Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Postglacial chronology of large earthflows in south-central British Columbia Jones, Penelope Sarah Ann 1988

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

Item Metadata

Download

Media
831-UBC_1988_A1 J66.pdf [ 38.38MB ]
Metadata
JSON: 831-1.0098061.json
JSON-LD: 831-1.0098061-ld.json
RDF/XML (Pretty): 831-1.0098061-rdf.xml
RDF/JSON: 831-1.0098061-rdf.json
Turtle: 831-1.0098061-turtle.txt
N-Triples: 831-1.0098061-rdf-ntriples.txt
Original Record: 831-1.0098061-source.json
Full Text
831-1.0098061-fulltext.txt
Citation
831-1.0098061.ris

Full Text

POSTGLACIAL CHRONOLOGY OF LARGE EARTHFLOWS IN SOUTH-CENTRAL BRITISH COLUMBIA BY PENELOPE SARAH ANN JONES B.A., THE UNIVERSITY OF CAMBRIDGE 1979 M.Sc, THE UNIVERSITY OF BRITISH COLUMBIA 1982 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF GRADUATE STUDIES (GEOGRAPHY) WE ACCEPT THIS THESIS AS CONFORMING TO THE REQUIRED STANDARD THE UNIVERSITY OF BRITISH COLUMBIA AUGUST, 1988 CcVENELOPE SARAH ANN JONES, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of <^eo^mft>PHW The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 i i ABSTRACT F i f t y - t h r e e earthflows in south-central B r i t i s h Columbia were i d e n t i -f i e d from an a i r photograph search and from a review of previous geologic reports. Many have p a r a l l e l 'en echelon 1 l a t e r a l deposits i n d i c a t i v e of se-ve r a l movement phases during the p o s t g l a c i a l period, and some have been ac-t i v e during the present century up to the time of the study. The purpose of t h i s study was to date phases of earthflow a c t i v i t y during the p o s t g l a c i a l period and r e l a t e them to c l i m a t i c f l u c t u a t i o n s . Earthflows i n the study area are concentrated i n serpentinised p e r i d o t i t e , basalt, sediments (mainly v o l c a n i c l a s t i c s ) , and other volcanics. Earthflows take place p r e f e r e n t i a l l y down dip i n sediments and many are as-sociated with f a u l t l i n e s . A l l earthflow materials, except those derived from serpentinised p e r i d o t i t e , weather to montmorillonite-rich material, and each y i e l d s a c h a r a c t e r i s t i c grain-size d i s t r i b u t i o n . Earthflow gradient i n the study area depends upon material type, i n d i c a t i n g that earthflow texture and mineralogy govern shearing resistance and hence p a r t i a l l y determine charac-t e r i s t i c slope angles. Earthflow movement during the past 60 years was investigated using a i r photograph chronosequences dating back to 1928. Reactivations of s i x flows were i d e n t i f i e d i n the period 1950 to 1960, and an analysis of p r e c i p i t a t i o n records from four stations around the study area showed that the l e v e l of winter p r e c i p i t a t i o n increased around 1950. It was concluded that the ob-served r e a c t i v a t i o n was a response to r i s i n g groundwater levels during a period of increased winter p r e c i p i t a t i o n . Earthflow movement and c l i m a t i c fluctuations over the l a s t 200 years were investigated i n an analysis of t r e e - r i n g width records at four s i t e s . Moist phases were recorded i n the periods 1800-1830, 1870-1920 and from 1945 to the present. Compression wood formation at three s i t e s corresponded with moist phases, so i t was concluded that, over the l a s t two centuries, earthflow movement was probably coincident with phases of higher p r e c i p i -t a t i o n . A p o s t g l a c i a l c l i m a t i c chronology was i n f e r r e d from published pollen analyses and from Neoglacial i c e f l u c t u a t i o n s , both within and outside the study area. New palynological data were c o l l e c t e d from Red Mountain, a high elevation s i t e i n the west of the study area. These showed a two-phase cooling period following the close of the Hypsithermal. The f i r s t cooling period took place around 6-7,000 BP, and the second around 3,000 BP. Earthflow movement during the p o s t g l a c i a l period was analysed using radiocarbon dates, tephrochronology, stratigraphy, carbonate accumulation, and earthflow morphology. Data from twenty-one s i t e s showed that movement was concentrated mainly a f t e r 7,000 BP, and at several s i t e s the st r a t i g r a p h i c p o s i t i o n of the Bridge River tephra dated movement around 2,400 BP. It was concluded that, at t h i s longer timescale, movement at many s i t e s could be attributed to post-Hypsithermal c l i m a t i c d e t e r i o r a t i o n . i v TABLE OF CONTENTS Abstract .' i i Table of contents i v L i s t of tables • x L i s t of figures x i Chapter 1 Introduction and l i t e r a t u r e review 1 1.1 Earthflows - d e f i n i t i o n 1 1.2 Earthflows - the d r i v i n g force 3 1.2.1 Earthflow movement re l a t e d to p r e c i p i t a t i o n and piezometric changes ... 4 1.2.2 Other causes of earthflow movement 12 1.3 Earthflows - scale dependent response 15 1.4 Hypothesis and objectives 18 Chapter 2 The study area 20 2.1 Earthflow locations i n south-central B r i t i s h Columbia 20 2.2 Topography . . . • 22 2.3 Regional geology 24 2.4 G l a c i a l h i s t o r y and Quaternary geology 29 2.5 Fraser r i v e r d i v e r s i o n 31 2.6 Climate 32 2.7 Vegetation . 35 Chapter 3 C h a r a c t e r i s t i c s of earthflows i n the study area 38 3 .1 Morphology . 38 3.1.1 Slump bowls 38 3.1.2 Flow track 39 3.1.3 Toe zone 46 3.2 Geology 47 V 3 . 2 . 1 Lithology 47 3 . 2 . 2 Dip 49 3 . 2 . 3 Faults 50 3 . 2 . 4 Groundwater flow 50 3 . 2 . 5 Geomorphic h i s t o r y 53 3 . 2 . 6 G l a c i a l h i s t o r y 55 3 . 2 . 7 Conclusions 55 3 .3 Geotechnical properties of earthflow materials 56 3 . 3 . 1 Introduction 56 3 . 3 . 2 Grain-size d i s t r i b u t i o n 57 3 . 3 . 3 Atterberg l i m i t s 61 3 . 3 . 4 Clay minerals and cations 63 3 . 3 . 5 Gradient and shear strength 65 3 . 3 . 6 Conclusions 72 Chapter 4 Movement of earthflows i n the study area i n h i s t o r i c time 74 4 . 1 Introduction 74 4 . 2 Earthflow movement as recorded i n a i r photographs and recent observations . . 74 4 . 2 . 1 The nature of the evidence 74 4 . 2 . 2 Black Dome earthflow group... 75 4 . 2 . 3 Churn Creek (A) earthflow 75 4 . 2 . 4 Drynoch (A) earthflow 77 4 . 2 . 5 Grinder Creek (A) earthflow 78 4 . 2 . 6 Additional s l i d e at Grinder Creek 78 4 . 2 . 7 Red Mountain (A) earthflow 79 4 . 2 . 8 Red Mountain B . C J D . E J G earthflows 79 4 . 2 . 9 Yodel earthflow 81 4 . 2 . 1 0 Earthflow movement - conclusions 83 4 . 3 Monitoring of movement at Grinder Creek (A) 83 v i 4.4 Climatic chronology 92 4.4.1 Introduction 92 4.4.2 Previous work 92 4.4.3 Quality of instrumental records . 94 4.4.4 Seasonal trends 100 4.5 A comparison of c l i m a t i c trends and earthflow a c t i v i t y 104 Chapter 5 The dendrochronologic record 109 5.1 Introduction 109 5.2 The c l i m a t i c record 109 5.2.1 Growth response of trees to c l i m a t i c variables 109 5.2.2 Tree-ring width chronologies - a survey of the relevant l i t e r a t u r e I l l 5.2.3 Sample c o l l e c t i o n 113 5.2.4 Sample processing 117 5.2.5 Operator variance 126 5.2.6 Quality of the records 126 5.2.7 Tree ring-width chronologies 137 5.2.8 Correlation with p r e c i p i t a t i o n 142 5.2.9 Summary 147 5.3 Compression wood chronology 149 5.3.1 Introduction 149 5.3.2 The nature and formation of compression wood 149 5.3.3 Previous applications of chronologies developed from compression wood formation 150 5.3.4 Data c o l l e c t i o n and processing ; 151 5.3.5 Movement chronologies and t h e i r r e l a t i o n s h i p s with the c l i m a t i c chronology 153 5.3.6 Spatial pattern of movement 159 v i i 5.4 Conclusions 163 Chapter 6 P o s t g l a c i a l climate and earthflow movement i n south-central B r i t i s h Columbia .... 165 6.1 Introduction 165 6.2 A survey of p o s t g l a c i a l c l i m a t i c f l u c t u a t i o n s i n south-central B r i t i s h Columbia...165 6.2.1 Deglaciation and subsequent f l u c t u a t i o n s i n i c e volume 165 6.2.2 Paleobotanical evidence for c l i m a t i c fluctuations i n the p o s t g l a c i a l period: an overview 166 6.2.3 P o s t g l a c i a l c l i m a t i c changes i n the study area 170 6.3 Pollen analysis of a core from Red Mountain (A) 171 6.3.1 Introduction 171 6.3.2 Description of the sampling s i t e at Red Mountain (A) 172 6.3.3 Sampling and laboratory techniques 176 6.3.4 Stratigraphy 177 6.3.5 Pollen frequency and percentage organic matter 178 6.3.6 I d e n t i f i c a t i o n of palynomorphs 181 6.3.7 Palynomorph zonation - methods 185 6.3.8 'Local' s i g n a l 189 6.3.9 Interpretation of the l o c a l s i g n a l 191 6.3.10 The regional p o l l e n s i g n a l 193 6.3.11 Interpretation of the regional s i g n a l 195 6.3.12 A comparison of Bog G with other p o l l e n analyses i n B r i t i s h Columbia ..201 6.3.13 Summary 202 Chapter 7 A chronology of p o s t g l a c i a l earthflow movement 7.1 Techniques for dating p o s t g l a c i a l earthflow movement 205 7.1.1 Introduction 205 7.1.2 Problems and l i m i t a t i o n s of radiocarbon dating 205 v i i i 7.1.3 Tephra i n the study area 206 7.1.4 Use of r e l a t i v e age dating techniques for dating earthflows ...209 7.2 P o s t g l a c i a l earthflow movement 210 7.2.1 Introduction 210 7.2.2 Big Bar Creek earthflow 210 7.2.3 Black Dome group of earthflows 215 7.2.4 Burkholder and Lac l a Mer earthflows 215 7.2.5 Camoo earthflow 215 7.2.6 Canoe Creek earthflow 217 7.2.7 Churn Creek earthflow , 218 7.2.8 Drynoch (A) earthflow 218 7.2.9 Drynoch (B) earthflow 226 7.2.10 Flapjack earthflow 229 7.2.11 Fountain, Gibbs Creek, G i l l o n Creek, Kettlebrook Creek and Tunnel earthflows ...232 7.2.12 Grinder Creek earthflow 232 7.2.13 Hat Creek earthflow 235 7.2.14 Heginbottora earthflow 235 7.2.15 P a v i l i o n (A) earthflow 241 7.2.16 Red Mountain (A) earthflow 245 7.2.17 Red Mountain B to E and G earthflows 255 7.2.18 Red Mountain (F) earthflow 255 7.2.19 Yalakom earthflows 256 7.3 A comparison of c l i m a t i c data and earthflow movement i n the p o s t g l a c i a l period 261 Chapter 8 Discussion and Conclusions 267 8.1 Introduction 267 8.2 Geologic factors 267 8.3 Scale considerations 270 i x 8.4 Observations on earthflow movement mechanisms 272 8.5 The r e l a t i o n s h i p between earthflow movement and c l i m a t i c fluctuations i n the study area 273 Bibliography 278 Appendix A Photographs used f o r a i r photograph i n t e r p r e t a t i o n 291 Appendix B F o s s i l palynomorphs recognised i n rock samples from Red Mountain (personal communication, G. Rouse).... 323 Appendix C Atterberg l i m i t s 324 Appendix D Miscellaneous tree r i n g data 326 Appendix E Cluster analysis showing l o c a l and regional palynomorph zonations 358 Appendix F Stratigraphy of cores from the study area 360 Appendix G Survey of published l i t e r a t u r e describing earthflows 373 Appendix H Earthflows i n the study area 378 Appendix I Detailed geology of selected flows within the study area 382 Appendix J A comparison of factors causing earthflows i n the study area with factors at other locations 400 Appendix K Laboratory techniques used to obtain geotechnical data from earthflows i n the study area 412 Appendix L A comparison of earthflow material properties within and outside the study area 419 Appendix M A detai l e d d e s c r i p t i o n of h i s t o r i c changes at Black Dome and Grinder Creek (A) and (B) ... 428 LIST OF TABLES x 2.1 Climatic s t a t i s t i c s from the study area 34 3.1 1/b r a t i o s of flow track zones for earthflows i n the study area . 42 3.2 Angles of shearing resistance from earthflow material i n the study area 70 4.1 Rates of movement for selected earthflows i n south-central B r i t i s h Columbia 76 4.2 Summary of movement at Grinder Creek 87 4.3 A comparison of winter p r e c i p i t a t i o n 1983/1984 and 1984/1985 .... 86 4.4 Comparison of means for Lytton2 and Lytton 99 4.5 Thornthwaite cal c u l a t i o n s 106 5.1 Trees used for chronology construction 122 5.2 Operator variance e f f e c t s , - 12,7 5.3 Tree-ring s t a t i s t i c s of chronology trees 129 5.4 F values for climate and tree e f f e c t s 133 5.5 Correlation of r i n g widths with cumulative departures and actual p r e c i p i t a t i o n measurements from Big Creek 144 6.1 Regional and l o c a l species on Bog G 187 x i LIST OF FIGURES 1.1 Major morphological features of earthflows 2 1.2 Cor r e l a t i o n between p r e c i p i t a t i o n and earthflow movement at Minnis North 2 1.3 Correlation between p r e c i p i t a t i o n and earthflow movement at the "20l" s l i d e 5 1.4 Cor r e l a t i o n between groundwater levels and cumulative movement at Drynoch landslide, B.C 5 1.5 Correlation between p r e c i p i t a t i o n and movement at Waerenga-O-Kuri 7 1.6 Cor r e l a t i o n between p r e c i p i t a t i o n and movement at Waerenga-O-Kuri 7 1.7 I n i t i a t i o n of the Wind Mountain landslide a f t e r a p r e c i p i t a t i o n increase around 1945 8 1.8 Disruption of a l o c a l groundwater flow net by landslide movement 10 1.9 Location of earthflow f a u l t scarp r e l a t i v e to Kj.rkwood earthflow. 10 1.10 Depth to shear plane and earthflow s i z e 16 2.1 Earthflow locations i n south-central B r i t i s h Columbia 21 2.2 Topography of the study area ..23 2.3 Si m p l i f i e d geologic map of the study area .25 2.4 Ice flow over the study area and f l u v i a l and f l u y i o g l a c i a l features 30 2.5 Seasonal p r e c i p i t a t i o n d i s t r i b u t i o n at AES stations within and close to the study area ....33 2.6 Vegetation zones within the study area . . . 36 x i i 3.1 S t r i a t i o n s on a former shear plane at Grinder Creek (A) 40 3.2 Selected long p r o f i l e s to show flow tracks with well developed l a t e r a l s ...43 3.3 The formation of l a t e r a l s at the Kirkwood earthflow .....44 3.4 "Crevassed" zone above lobe at Grinder Creek (B) 44 3.5 Lobe moving down Grinder Creek (A) earthflow 45 3.6 Diagram of tension zone associated with lobe 51 3.7 Disrupted drainage on Yalakom (A) 52 3.8 Hypothetical groundwater flow system at Red Mountain 54 3.9a Grain-size d i s t r i b u t i o n and geologic provenance of earthflow materials i n the study area 59 3.9b Grain-size d i s t r i b u t i o n and geologic provenance of earthflow materials i n the study area (other studies) 60 3.9c Grain-size d i s t r i b u t i o n and geologic provenance of earthflow materials from other locations 60 3.10 E l a s t i c i t y index, l i q u i d l i m i t s and geologic provenance of earthflows from the study area (data from t h i s study) 62 3.11 Relationship between earthflow area and flow angle for selected earthflows within the study area 67-69 4.1 Active and i n a c t i v e areas on the Red Mountain (A) earthflow 80 4.2 Stake array l o c a t i o n at Grinder Creek (A) 85 4.3 Stake array configuration with respect to earthflow motion 86 4.4 Zones of extending and compressing flow on the Grinder Creek (A) earthflow 90 4.5 Simple example to demonstrate accumulated departure c a l c u l a t i o n .93 4.6 P r e c i p i t a t i o n trends at Big Creek, Barrett, personal communication 95 4.7 Cumulative departures of Fraser discharge at Hope .96 x i i i 4.8 Month by month comparison of p r e c i p i t a t i o n at AES stations Lytton and Lytton 2 98 4.9a T o t a l and seasonal cumulative departures f o r Big Creek 101 4.9b Total and seasonal cumulative departures for Mamit Lake ........ 102 4.9c T o t a l and seasonal cumulative departures for Lytton 103 4.10 Correlation between earthflow a c t i v a t i o n and periods of above average p r e c i p i t a t i o n 107 5.1 Raw r i n g width chronology for P a v i l i o n Lake 114 5.2 Location of chronology trees sampled at Drynoch 115 5.3 Location of trees sampled for reaction wood at Drynoch (B) 116 5.4 Location of trees sampled for chronology construction at Grinder (A) . . .118 5.5 Location of trees sampled both f o r chronology and reaction wood at Heginbottom earthflow 119 5.6 Location of trees sampled both f o r chronology and reaction wood at P a v i l i o n (A) 120 5.7 Number of trees f i t t i n g exponential and polynomial models 125 5.8 Normal p r o b a b i l i t y p l o t of residuals a f t e r f i t t i n g a f a c t o r i a l model to the 350 year P a v i l i o n record . . , ,135 5.9 Drynoch 400 year chronology (exponential f i t ) 138 5.10 Drynoch 400 year chronology (polynomial f i t ) 138 5.11 Grinder Creek master chronology 139 5.12 Heginbottom 250 year chronology (exponential f i t ) 140 5.13 Heginbottom 250 year chronology (polynomial f i t ) 140 5.14 A v i s u a l c o r r e l a t i o n of "wet" and "dry" phases at a l l four chronology s i t e s 141 5.15 Cumulative departures of p r e c i p i t a t i o n at Hat Creek 145 5.16 Drynoch reaction wood 154 5.17 Heginbottom reaction wood 155 xiv 5.18 P a v i l i o n reaction wood 156 5.19 P a v i l i o n reaction wood 157 5.20 Tree at Drynoch s p l i t by an active tension crack 160 6.1 Location of s i t e s discussed i n chapter 8 167 6.2 Schematic diagram of vegetation zones on Bog G 173 6.3 Vegetation on Bog G ....175 6.4 Pollen diagram f o r Bog G showing " l o c a l " zonation 179 6.5 Percentage loss on i g n i t i o n at Bog G .180 6.6 Palynomorphs from Bog G . 182-184 6.7 Pollen diagrams showing " r e g i o n a l " zonation 194 7.1 Tephra i n the study area 207 7.2 Bovis (1985) i n t e r p r e t a t i o n of movement phases at Big Bar 211 7.3 Scarps cut i n terrace r i s e r s opposite Big Bar Creek earthflow ...212 7.4 Early phase of Big Bar earthflow overriding Fraser r i v e r terrace 214 7.5 G l a c i o f l u v i a l deposits on toe of Camoo Creek earthflow ..........216 7.6 Toe section on Drynoch (A) earthflow 219 7.7 Map of Drynoch (A) showing l o c a t i o n of s o i l p i t s 220 7.8a Calcium carbonate percentage i n Drynoch (B) l a t e r a l deposits ....222 7.8b Calcium carbonate percentage i n Drynoch (B) l a t e r a l deposits ....223 7.8c Calcium carbonate percentage i n Drynoch (B) l a t e r a l deposits ....224 7.9 Evaporite overridden by early Drynoch (B) earthflow movement .;..227 7.10 Calcium carbonate percentage i n Drynoch (B) l a t e r a l deposits ....228 7.11a Core from Flapjack earthflow ...230 7.11b Core from Beaver Lake 231 7.12 Bridge River tephra i n s o i l p i t at Grinder (A) 233 7.13 Movement phases at Grinder Creek (A) 234 7.14 Heginbottom earthflow movement phases 237 7.15 Buried s o i l overridden by Heginbottom earthflow 238 7.16 Bridge River tephra at Heginbottom 240 XV 7.17 Phases of earthflow movement at P a v i l i o n (A) 242 7.18 Movement phases at Red Mountain (A) 246 7.19a Stratigraphy of Lake F, core 1 249 7.19b Stratigraphy of Lake F, core 2 ...250 7.20 Formation of Lake K 252 7.21 Stratigraphy of Lake K ...254 7.22 Relative ages of movement at Yalakom (A) 258 7.23 Location of coring s i t e s on Yalakom (B) 259 7.24 Location of bog i n front of toe of Yalakom (C) flow 260 7.25 Summary of c l i m a t i c changes i n B.C. over the past 11,000 years ..262 7.26 Summary of earthflow movement over the l a s t 11,000 years .263 1 Chapter 1 Introduction and l i t e r a t u r e review  1.1 Earthflows - d e f i n i t i o n Earthflows are ' g l a c i e r - l i k e 1 , slow moving s l i d e s of earth and rock. E l i o t J Blackwelder (1912) f i r s t applied the term to the Gros Ventre flow. Howe (1909) previously had described the morphologically s i m i l a r Slumgullion 'mudflow', although his nomenclature invoked - probably i n c o r r e c t l y - a rapid, catastrophic genesis. Cross (1924) applied the term 'mudflow' to the Cucaracha s l i d e on the Panama Canal, where maximum sustained movement rates of 4.5m per day, and a surge of 23m i n 20 minutes were observed. A t h i r d term, 'mudslide', used to describe s i m i l a r features, entered common usage much la t e r (Hutchinson and Bhandari 1971), but has recently been adopted as the preferred nomenclature by many B r i t i s h workers (Brunsden 1984). In l i n e with North American custom 'mudslides' and 'mudflows' w i l l a l l be referred to here as 'earthflows' (Keefer and Johnson 1983). Earthflows i n the study area are defined by two c r i t e r i a : morphological features and movement c h a r a c t e r i s t i c s . Varnes (1978) supplies a simple d i -agram (adapted from Zaruba and Mencl, 1969, reproduced here as F i g 1.1) showing the three morphological zones common to a l l earthflows. These are: ( i ) a bowl-shaped 'source a r e a 1 ; ( i i ) the 'main track', or 'flow track' (Prior and Stephens (1971)), where an active flow i s confined to a channel bounded by slickensided shear surfaces, and ( i i i ) a 'depositional area' where material either accumulates either as a steep faced snout or else as a l a t -e r a l l y spreading fan. A l t e r n a t i v e l y i t may be removed by f l u v i a l , g l a c i a l or marine processes. Keefer and Johnson (1983) i d e n t i f y further morphological character-i s t i c s of earthflows. They describe these features as 'tongue or teardrop shaped, (with) rounded, bulging toes and si n u s o i d a l l o n g i t u d i n a l p r o f i l e s , concave upward near the head of the earthflow and convex upward near the toe.' They also recognise that many earthflows are flanked by 'immobile Source area Main track Depositional area F i g 1.1 Major morphological features of earthflows (adapted from Varnes, 1978) « 0.0 Weekly earthflow movement Weekly r a i n f a l l 1970 Fig 1.2 26 9 23 7 21 4 Aug Sept Oct Nov Correlation between p r e c i p i t a t i o n and earthflow movement at Minnis North (adapted from P r i o r and Stephens, 1971) 3 l a t e r a l ridges'. The morphological features of earthflows, c h i e f l y t h e i r bulging toes and convexo-concave long p r o f i l e s , suggest f l u i d flow. However, although some i n t e r n a l deformation of earthflows has been measured, i t ap-pears that movement occurs p r i n c i p a l l y by l a t e r a l shear along slickensided surfaces which characterise both the basal and l a t e r a l margins of an active earthflow. Several earthflows have been described from south-central B r i t i s h Columbia i n geological reports, notably earthflows at Drynoch (VanDine, 1980), P a v i l i o n (Bovis, 1980) and Hat Creek (Rawlings, 1984). Recent re-gional mass-movement surveys (Eisbacher (1979), Cruden (1984) and Bovis (1985)) show that, within B r i t i s h Columbia, earthflows are concentrated within the study area mapped i n F i g 2.1. Earthflows have also been described from many other locations around the world, notably from several s i t e s i n the western U.S.A.. A de t a i l e d d e s c r i p t i o n of these and many other earthflows reported i n the l i t e r a t u r e i s given i n Appendix G. 1.2 Earthflows - the d r i v i n g forces The g e o l o g i c a l l y , s p a t i a l l y and c l i m a t o l o g i c a l l y diverse earthflows summarised in Appendix G show varied responses to d i f f e r e n t frequency and magnitude p r e c i p i t a t i o n events. Some flows accelerate during or immediately a f t e r heavy r a i n f a l l , others display a seasonal movement regime, and de t a i l e d monitoring of two flows over more than a decade (Waerenga-O-Kuri and Wind Mountain), has revealed acceleration over a number of years. Several other factors have been shown to i n i t i a t e , maintain or accelerate earthflow movement. F i r s t , many earthflows are maintained by marine, g l a c i a l or f l u v i a l toe erosion which removes earthflow debris at the lower end of the flow, thereby maintaining flow gradient. Second, loading i n the upper part of a number of small flows has been shown to cause surging or r e a c t i -vation. Third, earthquakes may play a major r o l e , i f not i n i n i t i a t i o n , c e r t a i n l y i n r e a c t i v a t i o n . This may be achieved both by increasing 4 groundwater levels through disruption of the p r e - e x i s t i n g flow net and by causing l o c a l or regional tectonic warping, thereby increasing earthflow gradient. F i n a l l y , vegetation clearance may increase earthflow a c t i v i t y by decreasing evapotranspiration or decreasing i n f i l t r a t i o n time, although there i s i n s u f f i c i e n t evidence to draw even te n t a t i v e conclusions about the ef f e c t s of t h i s factor. The roles of p r e c i p i t a t i o n and groundwater and of the other major subsidiary factors are examined below. 1.2.1 Earthflow movement rela t e d to p r e c i p i t a t i o n and piezometric changes 5 2 Many small earthflows (<10 m ) show a rapid response to i n d i v i d u a l pre-c i p i t a t i o n events; F i g 1.2, reproduced from P r i o r and Stephens (1971) shows the c o r r e l a t i o n between p r e c i p i t a t i o n and earthflow movement at Minnis North 4 2 (#10 Appendix G, area c i r c a 10 m ) . The only medium or large flow i n Appendix G to show a rapid response to i n d i v i d u a l p r e c i p i t a t i o n events i s the "20l" 5 2 s l i d e on the Eel River (#46, area c i r c a 4.5x10 m ), where F i g 1.3 (reproduced from C a l i f o r n i a Department of Water Resources, 1973) indicates a 2 day lagged response. Above-average seasonal p r e c i p i t a t i o n has been documented as i n i t i a t i n g small earthflows i n many areas. For example, Sharpe and Dosche (1942) re-ported abnormally high January p r e c i p i t a t i o n i n 1937 which i n i t i a t e d several earthflows i n the Appalachian Plateau (#20, Appendix G); these advanced for 13 years and were thereafter i n a c t i v e . Grove (1952) att r i b u t e d s i m i l a r s h o r t - l i v e d features on Bredon H i l l i n the U.K. (#8, Appendix G) to high winter p r e c i p i t a t i o n i n 1950-51, while Keefer (1977) r e l a t e d earthflow i n i -t i a t i o n i n the San Francisco Bay area of C a l i f o r n i a (#15-18, Appendix G) to abnormally moist winter conditions. There i s only one example of a medium to large earthflow i n i t i a t i n g i n response to high p r e c i p i t a t i o n i n a s i n g l e season. At Handlovka (#25, Appendix G) Zaruba and Mencl (1969) recorded s l i d e movement i n 1960 i n i t i a t e d 5 F i g 1.3 Correlation between p r e c i p i t a t i o n and earthflow movement at the "201" s l i d e (adapted from C a l i f o r n i a Department of Water Resources, 1973) Water l e v e l in test hole Fi g 1.4 Correlation between groundwater levels and cumulative movement at Drynoch landslide, B.C. (adapted from VanDine, 1980) 6 by annual p r e c i p i t a t i o n ( f a l l i n g mainly i n the winter months) which exceeded by 50% the 1901-1960 mean. 5 6 2 Many moderately large earthflows (10 -10 m ), such as those described from Oregon by Swanson and Swanston (1977, #47 and #48, Appendix G), and i n the Van Duzen r i v e r basin by Kelsey (1977, #29-#34, Appendix G), are sea-sonally active. Kelsey stated that 1...movement does not occur with the onset of winter rains but each winter a f t e r f a l l and early winter rains have thoroughly wet the colluvium...'. However, a f t e r i n i t i a l wetting '...major episodes of movement coincide with periods of heavy r a i n f a l l . . . ' and periods of excessive p r e c i p i t a t i o n are important rather than i n d i v i d u a l storms. G i l and Kotarba (1977) quantify t h i s response, f i n d i n g a good (r=0.88-0.99) co r r e l a t i o n between flow movement and monthly p r e c i p i t a t i o n t o t a l s . 6 2 Large flows (>10 m ) show year-round movement, usually with pronounced seasonal maxima. For example, F i g 1.4, reproduced from VanDine (1980) shows that the Drynoch landslide (#56, Appendix G) moved most r a p i d l y 1-4 months afte r groundwater levels achieved t h e i r annual maxima, i n May-June, i n re-sponse to snowmelt. (It should be noted that t h i s apparent lag could be, at least i n part, an a r t e f a c t of the long response time of standpipe piezometers). Bovis (1985) and Nadler (1984) proposed a s i m i l a r time lag for the P a v i l i o n flow. Few earthflows have been monitored for an extended period: Wasson and Hall's (1982) work at Waerenga-O-Kuri and Fukuoka's measurements at Mt. Chausu, together with various unpublished records r e s u l t i n g from railway track realignment (the "201" s l i d e and others on the Eel River, Drynoch on the Thompson River) and Palmer's (1977) analysis of the Wind Mountain land-s l i d e i n Oregon, provide notable exceptions. The data from Waerenga-O-Kuri 4 2 (#52, Appendix G, 9x10 m ), a small to medium-sized feature, are p a r t i c u l a r l y i n t e r e s t i n g because they show two d i f f e r e n t response frequencies to precip-i t a t i o n input. F i g 1.5, from Campbell (1966) demonstrates that movement at 1963 1964/ 1965 200 100 R a i n f a l l Cumulative movement 1.5 Correlation between p r e c i p i t a t i o n and earthflow movement Waerenga-0-Kuri, af t e r Campbell (1966) 3000 2000 E O 1000 Cumulative movement Cumulative percentage departure from average monthly r a i n f a l l 1960 '-' 1965 1970 1975 1980 1.6 Correlation between p r e c i p i t a t i o n and earthflow movement at Waerenga-0-Kuri, after Wasson and Hall (1982) 8 F i g 1.7 I n i t i a t i o n of the Wind Mountain landslide increase c i r c a 1945, a f t e r Palmer 1977 a f t e r a p r e c i p i t a t i o n 9 Waerenga-O-Kuri follows wet periods a f t e r a 3-6 week time lag. However, on a longer timescale, Wasson and H a l l (1982) show that movement i s c l o s e l y related to p o s i t i v e cumulative departures from mean monthly p r e c i p i t a t i o n (Fig 1.6), and at t h i s longer timescale, lags of up to two years can be seen. Palmer (1977) describes s i m i l a r long term movements of the Wind Moun-t a i n landslide i n Oregon (#55, Appendix G), although he does not have such fine temporal resolution. F i g 1.7, reproduced from Palmer's report, shows that the upper part of the s l i d e reactivated i n the l a t e 1930's and accel-erated i n the mid 1940's i n response to a substantial increase i n mean annual p r e c i p i t a t i o n . The middle and lower parts showed marked acceleration i n 1970 af t e r heavy winter rains. In 1949 an earthquake of Richter magnitude 8.0 affected the earthflow (the proximity of the epicentre i s not stated by Palmer). However, i n t h i s case seismic forces were not a major influence on movement as the upper s l i d e accelerated 2-3 years previously while move-ment on the lower and middle portions did not occur u n t i l twenty years l a t e r . It i s clear from the many case studies presented above that earthflow movement i s , at many locations, c l o s e l y t i e d to p r e c i p i t a t i o n amount. How-ever, the response i s complex, showing lags ranging from one hour or less to a year or more. Moreover, there i s strong evidence that, at least for 5 2 many flows over 10 m , earthflow movement i s correlated with groundwater levels rather than d i r e c t l y with p r e c i p i t a t i o n . Groundwater levels a f f e c t mass movement by varying water pressure head at the shear plane. High groundwater levels i n an earthflow, engendered by above-average p r e c i p i t a t i o n , bring about a reduction i n e f f e c t i v e stress at the earthflow shear plane thereby reducing normal stress on the shear plane and decreasing the f r i c t i o n a l resistance. Consequently earthflows that are inactive may become active, and earthflows that are active may experience increases in the rate of movement i n response to increased groundwater lev-els . F i g I.81 Disruption of a l o c a l groundwater flow net by landslide movement (after Patton and Hendron 1974) Contours in feet Fig 1.9 Location of earthflow faul t scarp r e l a t i v e to Kirkwood earthflow (after Hadley, 1964) 11 Theoretical and applied work (Toth 1962, 1963) on regional groundwater flow patterns shows that diverse temporal responses mirror the complexity of l o c a l , intermediate and regional discharge zones. A l l three discharge zones may occur i n close proximity, but may each have d i f f e r e n t source areas, flow path length and hence lagged response to p r e c i p i t a t i o n inputs. Freeze and Witherspoon (1967) show t h e o r e t i c a l l y that the percentage of a watershed occupied by discharge areas i s '...between 7 and 40%...(usually)... near the lower end of the range. . . ' . They also recognise that ' . . .the e f f e c t of a long term increase i n the t o t a l annual p r e c i p i t a t i o n would be a (more) permanent r a i s i n g of the water table...', superimposed on seasonal fluctuations. Therefore, as discharge areas occupy only a small percentage of most watersheds, a small increase i n recharge has a much larger influence on piezometric levels i n discharge areas (provided, of course, that t h i s l e v e l i s not constrained by discharge to a permanent watercourse). There i s much evidence suggesting that medium and large-sized earthflows 5 2 (>10 m ) l i e i n areas of regional groundwater discharge. Many are found i n pronounced topographic depressions which are predisposed to be discharge areas. Many flows also have a poorly drained and chaotic surface, and earthflow movement may cause temporary or long-term ponding of surface drainage as, for example, at Handlovka (Zaruba and Mencl 1969, #25, Appendix G) and at the Stockpond earthflow (Kelsey 1977). A casual inspection of the a i r photographs i n Appendix A reveals many flows i n the study area with nu-merous sag ponds, marginal lakes and boggy depressions. Figure 1.8 demon-strates the hypothetical d i s r u p t i o n of a groundwater flow net by earthflow movement. Recent work suggests that rapid fluctuations i n groundwater l e v e l might play a s i g n i f i c a n t r o l e i n earthflow i n i t i a t i o n and r e a c t i v a t i o n . Okimura (1983, #24, Appendix G) finds that t h i s i s c r u c i a l i n the i n i t i a t i o n of small, shallow f a i l u r e s i n Japan. He does not, however, give an adequate 12 des c r i p t i o n of these features, so i t i s not c e r t a i n that they would be de-scribed as earthflows i n t h i s study. His conclusions are reinf o r c e d by the findings of McGreal and Craig (1977) who found that small coastal earthflows i n Northern Ireland reactivated c h i e f l y during the autumn when groundwater l e v e l s , although quite low, were r i s i n g r a p i d l y . No explanation was offe r e d for t h i s phenomenon and no analogous pattern was seen i n any other small, medium or large sized earthflows. At a l l other s i t e s movement took place a f t e r an i n i t i a l 'wetting-up' phase of up to several months. In summary, the majority of monitored earthflows show that movement de-pends either upon i n d i v i d u a l p r e c i p i t a t i o n events, monthly t o t a l s , the an-nual cycle or on long term p r e c i p i t a t i o n trends. Flows may show a response at more than one frequency. Indeed, i n Appendix G there are only two examples where a r e l a t i o n s h i p between groundwater/precipitation and movement i s not seen. These are Slumgullion (Crandell and Varnes 1961, #54, Appendix G) where movement takes place at a constant rate throughout the year and Kirkwood, Montana, (Hadley 1964, #49, Appendix G) where earthquake a c t i v i t y was un-doubtedly the only material factor. 1.2.2 Other causes of earthflow movement There are three other factors which have been shown to a f f e c t earthflow movement at some s i t e s . These are: toe erosion (marine, g l a c i a l or f l u v i a l ) , rate of debris supply, and earthquakes. Each i s considered i n turn below. Several stable earthflows terminate i n a bulbous snout or fan. It i s possible that t h i s feature buttresses the earthflow and, i n many cases, provides s u f f i c i e n t l a t e r a l earth pressure to prevent further movement. In other cases the earthflow toe i s characterised by overthrusting i n d i c a t i v e of passive f a i l u r e (that i s , f a i l u r e i n compression). I f the earthflow snout or fan i s removed, or prevented from forming by toe erosion, then extending flow may occur, associated with a state of ac t i v e ( i . e . unconfined) f a i l u r e . 13 Many earthflows debouching to r i v e r s s u f f e r s i g n i f i c a n t toe erosion and consequently provide a major proportion of t o t a l stream sediment load. For example Kelsey (1977) calculated that sediment discharge from nineteen earthflows to the Van Duzen River i n C a l i f o r n i a i s equivalent to Ira of sur-face lowering per century for the areas affected by earthflow. He also doc-umented a dramatic increase i n earthflow a c t i v i t y a f t e r a large flood i n December 1964; s i x out of eleven flows active p r i o r to the flood accelerated, while eight dormant earthflows reactivated. In the study area, the Drynoch earthflow i s s t e a d i l y eroded by Thompson r i v e r ; the o l d Trans-Canada highway has now p a r t i a l l y disappeared into the r i v e r demonstrating s i g n i f i c a n t toe erosion over the past 100 years. Although toe erosion i s an important factor i n the i n i t i a t i o n and main-tenance of earthflow movement at some s i t e s , several flows have reactivated i n t h e i r headscarp regions. Continued movement on earthflows with large fans (e.g. P a v i l i o n (A), Drynoch (B)) developed i n the absence of toe erosion, indicates that toe erosion may not be a major factor i n maintaining or i n i -t i a t i n g movement on large and medium-sized flows. Debris loading the upper part of an earthflow a f f e c t s movement i n two ways: f i r s t by increasing she.ar stress and second by bringing about un-drained f a i l u r e under conditions of undrained loading. There are f i v e doc-umented cases of earthflow r e a c t i v a t i o n or surging a f t e r loading of the upper part of the flow, and of these, three involved small earthflows and were man-induced. Barton (1973) documents possible earthflow a c c e l e r a t i o n due to the c l i f f - e d g e s i t i n g of wartime defence posts, and Brunsden and Jones (1972) consider that earthflows on the flanks of Stonebarrow Down may have been reactivated by loading provided by road construction. Reactivation of a s l i d e near B r i s t o l (Hawkins 1973) was d e f i n i t e l y the r e s u l t of increased overburden pressure when i t was used as a t i p p i n g s i t e . 14 Hutchinson and Bhandari (1971) describe an analogous, but not man-made, s i t u a t i o n at Beltinge (#1, Appendix G) where the upper end of a low-level earthflow was r a p i d l y loaded by higher elevation features. This generated excess pore pressures and brought about undrained f a i l u r e . The f i n a l example i s that of Handlovka, where Zaruba and Mencl (1969) describe the flow as o r i g i n a t i n g i n 'a shallow gulley f i l l e d with slope debris'. They describe how the 'root area' (bowl-shaped head) had a 15m thickness of debris reduced to 7m during movement. Major earthquakes such as the Alaskan earthquake of 1964, have been shown to t r i g g e r large scale rock f a i l u r e s (McSaveney i n Voight, 1978). Eisbacher (1979) suggests that t h i s type of f a i l u r e r e s u l t s from a reduction i n f r i c t i o n a l resistance following d i l a t i o n brought about by seismic waves. However, th i s type of f a i l u r e cannot occur i n clays, so a l t e r n a t i v e mech-anisms, such as disruption of the groundwater flow net, l i q u e f a c t i o n of s e n s i t i v e clays, or an increase i n gradient, must be proposed. There are only two examples (discussed below) of earthflows reactivated by earthquake ac-t i v i t y ; i n other cases a l t e r n a t i v e explanations are more convincing. The Kirkwood earthflow i n Montana (#49, Appendix G) reactivated 5 or more days after the Hebgen Lake earthquake of 1959- F i g 1.9 (reproduced from Hadley (1964)), shows that the f a u l t scarp traversed the earthflow; lm downthrow and also 6.6m/Km regional downwarping both increased earthflow gradient. Increased groundwater flow was,also reported. Kelsey (1977) re-ports that Shockwaves from the 1906 San Francisco earthquake caused the "201 s l i d e " (#46, Appendix G) to block the E e l River temporarily. Conversely, Palmer (1977) refutes the hypothesis that an earthquake i n 1949 i n i t i a t e d movement of the Wind River s l i d e (#55, Appendix G) which accelerated 2-3 years p r i o r to the quake. The role of earthquakes cannot be adequately assessed e i t h e r from the available l i t e r a t u r e or i n t h i s study. However, the circumstances of the 15 Kirkwood r e a c t i v a t i o n , p a r t i c u l a r l y the transverse l o c a t i o n of a f a u l t scarp, suggest that seismic a c t i v i t y may have a s i g n i f i c a n t r o l e i n deter-mining both the location and timing of earthflows i n the study area, as many occur p a r a l l e l or transverse to major f a u l t l i n e s . 1.3 Earthflows - scale dependent response Appendix G shows that earthflow area spans 5 orders of magnitude 2 7 2 (10 -10 m ). Depth to f a i l u r e plane i s less v a r i a b l e , although there i s a trend of increasing depth to shear plane with increasing earthflow s i z e (Fig 1.10, compiled from data i n Appendix G) . (It should be pointed out that many authors quoted i n Appendix G give only estimated depth to shear plane). Earthflow v e l o c i t y i s also scale dependent as smaller flows generally move seasonally and faster than larger features (Appendix G). Surging (rapid movement over a short period) i s observed mainly i n smaller flows, although there are notable exceptions such as Handlovka and the "201" E e l River s l i d e . Earthflow material type i s also scale dependent; both Keefer (1977) and Sharpe and Dosch (1942) emphasise that small earthflows i n t h e i r respective study areas occur i n an intensely weathered zone overlying sedimentary s t r a t a , while the majority of larger flows i n Appendix G entrain material at the headscarp by r o t a t i o n a l slumping of r e l a t i v e l y unweathered bedrock. The previous section discussed the ro l e of p r e c i p i t a t i o n and groundwater i n i n i t i a t i n g and maintaining earthflows. It seems that smaller, shallower 5 2 earthflows (<10 m ) respond to i n d i v i d u a l intense p r e c i p i t a t i o n events, 5 6 2 while moderate s i z e features (10 -10 m ) respond to the annual cycle. These moderate s i z e features require an i n i t i a l wetting-up phase before the onset of movement, although they may show, within the wet season, a response to high p r e c i p i t a t i o n months (Kelsey 1978). Measurements taken from the largest 6 2 features (>10 m ) show that they respond to seasonal (Drynoch, F i g 1.4) or long-term (Wind Mountain, F i g 1.7) p r e c i p i t a t i o n or groundwater l e v e l o 0..0 DEPTH TO SHERR PLRNE U 10-0 20.0 30.0 40.0 i I i I i 1 1 3 09 C0_ cn a a> rt cr rt O . w cr (D 3 a> p> 3 • p. fl> 0> H rt cr H i I—1 o 0) H -N O 0 ~" cn ZD rn ID o cn X X X H O 3 a-rt CT)_ O cr cn D — cn o O 9T 17: changes. The reasons for t h i s scale-dependent response probably l i e i n the speed with which moisture i s transmitted to the earthflow shear plane. A l l active earthflows have many tension cracks which have the p o t e n t i a l to provide pathways along which r a i n f a l l (and snowmelt) can i n f i l t r a t e . However, as most small inland flows are 2m deep or l e s s , tension and desiccation cracks feed water d i r e c t l y to the shear plane. At many s i t e s graphic descriptions of treacherous l i q u i d mud during flow s i g n i f y the rapid saturation of i n a c t i v e flow debris v i a an interconnected network of f i s s u r e s , shear planes, dessi c a t i o n cracks and other zones of high permeability (Grove 1952, Arber 1941, Lang 1943). Larger flows, most of which have been shown to respond to seasonal or longer term e f f e c t s (Figs 1.4 and 1.7) probably do so because t h e i r cracks and fiss u r e s do not penetrate to a s u f f i c i e n t depth to enable rapid i n f i l -t r a t i o n to the shear plane. However, they are capable of responding to rapid changes i n groundwater l e v e l , such as those caused by human interference. VanDine (1983) quotes a report by Huculak and Brawner (1961); 'At Drynoch landslide the rate of movement increased w i t h i n 24 hours following com-mencement of d r i l l i n g (with water) each Monday and decreased within 48 hours a f t e r cessation of d r i l l i n g each Friday' . The fa c t that no such n a t u r a l l y induced response i s observed from the majority of large flows (Appendix G) i s good evidence that p r e c i p i t a t i o n i s not able to penetrate r e a d i l y to the shear plane. Indeed, very slow i n f i l t r a t i o n , precluding rapid groundwater l e v e l changes, i s predictable as the hydraulic conductivity of most clayey -8 -9 s l i d e material i s probably very low (for example 10 to 10 m/s at P a v i l i o n (Bovis 1985)). Consequently most large flows are only a f f e c t e d by climat-i c a l l y induced piezometric f l u c t u a t i o n s , and not by i n d i v i d u a l p r e c i p i t i o n events. Several authors (Crandell and Varnes (1961), Keefer (1977), Rawlings (1984)) have speculated upon the age of i n a c t i v e portions of large earthflows 18 currently either i n a c t i v e or only marginally a c t i v e . In the study area Bovis (1985) noted several active earthflows which have stranded l a t e r a l deposits i n d i c a t i n g s ubstantial drawdown since the flow maximum; i n addition a number of flows are no longer active. There i s some s t r a t i g r a p h i c evidence suggesting long-term movement during a l l , or some, phases of the p o s t g l a c i a l . For example, Mazama tephra i s seen i n l a t e r a l deposits at P a v i l i o n (Bovis 1985), underlying a l a t e r a l at Drynoch (VanDine 1980) and blanketing the Lookout Creek flow i n Oregon (Swanson and James 1975). The write r has noted Bridge River tephra i n an inac t i v e shear plane at Big Bar earthflow and the Tunnel flow o v e r l i e s a s i l t y e o l ian horizon which i s Hypsithermal i n o r i g i n ( a l l t h i s information is summarised i n Bovis .(1985)). As several earthflows described i n the previous paragraph are now stable, e x i s t i n g chronologic evidence i n the study area suggests episodic, rather than continuous, earthflow movement has shaped the earthflows observed i n the study area. 1.4 Hypothesis and objectives Most earthflows within the study area are medium to large-sized; most 4 2 5 2 are over 10 m in area and the majority over 10 m . Therefore, the analysis i n the preceding part of t h i s section suggests that they respond to seasonal and/or longer term p r e c i p i t a t i o n f l u c t u a t i o n s . In addition, the morphology of the flows, i n d i c a t i n g , i n many cases, episodic movement, suggests a l t e r -nating periods of movement and quiescence. Therefore, the ce n t r a l hypothesis of t h i s thesis i s that large earthflows i n south-central B r i t i s h Columbia respond to environmental con-t r o l s i n such a way that movement phases correspond with periods of above average e f f e c t i v e p r e c i p i t a t i o n . Consequently, the s u i t e of earthflows i n the study area preserves a record of Holocene c l i m a t i c f l u c t u a t i o n s , a record analogous with that described from n e o g l a c i a l moraines (Ryder and Thomson, 19 1986) and h i s t o r i c g l a c i a l fluctuations (Luckman, Harding and Hamilton, 1986). The thesis w i l l proceed by o u t l i n i n g the phys i c a l environment of the study area. Next, morphological, geological and geotechnical properties of earthflows i n the study area w i l l be described b r i e f l y . The following three sections address the central hypothesis d i r e c t l y , demonstrating that, i n the study area, earthflow movement i s rela t e d to c l i m a t i c f l u c t u a t i o n s . The hypothesis i s treated at three timescales: h i s t o r i c time, (1928 to present), the past 300 years, and Holocene time. In Chapter 4, earthflow movement i s inf e r r e d from a i r photograph chronosequences, while c l i m a t i c fluctuations are obtained from an analysis of c l i m a t i c records. In Chapter 5 both earthflow movements and a c l i m a t i c chronology are i n f e r r e d from a study of tre e - r i n g widths. F i n a l l y , i n Chapters 6 and 7, earthflow movement over Holocene time i s investigated using a v a r i e t y of techniques including radiocarbon dating, tephrochronology and pol l e n analysis. 20 CHAPTER 2 THE STUDY AREA  2.1 Earthflow locations i n south-central B.C. A number of recent regional landslide analyses have demonstrated that there i s a concentration of large, slow moving earthflows i n the i n t e r i o r of south-central B r i t i s h Columbia (Eisbacher 1979, Cruden 1984, Bovis 1985). Individual earthflows, p a r t i c u l a r l y those at Drynoch and P a v i l i o n , which are r e a d i l y viewed from major transportation a r t e r i e s , have been noted i n many geological reports (Begbie (1871), Dawson (1896), Reinecke (1920), D u f f e l l and McTaggart (1952), Leech (1953), T r e t t i n (1961), Thurber Consultants Ltd (1971), Heginbottom (1972), Ryder (1976, 1981a), VanDine (1980, 1983), Mathews and Rouse (1984)). These reports a l l concern the 921 and 920 NTS map areas; however, a report by Evans (1982) describes a f a s t e r moving earthflow from the Quesnel area, somewhat further north. On the basis of t h i s i n f o r -mation, the author undertook an a i r photograph search of the 921,J,K,N,0,P and 93A,B,C,G map areas. Only the 921, 92J and 920 map areas y i e l d e d slow earthflows and recognition of these features formed the database for an analysis of common geologic and topographic conditions. F i f t y - t h r e e earthflows at twenty locations, of which over h a l f were neither mapped nor described i n p r e - e x i s t i n g reports. were i d e n t i f i e d as a r e s u l t of the search (Fig 2.1). At several locations the flows occurred too close together to be mapped i n d i v i d u a l l y , so these multiple features are indicated by a number i n parentheses following the c l u s t e r name i n the key. Previously unrecognised earthflows were named a f t e r nearby creeks, mountains or other natural features recognised on the 1:50,000 NTS topographic map or a f t e r homesteads not marked on the map, while a l l p r e - e x i s t i n g names were retained. In clusters of multiple earthflows the i n d i v i d u a l flows were recognised by an alphabetic s u f f i x (for example, at Red Mountain, A-G). 