediments are buried i n the section C-D (Figure 2) beneath a series of beach and lacustrine deposits, the Figure 3 L O G A R I T H M I C G R A I N S I Z E D I S T R I B U T I O N D I A G R A M S B. S A N D H E A D L A N D S 28 significance of which w i l l be discussed below. The deformed sediments reappear again further along the coast but most of the exposures are obscured by s u r f i c i a l slump deposits. Two measurements were recorded i n deformed ground ice bodies, exposed i n a large mudslump, where i t appears that the beds have a north or northwesterly dip of 20-30 degrees. Exposures i n the high b l u f f s along the south coast of the island are very limited due to the r e s t r i c t e d active recession of the c l u f f s and their mantling by debris. Two observations made in deformed beds of s i l t and clay, and one i n a small exposure of ground i c e , indicate a generally northeasterly dip of 15-25 degrees. Only one exposure of deformed sediments was recorded along the whole of the north coast, where a band of vein ice i n a mudslump appears to be gently folded and plunges almost due north at an angle of about 9 degrees. Direct evidence of deformation over the remainder of the island i s lacking, but a e r i a l photographs reveal the presence of a few marked lin e a t i o n features which may be s t r u c t u r a l l y controlled. Ground checks of these features however f a i l e d to confirm whether or not they were d e f i n i t e l y related to any underlying structures. Similar deformation of Pleistocene or e a r l i e r sediments has been recorded along adjacent sections of the mainland coast, from Herschel Island i n the west to the Nicholson Peninsula i n the east; a 2 distance exceeding 485 kilometres (300 miles). Three possible mechanisms - slumping, tectonic disturbance and ice thrusting - were ^Mackay, J. Ross (1956) "Deformation by Glacier Ice at Nicholson Peninsula, N.W.T., Canada", A r c t i c , Vol. 9, pp. 219-228. Mackay, J. Ross (1960) "Glacier Ice Thrust features of the Yukon Coast", Geographical B u l l e t i n , No. 13, pp. 5-21. presented to explain the deformation patterns. Although the conditions on Garry Island are often very conducive to mudslump development, the scale of the deformation, the r e g u l a r i t y of the pattern and the presence of deformation features.; i n areas where there i s no sign of past or present slump a c t i v i t y , are considered to be s u f f i c i e n t c r i t e r i a for eliminating this mechanism as a satisfactory explanation of the deforma-tion pattern. The most satisfactory interpretation of the deformation pattern i s that i t i s the result of the overriding action of g l a c i e r - i c e ; the same mechanism which produced the disturbed features on the adjacent mainland. Direct evidence of the g l a c i a t i o n of Garry Island i s d i f f i c u l t to assess, but the stony nature of the active layer suggests that many slopes are mantled by a thin veneer of g l a c i a l t i l l . G l a c i a l e r r a t i c s found on the slopes, i n s o l i f l u c t i o n deposits and along the beaches include granites, gneisses, quartzites, sandstones and slabs of f o s s i l i -ferous (Devonian ?) limestone. Large e r r a t i c s , up to 3 metres (10 feet) across, are found close to the highest summits of the island (Plate I-C),, but since the whole island l i e s below the upper marine l i m i t (see below), the p o s s i b i l i t y that they may have been ice-rafted cannot be excluded. Sand Headland Areas. Exposures on the north coast of the island (Figure 2) are dominated by thick deposits of sand which terminate i n coastal b l u f f s 6-11 metres (20-35 feet) high. The precise areal extent and thickness of the sands are d i f f i c u l t to determine since very few contacts with the •^Mackay, J. Ross (1960) "Glacier Ice Thrust features of the Yukon Coast", Geographical B u l l e t i n , No. 13, p. 5. 30 underlying sediments are exposed. The headland surfaces are gently sloping and devoid of any major r e l i e f features, and they contrast markedly with the more rugged topography of the area of deformed sediments to the south. This change i n the character of the l o c a l r e l i e f occurs at elevations of approximately 12-15 metres (40-50 feet) above present sea l e v e l . The sands are brown i n colour, horizontally s t r a t i f i e d with occasional signs of current bedding, and consist mainly of sands i n the medium- to fine-grain size category (Figure 3B) with inclusions of gravel lenses. The sands are not deformed, except for l o c a l upturning along the lines of the more prominent ice-wedges, and there i s no evidence of any g l a c i a l deposition on any of the headland surfaces. The sands contain iron-stained twig fragments and an abundant marine molluscan fauna. The following species were i d e n t i f i e d by Dr. Wagner: Astarte borealis Schumacher Astarte montagui (Dillwyn) Astarte montagui (Dillwyn) forma typica Astarte montagui var. s t r i a t a (Leach) Astarte montagui var. warhami Macoma calcarea Gmelin My a truncata Linnl" Trachoma balthica Linne Tachyrhyncus sp. probably T. erosum Trichatropis sp. A sample of these f o s s i l s submitted for radiocarbon dating yielded an age of >42,600 years (G.S.C.- 562). Many of these shells are i n an excellent state of preservation and frequently have their outer valve cover (periostracum) in t a c t . Consequently, although no hinged specimens of the shells were found, and despite the fact that the sands also contain many broken s h e l l fragments, the f o s s i l s probably have not been reworked from e a r l i e r deposits, and the above date i s therefore interpreted to be a 31 r e l i a b l e indicator of the minimum time of sand deposition. The textural s i m i l a r i t y of a l l the sand exposures (Figure 3B) may indicate that they represent the discontinuous remnants of a formerly much more extensive sand pl a i n formed when the r e l a t i v e l e v e l of the sea was 12-15 metres (40-50 feet) higher than at present. The discontinuity of the bodies of sand may be p a r t i a l l y inherent, r e f l e c t i n g irregular deposition, or i t may be e n t i r e l y related to the effects of marine erosion subsequent to deposition. In either case, the main problem concerns the o r i g i n a l source of the sand-size material and, in p a r t i c u l a r , whether i t was derived l o c a l l y or had an extraneous o r i g i n . One possible mode of o r i g i n for the sands, implying a l o c a l source, i s that they represent the residual accumulations of coarse material r e s u l t i n g from long-continued marine erosion of the island. The coastline of Garry Island i s currently undergoing rapid recession i n many places, and there i s evidence that the shoreline has retreated a con-siderable distance during the past. A comparison of the two grain size d i s t r i b u t i o n diagrams i l l u s t r a t e d i n Figure 3 i s shown in the triangular graph of Figure 4, indicating r e l a t i v e percentages of material of sand-, s i l t - and clay-size p a r t i c l e s of samples taken from both the sand head-lands and the areas of deformed sediments. The l a t t e r have an average sand content of only 7 per cent, and only one per cent i s i n the medium and coarse sand fractions (<0.50 mm.)- The major exception to these figures i s the f o s s i l i f e r o u s sand beds described in transect A-B which have an average sand content of 91 per cent, 10 per cent of which i s i n the medium sand or coarser range. These values contrast with the average of 91 per cent sand-size material i n samples taken from the sand head-lands and an average of 26 per cent of material with a grain size greater 32 Figure 4 G R A I N S I Z E D I S T R I B U T I O N F O R S O I L S A M P L E S T A K E N F R O M T H E S A N D H E A D L A N D S A N D A R E A S O F D E F O R M E D S E D I M E N T S CLAY (<0.004mm.) 100 % SAND (>0.06 mm.) SILT (0.06 to 0.004mm.) o Samples from Sand Headlands • Samples from Deformed Sediments 33 than 0.50 mm. Whilst the number of samples upon which these figures are based i s small, they do indicate that the volume of coarse material i n the deformed sediments is quite minor and, i f the sands are to be in t e r -preted as a residual accumulation produced by long periods of marine erosion of the isl a n d , they must r e f l e c t the disappearance of very extensive tracts of older sediments. Some of the sand and gravel, of course, may have been derived i n a similar manner through erosion of the s u r f i c i a l mantle of g l a c i a l t i l l and the stony-clays, containing the segregated ice bodies, exposed in the mudslumps. A l t e r n a t i v e l y , the sand headland material may represent the residual deposits r e s u l t i n g from the erosion of stratigraphic units which are poorly represented i n the e x i s t -ing succession. I f the land, which has subsequently been eroded away by the sea, contained greater quantities of the fine sands described i n the transect A-B, or was mantled by a greater thickness of g l a c i a l deposits, including t i l l and/or outwash material l a i d down following the retreat of the g l a c i e r ( s ) , the problem of a source material would be greatly al l e v i a t e d . As a corollary to this mode of o r i g i n , which interprets the sands as beach or spit deposits, i t i s interesting to note the d i s t r i b u t i o n of lakes on the island. As Figure 2 shows, some of them occur along the contact zone of the sand bodies and the older, deformed sediments. The lakes may thus have originated as lagoon features on the landward sides of the sandspits, i n much the same way as lagoons are currently being formed i n association with the present day sandspits. The major d i f f i c u l t y encountered with this hypothesis i s that the present lagoons are shallow, averaging 1.5-2.0 metres (5-6 feet) i n depth, whereas the lake bottoms are 34 as much as 12-15 metres (40-50 feet) below the surfaces of the sand head-lands. Part of this d i f f e r e n t i a l may possibly be explained by the thermo-karst or warming action of the lake water on the underlying sediments, since the thawing of any ice contained i n these materials could have resulted i n a s e t t l i n g of the lake f l o o r . An alternative mode of o r i g i n for the sands, inf e r r i n g an extraneous rather than a l o c a l source, i s that they represent coarse material brought down by streams following the retreat of the ice . These deposits may have been l a i d down under marine conditions when the r e l a t i v e l e v e l of the sea was 12-15 metres (40-50 feet) higher than at present, or i t may be that the marine f o s s i l s represent a post-depositional phase during which the sands were reworked. Such an o r i g i n seems more probable than the above-mentioned l o c a l source on the basis of comparisons of the compositions of the sand headlands and the present day beach deposits. The l a t t e r contain abundant cobbles and boulders, derived from erosion of the g l a c i a l t i l l , whereas similar material forms a very minor constituent in the rather uniform composition of the sand headlands. Since much of the sand material may have been l a i d down under the influence of wave action and moulded into large s p i t s , some of the lakes may s t i l l have originated as lagoon features. A l t e r n a t i v e l y , the lakes may represent deeper parts of the former r i v e r channels as i n the case of the lake situated immediately south of the sand headland o u t l i e r on the northwest coast of the island. Yet another o r i g i n may be postulated for the deep, steep-sided lake incorporated within one of the headlands, which possibly may be interpreted as a k e t t l e feature. Unfortunately, there are no good exposures of sand around the shores of this lake to check t h i s hypothesis. 35 THE EXTENT OF THE MARINE SUBMERGENCE The coastal lowlands of a r c t i c Canada were influenced, to vary-ing degrees, by the fluctuations in sea l e v e l produced by the waxing and waning of the ice sheets. Attempts have been made to determine the precise l i m i t s of the extent of the marine transgression using the four c r i t e r i a of: (1) the highest altitude at which marine shells occur; (2) the highest altitude at which strand-lines are preserved; (3) the lowest altitude at which undisturbed ground moraines can be recognized; and (4) the lowest 4 altitude at which perched boulders are found. As a resul t of research conducted along these l i n e s , enough evidence has been accumulated to pro-vide an overview of the pattern and extent of the post-glacial marine transgression i n northern Canada. Reference to the G l a c i a l Map of Canada reveals that the area affected by the post-glacial marine transgression was much less extensive in the v i c i n i t y of the Beaufort Sea than i n other sectors of the Canadian a r c t i c . This undoubtedly r e f l e c t s the fact that much of the eastern a r c t i c was, i n general,'an area of thick ice cover, whereas the Mackenzie Delta area, being at or close to the northern l i m i t of ice advance, was blanketed by a much thinner ice sheet. I t may also r e f l e c t the fact that the eastern a r c t i c has been more intensively investigated and thus there is a r e l a t i v e paucity of observations from the west. Also of significance Sim, V.W. (1960) "Maximum post-glacial marine submergence i n northern M e l v i l l e Peninsula", A r c t i c , Vol. 13, p. 180. ^Geological Survey of Canada (1968) G l a c i a l Map of Canada. Farrand, W.R. and Gajda, R.T. (1962) "Isobases on the Wisconsin marine l i m i t i n Canada", Geographical B u l l e t i n , No. 17, pp. 5-22. 36 may be the fact that the evidence of a marine transgression i n this westernmost sector of the Canadian a r c t i c i s by no means as obvious as i n i t s eastern and central counterparts. The d i s t i n c t i v e f l i g h t s of strand-l i n e s , which form prominent features i n the landscape of the central and eastern a r c t i c , are generally lacking i n the Beaufort Sea area. Most of the stratigraphy consists of unconsolidated s i l t s and clays: materials which are ea s i l y eroded but are not conducive to. the development of prominent and persistent wave-cut b l u f f s . The lack of persistence r e f l e c t s the high s u s c e p t i b i l i t y of these same sediments to frost action and mass movement, especially s o l i f l u c t i o n , and the operation of these processes, throughout at least post-glacial time, has tended to obscure much of the evidence of former submergence and emergence, i f indeed i t ever occurred. Intimately associated with this tendency i s the evidence which suggests that the area i n the v i c i n i t y of the Beaufort Sea was not glaciated during the late-Wisconsin, or most recent, stage of the Pleistocene period, thus presenting a longer time i n t e r v a l during which the traces of former sea levels could be obliterated. Even when the evidence for a marine transgression has been deciphered, there are further problems involved i n establishing a chronological sequence i n the Mackenzie Delta area. This sequence must take into consideration the complex i n t e r -action of an ove r a l l post-glacial r i s e i n sea l e v e l which may possibly be coupled with two opposing forces i n the earth's crust; an upward or positive movement representing a response to the removal of the weight of the i c e , and a downward or negative movement representing crustal depression i n response to the weight of the accumulating d e l t a i c sediments brought down by the Mackenzie River i n recent times. Because the conditions that favour the development and 37 preservation of de f i n i t e evidence of a marine transgression are so poor, and the interpretation of this evidence i s so complex, i t i s d i f f i c u l t to determine, with any r e a l degree of precision, how extensive i t was. This i s r eflected i n the existing l i t e r a t u r e which contains a number of widely-contrasting opinions. Richards claimed that there was no evidence of any 6 post-glacial marine beaches i n the v i c i n i t y of the Mackenzie Delta. Mackay, on the other hand, has described a number of extensive estuarine r i v e r terraces which suggest a r e l a t i v e submergence of approximately 15 metres (50 feet) when the coastal area f i r s t became free of ice.'' The emergence of the land in the subsequent period may have been reversed more recently to be replaced by a period of submergence i n the order of 3-6 metres (10-20 f e e t ) . 8 The magnitude of these apparent changes i n the land-sea relationships i s quite small compared to the evidence, a l b e i t more dubious, reported from the area lying to the west of the delta. O'Neill reported that the finding of marine f o s s i l s i n high-level g l a c i a l deposits, and high terraces on the mountains facing the A r c t i c Ocean, indicated that g l a c i a l or post-glacial submergence of the A r c t i c coast 9 extended at least to 152 metres (500 feet) above sea l e v e l . Richards, H.G. (1950) "Postglacial marine submergence of A r c t i c North America with special reference to the Mackenzie Delta", Amer. P h i l . Soc. Proc. , Vol. 94, p. 36. 7Mackay, J . Ross (1963) The Mackenzie Delta area, N.W.T., Geographical Branch Memoir, No. 8, p. 39. 8 I b i d . , p. 47. 9 0 ' N e i l l , J . J . (1924) "Geology of the A r c t i c coast of Canada, west of the Kent Peninsula", Report of the Canadian A r c t i c Expedition, 1913-1918, Vol. X I , Geology and Geography, Part A, p. 18A. 38 More recent investigations, using f o s s i l evidence and the mechanical properties of sediments found i n excavations at the Engigstciak. archeological s i t e , near the mouth of the F i r t h River, suggest an a l t e r -native explanation for evidence of such high-level submergence.^ Although marine clays at t h i s s i t e are found at an a l t i t u d e of 207 metres (680 f e e t ) , and f o s s i l s i n these sediments indicate that they l i v e d i n water depths of more than 30 metres (100 f e e t ) , i t i s believed that the clays were transported to thei r present elevation by the thrusting action of glacxer-xce. East of the Mackenzie Delta there also appears to be evidence of a more extensive marine transgression than i s found i n the immediate v i c i n i t y of the delta. Mackay has described a series of elevated beaches and gravel terraces r i s i n g to 61 metres (200 f e e t ) , and O'Neill observed similar features at elevations reaching 67 metres (220 feet) above sea 12 l e v e l . Further north, the evidence of a marine transgression on Banks Island has been found at a number of l o c a l i t i e s i n the neighbourhood of 177-183 metres (580-600 f e e t ) . 1 3 10 Mackay, J. Ross, Mathews, W.H. and MacNeish, R.S. (1961) "Geology of the Engigstciak archeological s i t e , Yukon T e r r i t o r y " , A r c t i c , Vol. 14, pp. 25-52. 1 1 I b i d . , p. 47. 1 2Mackay, J. Ross (1958) "The Anderson River map area, N.W.T.", Geographical Branch Memoir, No. 5, p. 38. O'Neill, J . J . (1924) o£. c i t . , p. 33A. 1 3 P o r s i l d , A.E. (1955) "The vascular plants of the Western Canadian A r c t i c Archipelago", Nat. Mus. Can. B u l l . , No. 146, p. 188. Manning, T.H. (1956) "Narrative of a Second Defense Research Board Expedition to Banks Island, with notes on the country and i t s history", A r c t i c , Vol. 9, pp. 3-77. 39 The i d e n t i f i c a t i o n and interpretation of evidence indicating changes i n the pattern of land-sea relationships was part of the f i e l d study programme on Garry Island. However, few of the afore-mentioned c r i t e r i a for delimiting the extent of a marine transgression are readily applicable to the conditions on Garry Island. I t i s v i r t u a l l y impossible to employ the c r i t e r i o n of undisturbed ground moraine, since i t appears that the whole island was affected by submergence, and therefore one cannot use a comparative investigation of disturbed and undisturbed t i l l deposits to determine the upper l i m i t of the marine transgression. No evidence was found i n the form of perched boulders but, again, i f the island were completely submerged i n the past i t i s u n l i k e l y that any of these would have survived. The use of marine f o s s i l s as evidence of a marine invasion of the land i s also extremely limited. Even where this c r i t e r i o n has been adopted i n other areas, there are problems i n determining whether the f o s s i l s are 'in situ' 1 or whether they were transported to their present elevations as '.shelly d r i f t ' by the thrusting action of g l a c i e r - i c e . Insomuch as there i s abundant evidence that the l a t t e r process has pro-foundly affected the strata on Garry Island, i t would require careful consideration. However, due to the fact that the late-Wisconsin ice sheet did not cover the island, the long period of subaerial exposure has probably resulted i n the destruction of much of the f o s s i l i f e r o u s evidence by weathering processes. In any case, the only marine f o s s i l s found i n association with the raised shorelines were r e s t r i c t e d to elevations around 7.5-10.5 metres (25-35 feet) above sea l e v e l . The presence of raised shoreline features affords the most direct evidence of a former submergence of the island. Unfortunately, 40 however, the conditions on Garry Island are similar to those found i n other parts of the delta, and were generally unfavourable to the formation and preservation of these features. Most of the strand-lines were either only weakly developed, or evidence of their presence has been obscured, or even obliterated, by the operation of p e r i g l a c i a l geomorphic processes i n the i n t e r v a l since the withdrawal of the sea. As a general r u l e , the raised shorelines do not exhibit strong topographic expressions and seldom are they backed by prominent wave-cut b l u f f s (Plate II.-A). .In.fact, the only surface expression of many of the strand-lines consists of a f a i n t bevel, or break of slope, the presence of which might go completely undetected i f i t were not accentuated by contrasts i n the vegetation pattern. Since most of the stratigraphic units on Garry Island are com-posed of r e l a t i v e l y fine-grained s i l t s and clays, the raised shoreline features are rarely characterized by impressive shingle or boulder ridges. Where these are developed they constitute the exception rather than the rul e . Mechanical probing of the active layer, however, frequently revealed concentrations of coarse sand, gravel and well-rounded, iron-stained, pebbles at depths ranging from 30-60 cms. (1-2 feet) below the ground sur-face. These deposits, found i n association with the f a i n t breaks of slope on the topographic p r o f i l e , can often be traced quite extensively along the contours, but for distances of only a few metres i n either an upslope or downslope dire c t i o n . They are interpreted as residual beach deposits buried beneath a mantle of s o l i f l u c t e d material. The locations of the strand-lines are often marked by the development of a vegetation association which includes the tussock-like forms of the Sheathed Cotton-grass (Eriophorum vaginatum). RAISED Plate I I SHORELINE FEATURES A. Wave-cut b l u f f at an el e v a t i o n of approximately 75 feet above sea l e v e l . B. Sandspit associated with the 25 foot r a i s e d strand-l i n e . C. Peat accumulation damming the o u t l e t of a lake at an e l e v a t i o n of 120-125 feet above sea l e v e l . D. I n t e r r u p t i o n i n the long-i t u d i n a l p r o f i l e of Stream 'B' at an e l e v a t i o n of 100 f e e t above sea l e v e l . 42 The d i s t r i b u t i o n of these raised shoreline features i s shown i n Figure 5. Although many of them can be traced for considerable distan-ces, the incompleteness of the pattern r e f l e c t s : (1) the number of observations made; (2) the degree to which the strand-lines have been destroyed by the combined action of mudslump development and coastal re-cession associated with lower sea levels than those at which they were formed; (3) the depth to which the evidence has been buried by s o l i f l u c -t i o n deposits; and (4) the type of sediment i n which the shorelines were developed. The. importance of the. f i r s t factor i s revealed i n Figure 5 where only a limited number of observations was made i n the central part of the island. A similar paucity of evidence along the south coast of the island can largely be attributed to i t s o b l i t e r a t i o n by slumping and coastal recession. The significance of the l a t t e r two factors i s related to the thickness and composition of the active layer. I f the residual accumulations of sand and pebbles l i e at depths greater than the thickness of the active layer, their presence can only be determined by d r i l l i n g into the permafrost beneath. Furthermore, i f the composition of the sub-strate consists of coarse-grained sediments, i t i s d i f f i c u l t to d i f f e r -entiate between the parent material and possible residual beach deposits. I t i s for t h i s reason that l i t t l e evidence of raised strand-line features, i f indeed they e x i s t , i s shown on the surfaces of the sand headlands. As Figure 5 indicates, the evidence suggests the presence of a series of raised strand-lines, at approximately 7.5 metre (25 foot) i n t e r -v a l s , reaching to heights of 46 metres (150 feet) above sea l e v e l . The approximate elevations of the strand-lines were determined using a telescopic alidade and an altimeter. Because; of the weak topographic expression of the raised shorelines, the degree to which the majority of Figure 5 44 them have been obscured by s o l i f l u c t i o n deposits and the v i r t u a l imposs-i b i l i t y of locating the o r i g i n a l breaks of slope beneath t h i s material, and i n the absence of a f i x e d datum for mean sea l e v e l r e s u l t i n g i n an error of 0.3-0.6 metres (1-2 feet) produced by t i d a l f l u c t u a t i o n s , i t i s almost impossible to determine the precise elevations of the strand-lines. The heights assigned to the raised shoreline features must consequently be regarded as approximations, and the 'regularity' of th e i r spacing should be interpreted as apparent rather than r e a l . The highest strand-lines, at the '46' metre (150 foot) and '38' metre (125 foot) l e v e l s , are the least extensively developed since few of the summit areas reach these elevations. They also have the weak-est topographic expression, and t h i s i s undoubtedly related to th e i r greater antiquity and their exposure to the modifying influences of p e r i -g l a c i a l or other geomorphic processes for longer periods of time. One noticeable aspect of the d i s t r i b u t i o n a l pattern of these highest shore-l i n e s , and to some extent that of the '30.5' metre (100 foot) shoreline, i s t h e i r relationship to the major topographic features of,the island. A reconstruction of the configuration of Garry Island when the r e l a t i v e l e v e l of the sea was 38 metres (125 feet) higher than at present, shows that i t would actually consist of nine small islands separated by open stretches of water. The manner i n which the elevated strand-lines can be traced on both the north and south sides of the i s l a n d , and along the low cols i n the summit areas, suggests that the main topographic features of Garry Island are of considerable antiquity, and that the effects of sub-mergence and subsequent emergence have had l i t t l e effect i n the remoulding of the landscape. Almost a l l the strand-lines developed at and below the 45 30.5 metre (100 foot) contour l e v e l have some topographic expression, and can be traced much more extensively along the length of the island. Each shoreline i s readily detectable around the interfluve areas, but seldom i s i t possible to trace them across the lines of the valley reentrants. The possible reason for this w i l l be discussed more f u l l y below. Just to the east of Stream 'K' i s a large, steep-sided depression across which there i s a prominent ridge, the elevation of which i s approximately 18.5-20 metres (60-65 feet) above sea l e v e l (Plate I-D)... This ridge cannot be linked with any of the adjacent strand-line features and i t i s composed of r e l a t i v e l y fine sands with few pebbles or boulders. Although no conclusive evidence could be found, the ridge i s interpreted to be a structural feature i n the deformed sediments behind which a small lake may have temporarily been ponded. Between the '15' metre (50 foot) and '23' metre (75 foot) strand-lines are a number of isolated boulder accumulations developed at the 17-18 metre (55-60 foot) l e v e l . These are p a r t i c u l a r l y noticeable on the north side of the i s l a n d , to the east of Stream 'B', and on the west side of the open valley above the head of Stream ' I 1 . Their composition, consisting of large, well-rounded, iron-stained boulders, often exceeding 30 cms. i n length, and their prominence contrasts markedly with the minor accumulations of beach deposits found i n association with the series of strand-lines. They appear to be much older features and may be isolated remnants of an e a r l i e r position of the sea related to the time of deposition of the material in the sand headlands. The evidence of strand-line development at the 7.5 metre (25 foot) l e v e l , where preserved, i s the most pronounced of a l l the elevated shorelines, but i t s limited occurrence bears witness to the 46 extent of coastal recession i n recent times. V i r t u a l l y nowhere along the south coast of the island i s there any indication of i t s presence, and i t i s best developed around the margins of bay-like depressions between the sand headlands on the north side of the island. Here the old shoreline feature can be traced quite readily around the l a t e r a l margins of the sand headlands and i t s development i n the coarse-grained sediments probably accounts for i t s prominence and preservation, since the sands are less susceptible to slumping and s o l i f l u c t i o n than the s i l t s and clays. Although, as mentioned previously, i t i s d i f f i c u l t to apply the c r i t e r i o n of residual beach deposits at depth to the sand headland areas, there appears to be no doubt that the '7.5' metre (25 foot) strand-line d e f i n i t e l y post-dates their formation. The '7.5' metre (25 foot) raised strand-line i s also developed around the edge of another large depression on the northwest coast of the island (Figure 5). In this l o c a l i t y , the retreat of the present coastline i s also responsible for i t s . limited occurrence, but the stratigraphic sequences exposed i n the coastal b l u f f s provide an excellent cross-section of the structure of the depression. The details of this cross-section constitute the p r o f i l e C-D shown i n Figure 6. The depression i s developed i n the deformed clay sediments which occur i n the large mudslump, where they also contain bodies of ground i c e , at the northeastern end of the transect B-C (Figure 2). These clays are only exposed i n the extreme southwestern part of the cross-section and elsewhere they are buried beneath 5.5-6.0 metres (18-20 feet) of pebbly-gravel. The altitude of the top of these gravels i s approximately 7.5 metres (25 feet) above sea l e v e l at point C; the same elevation as that of the shoreline feature found around the edge of Figure 6 STRATI GRAPHIC SECTION EXPOSED IN COASTAL BLUFFS ALONG NORTHWEST COAST OF GARRY ISLAND For locat ion of sect ion see Inset Figure 2 Hor izonta l Scale 1 inch = 350 feet Ve r t i c a l Sca le 1 inch = 35 feet Ver t i ca l Exaggera t ion x l O So l i f luct ion Ma te r i a l Peat Pebbly Gravel Lacustrine Sediments Beach Gravels Pebbly C lay 48 the depression. From this point the surface of the gravels slopes gently toward the centre of the basin, r i s i n g again to a height of 7.3 metres (24 feet) i n a well-defined ridge. This ridge can also be traced across the floor of the depression, where the gravels appear at the surface, and i t represents an old sp i t or bar associated with the '7.5' metre (25 foot) strand-line (Plate II-B). Beyond the s p i t the thickness of the gravels decreases u n t i l they disappear completely i n the centre of the section. The gravels reappear again beyond the area of polygonal ground and increase to a thickness of 4 metres (13 feet) again at point D. The gravels i n turn are overlain by a sequence of organic-rich clays which l o c a l l y exhibit a varve-like banding. These clays contain an abundant c o l l e c t i o n of gastropods, i d e n t i f i e d by Dr. F.J.E. Wagner as Lymnaea species, and these f o s s i l s indicate that the clays were deposited in freshwater conditions. The thickness of the lacustrine sediments, on both sides of the depression, increases towards the centre where they thin again and are replaced by thick accumulations of peat i n the form of a number of high-centred tundra polygons. These polygons have been dissected by deep trenches created by melting along the lines of former positions of the intervening ice-wedges. A specimen of peat from one of these polygons, taken from the coastal b l u f f at a depth of 1.5 metres (5 feet) below the surface, was submitted for radiocarbon dating and yielded an age of 4120 - 130 years (G.S.C. - 513). A similar sequence of lacustrine sediments was deposited in the smaller marginal basin created by the formation of the gravel s p i t . A peat sample, collected by Dr. J.G. Fyles, taken from the basal layer of this sequence near point C indicated an age of 10,330 ^ 150 years (G.S.C. - 517). At this same location, however, the lake sediments are 49 also buried beneath a further 1.5-1.8 metres (5-6 feet) of pebbly-gravel followed by an additional 0.6-1.2 metres (2-4 feet).of peat. This gravel, the upper surface of which i s at a height of 10.5 metres (35 feet) above sea l e v e l , contains a number of iron-stained twigs and wood fragments and a sample of these, also collected by Dr. Fyles, provided a further radio-carbon date of 9730 "r 160 years (G.S.C. - 575). The i n t e r s t r a t i f i c a t i o n of the gravels and organic material i s the only evidence found which suggests that the changing pattern of land-sea relationships on Garry Island was not a simple, progressive emergence of the land or withdrawal of the sea. The details of this stratigraphic cross-section provide an insight into the mode of formation of the elevated strand-line features on Garry Island. The configuration of the '7.5' metre (25 foot) strand-line is highly suggestive of the submergence of a pre-existing topography, and the submergence of the depressions i n the topography created a series of embayments i n the coastline. Residual deposits of sand and gravel, derived from the erosion of the adjacent sections of the coast, accumu-lated along the shoreline and on the floors of the depressions. Locally this material was concentrated into sandspits or bars developed across the mouths of the embayments creating a number of lagoons on their landward sides. As the bars extended completely across the mouths of the bays, the lagoons were transformed into freshwater lakes. which were gradually i n f i l l e d by the deposition of lacustrine sediments. The f i n a l stages of this i n f i l l i n g were accompanied by the development of tundra polygons. Although stratigraphic evidence of this sequential development can only be demonstrated conclusively at the one location on the northwest coast of the island, i t i s believed that the bay-like depressions on the north side 50 of the island represent a similar sequence of events, although the bays may not have been converted into freshwater lakes. The large bay-like depressions are p a r t i c u l a r l y well-developed in association with the '7.5' metre (25 foot) strand-line. Investigations of the longitudinal p r o f i l e s of some of the main stream courses indicate that similar features were formed, a l b e i t usually on a much smaller scale, i n association with the other raised shorelines. . Longitudinal Stream P r o f i l e s . Under favourable circumstances the longitudinal and cross-valley p r o f i l e s of major r i v e r valleys may record the evidence of r e l a t i v e changes i n base l e v e l . A negative or downward movement of the base le v e l causes a rejuvenation of a stream at i t s mouth and a regrading of the stream towards i t s new base l e v e l . Such rejuvenation may resul t i n the production of an interrupted or stepped p r o f i l e , with the breaks of slope, or nickpoints, representing the headward extent of the regrading process. The longitudinal breaks i n the p r o f i l e can also often be correlated with 'valley in v a l l e y 1 forms i n which r i v e r terraces, or simpler valley-side facets, mark the former levels of the valley f l o o r s . The existence of similar terraces in the estuarine sections of some of the main r i v e r s on 15 the adjacent mainland has already been.discussed., There are no large perennial streams on Garry Island. Indeed, the present watercourses carry only a very intermittent surface runoff during the spring and early summer when they receive water from the melting of the snow cover. Throughout the rest of the summer months, Mackay, J. Ross (1958) op_. c i t . , p. 38. 51 these channels receive moisture seepage, produced by thawing of the active layer on the adjoining v a l l e y sides, and this i s seldom s u f f i c i e n t to maintain surface runoff except when combined with the e f f e c t s of a pro-longed period of heavy r a i n f a l l . The i n e f f i c i e n c y of stream erosion i s demonstrated by studies of the stream channels. Over most of t h e i r length the surfaces of these channels are aggrading by the growth and accumulation of organic material. As a r e s u l t of th i s process most of the channels are i l l - d e f i n e d and t h e i r l i n e s can best be traced by contrasts i n the vegeta-t i o n pattern. In the absence of a well-defined channel, surface runoff, when i t occurs, i s e a s i l y diverted to other routes. This was amply demon-strated i n the case of Stream 'D' (see Figure 5) at the beginning of the 1965 f i e l d season, where snow meltwater no longer followed the stream course, but had been diverted along the l i n e of a well-trodden path between the base camp and the beach. Attempts to lower a r t i f i c i a l l y the lev e l s of two lakes on the i s l a n d , by means of ditches, provided yet another example of the weak erosive power of running water. The ditches were cut through peat and gravel b a r r i e r s at the outlets of the lakes (Plate II-C). I t was hoped that, once established, these channels would be excavated further by the flow of water draining from the lakes. At one of the lakes, however, i t was soon demonstrated that the flow was i n s u f f i c i e n t to excavate the underlying peat, gravel and cl a y , and the channel could only be maintained by constant digging. The l o n g i t u d i n a l p r o f i l e s of fourteen Garry Island stream courses were surveyed and the r e s u l t s are shown i n Figures 7 and 8. Only at t h e i r mouths, where they cut through the coastal b l u f f s , do these streams possess a well-defined channel bordered by steep banks approxi-mately one metre (3-4 feet) high. Over most of the i r lengths, as Figure 7 L O N G I T U D I N A L S T R E A M P R O F I L E S (I) Prof i le D Hor i zon ta l S c a l e 1 inch 1 0 0 0 feet Ver t i ca l Sca le 1 inch 2 0 0 feet Ver t i ca l Exaggera t ion x 5 I - V I C r o s s - V a l l e y P ro f i l e Loca t ions Figure 8 L O N G I T U D I N A L S T R E A M P R O F I L E S (2) Horizontal Sca le 1 inch = 1000 feet Vert ical Sca le 1 inch = 2 0 0 feet Vert ical E x a g g e r a t i o n x 5 54 mentioned i n the previous paragraph, the channels have been f i l l e d by the accumulation of organic material, and their positions are marked by contrasts i n the vegetation pattern where the higher moisture content of 16 the substrate favours the growth of willows, sedges and moss. The lack of d e f i n i t i o n of the stream courses becomes increasingly apparent towards the higher elevations. Each of the streams has an i l l - d e f i n e d source area i n e'ither one of the steep-sided depressions or f l a t - f l o o r e d cols i n the summit areas. From these in d e f i n i t e beginnings, as Figures 7 and 8 indicate, the descent to the coastline i s not a smooth curve but exhibits a char a c t e r i s t i c step-like form (Plate II-p_). Where best developed, the. interruptions on the longi-tudinal p r o f i l e are marked by the occurrence of small lakes or, more commonly, areas of tundra polygons on the valley f l o o r . Elsewhere, the breaks i n slope are less prominent but are accompanied by a widening of the valley f l o o r , as indicated by the vegetation pattern, to 2-3 times the normal width. The d i s t r i b u t i o n of the elevations of these interruptions in the longitudinal p r o f i l e s of the fourteen streams i s shown i n Table IV. The heights have been grouped into selected height-range intervals to see i f there is any correlation between the elevations of the breaks in slope and those of the elevated strand-lines. The altitudes i n parentheses indicate those which f a l l outside the selected height-range in t e r v a l s . As Table IV shows, eight of the fourteen streams have an-i n f l e c t i o n point on their longitudinal p r o f i l e s at an elevation of Further details on the vegetation patterns of the stream courses are included i n the following chapter. TABLE IV ELEVATIONS OF THE BREAKS OF SLOPE ON THE LONGITUDINAL PROFILES OF GARRY ISLAND STREAMS S tream Elevation at Source (Metres) 6-9 Selected Height Ranges (Metres) 14-17 21-24 29-32 36-39 44-47 A B C D E F G H I J K L M N 25.3 49.4 33.2 22.9 29.9 28.3 39.9 24.7 28.0 48.5 44.8 47.9 41.1 25.9 7.6 (1.2) 7.6 7.9 7.6 7.6 7.9 7.9 7.9 (11.3) 15.5 (12.2) 16.8 (11.6) (11.0) ( 9.8) 14.3 (18.6) 15.2 15.5 15.5 14.6 16.2 14.3 16.5 23.5 22.6 22.9 21.9 21.6 24.1 (26.5) 24.1 22.9 22.6 23.2 30.2 31.1 29.6 36.9 45.7 38.1 31.1 37.2 29.9 (33.8) 38.1 45.4 approximately 7.5 metres (25 feet) above sea l e v e l . Of the s i x streams which do not, three drain toward the south coast and one to the west coast where, i n each case, they terminate i n high coastal bluffs which have undergone considerable recession. As a re s u l t of this recession, the lower courses of these streams have become deeply entrenched and they reach the present shoreline through narrow 'V-shaped valleys. Interruptions i n the longitudinal p r o f i l e s at altitudes of both approxi-mately 15 and 23 metres (50 and 75 f e e t ) , are represented i n ten of the fourteen stream courses, but the evidence of similar breaks of slope at altitudes exceeding 30.5 metres (100 feet) above sea l e v e l i s more r e s t r i c t e d . Thus the numbers of i n f l e c t i o n points in the pr o f i l e s occurring at the 30.5, 38 and 46 metre (100, 125 and 150 foot) levels are only f i v e , four and two respectively. The limited number of interruptions in the p r o f i l e s at these levels can re a d i l y be explained by the fact that progressively fewer of the streams originate at these higher elevations (Table IV). For example, only eight of the fourteen streams shown i n Figures 7 and 8 originate at elevations of more than 30.5 metres (100 feet) above sea l e v e l ; similar figures for altitudes of 38 and 46 metres (125 and 150 feet) are s i x and three streams respectively. In a l l , Table IV l i s t s a t o t a l of 48 elevations, each of which represents a point of i n f l e c t i o n on a longitudinal stream p r o f i l e . Of this t o t a l , no fewer than 39, or more than eighty per cent, occur within the selected height-ranges. There may be some significance, however, to the fact that of the breaks i n slope which do not occur within these height ranges, more than one-half are developed at an elevation of approximately 10.5 metres (35 feet) above sea l e v e l . The evidence from the surveyed stream courses suggests that 57 there i s a correlation between the altitudes of the i n f l e c t i o n points on the longitudinal p r o f i l e s of these streams and the elevations of the raised shoreline features. I t i s postulated that the two sets of features are closely interrelated and that both are associated with former positions of the sea. When the r e l a t i v e l e v e l of the sea stood at each of the positions indicated by the strand-lines, i t i s postulated that the lines of the valleys formed small indentations or bays i n the shoreline. Residual accumulations of coarse sand and gravel, produced by the erosion of the coastal b l u f f s , were deposited at the shoreline, and l o c a l l y these were concentrated into small spits or bars, extending across the mouths of the bays. The lagoons, created by the formation of these bars, may eventually have been sealed off from the open sea and transformed into shallow, freshwater lakes. Some of these lakes may have been drained immediately when the r e l a t i v e l e v e l of the sea was lowered but many were l e f t occupying positions i n the newly-exposed valley f l o o r s . The larger and deeper lakes have persisted through to the present time, but many of the smaller, shallower ones have subsequently been f i l l e d by the accumu-l a t i o n of organic material and become the l o c i for tundra polygon develop-ment. The re p e t i t i o n of th i s sequence of events i s offered as an explanation of the stepped p r o f i l e so chara c t e r i s t i c of the stream courses on Garry Island. The sequential stages of development, offered to account for the d i s t r i b u t i o n and formation of the areas of polygonal ground, i s s i m i l a r , a l b e i t on a much smaller scale, to that proposed for the o r i g i n of the exposures of lacustrine sediments found along the northwest coast of the island. The hypothesis gains a certain c r e d i b i l i t y from two 58 aspects of studies of the contemporary shoreline. F i r s t l y , p r a c t i c a l l y a l l of the present streams have their outlets blocked by accumulations of gravel, driftwood and peat. Even during periods of r e l a t i v e l y strong runoff, the streams are only temporarily able to establish a channel through these materials to the sea. Such channels are very ephemeral features, however, and are soon choked by wave action. The second l i n e of support comes from the fact that polygonal ground i s currently developing around the margins of lagoons enclosed by the construction of the modern sandspits. In an attempt to establish further the v a l i d i t y of this hypo-thesis a number of cross-valley p r o f i l e s was surveyed at strategic points along Stream 'G'. These p r o f i l e s are shown i n Figure 9. At each of these points a series of d r i l l holes was made across the stream course to deter-mine the nature of the substrate and, i f possible, locate the presence of the buried gravel ridges required to support the hypothesis. The depths at which gravels were encountered i n these d r i l l holes are also recorded i n Figure 9. Cross-profile I was located at an elevation of approximately 4.5 metres (15 feet) above sea l e v e l where the stream course traverses the large bay-like reentrant i n the '7.5' metre (25 foot) strand-line. The i l l - d e f i n e d nature of the channel i s immediately apparent and the d r i l l i n g operations revealed no sign of gravels^ at least to depths of 2 metres (6 f e e t ) . Cross-profiles I I , I I I and IV were located at elevations of approximately 7.5, 15 and 23 metres (25, 50 and 75 feet) above sea l e v e l . As Figure 9 shows each of these locations was characterized by the presence of a f a i r l y prominent gravel ridge at depths of up to 1.5 metres (5 feet) below the surface. The details of one of these ridges were p a r t i c u l a r l y we11-developed at an altitude of 23 metres (75 feet) above Figure 9 C R O S S - V A L L E Y P R O F I L E S - S T R E A M ' G 1 Prof i le I Horizontal Scale 1 inch = 100 feet Vertical Scale 1 inch = 20 feet Vertical Exaggerat ion x5 w s Buried gravel Wi l lows Sedge/Moss 60 sea l e v e l where they could be traced l a t e r a l l y to prominent ridges on the side of the v a l l e y . Cross-profile V was located at an elevation of 27.5 metres (90 feet) above sea l e v e l . Although gravels were encountered beneath the stream course i t s e l f , they could not be detected at a l l on the sides of the v a l l e y . The f i n a l cross-profile was located just below an area of weakly-developed tundra polygons at an altitude of 38 metres (125 feet) above sea l e v e l . As the p r o f i l e shows, gravels were encounter-ed at shallow depths beneath the channel and for considerable distances on either side. Thus, although the subsurface p r o f i l e s were only determined at strategic points along one of the streams, the limited evidence pro-vided by these p r o f i l e s i s i n accordance with the hypothetical sequence outlined above. The studies of the longitudinal and cross-valley p r o f i l e s of the stream courses corroborate the opinion that a series of elevated strand-lines, spaced at intervals of approximately 7.5 metres (25 feet) and reaching elevations of 46 metres (150 feet) above sea l e v e l , can be i d e n t i f i e d on Garry Island. In the absence of d e f i n i t i v e f o s s i l evidence, however, i t i s reasonable to question the v a l i d i t y of a further conclusion that these strand-lines mark the positions of former levels of the sea. The prime reason for stating that these shorelines are marine features i s based e n t i r e l y on the apparent uniformity of their elevations on a l l sides of the island. Irrespective of the aspect of the slope on which they are developed, the shorelines seem to exhibit a strong degree of consistency with respect to their elevation above sea l e v e l . Such consistency i s not i n keeping with a hypothesis that they were developed around the margins of proglacial lakes. Moreover, the conditions on Garry Island were rarel y suitable for the development of these lakes. They could have only 61 developed, to any r e a l extent, along the south side of the island where meltwater may temporarily have been ponded between the coastline and the ice front as i t retreated to the south. A further hypothesis, which was also considered, i s that some of the shorelines were developed around lakes which were ponded by str u c t u r a l deformation features. The possible existence of one such lake was referred to previously. The absence of any marked anomalies i n the heights of the shoreline features would appear to indicate that neither of these possible modes of o r i g i n was more than of l o c a l significance, i f any at a l l . SUMMARY The stratigraphy of Garry Island consists of a sequence of sands, s i l t s , clays, and stony-clays, cemented by i c e , which have been intensively deformed by the thrusting action of gl a c i e r - i c e moving from the south. The deformed sediments are l o c a l l y overlain by sands and gravels, probably brought down by streams following the retreat of the ice. These materials are undisturbed and contain marine f o s s i l s dated at > 42,000 years. The absence of any signs of g l a c i a l t i l l on top of the sands suggests that Garry Island lay beyond the northwest l i m i t s of the Laurentide ice sheet during the late-Wisconsin g l a c i a t i o n . Attempts to determine the extent of changes i n the former r e l a t i v e positions of land and sea are complicated by the weak topo-graphic expression of raised shoreline features, and the degree to which they have been obscured, or even t o t a l l y o b l i t e r a t e d , by subsequent geo-morphic a c t i v i t y . In addition to these d i f f i c u l t i e s , the absence of a fixed datum for mean sea le v e l prevents a completely r e l i a b l e determina-62 tion of the precise altitudes of the shorelines. Despite these l i m i t a -tions, the evidence suggests the existence of a series of elevated strand-lines which occur at approximately 7.5 metre (25 feet) intervals to an alt i t u d e of 46 metres (150 fe e t ) . Some additional support for this view comes from surveys of the longitudinal p r o f i l e s of 14 stream courses on the island. These p r o f i l e s are c h a r a c t e r i s t i c a l l y s t e p - l i k e , and the altitudes of the i n f l e c t i o n points on the thalwegs exhibit a strong degree of s i m i l a r i t y to the elevations of the raised shoreline features. Further comments on the ove r a l l significance of these s t r a t i -graphic observations and their incorporation into a possible chronological sequence w i l l be discussed i n the f i n a l chapter of the thesis. CHAPTER I I I VEGETATION Garry Island i s located wholly beyond the northern l i m i t of trees which presently l i e s approximately 65-80 kilometres (40-50 miles), to the. south i n the modern delta of the Mackenzie River. The treeless character of the island places i t i n the a r c t i c tundra region and the vegetation i s composed primarily of dwarf shrubs, herbs, mosses and lichens. The objective of t h i s chapter i s to describe the major vegeta-tion types which occur on Garry Island and to examine the i n t e r r e l a t i o n -ships between the vegetation patterns and types of geomorphic a c t i v i t y . Attempts to define tundra vegetation communities on the basis of their plant composition are complicated by the fact that there are not always d i s t i n c t i v e , or diagnostic species i n each of the plant associations. Many species have a broad tolerance of environmental conditions, and therefore occur i n a wide variety of habitats. This fundamental inadequacy of a c l a s s i f i c a t i o n of a r c t i c vegetation on the basis of physiognomy or f l o r a l composition has been recognized by numerous botanists, as i s exemplified by the following quotations: "In the a r c t i c , differences are merely quantitative. The habitat preferences of the in d i v i d u a l species f i n d expression merely i n increased or decreased abundance i n more or less favourable habitats rather than by presence here or absence there.... A r c t i c vegetation must be described by reference to the physical conditions of the habitat rather than by an attempt to discover and deal with habitat preferences of the species present." Accordingly, the major purpose of the vegetation study was to examine the various factors contributing to the physical character of the plant habitats and, i n p a r t i c u l a r , the relationships of the vegetation pattern to geomorphic a c t i v i t y . Subsurface conditions were also checked, with s p e c i f i c reference to the r e l a t i v e amounts of the organic material and mineral s o i l f r a c t i o n s , and the depth of the active layer was recorded at various times throughout the summer for each of the habitats. A d i s -cussion of this data i s contained i n the following chapter. The vegetation was mapped i n the f i e l d on a scale of approxi-mately 1:25,000 and the delineation of the types was checked through sub-sequent studies of a e r i a l photographs and coloured transparencies. For each of the major habitats i d e n t i f i e d , an attempt was made to estimate the r e l a t i v e percentages of the area occupied by the dominant species. Since the emphasis was on the physical character of the habitat, however, this was done by v i s u a l estimation rather than by the employment of the more rigorous quadrat sampling technique. For completeness, and for readers having interests i n botany, the various species i d e n t i f i e d have been l i s t e d under each of the vegetation types, i n which they occur. Although i t i s not claimed to be exhaustive, Appendix I contains a summary of this information and l i s t s a t o t a l of 106 vascular species and 17 bryophytes Griggs, R. F. (1936) "The Vegetation of the Katmai D i s t r i c t " , Ecology, Vol. 17, pp. 381-382. 