21 B BI Bu C Ca Ch Cm D F Ft G Ashcroft ; G I H Ha LL Big Bar Black Dome (3) Burkholder (3) Canyon (2) Canoe Creek Churn Creek (6) i L Camoo Creek P Drynoch (2) R Flapjack T Fountain Y Grinder (5) Yo GilIon Heginbottom Hat Creek Lac l a Mer (3) Lone Cabin P a v i l i o n (2) Red Mountain (7) Tunnel (5) Yalakom (4) Yodel (2) Spences Bridge ScaleCKm) 10 10 30 F i g . 2 .1 E a r t h f l o w l o c a t i o n s i n s o u t h - c e n t r a l B r i t i s h C o l u m b i a Numbers i n p a r e n t h e s i s i n d i c a t e m u l t i p l e o c c u r r e n c e s 22 2.2 Topography Holland (1964) divides the study area into four physiographic regions: the southern plateau and mountain area (part of the I n t e r i o r System), and three mountain ranges, the Camelsfoot, the Clear Ranges and the Marble Range. These are shown i n F i g 2.2. The f i r s t of these, the I n t e r i o r Plateau, i s a remnant of a T e r t i a r y subaerial erosion surface, which, i n the study area i s preserved between 900-1500m. In the Pliocene, u p l i f t of the Coast Moun-tains resulted i n stream i n c i s i o n , which carved steep-sided rejuvenated vall e y s (for example the Fraser and the Thompson i n the study area) below the former ' r o l l i n g ' T e r t i a r y topography. G l a c i a l and f l u v i o g l a c i a l erosion have further dissected the former surface. Two of the three mountain ranges i n F i g 2.2, the Marble Ranges (highest point 2,260m), and the Clear Ranges (highest point 2,320m) were overridden by i c e and so have r e l a t i v e l y subdued topography. The fourth physiographic region, the Camelsfoot mountains, i s higher and shows a t r a n s i t i o n from northeast to southwest, from subdued topography (as at Red Mountain e l . 2,432m) to the more rugged t e r r a i n c h a r a c t e r i s t i c of the Coast Mountains (for example i n the Shulaps, highest points Big Dog Mountain, 2,846m, and Shulaps Peak, 2,862m). The Shulaps have many steep walled cirques (see F i g 2.4) and continuous ridge crests and l o c a l r e l i e f exceeding 2,000m i n the southern part of the Range (Leech 1953). The regional topography has undoubtedly contributed to earthflow con-centration i n the study area. Many flows occur on steep v a l l e y side slopes cut into the deeply dissected I n t e r i o r Plateau by both f l u v i a l and g l a c i a l erosion. Therefore, as i n c i s i o n dates only from the g e o l o g i c a l l y recent Pliocene, earthflow concentration i n the study area may be, i n part, a geomorphic response to topographic evolution. However, the occurrence of earthflows rather than other forms of mass movement i s due to geologic fac-tors explored i n the next section. 23 F i g 2.2 Topography of the study area, with inset showing loc.atioa_.of. ._. i Atmospheric Environment Service (AES) stations shown i n F i g 2.5/ 24 2.3 Regional geology Geologic units are shown i n F i g 2.3, adapted from Roddick, Muller and Okulitch (1979). For s i m p l i c i t y , several d i f f e r e n t rock units are combined i n each category i n F i g 2.3, either by v i r t u e of s i m i l a r l i t h o l o g y (for ex-ample, the granodiorite u n i t s ) , or because of s i m i l a r , or close, age (for example, most T r i a s s i c , Pennsylvanian and Permian rocks f a l l i nto one cate-gory). Letters on the map denote the earthflow c l u s t e r s recognised i n F i g 2.1. It can be seen from F i g 2.1 and Appendix H that most earthflows occur i n Eocene and Mio-Pliocene volcanics and associated sediments; i n Lower Cretaceous volcanics and associated sediments; or else i n T r i a s s i c u l t r a b a s i c s . Furthermore, the majority of earthflows occur i n , or j u s t outside, a wedge-shaped area bounded by the Fraser and Yalakom f a u l t s . The following discussion w i l l concentrate on explaining these earthflow concen-tr a t i o n s rather then providing a comprehensive geologic h i s t o r y of the study area. The oldest rocks i n the study area are the T r i a s s i c Shulaps ultramafic rocks and the Bridge River Group. Only one earthflow (Camoo) i s located i n Bridge River Group rocks whereas nine features were i d e n t i f i e d from the smaller area of ultramafics (see F i g 2.3). The Bridge River Group i s prob-ably part of a subduction complex and represents u p l i f t e d P a c i f i c ocean f l o o r (Coney et a l . 1980), while the Shulaps i s i n t r u s i v e . Serpentinisation of the Shulaps u l t r a b a s i c s i s of p a r t i c u l a r i n t e r e s t i n t h i s study because earthflows occur on serpentinised rocks rather than on t h e i r unaltered equivalents (dunite, p e r i d o t i t e ) . Leech (1953) records that serpentinisation i s uniform over large areas i n the Shulaps; however, although earthflows occur i n areas where the rocks are strongly altered, i t should also be noted that they traverse the Yalakom Fault zone. Therefore the respective influences of s e r p e n t i n i s a t i o n and f a u l t i n g cannot be deter-mined at t h i s location. F i g . 2.3 Simplified geologic Roddick, Muller and earthflow names are map of the Hutch i nson explained i study area, adapted from ( 1 9 7 9 ) • Abbreviations of n Figure 2 . 1 . Key for Figure 2.3 26 IMPvb Miocene - basalt flows, s i l l s and dykes Meg Miocene - gravels Mpw Miocene - shale, t u f f , greywacke Evr Eocene - r h y o l i t e , dacite {EOK Eocene and Oligocene - Kamloops Group, dacite, basalt ECB Eocene - Coldwater Beds, conglomerate, sandstone uKKW Upper Cretaceous - Kingsvale Group, arkose, conglomerate, greywacke, andesite, breccia, basalt 1KJM Lower Cretaceous - Jackass Mountain, greywacke, conglomerate 1KSB Lower Cretaceous - Spences Bridge Group, andesite, basalt rJKRM Jurassic and Cretaceous - Relay Mountain Group, shale, greywacke, conglomerate mJp Middle Jurassic - shale, conglomerate, sandstone mJH Middle Jurassic - Hazelton Group, andesite, basalt uTN Upper T r i a s s i c - Nicola Group, andesite, basa l t , limestone, a r g i l l i t e Mpc Mesozoic - g r a p h i t i c quartzose p h y l l i t e Msn Mesozoic - s c h i s t , gneiss 'PTP Paleozoic and T r i a s s i c - P a v i l i o n Group, chert, a r g i l l i t e , s i l t s t o n e , b a s a l t , andesite PTBR Paleozoic and T r i a s s i c - Bridge River Group, chert, a r g i l l i t e , c p h y l l i t e PMC Permian - Marble Canyon, limestone PPCC Pennsylvanian and Permian - Cache Creek Group - limestone 'ETgd Early T e r t i a r y lKgd Lower Cretaceous JJgd Jurassic granodiorite mJgd Middle Jurassic ^lTgd Late T r i a s s i c PTub Permian and T r i a s s i c - s e r p e n t i n i t e , p e r i d o t i t e 27 Three major l i t h o l o g i c a l units formed during the Cretaceous: the Jackass Mountain Group i n the west and the Spences Bridge and Kingsvale Groups i n the east. The Jackass Mountain Group i s a nonmarine sedimentary sequence of at least 5,150m thickness (Jeletsky and Tipper 1963) l a i d down on the northeast and east flank of the p a r t l y marine Tyaughton Trough. This was a long, narrow, northwest trending sedimentary basin bounded by landmasses on the southwest and northeast which provided g r a n i t i c source rocks. The upper and lower parts of the succession are s i m i l a r , being greywacke, shale, t h i n pods and lenses of conglomerate and arkose, whereas the middle zone succession comprises coarser rocks, notably the French Bar Formation, a boulder conglomerate, suggesting more rapid u p l i f t of the source area during t h i s t r a n s i t i o n a l phase. No earthflows were found i n Jackass Mountain Group rocks. The Spences Bridge Group occupies a wide b e l t trending northwest from the Nicoamen Plateau to the north of P a v i l i o n and comprises extremely v a r i -able lavas and p y r o c l a s t i c rocks ( D u f f e l l and McTaggart 1952). Lava flows of intermediate composition are most common but r h y o l i t e s , basalts, breccias and agglomerates a l l form s i g n i f i c a n t parts of the Group. Sediments occur as interbeds; sandstones, arkoses and t u f f s , often with large carbonised wood fragments, are found. The upper Cretaceous Kingsvale Group o v e r l i e s the Spences Bridge Group and i n the study area the u n i t comprises mainly b a s a l t i c and andesitic flows with minor sediments. Middle and upper Cretaceous con-t i n e n t a l sediments located near P a v i l i o n (Monger and McMillan 1984), not ascribed to the Kingsvale Group, have c l a s t s i n d i c a t i n g that they too are derived from volcanic source rocks. F i g 2.3 shows that earthflows occur i n the Spences Bridge and Kingsvale Groups, but not i n the broadly contemporaneous Jackass Mountain Group. This i s probably a consequence of the d i f f e r i n g provenance of the debris c o n s t i -t u t i n g each unit: the g r a n i t i c c l a s t s of the Jackass Mountain Group are slow 28 to weather, while those c l a s t s derived from volcanic and v o l c a n i c l a s t i c rocks, which comprise the sediments of the Kingsvale and Spences Bridge groups, r e a d i l y weather to the clay minerals which are a major constituent of most earthflow material. During the Eocene, as i n the Cretaceous, there was extensive volcanism. Mathews and Rouse (1984) describe v a r i a b l e Eocene volcanics from the study area, many dated at 45-50 m i l l i o n years. They also map assorted sediments including a prominent conglomerate u n i t , with predominantly volcanic c l a s t s , p a r a l l e l to and west of the Fraser Fault and o v e r l a i n by clays r i c h i n montmorillonite. Further east, s i m i l a r b e n t o n i t i c clays were deposited i n the Hat Creek basin (Rawlings 1984). Many earthflows occur i n Eocene sediments and volcanics (Fig 2.3), p a r t i c u l a r l y i n the sediments and more basic flows. As with the Cretaceous rocks, r e a d i l y weatherable volcanics and associated sediments provide c l a y - r i c h weathered debris which i s extremely susceptible to earthflow a c t i v i t y . Extensive Miocene outcrops of both volcanics and sediments are re-ported from the study area. Church (1980, 1982) gives K-Ar Miocene dates for b a s a l t i c lava and agglomerate on Black Dome Mountain and Flapjack Peak (Fig 2.2), while Tipper (1978) maps large areas of Miocene basalt west of the Fraser Fault zone, (although no K-Ar dates d e f i n i t i v e l y assign these l a t t e r deposits to the Miocene). Mathews and Rouse (1984) report poorly consolidated Miocene sediments from the Gang Ranch area, while Tipper maps Oligo-Miocene volcanics north of the Hungry Va l l e y F a u l t . Four earthflows are developed i n Miocene outcrops, three i n soft shales and one i n basalt. U p l i f t of the Coast Mountains took place i n the mid-Pliocene (Roddick 1960). Two major phases of volcanism occurred during t h i s period, r e s u l t i n g i n extensive basalt flows. They have been K-Ar dated at 6-10 m i l l i o n years and 2-3 m i l l i o n years (Bevier 1983). Only one earthflow (the Flapjack flow) i s developed i n Oligo/Mio/Pliocene b a s a l t s ; however, basalt cappings occur 29 adjacent to many earthflows and may play an important r o l e i n influencing groundwater flow. This i s explored i n the next chapter. 2.4 G l a c i a l h i s t o r y and Quaternary geology Tipper (1971) shows that during the Fraser G l a c i a t i o n c e n t r a l B r i t i s h Columbia was covered with i c e flowing from both the Coast and Cariboo Moun-tains (Fig 2.4). Tipper considers.that some g l a c i a l grooves at high el e -vations were not produced by the l a s t (Fraser) G l a c i a t i o n , because ice flow and groove orientations d i f f e r . This i n t e r p r e t a t i o n may also apply to er-r a t i c s found on Rex Peak (2,190m) i n the Shulaps Range (Leech 1953) above the maximum i c e elevation reached during the l a s t g l a c i a t i o n , and to r e l i c t patterned ground on the Clear Range (Ryder 1976) and on Red Mountain i n the study area. During the main part of the Fraser G l a c i a t i o n i c e movement was probably quite slow i n the northern part of the study area as i t represented an 'ice divide' where ic e sheets from the Coast and Cariboo Mountains met, r e s u l t i n g i n i c e flowing both north and south as shown i n F i g 2.4. Consequently, most of the study area suffered minimal g l a c i a l erosion from t h i s r e l a t i v e l y slow-moving i c e sheet. This i s r e f l e c t e d i n patchy ground moraine, v a r i a b l e i n both extent, depth and composition throughout the study area. Heginbottom (1972) characterises ground moraine i n the NTS 92 0 area as 'greyish brown, stony, sandy to sandy-loam t i l l , with a low cl a y content' and recognises discontinuous deposits from .3m to at least 5m t h i c k . The author has observed that i t generally r e f l e c t s l o c a l l i t h o l o g y e s p e c i a l l y at high elevations. Ablation i n the study area i s thought to have followed Fulton's (1967) four stage model, with gradual downwasting of the i c e sheet r e s u l t i n g i n , f i r s t , the emergence of upland areas, then a s e r i e s of stagnant i c e tongues with ice marginal drainage, and f i n a l l y dead i c e with e n g l a c i a l and s u b g l a c i a l drainage. The stagnant i c e phase caused two s t r i k i n g suites of features i n the study area. In upland v a l l e y s , kame terraces were l a i d down 2.4 Ice flow over the study area and g l a c i a l and f l u v i o g l a c i a l features and meltwater channels were cut, some through s o l i d rock and many across the grain of present topography, such as those near Burkholder Creek i n the Yalakom Valley. In lowland areas the stagnant i c e phase was characterised by p r o g l a c i a l lakes. In the study area g l a c i o l a c u s t r i n e s i l t s occur 350m above present-day r i v e r l e v e l s , suggesting that a s u b s t a n t i a l i c e plug must have occupied the early p o s t g l a c i a l Fraser Valley. Non-glacial Quaternary deposits are characterised by Pleistocene v a l l e y - f i l l sediments, c o l l u v i a l blankets, landslides and a l l u v i a l fans formed on a v a r i e t y of p o s t g l a c i a l surfaces (Ryder 1971). The l o c a t i o n of the Mazama ash (6,600 BP) near the surface of many p o s t g l a c i a l sedimentary deposits, suggests that i n i t i a l p o s t g l a c i a l sedimentation, the 'paraglacial' sedimentation cycle of Church and Ryder (1972), was rapid, with present day processes contributing l i t t l e sediment to what are e s s e n t i a l l y f o s s i l fea-tures . Deglaciation of the study area was probably complete before 10,000 BP. The p o s t g l a c i a l climate was, warmer than present u n t i l about 7,000 BP when a slow temperature decline began, bringing present day conditions by about 3,000 BP (Clague, 1981) . Evidence for the chronology of d e g l a c i a t i o n and p o s t g l a c i a l c l i m a t i c changes is analysed i n d e t a i l i n chapter 8. 2 .5 Fraser River diversion Tipper (1971) noted that many large t r i b u t a r i e s to the Fraser north of Williams Lake flow northwards to t h e i r confluence with the Fraser. This, together with observations of Miocene r i v e r gravels, indicates that a r i v e r system once discharged northwards through the Peace River system. Capture i s thought to have taken place between Williams Lake and Soda Lake, possibly near the C h i l c o t i n River confluence. Basalts flowing from the north created a lava dam (this i s indicated by foreset breccias near Canoe Creek, Mathews and Rouse 1984) i n d i c a t i n g northward r i v e r flow at 1.3-2.9 m i l l i o n years. Therefore diversion must have occurred during the l a s t 1-2 m i l l i o n years, 32 meaning that much of the Fraser v a l l e y i n the study area i s a r e l a t i v e l y recent feature. 2.6 Climate The climate of the study area, measured by the Atmospheric Environment Service stations shown i n F i g 2.2, i s strongly influenced by topography. Table 2.1, showing el e v a t i o n a l trends i n p r e c i p i t a t i o n , demonstrates that p r e c i p i t a t i o n i n inci s e d v a l l e y s ( i n c i s i o n c i r c a 300m) i s 15-20cm, while p r e c i p i t a t i o n on the plateau surface (c. 1,000m) i s 30-35cra. Totals for mountainous areas i n the west of the study area are doubtless considerably higher with a substantial proportion of winter p r e c i p i t a t i o n f a l l i n g as snow. In addition, summer temperatures within the study area are mainly determined by elevation. The study area straddles the boundary between a coastal (winter maximum) regime and one dominated by summer p r e c i p i t a t i o n , c h a r a c t e r i s i n g the Inte-r i o r Plateau of B.C.. The boundary occurs i n a northwest-southeast trending zone across the west of the study area (see Figs 2.2 and 2.5). Plots of monthly mean annual p r e c i p i t a t i o n (Fig 2.5) for Big Creek, L i l l o o e t and Lytton i l l u s t r a t e well the s h i f t from a predominantly coastal (winter/frontal) regime at Lytton, i n the south of the study area, through L i l l o o e t , with both winter and summer maxima, to Big Creek and Dog Creek, both located on the I n t e r i o r Plateau surface and both showing a d i s t i n c t summer p r e c i p i t a t i o n maximum. Another c h a r a c t e r i s t i c of s p a t i a l p r e c i p i -t a t i o n trends i s the increasingly important r o l e of winter p r e c i p i t a t i o n , and hence snow rather then r a i n , as elevation increases. For example, Ashcroft (305m) shows a d i s t i n c t summer maximum, while Mamit Lake (1,006m) has a s i g n i f i c a n t contribution from winter p r e c i p i t a t i o n and Highland Valley (BCCL) (1,470-1,494m) has a winter maximum. Mean annual temperatures show a predictable decline i n annual range with elevation, due mainly to a decline i n summer temperatures with elevation J F M A M J J A S O N D Lytton J F M A M J J A Li 1 looet S 0 N D J F M A M J J A S O N D Hat Creek r 100 50 cm J F M A M J J A S O N D Dog Creek A J F M A M J J A S O N D Shalath J F M A M J J A S O N D Highland Valley BCCL J F M A M J J A S O N D Mamit Lake J F M A M J J A S O N D Big Creek i I J J F M A M J J A S O N D Ashcroft J F M A M J J A S O N D Bralorne 100 •50 cm F i g 2.5 Seasonal p r e c i p i t a t i o n d i s t r i b u t i o n at AES stations within and close to the study area Table 2.1 Climatic s t a t i s t i c s from the study area. Annual Station pptn. • - j . Mean Record Yrs Temp Record Yrs E l Max. Min. Range Ashcroft V.CBU 151.3 1924-1969 33 9.0 1957-1969 9] 305 21.6 -6.1 27.7 Big Creek 318.5 1905-1980 68 2.7 1905-1980 71 1131 14.0 -11.5 25.5 Bralorne 620.8 1924-1963 28 5.0. 1934-1963 19 1015 15.6 -4.9 20.5 Dog Creek 388.4 1945-1960 15 3.8 1944-1960 14 1027 16.0 -10.4 20.4 Hat Creek 298.9 1961-1980 11 3.4 1961-1980 20 923 15.0 -10.6 25.6 Highland Valley 374.7 1967-1976 9 3.1 1967-1984 12 1482 14.1 -8.8 22.9 L i l l o o e t 251.7 1917-1970 23 . 9.8 1918-1942 25 290 21.9 -3.3 25.2 Lytton 1 458.9 1944-1983 38 10.0 1944-1970 24 258 22.0 -3.7 25.7 Lytton . 429.6 1917-1948 24 10.3 1926-1970 45 175 21.8 -2.9 24.7 Mamit Lake 329.3 1924-1965 32 3.6 1950-1965 10 1006 15.2 -9.0 24.2 Shalath 512.8 1963-1985 18 9.6 1963-1985 16 244 21.1 -3.3 24.4 35 (Table 2.1). Summer temperatures range from monthly maxima of 22°C at Lytton ( e l . 175m), to 14°C at Big Creek ( e l . 1131m). The winter range i s smaller for the same stations (2-3°C). As summer p r e c i p i t a t i o n d i f f e r s considerably between the stations i n F i g 2.5, low elevation s i t e s having low summer pre-c i p i t a t i o n (such as Lytton) experience a substantial moisture d e f i c i t over the summer months. These s t a t i s t i c s suggest that, i n the study area, most groundwater recharge occurs during the winter months at low elevations and from spring snowmelt at higher elevations ( i t i s shown i n Chapter 4 that summer p r e c i p i t a t i o n i s not s u f f i c i e n t to meet evapotranspiration require-ments). Marked temporal trends are seen i n p r e c i p i t a t i o n data from stations with long records (Barrett 1979), the main features being a warm, dry period i n the 1920's and 1930's. This pattern r e s u l t s i n a s l i g h t reduction of average p r e c i p i t a t i o n at s i t e s with long records i n comparison with stations dating back only to the 1950's or l a t e r . It i s not thought that these temporal fluctuations materially a f f e c t the conclusions about the s p a t i a l d i s t r i b -ution of p r e c i p i t a t i o n i n the study area. They do, however, have s i g n i f i c a n t implications for earthflow movement, and are explored i n depth i n Chapter 4. 2.7 Vegetation The major vegetation zones within the study area are shown i n F i g 2.6, adapted from Krajina (1969). Although the portions of the study area i n 921 and 92J have been mapped at a scale of 1:125,000, only a small part of the 920 sheet has been surveyed at t h i s scale. Therefore Krajina's small scale map was used as i t covered the whole study area with uniform d e t a i l . F i g 2.6 shows f i v e major vegetation zones, which are c o n t r o l l e d p r i m a r i l y by elevation (Fig 2.2). In semi-arid v a l l e y s such as the Fraser Valley near Big Bar, a r a i n shadow e f f e c t gives r i s e to very low p r e c i p i t a t i o n and high temperatures, producing an open Ponderosa pine-bunchgrass zone, with large 36 F i g 2.6 Vegetation zones within the study area t r a c t s of Artemesia t r i d e n t a t a (sagebrush) i n the d r i e s t areas. At a number of semi-arid earthflow s i t e s the regional vegetation i s modified by groundwater discharge. This enables moisture seeking vegetation, p a r t i c -u l a r l y aspen, ( Populus tremuloides ) to grow at most locations on active earthflow s i t e s (as for example at P a v i l i o n ) . Growth rates d i f f e r i n g by up to 10 times for s i m i l a r species only 50ra apart were noted at Heginbottom as a r e s u l t of l o c a l i s e d groundwater discharge zones. At higher elevations (above 1,000m) I n t e r i o r Douglas f i r ( Pseudotsuga  menziesii ) predominates, merging at high elevations into Engelmann spruce ( Picea engelmanii ) at about 1,500m. Kra j i n a shows the northern part of the study area as f a l l i n g within the Cariboo aspen-lodgepole p i n e - I n t e r i o r Douglas f i r zone. However, Artemesia and Ponderosa pine were observed to extend further north along Fraser River to the northern extreme of the study area. In addition, throughout much of the higher elevation parts of the study area, western white pine ( Pinus a l b i c a u l i s ) i s dominant from about 1,800m to t r e e l i n e , supplanting the mapped Engelmann spruce. The highest parts of the study area are occupied by alpine tundra; earthflows at Red Mountain and at Big Dog Mountain l i e p a r t i a l l y within t h i s zone. 38 CHAPTER 3 CHARACTERISTICS OF EARTHFLOWS IN THE STUDY AREA  3.1 Morphology Several authors (Zaruba and Mencl 1960, P r i o r et a l . 1968, Keefer, 1977) have recognised three morphological zones on earthflows. The f i r s t , and upper zone, i s a bowl-like source area; the second a flow track zone, characterised by l a t e r a l deposits and c l e a r l y marked l a t e r a l shear zones; and the t h i r d a toe zone, which may take the form of either a steep fronted snout, a fan, or which may be absent i f the earthflow toe i s eroded by a creek or r i v e r . This morphology i s independent of s i z e , as flows ranging from 0.001 2 2 km (Keefer i n C a l i f o r n i a ) to several km i n t h i s study (for example the P a v i l i o n and Red Mountain flows i n Appendix H) show the same three charac-t e r i s t i c zones. A l l three zones are i l l u s t r a t e d i n F i g 1.1, and sections 3.1.1 to 3.1.3 describe i n d e t a i l the morphology of each zone, i n order to demonstrate the location of s i t e s which were c r i t i c a l for obtaining chronologic information. 3.1.1 Slump bowls The majority of earthflows i n B r i t i s h Columbia head i n slump bowls, which are often described as ' c i r q u e - l i k e ' (Leech (1953), VanDine (1980), and many others). In long p r o f i l e the slump bowl has a steep back wall and a much f l a t t e r area downslope of the headscarp (see F i g 1.1). For the pur-poses of t h i s discussion, the upper zone terminates downslope where l a t e r a l deposits are found adjacent to the flow. Upslope from the headscarp tension cracks are often seen, frequently i n conjunction with a gradual b a c k t i l t i n g of s l i c e s before they became en-trained into the earthflow. The Grinder Creek (A) photographs i n Appendix A (p. 312-314) show well-marked, arcuate cracks above the a c t i v e headscarp; and Yalakom flows (A) and (B) have a s e r i e s of b a c k t i l t e d ridges above t h e i r headscarps, terminating i n a steep headwall which may be a former cirque wall (Appendix A). At these s i t e s movement takes place i n bedrock, while any 39 overlying unconsolidated sediments are rafted along. Depressions between s t a b i l i s e d headscarp s l i c e s , and depressions i n s t a b i l i s e d slump bowls pro-vide useful s i t e s for lake and bog coring. In addition, many trees growing on active headscarps have a useful record of movement recorded i n compression wood. On smaller flows developed i n poorly consolidated sediments (for ex-ample Black Dome (A)), and on scarps which mark the r e a c t i v a t i o n of some older flows (for example at Grinder Creek (A), Heginbottom and P a v i l i o n (A)), no steep headscarp i s seen. Instead, many tension cracks are seen transverse to the flow and the whole 'headscarp area' i s a jumble of small, loose seg-ments of flow material. Keefer (1977) c a l l e d these features 'pullaway slopes' and they are well i l l u s t r a t e d by a photograph i n Sharpe (1938). No useful chronologic information was obtained from the headscarp of flows i n t h i s category. 3.1.2 Flow track The flow track zone (Fig 1.1) i s an area where flow occurs i n a well defined channel for a distance several times the width of the flow zone. Within the flow track zone flow takes place mainly as t r a n s l a t i o n a l s l i d i n g on slickensided basal and l a t e r a l shear planes ( F i g 3.1). Slickensides show that flow can take place u p h i l l over short distances, p o s s i b l y to override small i r r e g u l a r i t i e s i n the flow path or to accomodate a swelling lobe ad-vancing down the flow track. Conversely, s l i c k e n s i d e s such as those i n F i g 3.1 dip down the flow more steeply than the p r e v a i l i n g earthflow gradient, implying drawdown. Latera l deposits, i n the form of long sinuous ridges, are found adja-cent to many flow tracks, up to 30m above the active flow. At several l o -cations successive sets are found, paired on e i t h e r side of the flow, stretching for up to 2 km. These l a t e r a l ridges range i n height from a few centimetres up to 15m (Yalakom A). Flows with many sets of l a t e r a l deposits, 40 F i g 3.1 S t r i a t i o n s on a fo rmer s h e a r p l a n e a t G r i n d e r C r e e k (A) 41 i n d i c a t i n g several phases of a c t i v i t y , have a high length to breadth (1/b) r a t i o for the flowtrack zone (notably Drynoch, Grinder and Red Mountain B, Table 3.1) and are a l l currently a c t i v e . Conversely, flows with no l a t e r a l s , or only one set, generally have a low 1/b r a t i o (Flapjack, Camoo Creek). The depressions between en echelon l a t e r a l deposits, and between l a t e r a l deposits and adjacent stable ground, provide useful coring s i t e s which date cutoff by l a t e r flow movements. Several earthflows have e a r l i e r phases with a low 1/b r a t i o and l a t e r phases, where only part of the flow has reactivated, with a higher 1/b r a t i o . In Table 3.1 Churn Creek, for example, i n i t i a l l y formed with a low 1/b r a t i o . Phase 4 (Bovis 1985), involving the upper part of the flow, created an earthflow with a much more c l e a r l y defined flow track. S i m i l a r l y , a number of small flows, one of which i s s t i l l a c tive, have formed on the main flow toe. These also have a higher 1/b r a t i o (Table 3.1) than does the main flow. This evidence suggests that long and narrow earthflows are most e f f i c i e n t , as they are able to move under conditions which do not produce movement i n flows with a low 1/b r a t i o . This i s i n t u i t i v e l y reasonable as, for a given flow volume, a cross-section approximating a s e m i c i r c l e w i l l have least f r i c t i o n a l resistance. Inspection of the long p r o f i l e s i n F i g 3.2 shows that within the flow track zone, l a t e r a l s are highest and most continuous where the gradient i s steepest (see Red Mountain B and Drynoch A). In the flow track, most movement takes place at or near the marginal (and basal) shear surfaces. However, zones of 'extending flow' (Glen 1970) comparable to i c e f a l l s on g l a c i e r s occur above sudden increases i n slope. 'Crevassed' zones are seen i n Grinder (A) and (B) (Fig 3.3) and at Yodel (Appendix A, pp 321-322); i t can be seen that they are oriented at various angles to the flow axis making a 'V' shape pointing upslope. Discrete lobes can be i d e n t i f i e d moving down well-defined flow tracks. Those at Grinder Creek (Figs M2a to f ) are the best developed, with lobes Table 3.1 !Length to breadth (1/b) r a t i o s of flow track zones  for earthflows i n the study area Flow 1/b r a t i o Big Bar 3-12 Black Dome (A) 7 (B) 5 (C) 5 Burkholder (A) 10 (B) 5 Canoe 5 Canyon 4-8 Camoo Poorly defined Churn (A) 1.6-2.9 (B) 13 (C) 3 •CD) 1.6 (E) 6-12 (F) 5-10 Drynoch (A) 2.6-21.7 (B) 1.5-6 Flapjack 4 Gibbs Creek 7.5-20.8 G i l l o n Creek 2-8.6 Grinder (A) 1.9-35 (B) 3.5-10 (C) 5 (D) 3-10 (E) Poorly defined Hat Creek Poorly defined Heginbottom 4-13 Kettlebrook : Creek 8-20 Lac l a Mer (A) Poorly defined (B) 15 (C) 2-6 Lone Cabin 4 P a v i l i o n (A) 1.2-5.2 (B) Poorly defined Red Mount- (A) 2.7-18 ain (B) 5.1-20 (C) 4.3 '(D) 4.3 (E) 6.7-20 (F) 12.5 (G) 10 Tunnel 3-12 Yalakom (A) 3-3.4 (B) 1.9-2.9 (C) 1.6-4.0 (D) 2.3 Yodel (A) 15 (B) 5 43 r 1200 900 Drynoch (A) L a t e r a l s highest and most continuous - 6O0 300 rt Eroded toe Flow track See l i n e of section of F i g 1.1 Red Mountain (B) lKm Snout Flow track See long p r o f i l e F i g 4.1 1800 F i g 3.2 Selected long p r o f i l e s to show flow tracks with well developed l a t e r a l s F i g _ 3 . 3 "Crevassed" zone above lobe at Grinder Creek. (B) F i g 3.4 Lobe moving down Grinder Creek (A) earthflow FLovJ d i l a -t e d and very loos« Many Flow a c c e l e r a t i o n e a r plane . o 5 Diagram o f t e n s i o n zone assoc i a t e d w i t U lobe F i g 46 overriding material beneath (Fig 3.4). At the top of the lobe i l l u s t r a t e d i n F i g 3.4, the active flow i s l e v e l , with several sets of l a t e r a l deposits, while at the base of the lobe i t i s 5-7m lower (Fig 3.5). Upslope from the lobe a f l a t t e r area i s found. This morphology suggests that the earthflow i s thicker at the lobe than elsewhere, a fact which was noted by the C a l i f o r n i a Dept of Water Resources (1973) i n the E e l River Basin. Lobate flow explains the observation of slickensides angled more and less steeply than the flow gradient, as those angled less steeply than the flow gradient are formed when the flow i s swelling, whilst those angled more steeply than the flow gradient are formed when the flow draws down. Iverson and Major (1987) present piezometric measurements which show that high movement rates on active lobes are associated with small-scale groundwater flow patterns within the earthflow. Such well defined mid-flow lobes are most common on earthflows with high 1/b r a t i o s (Table 3.1) such as Grinder (A), Red Mountain (B) and Black Dome (A) (Appendix A). At these s i t e s active lobes were characterised by standing water or phreatophyte vegetation, i n d i c a t i n g that they were found i n areas of groundwater d i s -charge. Similar, but less prominent lobes were recognised on earthflows with smaller 1/b ratios such as Red Mountain (A), Churn (A) and Big Bar, although at these s i t e s the flows d i d not occupy the entire flow track. 3.1.3 Toe zone Earthflows i n B r i t i s h Columbia terminate i n either a snout or a fan, unless they are eroded by a r i v e r or creek and thus have no depositional zone. A snout i s defined here as a steep-fronted, convex terminus, not spread l a t e r a l l y , while a fan comprises many snouts, spreading l a t e r a l l y and over-r i d i n g one another. Estimates of snout height, an i n d i r e c t i n d i c a t o r of flow thickness, range from 3m at Black Dome (A) to 30m for Red Mountain. Snouts are seen i n both active (Black Dome (A), Red Mountain (B), Yodel) and inac-t i v e flows (Yalakom (B)), (see Appendix A), where l i t t l e material has moved 47 down the flow track. They vary i n form from bulbous, l a t e r a l l y thickened snouts (Yalakom (C), Drynoch (B), see Appendix A), through to those which are semicircular i n plan and scalloped (Yalakom (A), see Appendix A). Many earthflows do not terminate i n a snout or a fan, but instead ma-t e r i a l i s eroded from the end of the flow track as i t debouches to a r i v e r or creek. A l l the flows adjacent to Fraser River have been eroded at the toe; i n addition the Heginbottom flow i s currently being eroded by Lone Cabin Creek. Flows Yalakom (A and D) feed into the Yalakom River, but erosion at these s i t e s has not been fast enough to prevent formation of a sub s t a n t i a l snout at Yalakom (A). The toe zones of earthflows have yielded few s i t e s of s t r a t i g r a p h i c value as they are mostly well-drained and therefore do not have lake and bog coring s i t e s . In addition, many active toe zones are too dry to support trees and hence do not provide useful s i t e s for dendrochronology. 3.2 Geology In Chapter 2 i t was shown that, within the study area, earthflows are concentrated i n three l i t h o s t r a t i g r a p h i c u n i t s : basic volcanics, sediments ( v o l c a n i c l a s t i c s and bentonite) and serpentinised p e r i d o t i t e . Within these units R number of subsidiary factors appear to control the precise location of earthflows. These factors are i d e n t i f i e d as f a u l t i n g , s t r a t i g r a p h i c con-f i g u r a t i o n , angle of dip, and regional groundwater flow. Each factor i s investigated i n the following sections 3.2.1 to 3.2.6. 3.2.1 Lithology Appendix H (summarised i n Appendix J, Table J2) tabulates the l i t h o l o g y of earthflows i n the study area. It can be seen that 56% of the earthflows occur i n sediments, or i n sediments intermixed with volcanics; 20% i n serpentinised p e r i d o t i t e , and 18% i n basalt. (The remaining 6% are i n other l i t h o l o g i e s or of unknown provenance). A d e t a i l e d d e s c r i p t i o n of the geology 48 of several flows included i n each l i t h o s t r a t i g r a p h i c category i s given i n Appendix I. Eight earthflows i n the study area are developed e n t i r e l y i n sediments. Those which were studied i n d e t a i l (Drynoch, Heginbottom, Appendix I) show a high proportion of volcanic c l a s t s i n the coarser sediments. Both flows showed a var i a b l e texture, ranging from conglomerate through to c l a y - s i z e material. Studies of other earthflows i n south-central B r i t i s h Columbia (Mathews and Rouse 1984, Nadler (1984), Rawlings 1984, Bovis 1985) found that fine-grained 'bentonite' beds within sedimentary bedrock were a major con-s t i t u e n t of earthflow material, although a l l flows were characterised by heterogeneous parent material. Bovis (1985) noted that at Gibbs Creek, Tunnel, Big Bar and Churn Creek, where t u f f s and sediments were admixed i n earthflow material, shear planes were concentrated i n clay r i c h layers derived from sediments. How-ever, f i e l d observations at Grinder Creek (A), where su b s t a n t i a l exposures of both past and present shear zones were observed, d i d not reveal any p r e f e r e n t i a l concentration of shear planes i n entrained sedimentary material rather than weathered t u f f s . In areas of volcanic rock, b a s a l t i c rocks are much more l i k e l y to de-velop earthflows than are those rocks with more a c i d i c l i t h o l o g i e s . A good example i s Red Mountain, i l l u s t r a t e d i n Appendix I, where basalt i s the parent rock providing most earthflow material. Rhyolite adjacent to the flow remains stable, forming steep c l i f f s bounding the earthflow. The concentration of earthflows i n basic v o l c a n i c rocks may be ex-plained by the differences i n weathering rates and products, which are seen i n basalt and r h y o l i t e . Birkeland (1974 p. 136) states that basalt weathers more rapidly to c l a y - s i z e d material than r h y o l i t e because the p r i n c i p a l constituent minerals of basalt ( o l i v i n e , hornblende/augite, plagioclase 49 feldspars) are less stable under subaerial weathering conditions than those of r h y o l i t e (quartz, orthoclase feld s p a r ) . Serpentinised p e r i d o t i t e forms a s u i t e of earthflows i n the Yalakom Valley, while adjacent volcanics do not p a r t i c i p a t e i n mass movement (see the d e t a i l e d d e s c r i p t i o n i n Appendix I ) . Serpentinised p e r i d o t i t e probably forms earthflows because the myriad small, curved s l i p planes formed during serpentinisation are r e a d i l y adopted as earthflow boundary shear zones (the author i s unaware of any engineering data to confirm that t h i s type of ma-t e r i a l has a low # value). Another p o s s i b i l i t y i s that serpentine weathers to c l a y - s i z e d montmorillonite due to leaching by neutral or a c i d i c groundwater (Wildman et a l . 1968). 3.2.2 Dip There are many cases where rockslides are c o n t r o l l e d by the dip angle of j o i n t s and other planar structures. However, the evidence for dip angle i n -fluencing earthflows either i n the study area or on other earthflows i s l i m i t e d due to the s c a r c i t y of dip measurements from adjacent s t r a t a . Rawlings (1984) shows that the Hat Creek s l i d e occurred i n a syncline, where at least the upper part of the s l i d e i s formed from beds dipping down the earthflow axis, as discussed by Bovis (1985). The two P a v i l i o n flows, both moving approximately northward, are found i n sediments with two dip meas-urements mapped by Monger and McMillan (1984) at 5° NNE and 18°E. These data are limited and inconclusive but ind i c a t e that flow may take place either down or at right angles to the dip (Bovis 1985). The most compelling evidence for movement affe c t e d by dip angle comes from the Fraser Fault zone between the Churn Creek and Big Bar earthflows. Mathews and Rouse (1984) note that the sediments dip gently to the east near Churn Creek, moderately to steeply at French Bar Creek and steeply (65°-90°) at Big Bar f e r r y (Fig 3.6). A d d i t i o n a l dip measurements from Grinder Creek (A) are given on F i g 3.6. U n d i f f e r e n t i a t e d sediments mapped 50 by Tipper on the east bank have comparable dips ranging from 10° near Grinder Creek to 65° near Big Bar, both to the east. As i d e n t i c a l sediments occur on both sides of Fraser River (Fig 3.6), while earthflows occur only on the west bank, i t was concluded that i n t h i s area the dip of the Eocene s t r a t a towards the deeply i n c i s e d Fraser River i s instrumental i n generating earthflows. 3.2.3 Faults Faults traverse 24 of the 53 earthflows i n the study area ( i . e . cross them approximately normal to the earthflow axis) while f i v e of the 53 flows occur p a r a l l e l to a f a u l t l i n e . Thirteen appear to be uninfluenced by f a u l t i n g , and i n s u f f i c e n t information was ava i l a b l e to form an opinion on the remaining 11. F i g 3.7 shows several supplementary examples of earthflows i n the study area which coincide with major f a u l t l i n e s , to add to those i l l u s t r a t e d i n Appendix I. There i s no evidence for p o s t g l a c i a l movement along the Fraser and Yalakom faults which may have i n i t i a t e d or maintained earthflow movement, although there are two instances of possible recent f a u l t i n g i n the 92J map area (Mathews, personal communication, 1985, and Roddick and Hutchinson, 1973) on faults unconnected with the Fraser.and Yalakom f a u l t s . Therefore recent major f a u l t i n g , increasing slope gradient and thereby i n i t i a t i n g or r e a c t i -vating earthflows, can be discounted. 3.2.4 Groundwater flow Many earthflows are located i n areas where groundwater levels are r e l -a t i v e l y high ( i . e . discharge zones). Two geological configurations are proposed which give hydrogeologic conditions favouring earthflow formation: layered rocks of d i f f e r i n g permeabilities and impermeable landslide material overlying permeable bedrock. The f i r s t configuration occurs where a well j o i n t e d (hence high conductivity) basalt o v e r l i e s Eocene volcanics and sediments with fewer 3200 Elevation ofbasalt unit in feet Fault trace Fi g 3.6 Earthflow location and dip of Eocene sediments, after Mathews and Rouse (1984) 52 Gibbs Creek Churn Creek A Fountain \ N Hat Creek \ \ P a v i l i o n A Kilometres Fig 3.? Location of fault traces with respect to selected earthflows 53 fractures (and hence lower hydraulic conductivity) as for example at Red Mountain flows (B) to (E) and (G). Two mechanisms combine to make t h i s geologic s e t t i n g p e c u l i a r l y susceptible to earthflow. F i r s t , i t i s possible that the basalt flows are anisotropic with higher conductivity p a r a l l e l with the flows than v e r t i c a l l y along j o i n t s . This would give r i s e to high porewater pressures as shown i n F i g 3.8a (adapted from a model given by Hodge and Freeze, 1977). Secondly, the presence of an impermeable layer beneath the capping basalts may also generate high piezometric pressures as i n F i g 3.8b (also adapted from a model given by Hodge and Freeze, 1977). The second model applies where material of low hydraulic conductivity, for example bentonite clay, o v e r l i e s well fractured rock, at a mid or low slope p o s i t i o n where the well fractured rock would n a t u r a l l y discharge. The P a v i l i o n earthflow provides a good example, as the hydraulic conductivity of the fractured bedrock of Mount Cole and the Cretaceous sediments com--3 p r i s i n g the earthflow, have been ca l c u l a t e d and estimated at 2.3x10 and _ g approximately 10 m/s r e s p e c t i v e l y (Nadler 1984). F i g 1.8 shows a hy-p o t h e t i c a l earthflow net which could account for artesian pore pressures such as those measured near P a v i l i o n earthflow (Bovis 1985). Faulting may create a analogous s i t u a t i o n where mechanically weak zones of shattered rock, with r e l a t i v e l y high hydraulic conductivity, underlie earthflows. 3.2.5 Geomorphic h i s t o r y Several geologic factors have been discussed which influence earthflow movement (dip angle, f a u l t i n g , l i t h o l o g y ) i n the study area. However, i n many cases conditions i d e n t i f i e d i n the study area extend over large areas of the I n t e r i o r . For example, T e r t i a r y v o l c a n i c l a s t i c s and bentonite extend northwards to Quesnel (Rouse and Mathews, 1979), adjacent to Fraser River. No earthflows were noted during an a i r photograph search of t h i s area (see section 2.1). E f f e c t of anisotropy which may be present at Red Mountain as a r e s u l t of layered basalt flows High piezometric pressure generated by gently dipping permeable rocks overlying less permeable s t r a t a . This s i t u a t i o n may occur at Red Mountain as a r e s u l t of basalt flows overlying Cretaceous sediments K=hydraulic conductivity. Lines are equipotentials| 3.8\. V Hypothetical groundwater flow system at Red Mountain leading to f a i l u r e on the dip slope (flows B,C,D,E and G ) , from Hodge and Freeze (1977) 55 It is proposed that the southward d i v e r s i o n of Fraser River, and the associated downcutting of i t s t r i b u t a r i e s during the l a s t m i l l i o n years, may account for the r e s t r i c t e d occurrence of earthflows within the i n t e r i o r of B r i t i s h Columbia. The concentration of earthflows adjacent to Fraser River south of the capture point, and on the side slopes of t r i b u t a r y v a l l e y s such as the Yalakom and Thompson suggests that stream i n c i s i o n over the l a s t m i l l i o n years may have influenced earthflow l o c a t i o n by providing steep sided valleys which dissected hitherto undisturbed rocks. Many other mass movement features, p a r t i c u l a r l y large tension cracks, are i n d i c a t i v e of regional h i l l s lope i n s t a b i l i t y . 3.2.6 G l a c i a l h i s t o r y Much of the study area f e l l i n the 'ice divide' region during the l a s t g l a c i a t i o n , where ic e flows from the Coast Mountains diverged northward and southward. This slow moving ice performed l i t t l e erosion. Therefore, i n the Interior, thick deposits of clay, t u f f s and basalt, deeply weathered during the warm T e r t i a r y period (Mathews and Rouse 1975), which would have been removed by g l a c i a l erosion i n the Coast Mountains, remained undisturbed. Hence the impotence of g l a c i a l erosion may account for the great thicknesses of deeply weathered and unconsolidated Cretaceous and T e r t i a r y material found in the study area. 3.2.7 Conelus ions This examination of the geologic conditions governing earthflow l o -cation in B r i t i s h Columbia, together with a l i t e r a t u r e review from other areas (Appendix J ) , has revealed a number of factors common to earthflows both within and outside the study area:-1) In south-central B r i t i s h Columbia, earthflows are developed mainly i n serpentine, basalt, and v o l c a n i c l a s t i c sediments (often admixed with t u f f s ) , whereas flows i n other areas occur predominantly i n sedimentary rocks. 56 2) Several earthflows i n B r i t i s h Columbia are probably influenced by dip angle, but examples from other areas are scarce or absent. 3) Many large flows, both i n B r i t i s h Columbia and elsewhere have well j o i n t e d , permeable rocks such as basalt and sandstone overlying v o l c a n i c l a s t i c s or mixed t u f f s and v o l c a n i c l a s t i c s . These permeable rocks provide recharge areas which discharge i n the earthflow basin-4) Earthflows do not usually occur i n d i v i d u a l l y , but i n c l u s t e r s within areas showing s i m i l a r l i t h o l o g y and structure. 5) Slope i n s t a b i l i t y over a large area may be indicated by other mass move-ment features or by i n c i p i e n t mass movement adjacent to the earthflows. 6) The factors a f f e c t i n g earthflow l o c a t i o n are size-dependent, with g l a c i a l h i s t o r y a f f e c t i n g the l o c a t i o n of small flows (none i n the study area), and angle of dip and f a u l t i n g a f f e c t i n g only the larger features. 7) In B r i t i s h Columbia slow-moving i c e during the past g l a c i a t i o n ensured that deeply weathered rocks were not s i g n i f i c a n t l y eroded, leaving a large volume of residual material i n s i t u . This weathered debris has subsequently formed earthflows. 8) In the study area and also i n the Columbia River Gorge, the Eel River area, the Van Duzen area and the Cascades, earthflows take place i n a g e o l o g i c a l l y 'young' landscape with s t r a t a d e s t a b i l i s e d by i n c i s e d r i v e r v a l l e y s . 3.3 Geotechnical properties of earthflow materials  3.3.1 Introduction The purpose of t h i s section i s to examine the range of engineering ma-t e r i a l s which form earthflows within the study area. Samples from ten earthflows, covering the three l i t h o s t r a t i g r a p h i c units described i n section 3.2, were characterised by analyses of grain s i z e d i s t r i b u t i o n , Atterberg l i m i t s and clay mineral species. A d e s c r i p t i o n of earthflow material of flows found outside the study area i s given i n Appendix L. It appears that g e o l o g i c a l l y d i f f e r e n t earthflow materials i n the study area each y i e l d c h a r a c t e r i s t i c earthflow gradients, thereby providing an i n d i r e c t measure of shearing resistance. Both VanDine (1980) and Rawlings (1984) have performed successful back c a l c u l a t i o n s from measured angles of shearing resistance for flows i n the study area. A comparison of earthflow s i z e , material and gradient suggests that earthflow gradient i s not only material dependent but may also (at least within the study area) be scale dependent. 3.3.2 Grain-size d i s t r i b u t i o n Forty-four samples taken from seven earthflows i n the study area, with contrasting geological parent material, were selected for grain s i z e analy-s i s . Sampling s i t e s were chosen judgementally at locations such as open tension cracks, active areas bare of vegetation, and stream cut sections, i n order to avoid s i t e s where pedogenic a l t e r a t i o n might a f f e c t the geotechnical data. Detailed sample des c r i p t i o n s , laboratory procedures and grain s i z e d i s t r i b u t i o n percentages are given i n Appendix K. Appendix K and F i g 3.9a show that the four rock types (basalt, other volcanics, sediments and serpentinised p e r i d o t i t e ) have d i f f e r e n t charac-t e r i s t i c grain s i z e d i s t r i b u t i o n s , although considerable overlap occurs with a l l but the serpentinised p e r i d o t i t e . These l a t t e r samples form a consider-ably coarser c l u s t e r than those from the other three groups, with most sam-ples having over 80% sand-sized p a r t i c l e s and less than 5% clay. The basalt samples are much more dispersed, but o v e r a l l show 35% or more clay and a correspondingly smaller proportion of sand. The samples derived from sedimentary rock, o r i g i n a t i n g from four d i f f e r e n t s i t e s , f a l l mainly between the other two groups, being coarser than the.serpentinised p e r i d o t i t e but having less clay than the basalt samples. The four samples derived from a c i d i c volcanic rock overlap with the sedimentary set, but are considerably lower i n clay content than t h e i r b a s a l t i c counterparts. 58 Bovis (1985) presents an exhaustive analysis of grain-size d i s t r i b u t i o n from seven other flows i n the study area, including data from VanDine's (1980) work at Drynoch. His work i s reproduced here as Figs 3.9b and 3.9c and a comparison with F i g 3.9a shows that the majority of samples c l a s s i f i e d as volcanics and sedimentary are broadly coincident with those from t h i s study (two of the regions i n F i g 3.9a are superimposed on Figs 3.9b and 3.9c to f a c i l i t a t e comparison). The Spences Bridge Group volcanics are extremely variable, but are mainly andesitic or more a c i d i c rocks; hence F i g 3.9c suggests that debris from a c i d i c volcanics i s generally coarser than that derived from either the more basic volcanic rocks or sediments. F i g 3.9c also shows more samples with a high clay content and low sand content than were obtained from the study area. It appears that, within the study area, three groupings of parent ma-t e r i a l can be recognised on the basis of grain s i z e d i s t r i b u t i o n (basalt, serpentinised p e r i d o t i t e , sediments and other v o l c a n i c s ) . The largest anom-aly i n the data from the study area i s the presence of coarse serpentinised p e r i d o t i t e , which i s completely d i f f e r e n t from any other earthflow material described i n the l i t e r a t u r e . The other grain s i z e groups are comparable with Bovis' data for both sediments, medium to a c i d i c volcanics and mixed sediments and volcanics. O v e r a l l , these data show that a l l earthflow mate-r i a l s have a substantial proportion of sand-sized and larger c l a s t s , which must provide considerable shearing resistance to flow. A l l material types except serpentinised p e r i d o t i t e also have a s i g n i f i c a n t (>10%) clay per-centage. The next two sections investigate the nature and influence of the clay f r a c t i o n . Fig 3.9b Grain size distribution and geologic provenance of earthflow materials in the study area (data from other studies, chief ly Bovis (1985)) ON o 61 3.3.3 Atterberg l i m i t s Atterberg l i m i t s were determined for a t o t a l of 75 samples from ten flows i n the study area. Samples were drawn from the same flows as for grain s i z e analyses (Black Dome (A), Canoe Creek, Grinder(A), Heginbottom, Lone Cabin, Red Mountain (A) and Yalakom (A and C)) and also from Big Bar and Churn Creek where Bovis (1985) had already obtained g r a i n - s i z e data. The samples were processed according to techniques described by Lambe (1951) to give p l a s t i c and l i q u i d l i m i t s f o r each sample. Atterberg s t a -t i s t i c s ( l i q u i d l i m i t (LL), p l a s t i c l i m i t (PL) and p l a s t i c i t y index (PI)) for a l l the samples are given i n Appendix C together with an e i g h t - f o l d c l a s s i f i c a t i o n of material types into f l u v i o g l a c i a l r i v e r terrace materials (Te), t u f f s ( T ) , basalt (B), other volcanics (V), serpentinised p e r i d o t i t e (SP) and three grades of sediments: conglomerate (SC), sandstone/siltstone (SS) and claystone (SB). A t o t a l of 12 samples were non-plastic including a l l those tested from the Yalakom flows. The remaining data from Appendix K were plotted i n F i g 3.10 together with the 'A-line' used for the U.S. U n i f i e d S o i l C l a s s i f i c a t i o n System (Craig 1978), which f a c i l i t a t e d a v i s u a l compar-ison with other data. F i g 3.10 displays several m a t e r i a l - s p e c i f i c groupings. A l l the f l u v i o g l a c i a l Fraser terrace material samples have r e l a t i v e l y low p l a s t i c i t y (highest value for PI 28), while the majority of t u f f s and basalt have PI values exceeding t h i s upper bound. Sediments cover a wide range of PI values with a s i g n i f i c a n t proportion (mainly sand and s i l t s t o n e s ) f a l l i n g into the lower region occupied c h i e f l y by f l u v i o g l a c i a l materials. Other sedimentary samples f a l l i n the higher PI range with t u f f s and basalt; these are either conglomerate or claystone. A s i n g l e conglomerate sample from Black Dome forms an o u t l i e r of exceptionally high PI and LL (89 and 118 r e s p e c t i v e l y ) . A comparison of data derived from t h i s study with those from other earthflows i n the study area (Bovis (1985)) reveals that the two data sets Jfi. B i g Bar B Black Dome L i q u i d L i m i t F i g 3.10 P l a s t i c i t y index, l i q u i d l i m i t and geologic provenance of earthflow materials from the study area (data from t h i s study) 63 show s i m i l a r d i s t r i b u t i o n s . Bovis maps t i l l with low values; t h i s study finds f l u v i o g l a c i a l sediments i n a s i m i l a r p o s i t i o n . Sediments from Drynoch, P a v i l i o n and G i l l o n Creek occupy a wide range of values as, i n t h i s study, do sediments from seven of the eight flow samples. Some of the sediments ( c h i e f l y from Grinder Creek), have markedly lower p l a s t i c i t y than any samples processed by Bovis. Bentonite-rich seams give Bovis three high PI and LL values which correspond well with the s i n g l e o u t l y i n g point i n Table K l (from Black Dome) i n t h i s study. Although the o u t l i e r i s derived from a conglom-erate, the matrix was bentonite r i c h and the rock had a smaller proportion of large c l a s t s than the other conglomerates sampled. The analysis shows that i n the study area flows occur mainly i n p l a s t i c materials, although those developed i n serpentinised p e r i d o t i t e constitute a notable exception. Many flows include a high proportion of t i l l and/or f l u v i o g l a c i a l sediments which have a uniformly low P I . O v e r a l l , those flows derived from basalt or t u f f s have higher Atterberg l i m i t s than many sedimentary samples, but the l a t t e r class show much v a r i a b i l i t y , and provide a few outlying, extremely high values. 3 . 3 . 4 Clay minerals and cations Eighteen samples from seven earthflows were processed for XRD (X-ray d i f f r a c t i o n ) . These included samples o f serpentinised p e r i d o t i t e from Yalakom (C); samples of claystone from the Canoe Creek and Grinder Creek flows; sandstone and conglomerate from Black Dome, Grinder and Heginbottom; and a vari e t y o f basalts from the Red Mountain (A) and Lone Cabin Creek areas. Details o f sample descriptions, preparation and clay mineral recog-n i t i o n , together with d e t a i l e d r e s u l t s , are given i n Appendix K. Appendix K l i s t s the main c l a y - s i z e d mineral species for each sample, together with r e l a t i v e abundance determined q u a l i t a t i v e l y from peak area. Thirteen of the eighteen samples l i s t montmorillonite as the most abundant mineral (no attempt was made to d i f f e r e n t i a t e between d i f f e r e n t species of 64 smectite); these included a l l the basalt samples and many of the sediments. Some traces had moderate amounts of k a o l i n i t e , c h l o r i t e or mica, and many showed a trace of quartz. Two of the samples dominated by vermiculite o r i g -inated from a prominent red claystone bed which cropped out at both Canoe Creek and Grinder Creek (A). As the traces for these two samples are iden-t i c a l , they may have been abstracted from one unit which stretches beneath the Pliocene basalt capping which separates the two flows. Serpentinised p e r i d o t i t e produced samples of a contrasting composition; they were domi-nated by serpentinite minerals with a u x i l i a r y c h l o r i t e (swelling i n one sample) and l e p i d o c r o c i t e . The predominance of swelling minerals at a l l flows except those i n serpentinised p e r i d o t i t e , suggests that earthflow material properties may be affected by exchangeable cations; i n montmorillonite a Na-saturated ex-change complex may cause very considerable expansion. However, i t appears 2+ that i n the study area Ca i s the dominant cation at a l l s i t e s except those in the Yalakom Valley. This i s reasonable, as most montmorillonites have 2+ Ca as t h e i r major exchangeable cation (Grim 1960) . The Yalakom samples are remarkable both for t h e i r unusually low t o t a l of exchangeable cations and 2+ 2+ also for the predominance of Mg . In most s o i l s the amount of Ca i s four 2+ or f i v e times that of Mg , while i n the Yalakom specimens the ions are re-versed i n abundance. Clay mineral analyses have been performed for a number of other earthflows i n the study area by VanDine (1980), Rawlings (1984) and Bovis (1985) as shown i n Appendix K. They a l l show that montmorillonite i s the predominant clay mineral. Results from other areas show that montmorillite is the most abundant mineral i n about h a l f of the flows analysed, including a l l the medium to large features. I l l i t e and k a o l i n i t e are the major minerals in a number of smaller flows and one feature, the Melendy Ranch flow i n the San Francisco Bay area (Keefer and Johnson, 1983) comprises serpentine and 65 t a l c minerals, i n d i c a t i n g a mineral composition s i m i l a r to that of the Yalakom flows. In summary, the data presented i n Appendix K show that within the study area earthflows occur either i n areas where the rocks weather to give ex-panding clay minerals (montmorillonite or, occasionally, v e r m i c u l i t e ) , or i n areas of serpentinised p e r i d o t i t e y i e l d i n g a s e r p e n t i n i t i c mineral s u i t e . Of a l l the clay minerals, montmorillonite has been shown to have the lowest angle of shearing resistance (Kenney 1967) and i t i s probably t h i s feature which explains why large earthflows are concentrated i n montmorillonite r i c h source rocks. The presence of large flows i n serpentinite minerals i s anomalous in that these have no swelling properties. It i s possible for serpentine rocks to weather to montmorillonite (Wildman et a l . 1968), but no evidence was seen for t h i s weathering sequence either d i r e c t l y through XRD or i n d i r e c t l y through p l a s t i c specimens i n Atterberg l i m i t t e s t s . 3.3.5 Gradient and shear strength No shear tests were performed i n t h i s study. However, the d i f f e r i n g clay mineral, Atterberg l i m i t s and grain s i z e c h a r a c t e r i s t i c s of samples from a var i e t y of flows, indicated that the (residual) angle of shear (0^) would probably d i f f e r considerably between flows formed from d i f f e r e n t materials. This contention was based upon c h a r a c t e r i s t i c angles of shearing resistance for d i f f e r e n t materials, both clay minerals and rock samples, such as those tabulated by Deere and Patton (1971). Several pertinent figures are given i n Appendix L (Table L2), and Appendix L (Table LI) l i s t s a few add i t i o n a l values, mainly from overconsolidated Eocene clays i n B r i t a i n . It can be seen that montmorillonite, the major clay mineral component of the majority of earthflow materials in the study area, has the lowest angle of shearing re-sistance (4-11°), while quartz and crushed sand (sand sized c l a s t s i n the study area are probably less r e s i s t a n t ) has a <j>^ value of 30-35°. As the e f f e c t i v e angle of shearing resistance (0 ) i s affected by both texture and 66 mineralogy, i t was thought that flows i n d i f f e r e n t materials might display d i f f e r e n t ranges of gradients, i n response to differences i n <p^. Gradients and areas of a l l the earthflows i n the study area were de-termined either from previous estimates (VanDine 1980, Rawlings 1984, Bovis 1985) or from a i r photographs (Appendix A) i n conjunction with the relevant NTS 1:50,000 maps. Appendix H l i s t s both s t a t i s t i c s , gradients ranging from 5.3° (Hat Creek) to 34° (Burkholder (B)), and areas from 0.002km2 (Yodel B) 2 to 6.5km (Camoo). These data were p l o t t e d on F i g 3.11 to see whether any consistent pattern was seen between a l l the flows i n the study area. Figs 3.11a to e show some r e l a t i o n s h i p between earthflow s i z e and slope 2 angle within areas of s i m i l a r l i t h o l o g y . Large earthflows (>1 km ) a t t a i n a maximum of 13° while the majority are less then 10°. Conversely, small 4 2 o o (<10 m ) features are a l l steeper than 10 , with several over 20 . Medium sized features show a much greater spread of values than large flows, whilst presenting an average value between those of the small and large flows. The r e l a t i o n s h i p of increasing flow angle with decreasing flow si z e i s most c l e a r l y seen within flow clusters of homogeneous material, such as the coarser sediments and basalt flows (Fig 3.11a to c ) . Figs 3. l i d and e did not show any clear r e l a t i o n s h i p . Four hypotheses were put forward to explain t h i s empirical r e l a t i o n -ship. F i r s t , the magnitude of boundary and basal shearing resistance r e l a t i v e to flow volume i s greatest i n small flows. Secondly, the r e l a t i o n s h i p between depth and earthflow s i z e demonstrated i n Chapter 1 (Fig 1.10) suggests that, 4 2 for flows less than 10 m i n a r e a , depth to shear plane i s i n s e n s i t i v e to changes i n flow volume, suggesting that the e f f e c t i v e downward component of movement i n smaller flows i s produced by increasing flow gradient rather than 4 2 by decreasing boundary shear, while for flows above 10 m , increases i n depth (and hence volume) decrease the r e l a t i v e importance of marginal shear. 67 27-w 25-CD CD i _ 23-O) CD •D 21-<: 19-o 17-O Q) 15-D) 13-C CO CD 11-O) co 9-CD > < 7-• GE •GD • GB • BOA BDA Black Dome (A) DA Drynoch (A) GA.GB etc Grinder Creek (A) and (B) etc He Heginbottom O sediments GA a • DA 103 10 4 1"05 Earthflow area Clog 10 1 0 6 axis] 10' 27 » 25 CD CD 23 CD -o 21-19-o 17-o CD 15-cn c 13" CO CD 11-o> CO L— 9-CD > < 7- s a RC F Flapjack LC Lone Cabin RA.RB etc Red Mountain flows • basalt • F • L C • HA 1'0'3 1 0 4 1 0 5 10 e Earthflow area C l o g 1 Q axis) 10' F i g 3.11 Relationship between earthflow area and flow angle for selected flows within the study area 68 35 334 31 5 29H D) 27 251 o 231 o 21 ® 19 o> « 17 6 15 ca 0 13 1 1 H 9 7 5 10c 29 271 g 251 g>23-» •o - 21 ^ l a -' s 17-<D o> 151 c 1 0 13 a> co 1 1 S> 91 < 7-1 10' OBuB BuA.BuB Burkholder flows Cm Camoo Creek LaA.LaB.LaC Lac la Mer flows YA,YB,YC,YD Yalakom flows O Serpentinised peridotite OLaB ° V C o L a C OYA OCm OLaA 10' 1 0 5 1 0 6 Earthflow area Clog 1 0 axis} 10' Sites ChA.ChB etc Churn Creek flows Gi G i l l o n Creek Ha Hat Creek PA.PB Pav i l i o n • Fine sediments • ChF • YoA , G i • ChE • ChC C « • PA • ChB »ChA 10' 10 : 10* 10' Earthflow area (log 1 0 axis} F i g 3.11 Relationship between earthflow area and flow angle for selected earthflows outside the study area • G BB Big Bar BDB.BDC Black Dome (B) and (C) Ca Canyon Fo Fountain Gb Gibbs Creek GC Grinder Creek (C) T Tunnel YoB Yodel (B) • BOB A Volcanics other than basalt • Unknown A YoB A B D C T rGb A C a A K 10> 10' 10 c 10< Earthflow area ( l o g 1 0 , axis] 10' F i g 3.11 Relationship between earthflow area and flow angle for selected flows within the study area Table 3.2 Angles of shearing resistance from earthflow material i n the study area Site * r (Back-calculated) S lope Reference Big Bar 7° 12° Bovis (1985) Churn 4° 7° Bovis (1985) Drynoch 15° 17° 9° VanDine (1980) G i l l o n Creek 11° 14.5° Bovis (1985) Hat Creek 8° 6.8° 5.3° Rawlings (1984) Pavilion 16-26° 9° Bovis (1985) T h i r d l y , large flows have the p o t e n t i a l to disrupt regional groundwater systems by o b l i t e r a t i n g established drainage networks over a r e l a t i v e l y large area. In smaller flows creeks are frequently diverted to the flow margins, thereby maintaining much of the drainage network on and around the flows (see the Grinder Creek a i r photographs for r a p i d l y evolving drainage networks i n Appendix A (p. 312-314). However i n the larger flows, diversion of pre- e x i s t i n g drainage around the flow cannot occur, and the re s u l t i n g interrupted groundwater flow system forms many extensive lakes and bogs on larger flows (Red Mountain A, P a v i l i o n (A), Hat Creek and Churn (A)). This suggests that on the larger flows, piezometric levels tend to be higher than on the smaller, better drained features. This may r e s u l t i n earthflows residing at a lower angle on these poorly drained large flows. The fourth p o s s i b i l i t y i s that, as many earthflows occupy t r i b u t a r y basins, Schumm's r e l i e f r a t i o e f f e c t may be important. This e f f e c t states that larger r i v e r basins have lesser gradients than t h e i r t r i b u t a r i e s . Therefore, smaller earthflows occupying smaller r i v e r v a l l e y s are assumed to r e f l e c t these steeper v a l l e y gradients. F i g 3.11 also demonstrates some differences i n slope angle related to flow material. The serpentinised p e r i d o t i t e flows are mostly steeper than those developed i n other materials, while those formed from basalt have the lowest angles. Sedimentary materials show considerable spread, but sand/siltstone flows generally l i e between basalt and serpentinised p e r i d o t i t e . These findings are reasonable, as the re s i d u a l angle of shearing resistance for montmorillonite clays i s 4-11° (according to section 4.2.3 the ultimate product of basalt weathering i n the study area). Measured an-gles of shearing resistance (Appendix K, Table K2) from sediments i n the study area span 4-26°, r e f l e c t i n g the v a r i e t y of sedimentary rocks p a r t i c -ipating i n earthflow. A comparison of Figs 3.11 and Appendix H reveals that i n many cases earthflow angle may approach 0^ . However, some of the larger flows have av-erage slopes lower than the measured tf^CPavilion, Hat Creek, Drynoch) im-p l y i n g that a high groundwater table and hence reduced e f f e c t i v e stress i s required for movement to take place. Back c a l c u l a t i o n s from f i e l d measure-ments of depth to shear plane and water table l e v e l s obtained from piezometers and boreholes at Drynoch and Hat Creek have given t h e o r e t i c a l 0^  values i n reasonable agreement with laboratory t e s t s . This, together with the observation of over 30m of excess hydraulic head close to the earthflow fan at P a v i l i o n (Bovis 1985), implies that groundwater levels have a c r u c i a l r o l e i n determining the l e v e l of earthflow a c t i v i t y . 3.3.6 Conelus ion 1. Earthflow materials i n the study area f a l l into f i v e major catego-r i e s with d i f f e r e n t mechanical and chemical properties. F i r s t are earthflows developed in serpentinised p e r i d o t i t e which have a sandy texture and are non-plastic. Second are earthflows formed from basalts where clay minerals are predominantly montmorillonites, and clay-textured p a r t i c l e s comprise >35% of flow material. Third and fourth re s p e c t i v e l y are flows developed from f i n e (clay textured) and coarser ( s i l t s t o n e to conglomerate) textured sedimentary parent material. The f i n a l category includes a range of volcanics, excepting basalt, which y i e l d flow material composed predomi-nantly of montmorillonite. These have a higher shearing resistance than the basalt samples due to a higher proportion of sand-sized c l a s t s . 2. F i g . 3.11 and Appendix L show that large and medium-sized earthflows in B r i t i s h Columbia and elsewhere i n the world have a size/gradient re-la t i o n s h i p such that larger flows are observed to move along gentler slopes than smaller flows. This r e l a t i o n s h i p does not hold for flows smaller than 4 2 approximately 10 m . Within areas of homogeneous rock, an approximate semi-logarithmic r e l a t i o n s h i p between earthflow area and flow angle was seen. 3. Reasonable estimates of re s i d u a l shear strength for rocks, residual s o i l s and clays of the same genre as those found i n earthflows show that, provided groundwater levels are close to the surface on some of the larger flows, movement on the observed slope can be explained i n terms of the shear strength of the earthflow material. 74 CHAPTER 4 MOVEMENT OF EARTHFLOWS IN THE STUDY AREA  IN HISTORIC TIME 4.1 Introduction The major objective of t h i s t h e s i s , f i r s t introduced i n Chapter 1, i s to examine the timing of earthflow movement i n the study area and to r e l a t e movement phases to c l i m a t i c f l u c t u a t i o n s . Three timescales are examined i n t h i s and the following chapters: h i s t o r i c time, the past few centuries and p o s t g l a c i a l time. This chapter examines the f i r s t of these, that of h i s t o r i c time. A movement chronology (section 4.2) was obtained from the a i r photograph record and the Grinder Creek record i s supplemented i n section 4.3 by ob-servations of movement measured by stake arrays. A corresponding p r e c i p i -t a t i o n chronology (section 4.4) was obtained from AES (Atmospheric Environment Service) records from several weather stations covering d i f f e r -ent time periods i n and around the study area (Table 2.1). In the f i n a l section of th i s chapter (section 4.5) the movement and p r e c i p i t a t i o n chronologies are compared i n order to investigate the influence of c l i m a t i c fluctuations on earthflow movement. 4.2 Earthflow movement as recorded i n a i r photographs and recent f i e l d ob- servations 4.2.1 The nature of the evidence A e r i a l photographs provide a useful record of h i s t o r i c change where s i g n i f i c a n t changes i n morphology have occurred between successive photo pairs (Brunsden and Jones 1976, Evans 1982). The a i r photograph record for the study area dates i n part from 1928, when a Federal Survey provided data for several earthflows adjacent to Fraser and Thompson r i v e r s . One or two sets of photographs are a v a i l a b l e f or most s i t e s between 1947 and 1950. Many s i t e s were photographed again i n May 1961, but unfortunately high elevation s i t e s s t i l l had su b s t a n t i a l snow cover i n t h i s s e r i e s . A l l s i t e s were pho-75 tographed again i n the periods 1965-1967 and 1974-1980, many several times. No photographs are ava i l a b l e for any s i t e s a f t e r 1980; for some of the Red Mountain s i t e s the l a t e s t photosets are those taken i n 1974. The paucity of recent photographic evidence necessitated on-site observations to sup-plement the recent chronology, although i t was not po s s i b l e to duplicate the comprehensive survey provided'by a e r i a l photographs. I t must be noted that the photographic record i s heavily biased towards areas with sparse vege-t a t i o n as r e l a t i v e l y minor changes could be observed at these s i t e s , whereas thick coniferous forest precluded observation of a l l but major changes at many other s i t e s . This section w i l l proceed by describing changes seen i n the a i r pho-tograph survey on a s i t e - b y - s i t e basis, supplemented, where relevant, by ground observations. A complete l i s t of a i r photographs for a l l s i t e s i s given i n Appendix A. 4.2.2 Black Dome earthflows The Black Dome (A) earthflow was i n a c t i v e i n a i r photographs taken i n 1948 and 1950. By 1961 the earthflow had reactivated, and between 1961 and 1967 Table 4.1 shows that maximum movement rates of 17m per year were seen. Movement continued up to the l a s t a i r photograph stereopair, taken i n 1979, al b e i t at decreasing annual rates (Table 4.1). A d e t a i l e d d e s c r i p t i o n of movement at t h i s s i t e , together with relevant maps, i s found i n Appendix M. 4.2.3 Churn Creek (A) earthflow A small portion of the north side of the large Churn Creek earthflow has been active during the l a s t 35 years, although most of the flow has re-mained in a c t i v e (Appendix A pp 299). This small part of the larger earthflow i s described i n the following chronosequence. Stereopairs taken i n 1928 and 1948 (Appendix A pp 310-311) show that the s i t e was i n a c t i v e , with the earthflow drained by deeply i n c i s e d creeks. Between 1948 and 1950 a major r e a c t i v a t i o n was seen (Appendix A p. 310), with 76 Table 4.1 Rates of movement for selected earthflows in South Central B r i t i s h Columbia Total movement Per year (m) Black Dome (A) 1947-1950 No movement n/a 1950-1961 In i t i a t i o n n/a 1961-1967 100m 17 1967-1974 25m 3.5 1974-1979 15m 3 Grinder (A) 1928-1950 None n/a 1950-1961 15m-reactivation n/a 1961-1967 40m 7 1967-1975 40m 5 1975-1980 15m 3 Grinder (B) 1928-1961 None n/a 1961-1967 Reactivation n/a 1967-1974 50m 7 1974-1980 5m 1 Drynoch 1871-1909 3 (VanDine 1983) 1928-1947 2.5 1951-1972 3 Yodel Before 1950 Inactive n/a 1950-1961 Reactivation n/a 1961-1965 35m 9 1965-1967 2-3m l-2m 1967-1974 2-3m 0.3-0.5m 1974-1980 2.3m 0.3-0.5m the creek course o b l i t e r a t e d and a new l a t e r a l deposit b u i l t on the north side of the flow. By 1961 (Appendix A p. 310) the flow had s t a b i l i s e d again and was once more dissected by a creek draining the main Churn Creek flow. Between 1961 and 1975 (Appendix A p. 311) further movement had taken place as the creek course was diverted to the north side of the earthflow track. When the flow was v i s i t e d i n 1983 i t appeared to be i n a c t i v e as no tension cracks or active shear surfaces were seen. 4.2.4 Drynoch (A) earthflow H i s t o r i c movements of the Drynoch earthflow are described by i n d e t a i l by VanDine (1983). The following discussion i s la r g e l y a summary of h i s ar-ch i v a l work, supplemented by a few ad d i t i o n a l observations. The Cariboo wagon road was b u i l t across the toe of Drynoch (A) i n 1862. Although no mention of s l i d e a c t i v i t y was made at the time of construction, Begbie (1871) noted i n s t a b i l i t y i n the v i c i n i t y of the s l i d e . The years 1877-1885 saw considerable s l i d e a c t i v i t y ; ' i n the course of a year i t (Drynoch landslide) has been known to s l i d e several feet' (anon. 1885 quoted by VanDine). In 1878 movement of 2.5 metres per year was measured during surveys for the Canadian P a c i f i c Railway, while 3 metres per year was quoted i n 1909. During the 1920's, 30's and 1940's some d i f f i c u l t y i n maintaining the highway was noted. However, no further d i r e c t measurements of movement rates were available u n t i l the Department of Highways established 'movement hubs', monitored for three years between 1957 and 1960. These registered movement averaging 3m/year (VanDine 1980). Subsequent drainage works and substantial excavations of s l i d e material have slowed down s l i d e movement at the base of the flow since 1961, when the present Trans-Canada highway was b u i l t . The f i r s t a i r photographs were Federal, taken i n 1928 (Appendix A, p. 300). Both the Canadian P a c i f i c Railway and the (now) Trans-Canada highway are seen traversing the toe of the flow. The next stereopair dates from 1947; 78 movement of 50m i n t o t a l i s estimated for the intervening period from the movement of four trees r e a d i l y recognised i n both stereopairs. This estimate gives an average movement rate of 2.5m/year for the period 1928-1947. VanDine shows, from a i r photographs taken between 1951 and 1972, that the toe of the s l i d e and a portion i n the flow track zone, have been moving at about 3m/year. There are no photographs to indic a t e movement since 1972. 4.2.5 Grinder Creek earthflows (A) and (B) Grinder Creek earthflows (A) and (B) were i n a c t i v e when viewed on the e a r l i e s t (1928) a i r photographs, with creeks flowing the length of both earthflows and cutting l a t e r a l deposits i n several places. By 1950, the date of the next usable stereopair, movement had just s t a r t e d i n the upper part of flow (A). Flow (B) was in a c t i v e i n the 1950 stereopair. Between 1950 and 1961 flow (A) underwent considerable morphologic change, with well defined lobes moving down both arms of the flow. Flow (B) was s u b s t a n t i a l l y unal-tered, with the exception of small morphological changes near the head of the flow, i n d i c a t i n g i n c i p i e n t movement. The period 1961 to 1967 saw maximum estimated rates of movement for Grinder (A) and a major r e a c t i v a t i o n of flow (B) (see Table 4.1). Movement has continued to the present time, although both flows show decreasing movement rates from 1967 to 1980. Stake measurements from Grinder (A), taken for the two year period 1983-1985, show maximum current rates of movement of 7m/year. A deta i l e d account of morphologic changes of both flows i s given i n Appendix M. The summary s t a t i s t i c s i n Table 4.1 were produced on the basis of that analysis. 4.2.6 Additional s l i d e at Grinder Creek Tension cracks were observed to open up between 1967 and 1975 on the steep slope adjacent to Fraser River north of the flow (the s i t e marked 'P' i n F i g 12). Between 1975 and 1980 the cracks widened and i n 1984-1985 a s l i d e 79 occurred (personal communication S. Oswald, 1985). The s i t e was v i s i t e d i n the summer of 1985 and many active tension cracks were seen. The s l i d e , however, took place down a very steep slope and the f a l l e n material had been removed by Fraser River. Although t h i s s l i d e was not part of an earthflow, i t was s u f f i c i e n t l y close to flows (A) and (B) to be influenced by the same regional groundwater flow system. 4.2.7 Red Mountain (A) earthflow No changes are v i s i b l e i n the a i r photograph sequence at t h i s s i t e which spans the period 1947-1977. However, two s i g n i f i c a n t changes have taken place since 1977. F i r s t a marginal depression shown as 'F* i n F i g 16 was occupied by water i n 1977, but t h i s had drained by 1982 when the s i t e was f i r s t v i s i t e d . The second change was the r e a c t i v a t i o n of a small basalt earthflow above the main earthflow (R i n F i g 16). A creek which formerly traversed t h i s small flow i s now ponded back i n several places and s h i f t e d drainage channels are seen. Severely t i l t e d t rees, without a bent trunk, a t t e s t to recent r e a c t i v a t i o n . 4.2.8 Red Mountain B,C,D,E, and G earthflow group Few a i r photographs are a v a i l a b l e for t h i s l o c a t i o n ; only 3 sets, taken i n 1947, 1964 and 1974 give s u f f i c i e n t d e f i n i t i o n to i d e n t i f y the subtle changes which took place over t h i s period. Sparse vegetation i n t h i s area enabled small changes i n surface topography to be seen. The r e l a t i v e p o s i t i o n of a l l f i v e flows i s indicated i n F i g 4.1. Between 1947 and 1964 (Appendix A, pp. 295-296), transverse tension cracks opened at the top of flow B; these had extended considerably further down the flow by 1974 (Appendix A, p. 296) and had reached a small lake i n the flow track by the time the s i t e was v i s i t e d i n 1985. The l a t e r a l shear zone was e a s i l y located during fieldwork and, comparing the appearance of t h i s shear zone with those at Grinder Creek (A), P a v i l i o n and Drynoch, where movement rates have been monitored, current displacement i s judged to be 80 81 several metres per year. However, no s i g n i f i c a n t displacement of trees growing on the flow was seen between 1947 and 1974, so movement at t h i s s i t e has accelerated since 1974. Sl i g h t movement of slump blocks at the top of flows D and E was observed between 1947 and 1964, although no s i g n i f i c a n t movement was seen between 1964 and 1974. In 1947, the lower parts of flows D and E had l i t t l e vegetation; by 1964 young conifers had increased i n s i z e and a commensurate increase i n cover was observed i n the 1974 photographs. The lower parts of these two flows are currently i n a c t i v e , with vegetation becoming firm l y established, although s i g n i f i c a n t movement i s s t i l l taking place i n the upper parts of D and E. The small flow G does not e x i s t on the 1947 photographs except as an obscure l a t e r a l (?) ridge. However, 1964 photographs show that the flow has reactivated and fresh l a t e r a l deposits had been b u i l t on either side. The bottom portion i s s t i l l i n active and i s cut by a stream. The period to 1974 saw continued movement with backward extension of the headscarp. However, by t h i s time some vegetation was established on the toe of the flow and a transverse step seen i n the previous photograph set had disappeared. The s i t e was not v i s i t e d , so i t s current a c t i v i t y status i s unknown. 4.2.9 Yodel (A) earthflow The e a r l i e s t photographs of t h i s flow date back to 1947 (Appendix A, pp 321-322). At that time the area which subsequently became a flow track was forested with vegetation s i m i l a r to that of the surrounding area. A scar was seen at the head of the i n c i p i e n t flow, but no l a t e r a l deposits, debris fan, or other manifestation of previous movement was v i s i b l e . A small creek flowed down what subsequently became the east margin of the flow. Well-forested former earthflow deposits were seen i n the v a l l e y immediately ad-jacent to the east (flow B) , but these d i d not p a r t i c i p a t e i n the l a t e s t flow phase. Moreover, the whole area had been aff e c t e d at some time by large scale 82 r o t a t i o n a l slumping. No change i n the landscape was seen between 1947 and 1950. The 1961 (Appendix A, p. 322) photographs show the o u t l i n e of the flow. Lat e r a l deposits and a shear zone are seen on the west side and a bulbous snout had formed which protruded a short distance into the v a l l e y . No ac-curate estimate could be made of the magnitude of t h i s movement because stable points could not be f i x e d r e l i a b l y on both stereopairs. Vegetation throughout the flow had been disturbed, with mature trees near the top of the flow replaced by saplings and deciduous vegetation. Two lobes had de-scended from the scarp at the top of the flow and the creek channel was blocked. The 1965 photographs (Appendix A, p. 322) show considerable movement at the toe of the flow, which had turned westward as i t debouched into the main va l l e y . Movement of the toe over the period 1961-1965 i s estimated at 35m. One-third of the way up the flow, large arcuate tension cracks with arcuate slump blocks feeding the flow created a break i n slope. Below t h i s zone the flow was steepest and most ac t i v e with l i t t l e vegetation and many transverse tension cracks. Drawdown below the break i n slope, presumably to feed the rapidly advancing lobe, r e s u l t e d i n a l a t e r a l deposit stranded on the west side of the flow. Above the break i n slope the flow and associated vegetation were not s u b s t a n t i a l l y d i f f e r e n t from those seen i n 1961, except that two lobes issuing from the headscarp were more c l e a r l y defined. A d i r t road was constructed between 1961 and 1965; i t was diverted around the snout of the flow. The 1967 photographs are too small scale to give much useful information about flow movement. The snout had advanced towards the edge of the road (estimated 2-3m since 1965). Several of the slump blocks below the break i n slope had moved downslope and there was more disruption of the remaining mature timber i n the upper portion of the flow. 83 R e l a t i v e l y l i t t l e movement took place between 1967 and 1974; the snout advanced to the edge of the road (estimated 2-3m since 1967). The break of slope was no longer c l e a r l y defined and had broadened to the east. No changes i n the upper part of the flow were seen since between 1967 and 1974. The road had become impassable by 1980 (personal communication S. Oswald, 1985), in d i c a t i n g another 2-3m advance of the flow i n the period since 1967, and grading was undertaken i n 1984. When the flow was v i s i t e d i n 1985, i t was active i n i t s e n t i r e t y , bounded by active shear zones, traversed i n many places by gaping tension cracks and with mature timber leaning at an assortment of angles, although i t had not yet encroached upon the newly graded road. 4.2.10 Earthflow movement - conclusions Only one s i t e , Drynoch (A), shows consistent movement rates throughout the-period of observation and measurement, while the other flows have move-ment rates varying throughout the observation period. The general pattern i s for quiescence u n t i l between 1950 and 1960, rapid movement during the 1950's and 1960's and slower movement to the present time. At Red Mountain (A) and (B) and at Grinder (A) there i s some evidence for increased movement since 1980, although the evidence comes from s i t e v i s i t s rather than from a i r photograph interpretation, which provide d e t a i l e d information but not a comprehensive overview. Section 4.4 w i l l consider c l i m a t i c influences which may have caused these fluctuations i n earthflow a c t i v i t y . 4.3 Monitoring of movement at the Grinder Creek (A) earthflow Flow movement rates were monitored at Grinder (A) for two annual pe-riods (1983/4 and 1984/5). This s i t e was chosen f o r three reasons. F i r s t , a preliminary v i s i t i n 1982 indicated that i t was moving fas t e r than any flows previously monitored i n the study area (Drynoch (A) and P a v i l i o n ) . Secondly, the flow exhibited well defined lobes (Figs M2a to f ) which appeared to be progressing more rapidly than the rest of the earthflow. F i n a l l y , the flow 84 bifurcated giving the opportunity to monitor two flow segments which expe-rienced s i m i l a r regional groundwater flow patterns. Twenty-eight stake arrays were placed along the two flow tracks at Grinder Creek (A) i n early May 1983 (see F i g 4.2 for stake locations). Distances between a l l possible pairs were measured with a tape measure i n late May 1984 and i n mid-June 1985. The stakes were placed so that 2 stakes were on stable ground (inactive l a t e r a l deposits or stable slopes bounding the earthflow) and two stakes were on the earthflow, with a large tension crack representing the l a t e r a l shear zone between the stake p a i r s (see F i g 4.3 for stake array po s i t i o n i n g ) . At two s i t e s (A6 and B6) a series of three stake pairs were placed. These were analysed as two separate pairs although at both s i t e s the uppermost pair showed n e g l i g i b l e movement. Nadler (1984) outlined the geometry of c a l c u l a t i n g the downslope com-ponent of earthflow movement (X component i n F i g 4.3) and l a t e r a l movement (Y component i n F i g 4.3) for the two t r i a n g l e s 134 and 234 i n F i g 4.3. The results given i n Table 4.2. represent the arithmetic mean of downslope and cross-slope movement computed from the two stake t r i a n g l e s . F i g 4.3 and Table 4.2 show that downslope movements on the south arm were considerably greater compared with the north arm i n both measurement periods. They were also more consistent, being s i m i l a r i n both measurement periods. Analogous consistent movement rates have been found over much longer periods for Drynoch (A), within the study area, and for Slumgullion (Crandell and Varnes, 1961). On the north arm movement for 1984/5 was approximately 4 times greater than for 1983/4 at s i t e s with the highest movement rates. P r e c i p i t a t i o n records show (Table 4.3) that r a i n f a l l over the 1984/5 winter season was greater than that over the 1983/4 winter season, so i t was con-cluded that, oyer the two years of measurement, earthflow movement was re-lated to p r e c i p i t a t i o n amount. No s i g n i f i c a n t differences i n flow gradient F i g 4.2,.. Stake array location at Grinder Creek (A) CO 86 Table 4.2;Summary of movement at Grinder Creek S i t e Ax 83/84 Ak 84/85 A y 83/84 Ay 84/85 Extending/ Compressing Al 0 154 0 .639 0 036 0 191 C A2 0 117 0 .616 0 .030 0 065 E A3 0 .203 1 .134 0 .049 0 118 n/a A4 0 .446 1 .360 -0 .024 -0 195 n/a A5 0 463 0 .871 0 .221 -0 151 E A6 0 134 0 207 0 047 0 026 E A7 0 447 1 592* -0 205 -1 167* E A8 0 001 0 .013 -0 .002 -0 003 n/a BI 0 122 1 .379 0 007 0 208 n/a B2 0 221 1 .079 0 .010 -0 088 n/a B3 0 228 1 .643 0 020 0 127 E B4 3 045 0 670 5 283 0 079 n/a B5 0. 344 1 055 -0 067 -0 300 E B6 0. 192 -0 366 E B7 0 .118 0 785 0 004 -0 012 -B8 1 057 4 217 0 246 0 921 E Cl -0 090 0 000 0 023 -0 005 n/a C2 5 547 7 618 -0 439 0 794 C C3 n/a 10 895 n/a -0 517 C C4 7 238 3 148 -2 233 0 035 E C5 3 709 3 365 0 774 0 800 C C6 3 700 3 722 -0 825 -0 524 E DI 0 007 0 044 0 005 0 021 C D2 5 923 6 862 0 524 0 784 C D3 n/a 8 375 n/a 1 237 E D4 3 335 3 194 -0 124 -0 520 E D5 3 627 3 516 1 487 1 521 C D6 0 107 0 103 0 017 0 039 E * Estimated as one peg was disturbed A x Average downslope displacement (average of two peg tr i a n g l e s ) A y Average movement normal to earthflow axis (average of two peg tr i a n g l e ) A l l measurements are i n metres 87 E a r t h f l o w U p s l o p e ir 1 2 ir it 3 4* Y+ -> X+ Y -Stable g r o u n d F i g 4.3 i Stake array configuration with respect to earthflow motion Idealised distance between stakes 7m; topography and earthflow movement dictated actual spacings ranging from 2-20m Table 4.3 A comparison of winter p r e c i p i t a t i o n from several  p r e c i p i t a t i o n stations i n and around the study area•for  1984/1984 and 1984/1985 1983/4 1984/5 Ashcroft 87.4 96.1 Wineglass ranch 120.6 130.5 Porcupine 187.0 223.4 L i l l o o e t (Seton) 133.1* 187.4 Lytton 203.3 253.8 Highland Valley 196.6* 217.4 (cm) These figures are averages for October to A p r i l i n c l u s i v e * Denotes 1 month data missing 88 or geometry were observed between the two arms, so the reason for d i f f e r e n t annual responses on the two arms i s not c l e a r . Movement rates (Table 4.2) also showed that both south and north arms experienced considerable i n t e r n a l v a r i a b i l i t y i n flow rates. On the south lobe, the upper part of the flow was moving at approximately 3.5m per year ( i t accelerated when moving down a major lobe front at D3 and C3, although at t h i s s i t e measurements were only a v a i l a b l e f or the 1984/5 year). At C3 and D3 8-10m of movement were recorded for 1984/5, the highest recorded i n t h i s study. Movement was also rapid immediately downslope of the lobe ( c i r c a 7m/year). However, the two lowest s i t e s on the south arm, DI and C l , showed n e g l i g i b l e movement, although a l a t e r a l shear zone was seen. The north arm showed s i m i l a r v a r i a b i l i t y i n movement rates. The headscarp (A6 and B6) was moving most slowly (0.2-0.6m i n 1984/5), while the flow track recorded rates of around lm i n the same year. S i t e A8, on the north side of the north arm was stable while B8 (4m i n 1984/5) recorded consider-ably higher values than any other s i t e . It seems that the flow track was deflected to the south i n t h i s toe zone, with the northern boundary of the active flow marked by a small l a t e r a l ridge. Reduction of the stake array data also allowed movement towards or away from the shear plane to be calculated ( i . e . y d i r e c t i o n movement i n F i g 4.3). On the south arm, vectors for the upper part of the lobe indicated movement towards the shear plane at C6, D5 and C5, and away at B4 and C4. As the former stake arrays were located above an active lobe, while the l a t t e r were located i n a closed depression, these data were interpreted as the e f f e c t of a lobe swelling i n the flow track at C6, D5 and C5 and the flow track emptying (C4 and D4) to feed a lobe lower down. Evidence for the l a t t e r mechanism was seen i n tension cracks traversing the flow 20m below C4. Sites C3, D3 and D2 a l l showed movement towards the outside of a curve i n the flow. On the north arm the pattern of 'y' component movement was equivocal at the top of the flow, 89 but i n the central part i t was uniformly outwards, suggesting swelling of the flow. Site A7 showed material moving towards the centre of the flow; t h i s was probably due to the diversion of the active flow track southwards round the recently formed mid-flow l a t e r a l . Measurements of the distance between stake p a i r s one and two, located on the moving earthflow, indicated whether the flow surface was extending (increasing i n distance) or compressing (pegs coming closer together). These r e l a t i v e movements are indicated on F i g 4.4 by 'E' and 'C1 respectively. Sites at the top of the two flow tracks, where large blocks of material slump into the flow, showed extending flow (A5, A6, B5, B6, D6). The two rapidly moving s i t e s towards the base of the north lobe (A7 and B8) also showed ex-tending flow; the earthflow at these points was broken into chaotic blocks bounded by deep tension cracks. On the south lobe, both s i t e s D5 and C5, D2 and C2, immediately below advancing lobes, showed flow compression. Conversely, s i t e s C3 and D3 on the most ra p i d l y advancing lobe front, showed flow extension which again corre-sponded with 'crevassed' t e r r a i n . S i t e C4, on the inside of a curve i n the flow, showed compressing flow, while D4, on the outside, showed extension. Hence i t seems that flow was extending both at the headscarp, where blocks broke o f f and entered the flow track, and on r a p i d l y moving lobes. In both places flow movement was the highest recorded. Conversely, where the flow was decelerating, compressive flow was seen. There are several problems and errors associated with stake array measurements. F i r s t , i n t h i s area there i s l i t t l e or no vegetation either on or beside the flow. Consequently, the earthflow material i s very loose and the pegs became loose between measurements. In addition, pegs were heaved, sometimes completely out of the ground. One of the pegs at s i t e A7 was found at the base of a recently opened tension crack; as a r e s u l t , movement for the year 1984/5 could only be estimated. Stakes at D6 were F i g 4.4 Zones of extending and compressing flow on the Grinder Creek earthflow NO o 91 mistakenly not positioned across the main shear plane; consequently movement rates at t h i s s i t e are not representative of measurements at the flow margin. Measurement errors due to tape, operator and loose pegs, were estimated to be less than 5cm as a l l measurements at the i n a c t i v e , or marginally active s i t e s at A8, DI and C l f e l l below t h i s threshold. The second, and more se-rious source of error arises from stake location; i n several cases, p a r t i c -u l a r l y on the south arm, i t was not possible to locate stakes on the flow at the same elevation as stakes on stable ground. This introduces a 'z' component into the c a l c u l a t i o n . No measurements of t h i s angle were made, but at s i t e s C3, D3 and C4 i t introduced a large, but unassessed, error into the movement ca l c u l a t i o n s . However, these s i t e s showed the greatest annual movement, of up to lOm/year; therefore i t was concluded that t h i s error does not invalidate the conclusions drawn from the observed flow patterns. A t h i r d source of error l i e s i n the possible movement of baseline pegs 3 and 4 which are assumed i n the c a l c u l a t i o n to be invariant from year to year. This error was less than 5 cms at a l l s i t e s except A4 (1 year only), B4 (where measurement error i s suspected), B8 (8cm), C2 (39cm), C3 (21cm), C6 (28cm), D2 (39cm) and D4 (39cm). Table 4.2 shows that these errors are very small in r e l a t i o n to t o t a l displacement except at A4 and B4. The C and D series were located p r i m a r i l y on knife-edge l a t e r a l deposits (no alterna-t i v e s i t e s were available) which appear not to provide s u f f i c e n t l y stable ground for long term stake measurements. Three broad conclusions may be drawn from the stake measurements. The f i r s t i s that within an active earthflow, there may be considerable v a r i a t i o n i n movement rates both temporally and s p a t i a l l y even within apparently ho-mogeneous morphological zones. Therefore, estimates of earthflow movement rates based upon a few stake arrays (as for example Crandell and Varnes (1961) at Slumgullion) may not provide an accurate p i c t u r e of o v e r a l l rates of movement. None of the data in Appendix G are derived from a comprehensive stake network, so the only s i t e with comparable measurements i s P a v i l i o n which has been monitored using both stake arrays and an EDM network (Bovis 1986). The second conclusion i s that the response of d i f f e r e n t parts of an earthflow to the same pattern of groundwater fluctuations may be d i f f e r e n t (this s p a t i a l l y v a r i a b l e response i s examined from a h i s t o r i c a l perspective i n chapter 5). F i n a l l y , these data show that earthflow movement does not occur uniformly as an undeformed block undergoing marginal shear, but that movement i s greatest where steep-fronted lobes occur within the earthflow track. Similar lobes are seen on many other earthflows, both active and inac-t i v e . The maps i n Figs M2a to f and Figs M3a to c show the location of d i s -t i n c t i v e lobe fronts at Grinder Creek (A and B), both active and i n a c t i v e , which were recognised on a i r photographs. Many more were found on other flows, but thick forest at many s i t e s prevents recognition of these features i n a i r photographs. 4.4 Climatic chronology 4.4.1 Introduction In this section fluctuations of p r e c i p i t a t i o n over the study area are examined in d e t a i l i n order to i d e n t i f y runs of years of above and below average r a i n f a l l which might be r e f l e c t e d in earthflow r e a c t i v a t i o n and movement. F i r s t , previous work on long term trends i s reviewed, p r i n c i p a l l y that of Barrett (1979). Subsequently, seasonal p r e c i p i t a t i o n patterns for three stations within and around the study area (Lytton, Mamit Lake and Big Creek) are analysed i n more d e t a i l . 