65 collected and i d e n t i f i e d on Garry Island. VEGETATION TYPES Figure 10 shows the areal extent and d i s t r i b u t i o n of the pri n c i p a l vegetation types found on Garry Island. The c l a s s i f i c a t i o n i s based primarily on the physical character of the habitat, but i t also includes the names of the s p e c i f i c plant species wherever they tend to dominate the particular vegetation type. I. Dryas-Hummock Type. Most of the drier tundra sit e s are characterized by an irregular surface where the ground i s covered by numerous, rounded earth hummocks, the size and spacing of which vary according to their position on the topographic p r o f i l e . These microrelief forms are found on a l l summit areas and valley-side slopes with moderate to good drainage con-ditions . (Plate III-A). Hummock p r o f i l e s consist of a surface layer of turf and organic material, beneath which there i s a domed core of mineral s o i l , but there i s l i t t l e or no s o i l beneath the intervening depressions which are the sites of organic accumulation. The vegetation i s composed primarily of low matted, woody shrubs together with lichens and mosses. This vegetation type i s the most An attempt was made i n the f i e l d " to identify the vascular plants using N. V. Polunin's book,, Circumpolar A r c t i c F l o r a , Clarendon Press, Oxford, 1959, and the format of Appendix I is based on this text. I am indebted however, to Dr. Eric Hulten of the Naturhistoriska Riksmusset, Stockholm, for his precise i d e n t i f i c a t i o n of the species l i s t e d . I am s i m i l a r l y indebted to Dr. Herman Personn, of the same i n s t i t u t e , for the i d e n t i f i c a t i o n s of the bryophytes. Unfortunately i t was not possible to identify the lichens. Dryas Hummocks Cass iope Snowpatches A lnus c r i spa E r i opho rum T u s s o c k s Hummock - T u s s o c k T rans i t i on S t r eam C o u r s e Willow T h i c k e t s Sedge - Moss F l a t s M u d s l u m p Commun i t i e s S t r and Commun i t i e s - Grave l Ba r s - Ma r sh - L a g o o n [ P o l y g o n s - Low C e n t r e d - High Cent red F i g u r e 10 G A R R Y I S L A N D V E G E T A T I O N T Y P E S S c a l e I A p p r o x t m a t e ) 1 Plate I I I VEGETAT ION TYPES A. DRYAS HUMMOCK. Hummock i n centre dominated by l i c h e n s , A r c t i c Avens, and Bigelow's Sedge. Darker colour of the inter-hummock depressions i s due to the dominance of the A r c t i c White Bell-heather and mosses. B. ALNUS CRISPA. Almost pure stands of the Mountain Alder growing i n a shallow, sheltered depression on the north side of Garry I s l a n d . C. CASSIOPE SNOWPATCH. T y p i c a l vegetation a s s o c i a t i o n found i n areas of l a t e - l y i n g snowpatches. The dominant species are the A r c t i c White Bell-heather and willows. D. ERIOPHORUM TUSSOCKS. C h a r a c t e r i s t i c tussock forms of the Sheathed Cotton-grass. The inte r - t u s s o c k areas are dominated by mosses and the A r c t i c White Be11-heather. 68 extensive found on the island and i t i s also the most d i v e r s i f i e d i n terms of i t s f l o r i s t i c composition. On the f l a t summit areas, the hummocks are quite subdued i n form, but the microrelief factor i s s t i l l s u f f i c i e n t to produce two d i s t i n c t plant habitats. The d r i e r , elevated hummock centres are dominated by the A r c t i c Avens (Dryas i n t e g r i f o l i a ) , after which the vegetation type i s named, together with mosses ( c h i e f l y Pieranum, Cinclidium and Bryum species) and lichens (the so-called 'reindeer moss'). Co l l e c t i v e l y these plants may account for 50-60 per cent of the vegetation cover on the hummock surface. Other major species found on the hummocks include Bigelow's Sedge (Carex b i g e l o w i i ) , A r c t i c Blueberry (Vaccinium  uliginosum var. alpinum), Mountain Cranberry (Vaccinium v i t i s - i d a e a ) , Common Crowberry (Empetrum nigrum), and the Glandular Birch (Betula  glandulosa). On the more sheltered flanks of the hummocks there i s an increase i n the percentage cover of these species, and a corresponding decrease in the cover of avens and lichens. Small willows ( S a l i x  r e t i c u l a t a , S_. glauca var. niphoclada) , rooted i n the hummocks, and the Narrow-leafed Labrador-tea (Ledum palustre ssp. decumbens) are also found on the sides of the hummocks. Less common species encountered growing on the hummock centres include the A r c t i c Wintergreen (Pyrola grandiflora) , Ar c t i c Meadow Grass (Poa a r c t i c a ) , Capitate Lousewort (Pedicularis  capitata), Alpine Bearberry (Arctostaphylos rubra), Veiny-leafed Willow (Salex phlebophylla), Northern Wood-rush (Luzula confusa), Woolly Lousewort (Pedicularis 1anata), Long-stalked Stitchwort ( S t e l l a r i a  longipes), Mountain Meadow B i s t o r t (Polygonum b i s t o r t a ) , Narrow-leafed Saussurea (S aussurea a n g u s t i f o l i a ) , Alpine B i s t o r t (Polygonum viviparum) , Lapland Rose-bay (Rhododendron lapponicum) and the Lake Louise Arnica (Arnica louiseana frigida) . 69 The inter-hummock depressions are sites of snow accumulation during the winter months and the organic substrate remains much moister than the more exposed hummock centres. These depressions support an e n t i r e l y different and less d i v e r s i f i e d plant assemblage. Mosses ( c h i e f l y Aulacomnium, Hylocomnium and Sphagnum species) and the A r c t i c White Bell-heather (Cassiope tetragona) are the dominant constituents of the f l o r a , and c o l l e c t i v e l y account for 60-70 per cent of the vegetation cover. Other ubiquitous, though minor, species found i n the depressions include the Lapland Butterbur (Petasites f r i g i d u s ) , Sudetan Lousewort (Pedicularis sudetica), Fragile Sedge (Carex membranacea), Sheathed Sedge (£. vaginata) , Radiate Saxifrage (Saxifraga r a d i a t a ) , Lapland Reedgrass (Calamagrostis. lapponica), Alpine Hedysarum (Hedysarum alpinum americanum) , Alpine F o x t a i l (Alopecurus alpinus) and the Naked-stemmed Parrya (Parrya  nudicaulis). The above descriptions pertain to characteristic plant assem-blages of hummocks and depressions on the upland surfaces. Changes i n these plant compositions appear to be most markedly affected by the degree of exposure and the moisture supply. On exposed, windswept slopes the hummocks are even more subdued than on the f l a t summit areas, the vegeta-tion cover i s thinner, and there i s patchy development of bare ground. The most noticeable effect of these conditions i s an increase i n the cover of the A r c t i c Avens, to a point where they are almost the only species found on the hummock surface. Contrasts i n the moisture conditions of the hummock centres and the depressions are less pronounced and the avens are frequently found growing on dry moss pads i n the l a t t e r areas. Other plants which appear to thrive i n this exposed environment include the Veiny-leafed Willow, Woolly Lousewort, Northern Wood-rush and the A r c t i c 70 Lupin (Lupinus arcticus) . Between the upland summit areas and the lower elevations there i s a gradual, but progressive, increase i n the size of the hummocks. Coincident with this increase in the micr o r e l i e f , the vegetation contrasts between the hummocks and depressions become more pronounced. This i s p a r t i c u l a r l y noticeable on the upper slopes, but towards the foot of the slope there are d i s t i n c t changes i n the vegetation cover on the hummocks. The lichens disappear almost completely and the A r c t i c Avens become only a minor constituent of the f l o r a . Mosses, Sedges, the Common Crowberry, A r c t i c Blueberry and Labrador-tea achieve much greater prominence and the Arc t i c White Bell-heather i s also found on the hummocks. The most notable change in the depressions i s an increase i n the amount of the moss cover. I t i s not surprising that the vegetation type which has the greatest areal extent, and the most d i v e r s i f i e d f l o r i s t i c composition, lends i t s e l f to the possible recognition of abundant sub-types. Some of these have already been alluded to with reference to the effects of exposure, moisture conditions, etc. The problem of d i f f e r e n t i a t i n g between a sub-type and a separate type i s not an easy one, but i n two cases the vegetation association that was produced was so d i s t i n c t i v e -the Alnus crispa and the Cassiope Snowpatch types - that separate vegeta-tion types were recognized. I I . Alnus crispa Type. In a very few sheltered locations, with a moderate to good moisture supply, the vegetation i s l o c a l l y dominated by pure stands of the Mountain Alder (Alnus cr i s p a ) . The largest stand of Alder was found on a north-facing slope and was composed of low bushes approximately one 71 metre (3 feet) high (Plate III-B). Each bush has a l a t e r a l spread of almost 2 metres (6.5 f e e t ) , and this e f f e c t i v e l y shades out the under-lying ground surface which is consequently devoid of any further plant cover and mantled by a l i t t e r of dead leaves. The stand of Mountain Alder shown in Plate 3JXL-B i s the only substantial one on the i s l a n d , although i n a number of i s o l a t e d , favourable spots individual bushes were noted. I I I . Cassiope Snowpatch Type. In the section on the Dryas-Hummock vegetation type, downslope changes in the f l o r a l composition were tentatively related to changes i n the moisture conditions of the s o i l . L o c a l l y , i n addition to the increase in moss cover, the hummocks and depressions al i k e are blanketed by a dense, almost pure cover of the A r c t i c White Bell-heather (Cassiope tetragona) and willows (Plate III-C). This vegetation pattern gives a d i s t i n c t i v e dark colouration to the ground surface which i s readily detectable on a e r i a l photographs. The lower slopes receive water seepage, derived from the thawing of the active layer, from the upper parts of the slopes, and are also the sit e s of l a t e - l y i n g snow patches. These effects are most pro-nounced i n the larger sheltered depressions and on northeast-facing slopes where the snow may remain on the ground well into the summer before i t disappears. IV. Eriophorum Tussock Type. Wet tundra habitats occur i n areas where the slopes are only of the order of 1-2 degrees and the re s u l t i n g poorly developed drainage conditions are reflected in a water table which i s close to the ground surface. This vegetation type i s found primarily on the floors of the major depressions and on the surfaces of the sand headlands, but i t also 72 occurs l o c a l l y along f l a t s associated with the raised beach lines and around the margins of some areas of polygonal ground. In these s i t e s the underlying s o i l s are t y p i c a l l y cold, grey s i l t s or clays, and the surface, i s mantled with a l i t t e r of raw humus. The vegetation of these areas i s dominated by plants of the sedge (Cyperaceae) family and the Sheathed Cotton-grass or 'Niggerhead' (Eriophorum vaginatum) i n par t i c u l a r . This plant exhibits a characteris-t i c tussock form (Plate III-D), up to 35-40 centimetres (14.0-15.5 ins.) high and as much as 30-35 centimetres (12-14 ins.) across at the crown. Each tussock i s a single plant and consists of a mass of dead and l i v i n g organic material overlying a small plug of mineral s o i l . The roots form a t i g h t l y woven mat penetrating down into t h i s mineral s o i l . The spacing of the tussocks i s variable and they are separated by shallow troughs ranging in width from 10-75 centimetres (4.0-29.5 i n s . ) . Where best developed, on slopes of 1-2 degrees, the tussocks occur i n close juxtaposition and account for as much as 90 per cent of the vegetation cover. The floors of the intervening troughs may be occupied by small pools of standing water, raw organic debris or saturated pads of moss ( c h i e f l y Sphagnum and Hypnum species). Other plants of minor importance found within this same habitat, and p a r t i c u l a r l y i n the s l i g h t l y drier sites such as elevated moss polders or the sides of the individual tussocks, include the A r c t i c White B e l l -heather (Cassiope tetragona) , Narrow-leafed Labrador-tea (Ledum palustre  ssp. decumbens), Glandular Birch (Betula glandulosa), Common Crowberry (Empetrum nigrum), A r c t i c Wintergreen (Pyrola grandiflora), and the willows ( S a l i x pulchra, S_. reti c u l a t a ) . 73 V. Hummock-Tussock Transition Type. Whereas the major extent of both the Dryas-Hummock and Eriophorum-Tussock vegetation types i s eas i l y i d e n t i f i a b l e , the margins are frequently blurred by extensive t r a n s i t i o n zones involving consider-able i n t e r d i g i t a t i o n with each other. On slopes of 2-4 degrees the tussocks become smaller and further apart, accounting for only 20-40 per cent of the plant cover. The inter-tussock areas become less moist, and there i s an increase i n the number and variety of other species and Cassiope tetragona i n pa r t i c u l a r . Large earth hummocks, with f l o r i s t i c associations similar to those described i n the Dryas-Hummock type, and bare mud bo i l s are intermingled with the tussock forms. F i n a l l y , on slopes above 4 degrees the tussocks are gradually eliminated and the vegetation becomes of the Dryas-Hummock type. VI. Stream Course Willow Thickets. The majority of the stream courses are occupied by narrow st r i p s of willow growing i n a substratum of moist organic material and occasionally sandy-gravel (Plate IV-A). Such s i t e s are nearly always amply supplied with water, receiving runoff from the melting snow i n the spring and early summer, and are kept moist throughout the rest of the summer by seepage from the thawing ground. The t y p i c a l vegetation association consists of a f a i r l y dense thicket of willows, c h i e f l y the Diamond-Leaf Willow ( S a l i x pulchra) and the Northern Willow (S. glauca  var. niphoclada), ranging from 0.5 to 1 metre (1.5 to 3.0 feet) high. The s t r i p i s commonly less than 5 metres (16 feet) wide, although i t frequently expands to many times this figure at stream junctions and at the intakes and outlets of some of the lakes. In these locations, the Plate IV V E G E T A T I O N TYPES A. STREAM COURSE WILLOW THICKETS. Stream course marked by a narrow ribbon of willow species. Darker patches on the slopes i n the centre of the photograph also show the Cassiope Snowpatch vegetation type. B. SEDGE - MOSS FLATS. T y p i c a l vegetation a s s o c i a t i o n found on the f l o o r s of the larger depressions, c o n s i s t i n g of an almost f e a t u r e l e s s mat of mosses together with the Creeping Sedge, A r c t i c Marsh Willow and Alpine Bearberry. C. MUDSLUMP COMMUNITIES . Dark area on the upper r i g h t side of the photograph represents an a c t i v e mudslump. The foreground i s dominated by almost pure stands of the Marsh Fleawort. The scar of a former mudslump, dominated by grasses, can be seen i n the centre background. D. MUDSLUMP COMMUNITIES . The vegetated f l o o r of a former mudslump dominated by grasses. The old headwall i n the background has been colonized by the Dryas Hummock vegetation. 75 willows may reach heights of as much as 2 metres (6.5 fe e t ) . Beneath the willows, the ground cover consists primarily of mosses ( c h i e f l y Sphagnum sp.) together with the Common Horsetail (Equisetum arvense) and the occasional 'niggerhead' tussock (Eriophorum vaginatum). Other species found i n this same habitat, but of minor importance, include the Lapland Reedgrass (Calamagrostis lapponica) , Lowly Fleabane (Erigeron humilis), Alpine F o x t a i l (Alopecurus alpinus) and the Northern Anemone (Anemone  parviflora) and Richardson's Anemone (A. r i c h a r d s o n i i ) . VII. Sedge-Moss F l a t s . Sedge-moss f l a t s are developed in areas of impeded drainage as found on the floors of the larger depressions, around the margins of areas of polygonal ground, and bordering the willow thickets of the stream courses. In the upper portions of the stream p r o f i l e s , this vegetation association frequently replaces the willow thicket type. The substratum is composed e n t i r e l y of organic material which i s usually saturated. Spongy mosses (Sphagnum squarosum, S_. warnstorfianum, Tomenthypnum nitens and Hypnum callichroum) are the dominant constituents of the f l o r a and l o c a l l y they may be the only plants present. Usually, however, they are found i n association with the Creeping Sedge (Carex chordorrhiza) and the A r c t i c Marsh Willow ( S a l i x arctophila) (Plate IV»B). Minor l o c a l r e l i e f i n th i s habitat i s provided by the growth of elevated moss polders, and these s l i g h t l y drier s i t e s support more extensive covers of the Alpine Bearberry (Arctostaphylos rubra), together with the A r c t i c Wintergreen (Pyrola  grandiflor a) and the A r c t i c Blueberry (Vaccinium uliginosum). VIII. Mudslump Communities. Mudslumps are large amphitheatre-like depressions formed by 76 the melting out of tabular bodies of segregated ground ice. Since coastal recession i s the major agency responsible for the exposure of these ice bodies, the vegetation type associated with these features i s largely con-fined to coastal locations. Because of the high ice to mineral s o i l r a t i o i n these exposures, melting produces large quantities of excess water and very l i t t l e debris to accumulate on the f l o o r of the slump. Prominent headwalls formed i n the slump persist for long periods after i t becomes inactive, so that a precise delineation of the habitat i s often a simple procedure. Vegetation associations i n these mudslumps appear to be related to colonization stages once the slump has become inac t i v e , and they can be subdivided into a number of sequential types. Active mudslumps.are characterized by large quantities of ice exposed i n the headwall, and the f l o o r of the slump i s covered to varying degrees, with a mobile layer of f l u i d mud. Most of the-surface i s bare except for the presence of clumps of vegetation, hummocks and willows, which have been detached intact from the rim of the slump. Away from the ice face, the l i q u i d mud i s generally restricted, to well-defined mudflows, and the. less active, mud surfaces are rapidly colonized by almost pure stands of yellow Marsh Fleawort (Senecio congestus) (Plate IV-C). This species grows gregariously even i n places where fresh mud and s i l t are continually being deposited by streams of water coming from the active slump face. Other plants found i n t h i s moist environment include the A r c t i c Butterbur (Petasites a r c t i c u s ) , Langsdorf's Lousewort (Pedicularis  l a n g s d o r f i i ) , Marsh Felwort (Lomatogonicum rotatum), A r c t i c Cotton-grass (Eriophorum scheuchzeri), Scentless Mayweed (Matricaria ambigua) and the Black-tipped Groundsel (Senecio lugens). In s l i g h t l y drier areas, though s t i l l moist, the above species are gradually replaced by the Alpine 77 Hedysarum (Hedysarum alpinum americanum), A r c t i c Dock (Rumex a r c t i c u s ) , T i l e s i u s ' s Wormwood (Artemesia t i l e s i i ) , Tawny Arctophila (Arctophila fulva) , Kotzebue's Grass of Parnassus (Parnassus kotzebuei) and Anderson's A l k a l i Grass. ( P u c c i n e l l i a andersonii). The next stage of the colonization process develops when the f l o o r of the mudslump i s no longer subjected to mudflows or to deposition by running water. . The surface of the mud becomes extremely hard and dry, and i s often traversed.by networks of desiccation cracks. S o i l conditions are t y p i c a l of the disturbed nature of the habitat, consisting primarily of mineral s o i l with patches of organic material i r r e g u l a r l y distributed at depth. The percentage of bare ground ranges from 20-50 per cent, being least where the surface has been inactive for longer periods of time. Where the mudslump as a whole i s no longer active, the old headwall may be p a r t i a l l y covered by Dryas-Hummock vegetation s l i d i n g down from the surface above. The. t y p i c a l vegetation of these areas i s a dense cover of grasses (Plate IV-D), of which the following species were the most prominent: V i o l e t Wheat Grass (Agropyron latiglume), Spiked Trisetum (Trisetum spicatum), American Hare's Ear (Bupleurum americanum), Smoothing Whitlow Grass (Draba g l a b e l l a ) , Arctagrostis (Arctagrostis l a t j f o l i a ) and the Sheathed A l k a l i Grass ( P u c c i n e l l i a vaginata). Other species found i n the same h i b i t a t include the Northern Asphodel ( T o f i e l d i a coccinea), A r c t i c Lychnis (Melandrium a f f i n e ) , Macoun's Poppy (Papaver k e e l e i ) , Pale Paint. Brush ( C a s t i l l e j a p a l l i d a ssp. elegans), Maydell's Oxytrope (Oxytropis maydeliana), Alpine Milk-vetch (Astragalus alpinus), Long-stalked Stitchwort ( S t e l l a r i a longipes), Fingered Buttercress (Cardamine  d i g i t a t a ) , Low Northern Rockcress (Braya humilis.ssp. a r c t i c a ) , Northern Tansy-mustard (Descuirainia sophioides), Mackenzie's Hedysarum (Hedysarum 78 mackenzii), Beringian Chickweed (Cerastium beeringianum), Dawson Hemlock Parsley (Conlosellnum c n i d i l f o l i u m ) , Acutish Jacob's Ladder (Polemonium  acutiflorurn), Reddish Sandwort (Arenaria rubella) and the M i l f o i l ( A c h i l l e a b o r e a l i s ) . Following even greater periods of s t a b i l i t y the ground supports a continuous cover of vegetation and the old headwall i s s i m i l a r l y mantled with Dryas-Hummocks. S o i l conditions s t i l l show the effects of the disturbed nature of the habitat, but they tend to be moister and there i s a thin organic accumulation at the surface. The grass vegetation i s eventually replaced by dense willow thickets composed mainly of the Northern Willow ( S a l i x glauca var. niphoclada), Alaskan or Felt-leaf Willow (S_. alaxensis) and the Net-veined Willow (S_. reticulata) . Beneath, and between, these willows i s a scant ground cover comprised mainly of mosses together with the Common Horsetail (Equisetum arvense) and the Arc t i c Lupin (Lupinus ar c t i c u s ) . This type of vegetation associa-tion i s also widely distributed along many of the s t a b i l i z e d b l u f f s , many of which formerly underwent active recession by a process similar to that operating i n active mudslumps. IX. Strand Communities. Strand communities also constitute one of the most d i s t i n c t i v e vegetation types and, l i k e the mudslumps, are easily i d e n t i f i a b l e on ae r i a l photographs. They are also r e l a t i v e l y simple to delineate since they generally abut quite sharply against the other vegetation types with l i t t l e or no t r a n s i t i o n zone i n between. Four sub-types, two major and two minor, make up the strand communities. The minor sub-types, too small to be shown e f f e c t i v e l y on 79 Figure 10, include the active beaches and coastal b l u f f s ; two locations which, together with the active mudslumps, account for most of the un-vegetated areas on the island. Active beaches, composed of coarse sands, gravels and boulders are well-drained and subjected to ice scouring i n the f a l l and winter months and wave abrasion during the summer. They also may be the sites of deep snow accumulation where i t is piled i n d r i f t s against the b l u f f s . The net effect of these adverse environmental conditions i s to keep the beaches almost completely devoid of any plant growth except for a few isolated clumps at the base of some of the c l i f f s . A c t i vely retreat-ing sections of the coastline are also vegetation free except for isolated clumps detached from the c l i f f edge during the retreat process. S t a b i l i z e d c l i f f s , usually resulting from the formation of protective sandspits at the foot of the c l i f f , support a vegetation cover proportionate to their period of s t a b i l i t y . The c l i f f s composed of fine-grained sediments, containing abundant ground i c e , have already been discussed under the heading of mud-slump communities. On the sand headlands, sections of the coastline that were actively retreating u n t i l r e l a t i v e l y recent times, s t i l l have 40-50 per cent of the c l i f f face bare of any vegetation cover. The remaining surface supports a l i g h t cover of mosses ( c h i e f l y Psilopilum cavifolium) together with the Radiate Saxifrage (S axifraga ra d i a t a ) , A r c t i c Avens (Dryas i n t e g r i f o l i a ) and the Sudeten Lousewort (Pedicularis sudetica). Bl u f f s which are also the sit e s of l a t e - l y i n g snowpatches, especially on northeast-facing exposures, are characterized by the appearance of such additional species as the Lowly Fleabane (Erigeron humilis) and the Small-flowered P r a i r i e Rocket (Erysimum inconspicuum). Towards the upper portions of these same c l i f f s the vegetation i s dominated by the Diamond-leaf Willow ( S a l i x pulchra) which appears to colonize the c l i f f s from the headland surface above. C l i f f s which have been protected from erosion for much longer periods of time are covered by dense willow thickets similar in composition to those found in the f i n a l colonization of the mudslumps. The greatest areal extent of the strand communities, i s developed i n the v i c i n i t y of the large sandspits (Figure 10) where two major habitats can be distinguished: gravel bars and marsh-lagoons (Plates V-A, V-B). The seaward margins of the sandspits have the same.un-favourable environment as the active beaches, and are subject to the same diurnal and seasonal fluctuations of ice-scouring and wave abrasion. On the higher, less-active portions of the gravel accumulations, as much as 90 per cent of the surface may s t i l l be devoid of any form of vegetation cover. Among the f i r s t plants to colonize these surfaces i s the Lyme Grass (Elymus arenarius ssp. m o l l i s ) , the fibrous roots of which penetrate deeply into the well-drained substratum. Other characteristic species i n this environment, and commonly forming prostrate mats on patches of finer material or r a f t s of washed organic material eroded from the adjacent sections of the coastline, include the Sea-beach Sandwort (Arenaria  peploides), Low Chickweed ( S t e l l a r i a humifusa), Beringian Chickweed (Cerastium beeringianum), Common Scurvy Grass (Cochlearia o f f i c i n a l i s ) and the Alpine Milk-vetch (Astragalus alpinus). On the inner margins of the sandspit, bordering the marsh-lagoon, the percentage of bare ground decreases, and there i s an increase in the number and variety of the plant species. The habitat i s s t i l l subject to periodic flooding during storm surges, but the sands are f i n e r , moister and frequently contain admixtures of organic material. Lyme Grass is less prominent i n these areas i n which the most common species are the A r c t i c Marsh Willow ( S a l i x a r c t ophila ), Creeping Sedge (Carex chordorrhiza), Beach Pea Plate V V E G E T A T I O N TYPES A. STRAND COMMUNITIES. S t e r i l e surface of the sandspit i s l o c a l l y colonized by t u f t s of Lyme Grass and r a f t s of the Sea-beach Sandwort. The marsh-lagoon habitat occurs between the sandspit and the s t a b i l i z e d c o a s t a l b l u f f s i n the background. C. LOW-CENTRED POLYGON. The margins of the c e n t r a l depress-ion, often occupied by water, and the is l a n d s of vegetation i n the centre are composed of Aquatic Sedge and F i s h e r ' s Dupontia. The r a i s e d edge of the polygon i s char a c t e r i z e d by lichens and mosses. B. STRAND COMMUNITIES . The vegetation cover increases on the f i n e r sands along the inner margins of the sandspit, and the species i n t e r d i g i t a t e with those of the lagoon-marsh h a b i t a t . D. HIGH-CENTRED POLYGON. Convex polygon centre i s covered by marbeloid hummocks of l i c h e n and moss together with the Glandular B i r c h and willow species. The wedge area i n the foreground i s almost completely overgrown with sedges, mosses and willows. 82 (Lathyrus maritimus), Alpine Hedysarum (Hedysarum alpinum amer i c anum), Scentless Mayweed (Matricaria ambigua), Dawson Hemlock Parsley (Conioselinum cnidiifolium) , Pale Paint Brush (Castillej;a p a l l i d a ssp. elegans), Seaside Crowfoot (Ranunculus cymbalaria var. alpinus), A r c t i c Fireweed (Epilobium l a t i f o l i u m ) , Roseroot (Sedum rosea ssp. i n t e g r i f o l i u m ) , Gentian (Gentiana arctophila) and the Greenland Primrose (Primula  egaliksensis). L o c a l l y , the junction between the sandspit and the coastal b l u f f i s marked by the growth of t a l l stands of the Alaskan or Fe l t - l e a f Willow ( S a l i x alaxensis) as much as 3-4 metres (9-12 feet) t a l l . The other major habitat within the strand communities includes the marsh-lagoons enclosed behind the gravel bars. These s i t e s , which are also subject to periodic flooding, have numerous bodies of shallow, open water which may or may not be connected to the sea. The s o i l , which is almost always saturated, contains abundant organic material, partly due to the natural accretion of vegetation debris, and partly as a result of the accumulation of waterborne material eroded from the coastal b l u f f s . Many of the plants bordering the moist sand areas are also found along the periphery of the lagoons but they are gradually replaced by a f l o r a in which the most ch a r a c t e r i s t i c species are the Lowly Fleabane (Erigeron  humilis), Common Horsetail (Equisetum arvense), Hair-grass-like Reedgrass (Calamagrostis deschampsioides), A r c t i c Rush (Juncus alpinus), A l k a l i Grasses ( P u c c i n e l l i a ander s o n i i , P . phryganodes , P_. v agin at a) , Common Mare's t a i l (Hippuris v u l g a r i s ) , Marsh Cinquefoil ( P o t e n t i l l a p a l u s t r i s ) , Sea-Pink (Armeria maritima ssp. a r c t i c a ) , Yellow Marsh Saxifrage (Saxifraga h i r c u l i s ) , Nodding Saxifrage (S_. cernua) , Pallas's Buttercup (Ranunculus p a l l a s i i ) and Fisher's Dupontia (Dupontia f i s h e r i ) . 83 X. Polygonal Ground. The plant communities associated with polygonal ground are intimately related to the stage of development of these features and can be c l a s s i f i e d accordingly. The e a r l i e s t stage with which a d i s t i n c t i v e vegetation pattern can be linked i s the low-centred type (Plate V-G), the characteristic form of which consists of a central depression surrounded by a raised rim of variable width and height. In many polygons of this type, the central depression may be occupied by a shallow body of open water which may be completely devoid of any vegetation. This water frequently has a d i s t i n c t i v e golden-brown colour due to certain algae, the accumulated remains of which form a layer of s o f t , oozy material on the floor of the pool. Occasionally, isolated tufts of the Aquatic Sedge (Carex aquatilis) and Fisher's Dupontia (Dupontia f i s h e r i ) occur as small islands of vegetation i n the water bodies. Around the shallower margins of the pool the vegetation i s dominated by the Tawny Arctophila (Arctophila fulva) together with the Aquatic Sedge and Fisher's Dupontia. The surrounding ridges are composed of peat, and this drier habitat i s reflected i n a completely different f l o r a l assemblage dominated by mosses and lichens. The major vascular species growing on the ridges include the Veiny-leafed Willow (S a l i x phlebophylla) , Alpine Bearberry (Arctostaphylos  rubra), Cloudberry (Rubus chamaemorus), A r c t i c Blueberry (Vaccinium  Uliginosum var. alpinum), Mountain Cranberry (Vaccinium v i t i s - i d a e a ) , Narrow-leafed Labrador-tea (Ledum palustre ssp. decumbens), Glandular Birch (Betula glandulosa), A r c t i c Avens (Dryas i n t e g r i f o l i a ) and the A r c t i c White Bell-heather (Cassiope tetragona). The troughs overlying the ice-wedges between the polygons also contain bodies of open water which may be p a r t i a l l y f i l l e d with a shallow 84 blanket of Sphagnum and sedge. In this aquatic habitat, the f l o r a i s similar to that found i n the central pools, together with the Gmelin's Buttercup (Ranunculus g m e l i n i i ) , Marsh Marigold (Caltha palustris ssp. a r c t i c a ) , Richardson's Anemone (Anemone r i c h a r d s o n i i ) , Meadow Bittercress (Cardamine pratensis), Nodding Saxifrage (Saxifraga cernua) and the Common Cotton-grass (Eriophorum angustifolium). With increasing passage of time the central pools of the low-centred polygons become progressively shallower and they are gradually colonized by the plants growing around their margins. The accumulation of vegetable debris, c h i e f l y the Aquatic Sedge and Fisher's Dupontia, raises the le v e l of the central area u n t i l i t i s f l a t and equal in height to the surrounding ridges. At the same time, there is a l a t e r a l spreading of the species from the dry ridges onto the central area, and this association gradually replaces the sedges. Further accumulation of peat i n t e n s i f i e s the dry nature of the habitat, and may eventually give the polygon the convex form of the high-centred type (Plate V-D). SUMMARY The a r c t i c tundra vegetation of Garry Island can be c l a s s i f i e d into ten major types which can be related to variations i n the physical character of the habitats. Drainage conditions and geomorphic a c t i v i t y , acting singly or, more often i n combination, appear to afford the best c r i t e r i a for the delineation of the habitats, with the possible exception of the 'Alnus crispa' type where shelter may be the most important factor. Seven of the habitats, corresponding to the 'Dryas-Hummocks', 'Cassiope Snowpatch', 'Eriophorum Tussock', 'Hummock-Tussock T r a n s i t i o n 1 , 8 5 'Stream Course Willow Thickets', and 'Sedge-Moss F l a t s ' vegetation types, are recognized primarily on the basis of variations i n the moisture con-ditions of the habitat. The influence of geomorphic a c t i v i t y i s also evident in each of these types however. Hummocks and tussocks are generally regarded as products of fros t action, and the microrelief produced by the development of these features i s s u f f i c i e n t to produce wet and dry habitats within the same vegetation type. In the case of the hummocks, the influence of the microrelief factor becomes less pronounced towards the lower, moister sections of the slopes, and i t v i r t u a l l y dis-appears, as i n the 'Cassiope Snowpatch' type, i n the sites of l a t e - l y i n g snow patches. The 'Stream Course Willow Thickets' and the 'Sedge-Moss F l a t s ' , vegetation types are found along the floors of the valleys and depressions into which intermittent runoff and moisture seepage are channelled throughout the summer months. In the 'Strand Communities', 'Mudslump Communities' areas of 'Polygonal Ground', the physical character of the habitat i s determined primarily by geomorphic a c t i v i t y . Within the 'Strand Communities' two contrasting habitats can be i d e n t i f i e d : these are the gravel bars, which are subject to diurnal and seasonal fluctuations of wave-abrasion and ice-scouring, and the marsh-lagoon areas, enclosed and protected by the bars, where the conditions are r e l a t i v e l y quiescent. Differences i n the vegetation patterns of these two environments may also be related to variations i n the moisture conditions of the substrate. In the 'Mudslump Communities' there appears to be a d i s t i n c t i v e sequence of plant colonization related to the length of time that the mudslump has been inactive. Thus the vegetated sections of active mudslumps are dominated by almost pure stands of the Marsh Fleawort (Senecio lugens). Once the floor of the mudslump i s no longer subjected to mudflows or deposition by running water, the surface becomes colonized by a cover of grasses and, following even greater periods of s t a b i l i t y , this grass vege-tation i s eventually replaced by willow thickets. In the areas of poly-gonal ground, the growth of ice-wedges also results i n the production of wet and dry habitats, and another d i s t i n c t i v e plant colonization sequence can be i d e n t i f i e d through which the polygons are transformed from low-centred to high-centred types. The s t a t i s t i c a l v a l i d i t y of the percentages for the dominant plant species i n each vegetation type i s limited due to the subjective method of v i s u a l estimation employed. Despite the crude nature of these percentages however, i t was possible to recognize four species, Dryas  i n t e g r i f o l i a , Alnus cr i s p a, Cassiope tetragona and Eriphorum vaginatum, as being diagnostic of s p e c i f i c vegetation types on the basis of their dominance of the plant composition. Though not dominated by a single species, the 'Strand Communities' were also characterized by a d i s t i n c t i v e f l o r a l assemblage probably due to the s t r i c t e r tolerance conditions produced by the presence of saline or brackish water. Furthermore, l i t t l e d i f f i c u l t y was encountered i n determining the areal extent of the communities i n the f i e l d . Even where the c l a s s i f i c a t i o n was based on changes i n the drainage conditions of the habitat, i t was found that the contacts between adjacent vegetation types were quite d i s t i n c t . Only where the lower slopes were gentle, and drainage conditions changed gradually, were there extensive t r a n s i t i o n zones necessitating the recognition of the 'Hummock-Tussock Transition' type. Where the habitat was determined primarily by geomorphic a c t i v i t y , the vegetation type usually formed a d i s t i n c t unit and there was no problem i n determining the 87 precise areal extent of the plant community. CHAPTER IV PERMAFROST CONDITIONS The term permafrost, or permanently frozen ground, i s used to describe a thickness of s o i l or other s u p e r f i c i a l deposit, or even of bed-rock, at a variable depth beneath the surface of the earth i n which a temperature below freezing has existed continually for a long time (from two to tens of thousands of years).''' This thermal condition may be wholly contemporaneous with the existing climate, or i t may be a r e l i c feature which was i n i t i a l l y developed during an e a r l i e r , colder climate and has been preserved under the negative mean annual temperatures of the present climate. Approximately o n e - f i f t h of the land area of the world, and as much as one-half of the area of Canada, i s underlain by permafrost i n 2 either a continuous or discontinuous d i s t r i b u t i o n . Whilst the areal extent of the perennially frozen ground i s reasonably w e l l established i n Canada, only a limited amount of data i s available r e l a t i n g to i t s v e r t i c a l thickness. Largely as a re s u l t of townsite investigations and o i l exploration d r i l l i n g programmes, more i s known about the thickness of ^Muller, S.W. (1947) Permafrost or permanently frozen ground  and related engineering problems, J.W. Edwards, Inc., Ann Arbor, Michigan, p. 3. 2 Jenness, J.L. (1949) "Permafrost i n Canada", A r c t i c , Vol. 2, p. 13. Geological Survey of Canada (1967) Permafrost i n Canada, Map 1246A, F i r s t E d i t i o n , Scale 1: 7,603,000. 89 the frozen ground i n the Mackenzie Delta area than i n other parts of the Canadian A r c t i c . Brown has reported permafrost to depths of 76-91 metres (250-300 feet) i n the v i c i n i t y of Inuvik, and Mackay thicknesses of 3 107-122 metres (350-400 feet) near A r c t i c Red River and Fort McPherson. Temperature data from a d r i l l hole near Tununuk on Richards Island reveal 4 that permafrost occurs to a depth of 366 metres (1200 fee t ) . Thus, although no information i s available concerning the thickness of perma-fro s t on Garry Island, i t may well be that the frozen ground extends to a depth of 305 metres (1000 feet) or more beneath the centre of the island. The precise thickness of the frozen ground i s of r e l a t i v e l y minor importance to the geomorphic processes operating i n permafrost areas. Of greater significance i s the active layer which thaws during the summer months and i s subject to alternate freezing and thawing during successive seasons. The thickness of this active layer i s quite variable and i s related to differences i n slope, natural drainage, aspect, type of vegetation cover and the nature of the material beneath the surface vegetation. The purpose of this chapter i s to describe a number of investigations made on Garry Island designed to: (1) determine the rate and depth of thaw i n the active layer under varying slope, exposure, JBrown, R.J.E. (1966) "Relation between mean annual a i r and ground temperatures i n the permafrost region of Canada", Proc: Permafrost International Conference, Lafayette, Indiana, November, 1963 Nat. Acad, of Sciences - Nat. Research Council, Washington Pu b l i c a t i o n , No. 1287, p. 243. Mackay, J. Ross (196 7) "Permafrost depths, Lower Mackenzie Valley, Northwest T e r r i t o r i e s " , A r c t i c , Vol. 20, pp. 21-26. 4 Mackay, J. Ross. Personal communication, October, 1968. 90 vegetation, moisture and microrelief conditions; and (2) measure the ground temperatures at selected sit e s i n th i s active layer i n order to better understand the processes involved i n the development of earth hummocks. These studies were accomplished primarily by making compara-tive investigations of the depths of thaw, at various times throughout the summer, beneath each of the major vegetation types i d e n t i f i e d i n Chapter I I I . The depth to the fr o s t table was determined by using a metal rod as a probe. Similar measurements, though using a network of wooden stakes and on a daily basis, were also made to examine changes i n the configura-tion of the f r o s t table beneath the surface of a small vegetation plot 5.6 square metres (60 square feet) i n area. To complement these observa-tions, excavations were made to determine the nature of the substrate. Five strings of thermistor cables were also i n s t a l l e d to different depths beneath the centres of an earth hummock, two inter-hummock depressions, and a mud b o i l and adjacent depression, i n an attempt to detect variations i n the thermal regime of the seasonally thawed layer. DEPTH OF THAW MEASUREMENTS A number of authors have investigated and documented the effects of a vegetation cover on the thermal regime of the underlying 91 ground."' These investigations have shown that the influences of the vegetation cover on permafrost are exceedingly complex, and that quantita-tiv e evaluations of these influences are extremely d i f f i c u l t to obtain. One of the most readily measurable characteristics of these relationships i s the va r i a t i o n i n the thickness of the active layer beneath contrasting surface covers. The data presented i n Table V show some of the variations in the thickness of the active layer beneath the major vegetation types on Garry Island as they were recorded at the end of the 1964 f i e l d season. As the table shows, the greatest depths of thaw were encountered i n areas having- l i t t l e or no vegetation cover, as exemplified by the bare ground of some of the mudslumps and active sand-s p i t s . In both these type l o c a l i t i e s , the position of the fro s t table was i n excess of one metre (3 feet) below the ground surface and, i n the case of the sandspit areas, could not be detected by probing. The influence of the type of mineral s o i l i n the substrate may also be reflected i n the greater depth of thaw i n the coarse sands and gravels of the strand areas compared with the s i l t - and clay-sized material found on the floors of the mudslumps. Comparisons of the thickness of the active layer beneath the Benninghof, W.S. (1952) "Interaction of Vegetation and S o i l Frost Phenomena", A r c t i c , Vol. 5, pp. 34-44. Tyrtikov, A.P. (1959) "Perennially Frozen Ground and Vegetation", P r i n c i p l e s of Geocryology, (Permafrost Studies), Part I , General Geocryology, Acad. S c i . U.S.S.R., Moscow, pp. 399-421. Translation by R.J.E. Brown, National Research Council of Canada, Technical Translation 1163, 1964, pp. 1-34. Brown, R.J.E. (1966) "Influence of Vegetation on Permafrost", Proc: Permafrost International Conference, Lafayette, Indiana, op_. c i t . , pp. 20-25. TABLE V DEPTHS OF THAW BENEATH THE MAJOR VEGETATION TYPES, GARRY ISLAND, SEPTEMBER 1, 1964. Vegetation Type Depth of Thaw (Cms.) Dryas-Hummock Hummocks Depressions Mud B o i l Eriophorum Tussocks Tussocks Depressions Sedge-Moss Flats Stream Course ) Willow Thickets Mudslump Communities Bare Ground Grass Cover Strand Communities Sandspits Lagoon Fl a t s 45 22 70 65 30 75 45 - 60 18 - 25 18 25 22 30 > 100 45 - 50 > 100 40 - 50 93 vegetated surfaces show that the amount of thawing was greatest beneath the domed centres of the Dryas-Hummocks and tussock-Like forms of the Sheathed Cotton-grass (Eriophorum vaginatum). In actual f a c t , i t i s extremely d i f f i c u l t to provide a meaningful average for these l o c a l i t i e s , and this i s i l l u s t r a t e d by the wide ranges indicated in Table V. The depth of thaw i s influenced considerably by the microrelief factor and, in general, the larger the hummocks and the more prominent the tussocks, the greater the depth of thaw. Because of the variations introduced by this microrelief factor, i t i s , however, v i r t u a l l y impossible, using simple probing techniques, to evaluate precisely the r e l a t i v e effects of the various vegetation types on the thermal regime of the underlying ground. One noticeable aspect of Table V i s the uniformly lower depths of thaw recorded beneath vegetation associations with a considerable per-centage of moss i n their f l o r a l composition. As described i n the previous chapter, the shallow depressions between the earth hummocks are f r e -quently dominated by mosses and accumulations of organic debris. Such l o c a l i t i e s were characterized by depths of thaw which were frequently less than one-half of those recorded i n the raised centres of the hummocks. This relationship was also observable on the summit areas and on south-facing slopes where the hummocks were'often more subdued i n form and, consequently, the microrelief factor was of r e l a t i v e l y less importance. A similar pattern was also observed i n the Eriophorum-Tussock communities where the inter-tussock depressions were again characterized by depths of thaw which were usually less than one-half of those measured i n the adjacent tussocks. Further evidence of the insu-l a t i n g property of a moss cover on the underlying ground was demonstrated 94 by the fact that the average depth of thaw beneath the major vegetation types was lowest, 18-22 cms. (7.0-8.5 i n s . ) , beneath the featureless surfaces of the sedge-moss f l a t s . The figures presented in Table V, despite their l i m i t a t i o n s , corroborate the findings of other investigators and.demonstrate that the thickness of the active layer i s greatest beneath bare ground surfaces, and that a cover of mosses allows less heat to penetrate to greater depths i n the underlying ground than does a vegetation cover dominated by vascular plants. The above descriptions are related to measurements of the thickness of the active layer at a number of selected, isolated points. To complement these observations, a vegetation plot was established during the 1964 f i e l d season, to examine the influences of some of these same factors on the configuration of the fros t table over a small contiguous area. The dimensions of this plot were approximately 0.9 metres by 6.1 metres (3 feet by 20 feet) , and i t was located on the lower part of the west-facing slope of a v a l l e y , close to the junction with the valley f l o o r . A series of wooden stakes, graduated i n centimetres and spaced at 30 cm. (1 foot) i n t e r v a l s , were i n s t a l l e d i n the plot and were driven into the ground u n t i l they were halted by the fros t table, at which point the depth of penetration was recorded. This procedure was repeated on a daily basis over the period July 9 - August 28, 1964. Figure 11 i l l u s t r a t e s the surface contours and the vegetation The vegetation plot constituted part of the f i e l d study programme of Dr. J . Ross Mackay and Dr. J.K. Stager, of the University of B r i t i s h Columbia, and I am most grateful to them for permission to refer to their studies i n this thesis. Figure 11 VEGETATION PLOT 96 cover of this plot. Most of the surface was dominated by the dwarf shrubs of the Ericaceae or Heath Family including the Alpine Bearberry (Arctostaphylos rubra), A r c t i c White Bell-heather (Cassiope tetragona), Narrow-leafed Labrador-tea (Ledum palustre ssp. decumbens), A r c t i c Blueberry (Vaccinium uliginosum var. alpinum), and Mountain Cranberry (Vaccinium v i t i s - i d a e a ) , together with minor quantities of lichens and Arc t i c Avens (Dryas i n t e g r i f o l i a ) . This f l o r a covered a number of sub-dued earth hummock forms, similar to those described i n the previous chapter, the central portions of which were elevated by 20-30 cms. (8-12 ins.) above the surrounding depressions. The plant associations occurring i n these depressions were composed p r i n c i p a l l y of mosses and the A r c t i c White Bell-heather. The other p r i n c i p a l constituent of the f l o r a of the vegetation plot was the d i s t i n c t i v e tussock-like forms of Sheathed Cotton-grass (Eriophorum vaginatum). The central part of the plot was dominated by the unvegetated surface of a prominent mud b o i l . The depth of thaw measurements are shown i n Figure 12, showing the configuration of the frost table at the beginning of August and September, 22 and 53 days after the i n s t a l l a t i o n of the stakes respectively. The isolin e s on these maps represent the depth of the frost table beneath the ground surface, and the higher values thus correspond to depressions in the frost table. The greater d e t a i l shown in the map for August 1, r e f l e c t s the number of observations made at intermediate points between the stakes, whereas the map for September 1 was based solely on the data obtained i n the excavations of the pr o f i l e s shown i n Figure 13. Reference to Figure 12 reveals that the depth of thaw was quite variable and imparted an irregular topography to the top of the ; 97 Figure 12 VEGETATION PLOT P O S I T I O N OF FROST TABLE (DEPTH BELOW G R O U N D S U R F A C E IN CMS.) Aug. 1, 1964 Sept. 1, 1964 Contour Interval 5 cms. Scale 0 20 40 60 80 Cm». 