4.4.2 Previous work The c h a r a c t e r i s t i c s of p r e c i p i t a t i o n f l u c t u a t i o n s i n B r i t i s h Columbia over the period of instrumental record have been examined by Barrett (1979) using the technique of accumulated departures from period means. This technique operates by c a l c u l a t i n g the period mean and then working out de-93 94 partures from the mean for each data point. The running t o t a l of p o s i t i v e and negative departures i s the accumulated departure. A simple example i s shown i n F i g 4.5. Barrett (1979) showed that, over much of southern B r i t i s h Columbia, a four phase pattern can be discerned. Big Creek, the clo s e s t meteorological s t a t i o n to the study area that was analysed (see F i g 2.2), i s shown here to i l l u s t r a t e t h i s pattern (Fig 4.6). This and a l l s i m i l a r subsequent plots are interpreted by a p o s i t i v e slope i n d i c a t i n g above average p r e c i p i t a t i o n , a negative slope i n d i c a t i n g below average p r e c i p i t a t i o n and a h o r i z o n t a l l i n e average p r e c i p i t a t i o n . The e a r l i e s t phase, Barrett recognised, ending close to the beginning of the record at Big Creek, was one of average or above average p r e c i p i -t a t i o n . It was succeeded by a phase of below average p r e c i p i t a t i o n , l a s t i n g at Big Creek from 1915 u n t i l about 1940. This was followed by 25 years of above average p r e c i p i t a t i o n , ending i n the l a s t 10 years to give a period of approximately average p r e c i p i t a t i o n . A concordant pattern i s observed i n the analysis of r i v e r discharge; McLean (personal communication, 1985, F i g 4.7) shows that the Fraser discharge (both mean annual and maximum d a i l y flow (flood)) underwent a pronounced decline in the l a t e 1930's and early to mid 1940's, followed by an equally pronounced increase i n the late 1950's to early 1970's. 4.4.3 Quality of instrumental records The locations of eleven AES stations i n and around the study area to-gether with summary s t a t i s t i c s have already been given i n Chapter 2 (Fig 2.2, Table 2.1). The study area was recognised as one of t r a n s i t i o n between the coastal (winter p r e c i p i t a t i o n dominant) and i n t e r i o r (summer p r e c i p i t a t i o n dominant) regimes. The boundary between these two d i s t i n c t c l i m a t i c regions was shown to trend NW/SE down the western side of the study area. Superim-posed on this dominant pattern i s a tendency for an increasing proportion B I G C R E E K - T O T A L P R E C I P I T A T I O N F i g 4.7 Cumulative departures of Fraser discharge at Hope, a f t e r McClean and Mannerstrom (1985) 97 of winter p r e c i p i t a t i o n at high elevations. The three c l i m a t i c stations with the longest and most complete records were selected for i n v e s t i g a t i o n (Big Creek, Lytton and Marait Lake). The records from Big Creek, situated 250 km north west ofMamit Lake (Fig 2.2), but showing the same monthly p r e c i p i t a t i o n d i s t r i b u t i o n , are the best available for the study area. They cover the period 1905-1984 (sporadic e a r l i e r records are ignored here), with only 7 incomplete years i n the period up to 1978. The record from Mamit Lake, although shorter (1924-1965), covers periods of both surplus and d e f i c i t p r e c i p i t a t i o n seen at Big Creek. However, l a t e r i n the record, major parts of 6 out of 7 adjacent years are missing (1947-1953), which suggests that the l a t t e r part of the record must be i n -terpreted with caution. The record at Lytton i s drawn from data from two s t a t i o n s : 'Lytton 2', operated between 1917 and 1949 and 'Lytton' recording the period 1944 to present (present i n t h i s study being 1984). Despite assurances from the AES that the stations were located so close as to make the records those of one s t a t i o n , an analysis of the overlap years indicated that 'Lytton' received higher p r e c i p i t a t i o n than 'Lytton 2' i n the summer months although no clear pattern was seen for the winter months, a pattern confirmed by the month-by-month comparison shown i n F i g 4.8. Table 4.4 i l l u s t r a t e s the problem of s h i f t i n g long term means, as over the long term 'Lytton' received more pre-c i p i t a t i o n than did 'Lytton 2'. F i g 4.6, p r e c i p i t a t i o n trends for Big Creek, shows that the period 1917-1949 (Lytton 2) was one of mainly below average p r e c i p i t a t i o n while that of Lytton, 1949-present, was one of above average p r e c i p i t a t i o n . This may explain the contradiction between observations i n the overlap years and long-term means. Therefore, i n the following analysis, the pattern observed for Lytton i s calculated from two d i f f e r e n t means, one (the e a r l i e r ) too low, and one (the l a t t e r ) too high for the whole period; t h i s has the effect of L y t t o n (cm) | l Too many p o i n t s t o be shown i n d i v i d u a l l y y Summer p r e c i p i t a t i o n v W i n t e r p r e c i p i t a t i o n F i g 4 . 8 Month by month comparison of p r e c i p i t a t i o n at AES s t a t ions ' L v t t o n ' and ' L y t t o n 2 ' 99 Table 4.4 Comparison of means for Lytton 2(1917-1949) and Lytton (1944-present) Month Lytton2 Lytton Jan 55 8 71.6 Feb 45 5 51.0 Mar 23 7 29.7 A p r i l 17 6 18. 7 May 21 3 15.2 June 24 4 21.5 July 14 9 12.6 August 18 0 18.6 September 25 3 24.4 October 42 5 46.2 November 52 1 68.9 December 90 7 77.6 Summer 103 4 92.3 Winter 334 4 360.5 Total 429 6 458.9 Overlap years SUMMER WINTER Lytton2 Lytton Lytton2 Lytton 1945 81.1 116.4 M 239 .3 1946 66.8 70.5 362.4 391.2 1947 92.3 96.8 283.5 253.6 1948 120.9 139.7 475.7 432.5 1949 79.0 78.7 276.9 319 .1 • MEAN 88.0 100.4 349.6 349 . 1 100 masking long-term trends. There are few missing years (7 i n t o t a l ) i n the two records. Missing years at a l l stations were given values derived from a re-gression analysis with nearby stations showing s i m i l a r annual p r e c i p i t a t i o n d i s t r i b u t i o n s . These included Hat Creek and Dog Creek (for Mamit Lake and Big Creek respectively) and Bralorne and Shalalth for Lytton. 4.4.4 Seasonal trends Annual data i n the following discussion have been broken down into summer (May to September) and winter (October to A p r i l ) seasons. This d i -v i s i o n was i n i t i a l l y selected by r e f e r r i n g to mean monthly temperatures; these show that during winter months the mean minimum temperature i s below 0°C. During the 'winter' months p r e c i p i t a t i o n was thought to be mainly f r o n t a l , while i n the 'summer' months i t was considered to be p r i n c i p a l l y convectional. Figs 4.9a to c show the accumulated departures from period means for t o t a l , summer and winter p r e c i p i t a t i o n at the three selected stations. It can be seen that the o v e r a l l and seasonal patterns at Lytton (Fig 4.9b) are very d i f f e r e n t from those at the other two stations. At Lytton t o t a l , summer, and winter p r e c i p i t a t i o n a l l reached minima i n 1930. However, the scaled deviations for t o t a l and winter p r e c i p i t a t i o n are small (-1.5 and -1), i n -di c a t i n g that the drying trend was less severe here than at Big Creek (Figs 4.6 and 4.9a). However, as has been explained above (section 4.1.3), t h i s may be an artefact of two d i f f e r e n t period means. The years 1930-1955 showed above average t o t a l p r e c i p i t a t i o n , followed by near average p r e c i p i t a t i o n u n t i l the late 1970's, and then a decline to 1984. This pattern mirrors that for winter p r e c i p i t a t i o n , which provides the bulk of Lytton's record. Summer p r e c i p i t a t i o n , a f t e r 12 years of above average p r e c i p i t a t i o n 1930-1942, has subsequently o s c i l l a t e d without any clear trend. F i g 4.9a Total and seasonal cumulative departures for Big Creek 3-4 Summer 1930 1940 1950 1960 R = values derived from regression analysis F i g 4.9c Total and seasonal cumulative departures for Mamit Lake 104 At Mamit Lake (Fig 4.9c) winter and summer p r e c i p i t a t i o n averages are almost equal (163.8 cm and 154.6 cm r e s p e c t i v e l y ) . Summer p r e c i p i t a t i o n changed from below average to above average i n 1940, winter p r e c i p i t a t i o n l a t e r i n 1945. The seasonal p r e c i p i t a t i o n d i s t r i b u t i o n at Big Creek (Fig 4.9a) i s s i m i l a r to that at Mamit Lake, although summer p r e c i p i t a t i o n i s s l i g h t l y more important here (winter p r e c i p i t a t i o n averaging 144.4 cm, summer p r e c i p i -t a t i o n 176.1 cm). To t a l p r e c i p i t a t i o n (Fig 4.6) i s below average from 1916 to 1940 and above average from 1940 onwards. Summer p r e c i p i t a t i o n i s above average from 1904-1916, below average 1916-1940, above average 1940-1948, and then o s c i l l a t e s with no clear pattern, although the o s c i l l a t i o n s match those at Lytton. However, winter p r e c i p i t a t i o n i s below average from the beginning of the record u n t i l 1943, when i t changes to above average u n t i l the end of useful record (1977). The cumulative departure for winter pre-c i p i t a t i o n at Big Creek (-4.3) i s the largest seen i n the three seasonal records. This analysis shows that a l l three stations switched from below average to above average annual p r e c i p i t a t i o n i n the period 1930-1945. Summer pre-c i p i t a t i o n at Lytton and Big Creek was considerably greater than, average at a l l three s i t e s for 10-12 years following the change, although no clear pattern i s seen i n the subsequent record. Winter p r e c i p i t a t i o n has been above average from 1930 (Lytton) or 1943-1945 (Big Creek and Mamit Lake) u n t i l the late 1970's. 4.5 A comparison of c l i m a t i c trends with earthflow a c t i v i t y In Chapter 1 i t was shown that movement of large earthflows i s sea-sonally dependent on groundwater l e v e l s . Moreover, Palmer (1977) showed that an increase in average p r e c i p i t a t i o n i n the Columbia River Gorge caused the large Wind Mountain earthflow to accelerate. Therefore i t seemed l i k e l y that the increased earthflow a c t i v i t y outlined above was a response to increased 105 p r e c i p i t a t i o n , which had, i n a d i f f e r e n t geologic s e t t i n g in B r i t i s h Columbia, already been put forward as the reason for increased landslide a c t i v i t y (Thomson and Mekechuk 1982). Thornthwaite (potential evapotranspiration) c a l c u l a t i o n s , for Big Creek, Mamit Lake and Lytton (Table 4.5) show that several summer months' p r e c i p i t a t i o n i s lower than p o t e n t i a l evapotranspiration and hence not available for groundwater recharge. Therefore, winter p r e c i p i t a t i o n , which shows an extended period of above average records 1945-1975 (for example at Big Creek) i s probably the main control on recharge, rather than the summer component which shows a shorter maximum to the mid 1950's, with a subsequent decline. A comparison of the Fraser discharge graph (Fig 4.7) and the pre-c i p i t a t i o n graphs from Big Creek (Fig 4.9a) i l l u s t r a t e s that the v i s u a l c o r r e l a t i o n with winter p r e c i p i t a t i o n i s considerably superior to that with summer p r e c i p i t a t i o n . The examination of discharge and p r e c i p i t a t i o n records undertaken i n the previous section shows trends consistent with levels of earthflow ac-t i v i t y . The o v e r a l l pattern i s for low p r e c i p i t a t i o n and discharge i n the 1920's and 1930's, followed by above average p r e c i p i t a t i o n and discharge from about the mid-1940's to mid 19 70's.. This pattern corresponds with observed earthflow r e a c t i v a t i o n around 1950, peaking i n the mid to late 1960's and decreasing at some s i t e s to the present time (Fig 4.10). Therefore, over the period of instrumental records, greatest earthflow a c t i v i t y has been recorded in the period with highest winter p r e c i p i t a t i o n . In Chapter 1 a number of alternate hypotheses were presented which might explain earthflow movement; that of p r e c i p i t a t i o n fluctuations has been ex-plored i n d e t a i l above. The two most important alternate hypotheses are earthquake shock and toe erosion. The earthquake shock hypothesis was d i s -carded because no s i g n i f i c a n t earthquakes took place in south-central B r i t i s h Columbia within or around the study area between 1945 and 1965 and Table 4.5 Thornthwaite c a l c u l a t i o n s Water surplus or d e f i c i t (cms) Month Lytton Mamit Big Lake Creek January 71. .6 29. .3 24 February 42. .2 21. , 1 16. ,9 March 9. . 1 18. .6 17. ,9 A p r i l -30. .3 7. .9 -0. ,4 May -71. .7 13. . 1 -15. ,4 June -109. . 1 6. .9 -19. .3 July -124, . 1 -18. .9 -56. .0 August -102, .4 -4, .0 . -31. 3 September -60, .4 8. .4 -12. .5 October 3 .4 13, .5 4, .3 November 58 .6 29, .4 21, .8 December 77 .6 34, . 7 27, .8 Accumulated d e p a r t u r e s Movement r a t e s ( i n metres p e r y e a r ) from T a b l e 4.1 00 ° ^ 108 furthermore, r e a c t i v a t i o n was not s u f f i c i e n t l y synchronous to be related to one s p e c i f i c event. The hypothesis of toe erosion was also rejected imme-di a t e l y for Black Dome, Red Mountain (B) and Yodel as they were not eroded by a creek or r i v e r at the toe. An examination of the photograph sequences for Grinder Creek (A) (Appendix A) leads to a s i m i l a r conclusion for th i s s i t e as both (A) and (B) i n i t i a l l y reactivated near the headscarp and f a i l u r e subsequently propagated downslope. Therefore, i t was concluded that both toe erosion and p r e c i p i t a t i o n levels influence earthflow movement rates, but that r e a c t i v a t i o n and d i f f e r i n g a c t i v i t y levels were the res u l t of p r e c i p i -t a t i o n fluctuations rather than seasonal toe erosion. 109 CHAPTER 5 THE DENDR0CHR0N0L0GIC RECORD 5.1 Introduction The f i r s t objective of t h i s chapter i s to e s t a b l i s h a c l i m a t i c chronology for the past 400 years, based on c l i m a t i c information recorded i n variable tree r i n g widths. This objective i s achieved by i n f e r r i n g past c l i m a t i c phases from the observed growth response during periods of instrumental c l i m a t i c records. Only long term (20-30 year) trends are relevant to th i s study, as the a i r photographs and instrumental records i n the previous chapter suggest that, at shorter timescales, earthflow a c t i v a t i o n i n B r i t i s h Columbia i s probably linked to 20-30 year fluctuations of p r e c i p i t a t i o n . The second objective of t h i s chapter i s to investigate the recent chronology of earthflow movement by c o r r e l a t i n g phases of reaction wood i n t i l t e d trees with the c l i m a t i c chronology obtained from veteran trees on stable ground adjacent to active earthflows. 5.2 The cli m a t i c record 5.2.1 Growth response of trees to c l i m a t i c variables The p r i n c i p l e s and techniques of dendrochronology are described i n de-t a i l i n Ferguson (1970), and F r i t t s (1976). Many coniferous and deciduous trees, as well as many woody shrubs (Roughton, 1962) form an annual growth-ring. This comprises two types of wood: earlywood, i n which c e l l s are large and t h i n walled, which forms i n periods of rapid r a d i a l growth; and latewood, in which c e l l s are thicker-walled, darker i n colour, and become progressively smaller to the outside of the annual r i n g . C e l l s i n conifers are mainly tracheid: long and t h i n with t h e i r long axes p a r a l l e l to the long axis of the trunk or branch. Hence they are seen i n (almost rectangular) cross-section i n discs and cores. Unless disrupted by f i r e , insect or other extraneous factors, annual growth depends on the tree receiving adequate warmth, moisture and food. At some s i t e s , adequate supplies of a l l three are a v a i l a b l e every year; trees 110 at such s i t e s are complacent, showing l i t t l e or no inter-annual v a r i a t i o n i n ring-width except for the trend associated with aging. Where moisture i s l i m i t i n g , as i n the a r i d southwest U.S.A., wide rings correspond to years of above average p r e c i p i t a t i o n and narrow rings to dry years. In subarctic and alpine climates, summer temperatures rather than moisture govern annual growth (Lamarche and Moony 1972, Jozsa et a l . 1984 and many others). Annual growth i s affected also by growth i n previous years. S i g n i f -icant correlations have been found between growth increments i n adjacent years (Peters et a l . 1982), i n d i c a t i n g that cold or drought i n a given year can reduce growth i n subsequent years, and i n some cases f r o s t or cold damage disrupts c e l l structure (Tapper et a l . 1978). In years when growth i s sup-pressed, the annual r i n g may not form over the whole growth 'cone' and con-sequently may be missing i n samples taken for dendrochronologic analysis. In alder, for example, 40% of the annual rings may be absent (Liese 1978) rendering t h i s species useless for dendrochronology. When the growing season is interrupted, p a r t i c u l a r l y i n the l a t t e r part, f a l s e rings may form giving two or more rings i n one year. These can usually be distinguished by c e l l s i z e , although i n some species double rings cannot be distinguished from annual rings, and thus they may not be used to reconstruct c l i m a t i c h i s t o -r i e s . Tree-ring width chronologies have been used i n many studies to inves-t i g a t e temperature and p r e c i p i t a t i o n f luctuations (for example, F r i t t s et a l . 1979, Kuivenen and Lawson, 1982). They have also been used to reconstruct past a g r i c u l t u r a l conditions (Schulman 1947) and to estimate long-term r i v e r flow levels together with flood p r o b a b i l i t i e s (Stockton and F r i t t s 1972, Stockton 1973). I l l 5.2.2 Tree-ring width chronology construction Samples for reconstruction of a c l i m a t i c sequence are taken from se-ve r a l trees i n an area subjected to either moisture or temperature s t r e s s . Only healthy, mature specimens are sampled. The trees should a l l be of one species as d i f f e r e n t species exhibit d i f f e r e n t degrees of s e n s i t i v i t y (an-nual r i n g width v a r i a b i l i t y , Huber 1970). They may also react to d i f f e r e n t stresses, or i n opposite senses to the same s t r e s s , and show marked d i f f e r -ences i n the time-lag or duration of growth response ( F r i t t s 1976). I f moisture i s l i m i t i n g , then specimens adjacent to lakes, v a l l e y s and on f l a t areas or other locations where s i g n i f i c a n t subsurface water may be present, should be avoided. The sampled trees should be growing as close together as possible, at the same elevation and on slopes with uniform aspect. In thi s way, v a r i a b i l i t y between trees due to s i t e factors i s minimised. Trees growing i n a closed canopy should be avoided because competition and suc-cession may mask c l i m a t i c factors. Samples may be di s c s , wedge cuts, or cores extracted with an increment borer. Duplicate samples should be obtained ( F r i t t s 1976) to investigate within-tree v a r i a b i l i t y , although Peters et a l . (1982) found that the best results were obtained by ignoring tree v a r i a b i l i t y and t r e a t i n g each core independently. This suggests that between-tree v a r i a b i l i t y i s as great as within-tree v a r i a b i l i t y i n homogeneous sampling areas. Analysis begins with cross-dating (the matching up of d i s t i n c t i v e r i n g width sequences). It i s h e l p f u l to have a p r e - e x i s t i n g chronology from an adjacent area to f i x major events such as droughts or f r o s t - k i l l . This procedure eliminates missing and double rings. Any cores which show aberrant growth are discarded at t h i s stage. Ring widths are then measured, and other data such as densitometric measurements (Jozsa et a l . 1984) may be obtained. Cross-dating has yielded chronologies dating back to approximately 6,200 BP 112 on bristlecone pines i n the U.S.A., and interrupted records to 8,500 BP from oaks i n Europe (Becker 1978). The raw r i n g widths are a function of systematic growth trend i n ad-d i t i o n to c l i m a t i c v a r i a b i l i t y (see f o r example raw data i n Appendix D l ) . The growth trend can be removed using an exponential, polynomial (2nd order or higher), moving average or other function to simulate growth at complacent s i t e s . Standardised deviations from the expected growth curve y i e l d a c l i -matic signal together with 'noise', which i s ascribed to s i t e v a r i a t i o n s , competition, succession, f i r e , insect attack or other extraneous factors. The deviations are standardised (see s e c t i o n 5.2.4) to give ring-width i n -dices. Various s t a t i s t i c s can then be calculated such as the mean, standard deviation, the 1st order or higher autocorrelation c o e f f i c i e n t which meas-ures the dependence of each value on the preceding value; and the mean sen-s i t i v i t y , which measures the differences i n r i n g width between successive years. F r i t t s (1976) uses analysis of variance to p a r t i t i o n ring-width var-iance between climate (the common s i g n a l ) , s i t e (between t r e e s ) , i n d i v i d u a l trees (within trees) and random factors (unexplained variance). However, the presence of temporal autocorrelation means that r e s u l t s from t h i s type of analysis should be treated with caution. A chronology can be constructed from the ring-width index by summing and averaging standardised r i n g widths for each year over several trees. However, a chronology obtained in t h i s way i s dominated by the more s e n s i t i v e trees. Peters et a l . (1982) suggest that t h i s i s i n e f f i c i e n t and that a 'better' chronology, one with a higher c o r r e l a t i o n with c l i m a t i c records, i s obtained from p r i n c i p a l components analysis, which emphasises cores with the highest common v a r i a b i l i t y . This procedure gave some of the best r e s u l t s from f i v e or fewer cores i n his study, whereas conventional methods of combination i d e a l l y u t i l i s e at least 20 (Huber 1970). Jozsa et a l . (1984) demonstrate 113 that the method can be extended to densitometric measurements with a con-comitant improvement i n c o r r e l a t i o n with c l i m a t i c measurements. 5.2.3 Sample c o l l e c t i o n Previous work i n the dry i n t e r i o r of B r i t i s h Columbia (Schulman 1947) indicated that both i n t e r i o r Douglas f i r (Pseudotsuga menziezii) and Ponderosa pine (Pinus ponderosa) produced s e n s i t i v e r i n g width sequences with few missing or f a l s e rings. Drew (1975) produced a chronology for Pa-v i l i o n Lake, 5 km from the P a v i l i o n earthflow (Fig 5.1) which, together with some unpublished data from L i l l o o e t (L. Jozsa personal communication), ena-bled marker years to be found at a l l the sampling s i t e s , t y i n g them i n with e x i s t i n g chronologies. It was assumed i n i t i a l l y that t r e e - r i n g widths i n the d r i e r parts of the study area would be c o n t r o l l e d by moisture s t r e s s , as at a l l four s i t e s trees occurred only i n moister places with sagebrush f l a t s occupying d r i e r areas. In addition moisture stress has been shown to be c r i t i c a l elsewhere i n southern B r i t i s h Columbia (Schulman 1947). Sampling s i t e s had to f u l f i l several c r i t e r i a : they had to have adequate forest cover (many earthflows are i n dry sagebrush country with no t r e e s ) , but not closed canopy forest, a requirement which eliminated a l l s i t e s above about 1500m. S u f f i c i e n t num-bers of mature, healthy i n t e r i o r Douglas f i r had to be a v a i l a b l e on dry slopes adjacent to active flows, which themselves bore s u f f i c i e n t mature trees for a useful movement record to be encoded i n reaction wood. Dry slopes adjacent to Drynoch (A and B), Grinder Creek (A), Heginbottom, and P a v i l i o n (A) s a t i s f i e d these c r i t e r i a . Sampling was c a r r i e d out with a 4mm diameter increment borer, either 12" or 15" long, depending on tree diameter. Two cores were taken from each tree, either on opposite sides cross-slope or, i f impractical, at 90°. The cores were stored i n l a b e l l e d drinking straws for transport back to the laboratory. F i g 5.2 Location of chronology trees sampled at Drynoch 116 Contours i n metres Tree numbers Dates show movement i n f e r r e d for groups of trees Fig 5.3 Location of trees sampled for reaction wood at Drynoch See Fig 5.2 for location of this figure 117 A t o t a l of 30 cores from three locations were sampled at Drynoch (Fig 5.2), with duplicate cores taken from each tree. Many trees had heart rot, and so yielded incomplete and broken cores. Several more were rejected be-cause of poor cross-dating q u a l i t i e s . Of the remainder, many had rings so narrow that r i n g boundaries could not be distinguished r e l i a b l y , even with a binocular microscope. Table 5.1 shows that 17 cores, or portions of cores, were retained, including f i v e r e p l i c a t e s . At Grinder Creek, 50 trees were sampled from 3 d i f f e r e n t locations (Fig 5.4), the majority successfully. However, most were younger than 100 years, and so were discarded. F i f t e e n cores from eight trees, each with 100 or more ri n g width measurements, were retained. Samples from the lowest s i t e ( l o -cation 3 i n F i g 5.4) were not used to construct the chronology because they did not correlate with trees from the other two s i t e s , or with trees from d i f f e r e n t flows. Fourteen trees from the Heginbottom s i t e (Fig 5.5) were sampled, y i e l d i n g 23 usable cores (Table 5.1). Fewer trees (7, with 5 r e p l i c a t e s y i e l d i n g 12 cores i n t o t a l ) were sampled from P a v i l i o n , and a l l were located i n the same part of the flow (Fig 5.6). This small sample was adequate because a chronology from P a v i l i o n Lake had already been constructed (Drew 1975) and samples from the flow were necessary only to e s t a b l i s h whether t h i s chronology could be applied to the earthflow s i t e . 5.2.4 Sample processing A l l cores, whether for chronology construction or for reaction wood re-cognition (section 5.3), were processed i n the same manner. They were mounted on plywood and sanded down with progressively f i n e r grades of sandpaper u n t i l a smooth surface showing c e l l structure was obtained. The processing of tree di s c s , used only for reaction wood evaluation, i s described i n section 5.3.4. Tree-ring widths were d i g i t i s e d on an ADD0-X machine; t h i s had a binocular microscope which was linked to a microcomputer. A l l measurements were per-118 Numbered lo c a t i o n s show the s i t u a t i o n of tr e e s used f o r chronology c o n s t r u c t i o n . F i g 5.4 Location of trees sampled f o r chronology, co n s t r u c t i o n at Grinder(A) 119 HEGINBOTTOM L o b e f r o n t ^ T r e e numbers £ ^ A r e a where t r e e s w e r e s a m p l e d f o r c h r o n o l o g y D a t e s g i v e movement i n f e r r e d f r o m r e a c t i o n wood Contours i n metres F i g 5.5 Location of trees sampled both for chronology and r e a c t i o n wood at Heginbottom earthflow F i g 5.6 Location of trees sampled both f o r chronology and r e a c t i o n wood at P a v i l i o n Chronology t r e e samples , — , r— 0 0.5 1.0 i n area of d i s c a n a l y s i s (|^  Area where cores were sampled Area where d i s c s were sampled HI A c t i v e area Base map a f t e r Bovis (1985) Contours i n f e e t 121 formed by the author, thereby eliminating operator variance. F i r s t , raw data plots were obtained for a l l cores, which enabled cross dating to be accom-plished e f f i c i e n t l y . The raw data were then manipulated to remove double rings and to add nominal values for missing rings p r i o r to curve f i t t i n g . (Many c u r v e - f i t t i n g routines are unable to proceed with zero values). This step i s reasonable, as scrutiny of several tree discs (see section 5.3.4) revealed that a number of narrow rings were l o c a l l y discontinuous although r e a d i l y d i s c e r n i b l e throughout most of t h e i r circumference. The chosen sub-s t i t u t e value was ( a r b i t r a r i l y ) that of the smallest adjacent r i n g width. This was probably an overestimate, thereby reducing the s e n s i t i v i t y of the record. It can be seen from Table 5.1 that the majority of trees have few rings missing, so these corrections a f f e c t only a very small proportion of the values used i n curve f i t t i n g . The same missing years are repeated i n many samples, for example 1798 to 1802. A l l cores which did not cross-date s a t i s f a c t o r i l y (throughout t h e i r record) were rejected for the chronology construction. This was a stringent requirement and led to the r e j e c t i o n of over 50% of the cores. After cross-dating, the growth trend was removed using four d i f f e r e n t techniques. The f i r s t and second methods f i t t e d a negative exponential curve to the data, f i r s t by a numeric routine (NL2SN0, from the UBC l i b r a r y ) and secondly by a regression package (SPSS), whereby the data were l o g a r i t h -mically transformed and a best f i t s t r a i g h t l i n e selected (see Appendix Dl for i l l u s t r a t i o n ) . Several studies have obtained a better f i t with a polynomial function, some of these u t i l i s i n g 3rd, 4th and higher order polynomials (Cropper 1984). As the l a t t e r functions do not provide a monotonic functional f i t they mask c l i m a t i c o s c i l l a t i o n s . In t h i s study a second-order polynomial was f i t t e d using the NL2SN0 routine. The three f i t --bx t i n g routines discussed above a l l had three parameters, ( i . e . ae + c and. 2 ax + bx + c) . The sum of squares and pattern of deviations from each f i t Table 5.1 Trees used for chronology construction Tree Replicate Years Missing years Best f i t DRYNOCH DB1A N 1613-1980 1928,1859 P D1BH Y 1791-1975 none E D1BU Y 1625-1979 1887, 1797-1803 P D2H Y 1507-1977 1917, 1771,1663 SPSS D2AH Y 1519-1974 none P D2ASN N 1814-1980 none E D3AEW Y 1764-1980 1796, 1787 P D3ANS Y 1705-1980 none E D4AEW Y 1766-1980 1799-1802 E D4ANS Y 1771-1978 1902 P DA5A N 1663-1908 1814-1820, 1760-1763, P 1702-1704 DAS6 N 1674-1977 1931, 1932, 1800 E DB7 N 1677-1980 none P DA8A N 1814-1980 1928 E DAS 10 N 1533-1980 1833-35, 1691 E DXHHA Y 1726-1979 1870-1871 P DXHHB Y 1769-1979 1843, 1832 P GRINDER B l l Y 1858-1980 none E B11A Y 1864-1980 none P G2CA Y 1853-1979 1884 P G2CB Y 1851-1979 1920 P GC5A Y 1858-1980 none E GC5B Y 1857-1979 1940, 1880, 1871-1873 E G6TA Y 1865-1980 1921-1899 E G6TB Y 1864-1962 1932 E G9TB Y 1869-1980 1961, 1902, 1901 E GC11A Y 1854-1980 none E GC11B Y 1846-1974 none E G14TA Y 1858-1920 1906, 1936, 1982 E G14TB Y 1856-1980 1895, 1923 E G35TA Y 1877-1980 none SPSS Tree Replicate Years Missing years Best f i t HEGINBOTTOM H100W Y 1748-1980 none E H100N Y 1758-1980 1801- 1806 E H200W N 1777-1980 1800 P H300W Y 1745-1980 1920 E H300N Y 1821-1980 1926, 1919, 1918 E H400S2 N 1830-1980 none E H500N N 1783-1974 none P H600S Y 1703-1980 1801, 1802 P H700E Y 1719-1980 1801, 1843, 1855 E H700N Y 1898-1980 none E H800E N 1734-1980 1801, 1802, 1763, 1764 P H900N Y 1657-1980 1945, 1844 E H900S Y 1647-1980 1964, 1843 P H1000E Y 1767-1980 none P HIOOON Y 1807-1979 1843 E HUB Y 1767-1980 1916, 1849 P H11U Y 1739-1980 1768- 1769 P H16SSA Y 1826-1980 none E H16SSB Y 1821-1980 none P H39UDA Y 1780-1980 1947, 1923, 1802 SI H39UDB Y 1780-1977 1823- 1825, 1863, 1856, E 1848, 1828, 1822, 1821 H48UDA Y 1741-1979 1798- 1801 P H48UDB Y 1743-1980 1800 P PAVILION PW1NW N 1521-1976 1793, 1759, 1702, 1604 E PW3U Y 1758-1976 none P PW3D Y 1781-1978 none P PW4U N 1775-1976 1803- 1804 P PW5D Y 1680-1978 none P PW5U Y 1646-1983 1914, 1892, 1748 P PW6D Y 1616-1974 none P PW6U Y 1613-1976 1888, 1889, 1696, 1684 P PW7U Y 1761-1950 1871, 1872, 1843 E PW7D Y 1764-1978 none P PW2U Y 1598-1978 1892, 1676, 1669 E PW2D Y 1602-1968 1927, 1821 P Key E=Exponent i a1 f i t P=Polynomial f i t S=SPSS package f i t 124 were examined to ascertain the "best" f i t . Table 5.1 summarises the best f i t and the range of values, while Appendix D6 l i s t s the three sums of squares. It can be seen that i n only three cases out of 69 does the SPSS routine give the best f i t . T h i r t y - f o u r cores f i t best with a polynomial and 32 with an exponential f i t t e d with the NL2SN0 routine. F i g 5.7 shows that the younger trees tended to f i t best with an exponential function, and older trees with a polynomial. An age of 200 - 240 years marks the s h i f t from exponential to polynomial. The scaled residuals (residual divided by expected value) at a l l s i t e s were plotted and compared with the raw ring-width data (Appendix DI shows some examples). The 13 year running mean i s also shown. It can be seen that a l l three f i t s give a near i d e n t i c a l pattern of residuals with the exception of both ends of the s e r i e s . Plots for cores DA8A, DXHHA, D1BH, D4AEW, D4ANS, D3AEW, DB7, DA5A, H16SSB, H48UDA, H600S, G14TA, G14TB, G2CB, PW2U and PW5D a l l show s i g n i f i c a n t differences i n f i t over the f i r s t 10 and/or l a s t 15-20 years of record. No consistent pattern was found i n the d i s t r i b u t i o n of residuals over the three functions, r u l i n g out any systematic bias between the f i t t i n g procedures. A fourth routine, advocated by Parker et a l . (1984) used a 99 year run-ning average as the 'expected' growth curve. The 'B' and 'C' chronologies were then constructed from residuals around t h i s curve. The 'B' chronology is a weighted 13-year mean of residuals a f t e r the growth trend i s removed (as advocated by F r i t t s 1976), and the 'C' chronology shows the remaining annual v a r i a b i l i t y . Data for four cores from Drynoch are shown i n Appendix DI for comparison with the other three methods. No meaningful sum of squares could be calculated for a quantitative comparison as values p l o t t e d i n Ap-pendix DI are for ri n g width indices, not scaled r e s i d u a l s . Exponential Polynomial 12 -i in a a u u a) Xi E 3 Z 10 8 -4 -2 -S \ \ o CO o vo o o CM O O O co CO o o VO -J-o CM O O O CO o vO o o o CM O o o o o o o o vo <f ^ o co vo ^ vo vo vo vo ir, in o o <f CM in m Age of tree F i g 5.7 Number of trees f i t t i n g exponential and polynomial models^ Ul 126 5.2.5 Operator variance Before constructing a chronology i t was necessary to tes t whether oper-ator or machine error could s i g n i f i c a n t l y a f f e c t r i n g width measurements. The ADDO-X machine measured r i n g widths to the nearest 0.1mm, which was ob-served to be f i n e r r e s o l u t i o n than the zone of blurred contact between ad-jacent rings. Operator error was assessed by the same operator (the author) d i g i t i s i n g a set of 10 cores on two d i f f e r e n t days. The sample was chosen to include specimens with d i f f e r e n t r i n g width frequencies; Table 5.2 shows the break-down of ri n g widths into three s i z e classes for each core. A n u l l hypothesis was formulated, s t a t i n g that the d i s t r i b u t i o n of differences between adja-cent r i n g widths and the d i s t r i b u t i o n of differences between repeated d i g i t i s a t i o n of the same r i n g , were s i m i l a r . An F tes t (Table 5.2) rejected the n u l l hypothesis and i t was concluded that measured year-to-year d i f f e r -ences were 're a l ' and consistent and not art e f a c t s of an imperfect measure-ment system or operator error. 5.2.6 Quality of the records A ring-width chronology i s the sum of many i n d i v i d u a l cores. Therefore, before examining the four s i t e chronologies, a q u a l i t a t i v e and quantitative analysis of the i n d i v i d u a l cores was c a r r i e d out. Table 5.3 gives the mean ring-width, standard deviation, f i r s t - o r d e r autocorrelation and s e n s i t i v i t y for the raw data. It can be seen that raw ri n g width i s extremely variable; both within and between s i t e s i t i s not ne c e s s a r i l y re l a t e d to tree age, although high average ring-width values at Grinder Creek may be explained by the predominance of r e l a t i v e l y young trees at that s i t e . Overall average r i n g widths are lowest at Drynoch and highest at Grinder Creek. This probably r e f l e c t s the eff e c t of a truncated growing season at higher elevation, as Drynoch i s the highest s i t e , P a v i l i o n and Heginbottom intermediate i n ele-vation and Grinder the lowest. Average s e n s i t i v i t y (defined as the average 127 Table 5,. 2 Operator variance e f f e c t s I Core Discrepancy due to mis-sing/double rings Size range no. rings Mean of operator differences Mean of year-to-year differences F value (2 t a i l probability) D1TX 0 <.3mm 17 .0306 .0205 .0894 .0800 9.90 (0) . 3 - . 7tnra 44 .0294 .0270 . 1096 . 1280 22.36 (0) .7-1mm 24 .0358 .0630 .3217 .2410 14.83? (0) W3U 0 <.3mm 13 .0208 .0160 .2331 . 1340 74.43 (0) .3-.7mm 31 .0165 .0140 . .2361 .1900 176.5 (0) .7-1mm 44 .0177 .0190 .4893 .3930 449.06 (0) DA6S 0 <• 3mra 122 .0207 .0230 .0893 . 1100 22.68 (0) .3-.7mm 83 .0220 .0180 . 1378 . 1440 62.85 (0) .7-1mm 31 .0281 .0320 .2826 .2110 43.48 (0) W5D 0 <. 3mm 31 .0132 .0090 . 1848 . 1550 161.13 (0) .3-.7mm 91 .0168 .0150 . 1687 .2070 201.44 (0) .7 -1mm 65 .0195 .0140 .2885 .2490 329.14 (0) S6 0 <. 3mm 145 .0223 .0250 .0851 .0910 12.97 (0) .3-.7mm 71 .0162 .0150 .1865 . 1840 159.43 (0) .7-1mm 40 .0330 .0280 .4410 .2070 242.58 (0) H7DUB 3 <.3mm 76 .0461 .0590 . 1175 . 1300 4.79 (0) .3-.7mm 60 .0423 .1732 12.26 128 .0590 .2070 •CO) .7-lmra 29 .0562 .1080 .2914 .2610 2.22 (.039) DX7HD 0 <.3mm 13 .0254 .0120 .1520 .1420 160.23 (0) .3-.7mm 45 .0296 .0230 . 1480 .1330 32.79 (0) .7-1mm 63 .0263 .0240 .2070 .2170 79.87 (0) j W10ACB 0 <.3mm 1 -• -.3-.7mm 21 .0152 .0140 .1586 . 1890 192.67 (0) .7-1mm 49 .0259 .0170 .1965 . 1860 125.66 (0) WIOACA 0 <. 3mm 4 .0150 .0060 .2600 .2200 14.64 (0) .3-.7mm 28 .0246 .0150 .1654 .1360 87.62 (0) .7-1mm 54 .0300 .0230 .2213 .0280 110.58 (0) H18A 0 <. 3mm 50 .0332 .0340 .2898 .3030 81.59 (0) .3-.7mm 94 .0386 .0360 .1068 . 1620 19.74 (0) .7 -1mm 64 .0272 .0252 . 1911 . 1930 59.54 (0) Table 5.3 Tree ri n g s t a t i s t i c s of chronology trees 129 Tree No. years Mean DRYNOCH DB1A 364 0.67 D1BH 185 1.16 D1BU 365 0.36 D2AH 455 0.50 D2ANS 166 0.54 D3AEW 217 0.62 D3ANS 276 0.54 D4AEW 215 0.64 D4ANS 208 0.48 DA5A 318 0.66 DAS6 304 0.54 DB7 304 0.48 DA8A 168 1.00 DAS10 448 1.20 DAS10N 411 0.49 DXHHA 284 0.47 DXHHB 210 0.48 AVERAGE 0.64 GRINDER G2CA 129 1.10 G2CB 130 1.16 GC5A 125 1.18 GC5B 124 1.02 G6TA 100 1.28 G6TB 113 1.21 G9TA 101 1.14 G9TB 113 1.17 B11A 118 0.92 B l l 124 0.94 GC11A 128 1.20 GC11B 128 1.02 GC13A 111 1.02 G14TA 127 1.04 G14TB 128 1.13 G35TA 105 1.57 G35TB 89 1.72 AVERAGE 1.17 Standard F i r s t order S e n s i t i v i t y deviation autocorrelation 0.42 0.84 0.30 0.67 0.76 0.30 0.27 0.64 0.34 0.32 0.81 0.27' 0.36 0.79 0.29 0.35 0.77 0.26 0.25 0.58 0.29 0.40 0.70 0.32 0.31 0.81 0.31 0.36 0.77 0.28 0.25 0.63 0.33 0.21 0.67 0.27 0.47 0.82 0.20 0.67 0.74 0.31 0.21 0.60 0.30 0.28 0.76 0.31 0.33 0.73 0.34 0.29 0.62 0.61 0.41 0.61 0.59 0.36 0.86 0.76 0.45 0.94 0.86 0.41 1.27 0.76 0.43 1.10 0.72 0.46 0.74 0.65 0.45 0.82 0.72 0.40 0.59 0.68 0.35 0.73 0.85 0.35 0.84 0.76 0.41 0.79 0.83 0.36 0.57 0.42 0.48 1.00 0.86 0.37 0.88 0.87 0.35 0.91 0.60 0.40 0.92 0.61 0.37 0.40 130 Tree No. years Mean HEGINBOTTOM " H100N 223 1.08 H100W 238 1.02 H200W 204 1.14 H300N 160 0.98 H300W 236 1.04 H400S 151 1.35 H500N 242 0.86 H600S 278 0.81 H700N 126 1.51 H700E 262 0.72 H800E 103 0.88 H900N 324 0.62 H900S 324 0.58 H1000E 214 0.98 H1000N 173 1.15 HUB 214 1.00 H11U 242 0.92 H16SSA 164 0.32 H16SSB 159 0.42 H39UDA 201 0.69 H39UDB 198 0.52 H48UDA 238 0.47 H49UDA 239 0.58 AVERAGE 0.85 PAVILION PW1NW 457 0.56 PW2D 367 0.88 PW2U 381 0.87 PW3D 199 1.27 PW3U 224 1.37 PW4U 205 0.95 PW5D 301 1.00 PW5U 309 1.01 PW6D 361 0.87 PW6U 362 0.57 PW7D 217 0.43 PW7U 213 0.44 AVERAGE 0.85 Standard F i r s t order S e n s i t i v i t y d e viation autocorrelation 0.76 0.65 0.43 0.66 0.60 0.44 0.81 0.60 0.45" 0.65 0.61 0.40 0.77 0.70 0.38 0.82 0.55 0.44 0.45 0.48 0.45 0.52 0.57 0.42 0.75 0.52 0.41 0.63 0.65 0.33 0.44 0.46 0.43 0.38 0.62 0.42 0.31 0.57 0.38 0.67 0.67 0.40 0.76 0.68 0.38 0.51 0.60 0.35 0.47 0.63 0.36 0.15 0.19 0.44 0.25 0.56 0.42 0.64 0.65 0.41 0.49 0.73 0.44 0.34 0.66 0.42 0.38 0.63 0.31 0.41 0.32 0.77 0.30 0.48 0.64 0.36 0.57 0.77 0.34 1.06 0.79 0.42 0.86 0.67 0.39 0.65 0.79 0.36 0.89 0.70 0.36 0.84 0.72 0.36 0.66 0.67 0.43 0.32 0.58 0.40 0.35 0.55 0.46 0.32 0.69 0.46 0.39 131 r a t i o of the absolute difference between each two successive r i n g widths divided by t h e i r mean, a f t e r Ferguson, 1970), a measure of the q u a l i t y of the r i n g width data, i s 0.29 at Drynoch and approximately 0.4 at the other three s i t e s . This suggests that tree growth at Drynoch i s less variable, and hence less s e n s i t i v e to c l i m a t i c influences, than at the other s i t e s . This could indicate either lower moisture s t r e s s , again the r e s u l t of higher elevation, or the influence of higher tree density and increased competitive e f f e c t s . Cross-dating q u a l i t i e s at Drynoch were greatly i n f e r i o r to those at the other three s i t e s . This was probably a r e s u l t of sampling -from three widely separated s i t e s at d i f f e r e n t elevations (Fig 5.2). Some differences were s i t e s p e c i f i c : for example, trees from the upper s i t e showed above average growth through the past 10 years and a large peak i n the f i r s t two decades of the 19th century; while those from the lower s i t e registered below average recent growth, and a more pronounced drought i n the 1920's to 1930's. At Grinder Creek no s i g n i f i c a n t differences were noted between the two sampling s i t e s as they were only separated by 100m i n elevation and 400m i n distance. The t h i r d s i t e , where cor r e l a t i o n s were poor, was located approx-imately the same distance and elevation away from the lower of the two useful s i t e s . However, a marked increase i n r i n g width at l o c a t i o n 3 (Fig 5.4) to-wards the end of the'record i s t e n t a t i v e l y ascribed to the increase in groundwater levels since 1940 which i s thought to have i n i t i a t e d earthflow movement. Neither Heginbottom nor P a v i l i o n showed any s p a t i a l trend or pattern in the sampled trees. The quality of each record was quantified by performing a f a c t o r i a l ef-fect analysis of variance on groups of trees at each s i t e selected according to t h e i r length of record. The model was:-132 Annual growth = Grand Mean + Year e f f e c t + Core e f f e c t + Interaction e f f e c t + Random error This model tested whether common annual v a r i a b i l i t y was distinguishable across a l l cores and also whether i n d i v i d u a l core e f f e c t s , such as competi-t i o n , succession, f i r e and insects, were important. Interaction e f f e c t s could not be quantified as no r e p l i c a t e measurements could be taken. However, Peters et a l . (1982) showed that duplicate cores from the same tree d i f f e r e d as much as i n d i v i d u a l cores from neighbouring trees (within a homogeneous sampling s i t e ) . V i s u a l inspection of the raw r i n g width plots confirmed his conclusions. Consequently a l l cores (even duplicate cores) are treated as independent. Groups of trees with s i m i l a r record length were selected (longer records were truncated), y i e l d i n g several groups with d i f f e r i n g but i n t e r n a l l y ho-mogeneous record lengths. F s t a t i s t i c s are given i n Table 5.4, together with the associated F p r o b a b i l i t y of r e j e c t i n g a true n u l l hypothesis ( i . e . no year/core e f f e c t ) . The 'year e f f e c t ' n u l l hypothesis i s rejected for a l l groups, showing that there i s a s i g n i f i c a n t component of common year-to-year v a r i a b i l i t y which i s most probably the r e s u l t of c l i m a t i c s t r e s s . The 'core e f f e c t ' n u l l hypothesis is accepted in al] cases except the shorter records at Drynoch, where the n u l l hypothesis i s rejected at the 95% s i g n i f i c a n c e l e v e l . There, are a number of assumptions underlying the model; v i o l a t i o n s of these assumptions could cause i n f l a t e d F values and so i n v a l i d a t e the anal-y s i s . The assumptions vary according to whether the f a c t o r i a l model i s ran-dom, fixed or mixed, although because there i s no r e p l i c a t i o n , the computational procedures are i d e n t i c a l . The model here i s regarded as mixed because years are a fixed e f f e c t ( i . e . a l l sampling p o s s i b i l i t i e s exhausted) while trees were randomly sampled ( i . e . a small number chosen from a large population). Therefore i t i s assumed that the random error i s normally d i s -Table 5.4 F values for climate and tree e f f e c t s Time-scale DRYNOCH 200 years 200 (overl-ap with 300) 300 years 300 (overl-ap with 400) 400 years GRINDER 110 years HEGINBOTTOM 200 350 Number of samples 15 9 Climate factor F value F p r o b a b i l i t y Tree factor F value F p r o b a b i l i t y 9 4 14 150 years 22 150 (overl- 5 ap with 250) 250 (poly) 5 250 (exp) 5 PAVILION 11 6 COMBINATIONS Pa v i l i o n and 12 Drynoch Heginbottom, 17 Pa v i l i o n and Drynoch P a v i l i o n and 12 Drynoch 7.1807 3.9519 3.7831 3.0916 3.6763 19.2027 19.0873 4.7984 5.3612 4.8018 4.7602 3.9055 6.5118 7.4319 5.6223 0 0 0 0 2.8133 3.5820 1.3728 5.1316 0.6813 1.0119 1.2000 0.6613 0.9043 0.6635 0.5598 0.8272 2.7748 1.7142 0.4894 0.0004 0.0005 0.2025 0.0018 0.5673 0.4369 0.2410 0.6222 0.4623 0.6191 0.8536 0.5320 0.0014 0.0374 0.9112 134 trib u t e d with a mean of zero. It i s also assumed that random error i s inde-pendent of both tree and year e f f e c t s and that the errors have the same variance for each treatment combination population. Hays and Winkler (1970) advise that the f i r s t two assumptions can be relaxed, as we have no reason to doubt the normality assumption, whilst with a balanced model (equal c e l l s izes) the assumption of equal error variance may be v i o l a t e d with l i t t l e r i s k to the experimental conclusions. It i s the t h i r d assumption, that of the independence of the random error, which i s the most important. As the records from d i f f e r e n t trees are temporally autocorrelated (otherwise the sequence would not be any use for dendrochronology), the main problem was to determine whether the 'random e f f e c t ' was indeed randomly and normally d i s t r i b u t e d (personal communi-cation, Department of S t a t i s t i c s , UBC). Histograms and associated normal p r o b a b i l i t y plots (Fig 5.8 and Appendix D3) of the random error term for Pa v i l i o n (200 and 350 years) show that the random error terms i s approxi-mately normally d i s t r i b u t e d , although there i s some bias , p a r t i c u l a r l y i n the 200 year record. Trees J (PW7D) and K (PW7U) i n Appendix D4 stand out as being biased and Table 5.3 shows that these are the most s e n s i t i v e cores of the P a v i l i o n s e r i e s . Scatter pl o t s of predicted values and residuals (Appendix D2) do not reveal any systematic trend. The P a v i l i o n residuals are p l o t t e d i n Appendix D5 to te s t for temporal autocorrelation. In some cases autocorrelation i s unequivocal (as for PW1NW, PW2D, PW3D and PW5U), but i n many other cases no pattern was r e a d i l y seen. Separate F tests were run on some sub-groups of the shorter records, to test whether older trees yielded a more consistent c l i m a t i c pattern than younger. This test could only be performed at Drynoch where core influences ( i . e . non-climatic) were found. The r e s u l t s (Table 5.4) were equivocal, with the 300 year record suggesting that tree factors were important i n those FC AO D " - 8 D B 0 - - - B 3 CD" • *BB' 0 0 A " 8 * A * • * B ADA FD • • • A -OFF E -8 0 • E ' C B 0 FSC B 0 FE C B E CF B CBE AF C FA 0 • • C E F - A 0 8 CE FA 0 0 BB • F • DD BB C ' A - D 0 BB C • " A D 8 B 8 f f " " A - 0 8 F " E " 0 BB F • • • * B A 8 AA • • • • A • A « » * « -.8 - J ,6 '-5 -.4 -'3 ,2 ^ 0 '1 .2 -3 .4 -5 .6 ? RESIDUAL F i g 5.8 Normal p r o b a b i l i t y p l o t of residuals a f t e r f i t t i n g a f a c t o r i a l model to the 350 year P a v i l i o n record KEY A PW1NW B PW2D C PW2U D PW5D E PW5U F PW6U • Too many points for individual identifiers to be shown 136 trees with longer records, contrary to the expectation of increased r e l i -a b i l i t y from older trees. Combined data from Heginbottom, P a v i l i o n and Drynoch for two d i f f e r e n t time periods (250 and 350 years) were tested to see whether a regional c l i -matic signal (year e f f e c t ) was present, or whether i t was masked by tree (or s i t e ) factors (core e f f e c t ) . Table 5.4 shows that shorter term re s u l t s for P a v i l i o n and Drynoch combined give s i g n i f i c a n t tree and core r e s u l t s , prob-ably due to the poorly cross-dated and i n s e n s i t i v e Drynoch cores. A s i m i l a r 250 year combination, but with Heginbottom included, and a 350 year combi-nation of P a v i l i o n and Drynoch both showed a s i g n i f i c a n t c l i m a t i c s i g n a l with no tree e f f e c t s at the 99% s i g n i f i c a n c e l e v e l , although the combination Drynoch/Heginbottom/Pavilion i s not s i g n i f i c a n t at the 95% l e v e l . Cost con-siderations precluded further analysis for other combinations, but these data support the v i s u a l impression gleaned from inspection of the chronology plots i n the next section (Figs 5.9 to 5.13 and Appendix D4) that temporal trends a re f a i r l y synchronous. (These a re considered i n d e t a i l i n the next section). This analysis shows that data from Grinder Creek, Heginbottom and Pa-v i l i o n show a s i m i l a r c l i m a t i c s i g n a l and a re s e n s i t i v e to year - t o-year c l i m a t i c fluctuations, despite h i g h f i r s t - o r d e r autocorrelation (Table 5.3). Data from Drynoch are suspect, p a r t i c u l a r l y the shorter cores which do not c r o s s - d a t e w e l l and i n w h i c h c o r e e f f e c t s , as w e l l as c l i m a t i c e f f e c t s , are important. The longer series from Drynoch are more s a t i s f a c t o r y , although in general as sample siz e increases ( i.e. more years and trees) the i n -creasing number o f degrees o f freedom means that year and tree e f f e c t s become more re a d i l y distinguished. Although most of the assumptions about error d i s t r i b u t i o n s are v a l i d , some temporal autocorrelation remains i n the error terms (Appendix D5). However, the strong signal o f c l i m a t i c influence and the weak or absent tree 137 e f f e c t , together with high s e n s i t i v i t y , suggested that the i n d i v i d u a l cores could be combined into tree ring-width chronologies. 5.2.7 Tree ring-width chronologies The residuals were summed and averaged and the chronologies from a l l four s i t e s are shown i n Figs 5.9 to 5.13 and Appendix D4. Chronologies were constructed for 200, 300 and 400 years at Drynoch, 110 years at Grinder, 150 and 250 years at Heginbottom and 200 and 350 years at P a v i l i o n , corresponding to the F tests i n Table 5.4. Although the F tests were performed on a mix of f i t s , chronologies for both exponential (NL2SN0) and polynomial f i t s were calculated for Drynoch, Heginbottom and P a v i l i o n (200 year record only). Inspection of the plots revealed that only very minor differences e x i s t . Therefore i t was concluded that the type of growth curve f i t t e d makes no p r a c t i c a l difference to the chronology, confirming the contention i n section 5.2 .4 that no systematic f i t t i n g bias was present. At Drynoch, a l l three ( d i f f e r e n t length) chronologies (Figs 5.9 and 5.10 and Appendix D4) showed the same trend over the common period, except over the last 10 years where the two shorter chronologies showed negative r e s i -duals while the longer chronology gave p o s i t i v e values. At Heginbottom and P a v i l i o n no major differences were seen between the longer and shorter chronologies, despite the small number of trees used in c a l c u l a t i n g the longer chronologies (Table 5 . 4 ) . Periods of above and below average growth are p l o t t e d i n F i g 5.14. This and Figs 5.9 to 5.13 show that a l l four records have broadly s i m i l a r patterns of wet and dry years, although the magnitude of the deviations does vary between s i t e s . Both Drynoch and P a v i l i o n show a low growth period from the 1620's or 1630's (this i s the beginning of the record at P a v i l i o n ) ending abruptly and synchronously i n 1692. A period of above average r i n g width follows, being most strongly expressed at Drynoch. This period gradually changes to one of average or below average r i n g width i n the 20 years p r i o r AVERAGE RESIDUAL AVERAGE RESIDUAL-1980 to Ul 09 a H o o o tJ* 4> o o v; (t> o> H o P4 t-i o ts o t—J o 00 o O 3 l-h F -r+ 1968 ' 1956 1944 1932 1920 ^ 1908 1896 . 1884 1872 -1860 -1848 1836 1824 4 K 1812 tn w 1800 1788 • 1776 • 1764 1752 1740 -i 1728 1716 -I 1704 1692 1680 1668 1656 1644 H 1632 1620 r) 1608 1596 1584 OP Ul vO a H << o o =r •p-o o CD 01 H O i-t o O 1—' o OP x o P ro 3 r+ H -H i r t F i g 5.11 Grinder Creek master chronology cr -c o a " LU cc _ LuS cc LU > -an d l JJD r£L . - i — i — i — i — i — i — i — i — i — i — i — ' — i 1 1 [TT1 1920 1914 1908 1902 1896 1890 1884 1878 1872 1990 i r 1974 1968 ~ i — r 1S62 i r 1956 —i r 1950 i r 1944 -1 r 1938 - l r 1932 1926. Year F i g 5.12 Heginbottom 250 year chronology ( exponential f i t ) 0.25 YEAR F i g 5.13 Heginbottom 250 year chronology ( polynomial f i t ) YEAR Grinder (A) Heginbottom P e r i o d s of above average grovth (weakly expressed) F i g 5.14 A v i s u a l c o r r e l a t i o n of "wet" and "dry" phases at a l l four chronology s i t e s 142 to 1800, with Heginbottom showing the most negative residuals. At a l l s i t e s the low growth period ends i n 1801 or 1802, followed by a run of several very high growth years. At P a v i l i o n the years 1803-1823 stand out as those with the most consistently high growth on record and at Drynoch the years 1802-1813 show analogous growth. A period of average to below average growth occurred i n the mid to late 1800*s. At Heginbottom a major high growth period i s seen 1876-1919, a period as favourable at that s i t e as the early 1800's. At P a v i l i o n , Grinder and Drynoch i t i s only a.small excursion and i s l a t e r i n s t a r t i n g (1898-1900). The period ends in 1916 at Drynoch and i n 1919 at P a v i l i o n and Grinder Creek. A major low growth period follows at P a v i l i o n and Drynoch l a s t i n g u n t i l 1936 (Drynoch) and 1944 ( P a v i l i o n ) . At these two s i t e s t h i s phase represents the most strongly negative residuals since the mid/late 1600's. This phase is seen at Heginbottom and Grinder Creek although i t ends e a r l i e r , around 1932, and i s smaller than dry phases of the mid 1870's (both s i t e s ) and the late 1700's (Heginbottom). A period of above average growth follows, ending around 1968 except at Heginbottom where i t ends e a r l i e r , around 1960. S l i g h t l y below average growth i s seen to 1980, the end of the record for the purposes of t h i s analysis. 5.2.8 Correlation with p r e c i p i t a t i o n A v i s u a l comparison of the r i n g width chronologies in Figs 5.9 to 5.13 and Appendix D4 with plots of cumulative departures of p r e c i p i t a t i o n records presented in the previous chapter (Figs 4.9a to c) shows that the general features of the plots correspond well. Average or above average p r e c i p i -t a t i o n was seen in the 1910's (only the Big Creek record covers t h i s period), while low p r e c i p i t a t i o n through the 1920 's and 1930's was mirrored by low tree-ring widths. In the 1940's and 1950's, as p r e c i p i t a t i o n increased, ring-width also increased. Average, or below average, p r e c i p i t a t i o n and ri n g width increments have been recorded since about 1968. 143 A closer examination of the Big Creek and Lytton records, looking at seasonal breakdowns, yielded some i n t e r e s t i n g r e s u l t s . F i g 4.9a shows that at Big Creek summer p r e c i p i t a t i o n was above average u n t i l about 1916-1918, while winter p r e c i p i t a t i o n was cons i s t e n t l y below average from the beginning of the record (1904) u n t i l the mid 1940's. The t r e e - r i n g record showed above average growth u n t i l about 1919, the year a f t e r summer p r e c i p i t a t i o n changed to a lower than average value, suggesting that tree growth was affected by summer rather than winter p r e c i p i t a t i o n . A l l s i t e s then showed low ring-width growth through the 1920's corresponding to average, or below average, sea-sonal and annual p r e c i p i t a t i o n a f t e r 1918. However, r i n g width increased from about 1932 at a l l s i t e s (to above average values at Grinder and Heginbottom). This does not correspond with t o t a l , summer, or winter p r e c i p i t a t i o n at Big Creek, although summer p r e c i p i t a t i o n at Lytton increased from 1930. At Pa-v i l i o n and Drynoch the s h i f t to higher r i n g widths occurred i n 1944 which does correspond with s h i f t s i n t o t a l and seasonal r a i n f a l l at Big Creek. This shows that the c o r r e l a t i o n between r i n g width and climate i s more complex than a d i r e c t c o r r e l a t i o n between seasonal p r e c i p i t a t i o n and r i n g width i n -crement . Above average winter and summer p r e c i p i t a t i o n was observed throughout the late 1940's and the 1950 's and t h i s corresponds at a l l s i t e s with above-average ring-widths. The period ended with a major decline i n summer p r e c i p i t a t i o n from 1969 (Big Creek) and 1972 (Lytton). In addition, sup-plementary data from Hat Creek (Fig 5.15), 30 km from P a v i l i o n , also show consistently low summer p r e c i p i t a t i o n from 1966, while the trend of winter p r e c i p i t a t i o n is indeterminate. However, winter p r e c i p i t a t i o n at Big Creek increased throughout the 1970's while that at Lytton (a s i t e heavily i n f l u -enced by the coastal regime) decreased from 1976. Therefore i t was concluded that, i n this l a t t e r part of the record, trees at a l l s i t e s were probably se n s i t i v e to changes i n summer, not winter, p r e c i p i t a t i o n . Table 5.5 Correlation of rin g widths with cumulative departures  of p r e c i p i t a t i o n and actual p r e c i p i t a t i o n measurements from Big Creek S i t e Season Correlation C o r r e l a t i o n t value cu. P r o b a b i l i t y with raw with cu. departures data departures Drynoch Total 0. ,039 0. 102 0.876 0.40 Summer 0. ,085 0. 197 1.717 0.09 Winter -0. ,104 -0. ,077 no c o r r e l a t i o n Grinder Total -0. ,010 -0. 085 no c o r r e l a t i o n Summer -0. ,014 0. 248 2.217 *0.03 Winter -0. ,003 -0. 224 1.964 0.06 Heginbottom Total -0. ,036 0. , 136 1.189 0.30 Summer -0. ,032 0. ,385 3.617 * <0.002 Winter -0, ,034 -0. ,117 1.007 0.30 Pa v i l i o n Total 0. , 138 0. 297 2.694 * 0.01 Summer 0. ,042 0. ,299 2.714 * 0.01 Winter 0. , 193 0. ,420 1.034 0.20 * S i g n i f i c a n t at 5% l e v e l -e-145 Hat Creek T o t a l CO Q) U 4-> u cd O H 0) -a > •rl 4-1 Cd to T J U O o a) u o CO •p •rH a • H O CO u +1 T m vo ON vO ON vO vO ON CO vO O N o ON CN ON ON vo O N CO ON —t Cl oo ON •1 4 Summer + 1 vo 00 O r-^  CO ON ON ON rH —I t-i i rH CN <r vO 00 O CN vO vO vo VO ON ON ON ON ON ON ON rH rH r-H r-l r—1 i - i • H -1 E l-i a> 4-> *-> U o CO +1 Winter CN <T vo CO vO VO vo vO ON ON ON ON rH i - l r-H rH •1 J Year F i g 5.15 Cumulative departures of p r e c i p i t a t i o n at Hat Cre 146 The r e l a t i o n s h i p between p r e c i p i t a t i o n and t r e e - r i n g width chronology was formally investigated by c a l c u l a t i n g the c o r r e l a t i o n c o e f f i c i e n t over the common period of record (Table 5.5). The scaled residuals were correlated with annual and seasonal p r e c i p i t a t i o n t o t a l s and also with cumulative de-partures of c l i m a t i c data (Fig 4.9a to c ) . The use of cumulative departures was j u s t i f i e d because of high f i r s t order-autocorrelation (Table 5.3) and the knowledge that second-order and higher autocorrelation effects are present. Table 5.5 shows the results of correlations of ring-width residuals with raw c l i m a t i c data for Big Creek and with cumulative departures, which sum departures over the whole record p r i o r to the data of i n t e r e s t . A l l values for r (the c o r r e l a t i o n c o e f f i c i e n t ) are very low. However, i t seems that, with the possible exception of P a v i l i o n , there i s only a weak association between annual p r e c i p i t a t i o n , summer p r e c i p i t a t i o n or winter p r e c i p i t a t i o n ( i . e . October of the previous year through to A p r i l i n c l u s i v e ) and the annual ring-width increment. Results for cumulative departures are s i m i l a r l y i n -conclusive although at a l l four s i t e s summer p r e c i p i t a t i o n departures give the best p o s i t i v e c o r r e l a t i o n s , followed by t o t a l p r e c i p i t a t i o n . Winter p r e c i p i t a t i o n shows no rel a t i o n s h i p or a negative c o r r e l a t i o n , probably the result of several periods (e.g. 1904-1917, 1950-1980) when the cumulative departure curves for summer and winter show d i f f e r i n g trends. Calculations of t values (Hays and Winkler 1970), i n d i c a t i n g whether the c o r r e l a t i o n co-e f f i c i e n t d i f f e r s s i g n i f i c a n t l y from 0 ( i . e . no c o r r e l a t i o n ) , reveal that three out of four s i t e s are s i g n i f i c a n t at the 5% l e v e l for summer p r e c i p i -t a t i o n . It must be recognised, however, that even i f a year-by-year c o r r e l a t i o n i s recognised, t h i s need not be rela t e d d i r e c t l y to earthflow movement. Photographic records reveal that flows move i n response to longer term trends, not to i n d i v i d u a l years of high p r e c i p i t a t i o n . A further problem i s 147 that t h i s analysis recognises that year-to-year tree growth i s more c l o s e l y r e l a t e d to summer than to winter p r e c i p i t a t i o n ; whereas earthflow movement i s probably related to trends i n winter p r e c i p i t a t i o n , since snowmelt prob-ably contributes most to groundwater recharge. These two relationships are i n t u i t i v e l y reasonable, as winter tree growth i s halted by sub-zero temper-atures, whilst summer p r e c i p i t a t i o n i s almost e n t i r e l y consumed by evapotranspiration. The question arises as to whether winter and summer p r e c i p i t a t i o n move i n concert, at least during periods of pronounced p o s i t i v e or negative de-v i a t i o n . The evidence of the l a s t 80 years i s that, over much of the record, they have moved synchronously, p a r t i c u l a r l y during the 1920's when both seasons showed very low values, and during the 1940's, 1950's and part of the 1960's when both seasons showed very high p o s i t i v e departures. Over other parts of the record, trends are divergent; for example, summer pre-c i p i t a t i o n declined during the 1970's, while winter p r e c i p i t a t i o n remained average to above average. In the early 1900's a sharp decline i n winter p r e c i p i t a t i o n was not matched u n t i l 1918 by a drop i n summer r a i n f a l l . Therefore, throughout the r e l a t i v e l y short period of instrumental records, periods of highest and lowest, p r e c i p i t a t i o n ( t o t a l and seasonal) are the re s u l t of winter and summer departures acting i n concert, while during t r a n s i t i o n periods, the two seasons often do not change synchronously. This conclusion is p i v o t a l to the analysis i n the next section and unfortunately the author i s unaware of any l i t e r a t u r e which explains the r e l a t i o n s h i p de-scribed q u a l i t a t i v e l y above. 5.2.9 Summary In t h i s section, methods of core c o l l e c t i o n , processing and the compi-l a t i o n of four ring-width chronologies have been described. Factor analysis was used to j u s t i f y the contention that c l i m a t i c e f f e c t s outweighed tree e f f e c t s at every s i t e (with the possible exception of Drynoch), confirming 148 that the high s e n s i t i v i t y seen i n the dendrochronologic record was the r e s u l t of widespread synchronous c l i m a t i c s h i f t s rather than l o c a l factors acting on i n d i v i d u a l trees, or at sing l e s i t e s . It was shown subsequently that ring-widths i n t h i s area are correlated with p r e c i p i t a t i o n , and most highly correlated with summer p r e c i p i t a t i o n . Therefore, the pattern of ring-widths, which extends back to before 1600 A.D. at Drynoch, can j u s t i f i a b l y be used to i d e n t i f y periods of above- and below-average summer p r e c i p i t a t i o n . The major regional patterns which emerged from t h i s analysis showed below-average summer p r e c i p i t a t i o n throughout the mid and l a t e 1600*s to 1692, when a sudden s h i f t to wetter conditions took place. The damp period graded gradually to average, or below-average, p r e c i p i t a t i o n conditions around 1750-1760. This p e r s i s t e d u n t i l a sudden s h i f t to very moist conditions i n 1801-1803, a phase which per s i s t e d u n t i l 1820-1825. A d r i e r period ended at most s i t e s i n the mid 1870's and was followed by wet years u n t i l about 1920. Then a sudden s h i f t to d r i e r conditions occurred, l a s t i n g u n t i l 1940, when again high p r e c i p i t a t i o n years dominated the record. The years since 1968 have seen decreasing summer p r e c i p i t a t i o n . Although a clear c l i m a t i c s i g n a l emerged from t h i s analysis of c l i m a t i c records, the best ring-width co r r e l a t i o n s are with summer rather than with winter p r e c i p i t a t i o n . However, l i t e r a t u r e reviewed i n Chapter 1 suggested that earthflows responded to winter p r e c i p i t a t i o n stored i n the snowpack rather than to summer p r e c i p i t a t i o n , which i s not u t i l i s e d for po t e n t i a l evapotranspiration (Chapter 4). Inspection of the p r e c i p i t a t i o n curves (Figs 4.9a to c) shows that summer and winter p r e c i p i t a t i o n over the past 80 years show the same o v e r a l l form, that of a major dry period succeeded by one of above average p r e c i p i t a t i o n , although the onset and termination of these periods i s not synchronous. Therefore, the tree ring record i s ac-cepted as a surrogate for cumulated summer p r e c i p i t a t i o n over the past 149 300-400 years, and very t e n t a t i v e l y as a general representation of major winter p r e c i p i t a t i o n f l u c t u a t i o n s . 5.3 Compression wood chronology 5.3.1 Introduction The basic p r i n c i p l e underlying t h i s part of the study i s that earthflow movement e n t a i l s headscarp retrogression and concomitant disruption of veg-etation. Whereas the flow track may remain active across a range of c l i m a t i c conditions, flow enlargement was assumed to be the r e s u l t of c l i m a t i c changes. T i l t e d trees located i n the flow track were not cored, as i t was uncertain whether they could be r e l a t e d to c l i m a t i c f l u c t u a t i o n s . In t h i s study only trees located on well defined slump blocks, or growing adjacent to active tension cracks, were used to construct the movement chronology, as i t was considered that these specimens provided the best record of i n -creased movement due to a more r a p i d l y moving and a c t i v e l y retrogressing earthflow. The other problem with trees growing on the main flow is that the shear plane(s) are usually well below root l e v e l and hence, once the earthflow i s moving 'en bloc', main-flow trees show l i t t l e t i l t i n g . Once a movement chronology was obtained, i t was correlated with wet and dry phases recorded in the tree r i n g record. In t h i s way c l i m a t i c f l u c t u -ations were related to earthflow movement using the assumption that r i n g -width chronology i s a surrogate for p r e c i p i t a t i o n records. 5.3.2 The nature and formation of compression wood Compression wood i s formed on the lower side of many species of older leaning gymnosperms which develop curved trunks as a p h y s i o l o g i c a l response to t i l t i n g . A l e s t a l o (1971) claimed that as l i t t l e as 1° of t i l t produces compression wood, a claim refuted by Schroder (1978) and the author's own observations. T i l t i n g also may be caused by a l i g h t stimulus on a h i l l s l o p e , e s p e c i a l l y in young trees (Phipps 1974). Indeed, as young trees respond 150 ra p i d l y to d i f f e r i n g l i g h t s t i m u l i , the f i r s t 20-30 years of the ring-width record dp not provide a r e l i a b l e record for geomorphic i n t e r p r e t a t i o n . Many review a r t i c l e s describe i n d e t a i l the nature and formation of compression wood i n conifers (Cote and Day 1965, Wardrop 1965, Robards 1969). Only a b r i e f summary of the relevant information i s included here. Growth rings i n compression wood are wider than those i n normal wood, giving an eccentric trunk cross-section. Where well developed, compression wood i s a reddish colour and the boundary between early and latewood i s hard to d i f f e r e n t i a t e . Microscopically, compression-wood tracheids have rounded cross-sections rather than the square, rectangular or hexagonal shapes seen i n normal wood. They also have abnormally thick walls, less c e l l u l o s e and more l i g n i n . Sometimes microscopic s l i p planes may be seen. Compression-wood tracheids d i f f e r e n t i a t e i n the cambium more quickly than normal c e l l s ( S c u r f i e l d 1973), and an intermediate type of wood forms i f t i l t i n g i s i n i -t i a t e d or stopped whilst d i f f e r e n t i a t i o n i s taking place. 5.3.3 Previous applications of chronologies developed from compression wood  formation Schroder (1978) and A l e s t a l o (1971) provide comprehensive reviews of the reaction of trees to many d i f f e r e n t types of st r e s s . Schroder (1978, p 173) recognised many responses, including compression wood formation, scar ti s s u e growing over abrasions and s p l i t s , growth suppression and release and sprouting. He also examined each response and l i s t e d other influences, be-sides mass movement, which might lead to each response. Confounding factors include layered s o i l s of d i f f e r i n g f e r t i l i t y , shading, succession and re-sponses delayed several years a f t e r a geomorphic event. In some cases missing and double rings may render precise dating impossible, although i n thi s study a l l compression wood samples were cross-dated with trees on stable ground. 151 5.3.4 Data c o l l e c t i o n and processing Schroder (1978) states that release, suppression, succession, and sprouting commonly cause a delayed response, and that they may be related to competition, c l i m a t i c or edaphic factors not linked d i r e c t l y to mass movement. Therefore, at the outset of t h i s part of the study s i g n i f i c a n t tree responses were li m i t e d to those of compression wood formation, together with associated 'opposite wood' (pale, narrow growth rings) and scar t i s s u e . Since compression wood occurs also under branches and non-vertical trunk segments, such sampling s i t e s were avoided. Tree cores were c o l l e c t e d at Drynoch (B), Heginbottom, and P a v i l i o n i n 1984 and 1985 with an increment borer, following the f i e l d procedures out-li n e d i n section 5.2.3. Cores were taken from the under and top sides of t i l t e d trees i n order to maximise the contrast between compression wood and 'opposite wood*. The trees selected for coring were a l l t i l t e d and many had curved or twisted trunks i n d i c a t i v e of past movements. Only trees associated with tension cracks or well defined slump blocks, or on active scarps, were sampled. At both Heginbottom and P a v i l i o n only i n t e r i o r Douglas f i r (Pseudotsuga menziesii) was sampled, although at Drynoch some Ponderosa pines (Pinus ponderosa) were also cored (29% of the sample). Z o l t a i (1975) found that older trees did not r e a d i l y produce compression wood and i t was found at Drynoch that older (over 50 years old) Ponderosa pines did not produce reaction wood and were unable to regain an erect stance a f t e r t i l t -ing. However, older specimens of i n t e r i o r Douglas f i r were able to produce reaction wood, although i t was harder to discern and c l a s s i f y i n older specimens as a l l the rings were narrower than i n younger trees. At P a v i l i o n , tree discs were cut from stumps remaining a f t e r a s e l e c t i v e logging operation i n the 1960's. In addition some cores from t h i s s i t e obtained during 1979 were given to the author for analysis by Dr. M.J. Bovis. 152 Tree cores were mounted and d i g i t i s e d as described i n section 5.2.4. Some d i f f i c u l t y was experienced i n mounting cores with a high proportion of reaction wood, as high and uneven compression wood shrinkage led to the twis t i n g of some cores during drying. Plots of r i n g width from both sides of the tree were then cross-dated with ring-width chronologies, and periods of compression wood production were noted. Three types of compression wood were recognised: 'strong' wood, where greatly enlarged rings had a reddish hue throughout and early and latewood could not be d i f f e r e n t i a t e d ; 'medium' wood, where enlarged rings were seen and the latewood appeared to be con-siderably thickened, giving a darker appearance to the rings; and 'weak' wood, where a marked increase i n r i n g width was seen but only a s l i g h t l y thickened and darkened latewood band. These gradings were subjective, but a l l were ca r r i e d out by the same i n d i v i d u a l (the author). The greatest problem was determination of the onset of t i l t i n g or the boundary between d i f f e r e n t classes of reaction wood, as i n many instances both the onset and termination of t i l t i n g was gradual. For th i s reason, re s u l t s are reported i n 5 year periods rather than assigned to i n d i v i d u a l years. Tree discs were polished with a b e l t sander to give a f l a t surface on which tree rings could be r e a d i l y seen. Cross-dating was e a s i l y c a r r i e d out without a microscope, as c h a r a c t e r i s t i c annual r i n g patterns were e a s i l y seen in a complete cross-section. The discs were e s p e c i a l l y valuable because i n a few cases two phases of t i l t i n g i n d i f f e r e n t d i r e c t i o n s were seen and the onset and end of compression wood could be seen more e a s i l y . A l l the discs were photographed to demonstrate the ease with which r i n g width patterns and reaction wood were seen. The response of i n d i v i d u a l trees at a l l s i t e s was p l o t t e d on charts (Appendix D7) which then displayed the number and percentage of trees showing reaction wood i n each f i v e year period of record. The number of trees sam-pled i n each time period, the percentage of those showing a l l three types 153 of reaction wood, and the number of trees showing new reaction wood, or a higher grade of reaction wood, are a l l p l o t t e d i n Figs 5.16 to 5.19. It can be seen that the longest record i s gleaned from tree discs at P a v i l i o n (Fig 5.19, from 1730), followed by cores from Heginbottom (Fig 5.17) and Drynoch (B) (Fig 5.16 from 1770), and f i n a l l y , tree cores from P a v i l i o n (Fig 5.18 from 1850). 5.3.5 Movement chronologies and t h e i r r e l a t i o n s h i p s with the c l i m a t i c  chronology The f i r s t major event to be recorded ( i n Figs 5.16 and 5.19), i s a marked increase i n a c t i v i t y at P a v i l i o n , Drynoch and Heginbottom i n 1800-1805. At Heginbottom and Drynoch the numbers of trees showing strong reaction wood, i n d i c a t i v e of severe t i l t i n g , remained high u n t i l 1830, although at P a v i l i o n the decline did not begin u n t i l 1850. At P a v i l i o n the number of discs showing compression wood i n t h i s period i s more than double the percentage showing compression wood at any subsequent time. A period of low compression wood production followed u n t i l 1860-1865 at Drynoch and P a v i l i o n and u n t i l 1870 at Heginbottom. Fluctuating, but moderately high le v e l s of continuing com-pression wood and compression wood i n i t i a t i o n were seen at Drynoch u n t i l 1910, while at P a v i l i o n and Heginbottom, i n i t i a t i o n rates and the amount of continuing strong reaction wood, increased throughout t h i s period to peak around 1900 and 1900-1915 res p e c t i v e l y . A quiescent period followed, terminating around 1940 at a l l s i t e s with a period of renewed a c t i v i t y . Movement has continued at high levels to the present time at Heginbottom, while at Drynoch (B) there has been a marked decline since 1960. Much of the P a v i l i o n earthflow was s e l e c t i v e l y logged i n 1964 (personal communication A.E. C o l l i n s and R.0, Tinder of L i l l o o e t Forest D i s t r i c t ) which resulted i n considerable disturbance to trees at t h i s s i t e . Therefore at P a v i l i o n , data from 1960 onward were excluded from the analysis . No of trees % trees showing movement 1 i n i t i a t i o n O % Reaction wood 1980 A I960 1940 1920 19O0 1880 H 1860 1040 1820 1000 1780 1760 % trees showing No of trees movement % Reaction wood i n i t i a t i o n o SSI 156 F i g 5.18 Pavilion reaction wood (core.s) % trees showing movement No of trees i n i t i a t i o n Reaction wood LSI 158 The general pattern of movement seen i n Figs 5.16 to 5.19 shows corre-spondence with the ring-width chronologies presented i n Figs 5.9 to 5.13 and Appendix D4, which i n turn have been compared with c l i m a t i c records i n the f i r s t part of t h i s chapter. A l l three s i t e s show a block of wet years beginning i n 1802 (see Figs 5.9 to 5.13). This phase i s the longest and highest i n magnitude throughout the whole record at P a v i l i o n (Appendix D4) and F i g 5.19 shows maximum re-corded movement i n t h i s period. The dry period i n the mid 1800's resulted i n low levels of compression wood formation (Fig 5.17), but increasing moisture from 1860-1870 onwards caused renewed development, increasing i n concert with increased ring-widths (and hence p r e c i p i t a t i o n ) to a secondary maximum during the period 1910-1915. The dry period during the 1920's and 1930's again i s associated with a decrease i n compression wood formation, while during the period 1940-1950 renewed development was seen, e s p e c i a l l y at Drynoch (Fig 5.16), where a major r e a c t i v a t i o n of much of the flow may have occurred. Climatic trends through the late 1960's to the present day show s l i g h t l y above average, or average, p r e c i p i t a t i o n . This i s a decline from the h i s t o r i c a l l y high levels of the 1940's and 1950's. A sharp decrease in reaction wood development at Drynoch i s seen during t h i s recent period, and unchanging or possibly d e c l i n i n g rates of development at Heginbottom (Figs 5.16 and 5.17). These data suggest that above average runs of r i n g widths correspond with increased reaction wood development and therefore with earthflow movement. This movement i s coincidental with phases of above-average p r e c i p i t a t i o n . The implication i s that cool or moist phases have, over the past 300-400 years, brought about phases of increased earthflow movement, at least i n the headscarp regions, many of which are currently active. It i s uncertain whether t h i s has accompanied concomitant acceleration of the entire earthflow complex, although the a i r photograph sequences discussed in Chap-159 ter 4 (for example those i n Appendix A pp. 291-322 show that contemporary reactivations have affected a major portion of several earthflow complexes. 5.3.6 S p a t i a l pattern of movement Careful f i e l d notes were kept to ensure that the movement of i n d i v i d u a l blocks, or movement at marginal tension cracks, was possible to reconstruct at a l l three s i t e s . Maps i n d i c a t i n g reconstructed dates for movement of groups of trees are shown i n Figs 5.3 to 5.5 for a l l s i t e s at which a com-pression wood record was obtained. The record for Drynoch (F i g 5.3) i s supplemented by wedge cuts from f i v e s p l i t trees, s t r a d d l i n g tension cracks, where scar tissue i s healing f i s s u r e s i n the trunk, f i s s u r e s which reach up to 2m above ground l e v e l (Fig 5.20). At Drynoch, samples taken from the top of the flow originated on s l i g h t l y backward-rotated slump blocks, some bounded by active tension cracks. Only two trees were seen to be t i l t e d into a h o r i z o n t a l p o s i t i o n , and the younger vegetation was l i t t l e disturbed. Most of the i n t e r i o r Douglas f i r s sampled on the upper part of the flow showed well developed trunk curvature, i n d i -cating that movement was not recent. They leaned at a v a r i e t y of angles across, up and down the flow. The s p l i t trees were located at the top of a marked break in slope. However, the s p l i t s were p a r t i a l l y healed by scar ti s s u e i n d i c a t i n g again that movement was not recent. Compression wood dates obtained from tree cores (see F i g 5.3 for d e t a i l e d locations) and from wedge cut sections confirmed the impression of past, rather than recent movement, most dating from the 1860-1915 moist phase. Trees and a lobe on the main flow above the b i f u r c a t i o n show r e a c t i v a t i o n i n 1950, as does the e n t i r e south arm, where trees are disrupted by many chaotic slump blocks bounded by fresh tension cracks. It was t h i s r e a c t i v a t i o n which brought about the peak i n movement 1940-1960, seen i n F i g 5.16. Lobes on the north arm were mainly inactive in the late 19th century and early 20th century, although trees on Fig 5.20 Tree at Drynoch sp l i t by an active tension crack 161 the fan have shown continued, but s l i g h t , movement throughout the period of record. At Heginbottom movement i s recorded at four l o c a t i o n s , shown i n F i g 5.5. Location A, with 35 trees sampled, i s currently the most active area on the earthflow, with many small trees t i l t e d at chaotic angles, and with roots spanning active tension cracks. Recent retrogression at the steep headscarp has entrained many larger trees, and these are i n the process of forming curved trunks to regain a v e r t i c a l growth p o s i t i o n . A l l trees on the middle and lower parts of the flow show undisturbed growth p r i o r to 1900, then a massive disturbance i n 1902. Later movement, beginning in the late 1960's and i n t e n s i f y i n g i n the 1970's and 1980's, i s seen mainly at the headscarp of t h i s t r i b u t a r y flow. Location B (Fig 5.5) comprises 18 trees on three lobes. These l i e immediately above a series of transverse ridges, separated by inactive tension cracks. Many of the trees were t i l t e d i n 1800-1840, then again in 1875-1920. Three of the trees show sporadic t i l t i n g post 1960; however, the s i t e presently appears i n a c t i v e , judging by the absence of ac-t i v e tension cracks. Forty-one trees were cored at l o c a t i o n C, i n an area of rapid retrogression. The area i c backed by a steep amphitheatre-shaped rock slope 50-100m high. The few trees growing on the upper and middle parts indicate considerable disturbance, and appear to be s l i d i n g and s l i p p i n g downslope. At the base of the slope, a series of large transverse tension cracks with many curved trees i s seen. Trees at t h i s s i t e were t i l t e d mainly i n the pe-r i o d 1880-1900, but indicate that s l i g h t movement has continued to the present time. Below the tension cracks two steep fronted lobes are seen overriding trees lower down the flow. Trees growing on the lobes indicate more recent movement, from 1950 onwards. Location D at Heginbottom comprises a complex seri e s of tension cracks and lobes above the west side of the flow (Fig 5.5). A series of lobes i s 162 r e a d i l y seen on the a i r photographs (Appendix A, p.302), flowing down the h i l l s i d e but not yet j o i n i n g the main flow. In 1985, 80 stressed trees i n t h i s area were cored. Five of the older trees, scattered throughout the sampling area, showed an older phase of r e a c t i v a t i o n i n 1800-1805. A major lobe on the north edge of t h i s area reactivated between 1935 and 1945. The upper part, characterised by slump blocks and transverse tension cracks, ceased movement around 1965, but the middle and lower parts remain active to the present. A smaller lobe, to the south of t h i s sampling zone, showed r e a c t i v a t i o n at the same time (around 1935), and also remains active to the present. At P a v i l i o n a l l discs and cores came from one area at the head of the active flow (Fig 5.6). Although retrogression has affected the most southwesterly portion of the headscarp, with concomitant slump blocks and tension cracks, many of the older trees and stumps upslope of the currently active area show t i l t i n g , t wisting, e c c e n t r i c i t y , and reaction wood i n d i c -ative of past movement. In addition, many large but inactive slump blocks could r e a d i l y be seen both on the a i r photographs and during ground recon-naissance, s t r e t c h i n g several hundred metres behind (southwest of) the cur-rent area affected by retrogression (Fig 5.6). Most of the discs show enhanced movement or major r e a c t i v a t i o n in 1802. However, those furthest back from the active flow showed a staggered onset of compression wood, from 1750, and many had s l i g h t l y developed compression wood going back to the date of establishment of the tree, i n many cases the period 1700-1736 (Fig 5.19). This sequence suggests r e a c t i v a t i o n at the old headscarp which has slowly propagated down flow; that i s progressive, not retrogressive f a i l u r e . Tree cores from P a v i l i o n were obtained from trees adjacent to currently ac-t i v e tension cracks at the head of the earthflow. Appendix D7 shows that the majority showed movement only in 1964 and the following years, movement interpreted as r e s u l t i n g from logging. However, a small number showed move-163 merit i n i t i a t i o n i n the period 1935-1940; and sporadic or continuous movement p r i o r to 1917. Although the onset of t h i s movement cannot be dated, abrupt termination i n most cases occurred around 1917. 5.4 Conclusions The movement chronology presented above shows three major periods of movement recorded at a l l three s i t e s : the periods 1802-1830, 1870 (or 1910) to 1917, and 1945 onwards. I d e n t i f i c a t i o n of the same three periods of ac-t i v i t y at a l l three s i t e s suggests that the factors causing flow r e a c t i v a t i o n are regional, not l o c a l . These three periods correspond well with periods of above-average ring-width production, which were shown i n section 5.2 to be the d i r e c t r e s u l t of high summer p r e c i p i t a t i o n . However, i t was also shown in section 5.2 that over the period of c l i m a t i c record, large fluctuations occur synchronously i n both summer and winter p r e c i p i t a t i o n . Consequently phases of s u b s t a n t i a l l y above-average ring-width production are shown here to correspond with increased earthflow a c t i v i t y , probably as a r e s u l t of increased piezometric l e v e l s . The s p a t i a l analysis of tree movement shows that, at a l l three flows, r e a c t i v a t i o n of d i f f e r e n t parts of the flow have occurred at d i f f e r e n t times. For example, at P a v i l i o n the active headscarp extended 200m north of the current headscarp during the early 1800's, while at Drynoch (B) one arm of the bifurcated flow has reactivated since 1950, while the other arm, which had moved slowly for the past 200 years, d i d not show comparable movement. A t h i r d conclusion relates the morphology of the three flows studied i n depth with t h e i r behaviour. Drynoch (B), Heginbottom and P a v i l i o n a l l have smaller 1/b ratios than those flows shown to surge i n the previous chapter (see Table 5.1). At both Heginbottom and P a v i l i o n , the flow track i s i l l -defined, and the earthflow may be considered to be a complex comprising se-veral independent flow tracks. At both Drynoch (B) and Heginbottom, many discrete lobe fronts are seen, each at the downslope edge of a lobe which 164 only occupies a small proportion of the width of the flow track, i f the flow track could be adequately defined. It appears that these r e l a t i v e l y wide flows are not able to move rap i d l y , at least i n response to recent c l i m a t i c f l u c t u a t i o n s . Instead, they respond by d i s j o i n t e d movement of unstable portions of the flow. D i f f e r e n t parts of the flow have reactivated h i s t o r -i c a l l y i n response to c l i m a t i c s h i f t s of s i m i l a r magnitude. The implication i s that stable portions of marginally active flows are l i a b l e to reactivate i n a s p a t i a l l y unpredictable response to increased groundwater l e v e l s . 165 CHAPTER 6 POSTGLACIAL CLIMATE IN SOUTH CENTRAL BRITISH COLUMBIA 6.1 Introduction As a necessary prelude to examining the p o s t g l a c i a l chronology of earthflow movement, t h i s chapter establishes the major c l i m a t i c fluctuations recorded over the Holocene i n south-central B r i t i s h Columbia. In the f i r s t part of t h i s chapter, evidence for c l i m a t i c changes i n southwest B r i t i s h Columbia and the surrounding areas i s drawn from p o s t g l a c i a l ice volume fluctuations and palaeobotanical changes since deg l a c i a t i o n . Much of the material i n t h i s f i r s t section i s discussed i n d e t a i l by Clague (1981) and Mathewes (1984). The second part presents a p o l l e n analysis of a core ex-tracted from a subalpine s i t e i n the study area, which was sampled to de-termine the nature of post-Hypsithermal cooling. 6.2 A survey of p o s t g l a c i a l c l i m a t i c f l u c t u a t i o n s i n south-central B r i t i s h  Columbia 6.2.1 Deglaciation and subsequent ne o g l a c i a l fluctuations in i c e volume The majority of radiocarbon dates marking d e g l a c i a t i o n i n B r i t i s h Columbia l i e between 10,000 and 12,500 BP. Although there are few dates i n southern B r i t i s h Columbia, the general pattern i s one of de g l a c i a t i o n at the margins of the C o r d i l l e r a n i c e sheet, p a r t i c u l a r l y at the coastal margins, before deglaciation of the i n t e r i o r . Within mountainous areas, high ele-vation s i t e s became ice free f i r s t , while stagnant i c e lingered i n g l a c i a l v a l l e y s . The chronology of de g l a c i a t i o n provides maximum dates for the i n -i t i a t i o n of many earthflows. Subsequent volume changes of v a l l e y g l a c i e r s can be used to i n f e r p o s t g l a c i a l c l i m a t i c fluctuations using the assumption that i c e volume i n -creased when conditions were cool and/or moist and decreased when they were warm and/or dry. Although there are few g l a c i e r s within the study area, data are available from the southern Coast Mountains and the southern Rockies. 166 The e a r l i e s t work i n the Coast Mountains was that of Mathews (1951) who i n -vestigated the g l a c i a l chronology of G a r i b a l d i Park, 80 km north of Vancouver. He found wood above the present day t r e e l i n e buried i n snowbanks, and rooted stumps sheared o f f by advancing g l a c i e r s . These macrofossils gave dates ranging from 7,640 ± 80 BP to 5,260 ± 200 BP. More recent work i n the Coast Mountains on the Tiedemann and G i l b e r t g l a c i e r s has yielded l a t e r dates for major p o s t g l a c i a l readvances, from 3,345 BP (or possibly e a r l i e r ) to 2,000 BP (Ryder and Thomson, 1986). A l l e y (1976) recognised a s i m i l a r late advance (3,200 - 2,300 BP) to the east of the study area. An e a r l i e r advance (the Dunn Peak) was dated by A l l e y at around 4,500 BP, although the chronology i s under dispute as Duford (1976) and Duford and Osborn (1978) suggest that i t predated the Mazama tephra ( c i r c a 6,600 BP). At a l l s i t e s evidence has been found for a r e l a t i v e l y recent (within the past 1,000 years) g l a c i a l expansion of greater magnitude than any of the other post-Hypsithermal advances. This i s the so c a l l e d " L i t t l e Ice Age" readvance. 6.2.2 Paleobotanical evidence for c l i m a t i c f l u c t u a t i o n s i n the p o s t g l a c i a l  period: an overview Most p o s t g l a c i a l p o l l e n analyses i n south-central B r i t i s h Columbia and the surrounding area show at least three d i f f e r e n t vegetation assemblages and by implication three d i f f e r e n t c l i m a t i c environments. The e a r l i e s t i s that of the late g l a c i a l , when vegetation suggests a cooler, damper climate ( i . e . higher elevation or higher l a t i t u d e ) than that experienced now, or else a cold steppe with no modern analogue. This period ended around 10,000 BP and was succeeded by the 'Hypsithermal', a r e l a t i v e l y warm, dry period (Mathewes and Heusser, 1981). The t h i r d period, that of the cooler, moister contemporary climate, ar r i v e d gradually between 7,000 and 3,000 BP. Around 10,500 - 10,000 BP the cool l a t e - g l a c i a l climate changed to give warmer and d r i e r conditions over the e n t i r e area shown in F i g 6.1. On the 167 1. Ga r i b a l d i Park 7. Fishblue Lake 2. Tiedemann and G i l b e r t g l a c i e r s 8. Finney Lake 3. Marion and Surprise Lakes 9. Kelowna Bog 4. Pinecrest and Squeah Lakes 10. Phair and C h i l h i l Lakes 5. Bonaparte Meadows 11. Simpsons F l a t s 6. Hagar Pond 12. Big Meadow F i g 6.1 S i t e l o c a t i o n 168 west coast i n southern B r i t i s h Columbia, p o l l e n diagrams from Marion, Sur-p r i s e , Pinecrest, and Squeah Lakes a l l show a sudden warming at 10,400 to 10,500 BP (Fig 6.1). Pseudotsuga appeared together with increased Alnus,  Pteridium and, at Marion and Surprise Lakes, Polypodiaceae. Other west coast s i t e s further south i n Washington show a comparable p o l l e n assemblage (Hansen and Easterbrook 1974) although there i s evidence that the Hypsithermal began s l i g h t l y l a t e r than i n B r i t i s h Columbia, between 10,000 BP (Barnosky, 1983) at Bonaparte Meadows (Mack et a l . 19 79) and 8,300 BP at Hagar Pond (Mack et a l . 1978). At Pinecrest and Squeah the Pseudotsuga peak i s accompanied by an i n -crease i n the non-arboreal component, mainly Gramineae, Artemisia and S e l a g i n e l l a wallacei , i n d i c a t i n g that the forest at Pinecrest and Squeah must have been considerably more open during the Hypsithermal, probably as a r e s u l t of moisture stress. At Marion and Surprise Lakes increased Alnus during t h i s period i n conjunction with Pteridium, i n a Pseudotsuga dominated forest, suggests disturbance, possibly r e s u l t i n g from longer and/or d r i e r summers. Corroborative evidence for t h i s i n t e r p r e t a t i o n comes from the ab-sence of moisture-loving species such as Tsuga heterophylla, Chamaecyparis and/or Thu ja from the Fraser lowlands, At the wettest west coast s i t e s (the Fraser lowland and Puget Trough), the f i r s t post-Hypsithermal vegetation change i s marked by an increase i n Tsuga heterophylla and a decrease in Pseudotsuga around 7-7,500 BP (Mathewes 1973, Hansen and Easterbrook 1974, Heusser 1974), and a l a t e r ( c i r c a 6,500 BP) r i s e i n Thuja/Chamaecyparis af t e r the Mazama tephra was deposited (6,600 BP) . An extended phase of p o s t g l a c i a l cooling i s suggested on the west coast by Hebda and Mathewes (1984) who show that red cedar reached maximum Holocene abundance at 2,000 to 3,000 BP, implying that the climate was cooler/moister than the present-day regime. 169 T r a n s i t i o n a l west c o a s t / i n t e r i o r s i t e s such as Pinecrest, Squeah and Fishblue show a Tsuga heterophylla increase before the Mazama tephra ( c i r c a 6,600 BP) implying cooling and increased p r e c i p i t a t i o n . A large r i s e i n Cupressaceae was seen around 3,000 BP, accompanied by a corresponding decline i n Gramineae. A l l the s i t e s i n i n t e r i o r B r i t i s h Columbia show an e a r l i e r temperature peak (and/or p r e c i p i t a t i o n minimum), with the onset of cooler and moister conditions around 8,500 - 8,000 BP. This change i s in f e r r e d from a decrease i n nonarboreal pollen , p a r t i c u l a r l y Artemisia together with an increase i n Pinus ponderosa / P. contorta (almost c e r t a i n l y P. ponderosa at low elevation s i t e s ) , and Pseudotsuga (Alley 1976, Hazell 1978, King 1980, Hebda 1982). Finney Lake, Kelowna Bog, Phair and C h i l h i l Lakes began accumulating peat or g y t t j a between 8,400 and 7,000 BP, which provides a d d i t i o n a l evidence for decreased evapotranspiration. At a l l these s i t e s (except Kelowna Bog) a second cooling phase i s recognised around 4,000 BP, marked by increased pine and the appearance of Tsuga heterophylla, Abies and Picea (King 1980, Hebda 1982). In northeast Washington, the Hypsithermal temperature peaked somewhat l a t e r , between 9,000 and 7.000-6,500 BP. At Simpsons Fl a t s (Sanpoil River V a l l e y ) , Hagar Pond and i n the Bonaparte Va l l e y (Mack et a l . 1978, 1979), there i s an unconformity before about 7,000 BP, interpreted as a period when bogs and ponds dried out and were subjected to d e f l a t i o n . Sites with a continuous record (Mack et a l . 1978, 1979) show increasing P. ponderosa and decreasing Gramineae and Artemisia from 7,000 BP. Increased moisture and decreased temperature i s implied at Simpsons F l a t s , Bonaparte Meadows and Creston around 4,000 BP by an increased proportion of haploxylon pine and more Pinus , Abies and Picea p o l l e n accompanied by a concomitant decline i n Gramineae (Mack et a l . 1976). Hagar Pond and Big Meadow (Mack et a l . 1978a, 170 1978b) show a second cooling phase somewhat l a t e r at 3,000 BP marked by i n -creased Picea and Abies. 6.2.3 P o s t g l a c i a l c l i m a t i c changes i n the study area The analysis of deglaciation dates i n 6.2.1 suggests that the study area became ic e free between 11,000 and 10,000 BP. Further c l i m a t i c amelioration occurred uniformly throughout B r i t i s h Columbia around 10,500 to 10,000 BP and s l i g h t l y l a t e r (9-10,000 BP) i n contiguous northeastern Washington^ Therefore i t i s l i k e l y that a l l s i t e s i n the study area experienced the onset of warm dry conditions synchronously around 10,000 BP. The timing and nature of the close of the Hypsithermal i s much less c e r t a i n . The study area straddles the boundary between the coast and the i n t e r i o r (Chapter 2), and so cooler conditions may have returned as early as 8,000-8,500 BP, as at i n t e r i o r s i t e s , or as l a t e as 7-7,500 BP as on the coast. An e a r l i e r date i s probable as t r a n s i t i o n a l s i t e s such as Pinecrest and Squeah show cooling underway well before the Mazama tephra. It i s l i k e l y that cooling took place in two stages, the f i r s t stage one of gradual cooling and a l a t e r stage, probably more abrupt, corresponding to the a r r i v a l of present-day conditions, at 3-4,000 BP. Although these c l i m a t i c fluctuations are i n f e r r e d mainly from the pollen record, the chronology of g l a c i a l f l u c t u a t i o n s provides independent evidence to corroborate the two-phase cooling i n f e r r e d from vegetation changes. Mathews' (1951) and Clagues' (1981) dates of 7,500-6,000 BP from snowbanks engulfing rooted trees and from dead wood stranded far above present-day t r e e l i n e correspond with the onset of cooler conditions, whilst Ryder and Thomsons' (1986) g l a c i a l maxima dated at 2-3,500 BP correspond to the second stage of c l i m a t i c d e t e r i o r a t i o n i d e n t i f i e d by p o l l e n analysis at most tran-s i t i o n a l and i n t e r i o r s i t e s . 