98 Figure 13 VEGETATION PLOT D E P T H O F T H A W P R O F I L E S Mineral Soil Scale 0 20 40 60 80 Cms. For locations of profiles see Figure 11 i i • 4 99 frozen ground. Comparison with Figure 11 shows that the contours of the frost table do not p a r a l l e l the configuration of the ground surface. In general, the upper surface of the frozen ground occurs at greater depths where the ground surface i s highest, and i t l i e s at shallower depths where the ground surface i s lowest. Thus the frost table i s frequently deeper beneath the earth hummocks and tussocks than beneath the i n t e r -vening depressions. The inverse relationship between the configurations of the ground surface and the permafrost surface undoubtedly r e f l e c t , at least in part, the significance of the microrelief factor. The elevated nature of the earth hummocks and tussocks allows heat to penetrate the ground l a t e r a l l y from their sides, as wel l as v e r t i c a l l y from the top, thus contributing to a greater rate of thawing. The undulations i n the frost table cannot, however, be attributed solely to the influence of variations i n the microrelief. The rate and t o t a l depth of thaw are also influenced by the type of material occurring i n the substrate. Figure 13 shows a t o t a l of 21 cross-sectional p r o f i l e s in the vegetation plot which were excavated at the beginning of September, 1964, when the position of the frost table was approximately coincident with the Upper surface of the perennially frozen ground. As these p r o f i l e s i l l u s t r a t e , there i s apparently also an inverse relationship between the t o t a l depth of thaw and the thickness of organic material i n the p r o f i l e . Thus the greatest depth of thaw, i n excess of 75 cms. (29.5 ins.) was recorded beneath the surface of the mud b o i l where the substrate was composed en t i r e l y of mineral s o i l , and the next greatest depths of thaw were recorded beneath the earth hummocks which were also composed pre-dominantly of mineral s o i l . As the r a t i o of organic material to mineral 100 s o i l i n the p r o f i l e s increased, the depth of thaw decreased accordingly, and the elevated parts of the permafrost surface, corresponding to the lowest depths of thaw i n Figure 12, occurred beneath the accumulations of organic material occupying the depressions i n the ground surface. Earth Hummock Experiments. In the preceding paragraphs the discussion was centred on variations i n the t o t a l depth of thaw beneath each of the major vegetation associations found on Garry Island. These va r i a t i o n s , whilst providing an indicator of the effects of contrasts i n the f l o r i s t i c composition, must equally r e f l e c t the numerous other climatic and t e r r a i n features, which also combine to influence the depth of thaw from the ground surface to the permafrost table. A number of experiments was undertaken during the 1965 f i e l d season to investigate some of the factors influencing the rate and depth of thaw beneath the centre of a t y p i c a l earth hummock. S p e c i f i c a l l y these experiments were designed to examine the influence of the vegetation cover, l i v i n g and dead, and the addition of controlled quantities of water on the thickness of the thawed layer. The f i r s t set of experiments was designed to examine the role of the vegetation cover alone. An attempt was also made to di f f e r e n t i a t e between the effects of the surface mat of l i v i n g vegetation and the under-lying accumulation of dead organic material which frequently overlies the domed core of mineral s o i l i n a t y p i c a l earth hummock. The success i n making this d i s t i n c t i o n was r e s t r i c t e d by the fact that, as others have noted, i t i s extremely d i f f i c u l t to delineate the boundary between these 101 l i v i n g and dead components of the organic layer.'' Six hummocks were selected for these studies, each of which occupied a similar position with respect to aspect and general location on the topographic slope p r o f i l e . ' In addition, to eliminate the possible influence of variations produced by the microrelief factor, each of the hummocks was of very similar dimensions, being about 70 cms. (27.5 ins.) in length, 55 cms. (21.5 ins.) in width and having a raised centre which projected about 35 cms. (14 ins.) above the l e v e l of the adjacent in t e r -hummock depressions. From two of these hummocks, the surface vegetation layer, averaging 7 cms. (3 ins.) in thickness, was removed leaving the raised centre s t i l l covered by a layer of 4-5 cms. (1.5-2.0 ins.) of peat material. From two of the other hummocks the complete organic cover of l i v i n g vegetation and peat was removed, exposing the mineral s o i l core. The remaining two hummocks were l e f t intact to act as controls. The depth from the ground surface to the frost table was recorded at weekly intervals over a six-week period and the results are shown i n Table VI. In this table, the values given for the hummocks from which the organic layer was either p a r t i a l l y or completely removed do not include the thicknesses of the removed layers. As the figures i n Table VI indicate the response to the p a r t i a l or complete removal of the organic material was quite rapid. During the f i r s t week of the experiment, although characterized by only Brown, R.J.E. (1966) "Influence of Vegetation on Permafrost", op. c i t . , p. 20. 102 TABLE VI DEPTHS OF THAW BENEATH EARTH.HUMMOCKS ON GARRY ISLAND, JULY 19 - AUGUST 30, 1965. Type of Cover Depth of Thaw (Cms.) J u l y August 19 26 2 9 16 23 30 T o t a l Change (a) N a t u r a l 51 54 55 58 59 60 61 52 54 54 57 58 60 62 + 10 + 10 (b) L i v i n g V egetation Removed 40 49 51 55 56 59 60 39 48 50 53 55 57 58 +20 +19 (c) Complete Organic Layer Removed 37 47 51 55 57 59 61 +24 36 47 50 54 55 58 59 +23 103 g 85 thawing degree-days, the lev e l of the frost table i n the control hummocks was lowered by 2-3 cms. (1 inch), whereas i n the hummocks from which the organic layer was either p a r t i a l l y or completely removed the corresponding values were 9 cms. (3.5 ins.) and 10-11 cms. (4.0-4.5 ins.) respectively. Over the remainder of the observation period the d i f f e r e n t i a l rates were not always as marked, though they remained quite s i g n i f i c a n t . The depth of thaw beneath the centres of the control hummocks increased by an average of 10 cms. (4 ins.) during the six-week period i n response to a t o t a l of 560 degree-days of thawing. The removal of the surface mat of l i v i n g vegetation resulted, however, i n a lowering of the frost table by an average of 19.5 cms. (7.5 i n s . ) , whilst the complete removal of the entire organic layer produced an average increase in the depth of thaw of 23.5 cms. (9.5 ins.) over the same time span. Assuming that the subsurface conditions.in each of the hummocks were the same, with respect to the type of mineral s o i l and frozen moisture content, these figures demonstrate, quite conclusively, the important influence which a s u r f i c i a l layer of organic material exerts on the thermal regime of the underlying ground. The complete removal of this insulating layer of vegetation increased the depth of thaw by 135 per cent over that recorded i n the control hummocks. Even where this layer was only p a r t i a l l y removed, the corresponding increase in the depth of thaw was 95 per cent. Due to the afore-mentioned problem Thompson, H.A. (1963) "Freezing and Thawing Indices i n Northern Canada" , Proc. of the F i r s t Canadian Conference on Permafrost, Ottawa, A p r i l 17-18, 1962, p. 21. The number of degree-days for any one day i s the difference between the average daily a i r temperature and 32°F and thawing degree-days occur when the former exceeds 32°F. 104 of distinguishing accurately the boundary between the l i v i n g and dead components of the organic material, i t cannot be stated affirmatively that the d i f f e r e n t i a l between the two percentage increases can be a t t r i b u -ted solely to the r e l a t i v e influences of these two constituents. I t does indicate, however, that a thin layer of peat, 4-5 cms. (1.5-2.0 ins.) i n thickness, had s u f f i c i e n t insulating effect to resu l t i n a depth of thaw which was almost 17 per cent less than that recorded i n the hummock from which the organic layer was removed e n t i r e l y . The second part of the experiment was designed to examine the influences of aspect and moisture, singly and i n combination, on the depth of thaw. A t o t a l of twenty earth hummocks was used for these investiga-tions and again, i n an attempt to achieve as much uniformity i n the other variables as possible, each hummock was of approximately the same dimen-sions and occupied a similar position on the topographic slope p r o f i l e . Nine of these hummocks were located on the southwest-facing slope of a va l l e y , and another nine were located on the opposing northeast-facing slope. Three of the hummocks i n each of these groupings were sprinkled with one l i t r e of water d a i l y , another three received two l i t r e s of water each day whilst the remaining three,left unwatered,were used as controls. In addition, two hummocks, also located on the southwest-facing slope, were p a r t i a l l y or completely stripped of their vegetation cover, as des-cribed i n the previous experiment, and were sprinkled with one l i t r e of water d a i l y . The temperature of the water was recorded before i t was applied to the hummock surfaces. These observations were continued over a five-week period, and the depths of thaw were measured at weekly int e r v a l s . One additional problem was introduced by these measurements, insomuch as the insertion of the metal probe to the frost table might 105 leave an open conduit presenting easy penetration of the water. An attempt was made to prevent this by plugging the hole at the surface with a length of wooden dowelling, but i t cannot be claimed that these pre-cautions were completely successful. The results of this project are presented i n Table V I I , and as the figures show the experiments yielded inconclusive r e s u l t s . Comparisons of the measurements i n the control hummocks on the two opposing slopes reveal no influence of aspect as the average lowering of the fr o s t table, 8.67 cms. (3.5 ins.) was i d e n t i c a l on both sides of the va l l e y . S i m i l a r l y , the addition of water to the hummock surfaces appeared to have few con-trasting effects on the two sides of the valley. On the southwest-facing slope the application of one l i t r e of water daily resulted i n a s l i g h t increase i n the amount of lowering of the frost table, whereas a doubling of the quantity of water applied produced a s l i g h t decrease. In each case the differences represented a deviation of less than one centimetre from the value recorded i n the control hummock. The application of water to the hummocks on the northeast-facing slope resulted i n s l i g h t decreases in the depth of thaw irrespective of the quantity of water added, though, once again the differences from the control value involved a maximum of one centimetre. Although the magnitude of the changes i s small, i t appears that the application of one l i t r e of water to the hummocks on the northeast-facing slope produced a larger reduction i n the depth of thaw than the application of two l i t r e s . Comparisons of Tables VI and VII give some indication of the effects of adding water to the surfaces of earth hummocks from which the vegetation cover had been either p a r t i a l l y or completely removed. In the hummock from which only the surface layer of l i v i n g vegetation was 106 TABLE VII EARTH HUMMOCK WATERING EXPERIMENT, GARRY ISLAND, JULY 26 - AUGUST 30, 1965. Type of Treatment Depth of Thaw (Cms.) July August Net 26 2 9 16 23 30 Changi A. Southwest-facing Slope (a) Control 50 51 53 55 57 58 + 8 53 55 58 60 61 62 + 9 52 53 57 56 58 61 + 9 (b) 1 L i t r e of water 48 49 52 54 56 57 + 9 added da i l y 50 52 53 54 56 58 + 8 52 53 55 56 58 62 +10 (c) 2 L i t r e s of water 48 50 50 52 53 55 + 7 added d a i l y 50 51 51 53 55 58 + 8 55 56 59 60 63 64 + 9 (d) Surface vegetation removed. 1 l i t r e 43 45 49 50 52 52 + 9 added d a i l y . (e) Complete organic layer removed. 1 l i t r e 44 45 48 51 53 54 +10 added d a i l y . B. Northeast-facing Slope (f) Control 52 52 55 56 60 60 + 8 49 51 55 56 58 59 +10 49 51 52 53 55 57 + 8 (g) 1 L i t r e of water 50 52 53 55 57 58 + 8 added da i l y 50 52 55 55 58 59 + 9 55 57 59 59 61 61 + 6 (h) 2 L i t r e s of water 55 57 60 61 64 65 +10 added da i l y 56 58 60 60 64 63 + 7 55 56 58 60 63 63 + 8 107 stripped, the observed lowering of the frost table amounted to 9 cms. (3.5 i n s . ) , whereas the removal of the entire organic material resulted in a corresponding figure of 10 cms. (4 i n s . ) . In each case these were 1.5 and 3.0 cms. (0.5 and 1.0 ins.) less than those recorded over the same time period i n the i d e n t i c a l l y prepared hummocks to which no water was added. The application of 35 or 70 l i t r e s of water respectively, with an average temperature of 8.25°C, to the two groups of hummocks over the five-week period represented a considerable potential source of heat to the underlying ground. Unfortunately, i t i s impossible to infer whether this water was able to penetrate into the mineral s o i l core of the hummock, or whether i t was absorbed by the s u p e r f i c i a l mantle of organic material. In the case of the hummock from which the vegetation cover was removed completely, i t was readily observed that, despite the li g h t sprinkling nature of the application, some of the water was lo s t by surface runoff. The effect of any water which was able to penetrate the organic layer would also depend on the existing moisture content of the s o i l . The replacement of the a i r i n the pores of the s o i l by water would have the effect of increasing the heat capacity of the s o i l , thereby reducing the amount of heat available to warm the s o i l at greater depths. I t seems reasonable to i n f e r , however, that most of the water was probably absorbed by the layers of organic material and was subse-quently lost by evaporation, including evapotranspiration. Since the evaporation process requires heat, which may be drawn from the surround-ing atmosphere, vegetation or s o i l , Tyrtikov has postulated that this may result i n a lowering of the a i r temperature near the ground surface and 108 consequently a reduction i n the warming of the s o i l . Whether or not this postulate can be applied to such a small area as the surface of a hummock is debatable, but, i f v a l i d , i t may account for the s l i g h t cooling effect of the water applied to the hummocks. GROUND TEMPERATURE MEASUREMENTS The preceding discussion has been based on observations made in the rate and t o t a l depth of thaw beneath each of the vegetation types. A series of observations was also made to determine more precisely the variations in the thermal regime of the thawed layer above the permafrost and the uppermost parts of the permafrost. Five sit e s were selected for these measurements to evaluate the s p e c i f i c effects of certain types of surface cover. Three of the sites were i n one of the t y p i c a l hummock-depression associations found i n the Dryas-Hummock vegetation type. The raised centre of the earth hummock, almost c i r c u l a r in plan and 60-70 cms. (23.5-27.5 ins.) i n diameter, reached elevations of 30-35 cms. (12-14 ins.) above the l e v e l of the surrounding depressions, and supported a surface cover dominated by A r c t i c Avens, Narrow-leafed Labrador-tea, A r c t i c White Bell-heather, together with a few lichens and dry moss pads. This vegetation was rooted i n a s u r f i c i a l layer of 10-15 cms. (4-6 ins.) of organic material overlying a prominent domed core of mineral s o i l extending down to the fro s t table. The adjacent depressions, 50-60 cms. (19.5-23.5 ins.) i n length and 20-40 cms« (8.0-15.5 ins.) across, were dominated by the growth of sphagnum mosses Tyrtikov, A.P. (1959), op_. c i t . , p. 5. 109 resting on a substratum of organic material which extended down almost to the base of the active layer. The remaining two sites were located i n a large mud b o i l and adjacent depression. The mud b o i l surface, 1.4-1.5 metres (4.5-5.0 feet) in diameter, was largely devoid of vegetation and, at the centre, rose approximately 20-25 cms. (8-10 ins.) above the surrounding moss-filled depressions. At each of the sites, ground temperatures were recorded daily 10 by cables of bead thermistors. The cables, each of which consisted of fiv e thermistors, were encased in lengths of rubber tubing which were sealed to prevent the entry of s o i l moisture. One of the cables was in s t a l l e d in the earth hummock and two in the adjacent inter-hummock depressions. The uppermost thermistor on each of these cables was placed at a depth of 10 cms. (4 ins.) below the ground surface. Beneath the raised centre of the earth hummock the other four thermistors were spaced at intervals of 25 cms. (10 ins.) to a depth of 110 cms. (43.5 ins.) whilst beneath each of the depressions the remaining thermistors were spaced at intervals of 15 cms. (6 ins.) to a depth of 75 cms. (29.5 i n s . ) . In the mud b o i l , the uppermost thermistor was i n s t a l l e d at a depth of 23 cms. (9 ins.) below the ground surface and additional beads were spaced at intervals of 30 cms. (12 ins.) to a depth of 113 cms. (44.5 i n s . ) . Only two thermistors were i n s t a l l e d i n the depression adjacent to the mud b o i l at depths of 14 cms. (5.5 ins.) and 24 cms. (9.5 ins.) beneath the A l l the thermistors used were manufactured by the Yellow-stone Springs Instrument Company, Series 401 beads, with a tolerance at 0°C of + 0.2°C. The thermistors were calibrated by the B.C. Research Council and by Dr. J. Ross Mackay of the Department of Geography, U.B.C. Fi e l d readings were taken with a small bridge, also constructed by Dr. Mackay, and calibrated to read with an operational range of "t 0.1°C. 110 ground surface. Figure 14 shows the mean ground temperature profile and the amplitude of the ground temperature fluctuations for four of these sites for the period July 1 to September 12, 1965.^ Table VIII also shows the mean, maximum and minimum ground temperatures, together with the date of occurrence of the maximum and minimum temperatures recorded by each of the thermistors. Each of the sites exhibits a similar pattern in which the highest mean and maximum ground temperatures, and the greatest amplitude of ground temperature fluctuations, were recorded by the thermistors located at the shallowest depth beneath the ground surface. S t r i c t comparisons of the thermal regimes at each of the sites are limited by the absence of ground surface temperature data. With this consideration in mind, Figure 14 shows that the highest mean temperature and greatest fluctuations were recorded at a depth of 10 cms. (4 ins.) beneath the surface of the earth hummock. Such a simple comparison of the profiles is misleading, however, since the uppermost thermistor in the cable installed in the mud b o i l was located at a depth more than twice that of the uppermost thermistor installed in the earth hummock. It is probable, therefore, that the mean summer ground temperature, and the amplitude of the temperature fluctuations, at a depth of 10 cms. (4 ins.) beneath the surface of the mud b o i l , were of even greater magnitude than those i iThe means and amplitudes shown for the mud boil are actually for the slightly longer period June 26 to September 12, 1965. In the depression adjacent to the mud b o l l , continuous readings were only recorded at a single depth, due to a faulty thermistor at a depth of 14 cms. (5.5 Ins.). As a result, the profile at the f i f t h site could not be plotted. I l l Figure 14 MEAN GROUND TEMPERATURES AND GROUND TEMPERATURE FLUCTUATIONS t M i i MUD BOIL _1_ 25 25 50 £ SO O 75 - 7 5 a. 100 100 : t: 10 15 125 125 20 -i 1 — T / E A R T H HUMMOCK J I L 10 15 Degrees Centigrade 5 10 16 Degrees Centigrade Levels of Thermistors. •" " - Mean Ground Temperature Profile • i Amplitude of Ground Temperature Fluctuations 112 TABLE VIII GROUND TEMPERATURES RECORDED AT DIFFERENT DEPTHS IN MUD BOIL, EARTH HUMMOCK AND INTER-HUMMOCK DEPRESSIONS. ( i n Degrees Centigrade) Depth below surface Mean Max. ( cms.) MUD BOIL: 23 4.3 8.2 53 2.4 5.0 83 -0.3 0.9 113 -1.4 -0.5 EARTH HUMMOCK: 10 7.6 15.9 35 2.8 5.8 60 -0.3 1.8 85 -1.0 -0.4 110 -1.9 -1.1 INTER-HUMMOCK DEPRESSION: 10 4.9 10.5 25 0.4 2.0 40 -0.5 0.1 55 -1.0 -0.4 70 -1.4 -0.8 INTER-HUMMOCK DEPRESSION: 10 4.1 10.3 25 0.7 2.7 40 -0.4 -0.2 55 -1.1 -0.4 70 -1.5 -0.8 Date of Min. Date of Ampliti Max. Min. Aug. 10 ^0.5 June 26 8.7 Aug. 24* -1.4 June 26 6.4 Aug. 25* -2.7 June 26 3.6 Sep. . 1* -3.7 June 26 3.2 July 8 -0.4 Sep. 3 16.3 Aug. 24 0.0 July 1* 5.8 Aug. 24 -1.4 July 1 3.2 Aug. 29* -2.4 July 1 2.0 Sep. 3* -3.4 July 1 2.3 Aug. 23 -0.2 Sep. 3 10.7 Aug. 23 -0.6 July 1 2.6 Aug. 21 -1.6 July 1 1.7 Aug. 31 -2.4 July 1 2.0 Aug. 31 -2.9 July 1 2.1 Aug. 23 -0.6 Sep. 3 10.9 Aug. 23 -0.7 July 1 3.4 Aug. 23 -1.7 July 1 1.9 Sep. 12 -2.4 July 1 2.0 Sep. 1 -2.9 July 1 2.1 E a r l i e s t date at which temperature was attained. 113 actually recorded at the hummock s i t e . The most v a l i d comparison of near surface temperatures, which can be made r e l i a b l y , i s between the hummock and adjoining inter-hummock depressions. The mean ground temperature recorded at a depth of 10 cms. (4 ins.) beneath the hummock was 2.7°C and 3.5°C higher than that recorded at an equivalent depth beneath the surfaces of the two depressions. The maximum ground temperature and the amplitude of the temperature fluctuations also exceeded those in the depressions by more than 5°C. Each of the sites also exhibits a similar pattern i n which the mean ground temperature and the amplitude of the temperature fluctuations decreases with increasing depth beneath the ground surface. As Figure 14 shows, however, the rate of decrease i s not the same at each of the s i t e s . The most gradual decrease occurred beneath the centre of the mud b o i l where the mean ground temperature did not reach 0°C u n t i l depths of 75-80 cms. (29.5-31.5 ins.) below the surface, and the amplitude of the fluctua-tions was s t i l l i n excess of 3°C at a depth of 113 cms. (44.5 i n s . ) . These figures contrast with those beneath the surface of the earth hummock where the mean ground temperature reached 0°C at a depth of 60-65 cms. (23.5-25.5 i n s . ) , and amplitudes i n excess of 3°C were not experienced below these depths. The most rapid decrease of ground temperatures, and diminishing of temperature fluctuations, occurred beneath the surfaces of the inter-hummock depressions. The mean ground temperature dropped to 0°C at the shallow depths of 30-35 cms. (12-14 ins.) and amplitudes in excess of 3°C were not recorded below 25-30 cms. (10-12 ins.) from the ground surface. The graphs i n Figure 14 and the values in Table VIII summarize the absolute magnitude of the ground temperature fluctuations at each of 114 the s i t e s over the whole of the observation period. The general picture conceals the number and depth of penetration of minor temperature f l u c t u a -tions which bear a d i s t i n c t temporal r e l a t i o n s h i p to the changes i n the mean d a i l y a i r temperature. Figures 15 and 16 present a more de t a i l e d picture of these minor f l u c t u a t i o n s i n the ground temperatures, by showing the pattern of isotherms over the same time period. The near surface layers at each of the s i t e s underwent a number of d e f i n i t e cycles of warming and cooling over time periods ranging from three to nine days. The largest f l u c t u a t i o n s were recorded beneath the surface of the earth hummock (Figure 15A), where temperature changes of as much as 7-8°C i n one day were observed at a depth of 10 cms. (4 i n s . ) . Maximum temperatures achieved during these cycles r e f l e c t e d an almost immediate response to a warming of the ambient a i r temperatures with lag factors involved being less than one day. The pattern of these minor f l u c t u a t i o n s also shows a decrease i n amplitude with increasing depth beneath the ground surface. At a depth of 35 cms. (14 ins.) , the largest f l u c t u a t i o n s , over s i m i l a r time periods, were, only of the order of 2-3°C or approximately 25 per cent of those recorded at a depth of 10 cms. (4 i n s . ) . Below depths of 50 cms. (19.5 ins.) most of these minor f l u c t u a t i o n s are damped out completely, and the thermal regime shows a generally progressive warming trend, with the extremes of the tempera-ture record occurring at the beginning (coldest) and end (warmest) of the observation period. The pattern of isotherms i n the ground beneath the surface of the mud b o i l (Figure 16A) i s d i f f i c u l t to compare with those of the earth hummock, due to the previously-mentioned difference i n the depth spacing of the temperature sensors. Broad comparisons with the thermal regime i n Figure 15 G R O U N D T E M P E R A T U R E P A T T E R N S IN A N E A R T H H U M M O C K A N D A D J A C E N T D E P R E S S I O N ( ° C ) Depth in Cms. 911 117 the earth hummock, however, show that the pattern of dai l y fluctuations i s similar and extends to greater depths. Figure 16A also shows the deeper penetration of the 0°C isotherm beneath the surface of the mud b o i l , almost to a depth of one metre, compared with a maximum depth of 70-75 cms. (27.5-29.5 ins.) beneath the centre of the earth hummock. The decrease of the minor temperature fluctuations beneath one of the inter-hummock depressions i s shown i n Figure 15B. Although the near surface layer experienced as many cycles of alternate warming and cooling as the earth hummock, comparisons of the temperature maxima, during corresponding time periods, show that the near surface ground temperatures beneath the depressions were 5-7°C cooler than at similar depths beneath the centre of the earth hummock. As was noted i n the case of the ove r a l l seasonal amplitudes, the minor temperature fluctuations penetrated to much shallower depths beneath the depressions than at any of the other s i t e s . Very few of the warm cycles, for example, were f e l t below a depth of 25 cms. (10 ins.), and the maximum penetration of the 0°C isotherm was to a depth of 40 cms. (15.5 ins.) during the la s t week of August. Below 25 cms., the ground temperatures indicated a gradual, progressive warming with the extreme temperatures, shown i n Table V I I I , being recorded at the beginning and end of the observation period. The data obtained i n the ground temperature studies are i n accordance with the results of the depth of thaw measurements. Just as vegetation, including the microrelief factor, exerts a marked influence on the t o t a l depth of thaw to the frost table, so i t also influences the ground temperature patterns i n the active layer and uppermost parts of the underlying permafrost. Thus the t o t a l depth of thaw was greatest, and ground temperatures were highest, beneath unvegetated surfaces and 118 earth hummocks where the substrate was composed predominantly of mineral s o i l . Conversely, the depth of thaw was lowest and the ground tempera-tures were several degrees cooler beneath areas covered by mosses, and i n which the substrate was composed almost e n t i r e l y of organic material. In such l o c a l i t i e s , these lower values are probably related to the greater insulating q u a l i t i e s of organic material compared to mineral s o i l , the shade produced by microrelief factors, and the afore-mentioned effects of the evaporation process. Freeze-Back i n the Active Layer, 1964. The graphs of the temperature penetrations and isotherms at each of the si t e s show a temporary re-freezing of the surface layers i n response to below-freezing a i r temperatures at the end of August and during early September, 1965. Ground temperatures beneath the hummock centre were at or just below 0°C to depths of 20-25 cms. (8-10 ins.) for three days. The cooling was not as pronounced i n the inter-hummock depressions where the ground was frozen for one to two days to depths of 10-15 cms. (4-6 ins.) below the ground surface. This cooling of the ground from the surface downwards was replaced by a period of above-zero temperatures as the mean air temperature rose above the freezing point again. Unfortunately, i t was not possible to remain i n the f i e l d and record the pattern of the freeze-back at each of the s i t e s . A p a r t i a l record-of the freeze-back was obtained, however, for the mud b o i l s i t e for the three-month period September to December 1, 1964, using tempera-ture values recorded by an a r c t i c thermograph. The f u l l significance of the following discussion of the pattern of isotherms shown in Figure 16B . 119 i s limited by the absence of any data pertaining to ambient a i r temperat-ures on the island. To p a r t i a l l y offset t h i s problem, the temperature data for Tuktoyaktuk have been used as a guide. The mean daily a i r temperature at Tuktoyaktuk f e l l below freezing during the second week i n September. The start of the freeze-back on Garry Island probably began at t h i s time and, as Figure 16B shows, the 0°C isotherm had penetrated to a depth of 23 cms. (9 ins.) by Sept-ember 17. Figure 16B also indicates that a further lowering of the ground temperature did not take place u n t i l the beginning of the second week of October. This slowdown i n the rate of downward penetration of the cold may be attributed to a s l i g h t warming i n the a i r temperatures, which fluctuated around 0°C for most of the second half of September, but i t may also be related to the moisture conditions of the s o i l . Once the s o i l temperatures reach the freezing point, a further loss of heat may tempor-a r i l y be compensated by the release of the latent heat of fusion as any moisture i n the s o i l i s converted to i c e . This condition has been termed 12 the 'zero curtain'. The duration of the zero curtain condition i s de-pendent primarily upon the quantity of moisture i n the s o i l . Cook has also suggested that i t may be aided by the development of hydrostatic pressure i n the unfrozen material, between the downward penetrating f r o s t l i n e and the underlying permafrost table, r e s u l t i n g i n a lowering of the 13 freezing point of the s o i l . Recent investigations by Mackay, using s o i l pressure c e l l s , indicate, however, that the role of the hydrostatic Muller, S.W. (1947) op_. c i t . , p. 17. 13 Cook, F.W. (1955) "Near surface s o i l temperature measurements at Resolute Bay, Northwest T e r r i t o r i e s " , A r c t i c , Vol. 8, p. 245. 120 pressure factor i s limited by the low s o i l strength, and hence any de-14 pression of the freezing point i s quite small. Any depression of the freezing point, either by hydrostatic pressure or by the presence of minerals i n the s o i l water, complicates the i d e n t i f i c a t i o n of the zero curtain condition i n the temperature record. Since the tolerance of the thermistors was also T. 0.2°C at 0°C, and the ground temperature records showed a number of days which f e l l i n the range -0.2°C to +0.2°C, the i d e n t i f i c a t i o n of the zero curtain condition i n F i g -ure 16B i s only an approximation. With these considerations i n mind, the duration of the zero curtain at a depth of 23 cms. (9 ins.) beneath the surface of the mud b o i l was interpreted as being i n the order of 18 days (September 18 - October 6 ) . Meteorological records for Tuktoyaktuk indicate that the mean daily a i r temperature dropped permanently below freezing on October 9, and i n response to the lower a i r temperatures the downward penetration of the cold was resumed. The 0°C isotherm reached a depth of 53 cms. ( 2 1 ins.) beneath the surface of the mud b o i l on October 26, and the position of the permafrost table, at a depth of approximately 85 cms. (33.5 ins . ) , , was reached on November 6. These dates correspond to indices of -245 and -536 degree-days of freezing respectively following the drop i n a i r temperat-ures permanently below 0°C. Approximate calculations of the duration of the zero curtain effect show that i t lasted for about 36 days (September 20 - October 26) at a depth of 53 cms. ( 2 1 i n s . ) , and achieved i t s maximum duration, of as much as 48 days, at the permafrost table. This general pattern of an increasing length of the zero curtain time period with depth Mackay, J. Ross. Personal communication, October, 1968. 121 can be related to a slowdown i n the rate of freezing, which i s proportional to the square root of time, and also probably to the presence of greater quantities of s o i l moisture i n the lower parts of the active layer. The slow penetration of the f r o s t l i n e provides abundant time for the formation of segregated ice lenses encountered i n excavations made during the summer months. The ground was frozen s o l i d l y by the end of the f i r s t week i n November and, as Figure 16B shows, the penetration of the cold continued with only minor fluctuations through to December 1. Once temperatures of -2 to -3°C were, attained, the lag factor at depth was gradually diminished to the order of seven days throughout the complete p r o f i l e . The temperat-ure record was terminated on December 1st, and no further data were available concerning the continuation of the cooling process. The record of ground temperatures from the moss-filled de-pression surrounding the mud b o i l was far less complete due to a mal-functioning of the recorder. From the extremely limited ground temperature data available, i t appears that the f r o s t l i n e i n t h i s depress-ion had penetrated to a depth of 24 cms. (9.5 ins.) by about the beginning of October. A comparison with the ground temperatures recorded at a similar depth beneath the unvegetated surface of the mud b o i l , indicates that the f r o s t l i n e had penetrated to t h i s l e v e l almost two weeks prior to this date. The evidence suggests, therefore, that a surface layer of vegetation, whilst reducing the amount of thawing which takes place during the summer months, also retards the i n i t i a l penetration of the cold temp-eratures during the winter. 122 SUMMARY The thickness of the active layer on Garry Island i s greatest, i n excess of one metre (3 f e e t ) , i n areas having l i t t l e or no vegetation cover and where the substrate i s composed primarily of mineral s o i l . In the vegetated areas of the island, the maximum depths of thaw,v.60-70 cms. (24-28 i n s . ) , occurred beneath the raised centres of the 'Dry as-Hummocks'1 and the tussock-like forms of the Sheathed Cotton-grass (Eriophorum vagin-atum) . In such l o c a l i t i e s , the thickness of the active layer i s influenced considerably by the microrelief factor and, i n general, the larger the hummocks or tussocks the greater was the depth to the permafrost table. Uniformly lower depths of thaw were recorded beneath vegetation associa-tions with a substantial component of moss i n their f l o r a l composition, as i n the 'Sedge-Moss F l a t s ' and the inter-hummock or inter-tussock depressions, where the permafrost table was encountered at depths ranging from 18-30 cms. (7-12 ins.) below the ground surface. The lower depths of thaw are believed to be due to a combination of an organic substrate be-neath the moss, a higher moisture content and greater shade produced by the microrelief. These observations were supported by evidence from studies of the changing position of the f r o s t table beneath a small vegetation plot approximately 0.9 by 6.1 metres (3 by 20 feet) i n dimension. Comparisons of contour maps of t h i s plot reveal an inverse relationship between the configurations of the ground surface and the permafrost table r e f l e c t i n g , at least i n part, the significance of the microrelief factor. Elevated parts of the ground surface allow heat to penetrate l a t e r a l l y from the sides as w e l l as v e r t i c a l l y from the top. Excavations of the plot also 123 reveal an inverse relationship between the depth of thaw and the r e l a t i v e abundance of organic material i n the substrate. The greatest depths of thaw, i n excess of 75 cms. (29,5 i n s . ) , occurred beneath the raised, unveg-etated surface of a mud b o i l where the substrate was composed ent i r e l y of mineral s o i l . As the r a t i o of organic material to mineral s o i l i n the substrate increased, the depth of thaw decreased accordingly, and the ele-vated parts of the permafrost table were located beneath accumulations of organic material f l o o r i n g depressions i n the ground surface. The complete removal of the vegetation cover from two earth hummocks resulted i n an increase i n the depth of thaw of 135 per cent over'. that recorded i n control hummocks during a six-week period. Where the vegetation cover was only p a r t i a l l y removed, i n an attempt to distinguish between the l i v i n g and dead components of the organic layer, the corres-ponding increase i n the depth of thaw was 95 per cent. D i f f i c u l t i e s i n determining accurately the boundary between the l i v i n g and dead organic material l i m i t the conclusiveness of the data, but the figures indicate that a layer of peat, 4-5 cms. (1.5-2.C/ ins.) thick, was s u f f i c i e n t to produce a depth of thaw which was 17 per cent less than that recorded i n the hummocks from which the organic layer was removed completely. The type of vegetation cover, microrelief and composition of the substrate exert a similar influence on the thermal regime of the active layer and uppermost parts of the underlying.permafrost. Ground temper-ature, data obtained for four sites on the island - a mud b o i l , an earth hummock and two inter-hummock depressions - exhibit a similar pattern i n which the mean summer ground temperatures, and the amplitudes of the temperature fluctuations, decreased with increasing depth beneath the ground surface. The rate at which these decreases take place, however, i s 124 far from uniform. The most gradual rate occurred i n mineral s o i l beneath the. unvegetated surface of the mud b o i l , where the mean ground temperature for the summer did not reach 0°C u n t i l a depth of 75-80 cms. (29.5-31.5 i n s . ) , and the amplitude of the temperature fluctuations was s t i l l i n ex-cess of 3°C at a depth of 113 cms. (44.5 i n s . ) . In comparison, a similar mean temperature and magnitude of amplitude occurred at a depth of 60-65 cms. (23.5-25.5 ins.) beneath the vegetated surface of the earth hummock. The most rapid decrease of ground temperatures and diminishing of the temperature fluctuations took place i n the organic substrate beneath the moss-covered surfaces of the inter-hummock depressions, where a mean summer ground temperature of 0°C occurred at the shallow depth of 30-35 cms. (12-14 i n s . ) , and amplitudes i n excess of 3°C were not recorded below 25-30 cms. (10-12 ins.) from the ground surface. Ground temperature data for the freeze-back i n the mud b o i l for the winter of 1964 indicate that the zero curtain condition lasted approximately 18 days at a depth of 23 cms. (9 ins.) below the ground sur-face, 36 days at a depth of 53 cms. (21 i n s . ) , and achieved a. maximum duration of as much as 48 days at the permafrost table. This general pattern of an increasing length of the zero curtain time period with depth can be related to a retardation of the rate of freezing and also probably to the presence of greater quantities of s o i l moisture i n the lower parts of the active layer. The slow penetration of the f r o s t l i n e also provides abundant time for the formation of the segregated ice lenses encountered i n excavations made during the summer months. CHAPTER V GEOMORPHOLOGICAL PROCESSES The purpose of th i s chapter, and one of the major aims of th i s thesis, i s to assess some of the contemporary geomorphic processes oper-ating on Garry Island. The sp e c i f i c processes considered involve problems associated with coastal recession, mudslumps, mudflows, and the genesis of certain types of patterned ground. In each case, wherever applicable, an emphasis i s placed on quantitative measurements of the ratesiof operation of these processes, and an evaluation of the various factors, especially the role of permafrost, influencing these rates. COASTAL RECESSION Rapid recession of many sections of the coastline between Point Barrow, Alaska and Langton Bay, N.W.T. , i n post- g l a c i a l and h i s t o r i c times, has been described by several authors c i t i n g both geomorphological and h i s t o r i c evidence. Leffingwell, using maps of the region drawn by Franklin i n 1826, reported erosion rates of as high as 30.5 metres (100 feet) a year at Cape Simpson, northern Alaska, but concluded that the average retreat was less than 1.2 metres (4 feet) per annum.''" MacCarthy measured rates of retreat at Point Barrow ranging from 0.0-4.5 metres Leffingwell, E de K. (1919) "The Canning River region, northern Alaska", U.S.G.S. Professional Paper, No. 109, pp. 169-171. 126 (0.0-14.7 feet) a year. Mackay has described both geomorphological and h i s t o r i c a l evidence indicating s i m i l a r l y rapid rates of recession along 3 sections of the coastline i n the Yukon and Northwest T e r r i t o r i e s . Despite the abundant evidence of coastal retreat, many of the rates quoted are at best approximations only. In the h i s t o r i c evidence there are prob-lems r e l a t i n g to the accuracy of the early maps. Unfortunately, many of the distances from the coastline to i d e n t i f i a b l e control points were est-imated, thus precluding the calculation of accurate rates of coastal recession. S i m i l a r l y , the lack of a detailed chronological scale for the area l i m i t s the v a l i d i t y of rates using geomorphic evidence based on the t o t a l recession which has taken place i n a p o s t - g l a c i a l period of, as yet, unknown duration. Accordingly, a programme of f i e l d studies was conducted on Garry Island for the dual purposes of: (1) providing exact data on the annual rates of recession along coastlines of varying lithology; and (2) investigating the nature and r e l a t i v e importance of the s p e c i f i c processes which contribute to the observed retreat values. Evidence of Coastal Recession. Active recession of the Garry Island coastline, apart from a few l o c a l mudslump features, currently i s r e s t r i c t e d primarily to the ex-posed west and northwest coasts, and to segments of the prominent sand 2 MacCarthy, G.R. (1953) "Recent changes i n the shoreline near Point Barrow, Alaska", A r c t i c , Vol. 6, pp. 44-51. 3 Mackay, J. Ross (1958) "The Anderson River Map Area, N.W.T.", Geographical Branch Memoir, No. 5, pp. 39-40. Mackay, J. Ross (1963) "Notes on the shoreline recession along the coast of the Yukon Te r r i t o r y " , A r c t i c , Vol. 16, pp. 195-197. 127 headlands on the north side of the island. In these l o c a l i t i e s , the b l u f f s are characterized by numerous fresh exposures with debris p i l e s on, and at the base of, the c l i f f face, and by an absence of vegetation on the faces. H i s t o r i c a l evidence of coastal recession on Garry Island i s available but does not y i e l d satisfactory quantitative data. In a recent survey, Captain Ages, of the C.S.S. 'Richardson 1, refers to a number of hydrographic control points, located i n 1930, i n the outer Mackenzie A Delta. Of twelve control points shown on P e l l y , Kendall and Whale (Grassy ?) Islands i n the o r i g i n a l survey, Captain Ages concluded that four, a l l located on P e l l y Island, had disappeared possibly due to erosion; i n s i x cases, only remnants of the markers remained with the collapse frequently attributable to caving i n of the ground; and only two of the markers were s t i l l i n t a c t . The p o s s i b i l i t y exists that one of the markers on Kendall Island had been moved by natives. The 1930 survey also shows the locations of three hydrographic control points on Garry Island. Ground checks made i n 1965 revealed no trace of one of these markers; a second (possibly dismantled), was found on the f l o o r of an inactive mudslump, but the t h i r d marker, located on the southeast t i p of the i s l a n d , was found collapsed on the c l i f f face (Plate VI-A);. Unfortunately, no figures are available indicating the distance of these control points from the o r i g -i n a l c l i f f edge, thereby preventing any r e l i a b l e estimate of the rate of coastal recession i n recent h i s t o r i c time. Additional evidence of coastal recession i n h i s t o r i c times can Ages, Captain A.B. Personal communication to Dr. J. Ross Mackay, July 4, 1965. The exact date of the hydrographic survey i n 1930 i s not mentioned. Plate VI C O A S T A L R E C E S S I O N A. Collapsed hydrographic marker on the c l i f f face along the south coast of Garry Island. B. Coastal recession along the northwest coast of Garry Island showing a truncated lagoon. C. Measurement of coastal recession rates. Stakes along the northwest coast of Garry Island. 129 possibly be inferred from the 'disappearance' of islands i n the outer delta area. S i r Alexander Mackenzie's account of the panoramic v i s t a from Whale (Garry) Island includes a reference to two small islands i n the ice lyi n g to the northwest by compass dir e c t i o n . ^ Even allowing for discrep-ancies i n Mackenzie's d i r e c t i o n a l observations, there are no islands i n that position today and i t i s possible that they have been removed subse-quently by wave action. A l t e r n a t i v e l y , as Mackay has suggested, Mack-enzie may have observed two patches of d i r t y ice covered with debris derived from the Mackenzie River break-up. 7 Albert Ol i v e r , a native of the area and our guide during the f i e l d seasons, also t e l l s of the former existence of a small i s l a n d , to the south of Hooper Island, which has been destroyed by wave action during his l i f e t i m e . Topographic and stratigraphic evidence of coastal recession over even longer periods of time yie l d s s i m i l a r l y inconclusive results with respect to actual rates of c l i f f retreat. Features developed along the northwest coast of the island include an example of a truncated lagoon, with a straight coastal shore barrier and actively receding c l i f f s on either side (Figure 17 and Plate VI-B) . Raised, shoreline, features can also be traced to the coast i n many places, but no evidence of them can be found Mackenzie, A. (1801) Voyages from Montreal on the River St. Laurence through the continent of North America to the Frozen and P a c i f i c  Oceans jn the years 1789 and 1793, T. Cadell, Jun. and W. Davies, Strand; Cobbett and Morgan, P a l l M a l l ; and W. Creech at Edinburgh. Reprinted edition (1966) i n March of America Facsimile Series, Number 52, University Microfilms Inc., Ann Arbor, Michigan, p. 63. Stager, J.K. (1965) "Alexander Mackenzie's exploration of the Grand River", Geographical B u l l e t i n , Vol. 7, p..,231. 7Mackay, J. Ross (1963) "The Mackenzie Delta area, N.W.T.", Geographical Branch Memoir, No. 8, p. 6. Figure 17 C O A S T A L R E C E S S I O N F E A T U R E S N O R T H W E S T C O A S T OF GARRY ISLAND T r u n c a t e d l a g Tundra p o l y g o n s on c l i f f top S t a k e #95 H i g h c e n t r e d p o l y g o n s S t a k e #31 T u n d r a p o l y g o n s on c l i f f top S t a k e #1 x * " ^ Mudslumps P o l y g o n a l G r o u n d S c a l e 1:17,000 (App rox ima te ! 131 i n stratigraphic sections exposed i n the c l i f f s . The longitudinal p r o f i l e s of stream courses draining towards the south coast of the island exhibit a char a c t e r i s t i c steepening of the gradient i n the lower parts, which may also r e f l e c t the effects of downcutting i n response to coastal recession combined with mudslump development. Exposures along the northwest coast of the island reveal a sequence of lacustrine and peat deposits (see Figure 6), and the tops of the high b l u f f s i n places along this same coastline are capped by a truncated series of tundra polygons (Figure 17). The lac-ustrine and peat deposits, exposed i n these sections, are ind i c a t i v e of freshwater conditions which could only have been produced as a resul t of ponding behind a ba r r i e r , since removed, on the seaward margins of the topographic depressions i n which they occur. Thus there i s abundant evidence that the coastline of Garry Island has undergone considerable recession i n recent geological and h i s t -o r i c a l time. Just how much recession has taken place, and at what rate, i s impossible to determine, but there may be some significance i n the fact that the depth of water for 16-24 kilometres (10-15 miles) off the shore of Garry Island averages only 4-5 fathoms. Such uniformly shallow depths may represent an extensive platform of marine planation, though the poss-i b i l i t y that i t i s p a r t i a l l y a product of delta i c sedimentation cannot be excluded. Coastal Stake Measurements. Approximately 2.5 kilometres (1.5 miles) of coastline were staked during the summer of 1964 to provide quantitative data on the current rates of coastal recession (Plate VI-C). . A t o t a l of 95 stakes was i n s t a l l e d along the exposed northwest coast where the b l u f f s l o c a l l y 132 exceed 30.5 metres (100 feet) i n height (Figure 17). The positions of these stakes were checked pe r i o d i c a l l y throughout each of the summers, and the amount of recession was recorded i n metres at the beginning and end of each summer. Since the exposures i n these b l u f f s covered a wide range of materials, ranging from fine-grained s i l t s and clays, which i n places con-tained large bodies of segregated ground i c e , to coarse sands and boulders, the data permitted a ready evaluation of the effect of lithology on the rate of recession. Comparisons of the amount of retreat recorded at the beginning of the summer with that recorded at the end of the previous summer also gave indices of variations i n the rate of recession through the year. Average retreat figures tabulated i n August, 1966 showed that the average rate of recession for the 95 stakes was 6.4 metres (20.8 f e e t ) , or 2.1 metres (6.9 feet) per annum. As expected, however, th i s average covered a wide range of ind i v i d u a l values which ranged from a maximum of 30.6 metres (100.4 feet) to a minimum of 0.1 metres (0.3 f e e t ) . Thirty-one of the stakes were located along a series of high b l u f f s ranging from 9.0-30.5 metres (30-100 feet) i n elevation. Sediments exposed i n t h i s section of the coastline were fi n e sands, s i l t s or clays, and bodies of segregated ground ice were v i s i b l e where the c l i f f s cut across the d i s t i n c t i v e scars of old mudslumps. In addition, the tops of the c l i f f s also cut across a series of well-developed, high-centred tundra polygons and associated network of ice-wedges. The highest retreat values were recorded where the c l i f f s intersected and reactivated a number of old mudslumps, and i n these l o c a l i t i e s the average recession amounted to 20.9 metres (68.6 f e e t ) , or 7 metres (23 feet) a year. (Plate VT.I.-A-) . ...The effects of segregated ground ice were shown by the variable rates of recession 1 3 3 Plate VII C O A S T A L R E C E S S I O N A. C o a s t a l r e c e s s i o n i s most r a p i d where wave a c t i o n has exposed b o d i e s of s e g r e g a t e d ground i c e i n t h e c l i f f s . B. The detachment of a h i g h -c e n t r e d t u n d r a p o l y g o n f r o m the c l i f f t o p as a r e s u l t of m e l t i n g a l o n g t h e l i n e s of the ice-wedges. C. M i n e r a l s o i l p i n n a c l e l e f t by the c o l l a p s e of a t u n d r a p o l y g o n . 134 recorded by f i v e stakes (#21-25), located, across one of these slumps. During the summer of 1964, no ground ice was exposed at these points and the average retreat during the summer was only 3.5 metres (11.5 f e e t ) . Towards the end of that summer, however, a large block was detached from the c l i f f face, and ground ice was v i s i b l e throughout the summers of 1965 and 1966, when the average rate of retreat increased to 10.5 metres (34.4 feet) per summer. High rates of recession were also observed i n the areas of tundra polygons, where melting along the lines of the ice-wedges r e s u l t -ed i n an average recession of the c l i f f face of 15.3 metres (50 f e e t ) , or s l i g h t l y more than 5 metres (16.4 feet) a year. The intervening polygonal peat units showed a much more variable rate of recession, averaging only 2.5 metres (8.2 feet) a year. Since the peat i n these polygons forms a coherent u n i t , which i s undermined by thawing of the i c e - r i c h basal layers, the rates of recession show a wide range from summer to summer. In any one summer, large overhangs, i n the order of several metres, may be pro-duced with l i t t l e or no material actually being detached from the c l i f f edge. By way of contrast, the following summer may be characterized by continued undermining and eventual collapse of the complete polygon, and a sudden retreat of the c l i f f edge by 7-10 metres (23-33 feet) (Plates VII.-B and VllfeC). Sediment types exposed i n the remaining section of the coast-l i n e were much more uniform, and consisted of coarse sands and gravels with minor variations caused by the development of polygonal ground. The average t o t a l rate of retreat along t h i s part of the coast amounted to only 2.6 metres (8.6 feet) or less than one metre (3 feet) per annum. The figure i s even less, .0.6 metres (2 feet) a year, i f a small slump (average 2.7 metres or 8.9 f e e t ) , ice-wedges (average 1.2 metres or 4 f e e t ) , and 135 peat sections (average 0.7 metres or 2.3 feet) are excluded from the c a l -culations. In addition to quantitative values of actual rates of recess-ion, the data also indicate that most of the retreat occurs during the short summer period. For most of the year, sea ice i s packed close to the shore, and the bl u f f s are buried by large snow d r i f t s which persist through late June and early July. The measurements show that f u l l y 60-70 per cent of the observed retreat occurred during the summer months. Unfortunately, i t was not possible to remain i n the f i e l d u n t i l freeze-up occurred, when this percentage would be even higher. Much of the retreat recorded between the l a s t observation of one summer and the f i r s t observation of the follow-ing summer probably represents further detachment of material from the c l i f f edge during the month of September. Sand Headland P r o f i l e Studies. Two prominent sand headlands on the north side of the island were selected as suitable sit e s for detailed studies of c l i f f p r o f i l e changes throughout the summer months. These headlands terminate i n abrupt c l i f f faces, 7-12 metres (23-39 feet) high, many sections of which are currently undergoing active recession as i s indicated by the numerous fresh exposures and lack of vegetation on the c l i f f face. Stable sections of the coastline, mantled to some degree with a vegetation cover, can be attributed largely to the development of a protective sandspit formation at the base of the c l i f f s (Plate VIIL-A). The c l i f f s have a f a i r l y straight, uniform appearance, and are generally devoid of gullying except for the presence of deep 'V'-shaped notches created by the melting out of ice-wedges. The material i n these 136 Plate VIII S A N D H E A D L A N D P R O F I L E S T U D I E S A. S t a b i l i z e d c l i f f s of the sand headland areas protected by sandspits. C. Crevices on the sand headland surface p r i o r to the detachment of hummock blocks from the edge of the c l i f f . B. P r o f i l e IV, J u l y 1965, showing the l o c a t i o n s of the stakes. D. P r o f i l e I, August 1965, a f t e r slumping of the thawed sand had taken place. 137 headlands i s composed predominantly of medium to coarse sands, with i n t e r -calated bands of pebbles, gravel, wood and s h e l l fragments. Boulders are generally lacking, but l o c a l l y they may occur i n s u f f i c i e n t quantities to dominate the whole exposure. Four locations, currently undergoing active recession, were chosen for the p r o f i l e studies. Each of these p r o f i l e s was surveyed at the beginning and end of the 1965 f i e l d season, and P r o f i l e IV was sur-veyed again during the summer of 1966. The results of these surveys are shown i n Figures 18 and 19. To obtain data on c l i f f p r o f i l e changes over 1-2 weeks, a procedure was adopted similar to that used by Twidale i n his Q study of r i v e r bank erosion i n South A u s t r a l i a . At each s i t e , wooden dowelling, approximately 45 cms. (18 ins.) i n length and 2.5 cms. (one inch) i n diameter, was driven into the c l i f f face, normal to the surface, and the stakes were spaced at approximately equal intervals with only a small portion of each l e f t exposed (Plate VIII'-B). The positions of these stakes on the four p r o f i l e s are also shown i n Figures 18 and 19. In each p r o f i l e , stake 1 was i n s t a l l e d on the headland surface to act as a control point, and the distance to the edge of the c l i f f was recorded. The remain-ing stakes on each p r o f i l e were numbered consecutively from top to bottom of the c l i f f . Reference to i n d i v i d u a l stakes i n the text also follows the system used by Twidale: e.g. stake (1,4) refers to P r o f i l e I , stake 4. The v i s i b l e length of the stake was measured and recorded, and at subsequent observations the procedure was repeated so that an increase, or decrease, i n the length of the stake exposed indicated whether erosion, °Twidale, C.R. (1964) "Erosion of an a l l u v i a l bank at Birdwood, South A u s t r a l i a " , Z e i t s c h r i f t fur Geomorphologie, Band 8, pp. 189-211. 138 Figure 18 S A N D H E A D L A N D P R O F I L E S (I) Profiles on August 31,1965 1 Positions of stakes SAND Figure 19 HEADLAND PROFILES (2) 139 Profiles on July 5, 1965 Profiles on August 31,1965 1 ! Positions of stakes — S.L.— Sea Level SCALE 0 5 METRES 140 or deposition, was taking place. These changes, measured over two-weekly periods, were too small to i l l u s t r a t e diagrammatically, and are summarized in Table IX. At the same time as these readings were being made, a metal probe was used to determine the approximate thickness of the thawed layer on the c l i f f face. When the p r o f i l e s were f i r s t surveyed i n early J u l y , 1965, l a t e - l y i n g snow d r i f t s s t i l l remained at the foot of the c l i f f s and, i n ' 9 P r o f i l e I , extended as high as stake (1,8). The absence of any marked concentrations of debris on the snow surface indicated that l i t t l e or no material had been dislodged from the c l i f f face prior to the i n s t a l l a t i o n of the stakes. Probing revealed that the thawed zone on the c l i f f face extended to depths of 15-20 cms. (6-8 i n s . ) , producing an unstable layer of loosely-packed sand at the surface. The general forms of the p r o f i l e s are shown i n Figures 18 and 19. The uppermost morphological facet i n each p r o f i l e consists of a small v e r t i c a l element, usually 0.5^1.0 metres (1.5-3.5 feet) high, and accom-panied by a short overhang r e f l e c t i n g the binding effect of the vegetation on the headland surface. Small earth hummocks dominate t h i s vegetation association, and i n many places lunate tension cracks p a r a l l e l the c l i f f edge on the landward sides of these hummocks (Plate VHI-C) . Below t h i s facet, the p r o f i l e consists es s e n t i a l l y of a f a i r l y uniformly sloping element, of 40-50° i n c l i n a t i o n , which i n many cases extends to the base of the p r o f i l e where i t terminates at the beach l e v e l . Ih other places, however, the junction with the beach may be more or less obscured by f a l l e n The other, lower stakes on t h i s p r o f i l e were i n s t a l l e d pro-gressively as the snowbank melted. 141 TABLE IX SAND HEADLAND PROFILE CHANGES S t a k e # P r o f i l e I P r o f i l e I I P r o f i l e I I I P r o f i l e IV . 1 0 0 0 0 ( P ) 2 -1 -4 -11 0 ( o ) 3 -8.5 -2 -11 - 1.5 (-29.5) 4 -9 -3 -10 - 7 (-16.0) 5 -7 -0.5 - 1 -11 (-22.0) 6 -4.5 -0.5 -10 - 4. (- 9.0) 7 -5 +1.5 - 9 - .0.5 (- 3.5) 8 -2.5 +1.5 + 2.5 - 2 (+ 3.0) 9 0 +0.5 + 3 + 6 (+23,0^  10 + 1 +9 + 6 + 7 (+20.5) 11 +2.5 +7 + 6 + 6 (+17.0) 12 +5 + 2 + 3 (+12. 0} 13 +6.5 0 + 3 (+10.0) 14 +4.5 0 + 6 (+14.0) 15 + 6 16 +4 17 0 The f i g u r e s a r e i n c e n t i m e t r e s and r e f e r t o t h e p e r i o d J u l y 5 A u g u s t 22, 1965, when t h e l a s t measurements were r e c o r d e d . The f i g u r e s i n b r a c k e t s f o r P r o f i l e IV i n d i c a t e t h e l e n g t h o f s t a k e exposed on Au g u s t 22, 1966. N e g a t i v e v a l u e s r e f e r t o e r o s i o n of t h e c l i f f f a c e , w h i l e p o s i t i v e f i g u r e s r e p r e s e n t t h e a c c u m u l a t i o n o f d e b r i s d e r i v e d f r o m upper s e c t i o n s of t h e p r o f i l e s . 142 debris composed of sand, detached hummocks which show d i s t i n c t signs of washing by wave action, and driftwood. Occasionally, some of the hummocks come to rest on the c l i f f face ;before reaching the beach l e v e l , and these are responsible for. most of the minor i r r e g u l a r i t i e s shown on the p r o f i l e s . Comparisons of Figures 18 and 19 and Table IX show that the changes, recorded by the stake measurements account for only a minor portion of the t o t a l p r o f i l e changes through the entire summer. For the period July 5 - August 22, 1965, of the 56 stakes i n s t a l l e d on the four p r o f i l e s , a t o t a l of 23 experienced an increase i n the length of stake exposed, indicating, erosion of the upper sections of the p r o f i l e s . The average rate of erosion recorded by these stakes amounted to 5.4 cms. (2.2 i n s , ) , with maxima of 11.0. cms. (4.3 i n s . ) . A further 24 stakes i n -dicated that deposition had taken place on the lower sections of the . p r o f i l e s , averaging 4.2 cms. (1.7 ins.) and reaching maxima of 6-7 cms. (2.4^2.8 i n s . ) . By way of contrast, measurements.taken from the surveyed diagrams of P r o f i l e s I - I I I show that the upper portions of the c l i f f faces:experienced a retreat of 75-100 cms. (29.5-39.5 ins.) over the s l i g h t l y longer period ending, August 31, 1965. The losses, or erosion, recorded by the stake.measurements therefore account for only 5-6 per cent of the t o t a l p r o f i l e changes, despite the fact that they cover approx-imately 85 per cent of the observation period. Similar percentages are also applicable to the rates of deposition recorded on the lower parts of the p r o f i l e s . Coincident with these.changes indicated by the.stake measure-ments, the depth of the thawed zone on the c l i f f face increased to 50-65 cms. (19.5-25,5 ins.) by August 1, and to 90-100 cms. (35.5-39.5 ins.) by 143 August 2 2 . Beginning i n late July and early August, prominent cracks, extending for several metres across the face of the c l i f f and to depths of 1 0 - 2 0 cms. (4-8 i n s . ) , appeared i n many places i n the upper sections of the p r o f i l e s . Locally, as i n P r o f i l e I , the appearance of these cracks was followed by minor slumping action. New cracks also opened up along the boundaries of the hummocks on the headland surface, and pre-existing cracks were widened and deepened as the hummocks t i l t e d bodily towards the c l i f f edge. In direct contrast to th i s long period of r e l a t i v e s t a b i l i t y , during which the p r o f i l e s underwent slow changes, the events observed i n P r o f i l e s I - I I I within the la s t week of August can only be described as catastrophic. By August 31, only the three stakes located on the headland surface as control points remained i n t a c t , and the remainder had been com-pletely obliterated by mass slumping of the thawed sand from the upper parts of the c l i f f faces (Plate VIII-D). In P r o f i l e s I and I I I , the earth hummock at the top of the p r o f i l e was also detached from the c l i f f edge. As Figure 19 and Table IX show however, slumping did not take place on P r o f i l e IV during that summer, nor the following summer of 1966, and most of the changes observed i n t h i s p r o f i l e are indicated by the additional figures i n Table IX. Towards the end of August, 1966, however, the hummock was also detached from the top of th i s p r o f i l e , and a number of stakes was either broken or t i l t e d during i t s f a l l . No attempt was made to record further changes in P r o f i l e s I - I I I following the slumping, but two stakes were relocated i n the upper part of P r o f i l e I to see whether slumping occurred again during the summer. These stakes were s t i l l i n place at the end of August, 1966. 144 Relevance of the P r o f i l e Studies to other Coastal Areas. The sand headlands were chosen for the p r o f i l e studies because the slow rate of change permitted ample time to study the processes involved. I t was evident, however, that similar changes and processes to those observed on the sand headlands also apply to the tracts of staked coastline on the northwest shore of the island. As noted previously, many of the c l i f f s along t h i s section of the coast are also composed of sands and gravels which l o c a l l y are capped by varve-like lacustrine sediments and i n t e r -sected by a network of ground ice-wedges. The retreat figures recorded i n these l o c a l i t i e s correspond we l l with those made on the sand headlands, and the general appearance of the c l i f f p r o f i l e s indicates that the method of c l i f f retreat i s the same. In the finer-grained sediments the processes are probably the same but d i f f e r i n degree. This difference can readily be attributed to the higher ice content of these sediments. The larger quantities of water released on thawing of t h i s material, together with the fact that the c l i f f faces are usually much steeper, and frequently v e r t i c a l , preclude the accumulation of any appreciable thickness of thawed material at the sur-face. Frequent probing of the c l i f f s showed that the thickness of t h i s thawed layer seldom exceeded 10-15 cms. (4-6 i n s . ) . The ice i n these sed-iments i s also predominantly pore i c e , l o c a l l y augmented by vein ice along bedding and shear planes, but moisture contents are commonly i n the order of 100 per cent by weight, and thawing produces a r e l a t i v e l y mobile mud which soon slides to the base of the c l i f f and provides a fresh exposure. Whereas slumping i n the sands may have a maximum frequency of only once during an entire summer, i t may occur d a i l y , or more often, i n these finer-grained sediments. The rate i s even higher i n the c l i f f s with 145 exposures of segregated ground ice which have moisture contents i n the order of several thousand per cent by weight. Thawing of the c l i f f s i n these l o c a l i t i e s produces large quantities of excess water, and the i n t e r -v a l between thawing and slumping of the material to the foot of the c l i f f reaches a minimum. Coastal recession i n areas where the c l i f f s intersect w e l l -developed networks of tundra polygons produces a>distinctive topography. In such l o c a l i t i e s , as was the case i n the sand headlands, recession takes place most rapidly along the lines of the ice-wedges. Where the polygons are of the high-centred type and are located on the tops of the b l u f f s , rapid melting also occurs i n the basal sections of the polygonal units which also contain large quantities of segregated ice,.-. Melting out of the ice-wedges, with an orientation normal to the c l i f f edge, exposes addition-a l wedges which l i e p a r a l l e l to the c l i f f , and the subsequent melting of these leads to an i s o l a t i o n of the polygonal unit from the main face of the b l u f f . In t h i s way a peat block i s formed which rests precariously on a pedestal of mineral s o i l . Further melting of the high ice-content basal sections of the polygon results i n further undermining of the peat block which f i n a l l y tumbles, usually i n t a c t , to the foot of the c l i f f , leaving behind d i s t i n c t i v e pinnacles of mineral s o i l (Plate VII-C). Under favour-able conditions, a complete tundra polygon, up to 10 metres (33 feet) i n diameter, may be removed i n t h i s manner during the course of a single summer. Processes at work i n Coastal Recession. From the preceding paragraphs, i t i s quite evident that the coastline of Garry isl a n d i s retreating primarily as a res u l t of 'thermal 146 erosion', or thawing of the frozen ground, accompanied by slumping of the thawed material. Coastal sections composed of sediments with a. high .. moisture content, p a r t i c u l a r l y i n the. form of ice segregations, are natur-a l l y the most susceptible to the sun's rays, and th i s i s readily substan-t i a t e d by the observed retreat figures. The material i n the sand headlands has a frozen moisture con-tent of only 15-20 per cent by weight, and i t was i n b l u f f s composed of these and coarser sediments that the lowest retreat values were recorded. In their frozen state, the water i s present i n the form of pore ice which acts as a strong cementing agent, but melting of t h i s ice r e s u l t s i n a considerable loss of strength and transforms the sand into a loosely-packed mass. As the summer progresses, the fr o s t table retreats further into the c l i f f face and, as the thickness of the thawed zone increases, the surface layer becomes increasingly unstable due to the fact that the i n c l i n a t i o n of the c l i f f face, 40-50 degrees, i s considerably steeper than the angle of repose of the sand grains. F r i c t i o n a l resistance between the sand grains may keep most of the thawed layer i n t a c t , but the opening of surface cracks and the changing attitude of the stakes i n s t a l l e d on the surface r e f l e c t l o c a l i z e d movements within t h i s layer and temporary stress r e l i e f . This process continues u n t i l the weight of the unfrozen sand reaches a c r i t i c a l l e v e l capable of overcoming the i n t e r n a l f r i c t i o n a l cohesion, and f a i l u r e occurs. Generally t h i s f a i l u r e takes place towards the end of each summer, when the thawed zone i s approximately one metre (3 feet) thick, although i t may occur l o c a l l y before t h i s thickness i s reached. The slumping usually takes place along f a i r l y well-defined ... planes at, or close to, the position of the f r o s t table. Minor slumps are primarily of the r o t a t i o n a l type, but the chaotic mixture of debris at 147 the foot of the larger slumps indicates considerable overturning and dis-integration of the sand masses during their descent. The above comments are mainly applicable to coarse-grained sediments with low ice contents, the thawing of which releases ne g l i g i b l e quantities of water. As the ice content increases however, the volume of water released upon melting becomes much more s i g n i f i c a n t . This i s par-t i c u l a r l y noticeable i n supersaturated sediments where thawing produces large quantities of excess water. In such sediments, the water released imparts an added mobility to the thawed layer and greatly f a c i l i t a t e s i t s removal from the c l i f f face. Consequently, the frequency of removal of the unfrozen material increases from a maximum of once or twice during the whole summer, as i n the case of the sand headlands, to an almost daily or even hourly occurrence. At the same time, as the volume of water increas-es, the mass movement, removal process gradually acquires the character-i s t i c s of a mudflow rather than a simple slump. On a short term basis, the major role of wave action appears to be i n the removal of slumped material from the base of the c l i f f s . Such removal i s necessary to prevent the accumulation of debris and b u r i a l of the lower sections of the p r o f i l e , and thus permit the maintenance of fresh exposures for the 'thermal erosion' process. In places where the beach and foreshore are narrow, as along the b l u f f s on the northwest coast of the isl a n d , or where mudflows carry material from coastal exposures of ground ice d i r e c t l y to the sea, removal by wave action may occur almost instantaneously despite the limited t i d a l range. Where the c l i f f s are more distant from the shoreline, the major part of the removal process i s accomplished during storm surges, especially i n the late summer and early fal l . 148 Although the chief role of wave action on a short term basis i s i n the removal of slumped debris from the base of the c l i f f s , direct undercutting by waves may l o c a l l y be of major importance during these same storm surges. The two lowermost stakes of P r o f i l e I I were removed by waves during a storm i n early August of 1965. Wave-cut notches are a common feature along the high b l u f f s along the northwest coast, and the most spec-tacular example of undercutting was observed i n t h i s area i n the summer.of 1964. In late July of that year, a deep c l e f t , 50 metres (165 feet) long and up to 7 metres (23 feet) deep, opened up i n the s i t e of an old mudslump at a distance of 5-7 metres (16-23 feet) from the edge of the c l i f f . The b l u f f s at t h i s point were 15-20 metres (50-65 feet) high, and had been undercut, at beach l e v e l for distances of 4-5 metres (13-16 f e e t ) . As the summer progressed, the c l e f t gradually became wider and deeper as the whole block t i l t e d bodily seaward u n t i l i t f i n a l l y collapsed into the sea. A l -though the formation of t h i s c l e f t exposed a large body of ground i c e , which undoubtedly f a c i l i t a t e d the collapse, i t was apparent that direct undercutting by the waves was primarily responsible for the removal or erosion of approximately 5,250 cubic metres (185,000 cubic feet) of coast-l i n e i n one single block. On a long term basis therefore, i n some areas, undercutting by wave action may be the most important process involved i n c l i f f r e t r e a t , p a r t i c u l a r l y during big storms when as much recession may take place as i n several years of "normal" erosion. Erosion due to wind action, through deflation of the f i n e r mat-e r i a l and mechanical dislodgement of i n d i v i d u a l p a r t i c l e s , i s of minor importance, but probably accounts for most of the changes recorded by the stakes on the sand headlands before slumping occurred. I n d i r e c t l y , wind action may contribute to coastal recession i n another way. Large tracts of 149 vegetation along the northwest coast have been k i l l e d off by s a l t spray, and the reduced binding effect would as s i s t i n the detachment of earth hummocks from the c l i f f edge i n these areas. Erosion by running water i s of l i t t l e or no consequence i n the coastal retreat process. The only minor exception was i n the v i c i n i t y of the notches i n the sand headlands, where further melting of the ice-wedges and snow accumulations temporarily provides s u f f i c i e n t quantities of water for the transportation of material, and the construction of miniature a l l u v i a l fans on the beach. MUDSLUMPS Many sections of the Garry Island coastline exhibit a d i s t i n c t -ive, scalloped appearance on a e r i a l photographs, r e f l e c t i n g the occurrence of numerous, large, crescent-shaped depressions. Their presence i s a r e l i a b l e indicator of the existence of massive bodies of segregated ground ic e , and they are i n fact large thermokarst features r e s u l t i n g from the exposure and melting of t h i s i c e . Similar features have been described i n other parts of the Canadian A r c t i c , where emphasis has been placed on the role of mudflows found i n the f l o o r s of the depressions and, i n some cases, the term mudflow (coulee de boue) was the only name applied to these 150 landforms.^ Mudflows are also found i n association with the amphitheatres on Garry Island, and their significance i n the c y c l i c development of the landforms w i l l be discussed below. Mackay, i n his studies i n the Mackenzie Delta area, has c l a s s i f i e d these features as slumps or ground ice slumps.^ l Since the overwhelming ch a r a c t e r i s t i c of active features of t h i s type i s the ubiquitous presence of a s u r f i c i a l layer of mud debris derived from the melting i c e , the term mudslump i s adopted here. A map showing the d i s t r i b u t i o n of mudslumps on Garry Island (Figure 20) suggests the most probable mode of o r i g i n of these features. The mudslumps are almost exclusively confined to coastal locations where the bodies of ground ice have been exposed i n the b l u f f s . The ice may be uncovered d i r e c t l y by wave action or i n d i r e c t l y as the resu l t of over-steepening of the c l i f f p r o f i l e , and subsequent mass movement of the vegetation layer from the c l i f f top. At one location on the south coast of the i s l a n d , such oversteepening had led to the downslope motion of a mass of vegetation extending as much as 25-30 metres (80-100 feet) along, and back from, the edge of the c l i f f (Plate IX-A). The layer remained f a i r l y •*-^ Washburn, A.L. (1947) "Reconnaissance Geology of portions of V i c t o r i a Island and adjacent regions i n A r c t i c Canada", Geol. Soc. America, Memoir 22, p. 142. Lamothe, C. and St-Onge, D. (1961) "A note on a p e r i g l a c i a l erosional process i n the Isachsen area, N.W.T.", Geographical B u l l e t i n , No. 16, pp. 104-113. St-Onge, D. (1965) "La geomorphologie de L ' l l e E l l e f Ringnes, T e r r i t o i r e s du Nord-Ouest, Canada", Geographical Branch-Paper, No, 38, p. 46. 1 : lMackay ? J . Ross (1963) "The Mackenzie Delta area, N.W.T.", Geographical Branch Memoir, No. 8, pp. 60-65. Mackay, J. Ross (1966) "Segregated Epigenetic Ice and Slumps in Permafrost, Mackenzie Delta area, N.W.T,", Geographical B u l l e t i n , No. 8, pp. 59-80. Figure 20 Plate IX M U D S L U M P S A. S l i d i n g of the vegetation layer produced by over steepen-ing of the c o a s t a l b l u f f s . C . Overburden of a f i r s t -generation mudslump showing the high i c e content. B. General view of Slump A with a prominent mudflow extending into the sea. D. Overburden, composed pre-dominantly of mineral s o i l , of a second- or late r - g e n e r a t i o n mudslump. 153 intact during i t s descent, and bold s t r i a t i o n s marked i t s path down the c l i f f face. The resultant depression was approximately 1.0-1.5 metres (3-5 feet) deep, so that f a i l u r e probably occurred along a plane more or less p a r a l l e l to the ground surface and at, or close to, the base of the active layer. The bare f l o o r of the depression i s being accentuated by deeper thawing and surface flowage r e l a t i v e to the adjacent vegetated sur-face, and f a l l e n hummocks around the margin of the hollow indicate that i t i s expanding a r e a l l y . A further continuation of these processes could feasibly lead to the eventual exposure of ground ice at depth, and the i n i t i a t i o n of slump development. A few smaller slumps are located around the edges of some of the larger lakes, where similar undercutting of the banks by wave action has exposed the i c e . Figure 20 also shows a c l a s s i f i c a t i o n of the mudslumps into active and inactive forms. Since the ground ice exposures frequently have ice contents of several hundred per cent by weight, melting of the perma-fr o s t leaves deep depressions which persist for long periods of time after either complete thawing, i f t h i s ever takes place, or p a r t i a l thawing and r e b u r i a l of the ice body. Apart from their persistence as negative r e l i e f features, the i d e n t i f i c a t i o n of these scars on a e r i a l photographs i s a comparatively easy task. Due to the disturbed nature of the substratum following periods of slump a c t i v i t y , the inactive mudslumps support a highly diagnostic vegetation association, dominated by grasses, as des-cribed i n Chapter I I I . The s t a b i l i z e d headwall scarp of an old mudslump, averaging 2-3 metres (6-10 feet) i n height, i s generally smaller than that of an active slump, and i t progressively loses some of i t s id e n t i t y as i t i s colonized by vegetation. In longitudinal p r o f i l e , the f l o o r s of many of both active and inactive depressions exhibit a d i s t i n c t i v e ribbed 154 appearance. The r i b s may be annual features, representing the deposits of mud derived from the melting of the ice face i n a summer, or they may mark the terminal positions of previous cycles of mudslump a c t i v i t y following renewed undercutting at the toe of an immediately preceding cycle. Many of the currently active slumps are second- or l a t e r -generation features located wholly, or p a r t i a l l y , within the confines of the older mudslumps. Others are first-generation forms cutting back into t e r r a i n which has not been affected previously by slumping processes. The headwalls of active slumps, ranging i n height up to 10 metres (33 f e e t ) , contain variable quantities of segregated ground i c e , and retreat rapidly during the summer months. The f l o o r s of the depressions are covered, to varying degrees, with a l i q u i d mud debris derived from the melting of the ice face and the overlying active layer or overburden. Depending upon the r a t i o of ice to mineral s o i l i n the headwall exposure, t h i s mud may accum-ulate as a r e l a t i v e l y s t a t i c or slowly advancing mud lobe at the base of the ice face, or, i n less viscous cases, i t may be concentrated into strong, well-defined mudflows extending across the f l o o r of the depression (Plate IX-B). Rates of Retreat. Lamothe and St-Onge considered this slumping process to be one of the most rapid erosional agents operating i n certain parts of the a r c t i c . Their observations during the summer of 1960 showed that the thermo-scarp retreated an average of 7 metres (23 f e e t ) , with a maximum recorded recession of 10 metres (33 f e e t ) . Erosion rates on the l a t e r a l walls of the depression, with less southerly aspect, were considerably lower, averaging 0.5 metres (1.5 feet) on a northeast-facing slope and 155 2 metres (6.5 feet) on a west-facing slope. Mackay, using f i e l d obser-vations i n the Mackenzie Delta area, concluded that the average retreat of active scarp faces i s variable but probably l i e s i n the range of 1.5-4.5 13 metres (5-15 feet) per annum. Three very active mudslumps, the locations of which are also shown i n Figure 20, were investigated on Garry Island, and data on the annual rates of retreat were obtained by i n s t a l l i n g a series of numbered stakes around the headwalls of each slump. These stakes were v i s i t e d per-i o d i c a l l y throughout the 1964 and 1965 f i e l d seasons, and one of them, Slump B> was r e v i s i t e d during the summer of 1966. The amount of recession was recorded i n metres and a summary of these observations i s presented i n Table X. Figure 21 also shows the surface configuration of Slump B i n d e t a i l , with the locations of the various stakes and the lines of retreat at selected intervals (the beginning and end of the i n d i v i d u a l f i e l d seasons) during the observation period. The figures i n Table X confirm Mackay 1s conclusions that the rates of retreat are highly variable and the average values, i n each of the mudslumps, approach or even exceed the upper l i m i t of his estimate. The highest average annual r e t r e a t , 6.4 metres (20.8 f e e t ) , was recorded i n Slump A on the south coast of the i s l a n d , and this value contrasts with annual averages of 4.6 metres (15.2 feet) and 3.9 metres (12.8 feet) for Slump B, located on the north coast, and Slump C, located on the west coast, respectively. In each of the mudslumps, more than 80 per cent of the observed recession occurred during the months of July and August. 'Lamothe, C. and St-Onge, D. (1961) op_. c i t . , p. 104. Wckay, J . Ross (1966) op. c i t . , p. 72. 156 TABLE X MUDSLUMPS - RATES OF RETREAT (Metres) SLUMP A. Average Maximum Minimum Summer 1964 5.51 10.40 0.0 F a l l 1964/65 0.84 3.70 0.0 Summer 1965 5.15 9.55 0.0 SLUMP B. Summer 1964 3.75 6.50 0.0 F a l l 1964/65 0.88 1.70 0.0 Summer 1965 3.68 6.20 0.0 (July 4/64 - Aug. 22/66) 11.66 21.10 0.0 *(Aug. 22/66 - Aug. 21/67) 4.26(5.96) 6.87 0.0 *(Aug. 21/67 - July 10/68) 1.54 2.90 0.0 SLUMP C. Summer 1964 3.14 5.20 0.0 F a l l 1964/65 0.75 1.90 0.0 Summer 1965 4.81 6.90 0.0 Representative annual retreat values can be obtained for each mudslump by adding together the average figures for Summer 1964' and F a l l 1964/65. * Information supplied by Dr. J. Ross Mackay. Personal communications August 1967 and October 1968. The rates ci t e d actually represent minimum values since they include measurements based on the la s t recorded positions of stakes which had been obliterated by headwall retreat. The figure i n brackets represents the average rate of retreat based solely on 21 stakes which were s t i l l i n position on August 21, 1967. By July 10, 1968, only 14 of the o r i g i n a l 27 markers were s t i l l i n t a c t , and of these only 4 were located along act i v e l y retreating sections of the headwall. A e r i a l reconnaissance of the island i n mid-June of 1965 revealed that the mudslumps were s t i l l largely f i l l e d with snow, and the paucity of fresh debris at the base of the ice face, or on the surface of the snow, at the beginning of Jul y , indicates that most of the retreat recorded during the f a l l represents continued recession during the preceding September before the onset of continuously freezing ;air temperatures. Since the maximum recession recorded by any of the ind i v i d u a l stakes .also occurred i n Slump A, i t may seem easy to conclude that slope aspect i s a dominant factor influencing scarp retreat. Comparisons made between stakes located around the same mudslump, however, revealed that the maximum retreat did not always occur on the slopes with the most southerly aspect. The f i e l d studies indicated that the rate of recession of a mudslump headwall r e f l e c t s the complex interplay of a number of factors including climate, the height and composition of the scarp face, and the rate of removal of the thawed debris. Of these, the climatic factor appeared l o g i c a l l y to be the most s i g n i f i c a n t , and an attempt was made to establish the relationship between the rate of retreat, as measured be-tween successive observation periods, and the number of thawing degree-days occurring during the same periods. Only ac t i v e l y - r e t r e a t i n g sections of the scarp face i n each of the three mudslumps were used for t h i s study, and the recession values thus represent the average readings taken from approximately 18-20 stakes at each of the slumps. These values were plotted against the thawing indices and are shown graphically i n Figure 22. The regression l i n e f i t t e d to these points yielded the following equation: Figure 22 RELATIONSHIP BETWEEN RATE OF RETREAT AND NUMBER OF THAWING DEGREE-DAYS FOR THREE GARRY I S LAND MUDSLUMPS 2.5 X *> 2.0 X V « • -*— < * c 1.5 * • ^_ 0 1> •I OC 1.0 « CO D X • > ^ ^ ^ ^ X <0.5 • Slump A Yc = 0.91 + 0.0022 X r = +0.95 • Slump B Yc = 0.53 • 0.0034 X r = +0.87 n X Slump C Yc = 0.39 + 0.0046X r = +0.87 v ( ) 100 200 300 400 Thawing Degree - Days 160 Y = 0.50 + 0.004 X c where Y^ i s the estimate of average retreat of the headwall i n metres for a given value of X, and X i s the number of thawing degree-days. Computation of the c o e f f i c i e n t of correlation yielded a value of +0.92 indicating a strong positive correlation between the two variables. The equations of the regression lines and the coefficients of correlation for measurements taken from the individual mudslumps are also given i n Figure 22. 2 By computing the c o e f f i c i e n t of determination (r ) i t can be seen that approximately 85 per cent of the v a r i a b i l i t y i n the observed rates of retreat can be 'explained' by variations i n the a i r temperature patterns. Each of the mudslumps investigated was similar insomuch as they were a l l p a r t i a l l y located, a l b e i t to varying degrees, within the l i m i t s of an older slump. For example, Slump A, when f i r s t v i s i t e d , was wholly located within an older slump. At the end of the observation period, recession of the ice face had completely eliminated the evidence of the previous cycle along a l l but a small portion of the rim. In Slump B, evidence of an e a r l i e r cycle was found between stakes 170-175 (see Figure 21), and i n Slump C approximately one-third of the rim was s i m i l a r l y located. The significance of this type of location i s primarily i n i t s influence on the structural composition of the headwall. In t e r r a i n which has not previously been affected by mudslumps, the overburden is v i r t u a l l y r e s t r i c t e d to the active layer, and rarely exceeds 1.0-1.5 metres (3-5 feet) in thickness although l o c a l l y this figure may be augmented by s o l i f l u c t i o n . Moreover, t h i s type of surface mantle may have a moderately high ice content (Plate IX-C). In contrast to t h i s , the composition of the headwall i n a second- or later-generation slump i s intimately related to 161 the sequence of events during the previous cycle. The amount of ice exposed i s p a r t i a l l y controlled by the l e v e l at which i t was planed off during the previous cycle, and i t may be mantled by an overburden of slumped material only a portion of which corresponds to the current active layer. This type of overburden, which has the appearance of a chaotic mixture of mud, stones, turf and willows, also frequently has a very low ice content (Plate IX-D). The amount and type of ground ice exposed i n the headwall, and the r a t i o of t h i s ice to the overburden above, i s a c r i t i c a l factor influencing the rates of recession. I t i s the ice face segment which undergoes the most rapid recession, and the maintenance of fresh exposures of the ice i s imperative for continuous retreat. In Table X, i t was shown that the maximum and minimum rates recorded i n each of the mud-slumps covered wide ranges. Minima, i n actual fact points at which no recession was observed, i n a l l cases coincide with headwall sections i n which no ground ice was exposed. At a l l other points ice was exposed for some time during the observation period and, i n general, the greatest recession occurred where the r a t i o of ground ice to overburden was highest, and the former was thus continually exposed throughout-most of the summer months. Another indirect influence of this r a t i o on the rates of recession i s related to i t s effect on the v i s c o s i t y of the thawed debris. High proportions of segregated ice i n the headwall represent a greater potential moisture supply affecting the mobility of the debris, and thereby f a c i l i t a t i n g i t s removal from the base of the ice face. In this respect, i t i s also s i g n i f i c a n t to point out that there was a con-siderable v a r i a t i o n i n the ice content of the ground ice exposures them-selves i n each of the mudslumps, and therefore important differences i n 162 the volumes of debris derived from the melting of these sections of the scarp face. Active removal of this debris, allowing the maintenance of fresh exposures, i s another important contributor to continued retreat. In addition, where the headwall i s high, the momentum gained by the debris during i t s descent aids i n i t s displacement from the immediate v i c i n i t y of the foot of the slope. I t i s now possible to re-evaluate the s t a t i s t i c s given i n Table X i n terms of the preceding statements. The highest average retreat values recorded in Slump A r e f l e c t the combination of the influence of a southerly aspect; the widespread occurrence of segregated ground ice with ice contents of several hundred per cent, including many bands of almost clear i c e ; an overburden which, although t h i c k , i n many places also con-tained large quantities of segregated ice; a high headwall ranging up to 10.0 metres (33 feet) i n height; and an extremely active rate of removal of the thawed debris i n the form of strong mudflows. The l a t t e r factor was extremely s i g n i f i c a n t i n sections of the mudslump where the headwall was dominated by thick deposits of slump debris from an e a r l i e r cycle. A b r i e f v i s i t to the same slump at the end of the 1966 f i e l d season showed that a weakening of the mudflow a c t i v i t y had led to an almost complete b u r i a l of the ground ice exposures in these sections. Higher retreat values recorded i n both the other mudslumps generally re f l e c t e d a similar combination of these conditions. The lower average figure for Slump B i s probably most related to i t s northerly aspect and the prevalence of a weak rate of removal of debris. Several factors account for the lowest recession rates recorded i n Slump C including a low headwall averaging only 2.0-3.0 metres (6-10 feet) i n height; ground ice exposures i n this headwall which were more akin to a frozen mud, with ice contents averaging less than one 163 hundred per cent; and the presence of numerous g u l l i e s on the f l o o r of the depression. Whilst these g u l l i e s assist the concentration of the debris into well-defined mudflows, some of which would break through with con-siderable force to the sea, they provided numerous constriction points, at which many of the flows became blocked, resulting in a highly intermittent, and sometimes complete lack of, removal of the debris. Ablation Studies. The retreat values cited i n the preceding paragraphs were related to the rate at which material, composed predomin-antly of mineral s o i l , was detached from the rim of the headwall. I t was also established that one of the key factors influencing this detachment process was the melting back of the segregated ice exposures i n the scarp face. To obtain additional data on this l a t t e r aspect of the retreat process a series of ablation studies was made during the 1965 f i e l d season. The technique employed for these ablation studies was similar to that used to record changes i n the c l i f f p r o f i l e s of the sand head-lands. A number of stakes was i n s t a l l e d i n the ice face, normal to the surface, and the amount of ablation was determined by measuring changes i n the exposed lengths of the stakes (Plate X-A). Several problems were encountered i n using this technique. F i r s t l y , the d r i l l i n g of the holes to accommodate the wooden dowelling was an arduous task due to the presence of numerous small pebbles i n the segregated ice body. These pebbles were of s u f f i c i e n t size to halt the penetration of the d r i l l before the required depth of 45 cms. (18 ins.) was achieved. Thus, on one occasion, a t o t a l of 27 holes was started before seven could be completed to i n s t a l l the stakes on one of the ablation p r o f i l e s . Secondly, even though the dowelling used for the stakes was 2.5 cms. (one inch) i n 1 6 4 Plate X M U D S L U M P S A. A b l a t i o n studies i n Slump B showing the loca t i o n s of the stakes i n P r o f i l e I. B. G u l l i e s produced by the d i f f e r e n t i a l melting of bands of ground i c e with c o n t r a s t i n g mineral s o i l contents. C. Meltwater g u l l y system on the headwall i n Slump B. D. Rejuvenation of Slump A caused by undercutting at the toe of the slump leading to renewed exposure of the ground i c e . 165 diameter, there were a number of instances where i t was either broken off or dislodged by hummocks and debris detached from the overhanging rim above. Thirdly, problems were also introduced i n the actual measurements of the amount of ablation due to the formation of small pit s produced by accelerated melting at the base of the exposed portion of each stake. To al l e v i a t e this problem and obtain representative ablation measurements therefore, a ruler was placed across these p i t s , flush with the adjacent ice surface, and the ablation l e v e l was interpreted as being the point of intersection of the ruler and the stake. For these ablation studies, two p r o f i l e s were established i n Slump B. Observations, covering periods of 3-4 days, were made at 14 intervals ranging from 2-3 weeks. P r o f i l e I was located midway between stakes 180 and 181 (see Figure 21), at a point where the headwall just exceeded 8.0 metres (26 feet) in height and the ice face was r e l a t i v e l y smooth with only incipient g u l l i e s at the base. P r o f i l e I I was situated midway between stakes 187 and 188 where the headwall was just under 6.0 metres (19.5 f e e t ) , high and the lower parts of the ice face were ribbed by a number of d i s t i n c t g u l l i e s . The forms of these p r o f i l e s , the locations of the ablation stakes on each, and the changes i n the positions of the ice face are i l l u s t r a t e d i n Figures 23 and 24. Each p r o f i l e was surveyed by stretching a tape between two rods, one anchored i n the ground surface above the rim and the other i n the ground ice floor of the slump, and taking plumb readings to the ice surface at approximately 0.5 metre (1.6 foot) i n t e r v a l s . This procedure On the majority of these days two sets of observations were made, but the changes were too s l i g h t to be recorded on Figures 23 and 24. L o c a t i o n of A b l a t i o n Stakes A July 8 5 30 P- m. B Ju l/ 9 4 30 P- m. C July 11 1 30 P m. D July 12 1 30 P m. E July 27 4 40 P m. F July 28 4 30 P m. G July 30 4 00 P m. H July 31 4 00 P m. 1 Aug . 15 10 45 P m. J Aug . 17 10 30 P m. K Aug . 19 10 30 P m. L Sep. 5 12 30 P- m. Figure 23 L o c a t i o n of A b l a t i o n S takes G A July 28 4 40 P-m. B July 30 4 00 P-m. C July 31 2 30 P-m. D Aug. 14 11 00 a m. E Aug. 1 5 10 55 a m. F Aug. 17 10 40 a m. G Sep. 5 12.45 p.m. Figure 24 ABLATION STUDIES IN SLUMP PROFILE H Scale in Metres 168 was repeated several times, usually at the beginning and end of each observation period, and the surveyed p r o f i l e s are indicated by the s o l i d lines i n Figures 23 and 24. The intermediate positions of the ice face, as determined by the ablation stake measurements, are represented by broken lines i n the same diagrams. A comparison of the individual ablation stake measurements for each observation period showed that i n P r o f i l e I there was very l i t t l e v a r i ation over the length of the p r o f i l e , and that the ice face underwent ess e n t i a l l y p a r a l l e l retreat (Figure 23). By way of contrast, a similar comparison of the values obtained i n P r o f i l e I I showed a progressive increase i n the amount of ablation from the top to the bottom of the p r o f i l e . This d i f f e r e n t i a l was most pronounced at the beginning of the observation period i n late J u l y , when i t was i n the order of 200 per cent, but i t s magnitude gradually diminished u n t i l , by late August, i t was less than 100 per cent. This miniature ablation-altitude gradient can probably be attributed to the presence of the gully system, extending over the lower part of the scarp face, which e f f e c t i v e l y concentrated meltwater, derived from the Upper part of the ice face, into well defined-channels. The two lower stakes i n P r o f i l e I I were deliberately positioned i n the floor of one of these channels. The greater rate of melting of the ice at these locations may therefore represent the effects of additional 15 mechanical erosion by the meltwater. Furthermore, the observed decrease in the d i f f e r e n t i a l rates of ablation may possibly be attributed to the ^~*Part of this d i f f e r e n t i a l may also be attributed to the fact that the lower stakes i n P r o f i l e I I were i n s t a l l e d v e r t i c a l l y rather than at rig h t angles to the ice surface. 