171 6.3 Pollen analysis of a core from Red Mountain (A)  6.3.1 Introduction The preceding l i t e r a t u r e review provides a general o u t l i n e of the p o s t g l a c i a l ( c l i m a t i c h i s t o r y of B r i t i s h Columbia and the surrounding areas. Most of the evidence i s eit h e r i n the form of po l l e n analyses, providing a picture of vegetation changes at low elevation s i t e s , or as volumetric changes of alpine g l a c i e r s . The closest s i t e s from which p o l l e n data are available are Pinecrest and Squeah Lakes i n the Fraser Canyon, Phair and C h i l h i l Lakes near L i l l o o e t , Horseshoe Lake near Pemberton and Finney Lake i n the Hat Creek Valley (Fig 6.1). As the highest s i t e s (Finney Lake, Phair and C h i l h i l ) l i e i n the rainshadow of the Coast Mountains (Chapter 2) and the lowest, Pinecrest and Squeah, are located i n the c l i m a t i c a l l y t r a n s i -t i o n a l zone of the Fraser Canyon, vegetation at a l l these s i t e s i s moisture stressed and hence very s e n s i t i v e to changes i n e f f e c t i v e p r e c i p i t a t i o n ( i . e . p r e c i p i t a t i o n minus evapotranspiration). The e x i s t i n g palynological data are probably applicable to the lower elevation earthflows i n the study area, p a r t i c u l a r l y as most occur i n the Fraser Valley and head at 1,000m or lower. There i s only one high elevation c l i m a t i c s i t e (Goat Meadows, F i g 6.1, unpublished data) while many earthflows in the study area either head above 1,500m or are fed by groundwater systems o r i g i n a t i n g from 1,500-2,000m or higher elevations (see F i g 2.1 and F i g 2.2). Section 2.6 showed that the winter p r e c i p i t a t i o n maximum i s more pronounced at high elevations, so i t i s possible that the high elevation s i t e s may have experienced c l i m a t i c changes s i m i l a r to those i n the west coast lowlands, rather than those experienced at i n t e r i o r or t r a n s i t i o n a l lowland s i t e s . Therefore a high elevation s i t e at Red Mountain was selected for pollen analysis to assess c l i m a t i c changes at high elevations within the study area. 172 6.3.2 Description of the sampling s i t e at Red Mountain (A) A seven-metre g y t t j a and peat core was extracted from Bog G on the Red Mountain (A) earthflow. The regional s e t t i n g i s shown i n F i g 15 and the flow morphology i n F i g 16, and the bog i s l a b e l l e d on the relevant a i r photographs i n Appendix A, p. 305. Sediment i n Bog G has accumulated i n a cut-off de-pression created when the south arm of the Red Mountain (A) flow dammed a small creek. Although the creek has cut a small, steep sided gorge about lOrii deep i n order to maintain i t s course, water le v e l s i n the depression were maintained u n t i l recently by a beaver dam. Therefore the bog i s bounded to the south and east by earthflow material (mainly weathered basalt) but to the north and west by r h y o l i t e flows which create a stable i s l a n d around which the flow bifurcates (Fig 16). Vegetation zones are shown i n the schematic diagram i n F i g 6.2, which also shows the location of the present-day stream channel and the sampling s i t e . Vegetation nomenclature i s taken from Hosie (1969) and Hitchcock and Cronquist (1973). The western half of the bog has abundant Sphagnum, i n -d i c a t i v e of an acid environment, and dead or dying stunted white spruce (Picea glauca) which are probably starved of nutrients by the t h r i v i n g Sphagnum. To the north the surface i s covered with white sand to cobble s i z e angular fragments derived from the erosion of r h y o l i t e flows described i n section 3.2.5. As the h i l l s l o p e surrounding the depression has very l i t t l e vegetation, the fragments probably r e s u l t from slopewash rather than from deep-seated mass movement or stream erosion. A low ridge, possibly a former beaver dam, separates the western Sphagnum /spruce area from the eastern portion where the stream divides around a large i s l a n d , the centre of which ov e r l i e s the deepest part of the bog (approx. 7m) maximum depth. The dampest parts of the study s i t e are occupied by Cyperaceae and Gramineae together with scattered composites and Apiaceae, notably Angelica  genuflora. A few scattered small Sphagnum mounds were seen on the wettest 173 Ef Earthflow R Rhyolite -_rr Old beaverdam '•'•'} Predominantly sphagnum with white spruce •tit:-. . • Zone 1A Salix (in drier areas) Zone IB Salix Cyperaceae Cyperaceae Angelica genuflora Grasses Engeron Bidens Bidens Compos it-ac Sphagnum Zone II Shepherdia canadensis Rosaceae Zone III Arctostophylos uva-ursi Rosaceae Zone IV P. albicaulis P. contorta Picea engelmanii Rosaceae N Zone IV F i g 6 .2 Schematic diagram of vegetation zones of Bog G 174 spots and small S a l i x bushes occupied d r i e r s i t e s i n Zone I. To the north, a narrow zone (Zone II) of S a l i x and soapberry (Shepherdia canadensis) fringes the bog, followed by a zone of Rosaceae and kinnikinnick (Arctostophylos uva-ursi) (Zone I I I ) . The alpine forest immediately sur-rounding the bog i s mainly pine (mostly Pinus a l b i c a u l i s with a l i t t l e Pinus  contorta), scattered spruce (Picea engelmanii) and dwarf juniper (Juniper  h o r i z o n t a l i s ) • In damper, shaded areas, such as the bog o u t l e t , fireweed (Epilobium alpinum) i s common i n the understory, while on d r i e r s i t e s Lupinus grows between mature, well spaced pines. F i g 6.3 i s a photograph of the north side of the bog showing these vegetation zones. The open forest surrounding the bog comprises a uniform white pine (Pinus  a l b i c a u l i s ) forest. At the lowest (northeasterly) extreme of the earthflow some trees reach t h e i r maximum of 25m (Hosie 1969). Tree height decreases with elevation, reaching 10m around Bog G and forming wide spreading, f l a t growing, twisted, spreading shrubs at high elevations, p a r t i c u l a r l y where exposed to the wind. Epilobium and Shepherdia are frequently part of the understory, together with other alpine composites. Picea glauca forms pure stands at moist s i t e s . No mountain hemlock (Tsuga mertensiana), alpine f i r (Abies .1 asiocarpa), larch (Larix) or alder (Alnus) were seen. At lower elevations i n t e r i o r Douglas f i r (Pseudotsuga menziesii var.  glauca) i s common together with some stands of western hemlock (Tsuga  heterophylla) , although pines are s t i l l a major component of the vegetation. Trembling aspen (Populus tremuloides) grows on dry h i l l s lopes while Alnus is common i n disturbed areas. At s t i l l lower elevations (below 1,200m) and to the east of the sampling s i t e , the vegetation i s that of x e r i c grasslands with sagebrush (Artemisia t r i d e n t a t a ) and grasses dotted with ponderosa pine (Pinus ponderosa) and i n t e r i o r Douglas f i r . 175 View to the South North side Fig 6.3. Vegetation on Bog G 176 6.3.3 Sampling and laboratory techniques Peat and g y t t j a samples were extracted from the deepest part of the bog with a H i l l e r corer, as attempts to core with a Livingstone corer f a i l e d because the g y t t j a was too s t i f f to be penetrated with a hand-operated core b a r r e l . The H i l l e r sampled i n 50cm sections and subsamples were taken at 12.5cm i n t e r v a l s . Sampling took place i n August 1984 and a supplementary basal sample was extracted i n September 1985 for an ad d i t i o n a l radiocarbon date. The subsamples used for pol l e n analysis were i n i t i a l l y preserved from decay by the addition of a few drops of phenol as no r e f r i g e r a t i o n f a c i l i t i e s were available i n the f i e l d . Multiple samples of g y t t j a were extracted from the base of the core and from a wood enriched horizon at 6.75 - 6.87m. These were a i r dried i n the f i e l d and l a t e r oven dried i n the laboratory and sub-mitted for radiocarbon dating. Laboratory procedures described by Mathewes and Rouse (19 75) and A l l e y (1976) were used to prepare microscope s l i d e s from the sampled material. F i r s t , exotic Lycopodium tablets were added to each weighed, dried sample (this enabled the subsequent determination of pol l e n frequency). Each sub-sample was screened and then treated with 5% KOH i f a s i g n i f i c a n t proportion of large organic fragments was present, and with HF i f the sample had a mineral component. If necessary ZnBr^ was used to separate out the heavy minerals followed by ac e t o l y s i s to remove non-polleniferous organic mate-r i a l . The samples were then sieved to remove any remaining f i n e fragments, the residue was bleached and stained pink with safranin 'O1 and mounted on microscope s l i d e s . The s l i d e s were then scanned at 400x magnification and the pollen grains i d e n t i f i e d and counted, although where necessary lOOOx magnification was used to aid i d e n t i f i c a t i o n . 300-350 p o l l e n grains were counted for each horizon as well as fungal spores and cysts which were, in some horizons, up to 10 times more abundant than po l l e n grains. 177 6.3.4 Stratigraphy A d e t a i l e d s t r a t i g r a p h i c diagram i s given i n Appendix F, pp 447-448. Fi g 6.4 shows a condensed version which can be compared with the p o l l e n assemblage. At the base of the core an angular gravel layer with scarce wood fragments was found: t h i s was penetrated to a maximum depth of 12.5cm. Im-mediately above the basal gravel, brown/black g y t t j a containing sparse wood and charcoal fragments was found. These basal clays yielded a g y t t j a radiocarbon date of 7,790 ± 80 BP (Beta 14273). A discontinuous layer of wood fragments was found at 6.75 to 6.8m. which yielded a radiocarbon date of 7,550 ± 90 BP (GSC 4181). At 6.5m rounded gravel and charcoal were found but no further coarse sediments were deposited in the g y t t j a . Root hairs were found at 4.5m and a l i g h t brown layer at 4.25 to 4.26m which was t e n t a t i v e l y i d e n t i f i e d as Mazama tephra. A sample was submitted for electron microprobe analysis but to date no r e s u l t s are available. F i g 7.1 shows that t h i s s i t e l i e s outside the recognised l i m i t s of the Mazama tephra f a l l , which would account for the meagre thickness of t h i s usually abundant tephra. Bridge River tephra shards (about 2,400 BP) are found at 1.96 to 2.08m. These, are sorted, with the coarsest basal c l a s t s reaching 2mm. Above the Bridge River tephra l i e s f l u v i a l l y deposited clay with two oxidised layers, followed by s i l t which i s also p a r t i a l l y oxidised and contains reworked shards of Bridge River tephra. Above 1.15m the core shows f l u v i a l sediments rather than the l a c u s t r i n e sequence seen lower i n the column. The f l u v i a l sediments have textures varying from clay to a gravel/clay mixture, i n t e r -mingled with large wood and/or charcoal fragments, roots and decayed sedge. The stratigraphy suggests that when the depression was i n i t i a l l y formed there were trees growing at or very close to the sampling s i t e . The de-pression was then occupied by a lake which became shallow, or dried out b r i e f l y , s l i g h t l y before the Mazama tephra f a l l . Deep-water conditions en-178 sued u n t i l the Bridge River tephra f a l l (2,400 BP), a f t e r which f l u v i a l de-posits dominated, i n d i c a t i n g e i t h e r that the lake had f i l l e d with sediment or that the o u t l e t had cut down to lake bed l e v e l . Large wood and charcoal fragments show that trees were probably present on t h i s p ortion of the bog during most of the past 2,400 years. The two radiocarbon dates and the po-s i t i o n of the two tephras indicate a f a i r l y constant rate of deposition over the post Mazama i n t e r v a l , but a much higher rate i n the lower part of th'e core. 6.3.5 Pollen frequency and percentage organic matter Pollen frequency, expressed as the number of p o l l e n grains per gram of sediment, shows that the depositional h i s t o r y of the lowest metre of the core d i f f e r e d from the upper 6 metres (Fig 6.4). The lower core has more than ten times the p o l l e n frequency found i n upper part of the core. This i s probably the r e s u l t of two fa c t o r s . F i r s t , the lower part of the core re-presented a period when bog G was a very shallow lake (this w i l l be discussed in 6.3.7) and forested around i t s perimeter. It i s possible that water flow was slow and that there was no o u t l e t . Under these conditions p o l l e n would not be flushed out of the basin. Secondly, the lower part of the core records pollen at the end of the Hypsithermal i n t e r v a l , when temperatures were s i g -n i f i c a n t l y higher than present, and hence p o l l e n production was possibly greater than present. Both factors would contribute to a higher p o l l e n percentage. Percentage loss on i g n i t i o n at 350°C was also determined (Fig 6.5). This showed three zones, broadly c o r r e l a t i v e with stratigraphy and p o l l e n f r e -quency. The lowest zone, from 6m to the base of the core, had a r e l a t i v e l y low organic material content (<15%) which corresponded with the zone of highest pollen frequency, and p a r t l y to a d e t r i t a l zone. This suggests a r e l a t i v e l y high i n f l u x of mineral material and p o l l e n , possibly due to slopewash and to trees growing on, or very close to, depositional s i t e s . The RED MOUNTAIN Zona I Zone II Zone I I I Zone IV Zone V 7 790180 BP |8* to so Peal [til Cnarcoal j Gyttja [ilj Gravel |^ /| wood Sartd and sill Number of cysts o 2oo 400 eoo eoo tooo pollen x 10 / g Fig - 6 . 4 P o l l e n diagram for Bog G showing " l o c a l " zonation F i g 6 .5 Percentage loss on i g n i t i o n , Bog G 181 numbers of recycled (Eocene) grains are also shown and i t can be seen that these were most frequent i n the mineral layer towards the base of the core. The basal layer was succeeded by 2m (depth 4-6 m) of continuously high organic g y t t j a (15-25% loss on i g n i t i o n ) . This i s interpreted as a conse-quence of high p r o d u c t i v i t y i n a l a c u s t r i n e environment during the l a t e Hypsithermal i n t e r v a l , and i s consistent with lower p o l l e n percentages i n the upper part of the core, which accumulated during cooler conditions characterised by a decreasing p o l l e n percentage. A zone of highly variable organic content and decreasing p o l l e n percentage followed, ceasing with de-p o s i t i o n of the Bridge River tephra. Organic content above, the tephra i s generally lower, p a r t i c u l a r l y i n the overlying one metre since f l u v i a l sediment comprises t h i s zone. Organic content increases again towards the surface (0-lm) as peat and woody remains form a large volumetric proportion of the f l u v i a l sediments. Increased pollen frequency here i s ascribed either to the e f f i c i e n c y of peat i n trapping pollen, or else to the termination of p o l l e n export from the depression as the former lake evolved into a bog. 6.3.6 I d e n t i f i c a t i o n of palynomorphs I d e n t i f i c a t i o n of palynomorphs was mad? with reference to the c o l l e c t i o n of Dr G.E. Rouse at the University of B r i t i s h Columbia. Many algal cysts were found in the core but no reference c o l l e c t i o n was a v a i l a b l e for iden-t i f i c a t i o n . As a l g a l cysts have now ceased accumulating, either because the pond changed to a bog or because of a degeneration i n climate, i t i s not now possible to determine the o r i g i n of the cysts. The main types of cysts were photographed in F i g 6.6. Two d i f f e r e n t grasses were seen, but they were not traced to p a r t i c u l a r species. Composites were apportioned to T u b u l i f l o r a e or L i g u l i f l o r a e on the basis of ornamentation, and Chenopodiaceae/Amaranthaceae were not d i f f e r e n -t i a t e d . Small amounts of pollen from a number of d i f f e r e n t species were seen F i g 6.6 contd 185 including Apiaceae, Arenaria, Actea rubra, Sagataria, P o t e n t i l l a ,  U r t i c a , Symphoricarpus, Heracleum, Polemonium and L i l i a c e a e . Pine p o l l e n was c a r e f u l l y examined under l,000x magnification to deter-mine whether i t was diploxylon (P.contorta or P. ponderosa) or haploxylon (P. a l b i c a u l i s ) . The haploxylon pines have warts on the leptoma whereas the diploxylon pines have no ornamentation i n that area (Ting 1975). The majority of pine palynomorphs at a l l horizons were haploxylon: these are a t t r i b u t e d to P. a l b i c a u l i s which i s the dominant species around Bog G. There are so few P. contorta grains that they are not shown separately i n the p o l l e n d i -agram (Fig 6.4). The Pseudotsuga category may include some Larix as the two grains could not be r e l i a b l y d i f f e r e n t i a t e d . Cupressaceae p o l l e n was sepa-rated into Juniperus and Chamaecyparis on the basis of ornamentation as de-scribed and i l l u s t r a t e d by (Kapp 1969) and by comparing i n d i v i d u a l grains with reference s l i d e s . Palynomorphs are reported as a percentage of t o t a l p o l l e n and cryptogam spores, excluding Potamogeton which i s an aquatic. Numbers of fungi and cysts were then standardised to the number per 300 p o l l e n grains (the minimum number counted). Consequently two d i f f e r e n t scales operate i n Figs 6.4 and 6.7, a percentage scale for pollen and an absolute scale for fungal spores and a l g a l cysts. 6.3.7 Palynomorph zonation - methods Inspection of the r e l a t i v e frequencies displayed i n F i g 6.4 reveals two d i f f e r e n t palynomorph assemblages within the one s i t e . The f i r s t i s the ' l o c a l ' assemblage (this zonation i s shown i n F i g 6.4), a zonation using pond fringe and bog loving species such as grasses, herbs, Spirea, Myrica,  Rosaceae, Potamogeton and Cyperaceae. These have responded p r i m a r i l y to changes i n water l e v e l and depositional environment which are i n f e r r e d from the s t r a t i g r a p h i c record, although post-Hypsithermal cooling also must have had an influence. The second pattern i s that of the conifers and angiosperms 186 i n the subalpine forest which are thought to record c l i m a t i c a l l y - i n d u c e d changes i n regional vegetation. Species r e l a t e d to the change i n environment at the bog/pond w i l l be referred to as the ' l o c a l ' s i g n a l , while species growing i n the surrounding forest and represented by p o l l e n blown i n from further a f i e l d w i l l be re-ferred to as the 'regional' s i g n a l . A l i s t of species apportioning them to the regional or l o c a l s i g n a l i s given i n Table 6.1. Tauber (1965) showed that for a bog of t h i s s i z e (100-200m i n diameter) 80% of the pollen i n f l u x would o r i g i n a t e from the 'trunk space' of the sur-rounding forest, 10% from the 'canopy space' and only 10% from long distance t r a v e l . Therefore most p o l l e n records show only the immediately surrounding forest rather than the regional p o l l e n r a i n . However, the forest surrounding the bog changes i n response to regional factors such as climate or suc-cession, so the 'regional' s i g n a l w i l l show a response to these larger scale changes rather than to changes in environment r e l a t e d to the bog or pond. At present the vegetation surrounding the bog i s the regional P. a l b i c a u l i s and i t is reasonable to assume that t h i s vegetation has been 'regional' i n nature throughout the Hypsithermal. Hence 'regional' i s used i n the sense of vegetation unaffected by l o c a l changes in drainage. Zonation of the entire p o l l e n diagram may be achieved by v i s u a l i n -spection of the p o l l e n diagram i n F i g 6.4. However, Birks and Birks (1980) emphasise that use of a computer technique i s preferable for several reasons including the bias of the i n v e s t i g a t o r , r e p l i c a t i o n by other workers and the p o s s i b i l i t y of using several d i f f e r e n t routines to obtain the most consistent zonation. They also emphasise that computer zonation benefits the i n v e s t i -gator by requiring him or her to formulate c r i t e r i a e x p l i c i t l y before zonation as well as d i r e c t i n g attention to aggregate zones, rather than those of one or two species. 187 Regional species Local species Pinus  Picea Tsuga heterophylla Pseudotsuga Alnus Juniperus Epilobium Sorbus si t c h e n s i s Tsuga mertensiana Betula Populus Polypodiaceae Ribes Abies Pteridium Chamaecyparis Corylus Prunus emarginata, Chenopodiaceae/Amaranthaceae Quercus S e l a g i n e l l a selago  Shepherdia Cyperaceae Poaceae S a l i x T u b u l i f l o r a e Lycopodium  Potamogeton  Spiraea  Myrica Fungi Cyst I Cyst II Cyst III Cyst IV Schia-osporus Cyst B Cyst C Cyst F Cyst G Table 6.1 Separation of p o l l e n grains into regional and l o c a l components 188 In t h i s study UBC CGROUP (from the UBC computer l i b r a r y ) was used as an aid to zonation of the p o l l e n diagram. It i s a routine which uses p r i n c i p a l components analysis to c l u s t e r items with common attrib u t e s of d i f f e r e n t magnitudes. The items are s a t i s f a c t o r i l y c lustered when the error term shows a large increase, although the nature of a s i g n i f i c a n t increase must be de-termined judgementally. The program i s able to include contiguity con-s t r a i n t s , f o r c i n g only s p e c i f i e d adjacent items to c l u s t e r . This l a t t e r feature was not useful because three depth i n t e r v a l s (1.75-1.87m, 3-3.12m and 4.5-4.62m) had a palynoassemblages s u f f i c i e n t l y d i s s i m i l a r from those of the surrounding depth i n t e r v a l s that they formed a r t i f i c i a l zone bounda-r i e s . In Appendix E the ' l o c a l ' palyno-zonation dendrogram i l l u s t r a t e s the problem of sing l e d i s s i m i l a r samples f o r c i n g adjacent items to cl u s t e r at a late stage. A major problem with t h i s type of analysis i s that each species i s given equal importance i r r e s p e c t i v e of whether i t comprises 99% or 0.5% of the t o t a l s i g n a l . Any cor r e c t i v e measure, however, would have to take into ac-count the vast l y d i f f e r i n g p o l l e n p r o d u c t i v i t y and d i s p e r s a l mechanisms of a l l the various species included i n the model and has never been attempted in any other published analyses. These two l i m i t a t i o n s , those of the con-t i g u i t y constraint, and of s c a l i n g a l l readings, mean that the objective c l a s s i f i c a t i o n achieved by the program must be tempered by experience and common sense in a r r i v i n g at the f i n a l zonation. Appendix E shows two dendrograms, one recording the c l u s t e r i n g of ' l o -c a l ' p o l l e n types representing environmental changes on and f r i n g i n g the bog or pond; and the second showing the zonation from which 'regional' pollen types were in f e r r e d . The next sections consider the composition of zones recognised by the two c l u s t e r analyses, and the environmental and paleoclimatic inferences which can be drawn from t h i s information. 189 6.3.8 'Local' s i g n a l Five zones are recognised p a r t l y on the basis of the c l u s t e r analysis i n Appendix E and p a r t l y by examination of the p o l l e n diagram i n F i g 6 .4. The species considered to give a ' l o c a l ' environmental signal are l i s t e d i n Table 6.1. Approximate dates for each zone are obtained by i n t e r p o l a t i o n between the basal radiocarbon dates, the Mazama and Bridge River tephras, and the surface. The dendrogram ('local signal') i n Appendix E shows that Zones I and V are homogeneous with adjacent samples c l u s t e r i n g at early stages. Zone II and IV both have some homogeneous c l u s t e r s , while zone III samples are most heterogeneous, with many c l u s t e r i n g at a l a t e stage. This diagram i l l u s t r a t e s well the problem of a r t i f i c i a l zone boundaries created by the three anomalous samples, as these prevent otherwise s i m i l a r items from c l u s t e r i n g success-f u l l y . Consequently the f i n a l zone boundaries were drawn judgementally a f t e r reviewing the p o l l e n diagram and dendrogram. A description of the species in each zone follows: (i ) Zone I (6.12m to the base of the core, 7,790 to c i r c a 7,350 BP) This lowest zone i s characterised by r e l a t i v e l y few cysts except for a small peak towards the top of the zone. Most are cyst I which is well re-presented here i n r e l a t i o n to i t s frequency i n other zones. Potamogeton and Cyperaceae are continuously present as are small amounts of Myrica and Spiraea (both <1%). S a l i x does not appear u n t i l the top of the zone and neither does Poaceae (1). Poaceae (2) i s present at the base but disappears rapidly and i s not seen again u n t i l Zone I I I . Throughout the zone numbers of fungal spores are very low. 190 ( i i ) Zone II (6.12-5m, c i r c a 7,350, to c i r c a 6,880 BP) This zone shows an increase i n a l g a l cysts although numbers are s t i l l low i n comparison with Zone III and Schizosporis i s not present. Cyst I continues to be prominent with cyst II increasing i n importance while cyst G appears only at the base of th i s zone and the top of Zone I ( i t does not appear elsewhere i n the core). Cyperaceae, Potamogeton, Myrica and Spiraea are s t i l l present although i n small amounts (<1% except Cyperaceae). S a l i x and Poaceae (1) became well established i n t h i s zone and there i s a s i g n i f i c a n t increase i n fungal spores. ( i i i ) Zone III (2.67-5m, c i r c a 6,880 to c i r c a 4,170 BP) Zone III has large numbers of cysts of a l l types. The base of the zone coincides with the reappearance of Schizosporis and greatly increased num-bers of cyst 2 and cyst 4. Two of the three occurrences of cyst B are recorded in t h i s zone. Potamogeton decreases and Cyperaceae p o l l e n i s at the lowest l e v e l recorded (<2%). Poaceae (2) reappears and the amount of Sal i x ,  Spiraea and Myrica pollen increases (maxima 3-4%). Fungi become more impor-tant from the middle of t h i s zone. (iv) Zone IV (1.5 to 2.87m, c i r c a 4,170 to c i r c a 1,925 BP) Cysts were s t i l l f l o u r i s h i n g throughout t h i s period;, cysts 2,3,4 and Schizosporis are as abundant as i n the previous zone. However, cyst 1 de-cl i n e s greatly, while several specimens of cyst C are seen. At the top of th i s zone and at the base of zone V cyst F i s seen. Cyperaceae increase from 2.6% at the base of the zone to nearly 7 % at the top. T u b u l i f l o r a e increase s l i g h t l y and both Poaceae (1) and (2) are present. Myrica, Spiraea and Sa l i x are consistently present although with lower percentages (<2%) than in Zone II I . 191 (v) Zone V (1.5m to surface, c i r c a 1,925 to present) The upper zone i s marked by a decreasing number of cysts, p a r t i c u l a r l y cysts 2 and 3. Potamogeton reappears and Cyperaceae continue at high, a l b e i t f l u c t u a t i n g , levels (up to 8%). Poaceae (1), Myrica and Spiraea are only sp o r a d i c a l l y present, while the percentages of S a l i x are s i m i l a r to those found i n Zones III and IV. Poaceae (2) reach t h e i r highest value (4%) and T u b u l i f l o r a e also maintain the l e v e l achieved i n Zone IV. 6.3.9 Interpretation of the ' l o c a l ' s i g n a l The stratigraphy, absolute dating framework, palynomorph frequency and percentage loss on i g n i t i o n i n Zone I a l l i n d i c a t e that deposition at the base of the core was rapid. The upper boundary at 6.12m i s marked by changes i n pollen frequency and percentage loss on i g n i t i o n as well as by changes i n the l o c a l pollen s i g n a l . The stratigraphy, however, shows that a pond i n which g y t t j a was deposited existed above 6.75m i n the s t r a t i g r a p h i c column. Shallow water i n Zone I i s suggested by Cyperaceae and Potamogeton which both grow in boggy areas. Myrica and Spiraea, t y p i c a l pond edge species, have low frequencies in t h i s basal zone which indicates a bog interspersed with forest rather than an open body o f water. As both plants are no longer present, i t i s not possible to determine p r e c i s e l y which species were present, although the Spiraea was probably Spiraea d o u g l a s i i which currently fringes ponds at lower elevations (Mathewes and Rouse 1975, Lyons 1976). Gyttja deposition proceeded r a p i d l y u n t i l the Mazama tephra f a l l (2.75m in 1,200 years). However, Zone I I , despite forming r a p i d l y , has the highest percentage organic matter i n the core. This could be an i n d i c a t i o n of high p r o d u c t i v i t y due to r e l a t i v e l y high temperatures and a longer growing season during the late Hypsithermal. The l o c a l p o l l e n s i g n a l , however, suggests that l o c a l vegetation remained unchanged except for the a r r i v a l of S a l i x and Poaceae (1), so no change i n depositional environment can be i n f e r r e d . 192 A temporary drying out phase i s shown by roots i n the s t r a t i g r a p h i c 3 column and by high p o l l e n frequency (236x10 grains/g sediment) although not i by any changes i n the p o l l e n record. Although cysts b r i e f l y decline, t h i s episode i s not distinguishable, from many other abrupt declines in cyst num-ber. The change between Zones II and III provides evidence for a markedly d i f f e r e n t pond environment. More favourable conditions for a l g a l growth be-gin at t h i s boundary with most cysts f l o u r i s h i n g . As Zone III spans the cooling phase at the end of the Hypsithermal (see section 6.3.10), i t appears that decreasing temperatures did not reduce cyst growth, but that i t was instead enhanced by increased water depth. Increased pond area and depth are also i n f e r r e d from the decline of Cyperaceae which grows i n moist locations, pond edges, or standing shallow water, but not i n deeper water. Myrica,  Spiraea and Salix increase i n Zone III suggesting that they fringed the pond. Algal cysts continue to f l o u r i s h i n Zone IV. However, the appearance of two new cysts (C and F), the disappearance of cyst I, increasing Cyperaceae and T u b u l i f l o r a e , together with lower Spiraea and Myrica, imply either a shallowing of the pond and/or decreased area. The trend continued into Zone V, where the decline i n cysts and the increase i n Cyperaceae mirror the stratigraphy which shows a change from a pond to bog environment. Myrica and Spiraea disappeared although S a l i x remained, growing on d r i e r spots i n and around the bog. It i s l i k e l y that the disappearance of Myrica and Spiraea was c l i m a t i c a l l y induced as neither i s seen growing at these elevations at present. Poaceae and T u b u l i f l o r a e show that the present environment is much more open than that seen at the base of the core i n Zone I. Therefore, a l -though the cyst record suggests a succession from shallow to deep to shallow water, the pollen record shows that the i n i t i a l shallow phase was characterised by few pond edge and understory species implying a much smaller 193 c l e a r i n g i n the forest. Conversely, the higher Cyperaceae, Poaceae, S a l i x and T u b u l i f l o r a e counts seen i n the upper lm of the p o l l e n record, r e f l e c t present-day vegetation found on and around the bog. 6.3.10 The regional p o l l e n s i g n a l Zonation of the regional p o l l e n diagram for a d e s c r i p t i o n of regional vegetation changes u t i l i s e d UBC CGROUP; however, for reasons explained above (6.3.6) the contiguity constraint d i d not give s a t i s f a c t o r y r e s u l t s . Therefore, the unconstrained data were grouped by eye ( F i g 6.7) a f t e r exam-inin g the dendrogram i n Appendix E ('regional s i g n a l ' ) . This gave four d i s t i n c t groups. It can be seen that many samples within the same zone are grouped to-gether but that the grouping step does not improve the error s i g n i f i c a n t l y more than non-grouping steps. The more heterogeneous zones are II and III as many i n d i v i d u a l samples do not j o i n u n t i l l a t e r steps. These d i s s i m i l a r samples are grouped towards the r i g h t hand side of the dendrogram. A de-s c r i p t i o n of species i n each zone follows: ( i ) Zone I (5.75-7.12m, 7,790 to c i r c a 7,220 BP) The lower part of the core i s s p l i t into two subzones, l a (6.37m to the base) and lb (5 . 75-6.37m). Zone l a i s characterised by high pine (80-90%), high Chenopodiaceae and Shepherdia canadensis, together with a low Alnus percentage (1-8%), increasing throughout the zone. No deciduous trees were present except Betula (probably Betula p a p y r i f e r a ) . S e l a g i n e l l a selago was also present, being highest at the bottom of the core. In Zone l b , Pinus declines from over 80-90% to below 80% and Alnus increases to 7-10%. Chenopodiaceae and Shepherdia are present and a number of deciduous species appear (Populus, Quercus, Sorbus s itchens i s , Corylus). RED MOUNTAIN "Peat £J] C harcoa l j Gyttja Jfl] Gravel w o o d h " I S*f>0 a w R'li pollen x 10 /g 0 200 400 600 800 1000 F i g 6 . 7 P o l l e n diagram showing "regional" zonation i n Bog G 195 ( i i ) Zone II (4-5.75m c i r c a 7,220 to c i r c a 6,160 BP) This zone has the lowest values for Pinus (54-80%) and the highest values for Alnus and Betula (max 17%, most samples over 10%), Quercus and Corylus. Chamaecyparis f i r s t appears as do a s i g n i f i c a n t number of other herbs. S e l a g i n e l l a disappears halfway through t h i s zone and Polypodiaceae increase. ( i i i ) Zone III (2.37-4m c i r c a 6,160 to c i r c a 3,300 BP) This zone has f l u c t u a t i n g l e v e l s of Pinus p o l l e n (60-80%) with, le v e l s generally higher than i n Zone I I . Alnus shows a marked decline with a l l but one subsample showing values of less than 10%. Betula also decreases, while Quercus and Corylus disappear. Chamaecyparis and Sorbus, although present i n small amounts (<1%), are most prominent i n t h i s zone and Populus becomes more important. Shepherdia i s absent from both, Zones III and IV, despite i t s presence in the present day pond-fringing vegetation. Polypodiaceae are continually present and a few grains of Rosaceae/Cruciferae are seen. (iv) Zone IV (2.37m to surface, c i r c a 3,200 BP to present) In Zone IV Pinus reaches i t s highest levels (76-88%), while Alnus de-c l i n e s to i t s lowest value ( t y p i c a l l y <3%). Betula and Sorbus also become less frequent. Chamaecyparis i s not seen i n the top lm, while Polypodiaceae show a marked decline in the top of t h i s zone. Levels of Populus, however, remain the same as i n Zones II and I I I , while Juniperus i s present consist-ently from the bottom of the zone. 6.3.11 Interpretation of the regional s i g n a l The base of the core i s dominated by coniferous species (Pinus, Picea,  Abies and Pseudotsuga), with Pinus a l b i c a u l i s dominant as i n the contemporary forest. P. a l b i c a u l i s characterises the i n t e r i o r alpine forest whereas P.  contorta dominates the coastal alpine forest i n Hosie's map of pine d i s -t r i b u t i o n . Tsuga mertensiana i s also absent, implying that t h i s s i t e i s c l i m a t i c a l l y ' i n t e r i o r ' rather than 'coastal', which again suggests a 'con-t i n e n t a l ' forest type (Barnosky 1983). This species requires at least 60 196 f r o s t - f r e e days and a thick snowpack because i t s seedlings cannot survive extended periods i n frozen ground. S e l a g i n e l l a selago, a small vascular plant common i n l a t e - g l a c i a l deposits (Erdtman 1945), i s found near the base of the core. S. selago suggests a rocky substrate or mineral s o i l , an assertion supported by a low percentage loss on i g n i t i o n (6.3.3) and by a Shepherdia peak at 6.5-6.62m. Shepherdia i s insect p o l l i n a t e d and hence a very low p o l l e n producer with no a b i l i t y to disseminate widely, so we can i n f e r that a considerable amount of Shepherdia was present i n the forest understory and probably on the bog, as at t h i s time the bog i s currently fringed with Shepherdia even though no poll e n i s found i n the upper part of the core. Shepherdia has recently been recognised from the basal sediments of cores from a v a r i e t y of locations i n and around B r i t i s h Columbia (R. Mathewes personal communication, 1986; Mathewes 1973; R i t c h i e 1976; Mack et a l . 1978; Mott and Jackson 1982). Mathewes (1984) attri b u t e s i t s successful early c o l o n i s a t i o n of mineral s o i l to an a b i l i t y to f i x nitrogen and hence to colonise mineral substrates. The presence of Shepherdia provides strong supporting evidence for the assertion i n 6.3.8 that the base of the core represents a bog intermingled with forest rather than an open pond environment. The s i g n i f i c a n c e of th i s i n t e r p r e t a t i o n i s that Tauber (1965) showed t h e o r e t i c a l l y that such a s i t e would record only the l o c a l p o l l e n r a i n rather than that of the surrounding region. Consequently no c l i m a t i c i n t e r p r e t a t i o n i s l a i d on the very high P. a l b i c a u l i s percentage at the base of the core: i t i s interpreted only as a r e f l e c t i o n of a closed or nearly closed canopy p o l l e n measurement lacking a s i g n i f i c a n t regional s i g n a l . Shepherdia, Potamogeton and Cyperaceae decline throughout Zone lb and Myrica, Spiraea and S a l i x increase, a pattern interpreted i n sec-t i o n 6.3.6 as an increasingly open pond environment. I f t h i s i n t e r p r e t a t i o n is correct, and a body of open water expanded to f i l l the basin observed 197 today, then Tauber's estimates of 90% l o c a l and 10% regional p o l l e n for a bog 100-200m i n diameter would apply, although the mountainous t e r r a i n of the surrounding area suggests that these proportions give only a conceptual guideline as to the r e l a t i v e importance of l o c a l and regional s i g n a l s . Throughout t h i s basal zone, recycled Eocene palynomorphs were seen, doubtless the r e s u l t of the r e l a t i v e l y high mineral content i n the basal zone. Most were r e a d i l y distinguished from modern p o l l e n by t h e i r dark colour, p i t t e d appearance and glassy l u s t r e . Some, however, had to be d i f -f e rentiated morphologically from modern-day grains (see F i g 6.6). Throughout the core many species of the west coast forest (Picea,  Tsuga heterophylla, Tsuga mertensiana, Pseudotsuga) are complacent and give no i n d i c a t i o n of c l i m a t i c change. However, Pinus, the dominant tree i n the subalpine forest surrounding the bog, and Alnus, together provide a synchronous record of vegetation fluctuations i n the surrounding forest. As no Alnus i s found i n the area today, the species i s uncertain, although Alnus  incana (mountain alder) i s a l i k e l y candidate because i t i s more tolerant of cold than are other species. Fluctuations in Alnus are thought to record temperature changes rather than p r e c i p i t a t i o n trends as Alnus i s r e l a t i v e l y thermophilous and only ranges northwards to southeast Alaska, whereas i t i s found as far south as Mexico (Heusser 1974). Below 6.37m Pinus levels are r e l a t i v e l y high and Alnus r e l a t i v e l y low compared with Zone II, although much higher than i n the top 2m of the core. However, these percentages must be interpreted cautiously, as at the base of the core the p o l l e n spectrum i s that of the forest f l o o r (6.3.9) while i n Zone II the area was more open with a higher regional contribution. Therefore the higher Alnus percentage i n Zone II (when compared with Zone I) i s not interpreted as the consequence of a warming trend, but instead as the r e s u l t of an increasing proportion of p o l l e n blown i n from more distant 198 (low elevation) areas. In both zones conditions were warmer than today as would be expected from the late-Hypsithermal basal date. Corroborative evidence f o r warmer c l i m a t i c conditions i n Zone II i s found i n a s l i g h t Betula maximum and also i n the appearance of Quercus and Corylus. Corylus was a l a t e p o s t g l a c i a l a r r i v a l ; i t appears i n Surprise and Marion Lakes around 8,000 BP, with maximum counts around 6,000 BP (Mathewes 1973). At Pinecrest and Squeah i t peaked s l i g h t l y a f t e r the Mazama tephra f a l l although i t i s s t i l l present i n the present day p o l l e n spectrum (Mathewes and Rouse, 1975). Both Fishblue and Horseshoe Lakes record a l a t e r a r r i v a l , with the f i r s t p o l l e n immediately postdating the Mazama tephra (Mathewes, 1984). Further south, i n the Columbia Basin, Corylus has been present throughout the p o s t g l a c i a l , period (Barnosky 1985), but unfortunately i t i s not possible to trace northward migration as p o l l e n diagrams for north Washington do not include i t (Heusser 1973, Mack et a l . 1978a, 1978b, 1979). On the basis of other cores the appearance of Corylus i n th i s area would be predicted at between 8,000 BP and the Mazama tephra. The f i r s t grain i s seen at 6.37m which, assuming uniform sedimentation before the Mazama tephra, would give a date of approximately 7,300 BP corresponding well with the ob-served a r r i v a l date elsewhere i n B r i t i s h Columbia. Quercus was s i m i l a r l y tardy i n migrating to B r i t i s h Columbia. How-ever, unlike Corylus which i s s t i l l present i n the contemporary vegetation, i t i s now seen at only a few edaphically dry s i t e s north of the U.S.A. border, for example at Yale i n the Fraser Canyon, on Sumas Mountain i n the Fraser lowland and on Vancouver Island (Hosie 1969). Quercus was present i n the Fraser lowland before 6,000 BP (Mathewes 1973), although here i t peaked s l i g h t l y a f t e r the Mazama tephra. It i s only s p o r a d i c a l l y recorded at Pinecrest Lake in the Fraser Canyon, and does not figure at a l l i n pollen diagrams from Fishblue, Horseshoe, Phair or C h i l h i l Lakes, or i n Kelowna Bog. 199 Mack (1976, 1979) does not record Quercus from s i t e s i n northeast Washington, although Heusser (1960) uses small amounts of Quercus to zone Pangborn Bog i n northwest Washington, and Hansen and Easterbrook (1974) also record small percentages i n the Puget Lowland. On the basis of these studies the provenance of Quercus i s uncertain i n the study area because i t i s absent or unrecorded i n lake sediments at a l l other s i t e s north of Yale. However, i t s b r i e f appearance at Red Mountain corresponds with i t s maximum elsewhere i n B r i t i s h Columbia and Washington. Chamaecyparis appears halfway through the Zone II before 6,600 BP and t h i s corresponds with i t s appearance i n the Fraser Lowland and i n the Fraser Canyon. It does not, however, show such a large increase at t h i s high elevation s i t e as that seen i n the Fraser Lowland a f t e r the Mazama tephra, or l a t e r at Pinecrest and Squeah lakes ( c i r c a 3-4000 BP). A sudden temperature decline at the close of the Hypsithermal i s pro-posed on the strength of a marked decline i n Alnus and a s i g n i f i c a n t increase i n Pinus pollen (Zone II1/II c i r c a 6,160 BP), although le v e l s remain s i g -n i f i c a n t l y higher than those seen i n sediments deposited since the Bridge River tephra. Although moisture stress could account for t h i s pattern, t h i s p o s s i b i l i t y is rejected because moisture s t r s s s is not l i m i t i n g under current cl i m a t i c conditions above about 1,000m. Although Quercus and Corylus were no longer present i n Zone I I I , Betula showed only a s l i g h t decline and Sorbus peaked. Therefore t h i s zone i s e s s e n t i a l l y t r a n s i t i o n a l between Hypsithermal f l o r a and the impoverished assemblage now seen around the bog. Alnus and Betula declined again immediately below the Bridge River tephra and Pinus increased; t h i s probably marks the disappearance of Alnus from the v i c i n i t y of the bog due to temperature s t r e s s , as none i s currently present. The percentage of Polypodiaceae, which i s present throughout the record u n t i l the upper metre, decreased; no Polypodiaceae were seen at the study s i t e during sampling which suggests that they have become extinct at 200 the study area i n the past 1,000-2,000 years, again implying a temperature decrease. However, Polypodiaceae are overrepresented i n the p o l l e n record (Banner et a l . 1983), so i t i s u n l i k e l y that they ever formed an important part of the regional vegetation. Sphagnum, an important element i n much of the bog, i s not represented i n the palynoassemblage, probably because i t mainly reproduces asexually. Juniperus shows a marked increase i n Zone V. It i s found f r i n g i n g the bog and elsewhere on the earthflow where the canopy i s broken, growing as a spreading bush with branches close to the ground. The appearance of Juniperus at the bottom of Zone IV cannot be a t t r i b u t e d to d i f f e r e n t i a l preservation of peat and g y t t j a , as Mathewes (1973) recorded good preservation i n g y t t j a at Marion and Surprise Lakes, whereas Hansen (1947) and Heusser (1960) found poor preservation in peat. At t h i s s i t e Juniperus p o l l e n i s found mainly i n peat where conditions for preservation are least favourable. The presence of Juniperus h o r i z o n t a l i s i s interpreted as a r e s u l t of c l i m a t i c d e t e r i -oration as i t occupies exposed s i t e s where larger trees cannot compete suc-c e s s f u l l y . Populus i s also s l i g h t l y more abundant towards the top of the core. However, t h i s i s probably an a r t e f a c t of better preservation of the species i n peat than in g y t t j a (Sangster and Dale 1961), so no c l i m a t i c i n -ferences were made by those authors. The decline i n Alnus, Betula, Polypodiaceae and other species to-gether with the increase i n Pinus and Juniperus mean that present day vege-t a t i o n i s depauperate i n comparison with the past f l o r a and that many species present at the s i t e during the Hypsithermal are no longer able to t o l e r a t e present-day temperature and/or p r e c i p i t a t i o n and/or snowpack conditions. Consequently a second marked temperature decline, leading to present day c l i m a t i c conditions, i s indicated from the Zone III/IV boundary dated at c i r c a 3,300 BP. 201 6.3.12 A comparison of Bog G with other p o l l e n analyses i n B r i t i s h Columbia A comparison of the c l i m a t i c pattern suggested by th i s core with c l i -matic inferences from other nearby s i t e s such as Pinecrest, Squeah, Fishblue, Horseshoe, Phair and C h i l h i l Lakes ( F i g 6.1), reveals good agreement with the major trends. A l l these s i t e s show the onset of cooling before 6,600 BP, the Mazama tephra f a l l , and imply a l a t e r period of cooling, either sudden or continuous. King (1980) considers that f a i r l y dry conditions p e r s i s t e d u n t i l 6,100-6,200 BP near L i l l o o e t , while at Pinecrest and Squeah Mathewes and Rouse (1975) showed that the Hypsithermal, characterised by Gramineae, Artemisia and S e l a g i n e l l a peaks, terminated around 7,500 BP, although they caution that other factors such as s o i l maturation, topography and f i r e s could cause the observed vegetation patterns rather than c l i m a t i c change. At Fishblue and Horseshoe Lakes, xerothermic conditions p e r s i s t e d u n t i l 8,000 BP (Mathewes 1984). Further west, i n the Fraser lowland, Mathewes (1973) also placed the end of the Hypsithermal at 7,500 BP, using transfer functions (Mathewes and Heusser 1981) to i d e n t i f y that the warmest and d r i e s t conditions occurred between 10,000 and 7,500 BP. Hebda (1982) found an ear-l i e r Hypsithermal temperature decline i n the i n t e r i o r o f B r i t i s h Columbia at 6,000 BP, based on unpublished data from Finney Lake and Pemberton H i l l Lake. The Bog G core i s too young (7,790 BP) to record f u l l Hypsithermal conditions, but i t does show that r e l a t i v e l y favourable conditions persisted u n t i l c i r c a 6,160 BP at which time sudden cooling occurred. This corresponds to a t r a n s i t i o n at Horseshoe, Fishblue, Phair and C h i l h i l Lakes which are the s i t e s closest to Red Mountain although these also record slow cooling beginning at 8,500 to 7,000 BP. The Red Mountain core records a period of moderately cooler and wetter climate from the time of the Mazama tephra, or s h o r t l y thereafter, u n t i l the 202 Zone III/IV boundary, dated at 3,300 BP, when present day cooler, raoister conditions were attained following a sudden abrupt cooling. Although t h i s boundary f i t s i n well with many dated Neoglacial i c e advances i n the Coast Mountains and surrounding areas (6.2.2), evidence for comparatively sudden, synchronous and pronounced change at low elevation s i t e s i s lacking. Hebda (1982) considers that the i n t e r i o r grasslands attained present-day conditions around 4,000-4,500 BP, a date which corresponds well with King's L i l l o o e t chronology (4,400 BP to 4,100 BP). However, some evidence for a l a t e r cooling phase i s seen at Pinecrest and Squeah lakes, where Thuja/Chamaecyparis percentages increase around 3-4,000 BP. On the west coast, Hebda and Mathewes (1984) suggest that red cedar reached maximum Holocene abundance on the west coast during the past 2-3,000 years. Also, Heusser (1977) proposed a 1-2° decline i n mean July temperature i n this pe-r i o d from a tra n s f e r function applied to f o s s i l assemblages on the west coast of Washington. This s i t e , however, i s unique i n that a sudden temperature decline is recognised around 3,500 to 3,000 BP, marked by percentage changes i n several species, thereby providing a c o r r e l a t i o n between Neoglacial i c e advances and c l i m a t i c a l l y induced vegetational changes. 6.3.13 Summary Analysis of the Red Mountain core has made two main contributions to p o l l e n a n a l y t i c a l work in B r i t i s h Columbia. The f i r s t i s the use of d i f f e r e n t species to enable zonation for both l o c a l and regional vegetation changes. The second i s the r e a l i s a t i o n that Hypsithermal cooling occurred in two stages, with r e l a t i v e l y stable vegetation during the intervening periods. D i v i s i o n of p o l l e n observed i n the core into ' l o c a l ' ( i . e . bog and pond related species) and 'regional' (surrounding forest and long distance t r a v e l species) yields two d i f f e r e n t patterns of p o l l e n zonation when the data were analysed by computer grouping techniques. The ' l o c a l ' s i g n a l showed envi-ronmental change as the depression f i l l e d to give a lake, which then f i l l e d 203 with sediment and evolved into a bog, while the 'regional' sign a l recorded Hypsithermal and post-Hypsithermal vegetation changes brought about by cooling. Separation of the two records i s also useful i n that i t indicates how changes i n l o c a l environment can a f f e c t the regional p o l l e n s i g n a l . At the bottom of the core, the paucity of cysts and pond edge species and the r e l -a t i v e abundance of moisture loving herbs ( Potamogeton and Cyperaceae) to-gether with macrofossils and coarse sediment i n the s t r a t i g r a p h i c column, show that the depression had not f i l l e d with a great depth of water. Sca r c i t y of grasses, composites and other herbs suggests that there was l i t t l e open ground with forest very close to or intermingled with the bog. This information enabled the regional s i g n a l to be interpreted as that o r i g i n a t i n g from closed f o r e s t , or from a small c l e a r i n g ; t h i s environment would receive very l i t t l e p ollen i n f l u x from s i t e s beyond the immediately surrounding forest. Therefore the 'regional' p o l l e n assemblage at the bottom of the bog records l i t t l e p ollen t r a v e l l e d over long distances, which would include pollen from lower elevations. The second contribution of the Red Mountain core i s the recognition of a staged Hypsithermal cooling pattern. A sudden c l i m a t i c d e t e r i o r a t i o n i s seen sh o r t l y a f t e r the Mazama tephra, mainly i n the Pinus/Alnus s i g n a l , a pattern seen at several other s i t e s i n B r i t i s h Columbia. However, the most marked temperature decline i s seen before the Bridge River tephra f a l l at about 3,300 BP, with the disappearance of Alnus, the decline i n Betula and the a r r i v a l of Juniperus. This pattern provides strong evidence for a major cooling phase which i s poorly expressed i n p o l l e n diagrams from lower ele-vation s i t e s , but which has been indicated i n a large number of neoglacial ice advances. Information from the Red Mountain core i s used i n the next chapter to provide a record of p o s t g l a c i a l c l i m a t i c change within the study area. It 204 provides c l i m a t i c data to t e s t the hypothesis that episodic earthflow move-ment correlates with moist periods i n the p o s t g l a c i a l period. 205 CHAPTER 7 A CHRONOLOGY OF POSTGLACIAL EARTHFLOW MOVEMENT  7.1 Techniques for dating p o s t g l a c i a l earthflow movement 7.1.1 Introduction This chapter sets out to date p o s t g l a c i a l earthflow movement i n south-central B r i t i s h Columbia and to r e l a t e i t to c l i m a t i c f l u c t u a t i o n s discussed i n the previous chapter. In the f i r s t part techniques used to date earthflow movement both absolutely (radiocarbon, tephrochronology) and r e l a t i v e l y (carbonate accumulation) are outlined. The second part u t i l i s e s the tech-niques described.in section 7.1 to document a chronology of earthflow move-ment i n the study area. This was based on f i e l d observations, the i n t e r p r e t a t i o n of a i r photographs, s o i l weathering ind i c e s , tephrochronology and radiocarbon dating. The f i n a l part of the chapter draws together the c l i m a t i c and earthflow chronologies to examine whether earthflow movement has been affected by c l i m a t i c f l u c t u a t i o n s . 7.1.2 Radiocarbon dating Radiocarbon dating, developed by W.F. Libby i n the l a t e 1940's, i s the most widely used method of dating Quaternary events and materials. There are several problems which can l i m i t the use of t h i s technique. These are related either to unresolved t e c h n i c a l questions, or to sample contam-inat i o n . Technical problems include underestimation of the Libby h a l f l i f e (Goudie, 1984), i n d i v i d u a l laboratory bias (Stuiver, 1982), and errors i n -14 13 troduced by f l u c t u a t i n g C /C r a t i o s , although i t i s now common pract i c e 13 for dates to be reported a f t e r adjusting for the C c o r r e c t i o n . Sampling problems ar i s e e i t h e r from the i n c l u s i o n of 'new' carbon i n the sample, for example by modern root penetration, or else from contamination by 'old' carbon, often the r e s u l t of carbonate accumulation. Five samples from t h i s study were submitted for radiocarbon dating, from locations where the s t r a t i g r a p h i c s i g n i f i c a n c e was unequivocal. Fur-206 thermore, a l l samples were thought to be free from modern contamination and also from calcium carbonate (this l a t t e r assumption was confirmed by the processing l a b o r a t o r i e s ) . 7.1.3 Tephras i n the study area Summary maps presented i n Clague (1981) (adapted i n F i g 7.1) show that two p o s t g l a c i a l tephras cover, or p a r t i a l l y cover, the study area. These are the Mazama tephra, o r i g i n a t i n g from Mount Mazama (Crater Lake) i n Oregon and the Bridge River tephra from the Meager Mountain area of B r i t i s h Columbia. The Mazama tephra i s dated at 6,640 ±250 BP (Rubin and Alexander 1960) from charcoal buried i n pumice near the source area. Within the study area i t i s reported from Jesmond Bog (Nasmith et a l . 1967), from a roadcut i n the P a v i l i o n earthflow (Nadler 1984) and from an a l l u v i a l fan overridden by Drynoch earthflow (VanDine 1980) as well as from many other p o s t g l a c i a l a l l u v i a l fans i n the middle Fraser V a l l e y (Ryder 1970, 1971). G a l l i e (1984) reports that the Mazama tephra i s disseminated throughout several centimetres of p o s t g l a c i a l alpine e o l i a n deposits near Pemberton, where i t is also seen i n sediments from Horseshoe Lake. Sites further south i n the Fraser Canyon at Phair Lake, Fishblue Lake (Mathewes and Westgate 1980) and at Pinecrest and Squeah lakes north of Hope (Mathewes and Rouse 1975) a l l y i e l d Mazama tephra (See F i g 6.1 for l o c a t i o n s ) . It has been reported from Williams Lake 30 Km north of the study area, but i s absent at Nazho Bog, west of Quesnel, 70 Km north of the study area, (personal communication, J.Clague, 1987) . Many of the earthflows investigated i n t h i s study l i e outside the area i n which the presence of Mazama tephra has been established. Two samples of pale, f i n e , s i l t y material found i n Bogs B and G at Red Mountain, were sub-mitted for electron microprobe analysis at UBC. No re s u l t s are yet a v a i l -able. No other unreported occurrences of Mazama tephra were noted i n th i s study. 207 Bl •JESMOND •jfV/BIG BAR CREEK kB B B i g B a r Bl B l a c k d o m e ( 3 ) Bu B u r k h o l d e r ( 3 ) C C a n y o n ( 2 ) Ca C a n o e C r e e k Ch C h u r n C r e e k ( 6 ) Cm C a m o o C r e e k D D r y n o c h (2D F F l a p j a c k Ft F o u n t a i n G G r i n d e r Gb G i b b s Gi G i l l o n C r e e k H H e g i n b o t t o m ( 5 ) Ha H a t C r e e k L L o n e C a b i n LL L a c l a M e r (3D P P a v i l i o n ( 2 ) R R e d M o u n t a i n ( 7 ) T T u n n e l ( 5 ) Y Y a l a k o m ( 4 ) Yo Y o d e l ( 2 ) 9 2 0 0 4 0 0 K m Bu L £ 5* NCnr-^Ft 1 0 i 2 0 3 0 _ l K m 4 0 i 5 0 I 'LYTTON' Bridge River tephra l i m i t s per Mathewes and Westgate (1980) A Bridge River tephra l i m i t s within the study area (selected) ..... Mazama tephra l i m i t s per Clague (1981) • Mazama tephra occurrences within the study area (selected) ie 7 1 Tephra occurrences within the study area 208 Nasmith et a l . (1967) describe the d i s t r i b u t i o n , source and grain si z e of the Bridge River tephra. They record blocks of pumice up to 10cm i n d i -ameter near the source, a vent near Meager Mountain northwest of Pemberton and thick deposits (0.5m) i n the Bridge River Valley. East of the Fraser River, the tephra i s discontinuous and only seen i n depositional environ-ments. Mathewes and Westgate (1980) show that the tephra i s present i n Horseshoe, Fishblue and Phair Lakes which a l l l i e southeast of Nasmith's plume area. In t h i s study Bridge River tephra was found to the north of the es-tablished plume at the Grinder Creek, Heginbottom, Flapjack and Red Mountain earthflows. It was recognised i n the f i e l d as sand and s i l t - s i z e d pumice fragments often seen i n graded beds f i n i n g upwards. It was distinguished from Mazama tephra by the presence of sand-sized (or larger) tephra frag-ments . Bridge River tephra at Grinder Creek Flapjack, Red Mountain and Heginbottom was 2-4cm thick but at Big Dog Mountain i t attained 13cm on a bog at the top of earthflow Yalakom (B) . As Big Dog Mountain and Red Mountain are almost equidistant from the Meager-Plinth vent, these measurements show that the northward extension of the plume received considerably less tephra than did the main depositional area. Several dates have been obtained for the Bridge River tephra. Lowdon and Blake (1978) dated wood from a burnt stump at 2,500 ± 150 BP and as the tree was 150 years o l d i n f e r r e d that the explosion took place around 2,300 ± 50 BP. Maxima include 2,440 ± 140 BP (Nasmith et a l . 1967), 2,450 ± 130 (Lowdon and Blake 1976) and 2,685 ± 180 BP (Mathewes and Westgate 1980). Lowdon and Blake (1976) provide a minimum age of 2,240 ± 130 BP. 209 7.1.4 Use of r e l a t i v e age dating techniques for dating earthflows Many relative-age dating techniques have been used to estimate the age of Neoglacial moraines or older g l a c i a l deposits including lichenometry, various indices of boulder weathering and measures of s o i l development. Of these, s o i l development was the most promising technique since mature s o i l p r o f i l e s were lacking on active flows, implying that p r o f i l e s developed on s t a b i l i s e d earthflow deposits r e f l e c t e d t h e i r true age and did not include the e f f e c t s of pedogenesis p r i o r to earthflow movement. Carbonate accumulation was found to be the most successful measure of s o i l development i n the study area (Arkley, 1962). The p r i n c i p l e i s that during seasons when p r e c i p i t a t i o n exceeds evapotranspiration, moisture i s available to i n f i l t r a t e , the depth of i n f i l t r a t i o n being determined by s o i l texture. More moisture passes through the top of the leaching p r o f i l e than through layers at greater depth; consequently more carbonate leaching i s achieved higher i n the s o i l p r o f i l e and carbonate accumulation i s seen below the zone of leaching (Gile et a l 1966). CaCO^ accumulations were investigated at Grinder Creek (A) and (C), at Drynoch (A) and (B), and at P a v i l i o n . At a l l s i t e s , series of "en echelon" l a t e r a l deposits were selected for study. P i t s were dug 100-120cm deep (where possible) i n the crest of the l a t e r a l deposits and samples were taken every 10cm. In the laboratory the samples were screened and the a i r d r i e d <2mm fr a c t i o n analysed for CaCO^ content using the gravimetric method described i n Black (1965). The r a t i o of >2mm to <2mm c l a s t s was also tabulated, but no further grain s i z e determinations were made owing to the small sample s i z e necessitated by f i e l d l o g i s t i c s . Results for each s i t e are given i n section 7.2.8. 210 7.2 P o s t g l a c i a l earthflow movement. 7.2.1 Introduction The chronology of movement at i n d i v i d u a l earthflows w i l l be described i n d e t a i l i n t h i s section using whichever techniques outlined i n section 7.1 are appropriate to i n d i v i d u a l features. At some flows, absolute dating or absolute maxima and minima are a v a i l a b l e , but on most, only r e l a t i v e dating information was obtained. Although r e l a t i v e ages i n themselves are not useful i n c l i m a t i c c o r r e l a t i o n s , they are h e l p f u l i n that comparison with dated flows may enable a probable or l i k e l y age to be assigned. 7.2.2 Big Bar Creek earthflow Bovis (1985) recognised s i x phases of the Big Bar earthflow (Fig 7.2). He noted that Phase 5 began sho r t l y a f t e r deposition of the Bridge River tephra as sorted a s h f a l l deposits were overridden by the earthflow. Two tephra-bearing sections are marked on his map, but many add i t i o n a l sections were seen by the author i n gulleys d i s s e c t i n g Phase 5. Phase 6 i s , according to Bovis, s t i l l active, although t h i s was not confirmed during f i e l d obser-vations by the author. Phase 1 was not dated. However, the configuration of r i v e r terraces opposite the flow suggests two phases of movement when the. Fraser River flowed at a higher l e v e l . The arcuate r i v e r terrace scars opposite the flow can be seen on the a i r photographs (Appendix A) and also on manuscript map EP10 (Fig 7.3). They occur at about 350m and 300m on EP10, 60m and 12m re-spectively above present r i v e r l e v e l . A terrace remnant matching the 350m le v e l is truncated by the flow at 'T' on F i g 7.3, and phase 1 o v e r l i e s (and hence postdates) the 12m terrace. F i g 7.4 shows the toe section where the flow overrides terrace material. Phase 5 at Big Bar provides a minimum date for the 12m terrace as Bridge River tephra i n the shear plane dates landslide movement which cuts through t h i s feature. F i g 7.2 Bovis (1985) i n t e r p r e t a t i o n of movement phases at Big Bar (reproduced with permission) 212 A Scarps cut in terrace r i s e r s opposite Big Bar Creek earthflow F i g 7.3 Extract from may EP10 sheet 7, B r i t i s h Columbia 'Department of Lands and Forests, Water Rights Branch, accompanying the report on flood control at Moran Canyon, Fraser River 213 The chronology of p o s t g l a c i a l Fraser River i n c i s i o n i s not firmly es-tablished. Ryder and Church (1986) dated a sequence of four terraces at L i l l o o e t , and t h e i r T2 (68m) and T4/L4 (8m) are clos e s t to the elevation of the landslide scars. A section with Mazama tephra gave a minimum of 6,700 BP for t h e i r highest terrace (103m); other minima for t h i s terrace include 7,000 BP for a 40m terrace near Spatsum (Ryder, 1976) and 7,500 BP for a 40m terrace at Drynoch (VanDine, 1980). The T4 terrace i s assigned a minimum date of c i r c a 1200 BP on the basis of buried Indian a r t e f a c t s . However, l a n d s l i d i n g downstream from L i l l o o e t at Texas Creek may have ponded back Fraser River as far upstream as L i l l o o e t , thereby creating l o c a l baselevels unconnected to regional i n c i s i o n . The r e l a t i v e ages of Phases 1 and 5 of over 6,700 BP and 2,300 BP or younger are r e f l e c t e d q u a l i t a t i v e l y i n the degree of d i s s e c t i o n i n the lower part of the flow. Phase 1 has a rounded, hummocky surface with tension crack lineations widely spaced. The l a t e r a l deposits have rounded crests and no steep faces ( a l l l a t e r a l s may be ascended without d i f f i c u l t y ) . Phases 2 and 3 have s u r f i c i a l textures s i m i l a r to Phase 1 and by implication are contemporaneous with Phase 1 rather than with the younger Phase 5. Phase 5, in contrast, has a c l o s e l y spaced pattern of tension cracks throughout i t s length but no rounded hummocks and depressions. There are no l a t e r a l s asso-ciated with Phase 5 with which those of Phase 1 may be compared. In summary, the evidence of r i v e r terrace scars suggests that the Big Bar earthflow was active during the ea r l y p o s t g l a c i a l period (before 7,000 B P ) , possibly d e s t a b i l i s e d by downcutting by Fraser River. Phases 1,2 and 3 date from t h i s early period. The age of Phase 4 was not determined, but Phase 5 took place s h o r t l y a f t e r the Bridge River tephra and Phase 6 may be recent. 214 F i g 7 .4 Phase I of Big Bar earthflow overriding Fraser r i v e r terraces 215 7.2.3 Black Dome group of earthflows Two of the three flows i n t h i s group, (B) and (C) both appear to have had only one movement phase. Flow (A), whose r e a c t i v a t i o n was described i n Chapter 4, had two e a r l i e r phases of movement. The f i r s t excavated the slump bowl which extends upslope to the ridge crest, while the second phase i n -volved a slump further down the flow, whose headscarp ponded back an ephemeral lake. 7.2.4 Burkholder and Lac La Mer earthflows These f i v e flows (Burkholder (A and B), Lac La Mer (A,B and C) occur close together and so are considered as a group ( F i g 1.7). Burkholder (A and B) are both active, but there i s evidence for e a r l i e r movement i n well forested, older deposits downslope from the active flow. Well rounded l a t -erals can also be seen. Lac La Mer (A) and (B) have only one phase of move-ment. However, Lac La Mer (C) has at least two phases as indicated on the photograph i n Appendix A. The chronology of these flows i s uncertain. They were not v i s i t e d , so no d i r e c t evidence i s ava i l a b l e . The lack of vegetation on Lac La Mer (A), when compared with the active parts of Burkholder (A and B) indicates r e l a -t i v e l y recent movement. However, on the other Lac La Mer flows and on the old Burkholder deposits, thick forest precludes an analysis of d i s s e c t i o n . 7.2.5 Camoo earthflow The toe of t h i s flow has undergone movement during p o s t g l a c i a l time. This i s shown by g l a c i o f l u v i a l deposits c l i n g i n g i n discontinuous patches to the steep c l i f f where Bridge River had eroded the toe (Fig 7.5). Movement is related to a slump at the toe of the flow rather than to movement of the entire earthflow. The main body of the earthflow has a t h i n and discontinuous t i l l cover and few l a t e r a l deposits were seen, while the few that are present have 216 F i g 7.5 G l a c i o f l u v i a l deposits on toe of Camoo Creek earthflow 217 l i t t l e topographic expression. No tension crack patterns, or chaotic de-pressions and raised blocks were seen on a i r photographs or on the earthflow. No conclusions were drawn about the probable age of t h i s flow. 7.2.6 Canoe Creek earthflow The flow has two phases seen i n the photograph i n Appendix A, the f i r s t of which formed a slump bowl at 670-750m and the second an arcuate scar where the flow narrows. L a t e r a l deposits are poorly defined and the flow terminates in a subdued snout at 600m, 270m above present r i v e r l e v e l . The slump bowl is bisected by a 30m deep i n c i s i o n by a t r i b u t a r y of Fraser River which i s thought to postdate the flow as i t s course i s unaffected by the flow morphology. The entire flow i s covered with a t h i n (<lm i n most places) t i l l layer. The flow surface has extremely subdued topography (see Appendix A). No open tension crack patterns are v i s i b l e , and the l a t e r a l deposits are rounded. Indeed, the flow surface has considerably less r e l i e f and d e t a i l than the flow at Big Bar Creek, whose older phases are estimated to date from 7,000 BP or e a r l i e r . The Canoe Creek flow disrupts g l a c i a l lake sediments at 670 - 700m, sediments presumed to be contemporaneous with those at a s i m i l a r elevation 7 km to the south where they are exposed in the headwall of Grinder Creek (A) . Consequently disturbances of g l a c i o l a c u s t r i n e sediments imply a p o s t g l a c i a l date for flow movement ( i . e . post c i r c a 10,500 BP). There are several small depressions in the headscarp of the flow. A test p i t was dug in the largest, but no stratigraphy was v i s i b l e . Therefore there is l i t t l e p o t e n t i a l for obtaining further chronologic evidence from th i s flow. 218 7.2.7 Churn Creek earthflow Bovis (1985) mapped 6 phases of movement at the largest Churn Creek flow (A) on the basis of morphological evidence. He described translocated and sheared t i l l , and t h i s writer observed t i l t e d and sheared r i v e r terrace ma-t e r i a l i n the lower part of the flow, below 450m. These observations imply p o s t g l a c i a l movement a f t e r at least the upper r i v e r terraces were formed. Recent movement of phase 6 has been described i n section 4.2.3. 7.2.8 Drynoch (A) earthflow Much of the Drynoch (A) flow i s cu r r e n t l y active with measurable movement (see section 4.2.4). However, paired sets of l a t e r a l s stranded up to 30m above the active flow attest to considerable drawdown since the flow f i r s t formed. The outer l a t e r a l deposits have a large t i l l component, par-t i c u l a r l y i n the upper part of the flow, while the inner l a t e r a l s comprise mainly sedimentary or volcanic rocks. VanDine (1980) described a gravel p i t section at the toe of the flow (reproduced i n F i g 7.7) where the outer l a t e r a l s override f l u v i a l deposits, eolian sand and f l u v i a l sand and gravels. The s t r a t i g r a p h i c i n t e r p r e t a t i o n given by VanDine i s that the lower f l u v i a l material was l a i d down by the Thompson River in the early p o s t g l a c i a l . During the Hypsithermal eolian sand covered the r i v e r terrace and t h i s was subsequently o v e r l a i n by poorly sorted f l u v i a l deposits of the Squianny Creek fan. Squianny Creek was l a t e r diverted by the flow and then by drainage works and now follows the north margin (Fig 7.7). Mazama tephra i n the fan deposits implies that the e a r l i e s t earthflow phase to reach the Thompson Valley axis s t r a t i g r a p h i c a l l y o v e r l i e s the Squianny Creek deposits, and therefore postdated 6,600 BP. A minimum date for i n i t i a t i o n i s provided by roots i n the s l i d e shear plane uncovered during road b u i l d i n g which were dated at 3,175 ± 150 BP (VanDine 1983). A l a t e r date (900 ± 50 BP) on charcoal found i n an abandoned Indian f i r e p i t (see Earthflow material _ _ Mazama tephra 6,600 BP^ 0  v ^7 ' o C 2 O O O Poorly sorted and washed gravel i n sandy matrix; Squianny Creek deposits E o l i a n sand containing c u l t u r a l implements dated at 7,530 BP F l u v i a l sand and gravel terrace F i g 7.6 Diagram of toe section of Drynoch (A) earthflow a f t e r VanDine (1980) r-" VO 220 O ISO SOO r>\<-1 r- * 5 F i g 7.7 Map of Drynoch (A) showing lo c a t i o n of s o i l p i t s 221 F i g 7.7 for location) indicates that the middle portion of the flow has not been greatly disturbed since that time. However, movement along basal and marginal shear zones, r a f t i n g a c e n t r a l plug, may have occurred without disturbing the archaeological s i t e s on the plug. Two sequences of p a r a l l e l , paired l a t e r a l deposits were selected to investigate whether r e l a t i v e dating using CaCO^ accumulation could be per-formed (Fig 7.7, s i t e s 1 and 2). Three a d d i t i o n a l samples from the oldest l a t e r a l (to which a range of absolute dates has been assigned above) were also taken. The s i t e s which were used, and the r e l a t i v e positions of the paired l a t e r a l s , are shown i n F i g 7.7. A lm deep p i t was dug at each sampling location and CaCO^ content determined every 10 cm (see section 7.1.4). CaCO^ content and l i t h o l o g y for each are shown i n F i g 7.8a to c. Two d u p l i -cate samples were taken from each p i t to ensure that the samples were uni-form. At a l l s i t e s except D16-D20 and D23 the CaCO^content i n duplicate samples lay within ±10%, while at the remaining s i t e s ( a l l with sedimentary parent material) the difference was within ±25%. Greater v a r i a b i l i t y was attributed to small c a l c i t e fragments. Three s o i l p i t s (D6, D12 and D27, the l a t t e r an a d d i t i o n a l p i t not shown on the cross p r o f i l e s ) in the outermost l a t e r a l s were dug in t i l l which contained a preponderance of fragments o r i g i n a t i n g from the igneous and metamorphic Coast Mountains complex. The majority of c l a s t s were green metamorphics. Neither D6 nor D12 showed a s i g n i f i c a n t CaCO^ horizon, while at D27 a well defined CaCO^ horizon at 20-30cm was measured. As t h i s type of parent material was not seen elsewhere on the flow, and the leaching/evapotranspiration processes did not lead to horizonation compara-ble with that seen on sedimentary parent material, no further consideration was given to these s i t e s . Other p i t s on the outermost l a t e r a l s (dated by VanDine (1983) at 3-6,000 BP) were D28•(not shown on the cross s e c t i o n s ) , D21 and D3. A l l Fig 7.8a Calcium carbonate percentage in Drynoch (A) lateral deposits (Site 1) F i g 7.8b Calcium carbonate percentage i n Drynoch (A) l a t e r a l deposits (Site 2) F i g 7.8c Calcium carbonate percentage i n Drynoch (A) l a t e r a l deposits (additional samples from outer l a t e r a l ) to 4> 225 showed CaCOg depletion i n the top 60cms and CaCO^ concentration at 60-90cm. Horizonation i n D3 was s i m i l a r to that seen i n D5, a p i t dug i n s i m i l a r parent material on adjacent stable t e r r a i n . It was concluded that the 3,000 to 6,000 years which elapsed since Drynoch (A) f i r s t moved are s u f f i c i e n t time for the formation of a CaCO^ horizon i n t h i s area. At the upper sampling s i t e ( s i t e 1) a l l the inner l a t e r a l s (D7 to D13) except D10 showed si m i l a r CaCO^ depletion i n the top 60cms and concentration at 60-90 cms. This was taken as evidence that the upper portion of the flow had not been active recently and, taken i n conjunction with VanDine's 900 BP f i r e p l a c e date, suggests that the upper part of the flow may have been stable for at least 1,000 years. An older date was corroborated by the lack of knife edge l a t e r a l ridges The upper portion of the flow i s now marginally active as some transverse tension cracks can be seen on the a i r photographs. The lower sampling s i t e ( s i t e 2) showed more va r i a b l e CaCO^ accumu-l a t i o n . CaCO^ accumulation i n the outer p i t s (at D21, D5 and D3) has already been described. Three of the innermost l a t e r a l s (D24 to D26) show l i t t l e CaCO^ horizonation, while D20, D22 and D23 on the south side of the flow, show depletion i n the top 20-30cm but no clear pattern of accumulation. The implication is that the innermost l a t e r a l s were formed r e l a t i v e l y recently although beyond the timespan of a i r photographs, the middle l a t e r a l s on the south side at an interim date, and the outermost l a t e r a l s between 3,000 and 6,000 BP. Continued or sporadic movement throughout t h i s period corresponds with contemporary movement at t h i s s i t e and the knife edge cross p r o f i l e of the innermost ridges i n comparison with the rounded appearance of the outer l a t e r a l s provides morphological evidence for a r e l a t i v e age difference. In summary, Drynoch i s known to have f i r s t moved i n the period 3,000 to 6,000 BP. This early phase i s marked by the outer rounded l a t e r a l s , up to 30m above the active flow and with a large component of t i l l . A second phase may be recorded i n a serie s of l a t e r a l s at the lower sampling s i t e 226 which were most c l e a r l y seen on early (1960) a i r photographs. Recent, a l -though not h i s t o r i c , movement with associated drawdown i s recorded by young, steep sided and unleached l a t e r a l s at the lower sampling s i t e s . 7.2.9 Drynoch (B) earthflow No absolute s t r a t i g r a p h i c information i s a v a i l a b l e from t h i s s i t e . However, evaporite deposits (Fig 7.9) i n lenses up to 10cm t h i c k were over-ridden by the i n i t i a l earthflow movement. These may be seen i n two wide tension cracks on the west arm of the flow. No evaporite i s seen forming at t h i s s i t e under present day c l i m a t i c conditions, so formation under warmer and d r i e r Hypsithermal c l i m a t i c conditions i s proposed based on s i m i l a r de-posits at P a v i l i o n (see 7.2.15). Therefore the flow must have overridden these deposits sometime a f t e r the close of the Hypsithermal ( i . e . after about 6,000 BP). Two sets of l a t e r a l deposits were sampled on the west arm to i n v e s t i -gate CaCO^ content (see F i g 7.10). The outer two (D15 and D17) showed CaCOg depletion i n the upper 40-60cm of the s o i l horizon followed by a well developed 20cm CaCO^ accumulation zone. The inner p a i r showed uniform CaCO^ content with depth, which was reasonable as they are currently forming i n response to the recent (1950 onwards) r e a c t i v a t i o n of the west arm (see Chapter 5). The outer set of l a t e r a l s were thought to be contemporaneous with the inactive lower part of the flow, where age was i n f e r r e d from the overridden evaporite deposits. S o i l p i t s sampled from paired l a t e r a l s on the east arm show some CaCO^ accumulation, i n d i c a t i n g that they are younger then the outer l a t e r a l s of the west arm. However, ground inspection revealed that the lower ha l f of t h i s arm probably accreted by overriding older deposits: hence the age of the l a t e r a l s i s probably less than that of the oldest parts of the east arm. 227 Fi g 7.9 Evaporite overridden by early Drynoch (B) earthflow movement East arm West arm 0 1 2 3 4 5 6 > II I n , . I 1 * 1 % Calcium carbonate by weight (unless indicated by a different scale) F i g 7 . 1C Calcium carbonate percentage in Drynoch (B) lateral deposits 00 229 7.2.10 Flapjack earthflow The a e r i a l photographs i n Appendix A show that t h i s flow had only one active phase. The snout of the flow traversed a v a l l e y d i v i d i n g a large lake into two unequal parts (Fig 7.11a). On the east side of the snout two shallow and small ponds connecting with the large lake to the east were formed. Three cores were extracted from these ponds using ABS piping f i l e d to form a sharp cuttin g edge: the pipe was forced into the basal sediments with a sledge hammer and extracted by hand. The cores were frozen and then the ABS piping was sawn l o n g i t u d i n a l l y and the core thawed, and cut to investigate the stratigraphy. F i g 7.11a gives the stratigraphy from one of the three cores (the second was s i m i l a r and the t h i r d d i d not extend to basal t i l l or landslide deposits). Bridge River tephra was found approximately halfway down the st r a t i g r a p h i c column at 26cm i n both complete cores. The lake (christened Beaver Lake) to the west of the Flapjack flow was also cored to investigate whether any s t r a t i g r a p h i c c o r r e l a t i o n s could be made. This lake was 2-3m deep, compared with l-2m for the small lakes which were cored (Fig 7.11b). A comparison of the two cores shows that the small Flapjack Lake lacks the pale cream g y t t j a and sn a i l s h e l l s which characterise the lower 2m of the other core. Only the top l-2m i s comparable and i t i s clear that sediment accumulated much more r a p i d l y at Beaver Lake i n the period before the Bridge River tephra. The implication i s that the Flapjack flow took place after the r e l a t i v e l y warm Hypsithermal, when rapid g y t t j a accumulation was re-corded i n other lake sediments and before the Bridge River tephra. Therefore, a maximum age is assigned as around 6,000 BP (the close of the Hypsithermal) and an absolute minimum of 2,400 BP. A basal date of 1,200 ± 80 BP (GSC-4149) was obtained for a basal sample of wood. This was rejected as the wood was found below the Bridge River tephra (2,400 BP). It i s thought that the large wood fragment was 230 Fig 7.11a Core from Flapjack earthflow 20- % 10 4) U 40« •P o s t-l •P a 4) Very fibrous, partly decomposed, mid to dark brown organic sediment with many large wood fragments Bridge River tephra Dull grey-green fine s i l t to very fine sand, coarser to base Dark brown/black gyttja with some fibres 60. 80-T i l l . Clay to large rounded clasts with long axes up to 5cm Clasts are mainly Coast Mountains Plutonic and weather green Schematic diagram of sampling sites V Cores 231 F i g 7.11|b Core from Beaver Lake 100 -200 -300 -Uniform black g y t t j a with f i n e roots Bridge River tephra Uniform g y t t j a with few f i n e roots Blocky black g y t t j a ( s t r u c t u r e destroyed by freezing) Organic layer - very f i b r o u s Blue green layer Dark brown g y t t j a with some roots Dark brown g y t t j a with many roots Buff/brown g y t t j a with many s h e l l fragments Marl with s n a i l s h e l l s i n t e r s p e r s e d with black sand and gravel lenses Gravel and large wood pieces Brown sedge Buff/brou.n marl with many s h e l l fragments 232 caught on the leading edge of the corer and thus moved to a lower s t r a t i g r a p h i c p o s i t i o n . 7.2.11 Fountain, Gibbs Creek, G i l l o n Creek, Kettlebrook Creek and Tunnel  earthflows No work was car r i e d out at these s i t e s . Bovis (1985) described the chronology of earthflow at these s i t e s . At Gibbs Creek he recognised three phases of movement. A l l o v e r l i e debris fan and f l u v i a l deposits l a i d down in the f i r s t 3,500 years or less of p o s t g l a c i a l time (Ryder 1976) and eolian s i l t containing artefacts dated at 4,730 ± 380 BP (J. Ryder quoted i n Bovis 1985). Therefore a l l three phases must postdate 4,730 BP. Bovis recognised three s i m i l a r phases at the Tunnel flow and on the basis that they o v e r l i e eolian s i l t they were also dated as post Hypsithermal. A small area of the Gibbs Creek earthflow i s s t i l l moving, but the bulk of a l l three flows ap-pears to be stable at present. This i s due, i n a large part, to drainage diss e c t i o n which prevents pore pressures from b u i l d i n g up at depth (personal communication, M. Bovis, 1987). 7.2.12 Grinder Creek earthflow Repeated movement has taken place at Grinder Creek (A) leaving up to fi v e pairs of en echelon l a t e r a l deposits. A s o i l p i t dug in the highest l a t e r a l on the Grinder Creek (A) south arm revealed a continuous band of sheared but sorted Bridge River tephra up to 3cm th i c k (Fig 7.12). The tephra implies that the e a r l i e s t movement at Grinder Creek took place at th i s time. Extensive excavations i n l a t e r a l cutoff depressions and a carbonate analysis of s o i l p r o f i l e s did not y i e l d any further s t r a t i g r a p h i c information. As the south arm of flow A overrides the north l a t e r a l s of Grinder Creek (C) near Fraser River, Grinder Creek (C) must predate 2,400 BP. As the la t e r a l s on flow (C) are degraded the flow i s probably considerably older than Grinder Creek (A). No dateable information was extracted from the flows Grinder Creek B,D or E. 233 F i g 7.12 Bridge River tephra i n former earthflow shear plane exposed i n a s o i l p i t at Grinder Creek (A) 234 GRINDER CREEK E a r l y p h a s e ( s ) | ( L a t e r phase(s) j — | A c t i v e p a r t s o f the f low B r i d g e R i v e r Ash ijL P i t 3.5m deep F i g 7.13/ Movement phases at Grinder Creek (A) 235 7.2.13 Hat Creek earthflow Neither Bovis (1985) or Rawlings (1984) off e r e d any s t r a t i g r a p h i c i n -formation from t h i s s i t e , although Bovis stated that the flow has been active i n p o s t g l a c i a l times and Rawlings (1984) measured movement on the upper part of the flow. Hebda's report on Finney Lake (personal communication, G. Rouse, 1985) has some i n c i d e n t a l information about Aleece Lake, the largest lake on Bovis' (1985) 'dormant area'. In h i s b r i e f d e s c r i p t i o n he refers to the 'post Mazama' part of the record thereby implying that the lake predated the Mazama tephra and hence that the s t a b i l i s e d part of the. flow i s older than 6,600 BP. This i n t e r p r e t a t i o n i s corroborated by the lack of tension crack d e t a i l and rounded topography of the flow. 7.2.14 Heginbottom earthflow The Heginbottom earthflow comprises at least three movement phases as shown i n F i g 7.14. This i n t e r p r e t a t i o n i s based on d i s r u p t i o n of kame t e r -races, the r e l i e f of the surface of d i f f e r e n t parts of the flow and the presence of Bridge River tephra (2,400 BP) in depressions on the flow, i n marginal depressions and i n sections cut by an ephemeral creek which cuts the east margin of the flow. The locations of a l l c r i t i c a l sections are given i n F i g 7.14. A large area on the east side of the toe of the earthflow has a subdued but hummocky topography and eroded.laterals, comparable with early phases at the Big Bar, Churn, Fraser and Hat Creek flows (see photographs in Ap-pendix A). It i s also at a higher elevation (10 - 20m) than the active part of the flow. At the toe of t h i s i n a c t i v e part of the flow a r i v e r terrace is cut in earthflow material 5m above the current l e v e l of Lone Cabin Creek. This implies that the flow took place before the creek had attained i t s current base l e v e l . Both morphology and possibly base l e v e l suggest an early p o s t g l a c i a l date for the higher (and older) part of the flow (see phases i n F i g 7.14). 236 Excavations i n two closed depressions (both at #1 i n F i g 7.14) on these older deposits revealed s o i l horizons buried by 6-7cm of Bridge River tephra shards ( s t r a t i g r a p h i c d e t a i l s are given i n Appendix K), confirming that the s t a b i l i s e d area predates 2,400 BP. A marginal depression cut o f f by a stranded l a t e r a l (#2 i n F i g 7.14) to the west of the east lobe was excavated. It revealed a buried s o i l with many charcoal pieces underlying a sorted layer of Bridge River tephra shards. This layer was o v e r l a i n by a t h i n organic layer which also contained charcoal, followed by unsorted shards together with pebbles from the s i l t s t o n e to conglomerate sedimentary succession which comprises the east arm of the flow. The upper layer was interpreted as slopewash (see Appendix K for a detai l e d d e s c r i p t i o n ) . Bridge River shards were found at a comparable marginal location (#3 i n F i g 7.14) on the west lobe and i n a depression near the head of the east lobe (#4 in F i g 7.14), impying that these areas have probably been stable since 2,400 B P . A buried s o i l (Fig 7.15, #5 i n F i g 7.14) was seen i n a stream cut section on the east side of the east lobe, o v e r l a i n by f l u v i a l and earthflow deposits dating from the damming of the stream by earthflow movement. The s o i l had a well developed CaCO^ horizon and a f a i n t l y brown/grey A horizon. No dateable material was found. The stream which flows down the east side of the flow was diverted by a series of arcuate slumps which can s t i l l be seen on the a i r photographs i n Appendix A. Three segments (#6 to #8 i n F i g 7.14) of the present incised stream channel show s t r a t i f i e d f l u v i a l deposits separated l a t e r a l l y by sections cut in earthflow material. In the lower sections (#6 and #7 i n F i g 7.14), Bridge River tephra i s seen near the base (Fig 7.16) implying diver-sion shortly before 2,400 B P . The tephra can be traced l a t e r a l l y at #6 for 237 Phase 1 See text for a de s c r i p t i o n Phase 2 of numbered locations Currently active Age unknown • • ? F i g 7.14 Heginbottom earthflow movement phases 239 20m. A sample of charcoal from 75cm below the top of t h i s section yielded a date of 1,000 ± 50 (GSC-4157). The implication of these sections i s that backward rotation of earthflow slump blocks diverted the marginal stream sh o r t l y before 2,400 BP to some time a f t e r 1,000 ± 50 BP. At t h i s time the earthflow reactivated, changing l o c a l base l e v e l and the creek was able to cut down through sediments accumulated during the past 1500 years. The sediments are not t i l t e d on the earthflow side of the section and match those on stable ground. Consequently the area adjacent to the stream must have been stable since before 2,400 BP. Much of the earthflow i s currently active (Fig 7.14). Several lobes, where t i l t e d trees were observed, have moved since 1800 A.D.. However, the dry lower part of the flow i s also thought to be act i v e , although no marginal shear zones were seen. Two mid-flow depressions were excavated (#7 and #10 in F i g 7.14) but no vestiges of Bridge River tephra were found, implying at least 3m and lm, respectively, of sedimentation since 2,400 BP. No shards were found i n any other smaller depressions on the lower part of the flow. In summary, at least three movement phases are recognised at Heginbottom (see F i g 7.14). The f i r s t phase(s) or.cnrred before 2,400 BP and may date from the early p o s t g l a c i a l as suggested by morphologic evidence. Subdued deposits to the east of the active flow as well as elevated l a t e r a l s on the west side of the flow were formed during t h i s early phase. A second protracted movement phase was marked by ponding of the east marginal stream shortly before 2,400 BP. A t h i r d phase, s t a r t i n g some time a f t e r 1,000 BP, marks a further r e a c t i v a t i o n . Present-day movement of di s c r e t e lobes as well as continuous slow movement of much of the lower part of the flow i s seen. 241 7.2.15 Pav i l i o n (A) earthflow Bovis (1985) recognised four phases of movement at t h i s s i t e (Fig 7.17). As at Drynoch (A) the outermost l a t e r a l deposits are t i l l covered while much of the inner part of the flow i s composed of a range of Cretaceous sediments (see Bovis (1985) for d e t a i l s ) . The e a r l i e s t movement phase b u i l t l a t e r a l s and fan deposits comprising t i l l for the most part. No r e l a t i v e or absolute evidence dating t h i s phase was found save that i t must predate phase 2 which occupies the crater at the head of the flow. Two depressions were selected for coring to e s t a b l i s h the age of this l a t e r phase; they are shown as #1 and #2 on F i g 7.17. The f i r s t depression (#1) i s currently a shallow pond on the flow which has a minimum water l e v e l of only 50cm during the summer months. Four cores were extracted by d r i v i n g bevelled ABS pipe into the lake bed with a sledge-hammer. The stratigraphy of the longest core i s shown i n Appendix F, together with a schematic diagram of the sampling locations. Basal sediments comprised grey green or oxidised, orange-brown g y t t j a , with many gravelly fragments. The gravel i s probably slopewash from the steep slopes surrounding the depression and i s derived from t i l l , not from under-lying sediments. A succession of peat and g y t t j a layers o v e r l i e the basal, gravelly gyttja i n a l l four cores, although the d e t a i l e d pattern varies from core to core. A sample from a peaty layer near the base of one core (Pav 2, stratigraphy given i n Appendix F) was dated at 1,810 ± 80 BP (Beta 16724). The second depression ascribed to phase 2 (#2 on F i g 7.17) was created by repeated slumping at the earthflow headscarp. It i s currently occupied by a bog, which was sampled with a H i l l e r corer. Detailed s t r a t i g r a p h i c information is given i n Appendix F. The lowest cored sediments at t h i s s i t e include many wood pieces, as the corer was stopped by wood rather than by gravel or sand. Gyttja i n the lower 5m of the core was very l i q u i d ; as a r e s u l t the author was not s a t i s f i e d 242 F i g 7.17 Phases of earthflow movement at P a v i l i o n (A) as mapped by Bovis (1985) 243 that an uncontaminated basal sample was obtained, as material leaked i n and out of the core b a r r e l as i t was extracted. Consequently, despite the dating p o t e n t i a l of t h i s s i t e , no r e l i a b l e samples were obtained for dating. The basal layer was succeeded by 40cm of brown-grey g y t t j a and then by 2m of cream "marl" with s n a i l s h e l l s and charcoal fragments. This was followed by 3.5m of mixed marl and organic layers g i v i n g way to Sphagnum and sedge i n the uppermost 3m. No Mazama or Bridge River tephra was p o s i t i v e l y i d e n t i f i e d . The Mazama tephra was observed i n a roadcut at t h i s s i t e , and thus would undoubtedly be present i n the bog were i t s u f f i c i e n t l y old. However, as the Mazama i s usually cream or yellow, and of s i l t y texture s i m i l a r to g y t t j a i t was not r e a d i l y d i f f e r e n t i a t e d from the "marl", espe-c i a l l y as the very l i q u i d lower sediments l o s t t h e i r structure during coring. It i s possible that l i t t l e or no Bridge River tephra f e l l at t h i s s i t e , as Hebda (G. Rouse, personal communication, 1985) d i d not record any at Finney Lake, although a g r i t t y orange layer at 235-236cm was t e n t a t i v e l y i d e n t i f i e d as Bridge River tephra. Stratigraphy at t h i s s i t e (#2, F i g 7.17, elevation 1,125m) can be compared with that at Phair and C h i l h i l Lakes (700 and 700m) and with Finney Lake (1300m). both locations shown in F i g 8.1. The vegetation at a l l three s i t e s is open Douglas f i r woodland, although Phair and C h i l h i l border on the ponderosa pine-bunchgrass zone of the Fraser Valley. Both Phair and C h i l h i l show a hiatus i n the early Hypsithermal, with permanent waterbodies becoming established shortly before the Mazama tephra (6,600 BP). At both s i t e s marl is seen most frequently i n the post-Hypsithermal stratigraphy, with g y t t j a or peat most common in more recent sediments. In contrast Finney Lake, with a continuous record from pre-Hypsithermal times, shows that marl only formed after p o s t g l a c i a l warming and before the Mazama tephra. The elevation of the P a v i l i o n s i t e f a l l s midway between Phair and C h i l h i l Lakes and Finney Lake. Therefore the continuous marl layer probably 244 formed during the Hypsithermal but before the Mazama tephra; while the d i s -continuous marl zone probably marks the long post-Hypsithermal cooling pe-r i o d . The onset of modern conditions i s indicated by the presence of Sphagnum and sedge, currently growing at the s i t e . If t h i s (rather speculative) i n t e r p r e t a t i o n of the core from s i t e Pa-v i l i o n #2 i s correct, i t implies that Phase 2 dates from the Hypsithermal or e a r l i e r . Although t h i s i n t e r p r e t a t i o n could be refuted by the 1,800 BP date from s i t e #1, t h i s l a t t e r date i s interpreted as recording the time at which t h i s r e l a t i v e l y shallow depression f i l l e d with water i n response to c l i m a t i c change rather than a date of formation. Also, there might have been minor r e a c t i v a t i o n near the headscarp. A further consequence of t h i s i n t e r p r e t a t i o n of Phase 2 i s that marl layers subsequently overridden by the earthflow are assigned a Hypsithermal,or s l i g h t l y e a r l i e r date, as no evaporites are forming i n the upper part of the flow at present. Locations where marl was observed are marked as #3 in F i g 7.17. Mazama tephra in a roadcut (#4 i n F i g 7.17) shows v e r t i c a l and l a t e r a l accretion o f the l a t e r a l deposit a f t e r 6,600 BP. Thus, although the earthflow at t h i s point predates the Mazama, there i s evidence o f a series o f large lobes having pushed through subsequently (M. Bovis, personal communication, 1987). The t h i r d phase recognised by Bovis (1985) i s only mapped on the east side o f the flow. No dendrochronologic work was c a r r i e d out at t h i s s i t e . However, v i s u a l examination of stumps remaining a f t e r s e l e c t i v e logging in 1962/1963 suggested that t h i s c u r r e n t l y i n a c t i v e area moved in the years succeeding 1800. This movement p a r a l l e l s that reported i n Chapter 5 for the area behind the current phase 4 scarp on the west side of the flow. Phase 4 comprises an area cu r r e n t l y a c t i v e , with movement rates measured by Bovis (1985) using e l e c t r o n i c distance measurements, and Nadler (1984) using mar-gin a l shear displacement indicated by stake arrays. 245 In summary, s t r a t i g r a p h i c evidence from the upper slump bowl suggests that the P a v i l i o n flow formed early i n the p o s t g l a c i a l . The flow was active at the time of the Mazama tephra and incorporation of evaporite deposits formed during the Hypsithermal into active flow zones shows that the flow has been active since the Hypsithermal. Unfortunately no evidence was found which indicated whether t h i s s i t e has been active throughout the p o s t g l a c i a l period or whether i t experienced period(s) of quiescence interspersed with movement phases. No chronologic work was undertaken at P a v i l i o n (B) (Fig 7.17) . 