169 gradual contraction of the gully system towards the end of the summer. This sequence of events i s apparent i n Figure 24 by the progressive 'straightening' of the ice p r o f i l e and a diminution i n the v e r t i c a l extent of the lower concave facet. An analysis of the rates of retreat for the rim of the headwall showed a strong positive correlation with air temperatures. Due to the short time periods (ranging from 7-48 hours) used in the ablation studies i t was impractical to employ the number of thawing degree-days as an index of a i r temperatures and, instead, the number, of thawing degree-hours was used. These were calculated by taking the hourly readings from the contin-uous temperature record at the climate s t a t i o n , subtracting 32°F from each, and cumulating the totals for the respective observation periods. Figure 25 i l l u s t r a t e s graphically the relationship between the ablation rates and a i r temperatures. The regression l i n e f i t t e d to the plotted points was determined by the method of least squares, and the resultant equation was: Y„ = 0.14 + 0.019 X c where Y c i s the estimate of the amount of ablation i n centimetres for a given value of X, and X i s the number of thawing degree-hours. The regression equations for the individual p r o f i l e data are also shown i n Figure 25. Computations of the coefficients of correlation (r) and the o c o e f f i c i e n t s of determination (r ) yielded uniformly high values of +0.99 and 0.98 respectively. Thus, approximately 98 per cent of the v a r i a b i l i t y i n the ablation data can be 'explained' by variations i n the air tempera-ture patterns. Figure 25 R E L A T I O N S H I P B E T W E E N A B L A T I O N R A T E S A N D THE N U M B E R O F T H A W I N G D E G R E E - H O U R S IN S L U M P B 171 The Recession Process and Evolution of the Gully System. Headwall recession i n a t y p i c a l mudslump involves the thermal and mechanical erosion of the scarp face, accompanied by the slumping of the thawed debris and free f a l l or s l i d i n g of the overlying active 16 layer. The r e l a t i v e importance of each of these processes depends primarily on the nature and composition of the scarp face, and i n particu-l a r , the type and quantity of ground ice exposed. The recession values, cited i n the preceding paragraphs, des-cribe the rates at which blocks of material were detached from the crown of the mudslump. Many actively-retreating headwalls include small sections i n which the o r i g i n a l ground ice exposures have been eliminated either by com-plete melting or by b u r i a l beneath deep mantles of thawed debris. Although active recession may s t i l l take place i n these sections, the rate is extremely slow. Small tension cracks are developed along the crown of the slump, p a r a l l e l to the rim, and blocks of turf or mud, bounded by these cracks, gradually t i l t forward and are ultimately detached from the head-wa l l . Since the thawing of these sections releases l i t t l e or no moisture, the debris merely accumulates on the scarp face and the recession i s eventually terminated. This process, the detachment and free f a l l i n g of the active layer, i s greatly f a c i l i t a t e d where underlying bodies of ground ice are exposed i n the main scarp face. A more rapid recession of the ice face segment leads to undermining at the base of the active layer, the development of large overhangs, and a sudden, rapid collapse of large blocks of debris to the foot of the scarp face. The thermal and mechanical erosion of the main scarp face i s Mackay, J. Ross ( 1966) op. c i t . , p. 74. 172 thus the dominant process influencing the rate of headwall recession, and one of the most apparent manifestations of this process i s the presence of a well-defined gully system covering sections of the ice face. Two types of gully can be distinguished i n the mudslumps on Garry Island. The f i r s t type, observed i n Slump A, i s structural and occurs i n heterogeneous bodies of ground ice composed of alternating, more or less v e r t i c a l bands of frozen mud and clear ice (Plate X-B). The frozen muds, with ice contents averaging only s l i g h t l y more than 100 hundred per cent by weight, melt quickly and form negative features in the ice face, while the clear ice bands, which melt more slowly, stand out as positive r i b features. D i f f e r e n t i a l melting of the bands therefore produces an embryonic gully system which i s accentuated by the channelling of meltwater streaming down the ice face. The width of the channel i s controlled by the width of the d i r t bands, but mechanical and thermal erosion by the running water results in a maximum amplitude of the g u l l i e s , i n the order of 15-20 cms. (4-6 i n s . ) , at the base of the ice face. The amplitude i s p a r t i a l l y con-t r o l l e d by the fact that excavation along the lines of the frozen mud results i n greater exposure and thus more rapid melting of the r i b s of clear i c e . Since the gully system i s s t r u c t u r a l l y controlled, there i s no evidence of c y c l i c development, and the pattern remains r e l a t i v e l y stable throughout the summer months. Where the ground ice has been strongly deformed, as i n Slump A, the orientation of the g u l l i e s i s not necessarily in the direction of steepest slope down the scarp face. In contrast to the diminutive, s t r u c t u r a l l y - c o n t r o l l e d features described above, the second type of gully system includes the pattern of large ridges and g u l l i e s which produces the d i s t i n c t i v e badland topography described recently by Mackay. I t i s best developed i n massive bodies of r e l a t i v e l y homogeneous ground i c e , devoid of any marked stru c t u r a l controls, and with ice contents ranging from 100-300 per cent by weight. These conditions occurred i n Slump B (Plate X-C), and the follow-ing description of the development of the badland topography i s related to a number of investigations made i n this mudslump. Unfortunately, the e a r l i e s t date at which i t was possible to v i s i t t his mudslump was at the beginning of July, when evidence of the 18 ridge and gully system was found between stakes 178 and 190. Between stakes 178 and 181, an incipient ridge and gully system was developed on the lower metre (3 feet) of the ice face while the upper portions of the headwall were smooth. The ridges, spaced at intervals of 30-35 cms. (12-14 ins.) exhibited a t y p i c a l buttress form, being approximately 20 cms. (8 ins.) across i n their upper portions and increasing to about twice this ) width at the base. The intervening g u l l i e s averaged 6.0 cms. (2.5 ins.) i n depth near their heads, and 6-8 cms. (2.5-3.0 ins.) i n depth at the base of the ice face, with maximum depths below the ridge crests of 10-12 cms. (4-5 i n s . ) . Small t r i c k l e s of water were being channelled down the g u l l i e s , which were floored by clean ice exposures, but the adjacent ridges were mantled by a s u r f i c i a l d i r t accumulation up to 3.0 cms. (one inch) thick. Between stakes 182 and 190, the ridges and g u l l i e s became progressively larger and covered an increasing proportion of the ice face u n t i l , by stake 190, they extended to within one metre (3 feet) of the base of the active layer and exemplified the t y p i c a l badland topography par excellence. At Mackay, J. Ross (1966) op_. c i t . , pp. 68-69. 'The stake numbers refer to the positions shown i n Figure 21. 174 this point, the ridges were approximately 3.0-3.5 metres (10-12 feet) apart, as measured between crests, and reached widths of 2-3 metres (6-10 feet) across near the base of the ice face. The depth of the g u l l i e s increased from an average of 25-35 cms. (10-14 ins.) near the head to a maximum of 1.0-1.5 metres (3-5 feet) at the foot of the slope. Between stakes 190 and 196, any ground ice exposures were s t i l l mantled by snow patches and debris, and the headwall was not affected by any form of gullying. In early August, one month l a t e r , several notable changes had occurred i n the appearance of the headwall. Ground ice was then exposed i n the scarp face between stakes 172 and 191, and the badland topography was only evid ent between stakes 182 and 191. A l l previous evidence of a gully*• ing pattern between stakes 178 and 182 had now been eliminated, and the headwall was now e s s e n t i a l l y a smooth ice face except for a few minor ripples at the base. Measurements of the s i z e , amplitude and areal extent of the ridge and gully system between stakes 182 and 191 showed that they covered an increasing proportion of the ice face as the l a t t e r marker was approached, and this was accompanied by a corresponding increase i n their dimensions similar to the pattern described above. No further extension of the badland topography occurred during the month of August, since no additional exposures of ground ice occurred beyond stake 191, and the bad-land topography was eliminated as far as stake 184. Whereas no evidence of a c y c l i c pattern was found i n the gully system i n Slump A, the recession of the headwall of Slump B exhibits a de f i n i t e cycle during which the badland topography i s developed and u l t i -mately disappears. During the winter months, the mudslumps are s i t e s of deep snow accumulation driven i n by the prevailing winds. With the 175 a r r i v a l of warmer temperatures in spring and early summer, the snow begins to melt and the uppermost portions of the scarp face are the f i r s t to be exposed. Rapid melting of the ice and the base of the active layer results i n the detachment of hummocks and debris from the rim of the slump which accumulate on the surface of the snow. Water, derived from the melting of the snow and thawing of the ice face, soon erodes a series of deep g u l l i e s in the ice surface. Once the gully pattern has been established on the ice face, i t becomes progressively larger and deeper as i t continues to channel the excess water released upon melting of the ice face. The system reaches i t s maximum development when almost the entire face i s covered by the badland topography with strong channels, 1.0-2.0 metres (3.0-6.5 feet) deep, separated by prominent ridges, of the buttress type, which may be as much as 2-3 metres (7-10 feet) across at the base. As noted previously however, the pattern of ridges and g u l l i e s r a r e l y reaches within more than one metre (3 feet) of the top of the ice face,and this probably r e f l e c t s the minimum surface area required to produce a s u f f i c i e n t -ly large quantity of water that can be collected together and channelled 19 into a surface flow ( c f . Schumm, 1956). Once the ridges reach the dimensions cited above,they occupy a large proportion of the ice face, and their large surface area, coupled with their additional exposure as positive r e l i e f features, reaches a size whereby melting releases s u f f i c i e n t water to form another set of g u l l i e s on the buttresses themselves. Entrenchment and headward extension of these new g u l l i e s gradually reduces, and f i n a l l y eliminates, the large ridges and produces the more closely spaced r i p p l e _ Schumm, S.A. (1956) "Evolution of drainage systems and slopes in badlands at Perth Amboy, N.J.", Geol. Soc. Amer. B u l l . , Vol. 67, p. 607. 176 remnants at the base of the scarp face. This entire sequence, the develop-ment of new channels at points formerly occupied by ridges during an e a r l i e r period of gullying, i s very similar to the process of gully-gravure described 20 by Kirk-Bryan i n the a r i d south-western parts of the United States. The length of time i n which this cycle i s completed of course i s much less than that c i t e d by Ki r k Bryan, and appears to vary according to the location within the mudslump. The smooth nature of the ice face between stakes 173 and 178, except for the minor ripples at the foot of the slope, indicated that the cycle had already been completed on these sections of headwall with a northwest or northern aspect, whereas, on a northeast facing slope the cycle was not completed u n t i l late September. On slopes with a more easterly aspect, the cycle may not be completed before freezing temperatures set i n , and remnants of the ridges and g u l l i e s may be buried beneath the d r i f t i n g snow. During the stage of maximum development of the badland top-ography, mechanical and thermal erosion of the ice face by running water i s a decisive factor influencing the rate of recession of the mudslump headwall. Besides contributing to active erosion, the water i s also an important agent i n quickly removing thawed debris from the ice surface, and thereby maintaining a clean exposure for more rapid thermal erosion. Since both thermal and mechanical erosion occur simultaneously, i t i s impossible to evaluate precisely the r e l a t i v e contributions of each, but comparisons of headwall recession values, over smaller intervals of time, at the indivi d u a l stake positions demonstrate the effic i e n c y of the mechanical 20 Bryan, K. (1940) "Gully gravure - A method of slope retreat", Journal of Geomorphology, Vol. 3, pp. 87-107. 177 erosion. These comparisons are shown i n Table XI. TABLE XI RATES OF RECESSION AT SELECTED STAKE POSITIONS IN SLUMP B (Metres). Stake # July 7 - 2 6 July 26 - Aug. 13 Aug. 13 - Sept. 9 179 182 190 2.05 3.30 1.35 1.70 1.80 1.85 1.20 1.10 2.30 The pattern for stakes 179 and 182 i s one i n which the greatest retreat was recorded during the f i r s t observation period i n J u l y , when the ridge and gully system was strongly developed, and a decreasing rate of recession, through August and into September, as the ridges and g u l l i e s disappeared and mechanical erosion was reduced. In comparison to t h i s , the pattern of retreat shown by stake 190 i s one of increasing rates through to late August and early September, at which time the badland topography had reached i t s maximum development. ice face undergoes more or less p a r a l l e l retreat. On the smooth ice face losses are due mainly to melting i n response to incoming short and r e f l e c t -ed long wave radiation, and the water, released by melting, follows a sinu-ous passage between p a r t i c l e s of ablation debris protruding from the ice surface. The recession process becomes even slower as the layer of surface d i r t accumulates to maximum thicknesses of 2-3 cms. (one inch). Once this thickness i s attained, the surface d i r t becomes unstable and moves as a mud s l i d e to the base of the ice face. Loc a l l y , i n s t a b i l i t y may be reached before these thicknesses are attained, being triggered by the impact of stones and debris detached from the overhanging active layer. An indicat-ion of the slow nature of th i s process was obtained by spreading a green With the eventual disappearance of the badland topography the 178 fluorescein dye on a clean section of the upper part of the c l i f f face. Despite the fact that the scarp face had an i n c l i n a t i o n of 66 degrees, the coloured water took a minimum of 15 minutes to seep down to the base, whereas a similar dye introduced into one of the s l i d i n g mud slimes reached the base i n a matter of only a few seconds. The Mudslump Cycle. The maintenance of fresh exposures of i c e , permitting the continued recession of the scarp face, i s a c r i t i c a l factor influencing the longevity of the mudslump cycle. The length of time that an exposure of ground ice i s maintained i n turn i s largely determined by the delicate balance between the rates at which debris i s supplied to, and removed from, the base of the scarp face. Figure 26 i s a composite p r o f i l e across a section of Slump B for the period July 4, 1964 to August 13, 1966, showing the i n d i v i d u a l mud lobes, representing the accumulated debris derived annually from melting of the scarp face, and the approximate position of the buried surface of the ground ice as determined by a series of d r i l l holes. Figure 27 i s a generalized diagram of the same p r o f i l e , showing the position of the ice 21 face at selected intervals during the same time period. Total recession of the headwall (r) i n t h i s time amounted to a horizontal distance of 16.5 metres (54 f e e t ) , during which the height of the scarp face (h) decreased from an i n i t i a l value (h-^ ) of 9.1 metres (30 feet) to only 5.7 metres (19 feet) - (h^ ?) - at the end of the observation period. This decrease, involving a reduction of 37.5 per cent, was due The p r o f i l e shown i n Figures 26 and 27 was located near stake 181 (see Figure 21) . Figure 26 GARRY I S L A N D MUDSLUMP PROFILE 1964-1966 S L U M P B |-« Mud Lobes -*-| 1 9 6 4 | 19 6 5 | 1 9 6 6 Ver t i ca l Exaggera t ion x2 Figure 27 RECESSION DIAGRAM OF A GARRY ISLAND MUDSLUMP 1964 - 1966 r H Depression c rea ted by thawing at base of scarp 1-6 A p p r o x i m a t e former positions of the ice face 1 July 4, 1964 (Surveyed) 4 August 10, 1965 (Surveyed) 2 September 4, 1964 5 September 6, J965 3 July 7, 1965 6 August 13, 1966 (Surveyed) 181 primarily to a slow rate of removal of debris, as indicated by the gently overlapping nature of the individual annual lobes, with the resultant b u r i a l of the basal sections of the ice face. The reduction would have been even greater, 44.0 per cent, but for the fact that the height of the scarp was augmented by a r i s i n g slope of two degrees on the surface above the rim of the mudslump. In a recent publication, Mackay has derived an equation for computing the approximate thawed volume of a unit section of a mudslump, 22 provided that the excess water i s free to escape. Assuming that the volume changes incurred during thawing of the active layer are ne g l i g i b l e , the thawed volume i s approximately equal to: ra + rV (h - a + d) where r = the horizontal distance of scarp retreat a = the average thickness of the active layer -1 V = the r a t i o of the i n i t i a l volume of frozen ground to the t o t a l volume of segregated ice h = the height of the scarp face and d = the mean depth of thaw at the scarp foot. The above equation was derived from a theoretical s i t u a t i o n i n which there was p a r a l l e l retreat of the ice face, with no a l t e r a t i o n i n height, across a perfectly horizontal surface. The major contrast between th i s situation and the one i l l u s t r a t -ed i n Figure 27 i s that the height of the scarp face (h) did not remain constant i n the l a t t e r . Since i t was found that the base points of the surveyed p r o f i l e s i n Figure 27 were located approximately along a straight l i n e , i t was possible to overcome this problem by averaging a series of measurements taken d i r e c t l y from t h i s diagram. The mean height of the 22 Mackay, J. Ross (1966) ££. c i t . , p. 71. 182 scarp face, derived i n t h i s manner, amounted to 7.5 metres (24.6 feet) and t h i s value incorporates the changes brought about by the sloping of the surface to the crown of the slump and basal mantling of the ice face by -1 thawed debris. The value of V was estimated using the nomograph included 23 in Mackay's a r t i c l e . A t o t a l of 40 samples (Table X I I ) , collected at different times from the ice face, gave an average ice content of 300 per cent (weight of ice to dry s o i l ) , and combining t h i s with an estimated porosity of the o r i g i n a l unfrozen ground of about 0.3 y i e l d s a value for -1 V of 0.15. Representative values of the average thickness of the active layer (a) and the mean depth of thaw (d), also taken from Figure 27, were 0.46 and 0.34 metres (1.5 and 1.1 feet) respectively. Substitution of these values into the equation gives an approximate thawed volume along a unit section of 0.3 metres (1 foot) width of 7.82 cubic metres (276 cubic f e e t ) . Planimetric measurement of the volume of debris shown i n Figure 27 gives a corresponding volume of 7.13 cubic metres (252 cubic f e e t ) . The close s i m i l a r i t y of these two figures indicates the prevalence of a very weak export of material, and the degree to which most of the thawed debris derived from the melting of the scarp face merely accumulated at the foot of the slope. Assuming that the present status quo remains es s e n t i a l l y unchanged, the longevity of the present cycle i n Slump B can be r e l i a b l y estimated. A continuation of the current rate of reduction i n the height of the ice exposure, one metre (3 feet) per annum, would r e s u l t i n the termination of the cycle by approximately 1972. 23 Mackay, J . Ross (1966) ££. c i t . , p. 72. 183 TABLE XII ICE CONTENTS (WEIGHT OF ICE TO DRY SOIL) OF SAMPLES TAKEN FROM EXPOSURE OF GROUND ICE IN SLUMP B. Sample Total Weight Weight of Ice Weight of Dry S o i l Ice Contei (grams) (grams) (grams) (%) 1 179.06 112.09 66.97 167.37 2 154.95 114.47 40.48 282.78 3 161.98 94.52 67.46 140.11 4 169.10 123.99 45.11 274.86 5 217.37 133.59 83.78 159.45 6 183.37 111.47 71.90 155.03 7 149.90 107.97 41.93 257.50 8 142.15 113.66 28.49 398.95 9 165.13 110.41 54.72 201.77 10 158.99 131.98 27.01 488.63 11 153.68 143.81 9.87 1,457.04 12 173.18 86.82 86.36 100.53 13 178,70 133.68 45.02 296.93 14 124.00 98.73 25.27 390.70 15 180.50 103.79 76.71 135.30 16 174.37 100.63 73.74 136.47 17 166.04 121.18 44.86 270.13 18 180.57 132.40 48.17 274.86 19 166.31 95.20 71.11 133.88 20 144.85 115.62 29.23 395.55 21 155.21 123.83 31.38 394.61 22 161.12 109.18 51.94 210.20 23 167.04 118.80 48.24 246.27 24 187.45 114.25 73.20 156.08 25 164.43 95.91 68.52 139.97 26 167.97 123.53 44.44 277.97 27 156.55 93,48 63.07 148.22 28 157.04 120.29 36.75 327.32 29 181.75 142.38 39.37 361.65 30 163.23 121.36 41.87 289.85 31 159.13 127,98 31.15 410.85 32 168.78 123.69 45.09 274.32 33 168.04 128.62 39.42 326.28 34 173.80 100.89 72.91 138.38 35 179.20 130.97 48.23 271.55 36 166,80 140.67 26.13 538.35 37 186.88 142.24 44.64 318.64 38 167.05 134.46 32.59 412.58 39 157.17 119.14 38.03 313.28 40 194.69 148.31 46.38 319.77 Average 167.69 299.85 184 The sequence of events i n Slump A indicates the manner i n which the period of mudslump a c t i v i t y may be prolonged. In August 1964, a second cycle of slumping was i n i t i a t e d even before the ex i s t i n g cycle had been terminated. Undercutting at the toe of the l a t t e r resulted i n the exposure of ground ice previously buried beneath the mud debris (Plate X-D). By the summer of 1966, the new exposure had retreated headward to j o i n up, and l o c a l l y eliminate, a l l evidence of the f i r s t cycle. Unless a similar sequence of events exposes the ground ice buried beneath the f l o o r of Slump B, a continu-ation of mudslump a c t i v i t y beyond the estimated six-year period appears u n l i k e l y , and the mud surfaces w i l l gradually be colonized by a vegetation 24 succession similar to that described i n Chapter I I I . MUDFLOWS The significance of mass-wasting as a major geomorphic process i n the moulding of landforms was slowly recognized as a r e s u l t of numerous, independent studies i n a wide variety of climatic environments. The important role of weathering and mass-wasting i n the sculpturing of A r c t i c landscapes i s generally acknowledged, and some authors regard the cumulative effect of a l l forms of mass-wasting as being the most important l e v e l l i n g 25 1 process operating i n these high latitudes. While t h i s may be undeniably true, these studies have been dominated by investigations of slow-moving forms of mass-movement under the general heading of s o l i f l u c t i o n features, whereas scant attention has been paid to more rapidly-moving forms. Most Z4" By the summer of 1968, Slump A was v i r t u a l l y inactive. Inform-ation supplied by Dr. J . Ross Mackay, personal communication, October, 1968. 25 Jenness, J.L. (1952) "Erosive Forces i n the Physiography of Western A r c t i c Canada", Geog. Review, Vol. 42, p. 247. 185 of the geomorphological l i t e r a t u r e pertaining to mudflows for example con-s i s t s of descriptions of these features from temperate la t i t u d e s , and a r i d 26 or semi-arid environments i n p a r t i c u l a r . Few authors have described mudflows i n A r c t i c lands, and one has even expressed surprise that they 27 should be important components of mass-wasting i n these latitudes. This certainly i s not the case i n permafrost areas underlain by unconsolidated sediments containing variable quantities of segregated ground i c e . During the warm summer months, such sediments are readily transformed from the i r frozen state to mobile mud s l u r r i e s , and the mudflow i s a ubiquitous feature i n these l o c a l i t i e s . In his c l a s s i c monograph dealing with a l l forms of mass-movement, Sharpe, following e a r l i e r work done by Blackwelder, l i s t e d four major conditions which appear to be most favourable for the occurrence of mudflows: 1. an abundant but intermittent water supply 2. the absence of a substantial vegetation cover 3. unconsolidated or deeply weathered material containing enough clay or s i l t to aid i n l u b r i c a t i o n of the mass, and 28 4. moderately steep slopes. In the case of mudflows on Garry Island, the moisture supply comes primarily from the thawing of the frozen ground, and of bodies of ground ice i n 26 Blackwelder, E. (1928) "Mudflow as a geologic agent i n semi-a r i d Mountains", Geol. Soc. Amer. B u l l . , Vol. 39, pp. 465-480. Sharpe, C.F.S. (1938) Landslides and Related Phenomena, Pageant Books Inc., New Jersey, p. 4. 27 Jenness, J.L. op_. c i t . , p. 28 Sharpe, C.F.S. op_. c i t . , p. 56. 186 p a r t i c u l a r . As long as fresh exposures of ice are maintained, there i s a f a i r l y steady water supply throughout the summer months. The vegetation associations range from a continuous mat of herbs, shrubs and sedges on the upland surfaces, to a variable cover of grasses i n the case of second- or later-generation slumps. The r e s t r i c t i n g influence of the vegetation cover i s limited however by the fact that most of the plants are only shallow-rooted i n the active layer, and the recession of the ice face frequently results i n an undermining of the plant cover. The f l o o r s of the depressions, over which the mudflows t r a v e l , are often devoid of any vegetation at a l l . Since large bodies of segregated ground ice are best developed i n s i l t s or f i n e sands, the predominant mineral sizes accord with those specified by Sharpe. The slope factor does not appear to be too c r i t i c a l i n Garry Island flows, since the ice faces frequently have ice contents of several hundred per cent (expressed as the weight of ice to dry s o i l ) , and thawing produces large quantities of excess water which greatly f a c i l i t a t e mudflow movement over very gentle slopes. Sharpe's c l a s s i f i c a t i o n of mudflows recognized three w e l l -defined types - semi-arid, alpine and volcanic - which he claimed were 29 created d i f f e r e n t l y and bore d i s s i m i l a r relations to other processes. The s i g n i f i c a n t relationship of mudflows to other processes i n the mudslump cycle has already been alluded to i n the preceding section. The perpetu-ation of t h i s cycle depends i n part upon the delicate balance between rates of supply of debris to, and removal of t h i s debris from, the base of the ice scarp. Unless the debris i s transported away, the base of the ice face i s buried and the height of the ice face, and consequently the potential 29 Sharpe, C.F.S. op_. c i t . p. 57. 187 volume of moisture supply which i s capable of contributing to the mobility of the flows, i s thus progressively attenuated. Mudflows are the primary agents responsible for the transportation and removal of th i s debris. In thi s respect, i t should be noted that i n a l l the mudflows studied i n the Mackenzie Delta area,the only functional role observed was that of transport-ation of debris, and there was nothing to indicate that mudflows are responsible for the actual excavation of the depressions as suggested by 30 Lamothe. and St-Onge. On th i s basis, using the c l a s s i f i c a t i o n c r i t e r i a established by Sharpe, i t seems v a l i d to distinguish yet another type of mudflow - the a r c t i c - the genesis of which, and relationship to other processes, i s uniquely or intimately related to s p e c i f i c permafrost conditions. A l l the mudflows observed i n the Mackenzie Delta area are produced by seasonal thawing of the permafrost. They occur i n a wide range of sizes depending primarily on the nature of, and volume of ice contained i n , the sediments. Small flows, often less than one metre (3 feet) wide, and considerably shallower i n depth, descend well-defined g u l l i e s i n the coastal b l u f f s and are generated by the melting out of ice-wedges. Since the r a t i o of ice to mineral s o i l i n these exposures may be quite small, and since almost a l l the debris comes from the sediments themselves, rather than from within the ground i c e , the importance of sediment type and slope i s probably much more c r i t i c a l than i n the case of the larger flows. Channel gradients i n these g u l l i e s frequently reach 30-40 degrees. The influence of sediment type i s also demonstrated i n these small flows. _ _ Lamothe, C. and St-Onge, D. (1961) "A note on a p e r i g l a c i a l erosional process i n the Isachsen area, N.W.T.", Geographical B u l l e t i n , No. 16, p. 104. 188 The ice-wedges found i n the s i l t - c l a y b l u f f s are larger than those found i n the sand headlands, but the thawed material found i n the l a t t e r l o c a l i t i e s i s not a f l u i d , mobile mud. The largest mudflows are found i n association with the mud-slumps, where ice c l i f f s i n the headwalls may extend for several tens of metres horizontally and from 5-10 metres (16-33 feet) v e r t i c a l l y . Figure 28 represents a contour map of one of these flows on the south side of Garry Island. The mudflow i s located within the confines of a second- or later-generation slump, and flows over deposits, having a minimum thickness of 2.75-3.0 metres (9-10 f e e t ) , l a i d down by former mudflows. Whilst the main body of the mudflow occupies a well-defined channel, t h i s i s generally a secondary ch a r a c t e r i s t i c rather than the r e s u l t of flowing along a pre-existing channel. The positions of two former mudflows are shown on the same map, and their temporal sequence was determined by the overlapping nature of the mud deposits. The positions of these successive flows are determined by the general slope trend of the slump f l o o r , and also by points of weakness i n the margins of the e a r l i e r flows. The mudflow i n Figure 28 originates at the headwall of a mud-slump i n which segregated ground i c e , 2.75-3.0 metres (9-10 feet) thick, i s overlain by an average of 1.5-1.8 metres (5-6 feet) of overburden. The mudflow extends for a distance of approximately 80 metres (260 feet) from the base of the ice face to the shoreline, and three d i s t i n c t sections can be i d e n t i f i e d i n t h i s distance. The uppermost section consists of a mud reservoir with a surface area of 135 square metres (1450 square feet) and an average surface gradient of 1 i n 12 (Plate XI-A). The main channel of the mudflow, which constitutes the second section, extends for a distance of approximately 43 metres (143 feet) from the outlet of the mud reservoir 189 Old Mud Levees Contour interval 0.5 metres S c a l e 5 Metres 10 15 190 Plate XI A G A R R Y I S L A N D M U D F L O W B. Main channel of the mudflow immediately below the outlet of the mud reservoir showing the locations of the upper and middle rows of markers. C. Terminal section of the mudflow showing i t s descent over a wave-cut b l u f f , two positions of the lower row of markers, and the surface of the mud lobe. 191 to the apex of the mud lobe (Plate XI-B). The width of the channel, meas-ured from the inner margins of the mud levees, ranges from 0.75-3.0 metres (2.5-10.0 f e e t ) , and has an average surface gradient of 1 i n 10. There are two noticeable sections of the channel however, where the surface gradient i s as steep as 1 i n 2. One of these sections i s just at the outlet of the mud reservoir, and the other i s at the lower end of the channel where the mudflow descends a wave-truncated bluff cut i n e a r l i e r mudflow deposits (Plate XI-C). The t h i r d , and f i n a l , element of the mudflow comprises an almost semicircular mud lobe which covers an area of just over 170 square metres (1830 square feet) on the foreshore (Plate XI-C). The irregular nature of the eastern side of th i s mud lobe (Figure 28) r e f l e c t s the re-s t r i c t i n g influence of two large driftwood logs on the beach. The mud lobe, which i s approximately 2 metres (6.5 feet) thick near the apex, has an average surface gradient of about 5 degrees increasing to 75 degrees along the terminal edge of the lobe. The central portions of the mud lobe have a fresh mud surface i n which a series of crudely concentric ridge patterns i s v i s i b l e . The d r i e r , peripheral regions of the lobe however are characterized by a well-developed system of transverse and r a d i a l crack patterns similar to those found at the snouts of many glaci e r s . Rates of Movement. Preliminary observations of the mudflows on Garry Island showed that they exhibited highly irregular rates of movement, and attempts were made to record these rates and determine the nature of the processes influencing them. Three lines of styrofoam b a l l s , 6.5 cms. (2.5 ins.) i n diameter, and two lines of s t i c k s , approximately 50 cms. (19.5 ins.) i n length, were i n s t a l l e d across the main channel of the mudflow. The 192 points at which the lines of b a l l s were i n s t a l l e d are shown i n Figure 28. Attempts to use table tennis b a l l s as markers were unsuccessful, as they were too buoyant and were easily moved by wind and surface streams of water.. The positions of the various markers were recorded at intervals over a 78-hour period, and the results are shown digrammatically i n Figure 29. In the descriptive comments pertaining to the mud lobe, an analogy was made between the pattern of cracks found on the surface of the mud lobe and those observed i n the snout regions of many glaci e r s . The flow patterns i n the mudflow, as indicated by the lines of markers, are also analagous to those found i n some glaci e r s . The greatest v e l o c i t i e s occurred i n the centre of the channel, and there was a decrease towards the margins r e f l e c t i n g f r i c t i o n a l drag against the bordering mud levees (Figure 29). In curved sections of the channel, the maximum average v e l -o c i t i e s occurred towards the outer side of the curve. The changing attitudes of the lines of sticks also indicated that the surface layers of the mudflow moved more rapidly than the mud at depth. As the lines of sticks moved down the mudflow, the markers were gradually rotated, be-coming increasingly inclined at angles and pointing i n a down stream di r e c t i o n , u n t i l eventually the st i c k s were l y i n g horizontally on the sur-face of the mud. Observations over a 78-hour period showed that the upper and middle lines of b a l l s moved at average rates of 8.5 and 6.0 cms. (3.3 and 2.4 ins.) per hour respectively, while the lower line; of b a l l s moved at Figure 29 MUDFLOW - R A T E S OF M O V E M E N T 194 31 an average rate of 28 cms. (11 ins.) per hour. These o v e r a l l averages are not t r u l y representative however, since values calculated f o r shorter lengths of time show that two d i s t i n c t patterns of movement were discern-i b l e during t h i s period. During the f i r s t 24 hours of observation, the average v e l o c i t i e s for the upper, middle and lower lines of b a l l s were 2.5, 7.5 and 28.0 cms. (1.0, 3.0 and 11.0 ins.) respectively, while for the l a s t 10 hours i n which these rates were recorded the average v e l o c i t i e s for the upper and middle lines were 37.0 and 15.0 cms. (14.6 and 5.9 ins.) respectively, and the lower l i n e of b a l l s experienced p r a c t i c a l l y no move-ment. These figures indicate that, during the 78-hour period, there was a change from an i n i t i a l state i n which the average ve l o c i t y of the mudflow increased f a i r l y regularly towards the terminal mud lobe, to a later state i n which the greatest average v e l o c i t i e s were recorded i n the section of the channel immediately below the outlet of the mud reservoir. The observation period to which the above values apply ended at 6.30 p.m. on August 20, 1966. The mudflow was not v i s i t e d again u n t i l 6.00 p.m. the following day when i t was found that several major changes had taken place. A l l the b a l l s , with the exception of two which had been l e f t stranded on the mud levees,' had been transported down onto the sur-face of the mud lobe, indicating that much greater v e l o c i t i e s had occurred during t h i s time period.(Plate XII-A). Markers located i n the upper l i n e of b a l l s f o r example, which had previously moved a t o t a l of 6.6 metres (21.7 feet) over a period of 78 hours or an average rate of 8.5 cms. These averages represent the mean values of each of the b a l l s i n each l i n e excluding those which were l e f t stranded on the mud levees. Since the lower l i n e of b a l l s was located near to the terminal section of the channel, the markers were frequently relocated and the rate c i t e d i s actually an average of three independent sets of measurements. 195 P l a t e X I I M U D F L O W - S U R G E P H E N O M E N A A. Appearance of t h e mud lobe f o l l o w i n g a p e r i o d of surge f l o w , showing the f r e s h mud s u r f a c e and the t r a n s p o r t e d markers. B. F r e s h s t r i a t i o n s and scou r marks on the i n n e r margins of a mud l e v e e , i n d i c a t i n g t h e l e v e l t o w h i c h mud r o s e i n the c h a n n e l d u r i n g the surg e . C. D e f o r m a t i o n of o r g a n i c m a t e r i a l on the beach produced by the a d v a n c i n g mud l o b e . 196 (3.4 ins.) per hour., had since been transported an additional 48.8 metres (160 feet) i n only 24 hours; an average rate of s l i g h t l y more than 2 metres (6.6 feet) per hour. Studies made on other mudflows on the i s -land show that the v e l o c i t i e s achieved during these surges are consider-ably higher than those indicated by t h i s average. During the summer of 1964', a mudflow of similar dimensions, located i n the f l o o r of another mudslump, reached v e l o c i t i e s ranging from 1.5-3.0 metres (5-10 feet) per second during one of these surges. As a r e s u l t of this surge, small blocks, previously bounded by desiccation cracks, were plucked from the inner margins of the mud levees. Fresh s t r i a t i o n s and scour marks on these same levees indicate that the l e v e l of the mud rose 45-60 cms. (17.5-23.5 ins.) within the confines of the channel (Plate XII-B). The edge of the mud lobe advanced by distances ranging from 1.75-3.5 metres (5.5-11.5 f e e t ) , pushing ahead of i t r a f t s of washed organic material on the beach (Plate XII-C). A large driftwood log, approximately 20.5 metres (67 feet) long and up to 0.6 metres (2 feet) i n diameter, which was anchored on the beach by a 0.9-1.2 metre (3-4- foot) root spread, was also pushed bodily forward by the advancing edge of the lobe. During t h i s particular surge, the mud lobe did not reach the sea, but on numerous other occasions i t was observed that mudflows had b u i l t prominent lobes out into the sea. Although these lobes are extremely soft underfoot, the mud i s highly tenacious, and many lobes may r e s i s t wave erosion for several weeks before they are f i n a l l y obliterated - usually during storm surges. The c h a r a c t e r i s t i c wave-like motion of mudflows has been 197 32 described by numerous authors. Sharp and Nobles, i n the i r description of the Wrightwood mudflow i n Southern C a l i f o r n i a , gave the following account of the flow: "The debris came down the channel above Wrightwood i n a succession of waves or, more appropriately, surges which usually started about 9:00 or 9:30 i n the morning, reached a peak of frequency i n the early afternoon, and tapered off to an end by late afternoon. F l u i d i t y was greatest at midday when the surges succeeded each other at intervals of a few seconds to tens of minutes. At other times, p a r t i c u l a r l y i n late, phases of the a c t i v i t y , hours intervened between surges".^3 In another study of the same mudflow, four possible explanations of the surges were offered: (1) periodic sloughing of debris i n the source area; (2) temporary choking of the channel; (3) caving of undercut banks; and 34 (4) f r i c t i o n between the moving debris and the channel. The f i r s t two factors, considered to be the most s i g n i f i c a n t i n the case of the Wright-wood mudflow, appear to be the most satisfactory explanations for the surge phenomena exhibited by the mudflows on Garry Island. The immediate cause of the Wrightwood flow was the melting of winter snow i n the head regions, and the p e r i o d i c i t y of the surges, with maximum frequency occurring during the daytime and a lack of flowage at night, was thought 35 to be related to diurnal variations i n the amount of melting. I t was hoped that i t would be possible to check t h i s relationship on Garry 3^For example, Blackwelder, E„ (1928) op_. c i t . , pp.. 465-480. 3 3Sharp, R CP. and Nobles, L„H. (1953) "Mudflow of 1941 at Wrightwood, Southern C a l i f o r n i a " , Geol. Soc. Amer. B u l l . , Vol. 64, p. 551. 34 I b i d . , p. 551. The a r t i c l e c i t e d i s Gleason, C.H. and Amidon, R.E. (1941) "Landslide and mudflow, Wrightwood, C a l i f o r n i a " , C a l i f o r n i a Forest and Range Experiment Station, Unpub. report, pp. 1-7. 3 5Sharp, R.P. and Nobles, L.H„ (1953) op_. c i t . , p. 551. 198 Island, since the occurrence of the mudflows i s intimately related to the thawing of ice bodies i n the permafrost. Accordingly, hourly observations of the amount of movement and rates of ablation of the ice face, measured by recording the exposed length of n a i l s driven into the i c e , were made continuously over a 32-hour period to determine whether or not there was any correlation. The values obtained during t h i s period showed that there was l i t t l e or no s t a t i s t i c a l correlation between the rate of movement of the mudflow and the rate of ablation of the ice face. The ablation measure-ments, crude as they were, did show evidence of a diurnal cycle, but there was no evidence of a similar cycle i n the mudflow v e l o c i t i e s . Indeed, the upper l i n e of b a l l s experienced p r a c t i c a l l y no movement at a l l . Figures for the maximum movements encountered i n both the middle and lower lines of b a l l s showed that there was a progressive decrease i n the rate of move-ment throughout the period. Attempts were also made to determine the influence of the vi s c o s i t y of the mud on the flow using the formula given by Sharp and Nobles: 3 6 2 n = dg sin© Z 0 where n = the co e f f i c i e n t of v i s c o s i t y d = the density of the f l u i d debris g = the gr a v i t a t i o n a l force © = the angle of slope of the ground ZQ= the thickness of the flow (cms.) Sharp, R.P. and Nobles, L.H. (1953) o£. c i t . , p. 552. 199 and V s = the velocity at the surface (cms./sec.) The use of t h i s formula involves simplifying assumptions, among which are Newtonian v i s c o s i t y , no marginal or terminal influences, no s l i p on the base and no shear stress on the upper surface, and laminar flow p a r a l l e l to the base. Of these assumptions, only the second one i s perhaps completely v a l i d , since the surface vel o c i t y value i s a maximum taken from the centre of the mudflow. Since the density of the f l u i d was not measured d i r e c t l y , an approximation was made using the data for the weight of mineral s o i l and weight of water i n each of the samples taken, and a value of 2.65 gm cm as the average unit weight of the s o i l debris. Using the l a t t e r f i g ure, an equivalent volume was obtained for the s o i l , and the o v e r a l l density of the sample was then calculated. The calculations of the c o e f f i c i e n t s of v i s -cosity of the mud at the three lines of b a l l s are shown i n Table X I I I . TABLE XIII COEFFICIENT OF VISCOSITY OF A GARRY ISLAND MUDFLOW d sin 9 z o n Upper B a l l s 1.7 .0698 88,4 .00088 5.17 x 10 8 poises Middle B a l l s 1.7 . 1045 82.3 .00124 4.72 x 10 8 poises Lower B a l l s 1.7 .1736 39.6 .00353 6.39 x 10 7 poises As these figures indicate, there i s an expected inverse re-lationship between v i s c o s i t y and rate of flow; i . e . the higher the v i s c o s i t y , the slower the movement. Since each of the samples taken had the same f l u i d density, the most s i g n i f i c a n t factors influencing the rate 200 of movement appear to be the angle of slope and the thickness of the flow. In summary, i t i s apparent that the mudflows on Garry Island exhibit two types of flow which occur i n an alternating sequence, a l b e i t with variable p e r i o d i c i t y . The c o n t r o l l i n g factor determining the type of flow i s the temporary blocking of the channel, i n this case just below the outlet of the mud reservoir. Once blockage ; of the channel occurs, either by stagnation of the mud deposit or by clumps of organic material, the downstream sections of the mudflow are deprived of additional supplies of debris, although continued l u b r i c a t i o n of the flow may be aided by streams of meltwater which percolate through and around the blockage. During t h i s period the mudflow exhibits a form of extending flow, as witnessed by a pattern of increasing v e l o c i t i e s i n a downstream di r e c t i o n ; an o v e r a l l general decrease i n a l l v e l o c i t i e s as the l e v e l of the mud i n the channel i s lowered; and the accompanying development of lunate tension cracks across the surface of the mud. Under these flow conditions, the influence of the gradient of the channel f l o o r appears to be more s i g n i f i c a n t than the thickness of the mudflow, since the lower l i n e of b a l l s moved faster than either of the other two lines despite the fact that the thickness of the flow was only one-half as great as at the other l o c a l i t i e s . The de-crease i n the v e l o c i t i e s towards the end of th i s phase of extending flow, besides r e f l e c t i n g the thinning of the flow, probably r e f l e c t s the fact that, as the mud levees were exposed, increasing quantities of water were channelled off through desiccation cracks. The phase of extending flow i s terminated f i r s t i n the upper reaches of the mudflow, as shown by the increased v e l o c i t i e s recorded i n thi s section towards the end of the observation period. This t r a n s i t i o n takes place when the accumulation of mud i n the reservoir builds up 201 s u f f i c i e n t pressure to force the blockage of debris downstream. The i n -creased v e l o c i t i e s recorded at the upper l i n e of b a l l s corresponded with the removal of debris from the c o n s t r i c t i o n just below t h i s l i n e . Once the material choking the channel has been cleared, the contents of the mud res-ervoir are discharged ra p i d l y , and the mudflow attains i t s greatest v e l o c i t i e s during these pressure surges. The frequency with which these surges occur depends on the size and nature of the blockage, the most fav-ourable locations for which are points of c o n s t r i c t i o n or slackening gradient i n the channel, and the rate of debris accumulation, and conse-quently pressure build-up,, i n the reservoir area or upstream sections of the channel. In addition to the pressure f a c t o r , the augmented v e l o c i t i e s during these surges may also be related to v i s c o s i t y changes as the thick-ness of the mudflow increases. Mud Levees. The edges of many mudflows, irrespective of their s i z e , are 37 marked by sharp, linear ridges termed mud levees. Figure 28 shows both active and inactive mud levees bordering a series of mudflows. Mud levees are generally symmetrically arranged on either side of the median channel, but there may be a marked asymmetry, with broader, higher levees on the outer curves, where the course of the mudflow i s sinuous. The ridges, the crests of which may be either sharp or rounded, vary i n height from only a few centimetres to almost one metre (3 f e e t ) . These heights normally increase towards the terminal portions of the flow, but the pattern i s by no means uniform since high ridges often occur at Sharp, R.P. (1942) "Mudflow Levees", Journal of Geomorphology, No. 5, pp. 222-227. 202 points of constriction i n the channel irrespective of their location along i t s length. The ind i v i d u a l levees are highly asymmetrical with the inner margins being shorter and steeper than the outer sides. This difference r e f l e c t s the fact that the inner margin of a mud levee i s i n i t i a l l y pro-duced by shearing, along more or less v e r t i c a l planes, developed between mud which has stagnated along the outer edges of the flow and less viscous mud s t i l l moving i n the a x i a l part of the flow. The inner faces of levees bordering active mudflows are often characterized by d i s t i n c t i v e scour marks or s t r i a t i o n s indicating that, once established, the i r slopes may be modified by either the erosive or plastering action of subsequent mudflows moving down the same channel. The outer slopes of the mud levees on the other hand are shaped ent i r e l y by deposition, and they frequently exhibit a multi-lobate character where small flows have topped the crests of the ridges and cascaded down the outer sides. Excavations of mud levees bordering inactive flows showed that they are composed of s i l t and clay with very few stone accumulations. The upper parts of the levees often exhibit a weak s t r a t i f i c a t i o n . Since the ridges are composed essenti a l l y of s o l i d i f i e d mud, their surfaces are f r e -quently covered with networks of desiccation cracks. Excavations were also made across the channels of inactive mudflows, and these often con-tained a larger number of stones than the adjacent levees. In the only d e f i n i t i v e paper r e l a t i n g to mud levees, Sharp, describing features i n the St. E l i a s Range, Yukon T e r r i t o r y , attributed them to be residual features of bouldery alluvium pushed aside by advan-38 cing streams of mud. The mudflows on Garry Island do not traverse 3 8 I b i d . 203 steep, boulder-strewn slopes similar to those described by Sharp i n the mountainous t e r r a i n of the St. E l i a s Range. The structure and composition of the mud levees on Garry Island suggest that they originate i n an en-t i r e l y d i fferent manner, and studies of active mudflows indicate that the levees are produced by a progressive bleeding of moisture from the mud:- .. flows. The mud loses water by direct surface runoff and by percolation into the underlying ground, a process which i s often aided by f r o s t and desiccation cracks covering the surface, u n t i l the v i s c o s i t y i s such that motion ceases. This stagnation process occurs f i r s t along the outer margins of the flow, while the central parts are s t i l l r e l a t i v e l y mobile and continue to move. Shear surfaces are developed along the inner margins of the stagnant mud, and this produces the central channel between the bordering ridges or levees. The height of the ridges may be increased by temporary choking of the channel and l a t e r a l s p i l l i n g of the mud, which explains the weak s t r a t i f i c a t i o n observed i n the excavations. In extreme cases, and especially during the pressure surges, the mud may com-pletely override the levees at low, or weak, points and establish an ent i r e l y new course. PATTERNED GROUND Patterned ground, which may be c l a s s i f i e d on the basis of geo-metric shape and presence or absence of sorting, i s a widely-adopted term for the more or less symmetrical forms, such as c i r c l e s , polygons, nets, steps and s t r i p e s , that are c h a r a c t e r i s t i c of, though not necessarily 204 39 confined to, a mantle subject to intensive f r o s t action. Patterned ground i n the Mackenzie Delta area i s r e s t r i c t e d primarily to non-sorted types. Although other factors may be involved, the absence of the sorted forms can largely be attributed to the fact that the mantle frequently lacks a s u f f i c i e n t concentration of stones to exhibit marked f r o s t 40 sorting. This discussion of patterned ground features on Garry Island i s r e s t r i c t e d to a consideration of some aspects of the development of two of the non-sorted forms: tundra, or ice-wedge, polygons and earth hummocks.^ Tundra Polygons. Of the non-sorted forms, tundra or ice-wedge polygons con-s t i t u t e one of the most widespread types of patterned ground i n the Mackenzie Delta area. Readily discernible on a e r i a l photographs, the ground exhibits a polygonal microrelief pattern formed by the intersection of shallow furrows underlain by ground ice-wedges. Le f f i n g w e l l , working on the coastal p l a i n of North Alaska, was among the f i r s t to postulate that the networks of tundra polygons were generated by contraction cracks i n the frozen ground, produced by intense stresses created as a re s u l t of 39 Washburn, A.L. (1956) " C l a s s i f i c a t i o n of Patterned Ground and review of suggested o r i g i n s " , Geol. Soc. Amer. B u l l . , Vol. 67, p. 284. 40 Mackay, J. Ross (1963) "The Mackenzie Delta area, N.W.T.", Geographical Branch Memoir, No. 8, p. 69. ^A few crudely-sorted stone c i r c l e s were discovered on the floo r s of a number of shallow, a r t i f i c i a l l y drained lakes on the island. See Mackay, J . Ross (1967) " Underwater patterned ground i n a r t i f i c i a l l y drained lakes, Garry Island, N.W.T.", Geographical B u l l e t i n , Vol. 9, pp. 33-44. 205 pronounced seasonal changes i n the ground temperature. His 'thermal contraction' theory was outlined as follows: "The permanently frozen ground contracts i n the cold A r c t i c winter and cracks are formed which divide the surface into polygonal blocks. In the spring these f r o s t cracks become f i l l e d with surface water which immediately freezes. In the expansion of the frozen ground as the temperature ri s e s i n summer, the vein of ice becomes more r i g i d than the country formation, and the readjustment takes place i n the l a t t e r . The res u l t i s to bulge up the inclosed block either bodily or else l o c a l l y along the sides of the ic e . During the next winter's cold wave a new crack forms at the same locus so that a continually growing wedge of ground ice i s formed. Thus the tundra becomes underlain by a network of ice-wedges, which inclose bodies of the o r i g i n a l formation',.'.^-} The general principles of L e f f i n g w e l l 1 s contraction theory have been accepted by most of the subsequent research workers inve s t i g -ating tundra polygons. Despite the voluminous l i t e r a t u r e on th i s subject, however, the precise details of their o r i g i n are s t i l l imperfectly under-stood. Black and Lachenbruch attr i b u t e some of th i s ignorance to an absence of quantitative, rather than q u a l i t a t i v e , data but i t may also r e f l e c t the paucity of observations describing the i n i t i a l development of 44 the f r o s t crack patterns. ^ L e f f i n g w e l l , E. de K. (1915) "Ground Ice-wedges. The Dominant form of Ground Ice on the North Coast of Alaska", Journal of  Geology, Vol. 23, pp. 635-654. Leffin g w e l l , E. de K. (1919) "The Canning River Region, Northern Alaska",. U.S.G.S. Professional Paper, No. 109, 251 p. 43 4 J L e f f i n g w e l l , E. de K. (1915) op_. c i t . , p. 654. ^ B l a c k , R.F. (1952) "Polygonal patterns and ground conditions from a e r i a l photographs", Photogrammetric Engineering, Vol. 18, p. 124. Lachenbruch, A.H. (1962) "Mechanics of Thermal Contraction Cracks and Ice-Wedge Polygons i n Permafrost", Geol. Soc. Amer. Special  Paper, No. 70, p. 5. 206 Incipient Frost Crack Patterns. An examination of the l i t e r -ature pertaining to tundra polygons reveals abundant references to, and descriptions of, the polygonal ground i n r e l a t i v e l y advanced stages of development, but surprisingly l i t t l e r e l a t i n g to the formation of the i n -i t i a l f r o s t crack patterns. Leffingwell shows i l l u s t r a t i o n s of i n c i p i e n t 45 cracks on the coastal p l a i n of north Alaska, and Black also makes brief 46 reference to similar features. Lachenbruch has made a theoretical study of f r o s t crack patterns i n his examination of L e f f i n g w e l l 1 s thermal con-traction hypothesis from the point of view of mechanics, but he does not 47 c i t e any f i e l d evidence to corroborate his conclusions. Washburn et a l . , to the writer's knowledge, have produced the only recent paper describing 48 the occurrence of f r o s t cracks, a l b e i t i n a non-arctic environment. Observations i n the Mackenzie Delta area during the summers of 1964 and 1965 revealed three locations where in c i p i e n t f r o s t cracks had developed on the ground surface. The locations, each characterized by an absence of any major r e l i e f features and elevations of less than one metre (3 feet) above mean sea l e v e l , included lake-strewn, a l l u v i a l f l a t s on Kendall and Grassy Islands and the f l a t s bordering the lagpons enclosed behind sandspits on Garry Island. A l l of these sit e s are frequently inun-dated during periods of high water, especially during storm surges. The fr o s t cracks were found to be equally w e l l developed on bare ground 4 5 L e f f i n g w e l l , E. de K. (1919) 0 £ . c i t . , 4 6 B l a c k , R.F. (1953) op_. c i t . , p. 130. ^Lachenbruch, A.H. (1962) op_. c i t . 4 8Washburn, A.L., Smith, D.D. and Goddard, R.H. (1963) "Frost Cracking i n a Middle-Latitude Climate", Biuletyn Peryglacjalny, Nr. 12, pp. 175-189. 207 surfaces (Plate XIII-A), and on f l a t s supporting a dense cover of grasses and sedges 20-40 cms. (8.0-15-5 ins.) t a l l (Plate XIII-B). In these l a t t e r areas, the vegetation was flattened, presumably by a combination of prevailing winds, snowfall and flood surges i n the preceding f a l l . The vegetation was cut by sharp, k n i f e - l i k e fractures and was probably frozen to the ground surface at the time of the f r o s t cracking to produce the clean break. In many places, the cracks were observed to extend beneath the surfaces of shallow lakes, though none were traced e n t i r e l y across the lake f l o o r . This indicates that the water, at least i n the shallower parts of the lakes, was frozen to the bottom although water i n the central sections of the lakes may have remained unfrozen at depth. The fact that the fissures on Kendall Island cut through the vegetation into the under-lyi n g mineral s o i l , would appear to substantiate the view that the cracks were produced by intensive f r o s t action, The cracks on the bare ground surfaces are interpreted as having originated i n a similar manner, although the p o s s i b i l i t y that they were produced by desiccation of the s o i l cannot d e f i n i t e l y be excluded. Excavations i n one of the sandspit-lagoon areas on Garry Island revealed that the f r o s t cracks were best developed i n organic-rich s i l t s and s i l t y loams. Mechanical analyses of the s o i l s showed that they were composed of 59.2 per cent s i l t , 21.3 per cent sand, and 19.5 per cent clay (Wentworth c l a s s i f i c a t i o n categories). The s o i l s were mantled by a thin layer of organic material, and the organic content of the s o i l samples averaged 5.5 per cent by weight of the dried sample. The s u r f i c -i a l organic layer and the organic matter at depth, often i n the form of thin i n t e r c a l a t i o n s , probably represent*washed peat derived through erosion of the adjacent coastal b l u f f s . The s o i l s also possessed a high Plate XIII I N C I P I E N T F R O S T C R A C K P A T T E R N S 209 moisture content, with frozen samples having an average ice content of 91.7 per cent (expressed as weight of ice to dry s o i l ) . Mackay has a l -ready documented the granulometric composition of the s o i l s underlying the sedge-covered f l a t s on Kendall Island, where the proportions of clay and sand were s l i g h t l y lower ( s i l t 79 per cent, sand 13 per cent and clay 8 v 49 per cent). Most of the f r o s t cracks exhibited l i t t l e or no topographic expression at the ground surface, but some of the larger fissures trav-ersing the unvegetated areas were marked by the presence of shallow troughs 30-70 cms. (12.0-27.5 ins.) across and 10-20 cms. (4-8 ins.) deep. When examined i n early August, 1964, the f r o s t cracks on Garry Island ex-hibited a wedge-like form extending down through the active layer and into the frozen ground at depths of 50 cms. (19.5 ins.) below the ground sur-face. At the surface, the open fissures were up to 4 cms. (1.5 ins.) across, and they remained open to depths of 15-25 cms. (6-10 ins.) but narrowed to only a few millimetres i n width at the l e v e l of the f r o s t table. In many l o c a l i t i e s , the cracks were i n f i l l e d at depth by sand-size material that had probably been blown or washed i n from the ground sur-face. These miniature sand-wedges are similar i n form to the larger scale features described by Pewe i n the McMurdo Sound area of Antarctica. Below the l e v e l of the f r o s t table, the cracks were occupied by small veins of i c e , approximately one millimetre across, which could be traced Mackay, J . Ross (1965) "Gas-domed mounds i n permafrost, Kendall Island, N.W.T.", Geographical B u l l e t i n , Vol. 7, p. 108. 5 0 ' ' Pewe, T.L. (1959) "Sand-Wedge Polygons (Tesselations) i n the McMurdo Sound Region, Antarctic -A Progress Report", Amer. Jour., S c i . , Vol.. 257, pp. 545-552. 210 to a depth of 76 cms. (30 ins.) below the ground surface. Rarely did the f r o s t cracks i n the sandspit-lagoon areas of Garry Island reveal any arrangement into a d e f i n i t e polygonal network. The majority of the fissures appeared to be randomly distributed over the ground surface, and showed no preferred d i r e c t i o n a l orientation except for a weak tendency to occur along lines developed at r i g h t angles and p a r a l l e l to the margins of the lagoons. The only other salient feature of the d i s -t r i b u t i o n was the contrast i n the density of the pattern between the s i l t y loam and coarser sand areas of the sandspit. Although the f r o s t cracks were present i n both of these areas, the density was much higher i n the s i l t y loam sections. The d i s t r i b u t i o n of the f r o s t cracks on Kendall and Grassy Islands showed a much greater tendency to be organized into crude polygon-a l patterns. Figure 30 i s a map showing the s p a t i a l arrangement of the f r o s t cracks on a part of the sedge-covered f l a t on Kendall Island. As t h i s diagram shows, the ground surface was subdivided into a number of highly ir r e g u l a r polygons of variable size, but averaging 2-3 metres (6.5-10.0 feet) across. The majority of these irregularly-shaped polygons were four- or five-sided, and hexagonal forms were notably conspicuous by their absence. Most of the fissures on Kendall Island also exhibited l i t t l e preferred d i r e c t i o n a l orientation except i n the v i c i n i t y of the larger water bodies. The s p a t i a l d i s t r i b u t i o n of the f r o s t cracks on Grassy Island demonstrated the existence of a much more regular polygonal network, a l -though on a considerably larger scale. On these bare a l l u v i a l f l a t s the indi v i d u a l polygons averaged 20-30 metres (65-100 feet) across, and t e t -ragonal forms were predominant. The pattern of these f r o s t cracks Figure 30 INC IP I ENT FROST C R A C K P A T T E R N K E N D A L L I S LAND , N.W.T. 211 212 moreover revealed much stronger trends i n their preferred d i r e c t i o n a l orientation. The larger f i s s u r e s , up to 5 cms. (2 ins.) wide and located i n the f l o o r s of shallow troughs i n the ground surface, were oriented at rig h t angles to the bank of a distributary of the Mackenzie River. The smaller cracks, less open and having almost no topographic expression on the ground surface, on the other hand were aligned more or less p a r a l l e l to the same r i v e r bank. These preferred orientations were especially noticeable within distances of approximately 50-60 metres (165-200 feet) -~ from the edge of the channel, but became less d i s t i n c t with increasing dis-tance from the bank. I t i s generally agreed that f r o s t cracks originate as a r e s u l t of large thermal stresses created by a sudden cooling of the ground. "When the t e n s i l e strength (of the ground) i s ex-ceeded near the surface, a tension crack forms and propogates downward. ... The formation of a crack causes a l o c a l r e l i e f of tension i n the s u r f i c i a l materials. ... Each crack i s , therefore, surrounded by a band i n which cracking has caused appreciable reduction of horizontal tension - the "zone of stress r e l i e f " . ... The component of thermal tension at the ground surface i n the d i r e c t i o n p a r a l l e l to the crack i s relieved only s l i g h t l y , by the cracking and, thus, large horizontal stress differences occur within the zone of stress r e l i e f . A second crack entering t h i s zone tends to a l i g n i t s e l f perpendicular to the d i r e c t i o n of greatest tension, and, hence, tends to intersect the f i r s t crack at r i g h t angles. Conversely, the occurrence of an orthogonal intersection generally implies that one of the cracks predated the other". Lachenbruch's conclusion that the angular intersections of a polygonal network of f r o s t cracks w i l l exhibit a preferred tendency toward an orthogonal pattern, contrasts with many descriptions of polygonal "lachenbruch, A.H. (1962) op_. c i t . , pp. 57-58. 213 ground i n which authors have expressed a tendency for hexagonal forms and 52 angular intersections of 120° to predominate. The implications of the hexagonal pattern, and angular intersections of 120°, are that the f r o s t cracks originated at a series of points and each crack developed more or less simultaneously. In an attempt to determine the v a l i d i t y of Lachen-bruch' s conclusion, particular attention was paid to the nature of the angular intersections of the f r o s t crack patterns on Garry and adjacent islands. A t o t a l of 101 angular measurements was recorded including those shown i n Figure 30. An additional f i f t y intersection patterns are. i l l u s -trated diagrammatically i n Figure 31, where an attempt has also been made to indicate the r e l a t i v e order of occurrence and propogational dire c t i o n of each of the f i s s u r e s . Primary f r o s t cracks, usually the larger, are defined as those which originated f i r s t at any location, and their prop-ogational directions were inferred, wherever possible, from thei r orientation with respect to the water bodies ( i . e . cracks which were oriented at r i g h t angles to, and propogated outward from, the body of water). Secondary f r o s t cracks are defined as those which developed later at each location, and these cracks terminate at, and propogate towards, pre-existing primary f i s s u r e s . Of the 101 angular measurements recorded, no fewer than 79, or eighty per cent, were of the orthogonal type. As Figures 30 and 31 indicate, most of these orthogonal intersections were formed by the For example: L e f f i n g w e l l , E. de K. (1915) op_. c i t . , p. 638. Black, R.F. (1953) op_. c i t . , p. 129. Conrad, V. (1946) "Polygon Nets and t h e i r Physical Development", Amer. Jour. S c i . , V ol. 244, pp. 277-296. 214 Figure 31 D I A G R A M M A T I C S K E T C H E S O F F R O S T C R A C K I N T E R S E C T I O N P A T T E R N S ILL. T I I P r i m a r y Frost C racks • Secondary Frost C racks • O r t h o g o n a l Intersections 60 Ang le of Intersect ion in Degrees —»- Inferred Pro p o g a t ional D i r e c t i ons of C r a c k s 215 junction of a primary and secondary f r o s t crack, and only rarely were two primary f r o s t cracks observed to intersect one another. The influence of a zone of stress r e l i e f i s also manifest i n the manner i n which many of the secondary fissures curve to intersect the primary cracks at r i g h t angles. Where the primary cracks were sinuous, the most favoured l o c i for the intersection points of secondary f r o s t cracks were located on the convex sides of the curves. This i s i n accordance with the d i s t r i b u t i o n a l pattern of stress r e l i e f on a curved section of a f r o s t crack as described by 53 Lachenbruch. Angular intersection: values of 60° were by f a r the most common of the non-orthogonal intersection patterns. This angle was most f r e -quently developed as a resu l t of the bifurca t i o n of a primary f r o s t crack, and at points of intersection where two f r o s t cracks of the same order approached one another obliquely. Very few of the angles measured were neither 90° nor 60°; and only two examples were found of t r i - r a d i a l i n t e r -sections, forming three obtuse angles of about 120°, suggesting that the fr o s t cracks originated at a point (Plate XIII-C). The f i e l d evidence collected i n the outer Mackenzie Delta area thus appears to substantiate the conclusions of Lachenbruch's the o r e t i c a l study. Primary f r o s t cracks were developed, essent i a l l y i n a random pattern, across the ground surface, and the junctions of secondary f r o s t cracks with these primary fissures showed a d e f i n i t e preferred tendency toward an orthogonal intersection pattern. According to Lachenbruch 1s c l a s s i f i c a t i o n scheme, the resultant crude polygonal network would Lachenbruch, A.H. (1962) op_. c i t . , p. 50. 216 54 therefore be c l a s s i f i e d as a 'random orthogonal system'. Only i n the v i c i n i t y of large bodies of water did preferred d i r e c t i o n a l orientations become s u f f i c i e n t l y pronounced to be c l a s s i f i e d as 'oriented orthogonal systems'. The apparent dichotomy between th i s evidence and the fact that most tundra polygons appear to be of a non-orthogonal type has been ex-plained by Lachenbruch to be the re s u l t of an obscuring of the intersection angles by the growth of large ice-wedges."^ The incip i e n t f r o s t crack pattern i s p r a c t i c a l l y the only stage i n the development of a network of polygonal ground therefore i n which the angular intersection patterns can be determined with any r e a l degree of accuracy. Moreover, these determin-ations can only be made i n the f i e l d through ground inspection, since the frost cracks are generally too small to be i d e n t i f i e d from a e r i a l photo-graphs. The transformation of an i n i t i a l pattern of f r o s t cracks into a network of tundra or ice-wedge polygons requires that recurrent fra c -turing takes place at the same l o c i . The evidence collected on Garry Island suggests that once a fracture i s formed, i t tends to persist as a permanent l i n e of weakness i n the mantle. Most of the larger cracks re-mained as open fissures throughout the summer months, possibly indicating that the ground was s u f f i c i e n t l y e l a s t i c to absorb the s t r a i n produced by i t s expansion under the summer's heat and consequently no deformation took 5 6 place. Where the ground surface was covered by a thin layer of washed 'Lachenbruch, A.H. (1962) op. c i t . , p. 46. 'Ibid. , Figure 13, p. 49. 'Leffingwell, E. de K. (1915) op_. c i t . , p. 640. 217 peat, i t i s also possible that the maintenance of the open f i s s u r e may have been aided l o c a l l y by a s l i g h t desiccation and shrinkage of t h i s layer. Even where the cracks were, closed, t h i s probably did not take place before some material had i n f i l t r a t e d from the ground surface, and the miniature sand-wedges, produced i n t h i s manner, also a s s i s t i n the preser-vation of the lines of weakness. Low-Centred Polygons. The f r o s t cracks extend down below the base of the active layer where they are occupied by a thi n vein of i c e . Recurrent fracturing at the same l o c i results i n the addition of success-ive increments of i c e , and the eventual formation of a large, f o l i a t e d ice-wedge. Accompanying the growth of these ice-wedges, and bounding them on either side, d i s t i n c t i v e ridges are formed which may p a r t i a l l y represent the upturning of strata adjacent to the growing w e d g e , T h u s the f i r s t polygons to form consist of a c e n t r a l , saucer-shaped depression enclosed between bordering ridges, and they are accordingly referred to as low-centred or raised-edge polygons. These polygons are best developed i n fine-grained sediments underlying poorly-drained f l a t s , but i f the drainage i s too poor they do not necessarily exhibit the c h a r a c t e r i s t i c saucer-shaped form. A number of polygons has developed around the edges of a lagoon enclosed by the sandspit on the north-central coast of Garry Island. This lagoon i s open to the sea, and the areas of polygonal ground are frequently inundated during periods of high water. Such polygons have very l i t t l e topographic "^Pewe, T„L. (1966) "Ice-Wedges i n Alaska - C l a s s i f i c a t i o n , D i s t r i b u t i o n , and Climatic Significance", Proc: Permafrost International  Conference, Lafayette, Indiana, November, 1963. Nat., Acad, of Sciences -Nat. Research Council, Washington Publication, No. 1287, p. 77. 218 expression and bordering ridges are non-existent. Figure 32 shows the surface contours and cross-sectional pro-f i l e s of a c l a s s i c a l low-centred polygon form, located at an elevation of 4.5-6.0 metres (15-20 feet) above sea l e v e l . The polygon i s surrounded by shallow troughs, marking the positions of the ice-wedges, ranging i n width from 3.0 to 4.5 metres (10-15 f e e t ) . In places, the troughs are occupied by water 0.5-1.0 metres (1.5-3.0 feet) deep (Plate XIV-A), but elsewhere they have been p a r t i a l l y i n f i l l e d by the accumulation of organic material which supports a vegetation cover composed predominantly of sedges. The polygon, ranges i n width from 10.5 to 14.5 metres (35-48 feet) as measured between the crests of the bounding ridges which reach maximum elevations of 65 cms. (25.5 ins.) above the l e v e l of the water i n the adjacent troughs. These same ridges enclose a shallow depression i n the central part of the polygon which, at the time of surveying, was floored by a marsh-sedge vegetation, but at other times during the summer was occupied by a shallow pool of standing water 5-10 cms. (2-4 ins.) deep. Central areas of other low-centred polygons i n the same general location were occupied by pools of water throughout the summer. The retention of t h i s water i s aided by the configuration of the f r o s t table which closely follows the surface microrelief as i l l u s t r a t e d i n the p r o f i l e i n Figure 32. The polygon shown i n Figure 32 was actually part of an exten-sive area of low-centred polygons located between the large lake and the truncated lagoon on the northwest coast of the island (see Figure 1). A larger grouping of these polygons i s shown i n Figure 33. As this map i l l u s t r a t e s , the most ch a r a c t e r i s t i c form of the unit polygons was a four-or five-sided figure. The sizes of the i n d i v i d u a l polygons ranged from Figure 32 S U R F A C E C O N T O U R S A N D C R O S S - S E C T I O N A L P R O F I L OF A T Y P I C A L L O W - C E N T R E D P O L Y G O N C R O S S - S E C T I O N A L P R O F I L E Standing water Standing water - ice f loor - muck floor Distance in metres 220 P l a t e XIV L O W - C E N T R E D P O L Y G O N S A. Low-centred or r a i s e d -edge t u n d r a p o l y g o n w i t h c e n t r a l d e p r e s s i o n o c c u p i e d by p o o l of s t a n d i n g w a t e r. B. I n f i l l i n g of c e n t r a l d e p r e s s i o n t o produce a f l a t - t o p p e d p o l y g o n which has been s u b d i v i d e d by the growth of a d d i t i o n a l i c e -wedges . C. O r i e n t e d , o r t h o g o n a l system of l o w - c e n t r e d p o l y g o n s d e v e l o p e d around a s m a l l l a k e . ( P h o t o g r a p h c o u r t e s y of Dr. J . Ross Mackay). Figure 33 222 9 to almost 24 metres (30-80 feet) across, but much of th i s v a r i a t i o n can be attributed toi the subdivision of the o r i g i n a l , larger polygonal units by the subsequent growth of additional ice-wedges (Plate XIV-B). I t i s v i r t u a l l y impossible from Figure 32 to determine the angular intersections of the polygonal network. Lines drawn along the central axes of the ice-wedge troughs could be interpreted to intersect orthogonally, as i n the case of the incipient f r o s t crack patterns, but d e f i n i t i v e conclusions cannot be made. In a few locations deep, open fissures i n the ice-wedges, presumably marking the positions of f r o s t cracks produced during the preceding winter, could be observed beneath the water, but, unfortunately, none of these were located at the ice-wedge intersection points. S i m i l a r l y , the presence of any preferred orientation in the alignment of the ice-wedges shown i n Figure 33 i s d i f f i c u l t to de-tect. A group of polygons around a small lake i n the same area of patterned ground showed a de f i n i t e orientation with respect to the lake, with the major fissures radiating outward from the lake l i k e the spokes of a wheel (Plate XIV-C). I t i s surprising that the polygonal network shows no pronounced orientation with the shore of the lake shown i n Figure 33. If any pattern i s discernible, i t i s perhaps suggestive of a crude r a d i a l and concentric arrangement around the fo c a l points A and B. I f th i s i n -terpretation i s correct, these f o c a l points may represent the former positions of two small lakes on the f l a t ground around each of which an oriented orthogonal(?) system was developed. Ice-wedge or tundra polygons are not r e s t r i c t e d i n their de-velopment to areas of moist, fine-grained sediments. Large networks of subdued polygons were also developed on Garry Island on the surfaces of the sand headlands (Figure 34). Although these polygons were also of the 223 Figure 34 LOW-CENTRED POLYGON NETWORK ON SAND HEADLANDS 224 low-centred type, there was a major contrast i n the surface microrelief. Thus the bounding ridges of the polygons barely rose more than 15-25 cms. (6-10 ins.) above the levels of the troughs or the central sections of each polygonal uni t . The troughs, marking the positions of the ice -wedges, contained no surface water and the ice was mantled by a layer of moss and sedge growing i n a peat substrate. Only near the edges of the c l i f f s did the troughs become more pronounced, where thawing of the ice -wedges had produced prominent 'V'-shaped notches, 1-3 metres (3.5-10.0 feet) deep, i n the c l i f f top. The polygons on the sand headland surface also exhibited a wide var i a t i o n i n si z e , r e f l e c t i n g a similar subdivision process to that described i n an e a r l i e r paragraph. The configuration of the more prominent surface expressions of the ice-wedges suggests that the headland surface was i n i t i a l l y divided into a series of irregular blocks, 25-35 metres (80-115 feet) across, and that these have subsequently been subdivided by the growth of additional ice-wedges into a number of pre-dominantly tetragonal forms, 6-10 metres (19.5-33.0 feet) across. The greater regularity and uniform widths of the troughs outlined i n Figure 34 make i t easier to infer the nature of their angles of intersection, which appear to exhibit a d e f i n i t e tendency toward a preferred orthogonal pattern. Thus the main effects of a coarse sand substrate on the net-work of patterned ground appear to be i n a s l i g h t l y coarser spacing of the primary crack patterns, and a less pronounced topographic expression. The wider spacing of the primary f r o s t cracks may r e f l e c t the lower c o e f f i c i e n t of expansion of frozen sand compared to fine-grained sediments, and the subdued microrelief may be related to a more r e s t r i c t e d moisture supply. 225 High-Centred Polygons. The major sequential stages i n the i n -f i l l i n g of a low-centred polygon, and i t s transformation into a high-centred form, were outlined i n Chapter I I I . Standing pools of water i n the central depressions of low-centred polygons are often floored by a thick layer of soft ooze, believed to be the accumulated remains of algae. The ooze deposit corresponds to the grey s i l t y material observed i n the basal sections of high-centred polygons exposed by wave action i n the coastal b l u f f s . The continuity of this deposit i n the same exposures also i n d i c -ates that the.accumulation of a l g a l remains i s a major factor i n the i n f i l l i n g of i n i t i a l lakes prior to the development of a system of i c e -wedges. The deposition of the ooze within the central areas of low-centred polygons leads to a shallowing of the water depths, and allows the areas to be colonized by vascular species i n the sequence also described i n Chapter I I I . The accumulation of peat results i n a further elevation of the.central, areas of the polygons r e l a t i v e to the levels of the bordering ridges ^and, since the grey s i l t layers often have ice contents of several hundred per cent (by weight), the formation of syngenetic (penecontempor-aneous), segregated ice lenses i n the ooze may also contribute s i g n i f -icantly to an elevation of the surfaces of the central depressions. The sequential stages i n t h i s shallowing process.were readily observable i n the group of low-centred polygons shown i n Figure 33, where, in general, the surfaces of the polygons located around the edge of the f l a t s were l e v e l with the bounding ridges. Nowhere, however, had t h i s i n -f i l l i n g process led to the production of the c h a r a c t e r i s t i c , dome-shaped, convex p r o f i l e of a t y p i c a l high-centred polygon (Plate XVVA). Figure 35 shows the surface contours and cross-sectional p r o f i l e of one of these high-centred forms, located at an elevation of approximately 21-23 metres 226 Plate XV H I G H - C E N T R E D P O L Y G O N A N D T H E R M O K A R S T F E A T U R E S A. C h a r a c t e r i s t i c , dome-shaped, convex p r o f i l e of a t y p i c a l high-centred polygon. B. Thermokarst features i n an area of high-centred polygons exposed by coastal recession. 227 Figure 35 S U R F A C E C O N T O U R S A N D C R O S S - S E C T I O N A L P R O F I L E O F A T Y P I C A L H I G H - C E N T R E D P O L Y G O N S U R F A C E C O N T O U R S A B O V E D A T U M (Cms.) A C o n t o u r i n t e r va l 15 cms Wa te r W i l l o w and sedge S e d g e C R O S S - S E C T I O N A L P R O F I L E Water f i l l ed wed ge F la t , dry and peaty hummocks with narrow l i t ter f i l l e d d e p r e s s i o n s Dry wedge w i l l ow and sedge Distance in metres 228 (70-75 feet) above sea l e v e l . Compared to thei r low-centred counterparts, these polygons were generally much more regular i n shape and averaged 6-8 metres (19.5-26.0 feet) i n diameter. At thei r highest points, usually i n the centres of the polygons, the surfaces were about 1.5-2.0 metres (5.0-6.5 feet) above the l e v e l of the surrounding troughs. The l a t t e r , marking the positions of the ice-wedges, were predominantly f i l l e d with moss and sedge although l o c a l l y they contained pools of water 15-50 cms. (6.0-19.5 ins.) deep. Most of the areas of tundra polygons on Garry Island at ele-vations exceeding approximately 15 metres (50 feet) above sea l e v e l are of the high-centred type. The height-distribution pattern of these areas appears to be intimately related to the elevations of raised shoreline features. I t i s postulated that most of the polygonal pattern was i n -i t i a l l y formed on lagoon-flats developed behind small sandspits, or bars, associated with former positions of the l e v e l of the sea. Many of these areas of tundra polygons are extremely limited i n their extent, and they t y p i c a l l y take the form of small pockets occupying the narrow f l o o r s of some of the va l l e y s . In these r e s t r i c t e d l o c a l i t i e s , however, the con-figurations of the networks of ice-wedges exhibit a very d e f i n i t e tendency toward the development of an oriented orthogonal system. Thus the most prominent wedge lines were invariably aligned along the axis of the valley f l o o r , where undoubtedly much of the reason for their prominence was their accentuation by intermittent surface runoff, and the other ice-wedges were frequently aligned at righ t angles to these main axes. Thermokarst Features. The term thermokarst i s used to describe surface hollows and depressions that originate through the melting of ground i c e . Because of the high ice content of many of the sediments, the 229 conditions on Garry Island are extremely favourable for the development of thermokarst topography, especially when these sediments are exposed by the recession of the coastline. Thus the most dramatic manifestations of t h i s process occur when the retreat of the c l i f f s exposes massive sheets of segregated ice r e s u l t i n g i n the development of the large mudslumps des-cribed e a r l i e r i n t h i s chapter. The shallow water depths, usually 1-2 metres (3.0-6.5 f e e t ) , around the margins of the larger lakes on the i s -land, and the concentrations of coarse gravel and boulders on their f l o o r s , may possibly be interpreted as evidence of thermokarst enlargement of the 58 lakes i n a manner similar to that described by Wallace i n eastern Alaska. A d i s t i n c t i v e type of thermokarst topography was developed on Garry Island where networks of tundra polygons, of both the low- and high-centred types, had been exposed i n the b l u f f s as a res u l t of coastal recession (Plates XV-B and XV-C). Melting along the lines of the i c e -wedges had transformed the shallow troughs into prominent trenches several metres deep. In the sand headland areas, the melting of the ice-wedges imparted a ch a r a c t e r i s t i c notched appearance to the c l i f f s . In the area of high-centred polygons, located on the northwest coast of the island (see Figure 6), the trenches were as much as 3-4 metres (10-13 feet) deep, and the indi v i d u a l polygonal units had been l e f t standing as isolated mounds of peat. Melting of the ice-wedges had also l e f t the sides of the polygons unsupported, and a considerable amount of slumping had taken place thereby accentuating the convexity of their surface p r o f i l e s . Since the majority of the areas of high-centred polygons on the island also 5 8Wallace, R.E. (1948) "Gave-in lakes i n the Nabesna, Chisana and Tanana r i v e r v a l l e y s , eastern Alaska",. Journal of Geology, Vol. 56, pp. 171-181. 230 appear to have been affected by thermokarst action, though usually to a lesser degree, i t seems reasonable to infer that the development of the t y p i c a l convex p r o f i l e of these polygons may represent the operation of this same slumping process. I t should be noted that the examples of thermokarst topography cit e d above were, or could have been, developed en t i r e l y by the operation of e x i s t i n g natural processes without any recourse to an amelioration of climate. Three further examples, observed during the f i e l d seasons, dem-onstrated the r a p i d i t y with which thermokarst features can be developed by an interference with the natural conditions. The f i r s t example was associated with attempts to lower a r t i f i c i a l l y the l e v e l of a lake by digging a ditch across the peat barrier damming, i t s outlet. The water was led out through t h i s a r t i f i c i a l channel, at an elevation of approximately 36.5 metres (120 feet) above sea l e v e l , and then allowed to seek i t s own course to lower elevations. Immediately prior to the opening of the channel, the depth of thaw just beyond i t s outlet was recorded as 30-35 cms. (12-14 i n s . ) , yet after only three weeks of intermittent flow t h i s had increased f i v e f o l d , to over 1.5 metres (5 f e e t ) , and the surface mat of l i v i n g vegetation was l i t e r a l l y f l o a t i n g on the thawed substrate. Without flooding, the position of the f r o s t table might have retreated by approximately 10 cms. (4 ins.) at the most during the same three-week period. The second example was much more impressive i n terms of i t s effect on the surface topography. As discussed e a r l i e r i n Chapter I I , the majority of the stream courses on Garry Island have i l l - d e f i n e d channels, and, consequently, the flow of water i s easily diverted from i t s normal path. An example of t h i s occurred when water was diverted from i t s 231 o r i g i n a l course to follow well-defined, sub-parallel paths between the base camp and the coast. At the coastline, t h i s path terminated at a series of tundra polygons, the existence of which was only feebly expressed at the ground surface. Surface runoff, derived p r i n c i p a l l y from melting snow, was observed to be flowing along these paths at the beginning of the 1965 f i e l d season. In the summers of 1967 and 1968 there was s l i g h t seepage down the paths, but no flow. The effects of th i s re-routing of the water were apparent by the summer of 1967, where rapid thermokarst erosion of the i c e -wedges had occurred over distances extending back to about 30-35 metres (100-115 feet) from the coastal b l u f f . This erosion had produced a number of prominent troughs, approximately 2 metres (6.5 feet) wide and, at the 59 intersection points of the ice-wedges, 2 metres (6.5 feet) deep. By 1968, the inter-ice-wedge areas of several polygons had been undermined by thaw to produce subsurface overhangs as much as 5 metres (16.5 feet) square 60 and 0.6-1.0 metres (2-3 feet) high. The t h i r d example of the development of thermokarst features occurred i n an area where a dog had been tethered during the 1965 f i e l d season. The vegetation cover of a c i r c u l a r area, approximately 3 metres (10 feet) i n diameter, was k i l l e d as a r e s u l t , and the accelerated thawing had led to a s e t t l i n g of the ground surface and the creation of a bowl-61 shaped depression, 20 cms. (8 ins.) deep, by the summer of 1968. 59 Information supplied by Dr. J. Ross Mackay, personal commun-ic a t i o n , August 27, 1968. 60 Information supplied by Dr.. J . Ross Mackay, personal commun-i c a t i o n , October, 1968. Ibid . 232 Earth Hummocks. The term earth hummock was f i r s t introduced by Sharp to de-scribe low, rounded knobs of f i n e material consisting of an earthen core 62 covered by a tight mat of moss, grass and scrubby plants. According to Washburn's c l a s s i f i c a t i o n scheme, they f a l l into the category of non-sorted 63 nets. On Garry Island, these microrelief features are generally confined to well-drained sit e s where they form a continuous, three-dimensional mesh on the gently-inclined upland surfaces and the steeper valley-side slopes. Although, as w i l l be discussed below, i t i s possible that similar micro-r e l i e f forms may be produced i n d i f f e r i n g environments by the operation of diss i m i l a r processes, this situation contrasts markedly with that described by Raup i n the Mesters Vig D i s t r i c t of northeast Greenland. Raup found that (turf) hummocks, which from his descriptions and i l l u s t r a t i o n s appear to be i d e n t i c a l to the earth hummocks on Garry Island, occurred only on sites abundantly supplied with gently-flowing surface water, derived from the melting of perennial snowdrifts or the thawing of frozen ground, 64 throughout most of the summer season. Perennial snowdrifts are lacking on Garry Island, and no incidence of surface runoff was observed on the hummock sites during three summers' observations. Moreover, earth hummocks were almost conspicuous by the i r absence on the lower, poorly-drained f l a t s °^Sharp, R.PC (1942) " S o i l structures i n the St. E l i a s Range, Yukon Te r r i t o r y " , Journal of Geomorphology, Vol. 5, pp. 282-283. 6 3Washburn, A.L. (1956) " C l a s s i f i c a t i o n of patterned ground and review of suggested or i g i n s " , Geol. Soc. Amer. B u l l . , Vol. 67, p. 830. 64 Raup, H.M. (1965) "The structure and development of turf hummocks i n the Mesters Vig D i s t r i c t , Northeast Greenland", Meddelseer om  Grinland, Bd. 166, Nr. 3, p. 5. Raup further states that turf hummocks may be common on drier s i t e s , but here they are always i n some stage of disintegration. 233 where they are usually replaced by the tussock forms of the Sheathed Cotton-grass (Eriophorum vaginatum) or featureless mats of moss and sedge. The apparent dichotomy between these environmental conditions i l l u s t r a t e s one of the most salient aspects of patterned ground studies. Hummocky ground, for example, has been extensively described i n both geo-morphological and botanical literature,, where numerous names have been applied to s u p e r f i c i a l l y similar forms. Anatomically, however, i t i s evident that the broad term 'hummocky ground1 has been used loosely to i n -corporate i n d i v i d u a l forms with widely d i f f e r i n g structures. Whilst i t i s possible that these d i s s i m i l a r structures may be nothing more than d i f f e r -ent expressions of the same, or s i m i l a r , genetic processes, i t i s equally probable that they each represent important modifications of these i processes i n response to s i g n i f i c a n t variations i n such s p e c i f i c environ-mental conditions as s o i l s , vegetation, slope and available moisture supply. Several authors have recognized that types of hummocky ground can be c l a s s i f i e d either on the basis of the plant species of which they are composed, or on the i r i n t e r n a l composition. . A c l a s s i f i c a t i o n accord-ing to plant species affords distinguishing c r i t e r i a for those microrelief features which have a tussock form. Individual tussocks are developed by the upward growth of a single plant on a columnar base of i t s own dead leaves and roots, and the height of these tussocks i s further accentuated by d i f f e r e n t i a l f r o s t action and the doming-up of mineral s o i l beneath the plant. On Garry Island, the Sheathed Cotton-grass (Eriophorum vaginatum) i s the. major tussock-forming species, and i t thrives on f l a t to gently-sloping, poorly-drained surfaces. Since the structure and development of tussocks have been described i n d e t a i l by Hopkins and Sigafoos, they w i l l 234 not be discussed further h e r e . ^ A c l a s s i f i c a t i o n of hummocks on the basis of their i n t e r n a l composition i s much more complex, but i t would seem l o g i c a l to make a fundamental subdivision r e l a t i n g to the degree of homogeneity of the sub-strate. In t h i s way the hummocks which are composed en t i r e l y of peat could be d i f f e r e n t i a t e d from those which have a prominent mineral core. I t i s into t h i s l a t t e r category that the hummocks on Garry Island f a l l and, since the domed mineral core i s the most diagnostic feature of these microrelief forms, Sharp's nomenclature of earth hummocks has been adopted. The Structure of Earth Hummocks. Information pertaining to the structure of earth hummocks was collected using the following methods. Three locations were selected subjectively to i l l u s t r a t e diagnostic hummock features at sp e c i f i c locations on a t y p i c a l slope p r o f i l e , and a fourth location where the hummocks had coalesced into stripe forms. At each of these locations the surface configuration was obtained by superimposing a 1.8 metre (6 foot) square- grid horizontally above the ground surface, and., taking plumb readings to the surface at horizontal intervals of 10. cms.» (4 i n s . ) . This data was used to prepare contour maps with a 2.5 cm. (one inch) contour i n t e r v a l , rough copies of which were made at the time to permit f i e l d checks. On the slope p r o f i l e , the grid size was usually s u f f i c i e n t to show a group of hummocks, but where the stripe features were developed the process was repeated four times to cover a 3.6 metre (12 foot) Hopkins, D.M. and Sigafoos, R.S. (1951) "Frost action and vegetation patterns on Seward Peninsula, Alaska", U.S_.G.S_. Bulletins, No. 974-C, pp. 51-101. Hopkins, D.M. and Sigafoos, R.S. (1954) "Role of f r o s t thrusting i n the formation of tussocks",. Amer. ...Jour. S c i . , Vol. 252, pp. 55-59. 235 square plot. Before the grid was removed, the, dominant aspects of the vegetation cover were also mapped. P r o f i l e lines were selected to give representative longitudinal- and cross-sections of a t y p i c a l hummock i n each of the locations, and excavations to the f r o s t table i n late August provided the structural data. Figure 35 and Plate XVI-A show earth hummocks on the gently-inc l i n e d upper portions of a slope p r o f i l e . Individual hummocks rose about 12-14 cms. (4.5-5.5 ins.) above the mean l e v e l of the surrounding depressions, and they were approximately c i r c u l a r i n outline, averaging about 50-60 cms. (19.5-23.5 ins.) i n basal diameter. In p r o f i l e the hummocks were broadly convex, and generally they did not exhibit any marked asymmetry. The intervening depressions commonly took on the form of shallow, open troughs. There were quite pronounced differences between the. vegetation associations of the hummocks and the depressions as shown i n Figure 37. The higher, drier s i t e s of the hummock centres were dominated by the A r c t i c Avens (Dryas i n t e g r i f o l i a ) , A r c t i c Blueberry (Vaccinium uliginosum var. alpinum), small sedges (Carex sp.), lichens and dry moss pads ( c h i e f l y Dicranum sp.). The A r c t i c Blueberry was p a r t i c u l a r l y abundant on the sides of the hummocks and sometimes extended down into the flo o r s of the. depress-ions, but these were dominated by the A r c t i c White Bell-heather (Cassiope  tetragona) growing i n a damp substratum of mosses (c h i e f l y Aulacomnium and Sphagnum sp.). Small willows (Salix arctica) were occasionally found on the surface of the hummocks, but they were most abundant on the lower parts of the sides. There were also appreciable differences, between hummocks and depressions, i n the relationship of this vegetation cover to the underlying surface. The plants on the hummocks formed a t i g h t , i n t e r -236 Figure 36 EARTH HUMMOCK LOCATED ON UPPER PART OF SLOPE PROFILE S U R F A C E C O N T O U R S A B O V E D A T U M S T R U C T U R A L P R O F I L E S I I I 11 Living vegetat ion Peat t'Xvlvl Minera l soil SCALE 0 20 40 60 Cms. Contour interval 2.5 cms. r = = ^ ^ K = = B ^ H = = 3 I I B B 237 Figure 37 E A R T H H U M M O C K L O C A T E D O N U P P E R P A R T O F S L O P E P R O F I L E V E G E T A T I O N T Y P E S i — i — i Edge of Hummock V * " | L i c h e n / a v e n s / s e d g e a s s o c i a t i o n Ave n s Sedge spec ies [•:•:';:} Moss a n d / o r Hea ther fS3r3 W i l low B lueber ry p a t c h S c a l e in cms . 20 40 60 P l a t e XVI E A R T H H U M M O C K S all' •UBI :—, A. Earth hummock located on the upper part of a slope p r o f i l e showing the almost c i r c u l a r o u t l i n e . B. Earth hummock located on the middle part of a slope p r o f i l e showing the elongation i n a downslope d i r e c t i o n . C. Earth hummock located on the lower part of a slope p r o f i l e e x h i b i t i n g an almost p e r f e c t c i r c u l a r form. D. Structure of a t y p i c a l earth hummock showing the mineral core, the wedges of organic m a t e r i a l beneath the troughs, and an i n t e r -mittent buried organic layer near the top of the perma-f r o s t . 239 woven mat which was intimately bound to the ground below by root systems which extended down into the peat and, i n some cases, into the mineral core. Plants occupying :the depressions on the other hand tended to be very shallow rooted, and were only loosely anchored i n the moss which i n turn rested loosely on the underlying peat. Figure 36 also shows the structural features of these hummocks along the indicated p r o f i l e l i n e s . The l i v i n g turf mat, including a zone of brownish, loosely-packed organic material, averaged 4-8 cms. (1.5-3.0 ins.) i n thickness, but increased l o c a l l y to almost 15 cms. (6 ins.) i n some of the larger depressions. Beneath t h i s surface mat was a layer of t i g h t , black peat averaging 5-8 cms. (2^3 ins.) i n thickness i n the central mound areas, but thickening to form organic wedges penetrating down to the fro s t table beneath the depressions. I t i s perhaps s i g n i f i c a n t that the positions of these organic wedges did not always coincide exactly with those of the surface depressions. Moreover, these organic wedges f r e -quently did not extend down either v e r t i c a l l y , or with uniform width, to the f r o s t table, but were often decidedly convex i n a downslope d i r e c t i o n , tapering to thin s t r i p s towards the base of the active layer. Local thickenings of the same peat layer were also evident beneath some of the central mounds, often i n close association with minor depressions on the hummock surface. The central core of the hummocks consisted of a promin-ent dome of mineral s o i l . This core reached i t s maximum thickness, approximately 35 cms. (14 ins.) under the centres of the hummocks, but .. l a t e r a l l y i t was greatly attenuated and occasionally was absent completely beneath the depression areas. The basal, s t r u c t u r a l feature of these hummocks was a buried organic layer which occupied a position close to the fro s t table. This layer always attained i t s greatest prominence towards 240 the l a t e r a l margins of the hummocks where i t was nearly always observed to unite with the organic wedges. In longitudinal, or downslope, sections i t was best developed on the upslope sides of these same wedges. The basal organic layer reached i t s maximum thickness, 2-3 cms. (0.75-1.25 i n s . ) , i n these l a t e r a l positions, but i t thinned appreciably towards the centres of the hummocks where only thin traces were sometimes present. In some i n -stances the organic layer could be traced e n t i r e l y beneath the complete hummock structure, but i n others i t appeared to be absent i n the central a r e a s . ^ The material beneath the organic layer consisted of mineral s o i l similar i n composition to that forming the hummock cores. Figure 38 and Plate XVI-B show a t y p i c a l earth hummock located on the middle slope region of the topographic p r o f i l e . The hummock rose 15-20 cms. (6-8 ins.) above the surfaces of the depressions, but rather than being c i r c u l a r i n outline, i t exhibited a de f i n i t e elongation i n a downslope di r e c t i o n . Whilst the hummock was s t i l l broadly convex i n pro-f i l e , l ongitudinally there was a de f i n i t e asymmetry with the downhill face steeper than that on the u p h i l l side. The intervening depressions were i stronger features than those described on the upper slope, and were more continuous, narrower and had steeper sides. The str u c t u r a l diagrams i n Figure 38 show that t h i s hummock was similar i n composition to the one described on the upper slope, and the vegetation associations of the hummock and depression areas were also s i m i l a r , as i s indicated by the map Since the f r o s t table continues to recede into September, the position of the buried organic layer i s perhaps more aptly described as at, or just above, the base of the active layer. Additional excavations of these and other hummocks to depths below the f r o s t table i n late August demonstrated a greater extent of th i s layer, but confirmed that i n some i n -stances i t i s completely lacking beneath the centres of some hummocks. Figure 38 EARTH HUMMOCK LOCATED ON MIDDLE PART OF SLOPE PROFILE S U R F A C E C O N T O U R S A B O V E D A T U M S T R U C T U R A L P R O F I L E S | | | 11 Living vegetat ion Peat rlvMv] Minera l soil Contour interval 2.5 cms. SC At E 20 40 60 Cms. Figure 39 E A R T H H U M M O C K L O C A T E D O N M I D D L E PART O F S L O P E P R O F I L E V E G E T A T I O N T Y P E S i — i — i Edge of Hummock | " , | L ichen / a ve n s / sedg e a s s o c i a t i o n |."Q°,°| A v e n s Sedge spec [v.y.'vj Moss a n d / o r Heather W i l l ow B l u e b e r r y p a t c h Sca le in cms. 0 20 40 60 243 of vegetation types shown i n Figure 39. Figure 40 and Plate XVI-C show a large earth hummock located on the lower slope near the base of the topographic p r o f i l e . This hummock rose as much as 30-35 cms. (12-14 ins.) above the adjacent depressions, and i t was approximately c i r c u l a r i n outline, averaging 70-75 cms. (27,5-29.5 ins.) i n basal diameter. The structural and vegetation patterns (Figure 40, Plate XVI-D and Figure 41) of th i s hummock and the neighbouring depress-ions also followed the same pattern as that described higher up the slope, but a few notable exceptions are worthy of mention. The proportion of the hummock occupied by the mineral core was greater, and the buried organic layer reached thicknesses of 5-8 cms. (2^3 ins.) beneath the hummock areas. In the depressions, the.proportion of moss was much higher and they were also much damper than the depressions higher up the slope. Other small changes i n the vegetation types included the presence of the Narrow-leafed Labrador-tea (Ledum palustre ssp. decumbens) and Mountain Cranberry (Vaccinium v i t i s - i d a e a ) with the lichens/avens/sedge association; the i n -clusion of the Common Crowberry (Empetrum nigrum) and Mountain Cranberry with the sedge species; and the appearance of tussocks and the Glandular Birch (Betula glandulosa) which are largely absent on the higher slopes. Figure 42 shows another hummock p r o f i l e excavated on the lower slope close to the junction with the f l o o r of the depression. This hummock was also almost c i r c u l a r i n outline, but averaged almost 1.2 metres (4> feet) i n diameter. The height of the hummock, 20^25 cms. (8-10 ins.) was consid-erably less than the one shown i n Figure 40, although two large tussocks of the Sheathed Cotton-grass (Eriophorum vaginatum) rose several centimetres above i t s general surface. The continuity of the vegetation mat was broken by the growth of these tussocks, but even i n between i t was extremely t h i n . Figure 40 EARTH HUMMOCK LOCATED ON LOWER PART OF SLOPE PROFILE S U R F A C E C O N T O U R S A B O V E D A T U M S T R U C T U R A L P R O F I L E S Living vegetat ion Peat ESD M ineral soil Contour interval 2.5 cms. 245 Figure 41 E A R T H H U M M O C K L O C A T E D O N L O W E R P A R T O F S L O P E P R O F I L E V E G E T A T I O N T Y P E S •*•»•« A V V *•' — « 1 > *. <•' i — i — i Edge of Hummock or Tussock \i2*\ L i c h e n / o v e n s / s e d g e a s s o c i a t i o n wi th L a b r a d o r tea and C r a n b e r r y EZ3 A v e n s S e d g e spec ies with C r a n b e r r y and C r o w b e r r y tol^ S e d g e tussock [::*.Vvl M o s s a n d / o r H e a t h e r W i l l o w rilSill G r o u n d b i rch B l u e b e r r y p a t c h Scale in cms. 0 20 40 60 246 Figure 42 S T R U C T U R A L P R O F I L E OF A N E A R T H H U M M O C K L O C A T E D A T THE F O O T O F A S L O P E T u s s o c k s Sca le 0 20 40 60 cms. Figure 43 M U D - B O I L S U R F A C E C O N T O U R S A B O V E D A T U M 247 The underlying peat layer was only a few centimetres thick over the centre of the hummock, but i t thickened l a t e r a l l y and extended down i n t y p i c a l peat wedges beneath the adjacent depressions. The size of the mineral core i s also much greater than that shown i n Figure 40, and the mineral s o i l was by f a r the dominant constituent of the hummock. The buried or-ganic layer exhibited the same thickening towards the margins of the troughs, but i t was even thinner and less continuous beneath the centre of the hummock. The appearance of the tussocks r e f l e c t s an obvious change i n the plant composition of the vegetation cover, and i n the. depressions the build-up of a substantial layer of moss was more pronounced than at loca-tions higher up the slope. Figure 43 shows a large mud b o i l located on the f l o o r of the depression close to the junction with the slope above. The mud b o i l , t y p i c a l of many developed i n similar locations, was almost c i r c u l a r i n outline,, averaging 1.2-1.4 metres (4-5 feet) i n diameter, and i t s central area rose 25-35 cms. (10-14' ins.) above the adjacent depressions. A large part of the surface of the mud b o i l was completely devoid of, or only scantily covered with, vegetation (Figure 44). The bare mud areas were often covered with a. network of miniature cracks which gave a blocky appearance to the surface. In places these cracks had developed into strong r a d i a l furrows 10-15 cms. (4-6 ins.) deep. Small sedges (Carex sp.) were the dominant species of the sparsely vegetated areas of the mud b o i l surface. The Common Crowberry (Empetrum nigrum) and Mountain Cranberry (Vaccinium v i t i s - i d a e a ) were, prominent i n the larger surface furrows and along the margins of the mud b o i l where they were found i n association with the Narrow-leafed Labrador-tea (Ledum, palustre ssp. decumbens). Also flanking the outer slopes of the mud b o i l was a f l o r a characterized by 248 Figure 44 M U D - B O I L V E G E T A T I O N T Y P E S i — i — i Edge of M u d - B o i l or Tussock H^ v^  S e d g e s p e c i e s with C r a n b e r r y and C r o w b e r r y p2-;T>j S e d g e tussock lyXXX] M o s s w i th S e d g e s p e c i e s E%?%1 H e a t h e r W i Mow G r o u n d b i r ch B l u e b e r r y p a t c h Bare g r o u n d 249 willows (Salix arctica),, A r c t i c Blueberry (Vaccinium uliginosum) , Gland-ular Birch (Betula glandulosa) and small tussocks of the Sheathed Cotton-grass (Eriophorum.vaginatum). Sedges, A r c t i c White Bell-heather, and the Sheathed Cotton-grass, along with several moss species, were the.dominant plants of the surrounding depression areas. S t r u c t u r a l l y , the mud b o i l s were made up en t i r e l y of mineral s o i l . The surface mantle of vegetation and peat was very thi n and discon-tinuous, but i t thickened towards the margins where similar peat wedges occurred beneath the depressions. The buried organic layer was also ex-tremely patchy, and i n many places was non-existent. One other notable feature of the mud b o i l was the occasional presence of small quantities of dark, organic material embedded i n the mineral s o i l at various le v e l s . Individual hummocks were generally to be found from top to bottom of most slope p r o f i l e s , and on a l l but the gentlest slopes the hummocks were aligned i n a downslope direc t i o n . L o c a l l y , however, the hummocks had coalesced to produce d i s t i n c t i v e , continuous vegetated stripe features which were also aligned i n the same manner. These stripe features were often readily discernible through the contrasting vegetation patterns of the raised centres and the intervening depressions. On gently-sloping t e r r a i n , the difference i n elevation between the ridges and the furrows was seldom more than a few centimetres. On steeper slopes, averaging 10-15 degrees, especially those which were south-facing or the sit e s of l a t e -l y i n g snow patches, the stripe forms achieved much greater prominence. Figure 45 and Plate XVII i l l u s t r a t e some of the aspects of these earth hummock stripes. The ridges, which were sometimes continuous over distances of about 10-20 metres (30-65 f e e t ) , were approximately 70-80 cms. (27.5-31.5 ins.) i n width and as much as 50-60 cms. (19.5-23.5 ins.) 250 Figure 45 EARTH HUMMOCK STRIPES ] Minera l soil 251 Figure 46 E A R T H H U M M O C K S T R I P E S V E G E T A T I O N T Y P E S 1 Met re i—i—i Edge of Hummock Stripes [JVT j L i c h en / a ve n s / sed g e a s s o c i a t i o n K"°'°*l Avens Sedge spec ies and Bea r ber ry \yy//ss] Moss H e a t h e r wi th S e d g e spec ies W i l l o w B lueber ry p a t c h Plate XVII E A R T H H U M M O C K S T R I P E S A. S t r i p e features prod-uced by the coalescence of earth hummocks. B. Excavation along the lon g i t u d i n a l axis of an earth hummock s t r i p e , showing the progressive b u r i a l of the organic material. (Tape extended to 1 foot f o r s c a l e ) . C. Transverse or cross-slope excavation of an earth hummock s t r i p e , showing the mineral core and wedges of organic material beneath the troughs. (Pocket tape case, 2 inches, shows sc a l e ) . 253 high. Seldom, however, were these ridges simple straight features, and they frequently exhibited a beaded form involving a d i s t i n c t l a t e r a l d i s -placement across the slope. The intervening troughs were very strongly developed, and often took on the form of narrow, steep-sided c l e f t s . The vegetation cover of the stripes (Figure 46) was si m i l a r to that described for the. hummocks, except for higher percentages of Cassiope  tetragona and moss occurring i n small depressions on the surfaces.of the ridges, and the absence of Eriophorum vaginaturn and Betula glandulosa from the vegetation types. The underlying peat layer was continuous but highly irregular i n thickness, ranging from a few centimetres to 45 cms. (17.5 i n s . ) . Local thickenings of peat material occurred i n close association with the hollows on the stripe surface and with the narrower widths i n the beaded forms. The mineral core of the ridges (see Figure 45), varied i n thickness from 20-45 cms. (8.0-17.5 i n s . ) , and was thicker near the f r o n t a l downslope end. This mineral core frequently contained t h i n , taper-ing wedges of organic material, which extended down below the thicker portions of the. peat layer above. These wedges dipped steeply downward i n an upslope d i r e c t i o n , and detached,elongated s t r i p s of organic material, wholly incorporated within the mineral s o i l , showed the same general orientation. The f r o n t a l lobe of the stripe showed a wedge of organic material i n the process of being buried by the mineral core (Plate XVII-B), and this could be traced upslope into the buried organic layer. In cross-section, the p r o f i l e s revealed wedges of organic material extending down in places to unite with organic material at the f r o s t table (Plate XVII-C). Apart from the.presence of the_buried organic layer, these features show a d i s t i n c t s i m i l a r i t y to those described by Sharp, and t h i s would seem to j u s t i f y the adoption of his nomenclature for the microrelief.features. 254 The Size and Form of Earth Hummocks. The l i t e r a t u r e contains few references to the sp e c i f i c s i z e , spacing or form of hummocks, and, when these dimensions are given, they are frequently extremely generalized and descriptive. For example, Sharp, describing earth hummocks i n the St. E l i a s Range, wrote: "The Wolf Creek hummocks, are 1 to 2 feet high with ground dimensions of 1 to 5 feet. On f l a t s or slopes of 5 degrees the hummocks are crudely hemi-spherical, but on steeper slopes they develop an elongation across the slope. The downhill side i s higher and steeper than the u p h i l l side, and i n a few places the upper surface grades back into the h i l l slope without an intervening depression, thus forming a small t e r r a c e " . ^ The diagrammatic i l l u s t r a t i o n of an earth hummock i n Sharp's paper suggests that they are similar to the microrelief features found on Garry Island, and the two structures may indeed have a similar o r i g i n . The change from a roughly hemispherical to an elongated form was also ob-served on Garry Island, but the direc t i o n of elongation appeared to be downslope rather than across i t . Preliminary observations i n the f i e l d suggested that there might be a correlation between the form of the hummock and the angle of the slope. Attempts to establish such a simple correlation proved f r u i t l e s s , however, since the prevalence of smooth, convex-concave slope p r o f i l e s pro-vided very few slope segments of constant i n c l i n a t i o n where s u f f i c i e n t l y large samples could be obtained to examine this relationship. However, these observations suggested a change i n the size and form of the hummocks between the tops and bottoms of these same slope p r o f i l e s . 0 /Sharp, R.P. (1942) " S o i l structures i n the St. E l i a s Range, Yukon Te r r i t o r y " , Journal of Geomorphology, Vol. 5, p. 283. 255 The f i e l d method adopted to investigate these changes was 68 based on that used by Strahler and Koons to measure t e r r a i n roughness. Five slope p r o f i l e s were selected to cover the possible variations i n as-pect. The cross-profiles of the slopes were then surveyed and, on each of them, stations were established for sight lines across the slope. The slope p r o f i l e s and the locations of the sampling stations are shown i n Figure 47. The locations of the stations were determined subjectively, rather than objectively as i n Strahler's study, at points where the hummock form appeared to be v i s u a l l y different from that at the adjacent upslope station. At each of these points an alidade was set up, and a horizontal sight l i n e was made along the contour to a stadia rod. Since the hummocks were not aligned perfectly along the contours, i t was necessary to i n t e r -pret the sight l i n e as an angular sector. The stadia was moved within t h i s sector to the highest or lowest points of hummocks or depressions respectively, parts of which actually crossed the sight l i n e . A tape was also stretched along the contour to provide data on the spacing of these microrelief features. In order to provide s u f f i c i e n t l y large samples for s t a t i s t i c a l comparisons, the procedure-was repeated at each station with another sight l i n e along the contour i n the opposite d i r e c t i o n . Data on hummock heights were obtained by subtraction of successive stadia height readings as representing the v e r t i c a l distance through which the ground p r o f i l e i s displaced between adjacent hummocks and depressions along the angular sight l i n e sector. The data were then Strahler, A.N. and Koons, D. (1960) "Objective and. Quantitat-ive Methods of Terrain Analysis", U.S_. Dept. of the Navy, Office of Naval  Research, Geography Branch, Project NR 387 - 021, Contract Nonr 266 - 50, 51 p. Figure 47 SLOPE PROFILES SHOWING LOCATIONS OF EARTH HUMMOCK SAMPLING STATIONS 256 1 O 2 4 6 e 10 Distance in Metres 257 processed by frequency d i s t r i b u t i o n analysis to provide indices of mean height, variance and standard deviation. Since most of the frequency dis-tributions were not markedly skewed (Figures 48-52), i t was not considered necessary to. transform the data before further s t a t i s t i c a l tests could be performed. The mean values for the hummock heights are presented i n Table XIV. The means were further analysed by running paired 't'-tests on each p r o f i l e to determine whether or not the differences between the means for adjacent sampling stations were s i g n i f i c a n t at the .01 or .05 significance levels. The procedure for the 't'-test followed that outlined by Croxton and Crowden, where i t i s assumed that the samples are independent, and 69 that the sample means follow a normal d i s t r i b u t i o n . Reference to the range values i n Table XIV shows that at each of the stations there was a considerable v a r i a t i o n i n the heights of the hummocks. This range value, calculated as the difference i n centimetres between the height of the largest and smallest hummocks observed along the indivi d u a l sight l i n e s , increased i n a downslope di r e c t i o n . The greater uniformity of hummock heights i n the upper portions of the slope p r o f i l e s was due to the absence of any extremely large hummocks i n these locations. Despite these v a r i a t i o n s , each of the f i v e p r o f i l e s shows that there was a gradual increase i n the mean height of the hummocks from the top to the bottom of the slope. The majority of these changes were found to be s i g -n i f i c a n t at the .01 l e v e l (Table XIV). I t i s also quite noticeable that most of the non-significant values encountered were i n tests involving Croxton, F.E. and Crowden, D..J. (1955) Applied General  S t a t i s t i c s , Prentice-Hall, Inc., Englewood C l i f f s , N.J., Second E d i t i o n , p. 651. 258 Figure 48 H I S T O G R A M S OF H U M M O C K H E I G H T S - P R O F I L E 1 S T A T I O N 1 S T A T I O N 2 25 r 20 15 oi 10 X = 7 92 n =51 0 6 12 Height (Cms.) 20 15 -v 10 X = 10.06 n = 53 0 6 1? 18 Height (Cms.) S T A T I O N 3 2 0 -u 15 -10 X = 13.11 n = 60 0 6 12 18 24 Height (Cms.) S T A T I O N 4 20 X = 17. 68 n = 50 6 12 18 24 30 Height (Cms.) S T A T I O N 5 S T A T I O N 6 5 -X =21.03 n = 54 0 6 12 18 24 30 36 Height (Cms.) 0 I I-X =16.15 n = 65 18 24 30 He igh t (C r 259 Figure 49 H I S T O G R A M S O F H U M M O C K H E I G H T S - P R O F I L E II S T A T I O N 1 S T A T I O N 2 X = 7. 3 2 n = 55 0 6 12 18 Height (Cms.) 25 20 u 15 . X = 9.75 n = 68 0 6 12 18 Height (Cms.) S T A T I O N 3 S T A T I O N 4 » 5 -X = 12.50 n = 55 • 0 6 12 18 24 Height (Cms.) 15 r » 5 -X =15.24 n = 56 0 6 12 18 24 30 Height (Cms.) S T A T I O N 5 S T A T I O N 6 '5 r " 10 X = 19.81 n = 51 0 6 12 18 24 30 Height (Cms.) 15 X =15.54 n = 67 0 6 12 18 24 30 He igh t (Cms.) Figure 50 H I S T O G R A M S OF H U M M O C K H E I G H T S - PROF ILE III 260 S T A T I O N 1 20 10 • 0 6 12 He igh t ICms. X » 7.01 n = 59 S T A T I O N 20 15 • 10 5 • 6 H e i g h t • 2 18 (Cms.) X =10.97 n - 67 S T A T I O N 3 25 20 * is cr e 10 X = 12.80 n = 68 6 12 18 24 H e i g h t (Cms.) S T A T I O N 4 15 r o 10 4) 5 . X = 17.37 n - 73 • b 6 12 18 24 30 H e i g h t (Cms.) S T A T I O N 6 S T A T I O N 5 io r « 5 • X = 23.47 n s 61 LrftL 0 6 12 18 24 30 36 42 H e i g h t (Cms.) 20 r 15 • 10 • r X • 22.25 n » 69 12 18 2 4 30 36 4 2 H e i g h t (Cms.) 261 Figure 51 H I S T O G R A M S OF H U M M O C K H E I G H T S - P R O F I L E IV 30 r 25 20 15 10 S T A T I O N 1 X « 6.71 n . 80 0 6 12 18 Height (Cms.) S T A T I O N 20 15 10 X = 9.45 n * 6 3 0 6 12 18 Height (Cms.) S T A T I O N 3 25 20 " 15 Z 10 X = 12.80 n • 5 6 0 6 12 18 24 Height (Cms.) S T A T I O N 5 10 X -21.03 n • 52 t f b 0 6 12 18 24 30 36 Height (Cms.) S T A T I O N 4 15 10 ar « 5 X = 15.54 n - 50 6 12 18 24 Height (Cms.) S T A T I O N 6 10 • X -21.64 n - 50 Lb 0 6 12 18 24 30 36 Height (Cms.) 262 Figure 52 H I S T O G R A M S OF H U M M O C K H E I G H T S - PROF ILE V S T A T I O N 1 30 « 15 10 5 -X = 7.32 n - 70 0 6 12 H e i g h t , (Cms.) S T A T I O N 2 20 15 X = 7.62 n s 6 6 0 6 12 18 H e i g h t (Cms.) S T A T I O N 3 20 10 X =11.28 n = 63 0 6 12 18 24 H e i g h t (Cms.) S T A T I O N 4 10 -r l X = 13.72 n = 58 JZ3 0 6 12 18 24 30 H e i g h t (Cms.) 20 r 15 -10 S T A T I O N 5 X = 17. 3 7 n = 72 6 12 18 24 30 H e i g h t (Cms.) S T A T I O N 6 15 u 10 X = 21.03 n - 63 0 6 12 18 24 30 36 H e i g h t (Cms.) 263 TABLE XIV ANALYSIS OF GARRY ISLAND EARTH HUMMOCK DATA - HEIGHTS (Cms.). PROFILE I (Azimuth • - 360 ) Station Sample Range X s t Significance Level Size .01 .05 1 (Top) 51 11.28 7.92 3.12 3.66 S. 2 53 11.28 10.06 2.74 7.35 S. 3 60 21.03 13.11 3.99 5.08 S. 4 50 23.16 17.68 5.33 2.43 N.S. S. 5 54 25.30 .21.03 6.00 4.21 S. 6 (Bottom) 65 28.65 16.15 6.33 PROFILE I I (Azimuth - 285 ) 1 (Top) 55 14.63 7.32 2.66 4.35 s. 2 68 13.72 9.75 3.37 3.44 s. 3 55 21.34 12.50 5.33 2.69 S. 4. 56 27.43 15.24 5.35 3.94 S. 5 51 26.52 19.81 6.31 3.69 S. 6 (Bottom) 67 25.30 15.54 6.07 PROFILE I I I (Azimuth - 180 ) 1 (Top) 59 11.28 7.01 2.59 6.60 S. 2 67 17.68 10.97 3.89 0.80 N.S. 3 68 18.90 12.80 4.45 5.36 S. 4. 73 27.13 17.37 5.54 4.81 S. 5 61 34.75 23.47 8.88 . 0.89 N.S. 6 (Bottom) 69 28.35 22.25 6.53 264 TABLE XIV (Continued) PROFILE IV (Azimuth - 55 ) Station Sample Range X s t Significance Level Size .01 .05 1 (Top) 80 14.33 6.71 3.00 4.76 S. 2 63 13.41 9.45 3.84 4.64 S. 3 56 17.68 12.80 2.20 3.86 S. 4 50 21.34 15.54 4.66 4.83 S. 5 52 24.69 21.03 6.49 0.51 N.S. 6 (Bottom) 50 22.56 21.64 5.47 PROFILE V (Azimuth - 230 ) 1 (Top) 70 13.72 7.32 2.57 0.90 N.S. 2 66 15.54 7.62 3.72 5.08 S. 3 63 24.08 11.28 4.36 2.78 S. 4 58 25.30 13.72 5.21 3.70 S. 5 72 24.08 17.37 5.81 3.28 S. 6 (Bottom) 63 27.43 21.03 7.03 265 paired means at the terminal parts of the p r o f i l e s . This probably r e f l e c t s the fact that an attempt was made to establish s i x stations, approximately equally spaced, on each of the p r o f i l e s . Where the topographic p r o f i l e consisted of a short, steep mid-section and r e l a t i v e l y longer upper convexities and lower concavities, as i n P r o f i l e I I I , the stations i n these terminal positions were probably too close together. Three of the p r o f i l e s show that towards the base there was a tendency for the mean height of the hummocks to decrease again, and two of these cases were s i g n i f i c a n t at the .01 l e v e l . On the other two p r o f i l e s , IV and V, the mean height of the hummocks increased right to the base of the p r o f i l e . Variations i n aspect may be p a r t i a l l y res-ponsible for t h i s , but i t appears to be more related to the drainage conditions at the foot of the slope. Where i t was observed that there was a decrease i n hummock height at the base of the slope, the p r o f i l e s terminated i n semi-closed depressions with poor drainage. There was considerable evidence of a moss-peat accumulation on the f l o o r of the depression, and the trough areas between the indiv i d u a l hummocks showed moss accumulations as described i n the preceding section. I t i s th i s i n f i l l i n g of the depressions that i s mainly responsible for the apparent decrease i n height of the hummocks. In the case of p r o f i l e s IV and V 'active' stream channels drained the foot of the slope, and there was l i t t l e evidence of the buil d up of organic material. Data on the sizes of the hummocks were obtained by combining the horizontal distance measurements between each depression, and these data were subjected to the same form of analysis. The results of the analyses are presented i n Table XV. As i n the case of the height data, there was a considerable range i n the sizes at each station , but the 266 TABLE XV ANALYSIS OF GARRY ISLAND EARTH HUMMOCK DATA PROFILE I Station Sample Range X s t Size 1 (Top) 50 107 73.20 22.23 2.66 2 52 99 86.06 25.88 0.23 3 58 122 87.22 27.76 1.41 4 48 130 94.77 26.49 0.73 5 47 122 99.16 31.70 0.12 6 (Bottom) 64 147 99,92 31.72 PROFILE I I 1 (Top) 55 99 77.90 24.41 0.81 2 66 99 81.66 25.88 2.41 3 54 155 95.15 34.82 0.54 4 55 145 98.60 . 31.37 1.89 5 49 170 112.60 42.80 0.17 6 (Bottom) 66 147 111.30 31.12 PROFILE I I I 1 (Top) 2 3 4. 5 6 (Bottom) SIZES (Cms.). Significance Level .01 .05 S. N.S. N.S. N.S. N.S. N.S. N.S. S. N.S. N.S. N.S. 57 102 76.43 17.86 0.26 N.S. 65 91 77.09 19.66 0.85 N.S. 66 76 80.24 22.25 0.56 N.S. 71 84 82.27 20.09 3.44 S. 59 97 94.84 21.23 1.20 N.S. 67 127 99.67 23.24 267 TABLE XV (Continued) PROFILE IV Station Sample Range X s Size 1 (Top) 78 107 69.98 21.77 2 61 102 74.42 19.56 3 54 97 85.88 22.81 4 48 112 89.23 26.44 5 48 107 99.95 23.19 6 (Bottom) 48 114 97.49 26.39 PROFILE V 1 (Top) 68 137 71.98 22.81 2 64 94 76.28 20.14 3 61 142 81.56 25.81 4 56 122 82.60 25.98 5 70 119 88.98 20.27 6 (Bottom) 61 140 99.26 22.78 t Significance Level .01 .05 1.24 N.S. 2.87 S. 0.68 N.S. 2.09 N.S. S. 0.48 N.S. 1.13 N.S. 1.27 N.S. 0.22 N.S. 1.53 N.S. 2.71 S. 268 increase i n range values i n a downslope direction was not as marked. As Table XV shows, there was an increase i n the size of the hummocks to-wards the base of the slope, but tests of paired means yielded very few sig n i f i c a n t values. Tests on the mean hummock size of samples 1 and 6 of each of these p r o f i l e s , however, yielded ' t ' values of 5.00, 6.42, 6.10, 6.30 and 6.73 for p r o f i l e s I to VI respectively, a l l of which were s i g -n i f i c a n t at the .01 l e v e l , and indicate that there i s a s i g n i f i c a n t increase i n the size of hummocks i n a dpwnslope direction. Although moderate success, demonstrating changes i n the size of hummocks, was achieved using t h i s method, i t was hoped that more success could be obtained using the idea of an 'index of c i r c u l a r i t y 1 . The size measurements were made by taking readings d i r e c t l y from the metal tape. Insomuch as the tape was fixed along the sight l i n e , while the points establishing hummock and depression positions were considered within an angular sector, the linear distances obtained were not a true indicator of the size of the hummocks. Moreover, these readings measured only the variations i n the cross-slope dimensions of the hummocks, and gave no insight into corresponding changes i n the downslope dimensions. The idea of an index of c i r c u l a r i t y was introduced therefore to i n -vestigate changes i n the r a t i o of downslope to cross-slope dimensions of hummocks at different positions on the topographic p r o f i l e . Equi-dimensional hummocks would have an index of 1.0, whilst hummocks which were elongated down or across the slope would have indices greater or less that 1.0 respectively. P r o f i l e V was selected to investigate this index as w e l l as certain angular properties of the hummock faces. Three sample plots were established near the top, middle, and base of this p r o f i l e . Each plot 269 extended for a distance of approximately 8 metres (26 feet) i n a down-slope d i r e c t i o n , and for variable distances along the contours on either side of the p r o f i l e l i n e . The sampling procedure consisted of taking the f i r s t 25 hummocks i n th i s plot on either side of the p r o f i l e l i n e , and measuring their downslope and cross-slope dimensions. These values were then used to calculate the index of c i r c u l a r i t y . The mean index of c i r c u l a r i t y for 50 hummocks at the upper part of the slope p r o f i l e was 1.05 with a maximum and minimum of 1.47 and 0.82 respectively. Eighty per cent of the hummocks i n this plot had an index greater than 1.0, and only eight per cent had an index of less than 1.0. The mean index for 50 hummocks i n the middle section of the p r o f i l e however was 1.95. The maximum and minimum values of 2.58 and 1.42 show that a l l the hummocks, without exception, exhibited an elongation i n the downslope direction. In the t h i r d p l o t , located at the base of the p r o f i l e , the mean index of c i r c u l a r i t y was 1.09, with a maximum and minimum of 1.31 and 0.82 respectively. Once again only eight per cent of the hummocks had an index value of less than 1.0. Paired tests of the mean index values were run using the value for the middle plot against the values for the plots i n the terminal positions of the p r o f i l e , and yielded ' t 1 values of 21.57 and 24.40. Both these values are s i g n i f i c a n t at the .01 l e v e l . These figures indicate that, i n addition to changes i n the height and size of the hummocks i n a downslope d i r e c t i o n , there i s also a s i g n i f i c a n t change i n the form of the hummocks from an almost c i r c u l a r outline at the upper slope position, to an elongated form on the steeper portions of the slopes, and a return to an almost c i r c u l a r shape at the foot of the slope. Since only eight of the one hundred and f i f t y hummocks investigated showed an elongation across the slope, these findings contrast 270 70 with those observed by Sharp. At the same time as the index of c i r c u l a r i t y measurements were being made, the in c l i n a t i o n s of the downslope and upslope faces of the indi v i d u a l hummocks were also recorded. An arbitrary decision was made that i f the two in c l i n a t i o n s differed by less than f i v e degrees the hummock would be classed as symmetrical. For the 50 hummocks i n the upper p l o t , i t was found that 72 per cent were steeper on the downslope face, 18 per cent were symmetrical and only 10 per cent had steeper faces on the upslope side. The mean angle of the downslope face was 48 degrees compared to 31 degrees for the upslope face. In the middle plot the c l a s s i f i c a t i o n gave corresponding values of 72, 18 and 10 per cent and 55 and 42 degrees. Most of the hummocks thus exhibited a marked asymmetry similar to. that observed by Sharp, but decidedly different from the observations of Raup who found that turf hummocks have steeper upslope 71 faces. I t was also noticeable that many of the hummocks were not aligned perfectly i n a simple upslope-downslope d i r e c t i o n , but were sometimes obliquely oriented to the general dire c t i o n of the slope. Furthermore the direction of the oblique tendency was found to be highly variable over even small areas. The studies of the angles of the hummock faces also revealed the influence of lemming a c t i v i t y , which was especially evident i n sites frequently occupied by l a t e - l y i n g snow patches. At these sites the downhill faces of the hummocks were frequently ..vertical, or even overhanging, and were punctured by the burrows of these small 70, 71, 'sharp, R.P. (1942) OJJ. c i t . , p. 283. "Sharp, R.P. (1942) op_. c i t . , p. 283. Raup, H.M. (1965) op_. c i t . , p. 105. 271 rodents. I t was mentioned e a r l i e r that the locations of the f i v e p r o f i l e s were carefully selected ,to try to evaluate the varying form of the hummocks on slopes with different aspects. As Table XV shows, the results were not p a r t i c u l a r l y i lluminating, although i t may be s i g n i f i c a n t that the largest hummocks encountered were on the south-facing slope or conversely on northeast-facing slopes, the most favourable sit e s for la t e -lying snow patches. I t was i n these same two environments that the hummocks were v i s u a l l y the most distinct,and where the tendency to coalesce into a stripe form was best developed. On a smaller scale, i t was found that the influence of aspect gave a further quality of asymmetry to the ind i v i d u a l hummocks. One hundred hummocks were studied on a west-facing slope and i t was found that, using similar c r i t e r i a to the preceding i n c l i n a t i o n studies, 14 per cent of the hummocks could be classed as symmetrical, 10 per cent had steeper faces on the south-facing side, while 76 per cent of the hummocks had a steeper slope on the north-facing side. At the same time i t was noticed that the north-facing sides tended to be shorter and higher when compared to the gentler, longer p r o f i l e s facing in a southerly direction. ' In summary, these studies show that there i s a considerable change i n the forms of indiv i d u a l earth hummocks i n a downslope di r e c t i o n . Hummocks show s i g n i f i c a n t increases i n height and size from the top to the base of a slope, and appear to go through a c y c l i c a l form change from c i r c u l a r through elongated and back to a c i r c u l a r form. Individual hummocks exhibit a preferred orientation i n a downslope d i r e c t i o n , and where they coalesce to form stripes they demonstrate the same pref e r e n t i a l alignment. Downslope faces of the hummocks are steeper than corresponding 272 upslope faces by an average of 15 degrees. On west-facing slopes the influence of varying aspect imparts an additional factor of asymmetry with north-facing sides of the hummocks being shorter, higher and an average of 15 degrees steeper than the opposing south-facing sides. Each of these facts deserves f u l l consideration i n any discussion concerning the o r i g i n and development of these microrelief features. The o r i g i n and development of Earth Hummocks. In the voluminous l i t e r a t u r e that has been published describing various forms of patterned ground, numerous theories have been proposed concerning their o r i g i n . Although i t i s now almost universally agreed that they are the result of f r o s t action, there i s s t i l l a great deal of controversy con-cerning the precise mode of o r i g i n of many of these forms. Washburn, i n his c l a s s i c a l review of these features, l i s t e d no fewer than nineteen different hypotheses, summarized according to the dominant processes, that various authors have proposed to explain th e i r genesis, and concluded that the o r i g i n of most forms i s uncertain and, i n the majority of cases, i s • 7 2 polygenetic. The most commonly accepted o r i g i n for the doming-up of the mineral s o i l to form the earth cores of hummocks follows the theories pro-73 posed by Thoroddsen and Beskow. In his discussion of the formation of 'thufurj, Thoroddsen concluded that the mineral s o i l p a r t i c l e s were 7 2Washburn, A.L. (1956) op_. c i t . 73 Thoroddsen, Th. (1914) "An account of the physical geography of Iceland with special reference to the plant l i f e " , pp. 187-343, i n Kolderup Rosenvinge, L. and Warming, E. The Botany of Iceland, Vol. 1, 675 p. Beskow, G. (1930) "Erdfliessen und Strukturboden der Hoch-gebirge im Lic h t der Frosthebung, Geol. Foren. Stockholm, Forh., Bd. 52, pp. 622-638. Information c i t e d i n Raup, H.M. (1965) op_. c i t . , p. 15. 273 gradually moved upward under the hummocks through the combined action of deep freezing i n the troughs and the upward movement of c a p i l l a r y water i n 74 the hummock centres. Beskow claimed that deeper, more-rapid freezing of the damp trough areas took place while the centres of the hummocks remained unfrozen at depth, and the re s u l t i n g pressures forced material into the unfrozen c o r e s . ^ The d i s t r i b u t i o n of snowfall was considered to be an important factor by Griggs, who suggested that i t s effect would be to create differences i n the growing season period between hummocks and 76 troughs, i n favour of more luxuriant growth on the hummock centres. He further claimed that the snow would exert l a t e r a l pressures against the sides of the hummocks thereby squeezing them higher. In the chapter on permafrost conditions, i t was shown that the raised hummock centres are sites of deeper thaw during the summer than the adjacent troughs. Excavations of different hummocks through the summer months showed that, i n each case, there was a gradual increase i n the water content down to the position of the f r o s t table, and this was accompanied by an increase i n the percentage of fi n e material with depth. Where these excavations: were continued below the position of the fr o s t table, i t was evident that the mineral s o i l contained many small ice lenses of the s i r -l o i n type. The size of these lenses increased with depth and with the size of the hummock, and i n the mud bo i l s there was a d i s t i n c t layer of clear ice near the base of the active layer. Unfortunately i t was not 74 Thoroddsen, Th. (1914) op_. c i t . 7 5Beskow, G. (1930) op_. c i t . 76 Griggs, R.F. (1936) "The vegetation of the Katmai D i s t r i c t " , Ecology, Vol. 17, pp. 380-417. 274 possible to obtain a complete picture of the penetration of the f r o s t l i n e during the f a l l and winter months, but, from the limited data available, i t appears that freezing of the ground surface occurs most rapidly i n the trough areas where the moist organic material acts as a good conductor of heat. The penetration of the f r o s t l i n e i s slower under the hummocks and once the troughs are completely frozen, any confined pressures could re s u l t in further continued slow penetration below the hummocks. The combinations of slow freezing and fine material i s extremely favourable for the formation of ice lenses and, provided that there i s a source of water, the growth of these lenses may cause a general doming of the hummock surface; Assuming that the doming.of the hummocks i s primarily a con-sequence of d i f f e r e n t i a l f r o s t action between the hummock.centres and the adjacent troughs, several authors have expressed concern over the type of topography upon which t h i s process could work during the i n i t i a l stages of hummock development. Beskow thought that t h i s r e l i e f could develop from chance variations i n the s o i l surface and vegetation cover, but the ubiquitous occurrence of these features would seem to require much more than simple chance variations i n the surface form.^ Raup suggested that these micro-elevations could be achieved i n a number of ways such as cobbles, boulders, l o c a l sand and s i l t deposits i n stream beds, moss polsters i n snow beds, upfrozen stones, pre-existing g e l i f l u c t i o n features or primarily through the development of a normal irregular surface i n mats 78 of aquatic mosses. The general paucity of stones and the occurrence of hummocks i n locations which are not, and never were, occupied by stream Cited i n Raup, H.M. (1965) o£. c i t . , pp. 15-16. 'Raup,, H.M. (1965) op_. c i t . , pp. 106-107. 275 channels would appear to eliminate the p o s s i b i l i t y that the earth hummocks on Garry Island had a similar mode of o r i g i n . . Attempts to lower the le v e l of a lake on the island during the ;summer of 1965 provided a valuable opportunity to study t h i s problem. A lowering; of the water l e v e l exposed a gently-sloping, vegetation-free lake bottom which, towards the shoreline, was l o c a l l y mantled by a l i t t e r of peat debris 4-5 cms. (1.5-2.0 ins.) thick. By the end of the summer, this surface was covered by a network of desiccation cracks. Observations during the summer of 1966 showed that some of these cracks had been accentuated into a trough-like form, probably i n part by. f r o s t action during the preceding f a l l , but also by running water derived from spring melting of snow. The largest troughs, as much as 20-30 cms. (8-12 ins.) wide and 15-20 cms. (6-8 ins.) deep, ran d i r e c t l y down the slope and ex-tended as far as the new. shoreline on the peat-free surface. In other places, consisting of bare mineral s o i l , a s i m i l a r , though less prominent, network of troughs was also evident. Although the surface was s t i l l largely devoid of any vegetation cover, a few small sedges had taken root on the peat mat. Only time w i l l t e l l whether or not this pattern of cracks w i l l develop into a system of earth hummocks, but these observations suggest that the formation of earth hummocks may be polygenetic from even the most incipient stages. The i n i t i a l topography may be produced by a combination of desiccation and fr o s t action l o c a l l y accentuated by r i l l w o r k . In t h i s respect the former two processes re s u l t i n a subdivision of the ground into a block form, and subsequent development of the major cracks provides the framework for subsequent hummock development. Some of the i n i t i a l r e l i e f and alignment pattern of the hummocks however may be the resul t of running 276 water during t h i s i n i t i a l stage. I t i s also quite probable that an analogy may be drawn at t h i s stage between earth hummock formation and the incipient f r o s t crack stage of the tundra polygon development. The accentuation of the major cracks at the expense of the weaker ones to give the basic hummock form, may possibly be the re s u l t of the formation of a miniature ice-wedge pattern, and the subsequent growth of these wedge areas may serve to increase the height of the hummocks during their early period of growth. Once the.production of an i n i t i a l topography has provided a surface upon which d i f f e r e n t i a l f r o s t action can take place, the role of s o l i f l u c t i o n becomes of major importance i n the further development of earth hummocks. As noted previously, the presence of a buried organic layer was detected i n a l l the hummock excavations that were made. A number of theories has also been proposed concerning the o r i g i n of this organic material, including climatic change, convectional movements within the s o i l and progressive b u r i a l by, the downslope movement of material. With the possible exception of the small quantities of organic material incorporated within the mineral s o i l of the mud b o i l , there i s l i t t l e or no evidence to suggest that any type of convectional movement i s involved i n the earth hummocks on Garry Island, similar to that described by Hopkins 79 and Sigafoos. The fact that, i n a number of the excavations made, the wedge of organic material beneath the depresssions was strongly convex i n a downslope d i r e c t i o n , and i n many places was found to taper upslope again beneath the hummock centres, i s the most convincing evidence that, as Mackay has suggested, the organic layer at depth i s the r e s u l t of prog-Hopkins, D.M. and Sigafoos, R.S. (1951) "Frost Action and Vegetation Patterns on Seward Peninsula, Alaska", U.S. Geol. Survey  B u l l e t i n , No. 974-C, 101 p. 277 80 ressive b u r i a l . This i n turn suggests that s o l i f l u c t i o n plays an important role i n further earth hummock development. The relevance of these statements w i l l now be applied to a complete interpretation of the f i e l d data. Summary. The earth hummocks on Garry Island probably originated through the formation of miniature desiccation/frost crack patterns on newly-exposed surfaces. • In the past, such surfaces would have been common following the marine submergence of the island and the sub-sequent withdrawal of the sea. L o c a l l y , the action of running.water may have contributed to an accentuation of the major cracks, and produced the dominant alignment i n a downslope direction. Contrasting vegetation associations, established over the hummock areas and intervening troughs, would lead to d i f f e r e n t i a l f r o s t action due to their varying insulating properties on the underlying ground. By deeper, more rapid freezing of the trough areas, pressures may have been established beneath the hummocks, and additional quantities of fi n e clay may have been carried up beneath the hummocks thereby accentuating ..the domed form. The formation of ice lenses i n the mineral core results i n an expansion of the substrate and a heaving of the ground surface. During the thaw period of the following summer, melting occurs most rapidly beneath the hummock centres, and melting of the ice results i n s e t t l i n g of the material under the influence of gravity. Brown has also pointed out that during the spring, the mineral s o i l i s close to saturation, the fr o s t Mackay, J . Ross (1958) "A subsurface organic layer associated with permafrost i n the Western A r c t i c " , Geographical Branch, Ottawa, Geographical Paper, No. 18, 21 p. 278 81 table i s close to the surface, and the s o i l flows more re a d i l y , and Everett has produced quantitative data indicating that the greatest move-82 ments i n similar features i n Alaska occur during the freeze-up and thaw. On the gently-sloping surfaces, the differences between the elevations of the hummocks and adjacent troughs are quite small, and the available moisture supply i s most limited. Consequently the amounts of ice lensing i n these locations w i l l be minimal, and the low angle of slope w i l l further result i n very slow downslope migration of these features i n the s o l i -f l u c t i o n process. Once these hummocks reach steeper slopes, however, the process becomes r e l a t i v e l y more rapid. On the steeper slopes the influence of gravity becomes more pronounced,.and the downslope movement of the material i s reflected i n an elongation of the hummock form, i n t h i s direction. Downslope faces of the hummock advance over the vegetation of the depressions, and t h i s organic material i s buried at depth. Since the amount of organic accumulation i n the troughs, located on the upper parts of the slope, i s small compared to similar situations on the wetter, lower slopes, this organic layer i s quite thin and may be highly discontinuous. Furthermore as the height of the hummock increases, the augmented exposure of the hummock centres results i n deeper penetration of the summer thaw and the formation of larger dep-ressions i n the f r o s t table, r e l a t i v e to the surrounding troughs. These 81 Brown, J . (1966) " S o i l s of the Okpilak River Region, Alaska", U.S. Army, Cold Regions Research and Engineering Laboratory, Research  Report, No. 188, p. 11. 82 Everett, K.R. (1966) "Slope-movement and related phenomena", Chapter XII i n N.J. Wilimovsky (Ed) Environment of the Cape Thompson Region, Alaska, U.S. Atomic Energy Commission, Division of Technical Information, Washington, D.C., p. 175. 279 depresssi oris are favourable sit e s for the accumulation of water which may seep down through the s o i l from up-slope. This i n turn creates a situation whereby more moisture i s available for ice lens formation during the next f a l l . Differences i n the homogeneity of the s o i l , and i n the available moisture supply, may be primarily responsible for the wide range of hummock heights encountered along the sample transects. As a re s u l t of these differences, there could be variations i n the amounts of heaving and rates of downhill movement of the ind i v i d u a l hummocks. Consequently, some of the larger hummocks may tend to override smaller hummocks immediately below them, and coalesce into stripe forms. The impeding nature of some of the hummocks may modify the downslope movement of others, so that an oblique component may be introduced. This factor may p a r t i a l l y explain the oblique nature of many hummocks and the alignment of their downslope faces observed i n the plot studies. An additional factor influencing the direct i o n of movement could be variations i n aspect. As noted i n the plot studies, there was a marked asymmetry between north- and south-facing sides of hummocks. These differences may be due i n d i r e c t l y to contrasts i n the vegetation patterns, but the steeper, higher and shorter faces on the north sides of the hummocks may also r e f l e c t a more limited amount of insolation received which, combined with the effects of the vegetation, may resul t i n slow rates of thaw. On the south-facing sides of the hummock, the amount and rate of thaw may be greater, r e s u l t i n g i n flowage of the material on that side of the hummock. The influences of varying aspect may be f e l t on a l l hummocks, and produce deflections of the downhill movement from the direction of the steepest slope. At the foot of the slope, the available moisture supply reaches a maximum and, consequently, the potential amount of heaving i s also 280 greatest. With the gradual removal of the slope factor, these heaving forces are directed v e r t i c a l l y , and may exceed the binding influence of the surface vegetation. This cover i s thus stretched and thinned, u n t i l l o c a l l y i t may be ruptured exposing the bare mineral s o i l of the mud b o i l areas. With the thinning and eventual breaking of the surface vegetation cover, the summer thaw proceeds to even greater depths i n the exposed mineral s o i l , and i t may extend below the le v e l of the buried organic layer. As additional mineral s o i l i s drawn up into the mud b o i l , the organic material may become incorporated into i t . In the above scheme, the mud b o i l i s regarded as a possible disintegration stage of earth hummock development. The d i s t r i b u t i o n of mud b o i l s however shows another s i g n i f i c a n t feature. Mud b o i l s are also found on the f l a t upland surfaces, where they are much smaller than those described above. On such f l a t surfaces, the forces of heaving exerted beneath a hummock surface w i l l be directed almost v e r t i c a l l y , and the hummock w i l l not undergo a pronounced downslope movement. Thinning of the vegetation on the surfaces of these hummocks may also be due p a r t i a l l y to the high degree of exposure to strong winds i n these locations, and to the action of ptarmigan. Under t h i s proposed system of earth hummock develop-ment, there w i l l be a gradual movement of a hummock from the upper slope to the lower slope. The formation of new hummocks on the upland surface w i l l not be i d e n t i c a l to that described i n the i n i t i a l stages, but w i l l more l i k e l y r e s u l t from variations i n the surface cover created by the downhill movement of these features. No attempt was made to determine the age of these earth hummocks, but i t i s quite evident that they are the result of very slow processes operating over many thousands of years. Similar features have 281 been dated i n other areas by radiocarbon dates of the buried organic Layer, or by the analysis of the ages of willows growing" i n the hummocks. Recent measurements from the lower part of a slope on Garry Island indicate a downslope movement of the hummocks at a rate of approximately 0.25-1.25 83 cms. (0.1-0.5 ins.) per annum. Information supplied by Dr. J. Ross Mackay, personal communication, October, 1968. CHAPTER VI SOME OBSERVATIONS ON THE GEOMORPHOLOGICAL EVOLUTION OF GARRY ISLAND In this f i n a l chapter an attempt w i l l be made to synthesize some of the material contained i n the preceding chapters by ou t l i n i n g the major episodes, including their r e l a t i v e arrangement into a chronological sequence, i n the geomorphological evolution of Garry Island. PLEISTOCENE.DEPOSITS AND GLACIATION This evolution began with the deposition of the fine sands, S i l t s and s i l t y clays which, apart from the sand headland areas, constitute the bulk of the stratigraphic succession found on the island. The beds of fine sand included i n this succession provide.the only i d e n t i f i a b l e clues as to the type of depositional environment i n which the sediments were formed. F o s s i l i f e r o u s evidence, and the abundant fragments of washed wood and peat, are indicative of a marine delt a i c environment, and the sediments were probably, deposited by an ancestral version of the present Mackenzie River. Unfortunately, no age determinations were made for any material i n these strata and the geological age of the sediments therefore cannot be stated with any degree of accuracy, but.presumably they are Pleistocene. Subsequently these sediments have been deformed by the over-r i d i n g action of glacier i c e , and the evidence suggests that the sediments were frozen at. the time of deformation. This evidence comes from crude petrographic examinations of deformed bodies of segregated ground ice 283 contained i n some of the fine-grained sediments i n the stratigraphic suc-cession. In places t h i s ground ice has a d i s t i n c t i v e banded appearance caused by an alternation of bands of di r t y and clear i c e . Examinations of Tyndall figures and the orientations of elongated bubbles i n specimens of the clear ice showed that the 'c'-axes of the ice crystals were oriented 1 at rig h t angles to the layering structures. The orientation of these axes i n ice crystals forming at the l e v e l of a slowly-penetrating freezing plane i s usually p a r a l l e l to the direction of heat flow, i . e . , normal to the freezing,plane. Since this freezing plane i n turn penetrates downward from, and approximately p a r a l l e l to, the ground surface, t h i s suggests that the banded ice structures.were o r i g i n a l l y p a r a l l e l to the overlying ground surface. Another notable aspect of the segregated ground ice i s that i t contains a. surprisingly large number of pebbles and boulders, the only source regions for which are the mainland areas to the south. Three possible origins may be inferred for these pebbles and boulders: (1) they may have been brought down by the same r i v e r which deposited the fine-grained sediments; (2) the boulders may have been ice-rafted into the area; and (3) the segregated ice may have developed, at least, p a r t i a l l y , i n a layer of g l a c i a l t i l l . Whichever of these hypotheses i s correct, i t i s now possible to infer something about the i n t e r v a l between the deposition of the sedi-ments and their subsequent deformation. At some time prior to the deform-I Mackay, J . Ross and Stager, J.K. (1966). "Thick t i l t e d beds of segregated i c e , Mackenzie Delta area, N.W.T.", Biuletyn Peryglacjalny, No. 15, pp. 39-43. 284 ation there was a r e l a t i v e emergence of the land which may have occurred as a result of shoaling i n the ancestral delta or i t may have been pro-duced by glacio-eustatic fluctuations i n the l e v e l of the sea. When the sediments were exposed, the climate must have been s i m i l a r , or possibly even colder, than the present climate, and the exposure was accompanied by the formation of permafrost, including the development of segregated ground ice. If the hypothesis that t h i s ice i s i n fact developed i n g l a c i a l t i l l i s correct, the exposure of the sediments may have been the resu l t of a lowering of sea l e v e l during a g l a c i a l advance, and some freezing of the ground may have taken place beneath a cold gla c i e r . There i s no evidence however to indicate when freezing of the sediments f i r s t occurred, nor to what depths. Although the deformed ground ice i s exposed over 610-915 metres (2,000-3,000 feet) of coastline along the southwest part of the island, t h i s figure gives no r e l i a b l e indication of the thickness of the permafrost at the time of deformation since i t undoubtedly includes a considerable amount of imbrication produced during the deformation process. There i s l i t t l e doubt that t h i s deformation was. produced by the overriding and thrusting action of glacier i c e . G l a c i a l lineation features show that an arm of the Laurentide ice sheet moved from the v i c i n i t y of Great Bear Lake, along the Mackenzie River, valley and into 2 the delta area. The northernmost extent of t h i s ice sheet i s imperfectly known, but i t seems definite that at some time the ice extended as a lobe, _ Prest, V.K. et a l . , (1968), G l a c i a l Map of Canada, Geological Survey of Canada, Map 1253A, Scale 1:5,000,000. 285 3 or a series of lobes, across the f l o o r of the Beaufort Sea. Although the pattern of ice flow i n the v i c i n i t y of the Mackenzie Delta i s clear, there are c o n f l i c t i n g opinions as to the number and r e l a t i v e extent of ice advances responsible for t h i s pattern. Muller, i n his description of the. stratigraphic cover of the,Ibyuk Pingo near Tuktoyaktuk, has i d e n t i f i e d a t i l l , possibly deposited subaqueously, under-l a i n by sand and gravel, containing a driftwood log 28,000 200 years old (Be - 49), and overlain by clayey and sandy s i l t s which also contained 4 driftwood dated at 12,000 - 300 years.(S - 6 9 ) . From th i s section he claimed that the Mackenzie Delta area was occupied by an ice sheet which lasted from 25,00 B.P. to 15,000 B.P. and therefore belonged to the la t e -5 Wisconsin period. Mackay has subsequently suggested that the 'Ibyuk T i l l ' 6 may be a 1 p s e u d o - t i l l 1 r e s u l t i n g from the melting of a ground ice sheet. Johnston and Brown have recently reported the occurrence of a pebble-rich s i l t y clay, tentatively i d e n t i f i e d as a t i l l , at a depth of 67.4-70.1 7 metres (221-230 feet), below the surface i n a d r i l l hole near Inuvik. 3 Carsola, A.J. (1954) "Extent of Glaciation on the Continental Shelf i n the Beaufort Sea", Amer. Jour. Science, Vol. 252, pp. 366-371. Mackay, J . Ross (1963) "The Mackenzie Delta area, N.W.T.", Geographical Branch Memoir, No. 8, p. 22. 4 Muller, F. (1962) "Analysis of some Stratigraphic Observations and Radiocarbon Dates from two.Pingos i n the Mackenzie Delta Area, N.W.T.", A r c t i c , Vol. 15, pp. 279-288. 5 Ibi d , p. 285. 6 Mackay, J . Ross (1963) op_. c i t . , pp. 21-22. 7 Johnston, G.H. and Brown, R.J.E. (1965) "Stratigraphy of the Mackenzie River Delta, Northwest T e r r i t o r i e s , Canada", ,,Geol. Soc. Amer. B u l l . , Vol. 76, pp. 103-112. 286 The t i l l was underlain by bedrock and overlain by another 12.5 metres (41 feet) of s i l t y - c l a y , interpreted to have been deposited under glacio-marine or estuarine conditions, followed by 55 metres (180 feet) of de l t a i c sediments deposited by the present: Mackenzie River and i t s p o s t - g l a c i a l 8 ancestors. The delta i c sediments at a depth of 38 metres (125 feet) were dated at 6900 * 110 years (G.S.C. - 54), but no age was suggested for the t i l l . I t would seem u n l i k e l y , however, that the i n t e r v a l between the evacuation of the area by the ice sheet and i t s subsequent occupation by the r i v e r was of any great length, and a late-Wisconsin age for the t i l l could probably be inferred. Craig and Fyles on the other hand have suggested that a coastal s t r i p of the mainland on both sides of the Mackenzie River lay within the maximum l i m i t s of the area affected by the Laurentide ice sheet, but beyond the northwestern l i m i t of the ice through-9 out the Wisconsin gl a c i a t i o n . Opinions as to the date of the la s t ice advance to affect-the Mackenzie Delta area thus range from pre- to late-Wisconsin. The only evidence as to the date of the ice advance which produced the deformation of the sediments on Garry Island comes from the marine f o s s i l s collected from the sand headland areas. These shells are i n an excellent state of preservation, many with their periostracum i n t a c t , and, although no hinged specimens were found, i t i s believed that the f o s s i l s have not been re-worked from an e a r l i e r deposit. The age determination of > 42,600 years (G.S.C. - 562) i s consequently interpreted to be a r e l i a b l e indicator of the minimum time of sand deposition. 8 Ibid. 9 Craig, B.G. and Fyles, J.G. (1960) "Pleistocene Geology of Ar c t i c Canada", Geol. Survey of Canada, Paper 60 - 10, p. 2. 287 The relevance of t h i s age determination to the time of deform-ation of the fine-grained sediments l i e s i n the complete absence of any signs of similar deformation i n these sands. I t may be possible that the frozen sands responded quite d i f f e r e n t l y to the thrusting action of the glacier i c e , and thus exhibit l i t t l e or no deformation, but t h i s i s considered to be most improbable. Moreover, the surfaces of the sand headlands show no evidence of a t i l l deposit unless i t i s : very fine-grained and almost completely stone-free. Such a deposit may formerly have existed and subsequently have been removed during the later submergence of the island, but t h i s again i s considered to be u n l i k e l y . The inference i s therefore,that these sand deposits post-date the deformation of the other sediments on the island and, furthermore, they probably post-date the last ice advance to reach Garry Island. In the absence of additional age determinations, attempts to correlate t h i s ice advance with the Pleistocene chronology established along the southern margins of the Laurentide ice sheet can only be inferred on a tentative basis. The duration of the "Wisconsin state of the Pleistocene sequence i n central North America has been progressively extended, and events believed to have occurred between 70,000 and 10,000 years ago are 10 now assigned to the Wisconsin g l a c i a t i o n . The ice advance responsible for the deformation of the sediments on Garry Island may therefore tentatively be i d e n t i f i e d as early-Wisconsin or possibly pre-Wisconsin i n age. A possible correlation with the Pleistocene sequence - established lb" F l i n t , R.F. (1963) "Status of the Pleistocene Wisconsin Stage i n Central North America", Science, Vol. 139, pp. 402-404. 288 for the central Brooks Range and the A r c t i c Slope of Alaska would be with the Sagavanirktok g l a c i a t i o n , considered to be pre-Wisconsin ( I l l i n o i a n ) 11 i n age. The next i d e n t i f i a b l e episode i n the evolution of Garry Island, following the withdrawal of t h i s ice sheet, was the deposition of the sands and gravels exposed i n the sand headlands on the north side of the island. Granulometric analyses of these deposits, and the s i m i l a r i t y i n radiocarbon dates, indicate that they are broadly contemporaneous with the terraces described by Fyles along the west side of Richards Island and along the east bank of the East Channel of the Mackenzie River which were dated at > 40,000 years (G.S.C. - 709). Fyles has also dated marine s h e l l s , found i n a lake shore bluff located 10 miles south of Kendall Island, which yielded an age of > 42,600 years (G.S.C. - 690), and Mackay obtained a radiocarbon date of >38,000 years (I - 482) from material 12 contained i n a terrace at the mouth of the Kugaluk River. The sand and gravel deposits on Garry Island also appear to be similar i n age and 13 texture to part of the Gubik formation of northern Alaska. The s t r i k i n g d i s s i m i l a r i t y between the texture of these sands and gravels and any other sediments on Garry Island, suggests that they are not simply residual accumulations derived through long-continued erosion 11 Detterman, R.L., Bowsher, A.L., and Dutro, J.T., J r . , (1958) "Glaciation on the A r c t i c Slope of the Brooks Range, Northern Alaska", A r c t i c , Vol. 11, pp. 43-61. Porter, S. (1964) "Geologic history of the Anaktuvuk Pass area, Brooks Range, Alaska", Amer. Jour. Science, Vol..262, pp. 446-460. 12 Mackay, J . Ross (1963) op_. c i t . pp. 38-39. 13 Coulter, H.W., Hussey, K.M., and O'Sullivan, J.B. (1960) Radiocarbon dates r e l a t i n g to the Gubik Formation, Northern Alaska", U.S. Geol. Survey, Professional Paper, No. 400-B pp. B350-B351. 289 of the island. Rather, an extraneous source appears to be indicated. The most probable o r i g i n of the sands i s that they represent deposits brought down by streams, following the-withdrawal of the i c e , and deposited under marine conditions when the,relative l e v e l of the sea was at least 9-15 metres (30-50 feet) higher than at present. These deposits may have accumulated i n the form of large sandspits, but the discontinuous nature of the sand headlands probably also represents the effects of subsequent erosion. The age of the sands, > 42,600 years (G.S.C. - 562) , may represent an i n t e r s t a d i a l period i n the early-Wisconsin or an even e a r l i e r i n t e r -g l a c i a l (Sangammon ?) period. Garry Island probably lay beyond the northwest l i m i t s of the Laurentide ice sheet during the late-Wisconsin period, although the adjacent mainland and the area of the modern Mackenzie Delta to the south may have been occupied by ice at t h i s time. The fresh, unaltered appearance of a number of large e r r a t i c s on the i s l a n d , however, may indicate deposition by ice of late-Wisconsin age, but since they a l l occur w e l l below the marine l i m i t , they may have been rafted i n by d r i f t i n g icebergs. THE EXTENT OF THE MARINE TRANSGRESSION The existence of a series of weakly developed raised strand-l i n e features, up to heights of 46 metres (lSO^feet) above sea l e v e l , indicates that the island was completely submerged by the p o s t - g l a c i a l r i s e i n sea l e v e l . Moreover, the manner i n which these elevated shore-lines are developed around the major elements of the topography, with pronounced re-entrants along the valleys and depressions, suggests that the.pattern was produced by the drowning of a pre-existing topography. 290 This i n turn eould have been developed during pre-late-Wisconsin times. This submergence was then followed by a progressive emergence of the land r e l a t i v e to the l e v e l of the sea; a process which was halted intermittently at elevations of approximately 38, 30.5, 23, 15 and 7.5 metres (125, .100, 75, 50 and 25 feet) above present sea l e v e l . None of these periods of s t a n d s t i l l were.of any great duration as witnessed by the weak expression of the strand-lines, even allowing for the obscuring effects of later s o l i f l u c t i o n processes. During these h a l t s , however, small sandspits or bars, composed of residual accumulations of material derived through erosion of the-adjacent b l u f f s , were constructed across the mouths of some of the valleys. The formation of small lagoons or f l a t s , behind these spits or bars, provided numerous si t e s for the develop-ment of tundra polygons which have produced the d i s t i n c t i v e stepped p r o f i l e s of the. present stream courses. The larger embayments i n the.elevated shorelines were occupied by large freshwater lakes i n which a series of lacustrine sediments were deposited. I t i s impossible, from evidence presently available, to establish an absolute chronological scale for these fluctuations i n the r e l a t i v e positions of the land and sea. I f the hypothesis that the tundra polygons are intimately related to the shorelines i s correct however, radiocarbon dates from peat exposed i n some of the high-centres polygons provide certain clues. Samples of peat taken from a depth of about 3.7 metres (10.0 feet) below thesurface of a high-centred polygon on the northwest coast of the island yielded an age of 11,700 * 250 years (S - 276). These polygons are developed i n a small depression floored by gravels which occur at an elevation of about 23 metres (75 feet) above sea l e v e l . As the peat i n these polygons shows no signs of having been buried, i t 291 would seem l o g i c a l to assume that the r e l a t i v e l e v e l of the sea could not have exceeded the 38 metre (75 foot) elevation since the.peat began to accumulate. This implies a minimum date of >11,500 years for the 38 metre (75 foot) raised strand-line. No age determinations were made for any organic material occurring at higher elevations, and therefore no dates can be suggested for the higher strand-lines. Similar reasoning can be applied however to age determinations made i n peat deposits occur-ring at lower elevations. A sample of peat occurring i n the basal part of a series of lacustrine sediments, which rest on beach gravels at an elevation of approximately 7.5 metres (25 feet) above sea l e v e l , yielded an age of 10,330 * 150 years (G.S.C. - 516). These lake sediments are overlain by a layer of sand and pebbly gravel containing iron-stained wood fragments dated at 9730 * 160 years (G.S.C. -575). These figures suggest that the emergence of the island continued u n t i l the r e l a t i v e l e v e l of the sea. stood only 7.5 metres (25 feet) higher than at present at least 10,000 - 10,500 years ago. This emergence was then temporarily halted by a period of renewed submergence which affected areas up to elevations of approximately 10.5 metres (35 feet) above present sea l e v e l , (the a l t i t u d e of the upper surface of the sand and pebble gravel). The radiocarbon date of 9730 years may indicate the date of th i s submergence but i t i s possible that the wood has been reworked from e a r l i e r sediments. Since that time there has been a renewed emergence of the island which has probably continued through to the present day. Another age determin-ation made on a peat sample occurring i n a high-centred polygon at an elevation of approximately 3.5-4.0 metres (12-13 feet) above sea le v e l yielded an age of 4,120 - 130 years (G.S.C. - 517). The absence of any signs of b u r i a l on these polygons provides a minimum date for the 292 afore-mentioned period of submergence, and further suggests that the r e l a t i v e l e v e l of the sea has not exceeded t h i s elevation since that time. Mackay has indicated that the last episode i n the changing relationship of the land and sea i n the delta area may have been a s l i g h t 14 submergence of 3, 6 or more metres (10, 20 or more fee t ) . There i s no conclusive evidence that a similar submergence, has affected Garry Island. The existence of a channel outlet from the. large lake i n one of the sand headlands on the north side of the island may be relevant however. The outlet, several tens of metres long and maintaining a depth of 6 metres (20 feet) almost to the present coastline, may indicate the submergence of a channel graded to a former, lower sea l e v e l . . A l t e r n a t i v e l y , the depth may be attributed to thermokarst melting of the underlying sediments. Throughout t h i s period of submergence, the mean temperature of the sea water must have been close to 0°C, for i t appears that very l i t t l e thawing of the underlying permafrost took place. The present outline of Garry Island has been produced by coastal recession associated with the present sea l e v e l . Stratigraphic, geomorphic and h i s t o r i c evidence shows that a considerable amount of coastal recession has occurred i n recent times. I f the latest episode i n the fluctuations of sea l e v e l has been a period of submergence, t h i s would have contributed to the continued recession of the shoreline by increasing the offshore depths of water. As the land surface emerged from beneath the sea i t was sub-jected to intense thermal stresses produced by the seasonal temperature variations of the a r c t i c climate. Contraction of the ground upon freezing 14 Mackay, J. Ross (1963) ££. c i t . , p. 41. 293 during the winter produced networks of fr o s t cracks, and recurrent fracturing at the same l o c i led to the formation of ice-wedges. On the lower, f l a t t e r ground, the networks of ice-wedges led to the development of systems of orthogonal, s l i g h t l y - o r i e n t e d , tundra polygons and, with increasing passage of time, these polygons were transformed from low- to high-centred forms. On steeper, drier s i t e s the newly-exposed ground surface was probably covered by a miniature network of similar f r o s t or desiccation cracks. These networks have been accentuated to produce the microrelief of the t y p i c a l earth hummocks. Once established, the sub-sequent development of these earth hummocks was largely the resu l t of d i f f e r e n t i a l f r o s t action and s o l i f l u c t i o n as evidenced by the buried organic layer at depth. CURRENT GEOMORPHOLOGICAL PROCESSES Throughout the period of post - g l a c i a l time, the geomorphic processes, with the possible exception of f l u v i a l action, shaping and remoulding the surface topography of Garry Island have probably remained the same, though they may have varied i n their r e l a t i v e intensity from time to time. Observations of the geomorphic processes currently operating on the island reveal the dominant influence of the underlying permafrost on the rates at which they operate. Continued recession of the coastline i s constantly re-shaping the outline of the island. Actively retreating parts of the coast are presently r e s t r i c t e d to sections of the northwest coast of the isl a n d , and parts of the sand headlands which have not been protected by the form-ation of sandspits. Observations on the processes involved i n the retreat of the c l i f f s demonstrate the importance of permafrost and especially i t s 294 composition. The dominant process by which material i s detached from the c l i f f face i s a thermal erosion of the frozen ground, which i s extremely susceptible to the sun's rays when exposed. Melting of the i c e , which acts as a strong cementing agent, leaves the sediments unsupported and often at angles higher than the angle of repose, so that they slump to the^base of the c l i f f s . The rate of retreat i s profoundly affected by the composition of the frozen sediments. Where the sediments are coarse-grained, and have a low ice content, the rate of retreat i s usually less than one metre (3 feet) a year, but where the frozen ground contains large masses of seg-regated ground ice the rate may be as high as 30 metres (98 feet) per annum. The major role of wave action, on a short term basis, appears to be i n the removal of the slumped material from the base of the c l i f f s , and th i s i s p a r t i c u l a r l y effective during storm surges. L o c a l l y , however, and on a long term basis, direct undercutting by wave action may assume the most important role, i n the c l i f f retreat process. The exposure and subsequent melting of the permafrost i s also responsible for the production of one of the most d i s t i n c t i v e landforms found on Garry I s l a n d - the mudslump. These are amphitheatre-like depressions produced by the melting out of tabular bodies of segregated ground ice. Once again, the rate of retreat of the mudslump headwall i s profoundly influenced by the composition of the frozen ground. The highest rates of recession, i n excess of 10 metres (33 feet) per annum, occur i n frozen sediments with a high ice content, and as the proportion of mineral s o i l i n the headwall decreases the rate of recession decreases to a point where, i f no ice i s exposed, l i t t l e or no retreat takes place. The r a t i o of ice to mineral s o i l i n the frozen ground also determines, i n part, the longevity of the mudslump a c t i v i t y , and the degree to which the scars 295 persist as recognizable features i n the landscape. In general, the higher the ice content the deeper i s the depression produced by melting, and the longer t h i s depression persists before i t i s obscured by vegetation. The longevity of the period of mudslump a c t i v i t y i s also affected by the degree to which thawed material i s removed from the base of the headwall, thus permitting, as i n the. coastal recession processes, the maintenance of fresh exposures of ground ice and continued thermal erosion. Where the ice contents are high, releasing large quantities of excess water upon melting, the thawed debris i s e f f i c i e n t l y removed by strong mudflows, but i f the amount of water released i s small the thawed debris may accumulate at the foot of the headwall. The action of running water as an active geomorphic agent on Garry Island i s limite d , r e f l e c t i n g a combination of low p r e c i p i t a t i o n , small catchment areas and the presence of an almost continuous cover of vegetation. The exi s t i n g stream courses are poorly defined, and serve primarily as conduits for surface runoff derived from melting snow and the thawing of the active layer during the spring and early summer. Throughout the remainder of the summer, these channels are kept moist by seepage from the thawing ground, but surface runoff i s generally lacking except for a short time following prolonged periods of heavy r a i n f a l l . The r a p i d i t y of surface runoff after such r a i n f a l l r e f l e c t s the presence of permafrost at shallow depths beneath the ground surface. The fact that the i l l - d e f i n e d channels are located i n a number of well-defined 'V'-shaped valleys may bear witness to the greater significance of the action of running water i n the past. Additional evidence of the present insignificance of running water comes from the fact that the fl o o r s of many of the stream courses are being raised by the accumulation of organic material. The most important, single c r i t e r i o n r e s t r i c t i n g the effectiveness of running water today, however, appears to be the continuous vegetation cover. Once t h i s cover i s breached, exposing the underlying frozen ground, and especially where t h i s has a high ice content, a considerable amount of thermokarst erosion may be accomplished by only small quantities of water i n a very short time. The Concept of a P e r i g l a c i a l Morphogenetic Region. F i n a l l y , from these observations an attempt can be made to examine the concept of a p e r i g l a c i a l morphogenetic region. According to P e l t i e r , the characteristics of the various geomorphic processes operating i n p e r i g l a c i a l regions are strong mass movement, moderate to strong wind action and a weak effect of running water.^ Although i t was not studied in d e t a i l , s o l i f l u c t i o n i s undoubtedly one of the most widespread processes moulding the landscape of Garry Island. Reference has been made to i t s importance i n the p a r t i a l o b l i t e r a t i o n of raised strand-line features and the growth and gradual downslope movement of earth hummocks. The l o c a l presence of s o l i f l u c t i o n lobes, and the ubiquitous occurrence of a buried organic layer, t e s t i f y to i t s significance on a l l slopes of the island. The inclusion of mudslumps and mudflows supports the contention that the phenomena of mass movement are the most important geomorphic processes operating i n p e r i g l a c i a l regions. Certain q u a l i f i c a t i o n s are necessary however. Observations of some of these processes on Garry Island demon-strate the influence of the composition of the underlying frozen ground P e l t i e r , L.C. (1950) "The geographical cycle i n p e r i g l a c i a l regions as i t i s related to climatic geomorphology", Ann. Assoc. Amer. Geog., Vol. 40, p. 215. 297 on the rate at which they operate. If additional credence i s to be given to the concept of morphogenetic regions on the basis of quantitative measurements, there w i l l be a d e f i n i t e need to distinguish between bedrock and areas of unconsolidated sediments and between areas of wet and dry permafrost. Otherwise, there w i l be as much va r i a t i o n between the rates of operation of mass movement processes within p e r i g l a c i a l regions as between contrasting climatic environments. In t h i s respect, P e l t i e r ' s paper i s deficient since i t infers the presence of bedrock, and makes no reference to the effects of changes i n the composition of the frozen ground. Furthermore, even i n areas which are underlain by a r e l a t i v e l y homogeneous substrate, and especially where th i s consists of unconsolidated sediments, there i s s t i l l a need to consider the areal scale problem. This problem i s very w e l l i l l u s t r a t e d where thermal erosion i s an important process since variations i n aspect may also contribute to s i g n i f i c a n t l y wide ranges i n the rates at which t h i s process operates. A l l of these factors must be given due consideration i n any attempt to establish the concept of morphogenetic regions on a quantitative basis. Wind action on Garry Island plays a very minor role despite i t s coastal location and exposure to strong winds. This possibly r e f l e c t s the presence of a continuous vegetation cover, and there i s a further need i n P e l t i e r ' s paper to d i f f e r e n t i a t e between vegetated and unvegetated areas. The vegetation cover i s probably also the most s i g n i f i c a n t factor contrib-uting to the r e s t r i c t e d e f f i c i e n c y of f l u v i a l action on the i s l a n d , together with the absence of perennial streams. The rapid thermokarst erosion of the frozen ground by only small amounts of water, however, probably indicates an underestimation of the role of running water i n the hierarchy of r e l a t i v e importance of geomorphic processes. Since the 298 existence of a number of well-defined 'V'-shaped valleys on the island may bear witness to the greater significance of the action of running water i n the past, there i s also a need to recognize an additional scale problem involving the time period over which these present geomorphic processes have operated at the same absolute or r e l a t i v e intensity as at present. The action of waves was excluded from P e l t i e r ' s paper because 16 of i t s a p p l i c a b i l i t y only to coastal locations. On Garry Island, the recession of the coastline, especially where i t exposes masses of ground i c e , represents one of the more spectacular aspects of the operation of geomorphological processes. Furthermore, the stratigraphic, geomorphic and h i s t o r i c evidence indicates that coastal recession has substantially altered the outline of the island throughout recent geological time. I b i d . , p. 218. 299 BIBLIOGRAPHY Benninghof, W.S. "Interaction of Vegetation and S o i l Frost Phenomena", A r c t i c , Vol. 5 (1952), pp. 34-44. B i r d , J . Brian. The Physiography of. A r c t i c Canada: with Special Reference  to the Area South of Parry Channel. ".'•Baltimore, Maryland: The Johns Hopkins Press, 1967. Black, R.F. "Polygonal patterns and ground conditions from a e r i a l photo-graphs", Photogrammetric Engineering, Vol. 18 (1952), pp. 123-134. Blackwelder, E. "Mudflow as a geologic agent i n semi-arid Mountains", Geol. Soc. Amer. B u l l . , Vol. 39 (1928), pp. 465-480. Bredin,. T.F. "Whale Island and the Mackenzie Delta: charted errors and unmapped discoveries.1789 to 1850", A r c t i c , Vol. 15 (1962), pp. 51-65. Brown, J . " S o i l s of the Okpilak River Region, Alaska", . U.S. Army Materiel Command, Cold Regions Research & Engineering Laboratory, Research  Paper 188, 1966, 49 p. Brown,. R.J,E. "A Review of permafrost investigations i n Canada", The  Canadian Geographer, No. 6 (1962), pp. 162-165. Brown, R.J.E. "Influence of Vegetation on Permafrost", Proc: Permafrost  International Conference, Lafayette, Indiana, November, 1963. Nat. Acad. S c i . - Nat. Research Council,, Washington.Publication, No. 1287, 1966, pp. 20-25. Brown, R.J.E. "Relation between mean annual a i r and ground temperatures i n the. permafrost region of Canada", Proc; . Permafrost International  Conference, Lafayette, Indiana, November, 1963. Nat.Acad. S c i . - Nat. Research Council, Washington Publication, No. 1287, 1966, pp. 241.246. Bryan, K. "Gully gravure - A method of slope retreat", Journal of Geo-morphology, Vol. 3 (1940), pp. 87-107. Carsola, A.J. "Extent of Glaciation on the Continental Shelf i n the Beaufort Sea", Amer. Jour. Science, Vol. 252 (1954), pp. 366-371. Conrad, V. "Polygon Nets and their Physical Development", Amer. Jour. Science, Vol. 244 (1946), pp. 277-296. Cook, F.W. "Near surface s o i l temperature measurements at Resolute Bay, Northwest T e r r i t o r i e s " , A r c t i c , Vol. 8 (1955), pp. 237-249. Coulter, H.W., Hussey, K.M. and O'Sullivan, J.B. "Radiocarbon dates r e l a t i n g to the Gubik Formation, Northern Alaska", U.S. Geol. Survey, Professional Paper, No. 400-B (1960), pp. B350-B351. 300 Craig, B.G. and Fyles, J.G. "Pleistocene Geology of A r c t i c Canada", Geol. Survey of Canada, Paper 60-10, 21 p. Croxton, F.E. and Crowden, D.J. Applied General S t a t i s t i c s . Englewood C l i f f s , N.J„: Prentice-Hall, Inc., Second Edi t i o n , 1955. Detterman, R.L„, Bowsher, A.L. and Dutro J r . , J.T. "Glaciation on the A r c t i c Slope of the Brooks Range, Northern Alaska", A r c t i c , Vol. 11 (1958), pp. 43-61 Everett, K.R. "Slope Movement and Related Phenomena". Chap. 12 i n N.J. Wilimovsky (Ed.), Environment of the Cape Thompson Region, Alaska. Washington, D.C.: U.S. Atomic Energy Commission, Division of Technical Information, 1966, pp. 175-213. Farrand, W.R. and Gajda, R.T. "Isobases on the Wisconsin marine l i m i t i n Canada", Geographical B u l l e t i n , No. 17 (1962), pp. 5-22. F l i n t , R.F. "Status of the Pleistocene Wisconsin Stage i n Central North America", Science, Vol.. 139 (1963), pp. 402-404. Fra n k l i n , S i r John. Narrative of a second expedition to the shores of the  Polar Sea i n the years 1825, 1826 and.1827. Philadelphia: Carey, Lea and Carey, 1828. Geological Association of Canada. G l a c i a l Map of Canada. 1958. Geological Survey of Canada. Permafrost i n Canada. Map 1246A, 1967. Griggs, R.F. "The, vegetation of the.Katmai D i s t r i c t " , Ecology, Vol. 17 (1936), pp. 380-417. Hopkins, D.M. and Sigafoos, R.S; "Frost action and vegetation patterns on Seward Peninsula, Alaska", U.S. Geol. Survey, B u l l e t i n , No. 974-C (1951) ,. pp. 51-101. Hopkins, D.M. and Sigafoos, R.S. "Role of f r o s t thrusting i n the formation of tussocks", Amer. Jour. Science, Vol. 252 (1954), pp. 55-59. Jenness, J.L. "Permafrost i n Canada", A r c t i c , Vol. 2,(1949), pp. 13-27. Jenness, J.L. "Erosive Forces i n the Physiography of Western A r c t i c Canada", Geographical Review, Vol. 42 (1952), pp. 238-252. Johnson, A.W. et. a l . "Vegetation and F l o r a " . Chap. 14 i n N.J. Wilimovsky (Ed.), Environment of. the Cape Thompson Region, Alaska. Washington, D.C.: U.S. Atomic Energy Commission, Division of Technical Information, 1966. Johnston, G.H. and Brown, R.J.E. "Stratigraphy of the Mackenzie River Delta, Northwest;Territories, Canada", Geol. Soc. Amer., B u l l . , Vol. 76 (1965), pp. 103-112. 301 Lachenbruch, A„H„ "Mechanics of Thermal Contraction Cracks and Ice-Wedge Polygons i n Permafrost", Geol. Soc. Amer., Special Paper, No. 70 (1962), 69 p. Lamothe, C. and St-Onge, D. "A note on a p e r i g l a c i a l erosional process i n the Isachsen area, N.W.T.", Geographical B u l l e t i n , No. 16 (1961), pp. 104-113. Lef f i n g w e l l , E. de K. "Ground Ice-Wedges. The Dominant form of Ground Ice on the North Coast of Alaska", Journal of Geology, Vol.. 23 (1915), pp. 635-654. Leffin g w e l l , E. de K. "The Canning River Region, Northern Alaska", U.S. Geol. Survey, Professional Paper, No. 109 (1919), 251 p. MacCarthy, G.R. "Recent changes i n the shoreline, near Point Barrow, Alaska", A r c t i c , Vol. 6 (1953), pp. 44-51. Mackay, J . Ross. "Mackenzie Deltas - A Progress Report", The Canadian  Geographer, No. 7 (1956), pp. 3-7. Mackay, J . Ross. "Deformation by Glacier Ice at Nicholson Peninsula, N.W.T., Canada", A r c t i c , Vol. 9 (1956), pp.. 219-228. Mackay, J . Ross. "A subsurface organic layer associated with permafrost i n the Western A r c t i c " , Geographical Branch, Ottawa, Geographical Paper, No. 18 (1958), 21 p. Mackay, J. Ross. "The Anderson River map area, N.W.T.", Geographical Branch, Ottawa, Memoir, No. 5 (1958), 137 p. Mackay, J. Ross. "Glacier Ice Thrust Features of the Yukon Coast", Geographical B u l l e t i n , No. 13 (I960),, pp. 5-21. Mackay, J . Ross, Mathews, W.H. and MacNeish, R.S. "Geology of the Engig-stciak archeological s i t e , Yukon Territory", . A r c t i c , Vol. 14 (1961), pp. 25-52. Mackay, J. Ross. "Notes on the shoreline recession along the coast of the Yukon Territory", . A r c t i c , Vol. 16 (1963), pp. 195-197. Mackay, J. Ross. "The Mackenzie Delta area, N.W.T.", Geographical Branch, Ottawa, Memoir, No. 8 (1963), 202 p. Mackay, J . Ross. "Gas-domed mounds i n permafrost, Kendall Island, N.W.T.", Geographical B u l l e t i n , Vol. 7 (1965), pp. 105-115. Mackay, J . Ross. "Segregated Epigenetic Ice and Slumps i n Permafrost, Mackenzie Delta area, N.W.T.", Geographical B u l l e t i n , Vol. 8 (1966), pp. 59-80. 302 Mackay, J. Ross, and Stager, J.K. "Thick t i l t e d beds of segregated i c e , Mackenzie Delta area, N.W.T.", Biuletyn Peryglacjalny, No. 15 (1966), pp. 39-43. Mackay, J. Ross. "Underwater patterned ground i n a r t i f i c i a l l y drained lakes, Garry Island, N.W.T.", Geographical B u l l e t i n , Vol. 9 (1967), pp. 33-44. Mackay, J . Ross. "Permafrost depths, Lower Mackenzie Valley, Northwest T e r r i t o r i e s " , A r c t i c , Vol. 20 (1967), pp. 21-26. Mackenzie, A. Voyage from Montreal on the River St. Laurence through the  continent of North America to the Frozen and P a c i f i c Oceans i n the  years 1789 and 1793. London and Edinburgh: T. Cadell, Jun. and W. Davies, Cobbett and Morgan, and W. Creech, 1801. Reprinted e d i t i o n , March of America Facsimile Series, Number 52. Ann Arbor, Michigan: University Microfilms Inc., 1966. Manning, T.H. "Narrative of a Second Defense Research Board Expedition to Banks Island, with notes on the country and i t s history", A r c t i c , Vol. 9 (1956), pp. 3-77. Muller, F. "Analysis of some Stratigraphic Observations and Radiocarbon Dates from two Pingos i n the Mackenzie Delta Area, N.W.T.", A r c t i c , Vol. 15 (1962), pp. 279-288. Muller, S.W. Permafrost or permanently frozen ground and related engin-eering problems. Ann Arbor, Michigan: J.W. Edwards, Inc., 1947. O'Neill, J . J . "Geology of the A r c t i c Coast of Canada, west of the Kent Peninsula"". Report of the Canadian. A r c t i c Expedition, 1913-1918, Vol. XI, Geography and Geology, Part A. Ottawa: King's P r i n t e r , 1924. P e l t i e r , L.C. "The geographical cycle i n p e r i g l a c i a l regions as i t i s related to climatic geomorphology", Ann. Assoc. Amer. Geog., Vol. 40 (1950), pp. 214-236. Pewe, T.L. "Sand-Wedge Polygons (Tesselatipns) i n the McMurdo Sound Region, Antarctica - A Progress Report", Amer. Jour. Science, Vol. 257 (1959), pp. 545-552. Pewe, T.L. "Ice-Wedges i n Alaska - C l a s s i f i c a t i o n , D i s t r i b u t i o n and Climatic Significance". Proc: Permafrost International Conference, Lafayette, Indiana, November, 1963. Nat., Acad. S c i . - Nat. Research Council, Washington Publication, No. 1287 (1966), pp. 76-81. Polunin, N.V. Circumpolar. A r c t i c F l o r a . Oxford: Clarendon Press, 1959. P o r s i l d , A.E. "The vascular plants of the Western Canadian A r c t i c Archipelago", Nat. Mus. Can. B u l l . , No. 146 (1955), 218 p. Porter, S. "Geologic history of the Anaktuvuk Pass area, Brooks Range, Alaska". Amer. Jour.'Science, Vol. 262 (1964), pp. 446-460. 303 Raup, H.M. "The structure and development of turf hummocks i n the Mesters Vig D i s t r i c t , Northeast Greenland", Meddelseer om Gr^nland, Bd. 166 (1965), pp. 1-112. Richards, H.G. "Postglacial marine submergence of A r c t i c North America with special reference to the Mackenzie Delta", Amer. P h i l . Soc. Proc., Vol. 94 (19.50), pp. 31-37. Schumm, S.A. "Evolution of drainage systems and slopes i n badlands at Perth Amboy, New Jersey", Geol. Soc. Amer. B u l l . , Vol. 67 (1956), pp. 597-646. Sharp, R.P. "Mudf low Levees", Journal of Geomorphology, Vol... 5 (1942), pp. 222^227. Sharp, R.P. " S o i l structures i n the St. E l i a s Range, Yukon Ter r i t o r y " , Journal of Geomorphology, Vol. 5 (1942), pp. 275-301. Sharp, R.P. and Nobles, L.M. "Mudflow of 1941 at. Wrightwood, Southern C a l i f o r n i a " , Geol. Soc. Amer. B u l l . , Vol. 64 (1953), pp. 547-560. Sharpe, C.F.S. Landslides and Related Phenomena. New Jersey: Pageant Books, Inc., 1938. Sim, V.W. "Maximum post- g l a c i a l marine submergence i n northern M e l v i l l e Peninsula", A r c t i c , Vol. 13 (1960), pp. 178-193. Stager, J.K. "Alexander Mackenzie's exploration of the Grand River", Geographical B u l l e t i n , Vol.. 7 (1965), pp. 213-241., St-Onge, D. "La geomorphologie de L ' l l e E l l e f Ringnes, T e r r i t o i r e s du Nord-Ouest, Canada", Geographical Branch, Ottawa, Geographical Paper, No. 38 (1965), 58 p. Strahler, A.N. and Koons, D. "Objective and Quantitative Methods of Terrain Analysis", U.S. Dept. of the Navy, Office of Naval Research, Geography Branch, Project NR 387-021, Contract Nonr 266-50 (1960), 51 p. Thompson, H.A. "Freezing and Thawing Indices i n Northern Canada", Proc. of the F i r s t Canadian Conference, on Permafrost, Ottawa, A p r i l 17-18', 1962, (1963), pp. 18-42. Thoroddsen, Th. "An account of the physical geography of Iceland with special reference to plant l i f e " , In Kolderup Rosenvinge, L. and Warming, E. (Eds.), The Botany of Iceland. Vol. 1. Copenhagen, 1914. Twidale, C.R. "Erosion of an a l l u v i a l bank at Birdwood, South A u s t r a l i a " , Z e i t s c h r i f t fur Geomorphologie, Bd. 8 (1964), pp. 189-211.. 304 Tyrtikov, A.I. "Perennially Frozen Ground and Vegetation", P r i n c i p l e s of  Geocryology, (Permafrost Studies), Part I , General Geocryology. Moscow: Acad. S c i . U.S.S.R., 1959, pp. 399-421. Translation by R.J.E. Brown, National Research Council of Canada, Technical Translation 1163, 1964, pp. 1-34. Wallace, R.E, "Cave-in lakes i n the Nabesna, Chisana and Tanana r i v e r v a l l e y s , eastern Alaska", Journal of Geology, Vol. 56 (1948), pp. 171-181. Washburn, A.L. "Reconnaissance Geology of portions of V i c t o r i a Island and adjacent sections i n A r c t i c Canada", Geol. Soc. Amer. Memoir, No. 22 . (1947), 142 p. Washburn, A.L. " C l a s s i f i c a t i o n of patterned ground and review of suggested origins", Geol. S o c Amer. B u l l , , Vol. 67 (1956), pp. 823-866. Washburn,. A.L., Smith, D.D. and Goddard, R.H. "Frost Cracking i n a Middle-Latitude Climate", Biuletyn Peryglacjalny, Nr. 12 (1963), pp. 175-189. 305 APPENDIX I A. LIST OF VASCULAR SPECIES FOUND ON GARRY ISLAND 1. PTERIDOPHYTA 2. SPERMATOPHYTA Equisetaceae Monocotyledonae Graminae Equisetum arvense Cyperaceae Dicotyledonae Juncaceae Salicaceae Agropyron latiglume Alopecurus alpinus Arctagrostis l a t i f o l i a Arctophila f u l v a Calamagrostis deschampsioides Calamagrostis lapponica Calamagrostis purpurascens Dupontia f i s h e r i Elymus.arertarius ssp. moll i s Pba a r c t i c a P u c c i n e l l i a andersonii P u c c i n e l l i a phryganodes P u c c i n e l l i a vaginata Trisetum spicatum Carex a q u a t i l i s Carex bigelowii Carex chordorrhiza Carex membranacea Carex vaginata Eriophorum angustifolium Eriophorum scheuchzeri Juncus alpinus Luzula confusa S a l i x alaxensis S a l i x arctophila S a l i x glauca var. niphoclada S a l i x phlebophylla S a l i x pulchra S a l i x r e t i c u l a t a 306 Betulaceae Alnus crispa Betula glandulosa Polygonaceae Oxyria dignya Polygonum bistortum Polygonum viviparum Rumex arcticus Garyophyllaceae Arenaria peploides Arenaria r u b e l l a Cerastium beeringianum Melandrium aff i n e S t e l l a r i a humifusa S t e l l a r i a longipes Ranunculaceae Papaveraceae Gruciferae Crassulaceae .Saxifragaeeae Rosaceae Anemone p a r v i f l o r a Anemone ri c h a r d s o n i i Caltha p a l u s t r i s ssp. a r c t i c a Ranunculus eymbalaria var. alpinus Ranunculus gmelinii Ranunculus p a l l a s i i Papaver keelei Braya humilis ssp. ar c t i c a Cardamine.digitata Cardamine pratensis Descurainia sophoides Draba gla b e l l a Erysimum inconspicuum Parrya nudicaulis Sedum rosea.ssp. int e g r i f o l i u m Parnassia kotzebuei Saxifraga cernua Saxifraga hircuius Saxifraga radiata Dryas i n t e g r i f o l i a P p t e n t i l l a p a l u s t r i s Rubus chamaemorus Leguminosae Astragalus alpinus Hedysarum alpinum americanum Hedysarum mackenzii L.athyrus maritimus Lupinus arcticus Oxytropis maydelliana Empetraceae Empetrum nigrum 307 Onagraceae Hippuridaceae Umbelliferae Ericaceae Primulaceae Plumbaginaceae Gentianaceae Polemoniaceae Scrophulariaceae Compositae Epilobium l a t i f o l i u m Hippuris vulgaris Bupleurum americanum Conioselinum c n i d i i f o l i u m Pyrola grandif lora Arctostaphylos rubra Gassiope tetragona Ledum palustre ssp. decumbens Rhododendron lapponicum Vaccinium uliginosum var. alpinus Vaccinium v i t i s - i d a e a Primula egaliksensis Armeria maritima. ssp. a r c t i c a Gentiana arctophila Lomatogonicum rotatum Polemonium acutiflorum C a s t i l l e j a p a l l i d a ssp. elegans Pedicularis capitata Pedicularis lanata Pedicularis l a n g s d o r f i i Pedicularis sudetica A c h i l l e a borealis Arnica louiseana ssp. f r i g i d a Artemisia t i l e s i i Erigeron humilis M a t r i c a r i a ambigua Petasites arctieus Petasites f r i g i d u s Saussurea a n g u s t i f o l i a Senecio atropurpureus Senecio.congestus Senecio lugens B. LIST OF BRYOPHYTES FOUND ON GARRY ISLAND A b i e t i n e l l a abietina (Schwaegr.) C.M. (Thuidium abietinum) Aulacomnium palustre (Hedw.) Schwaegr. Aulacomnium turgidum (Wg.) Schreb. Bryum ovaturn Jur. Galliergon Richardsonii (Mitt.) Kindb. Cinclidium stygium Sw. Dicranum elongatum Schleich. c. f r . Dicranum fuscescens Turn. Drepanocladus revolvens (Sw.) Kindb. k o l l . Hylocomium alaskanum (Lesq. & James) Kindb. Hypnum callichroum (Brid.) Br. & Sch. Pogonatum alpinum Hedw. P s i lop Hum cavif olium (Wils.). Hag. Sphagnum squarosum Pers. Sphagnum Warnstorf ianum DR. (S. W.arnstorfii Russ. non Roll) Tomenthypnum nitens (Schreb.). Loeske Tritomaria quinquedentata (Huds.) Buch N X S . NO. 107 & U7 145° AERONAUTICAL IN ^ AERONAUTICAL SYMBOLS AERONAUTICAL SYMBOLS A E R O D R O M E S AERODROMES WITH FACILITIES AERODROMES WITH EMERGENCY OR NO FACILITIES i — [ O T T A W A ! 357 LH 88 T O W E R 278 126.18mc G C A L A N D W A T E R L A N D W A T E R 0 •0- Civil O A o W Joint Civil And Military © ® Military A E R O D R O M E D A T A LAND WATER 357 Elevation In Feet 00 Elevation In Feet (Sea Level) L Minimum Lighting: Obstruction, Boundary Or Runway Lights, And Lighted Wind Indicator H Hani Surfaced Runway 88 Length Of Longest Runway ln Hundreds of Feet S 99 Minimum Lighting: Obstruction, Boundary Or Channel Lights, And Lighted Wind Indicator Sheltered Take-Off Area Length Of Longest Alighting Run 9.900 Feet Or Longer iPATRICIA BAY! Ob LS 99' When specific infnrmation prelaining to landing <laca is lacking the respective character will bfl replaced by a dash ( —) [ O T T A W A ; Customs Aerodrome T O W E R Call "Ottawa Tower" 278 I26.i8mc Control Tower Frequencies G C A Ground Controlled Approach System R A D I O F A C I L I T I E S Use Of The Word "Radio" Within The Box Indicates Voice Facilities. All Frequencies Are Kilocycles Unless Otherwise Stated Radio Range — time of weather broadcast 0 in minutes after the hour T O R O N T O R A D I O 368 Y Z 25 &55" Non-Directional Radio Beacon.Times shown for Marine NDB (marine in top line of box) indicate the start of one minute broadcasts in each hour. VHF Omni-Directiona Radio Range (with voice) Compass rose is oriented on mag-netic north. O B R A N D O N 233 B R T-L7 Radio Communication Station (with voice) Radio Communication Station (without voice) -cs-EMBARRAS RADIO 5420 -CS-HARMON 6475 AKX2 VOR-Commereia Station Radio Fan Marker / -Beacons 1 Broadcast im 100 watts r-BS-, C B O 910 - H E M M I N G F 0 R D C A M A N 0 Radio Direction Finding Station (with voice) O v OF 1 G O O S E H O M E R 137.7 mc 5 watts Radar Beacon 0 -Point To Point Radio Station (with voice) available to aircraft with prior permission COTE ST LUC RACON 9310mc 1-2-2 A I R N A V I G A T I O N L I G H T S Rotating Light Rotating Light (with flashing code lights) . . A Rotating Light (with course lights and. . . AL site number) I Flashing Light .^ , Fi -k * Flashing Light (with code) fi. * Lightship . . 4? Marine Light Occ w R G • F-Fixed (,)k F l < > i c k Flashing Occ-Occulting G p - C r o u p W - W h i t e B - B l u e FI .Flashing (3) number of flashes for period indicated Alt-Alternating R - R e d G - G r e e n SEC-Sector sec-Second Marine lights are white unless colors are stated. Marine alternating lights are red and white unless otherwise indicated. Isogonic Line M I S C E L L A N E O U S 4°£- Prohibited Area Prominent Transmission Line . V/////////////A Restricted or Warning Area ^ C5R2 2 (C5R2) C5 - Province of Canada o r R 2 o r W 2 Race Track . . ; O Danger or Caution or Intensive . . .K C5D2 \ J Jet Training Area ; orezXl 1425 1425 <C5D2> C 5 - Province of Canada U 2 S u 2 $ Obstruction and group obstruction (lighted) A 7A\~ Obstruction and group obstruction (unlighted) A M (280M280) . (280) (280) Numerals in italics indicate elevation above sea level of top. Heights above ground are shown in upright type in parenthesis. A I R W A Y S , AIR R O U T E S A N D C O N T R O L UNCONTROLLED AIRSPACE Air Route Identification Control Zone -e- 2 8 8 £*-Control Area ixtension Airway Identification *5> Air Route Bou mlarv •V360 O N T R E A L A T C . M O N C T O N A T C VHF/UHF Airway Controlling Control Centre Line Agency Separation LF/MF Airway Centre Line ,q0 \ • O } Aerodrome Traffic Zone Mid point bearings are magnetic and are based on mid point variation UNCONTROLLED AIRSPACE Designated Intersections A Name. . . .Compulsory Reporting Points A Name. . . .On Request Reporting Points T O R O N T O R A D I O 368 Y Z -z:=r " 25-455 LF/MF RADIO RANGE (AURAL) A -273° 093. N N •-m° 093°-* A N \ A T h e heavy line indicates the " N " quadrant. T h e bearings shown are magnetic GROUND-AIR EMERGENCY CODE LAY OUT SYMBOLS IN OPEN GROUND AS FOLLOW! 3. Make symbols as large as possible. They should be at least 8 fee 4. Lay out symbols exactly as depicted to avoid confusion with othf 5. Separate the parts of symbols 2. 7. 10. 14 and 17 by 10 feet if p TO ATTRACT ATTENTION 1. Pour oil on rags and make a smudge. Use other materials to make smoke 2 Make large SOS in open ground. In snow outline with boughs or moss. 3. Make trails in virgin snow. These can be seen readily from aircraft. 4. Lay your cowlings out so that they shine in the sun. 5. Keep your aircraft clear of snow or brush. 6. Using radio Call at H+15 to 18 and H +45 to 48. Save your battery as much as possible. 7. Point flashlight at approaching aircraft and send SOS. X F • REQUIRE DOCTOR SERIOUS INJURIES REQUIRE MEDICAL SUPPLIES UNABLE TO PROCEED 7 V REQUIRE FOOD AND WATER REQUIRE FIREARMS AND AMMUNITION REQUIRE MAP AND COMPASS LO AIRCRAFT ACKNOWLEDGEMENT SIGNALS UNDERSTOOD (a) rocking from side to side. GROUND SEARCH PARTY SIGNAL CODE 3 LIT OPERATION ENDED RETURN TO YOUR BAS • A +-+ REQUIRE SIGNAL LAMP WITH BATTERY AND RADIO INDICATE DIRECTION TO PROCEED L AM PROCEEDINC WILL ATTEMPT TAKE OFF AIRCRAF DAMAGED PROBABLY SAFE TO LAND HERE 14 16 17 18 LL N Y JL W REQUIRE FUEL AND OIL ALL WELL YES NOT UNDERSTOOD REQUIRE ENGINEER AIRCRAFT ACKNOWLEDGEMENT SIGNALS NOT UNDERSTOOD (a) complete right hand circuit, or (b) red flashes on signalling lamp. XX SOME PERSONNEL FOUND UNABLE TO CONTINU RETURNING TO BASE DIVIDED INTO TWO GROUPS PROCEEDING AS INDICATED INFORMATION SAYS AIRCRAFT IN THIS DIRECTION NOTHING FOUND WILL CONTINUE SEARCH Ground parties should have signal kits for these codes. TRAFFIC CONTROL LIGHT SIGNALS COLOUR GREEN CHARACTER — STEADY... — FLASHES.. IN .IGL ..CLEARED TO LAND.. RETURN FOR IANDINl ON AERODROME CLEARED FOR TAKE OFF ,.. .CLEARED TO TAXI RED STEADY/Trv***. .GIVE WAY TO OTHER AIRCRAFT AND CONTINUE CIRCLING FLASHES AIRPORT UNSAFE.If^Y. v: DO NOT LAND STOP .. TAXI CLEAR OF AN DING AREA IN USE WHITE FLASHES RETURN TO STARTING POINT ON AERODROME RED PYROTECHNICAL DO NOT LAND FOR THE TIME BEING INDEX TO WORLD AERONAUTICAL CHARTS Scale 1 : 1,000,000 16-2 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0102152/manifest

Comment

Related Items