7.2.16 Red Mountain (A) earthflow Red Mountain (A) i s the largest of seven flows i n the Red Mountain group (Fig 7.5). The a i r photographs i n Appendix A show that i t has a large number of marginal cutoff and mid-flow depressions which are occupied by either seasonal or shallow permanent lakes, or else by bogs. Therefore t h i s s i t e yielded a large number of s i t e s from which s t r a t i g r a p h i c d e t a i l y i e l d i n g both r e l a t i v e and absolute dates could be obtained. A t o t a l of nine depressions were cored (indicated by a t r i a n g l e on F i g 7.18) . Lakes A,C,D and K were cored with a Livingstone corer, lakes A and K from a r a f t , while Lakes C and D were s u f f i c i e n t l y shallow to be cored by wading. Lake M was cored with bevelled ABS pipe driven i n by sledge-hammer while Lake F, which appeared on the most recent a i r photographs (1977) as a lake, had drained by the summer of 1984. A p i t was dug i n t h i s former lake bed. The three bog s i t e s (B,E and G) were sampled with a H i l l e r corer. De-t a i l e d stratigraphy from a l l s i t e s , except Bog G, i s given i n Appendix F. The stratigraphy and palynology of Bog G has already been discussed i n sec-t i o n 6.3.4. Lakes A and C both receive sediment from a small creek o r i g i n a t i n g near the active upper part of the flow, and both have a s u b s t a n t i a l proportion of sand and gravel horizons. Both s i t e s show a high proportion of Bridge 246 A Coring s i t e s F i g 7.18 Movement phases at Red Mountain (A) 247 River tephra shards i n most coarse horizons. At Lake A only lm of sediment was extracted ( i n four segments), the lowest segment only 5cm long as coarse sediments presented considerable f r i c t i o n on the core b a r r e l . Coring at Lake C was more successful with 425cm recovered from 7 cores. At t h i s l a t t e r s i t e the stratigraphy suggests that basal sediments may have been reached as gravel succeeded by sedge and then g y t t j a i n the bottom lm probably marks the gradual f i l l i n g of the lake basin. From 170cm to the top of the core, r a p i d l y a l t e r n a t i n g layers of sand, g y t t j a and sedge record the shallowing of the lake. The g y t t j a i s mainly purple r e f l e c t i n g the black to mauve basalt source rock. Rapid accumulation i s suggested by thick laminations, a minimum of 2mm thick i n the fi n e sediments, where they are seen as a l t e r n a t i n g purple and blue-grey bands, and considerably coarser i n the upper part of the core where an a l t e r n a t i n g sand-gyttja, or sedge, sequence i s seen. Some Bridge River tephra shards were present i n the basal sediments so the depression was thought to postdate 2,400 BP. The depression l a b e l l e d 'B1, now a meadow traversed by a meandering stream, shows a s i m i l a r stratigraphy to that of Bog G, although the base of the depression was not reached (see Appendix F for a d e t a i l e d stratigraphy). At t h i s s i t e , uniform, featureless black g y t t j a from the base of the core at 8m, to coarser sediments at 3m, indicates that the depression was i n i -t i a l l y occupied by a lake. At 2.3m the black g y t t j a changes to grey, and then to purple, (s i m i l a r to Lake C) i n d i c a t i n g e i t h e r a change i n source area, in lake environment, or both. The f i r s t coarse sediment appears at 2m: sorted Bridge River tephra i s seen at 170cm, and above t h i s i s an increasing amount of sedge and coarse sediments. Sedge without intervening mineral layers i s only seen i n the top 20cm of sediment; t h i s may be a response to the aban-donment of a large beaver dam which formerly ponded back the creek and probably promoted overbank flooding. At t h i s s i t e the base could not be 248 reached because s t i f f g y t t j a precluded further penetration of a manually operated H i l l e r corer. A mid-brown layer, 2cm thick at a depth of 572-4 cm, was submitted for analysis to determine whether i t was Mazama ash (no results are yet a v a i l a b l e ) . The evidence presented above suggests that the depression was formed at the same time as, or before Bog G, ( c i r c a 7,500-8,000 BP) i n view of the thickness of sediments which have accumulated below the Bridge River tephra. As t h i s depression stretches much of the width of the flow, i t was concluded that the middle part of the flow had been i n a c t i v e since at least the end of the Hypsithermal. The change i n sedimentary regime at about 2m, however, could be interpreted as a r e a c t i v a t i o n of the Upper part of the flow as th i s provides sediment for the creek traversing Bog B; a s i m i l a r change i s seen at Bog G, although there i t occurs a f t e r the Bridge River tephra was depos-i t e d . Valuable s t r a t i g r a p h i c information was extracted from the bogs and ponds along the north side of the north arm and at the lower end of the flow. Lake D and Bog E both showed Bridge River tephra close to the basal sediments: at Lake D (Appendix F) gravel was found 25cm beneath Bridge River tephra shards, and at Bog E wood prevented further penetration by the corer 10cm below the tephra (no stratigraphy was noted at Bog C beyond the location of the Bridge River tephra). Therefore both depressions were dated s l i g h t l y before 2,400 BP. Lake F, although at a s i m i l a r marginal l o c a t i o n , i s considerably older than Lakes D and E. Charcoal was extracted from a buried s o i l near the base of the p i t dug i n th i s dried lake bed (at 138cm, F i g 7.19), giving an age of 7,280 ± 100 BP (GSC-4164). A small diagram i l l u s t r a t i n g the location of the section i s shown i n F i g 7.19: i t can be seen that i t i s located on an a l l u v i a l fan b u i l t up by an ephemeral creek (not currently active) which exploited the depression between the flow material and the h i l l s l o p e . A l l a n uu 50 100 . t +• + t G.c 4164 7,520 t 100 • 150 O o 0 0 O 0 O . o. Dark brown sand with organic m a t e r i a l i n c r e a s i n g towards ground l e v e l . Many roots. Bridge R i v e r Ash parent m a t e r i a l Buff sand with many roots. Parent m a t e r i a l white coarse sand grains (l-2mm), a l t e r e d Bridge R i v e r Ash. Gravel w i t h angular stones up to 10cm w i t h long axes h o r i z o n t a l Organic m a t e r i a l Buff layer derived from Bridge R i v e r Ash shards. White, uniform, subrounded, sand grains. Stream channel cut i n t h i s l ayer Charcoal Clay w i t h some s i l t . Small c h a r c o a l fragments throughout. Top of u n i t r u s t coloured grading through mid brown t o dark brown at base Yellow-brown, poorly Borted s t i c k y , p l a s t i c c l a y t o g r a v e l Rounded c l a s t s Pale orange/yellow/buff f i n e s i l t and s t i c k y , p l a s t i c c l a y Clay to f i n e rounded j r a v e l . Yellow/brown m a t r i x w i t h dark orange mottles. Charcoal Yellow/brown, poorly sorted s t i c k y , p l a s t i c c l a y to rounded gravel Discontinuous gravel lenses Sand and g r a v e l , rounded to subrounded. H i g h l y coloured and variable v o l c a n i c c l a s t s . Fig 7.19a Stratigraphy of Lake F, core 1 Dark brown sand derived from Bridge River tephra Much decomposed organic material I Flow d i r e c t i o n of ephemeral creek Bridge River tephra shards./ Brown clay with s i l t and f i n e l y divided charcoal Stable ground Earthflow Yellow/brown, poorly sorted gravel to clay Schematic diagram of Lake F showing coring s i t e s F i g 7.19b Stratigraphy of Lake F 251 s t r a t a i n the p i t dip to the east (apparent dip 6°), and a subsidiary p i t dug i n the east side of the lake (Fig 7.19) confirmed t h i s i n t e r p r e t a t i o n as most of the layers were duplicated, but were considerably narrower. The stratigraphy i n F i g 7.19 shows non-plastic basal sediments of rounded to subrounded varicoloured volcanic c l a s t s . This material may have originated from slopewash, and the dated charcoal horizon i s found towards the top. The next layer i s coarse rounded material i n a s t i c k y , p l a s t i c ma-t r i x and i s interpreted as a l l u v i a l fan sediments l a i d down when earthflow marginal erosion rates were high, probably when the flow was moving, suc-ceeded by a more quiescent sedimentary environment, probably the r e s u l t of low erosion rates once the flow s t a b i l i s e d , when 40cm of clay was deposited. The Bridge River tephra i s seen as a layer 30cm thick, sorted i n the bottom 5cm. It indicates that a considerable volume of tephra was blown or washed into t h i s depression a f t e r deposition. The shards were probably deposited i n a shallow lake, as an i n f i l l e d stream channel i s seen near the top of the Bridge River tephra layer. Most subsequent deposition has been reworked Bridge River tephra shards. F i g 7.18 shows that the complex topography and drainage patterns at the t oe o f the flow may be explained by the chronology o f earthflow movement. F i r s t , the south arm, which i s dated by the cutoff of Bog G at or before 7,790 BP advanced down a Pleistocene v a l l e y (phase 1 i n F i g 7.20). At t h i s time Lake K d i d not e x i s t ; i t was formed when the north branch of the north lobe dammed a creek draining the south branch of the flow. The creek draining lake K i s a c t i v e l y eroding the outflow channel. Lake K was cored to determine i t s age, thereby dating movement on the north arm of the west lobe. Three cores were taken from three d i f f e r e n t l o -cation in Lake K in 3-5m of water; a p a r t i c u l a r concern at t h i s s i t e was to avoid the south end of the lake where a sizeable d e l t a had b u i l t up (see Appendix A). It was also thought prudent to avoid the west slope, which was 2. Earthflo« • o v M i t circa 7,000 to 8,000 BP cam** I. Poatglaclal topography th* moat to occupy th« pr«*xiiting vallay. Th« cr»»k It dlwt*d around th* «»»t aid* of th* flow F i g 7.20 Formation of Lake K 3. Shortly bafor* 2,400 IP th* craak dralala* th* South a n !• daaMd to (iv* Lak* K ho Ln 253 steep and had very l i t t l e vegetation, and was therefore p o t e n t i a l l y a source of boulders and gravel which might prevent a complete core being extracted. A l l three cores showed the same stratigraphy (one i s shown i n F i g 7.21). The base of each core was sandy g y t t j a with angular gravel and no obvious stratigraphy. Diminishing amounts of gravel p e r s i s t e d to 10cm below the Bridge River tephra which was 2-3cm thi c k , f i n i n g upwards. The upper part of the core was g y t t j a with a l l three cores showing a higher proportion of organic material towards the top. It was concluded from these cores that Lake K had formed shortly before the Bridge River tephra ( i . e . before 2,400 BP) and that t h i s date corresponded with major movement recorcjed i n two marginal depressions on the north side of the west part of the flow. This det a i l e d analysis of a large number of depressions has enabled a chronology of movement to be constructed. The f i r s t phase of movement, cre-ating both the south and west parts of the flow, took place at or before the two radiocarbon dates of 7,290 ± 100 BP and 7,790 ± 90 BP. Much of the southern part of the flow, together with the area above the b i f u r c a t i o n point on the west part, has been inactive since that time. Major r e a c t i v a t i o n of the north side of the b i f u r c a t i o n and of the upper part of the west part, took place just before the Bridge River tephra (2,400 BP) and th i s i s re-corded both i n lake bottom dates and i n increased c l a s t i c sediment fed into Bogs B and G. Those parts of the flow which experienced r e a c t i v a t i o n , to-gether with a small part of the south lobe, are active at the present time. It i s not known whether a c t i v i t y has been continuous since Bridge River times or whether recent movement i s a response to increased p r e c i p i t a t i o n since 1950. F i g 7.21 Stratigraphy of Lake K 0 • 20-40 • 60-80' 100-120 • 140 • 160 180 • 200 4 M l Organic Blue grey g y t t j a Blue grey g y t t j a with roots Blue grey g y t t j a Brown grey g y t t j a Bridge River tephra shards Green blue g y t t j a with angular gravel and sand becoming coarser with depth. No stratigraphy seen. 220. 255 7.2.17 Red Mountain B to E and G earthflows No d e t a i l e d s t r a t i g r a p h i c evidence was obtained to date phases of movement of these flows as they were inaccessible without helicopter support. P i t s were dug to approximately 1.5m i n two marginal depressions to the east of flow B but only uniform, featureless, p l a s t i c orange-pink clay derived from weathered basalt was found. As the flow surface was dominated by boul-ders, there were no depressions i n which Bridge River tephra might have been sought. The meadow at the bottom of Flow B was cored to investigate whether any changes i n sedimentary regime had taken place as sediment was deposited by flooding of streams draining the flow. The s t r a t i g r a p h i c column i s re-corded i n Appendix F: only one exploratory core was extracted as the s i t e was a f u l l day t r i p from base camp. It can be seen that Bridge River tephra shards are recorded at approximately 25cm: the core i s about lm long. Therefore the bog predates the Bridge River tephra and must be considerably older. No chronologic observations were made on flows B to E and flow G was not v i s i t e d . 7.2.18 Red Mountain F earthflow A shallow lake and marginal swamp at the top of t h i s flow was cored using ABS pipe driven i n with a sledge-hammer. Two cores were taken: the stratigraphy of the deepest core i s shown i n Appendix F. It can be seen that the core was only driven to the depth of the Bridge River tephra as at t h i s s i t e shards (probably washed i n from the surrounding area) provided s u f f i -cient resistance to prevent a deeper sample being extracted. Therefore 2,400 BP i s a minimum age for t h i s flow, unless the tephra i s reworked. 256 7.2.19 Yalakom earthflows A l l four Yalakom flows (Fig 1.7) comprise uniform serpentinised p e r i d o t i t e . A l l the flows bear much vegetation s i m i l a r i n appearance to that of the surrounding forest. However, parts of both Yalakom (A) and (D) bear sparse, stunted trees i n comparison with other flows. Ground inspection at Yalakom (A) revealed that although the flow had been mantled with a thick (15cm) blanket of Bridge River ash, on steep slopes, and on some active parts of the flow, t h i s cover had been l o s t . Bare serpentinite was colonised predominantly by stunted Pinus  contorta whereas Pinus a l b i c a u l i s with many grasses, sedges, Apiaceae and composites, t h r i v e where tephra o v e r l i e s the s e r p e n t i n i t e . Yalakom (A) was zoned into post-and pre-Bridge River zones of movement on the basis of presence or absence of tephra and accompanying vegetation (Fig 7.22). It can be seen that post-Bridge River movement occurs at the toe of the flow where slumping, rather than earthflow i s seen. A l l the l a t -e ral deposits were formed p r i o r to 2,400 BP as they are covered by Bridge River tephra. It can also be seen that large "sackung" features and tension cracks at the top of the flow were formed before 2,400 BP since, although the tephra has been washed away from the steep block-covered slopes at the top of the flow, i t was trapped i n small depressions created by backwards rotated slump blocks. Therefore a series of narrow arcuate vegetation zones have formed above the nominal t r e e l i n e (Appendix A) where thick tephra ac-cumulations are found. The vegetation at these s i t e s i s stunted P.  a l b i c a u l i s . Bridge River tephra accumulation on l a t e r a l deposits shows that con-siderable movement of the Yalakom (A) flow took place before 2,400 BP, a l -though i t i s not possible to e s t a b l i s h a maximum date. The flow i t s e l f may have been active, but s t i l l maintained a covering of Bridge River tephra i f l i t t l e i n t e r n a l deformation took place. Pre-Bridge River movement on Yalakom 257 (A) must be p o s t g l a c i a l since i t cuts a series of kame terraces and has forced the Yalakom River to d i v e r t around the flow toe. Yalakom (B) i s mantled by tephra throughout, i n d i c a t i n g either s t a -b i l i t y from 2,400 BP or 'en bloc' movement with l i t t l e or no i n t e r n a l de-formation. Two lakes and a bog occupy depressions i n the slump bowl (Fig 7.23). The two lower lakes (#1 and #2 i n F i g 7.23) were cored using ABS pipe while two sections were dug i n the bog (#3). Descriptions of these sections are given i n Appendix F. In a l l cases a considerable thickness of Bridge River tephra was seen, 15cm i n the bog, 30-40cm i n Lake 2 and 60cm i n Lake 1. Abundant organic matter o v e r l i e s the tephra, but beneath there i s no evidence of weathering, s o i l formation or organic matter accumulation. It appears that at t h i s elevation (1700m) u l t r a b a s i c rocks were i n i m i c a l to vegetation growth during the more favourable growing conditions of the Hypsithermal. Yalakom (C) (Fig 7.24) i s mantled by Bridge River tephra, implying an o r i g i n before 2,400 BP. Some post-Bridge River slumping has taken place on the steeply-sloping snout front, and a few active slumps are seen on the flow i t s e l f , breaking the uniform mantle of Bridge River ash. There i s a bog i n front of the flow. A creek draining the north earthflow margin crosses the bog (Fig 7.24) so i t was thought that the r e s u l t i n g sedimentary record might provide valuable s t r a t i g r a p h i c information about cessation of movement of the earthflow. Six cores were extracted using a lm p l e x i g l a s s piston corer with cutting teeth, The stratigraphy of the longest core, Ql, i s shown i n Appendix F. It can be seen that before the Bridge River tephra, the sediments were mainly mineral, f i n i n g upwards, with a high proportion of gravel and only t h i n sedge layers. A f t e r the tephra (11cm thick and sorted i n t h i s core), the sediment i s e x c l u s i v e l y organic. A l l the other cores revealed a s i m i l a r stratigraphy. 258 F i g 7.22 R e l a t i v e ages of movement at Yalakom (A) 259 F i g 7.23T Location of coring s i t e s on Yalakom (B) 260 261 Sedge from the base of the core above the gravel and pebbles was dated at 3,790 ± 100 BP (Beta 16723). This date i s thought (speculatively) to represent cessation of earthflow movement on the premise that coarse sediment corresponds to erosion by marginal streams of an active earthflow thereby providing a minimum age. No maximum age was av a i l a b l e at t h i s s i t e . 7.3 A comparison of c l i m a t i c data and earthflow movement i n the p o s t g l a c i a l period The c l i m a t i c data presented i n Chapter 6 derive mainly from pollen analysis and a survey of Neoglacial f l u c t u a t i o n s . Both sources indicated a warm period p r i o r to 7,000 to 8,000 BP, succeeded by slow cooling to modern conditions. However, because the d e t a i l of flu c t u a t i o n s within t h i s broad framework was much less c l e a r , the core at Red Mountain (Bog G) , from a high elevation s i t e within the study area, provided valuable supplementary e v i -dence i n timing post-Hypsithermal c l i m a t i c changes. Estimates for the onset of Hypsithermal cooling vary according to l o c a t i o n i n southern B r i t i s h Columbia. Bog G shows that a s i g n i f i c a n t cooling phase occurred sh o r t l y after the Mazama tephra, although c l i m a t i c conditions were much warmer than at the present time u n t i l 3,200 BP. A second sudden temperature decline took place around 3,200 BP, followed by uniformly cool conditions u n t i l the present time. This chronology correlates well with Neoglacial ice advances beginning 3,000 to 3,500 BP, indeed better than any other palynological studies i n south central B r i t i s h Columbia. It i s possible that as t h i s s i t e i s con-siderably higher than any other published s i t e s , i t r e f l e c t s temperature fluctuations rather than p r e c i p i t a t i o n trends. This would explain the weaker signal recorded at lowland s i t e s more s e n s i t i v e to evapotranspiration i n the B r i t i s h Columbian i n t e r i o r . Z9Z . BIG BAR CAMOO CANOE CHURN DRYNOCH (A) DRYNOCH (B) • 6 >• 6 • 3 • • 2 FLAPJACK GIBBS CREEK 1 KETTLEBROOK CREEK i TUNNELj * GILLON CREEK GRINDER (A) GRINDER (C) HAT CREEK HEGINBOTTOM | ». 3 j> I PAVILION (A) I ! • 3 RED MOUNTAIN (A) RED MOUNTAIN (F) YALAKOM (A and B) YALAKOM (C) Thousands of years 5 6 7 8 ? 5 4 3 2 . U M MI • 3 2 1 4 M 4 l * 1 4 2 1 10 11 —4-•* • 2 14 Radiocarbon or tephra date • Relatively or absolutely dated range M Open ended maximum or minimum Numbers refer to phases discussed i n text Fig 7.26 Summary of earthflow movement over the l a s t 11,000 years 264 It i s d i f f i c u l t to correlate c l i m a t i c fluctuations with earthflow movement, since at many s i t e s movement cannot be determined to better than several thousand years. In F i g 7.26 maxima and minima have been c l a s s i f i e d judgementally into those giving a r e a l i s t i c (say, ± 500 years) estimate of movement, those giving a useful range and open ended maxima and minima too broad to give useful information. At two s i t e s (Red Mountain A and Heginbottom), the Bridge River tephra gives a useful minimum date as major movement ended sho r t l y before 2,400 BP. Useful maxima are given by the Bridge River tephra at Big Bar and Grinder Creek (A). At the former s i t e t h i s l a t e r e a c t i v a t i o n i s morphologically d i s t i n c t from older phases i n the same flow, implying a prolonged quiescent period p r i o r to 2,400 BP. At Grinder Creek (A) the tephra dates i n i t i a t i o n of movement. Post-Hypsithermal dates are in f e r r e d from several other s i t e s , notably Drynoch (A and B), Flapjack, Gibbs Creek, Kettlebrook Creek and Tunnel earthflow and GilIon Creek on the basis of radiocarbon dating and the posi-t i o n of the Mazama tephra and sedimentation type and rate. None of these s i t e s have e a r l i e r phases i n f e r r e d from either morphology or stratigraphy. Red Mountain (A) i s the e a r l i e s t movement phase which has been firmly dated. Cessation of movement i s given by radiocarbon dates from two marginal depressions, one on either arm of the flow, between approximately 7,300 and 7,800 BP. A number of broadly contemporaneous minima are also shown (Big Bar, Churn Creek and Hat Creek), but at a l l these s i t e s a s u b s t a n t i a l l y older date than 7,000 BP i s possible. It should be possible to obtain maxima for Big Bar and Churn Creek once the Fraser River terraces are r e l i a b l y dated, and possibly a minimum for Hat Creek. At a l l these s i t e s a date substan-t i a l l y older than 7,000 BP i s possible. P a v i l i o n phases 1 and 2 are placed i n the early p o s t g l a c i a l on the strength of the sedimentary record from a lake record i n a slump bowl de-265 pression. As no basal date was obtained, t h i s i n t e r p r e t a t i o n may be incor-rect, although i n the author's judgement, the p a r a l l e l s between the sedimentary record seen at t h i s and other nearby and comparable s i t e s enable a v a l i d comparison to be made. Early p o s t g l a c i a l dates for Canoe Creek and Camoo are speculative and are not given any further consideration. Movement around 2,400 BP may be a t t r i b u t e d to the sharp temperature decline and (or) p r e c i p i t a t i o n increase recorded at Bog G. This movement overlaps with a documented phase of neoglaciation of the Coast Mountains. Although g l a c i a l readvance i s known to have begun e a r l i e r , around 3,000 -3,500 BP, neoglaciation probably continued to around 2,000 BP. At Red Moun-t a i n and Heginbottom the dates are minima, so a s l i g h t l y e a r l i e r date i s implied at these s i t e s . Movement dated as post-Mazama tephra can also be att r i b u t e d to temper-ature decline and/or p r e c i p i t a t i o n increase, although the chronological un-certainty of several thousand years i n dating most earthflows precludes a di r e c t comparison with climate. However, the date of 7,800 BP from Red Mountain, which i s accepted on the basis of three c l o s e l y a l l i e d radiocarbon dates from two d i f f e r e n t materials, and from two d i f f e r e n t s i t e s on the earthflow, i s problematic. It i s well established from many studies that the climate at th i s time was warmer and d r i e r than that experienced i n the post-Hypsithermal period. Although the I n t e r i o r may have started to cool as early as 8,500 BP, there i s l i t t l e or no evidence for a s i g n i f i c a n t vege-t a t i o n s h i f t u n t i l a f t e r 6,700 BP (Mazama tephra). Therefore, a connection between earthflow movement at Red Mountain and c l i m a t i c fluctuations i s not cert a i n . In summary, most earthflows which have been dated f a l l within the post-Hypsithermal (say 7,000 BP) period of cooling. A peak of a c t i v i t y i s seen around 2,400 BP which overlaps with a major period of neoglaciation. However, the oldest s i t e which was fi r m l y dated (Red Mountain (A)) was f i r s t 266 active towards the end of the Hypsithermal ( i . e . c i r c a 7,800 BP). As a r e s u l t of uncertainty i n dating the onset of cooler, moister conditions a f t e r the Hypsithermal temperature peak, movement at t h i s s i t e cannot be correlated with climate. 267 CHAPTER 8 DISCUSSION AND CONCLUSIONS 8.1 Introduction In chapter one the objective of the thesis was put forward. It was proposed to i d e n t i f y phases of earthflow a c t i v i t y and then c o r r e l a t e these with c l i m a t i c fluctuations recognised both from analyses i n t h i s study and also from other research i n southern B r i t i s h Columbia. A subsidiary issue, which arose out the need to describe and c l a s s i f y earthflows, was a compar-ison of geologic conditions and earthflow materials i n south-central B r i t i s h Columbia with those described elsewhere i n the world i n order to i d e n t i f y common topographic and geologic fa c t o r s . In the course of t h i s work and the associated l i t e r a t u r e review, some major scale e f f e c t s governing the behaviour of earthflows were observed. 8.2 Geologic factors Three broad categories of geologic factors were i d e n t i f i e d i n t h i s study as contributing to the observed concentration of earthflows i n the study area: rock type, geologic structure and geomorphic h i s t o r y . The f i r s t category, rock type, embraces both l i t h o l o g y and rock weathering products, while the second category includes other factors such as dip angle, f a u l t i n g and geologic succession. The t h i r d category, geomorphic h i s t o r y , analyses the influence of regional u p l i f t and subsequent d i s s e c t i o n , and of g l a c i a l erosion. Earthflows i n the study area were developed i n serpentinised p e r i d o t i t e (20%), basalt (18%), sediments (16%), other volcanics (excluding b a s a l t ) , and sediments (42%). This i s a contrast with other areas where only one flow (Melendy Ranch, San Francisco Bay Area, Keefer, 1977) was recorded i n serpentinised p e r i d o t i t e , although t h i s rock type formed a s i g n i f i c a n t proportion of several of the Franciscan melange flows (mixed sediments and serpentinite comprise 14% of the t o t a l outside the study area). No earthflows 268 developed i n b a s a l t i c materials were recorded from outside the study area, and earthflows developed i n other volcanics formed only a small proportion of the t o t a l . The major rock category which formed earthflows outside B r i t i s h Columbia was sediments (66%), ranging from clays and shales to conglomerate. In many cases, both within and outside the study area, sediments were v o l c a n i c l a s t i c s . Within the study area a l l the rock types forming earthflows, with the exception of serpentinised p e r i d o t i t e , weather to montmorillonite with small quantities of other clay minerals (Chapter 3) . It seems that earthflows form p r e f e r e n t i a l l y i n montmorillonite-rich debris because t h i s clay mineral has a very low angle of shearing resistance. Analyses from outside the study area (Table L l ) show a much wider range of dominant cla y minerals, including several flows with a high proportion of k a o l i n i t e and i l l i t e , neither of which i s an expanding clay mineral. However, as the majority of flows given in Table L l are 'small' (area less than 10^), while the majority i n the study area are 'medium-sized' or 'large' (area over 10 ), considerations of scale, discussed below, suggest that a d i r e c t comparison i s inappropriate. The location of earthflows i n serpentinised p e r i d o t i t e cannot be as-cribed to the presence of expanding clays with a low angle of shearing re-sistance. Grain si z e analyses (Fig 3.9) show that there i s an extremely small proportion of clay- and s i l t - s i z e d material i n these flows, 75-90% of the <2mm clasts being sand-sized. Moreover, the small clay percentage comprises mainly serpentine minerals and c h l o r i t e . Therefore i t i s proposed that earthflows do not form i n serpentinised p e r i d o t i t e as a r e s u l t of weathering products, but because the i n t a c t rock has many small, curved, s l i p planes, which re s u l t i n a low angle of shearing resistance at the earthflow shear plane. There are no geotechnical data which t e s t t h i s assertion. A secondary e f f e c t of weathering rates, rather than products, i s shown in Figs 3.11. In these diagrams, plo t s of earthflow s i z e and angle for 269 basalt, sediments and serpentinised p e r i d o t i t e , show substantial differences i n the size/angle r e l a t i o n s h i p . In basalt earthflows, for example, movement takes place, on average, on s l i g h t l y lower angle slopes than for sediments. This i s probably a r e s u l t of higher f r i c t i o n a l resistance from the sediments which are, on average, coarser textured than basalt weathering products (see F i g 3.9). This argument i s reinforced by the markedly d i f f e r e n t size/angle r e l a t i o n s h i p displayed by serpentinised p e r i d o t i t e , which requires steeper slopes to form earthflows, and i s shown i n F i g 3.9 to be considerably coarser than any other material forming earthflows. Within the study area there are geologic s t r u c t u r a l features which may influence earthflow location. For example, 36 of the 49 flows examined i n Appendix H occur p a r a l l e l with or transverse to major f a u l t s . This may be a coincidence, as several f a u l t s , such as the Hungry Va l l e y and Yalakom f a u l t s , coincided with deeply i n c i s e d v a l l e y s . However, f a u l t i n g may also a f f e c t groundwater flow, either by e x p l o i t i n g a shattered, r e l a t i v e l y permeable f a u l t zone, or by bringing permeable and impermeable s t r a t a into f a u l t con-ta c t . Outside the study area the role of f a u l t i n g cannot be assessed as few authors record whether the features they studied occurred in a f a u l t zone.. In only one case, that of the Kirkwood earthflow, i s f a u l t i n g unequivocally linked with an earthflow. At t h i s s i t e an earthquake formed a f a u l t scarp across the flow, thereby increasing flow gradient. Greatly increased groundwater flow was also recorded at Kirkwood by v i s u a l observations, sug-gesting disruption of the established drainage network by f a u l t action. It was concluded, however, that the majority of earthflows were unaffected by recent f a u l t a c t i v i t y Within the study area the d i s t r i b u t i o n of earthflows adjacent to Fraser River (sediments, F i g 3.6) and on Red Mountain (basalt, F i g 15), suggests that earthflows move p r e f e r e n t i a l l y p a r a l l e l with regional dip angle. It 270 i s possible that t h i s pattern of movement res u l t s from high piezometric pressures which have been shown t h e o r e t i c a l l y to form on the dip slope (see F i g 3.8). Outside the study area few earthflows show any r e l a t i o n s h i p to dip angle, which i s not described at the majority of s i t e s . Of those flows outside the study area which do take place down dip, only one i s small, one medium-sized and two (out of a t o t a l of three) are large. Although the sample sizes are too small to draw general conclusions, i t i s possible that the e f f e c t of dip angle on earthflow l o c a t i o n i s l i m i t e d to medium- and large-siz e d features. Within the study area, r i v e r i n c i s i o n has taken place since u p l i f t of the Coast Mountains over the past m i l l i o n years. This has created steep v a l l e y sides upon which earthflows form. This factor i s also present in other areas of medium and large-sized earthflow concentration, including the Columbia River Gorge, the E e l River area, the Cascades and the Van Duzen River basin. Many smaller flows outside the study area are located on eroding coastal s i t e s which provide l a t e r a l (not v e r t i c a l ) toe erosion which has a s i m i l a r e f f e c t of increasing earthflow gradient. The influence of g l a c i a t i o n on earthflow a c t i v i t y i s uncertain. Sharpe and Dosche (1942), on the Appalachian Plateau, show that earthflows only form outside the g l a c i a l l i m i t , implying that g l a c i a l erosion scoured out weath-ered material which provided p o t e n t i a l earthflow debris. However, as the whole of B r i t i s h Columbia was g l a c i a t e d , earthflows i n the study area cannot be a t t r i b u t e d to the absence of g l a c i a l erosion. It i s possible that, as ice i n the i n t e r i o r of B r i t i s h Columbia was probably slow moving, i t may have performed less erosion than that at the more r a p i d l y moving ic e marginal locations on the coast. 8.3 Scale considerations In Chapter 1 the scale-dependent r e l a t i o n s h i p between earthflow move-ment and p r e c i p i t a t i o n was explored. Appendix G shows that the majority of 271 flows which respond to i n d i v i d u a l p r e c i p i t a t i o n events are c l a s s i f i e d as 'small' (area <10 -10 m ). Those shown to respond to longer term c l i m a t i c changes, both i n Appendix G (Waerenga-O-Kuri and Wind Mountain), and i n the 5 6 2 present study, are 'medium-sized' (area 10 -10 m ) or 'large' (area >10 6m 2). A further d i f f e r e n c e i n behaviour between 'small' (<10^-10^) and other flows i s also suggested by F i g L3 which shows that, for medium and large flows outside the study area, earthflow gradient i s r e l a t e d to earthflow s i z e 5 6 2 6 2 for medium (10 -10 m ) and large (>10 m ) flows. The reasons for t h i s re-lation s h i p are not c l e a r , but four hypotheses concerning groundwater flow disruption, marginal and basal shear, and regional topography, are discussed i n Chapter 3. For small flows, however, gradients are considerably lower and no r e l a t i o n s h i p between s i z e and flow angle can be seen. Scale consider-ations may also be r e f l e c t e d i n flow mineralogy. In the study area, where flows are developed mainly i n montmorillonite, flow angle i s related to earthflow s i z e . Earthflows i n other areas, mainly small flows with a v a r i e t y of minerals as shown i n Table LI, do not show any r e l a t i o n s h i p between flow angle and s i z e . On the basis of t h i s evidence, i t was concluded that a fundamental d i s t i n c t i o n must be made between the behaviour of 'small' flows and others. Small flows are found predominantly at coastal locations and a c t i v e l y undercut v a l l e y side slopes, where they are maintained by toe erosion. They respond r a p i d l y to i n d i v i d u a l p r e c i p i t a t i o n events whose moisture i s fed d i r e c t l y to the shear plane, causing a rapid r i s e i n the l o c a l piezometric l e v e l , and consequently rapid movement i n the short term. In contrast, medium-sized and large flows, which include the majority of flows i n the study area, do not usually respond to i n d i v i d u a l p r e c i p i t a t i o n events. They respond to seasonal and longer p r e c i p i t a t i o n f l u c t u a t i o n s , they move r e l a -272 t i v e l y slowly, and form mainly where a high proportion of expanding clay minerals permit creep on r e l a t i v e l y low angle slopes. Examination of flow depth and area showed that a s i m i l a r scale-dependent d i s t i n c t i o n could be made between 'small' (<10^-10^m^) and larger earthflows (Fig 1.10). 'Small' flows were less than 5m deep, while medium and large-sized flows showed a weak size-dependent depth r e l a t i o n s h i p . As tension cracks frequently penetrate earthflows to at least 3m, the shear plane on these smaller flows may be fed d i r e c t l y from cracks i n the flow surface, and by streams t r a v e r s i n g the flow, as well as by groundwater d i s -charge. Therefore 'small' flows have responded r a p i d l y to p r e c i p i t a t i o n inputs as these are de l i v e r e d d i r e c t l y to the shear plane rather than f i l -t e r i n g through the groundwater flow system. In contrast, larger flows have responded to changes i n groundwater l e v e l . Seasonal acceleration i s well i l l u s t r a t e d by F i g 1.4, which shows how Drynoch earthflow i n the study area has responded to snowmelt. Longer term changes in groundwater l e v e l , i n response to changes i n seasonal mean pre-c i p i t a t i o n , have not been monitored. However, i n view of the seasonal pat-tern at Drynoch i t i s reasonable to propose that an o v e r a l l increase i n groundwater level might be seen in response tc increased winter p r e c i p i -t a t i o n . It i s postulated that t h i s mechanism has driven the long term f l u c -tuations assessed i n section 8.5. 8.4 Observations on earthflow movement mechanisms Monitoring of earthflow marginal stake arrays at Grinder Creek (A) indicated that, at t h i s s i t e , earthflow movement was s p a t i a l l y v a r i a b l e . Moreover, the most r a p i d l y moving s i t e s occurred at, or j u s t beyond, mid-flow lobes. Most of the earthflow length was characterised by 'plug' type move-ment, where the main body of the flow was ra f t e d along, and deformation took place at a marginal shear plane and i n a marginal deformation zone about lm in width. Where lobes occurred, the earthflow d i l a t e d at the lobe c r e s t , 273 where tension created by the accelerating flow broke the 'plug' into a series of chaotic blocks separated by deep tension cracks. As the flow accelerated beyond the crest, drawdown below the crest of the l a t e r a l deposits was seen and 'plug' structure was absent as the e n t i r e flow underwent i n t e r n a l de-formation. Craig (1981) showed that both types of flow ('plug'and 'viscous') are observed i n d i f f e r e n t seasons i n Northern Ireland, but there are no re-ports of comparable temporally v a r i a b l e rheology from any other flows outside the study area. 8.5 The r e l a t i o n s h i p between earthflow movement and c l i m a t i c fluctuations  i n the study area Many studies have correlated earthflow movement with p r e c i p i t a t i o n or groundwater levels (see Chapter 1) over a period of a few months (Figs 1.2 to 1.6) and Wasson and H a l l (1982) and Palmer (1977) have demonstrated that a longer term r e l a t i o n s h i p i s seen between earthflow movement and periods of high r a i n f a l l over several decades. Within the study area paired l a t e r a l deposits indicate that at many s i t e s repeated movement phases have taken place. One objective of t h i s study was to investigate earthflow movement at three longer timescales: the l a s t 50 years; the l a s t 300 years; and the p o s t g l a c i a l period. Earthflow movement over the l a s t 50 years was interpreted from a i r photographs which went back at some s i t e s to 1928, as well as stake meas-urements made over a two year period. At a l l s i t e s a i r photograph stereopairs were av a i l a b l e from about 1940 onwards, and c a r e f u l examination of photo sequences indicated that at seven s i t e s : Black Dome, Churn Creek, Grinder Creek (A) and (B), Red Mountain (B) and (G) and Yodel (A), major reactivations were seen between 1950 and 1960, followed by rapid movement during the 1950's and 1960's, with a trend at some s i t e s for decreasing movement i n the 1970's. The onset of earthflow movement p a r a l l e l s that noted in studies of recent g l a c i a l advances i n the Coast Mountains, the Rocky 274 Mountains and the i n t e r i o r of B r i t i s h Columbia (Luckman, Harding and Hamilton, 1986). No p r e c i p i t a t i o n records of a useful length were av a i l a b l e within the study area, so c l i m a t i c fluctuations were i n f e r r e d from an analysis of pre-c i p i t a t i o n records from three AES stations close to the study area (Lytton, Mamit Lake and Big Creek). Of these, Big Creek had the longest and most complete record, and also had a regime midway between coastal (winter pre-c i p i t a t i o n dominant) and i n t e r i o r (summer p r e c i p i t a t i o n dominant). The ge-ographic location and vegetation assemblage of the study area suggests that i t , too, i s influenced by both regimes. P r e c i p i t a t i o n records at a l l three s i t e s showed an extended period of above-average winter p r e c i p i t a t i o n 1945-1970. This probably brought about the observed earthflow r e a c t i v a t i o n , as Thornthwaite c a l c u l a t i o n s show that summer p r e c i p i t a t i o n is not available for recharge at a l l four c l i m a t i c stations. Similar r e a c t i v a t i o n i n response to a long term increase i n pre-c i p i t a t i o n was seen at Waerenga-O-Kuri (Wasson and H a l l , 1982) and i n Oregon (Palmer, 1977), although at these s i t e s t o t a l p r e c i p i t a t i o n , not winter p r e c i p i t a t i o n , was correlated with movement. No p r e c i p i t a t i o n records were av a i l a b l e before about 1900, so longer term movement was assessed using t r e e - r i n g width chronologies from four s i t e s adjacent to active earthflows. These, however, were found to co r r e l a t e with summer rather than winter p r e c i p i t a t i o n , which was shown i n the t r e e - r i n g analysis to control earthflow movement. However, further analysis of the p r e c i p i t a t i o n data revealed that summer and winter p r e c i p i t a t i o n had, over the period of record, coincident maxima and minima, but showed d i f f e r i n g trends and fluctuations between these p r e c i p i t a t i o n extremes. Therefore, on t h i s basis, the t r e e - r i n g record was accepted as a surrogate for long-term winter p r e c i p i t a t i o n trends. The i n f e r r e d c l i m a t i c pattern varied i n d e t a i l 275 from s i t e - t o - s i t e , but the o v e r a l l trend was for periods of above-average p r e c i p i t a t i o n 1690-1720, 1800-1820, 1880-1920 and 1940-1970. A movement chronology was i n f e r r e d from reaction wood i n cores and tree discs extracted from t i l t e d trees growing i n active headscarp areas at four s i t e s . These revealed movement maxima which correlated with moist periods i n the c l i m a t i c record. Therefore i t was concluded that earthflows have moved over the past two centuries i n response to 20-50 year p r e c i p i t a t i o n cycles. The p o s t g l a c i a l c l i m a t i c chronology was assessed both from an exten-_sive l i t e r a t u r e review, and from a d d i t i o n a l work i n Chapter 6. Climatic changes were i n f e r r e d from pollen analyses at low elevation s i t e s around and within the study area. The o v e r a l l pattern was for the warm, dry, Hypsithermal i n t e r v a l from 10,000-7,000 BP, followed by slow cooling to around 2,500 BP at which time present-day conditions were established. However, as many earthflows occurred above 1800m, a supplementary po l l e n analysis was performed on a core from Bog G at Red Mountain. A two-phase p o s t g l a c i a l cooling period was recognised, the f i r s t phase taking place sho r t l y after the Mazama tephra, and the second, not c l e a r l y recognised i n other palynological work from lower s i t e s , around 3,200 BP. The second cooling phase corresponds with the onset of a major Neoglacial ice advance dated at a number of s i t e s i n B r i t i s h Columbia and the surrounding area (Ryder and Thomson 1986). The chronology of p o s t g l a c i a l earthflow movement was gleaned from a large number of sources. F i r s t , Mazama and Bridge River tephras were iden-t i f i e d at many s i t e s i n c r i t i c a l locations such as mid-flow depressions, marginal depressions, and also embedded in former shear planes now exposed by stream i n c i s i o n . At a number of s i t e s samples were submitted for radiocarbon dating, and elsewhere morphologic and pedogenic evidence was used to in f e r the r e l a t i v e age of earthflow movement. 276 The oldest earthflow r e a c t i v a t i o n to be dated was the Red Mountain earthflow, at around 7,500 BP. As t h i s was based on 3 dates from two l o -cations, i n i t i a t i o n or a major r e a c t i v a t i o n of t h i s feature must have taken place at t h i s time. However, paleoclimatic evidence shows that t h i s date i s at the close of the Hypsithermal i n t e r v a l , and so temperatures were higher and possibly p r e c i p i t a t i o n lower than at present. This evidence suggests that t h i s feature may not have reactivated i n response to d e t e r i o r a t i n g c l i m a t i c conditions, although some work from the i n t e r i o r , summarised i n Mathewes (1984), does suggest that Hypsithermal cooling may have begun as early as 8,500 BP. Post-Hypsithermal dates ( i . e . 6,600 to around 4,000 BP) are i n f e r r e d from seven other s i t e s on the basis of radiocarbon dates and the Mazama tephra. None of these s i t e s has older e a r l i e r deposits, so earthflows prob-ably i n i t i a t e d here in response to the cooling and moistening climate a f t e r 6,600 BP. Four s i t e s show major reactivations around 2,400 BP, marked by the Bridge River tephra, and t h i s l a t t e r peak of a c t i v i t y can be correlated with a Neoglacial ice maximum of 2-3,200 BP. This analysis of c l i m a t i c and movement chronologies shows that, at three d i f f e r e n t timescales, earthflow movement can be correlated with p e r i -ods of above-average p r e c i p i t a t i o n . However, the analysis has broader im-p l i c a t i o n s because recent earthflow a c t i v a t i o n s (those of the l a s t 30 years), and also earthflow movements over the past 6,000 years, co r r e l a t e broadly with periods of g l a c i a l expansion. 8.6 Further research Within the study area there i s some p o t e n t i a l for extending the chronology presented i n t h i s t h e s i s . Some s i t e s (notably the Burkholder, Canyon and Lac l a Mer flows) were not v i s i t e d ; a l l have depressions with the po t e n t i a l to y i e l d chronologic information. At other s i t e s (Red Mountain (A), Pavilion) l i m i t a t i o n s on a v a i l a b l e coring devices r e s t r i c t e d the value of 277 chronologic information analysed i n th i s study. F i n a l l y , as most earthflows l i e within the Bridge River tephra plume, i t may be possible to develop f i e l d or laboratory techniques for recognising whether tephra i s present or absent i n the earthflow s o i l p r o f i l e s , thereby dating movement at many s i t e s as post or predating 2,400 BP. The analysis i n t h i s thesis demonstrates that, within south-central B r i t i s h Columbia, earthflows are s e n s i t i v e to small c l i m a t i c f l u c t u a t i o n s . Indeed, within the study area they have yi e l d e d a movement chronology which broadly p a r a l l e l s that obtained from alpine g l a c i a l expansion. Moreover, where both g l a c i a l and earthflow chronologies are a v a i l a b l e , earthflows oc-curring below t r e e l i n e provide many more s i t e s with the p o t e n t i a l f o r re-covering datable material than do g l a c i e r s . As earthflow groups are found i n many other areas with a paucity of chronologic data, they may be used to provide a c l i m a t i c record analogous to that obtained from g l a c i a l f l u c t u -ations. 278 BIBLIOGRAPHY Al e s t a l o , J . , 1971, Dendrochronological i n t e r p r e t a t i o n of geomorphic proc-esses: Fennia, 105, pp 1- 140. A l l e y , N., 1976, The palynology and paleoclimatic s i g n i f i c a n c e of a dated core of Holocene peat, Okanagan v a l l e y , southern B r i t i s h Columbia: Can. J. Earth Science, 13, pp 1131-1144. Arber, M.A., 1940, The coastal landslips of south-east Devon: Proc. Geol.  Ass., LI, pp 257-271. Arber, M.A., 1941, The coastal landslips of west Dorset: Proc. Geol. Ass•, LII, pp 273-283. Arber, M.A., 1973, Landslips near Lyme Regis: Proc. Geol. Ass., 84, pp 121-133. Arkley, R.J., 1962, S o i l Survey of Merced Area, C a l i f o r n i a , U.S. Dept. Ag.  S o i l Survey, 7, 131 pp. Auger, P. and Mary, G. , 1968, Glissements et coulees boueuses en Basse-Normandie: Rev. Geog. Phy. Geol. Dyn., 10, pp 213-224. Balteanu, D. , 1976, Two case studies of mudflows i n the Buzau Subcarpathians: Geog. Ann., 58A, pp 165-171. Banner, A., Pojar, J. and Rouse, G.E., 1983, P o s t g l a c i a l paleoecology and successional relationships of a bog woodland near Prince Rupert, B r i t i s h Columbia: Can. J. Forest Research, 15, pp 938-947. Barnosky, C.W., 1983, Late Pleistocene and early Holocene environmental h i s t o r y of southwestern Washington state, U.S.A.: Can. J. Earth S c i . , 21, pp 619-629. Barnosky, C.W., 1985a, Late Quaternary vegetation i n the southwestern Columbia Basin, Washington: Quaternary Research, 23, pp 109-122. Barrett, G., 1979, Changes i n the discharge of selected r i v e r s i n B r i t i s h Columbia during the period of instrumental records: B.Sc. t h e s i s , University of B r i t i s h Columbia. Barton, M.E., 1973, The degradation of the Barton Clay c l i f f s of Hampshire, Q. J l . Eng. Geol., 6, pp 423-440. Becker, B., 1978, Dendrochronological evidence for Holocene changes i n the drainge system of south c e n t r a l Europe i n Dendrochronology i n Europe, ed. J. Fletcher. Begbie, M.P., 1871, On the 'benches' or v a l l e y terraces of B r i t i s h Columbia: Proc. Roy. Geogr. S o c , Session 1870-1871, 15, pp 133-145. 279 Benson, W.N., 1940, Landslides and a l l i e d features i n the Dunedin d i s t r i c t i n r e l a t i o n to geological structure, topography, and engineering: Trans.  Roy. Soc. New Zealand, 70, pp 249-263. Bevier, M.L., 1983, Regional stratigraphy and age of C h i l c o t i n Group basalts, south-central B r i t i s h Columbia: Can. J. Earth S c i . , 20, pp 515-524. Birkeland, P.W., 1974, Pedology, weathering and geomorphological reseach: Oxford University Press, London. Birks, H.J.B. and Birks, H.H., 1980, Quaternary paleoecology: Edward Arnold, London. Bishop, A.W., 1955, The use of the s l i p c i r c l e i n the s t a b i l i t y analysis of slopes: Geotechnique, 5, pp 7-17. Black, C.A., 1965, Methods of s o i l analysis: Am. Soc. Agronomy. Blackwelder, E., 1912, The Gros Ventre s l i d e , an active earthflow: B u l l .  Geol. Soc. Am., 23, pp 487-492. ... Bovis, M.J., 1980, The character and extent of earthflow-type mass movement, Fraser Valley, B r i t i s h Columbia, Canada (Extended abs t r a c t ) : International  Geographical Union Symposium on f i e l d experiments i n geomorphology, Kyoto University, Japan, August 25-30, 1980. Bovis, M.J., 1985, Earthflows i n the I n t e r i o r Plateau, southwest B r i t i s h Columbia: Can. Geotech. J . , 22, pp 294-312. Brindley, G.W. and Brown, G., 1980, C r y s t a l structures of clay minerals and  t h e i r X-ray i n t e r p r e t a t i o n : Mineralogical Society, London, 495 pp. Bromhead, E.N., 1978, Large landslides i n London clay at Heme Bay, Kent: Q. J l . Eng. Geol., 11, pp 291-304. Brunsden, D., 1979, Mass movement: i n Processes i n geomorphology, edited by C.Embleton and J.B. Thornes, Edward Arnold, pp 130-186. Brunsden, D., 1984, Mudslides: i n Slope I n s t a b i l i t y , edited by D. Brunsden and D.B. Pri o r , John Wiley and Sons Ltd.. Brunsden, D. and Jones, D.K.C., 1972, The morphology of degraded landslide slopes i n south west Dorset: Q. J l . Eng. Geol., 5, pp 205-222. Brunsden, D. and Jones, D.K.C., 1974, The evolution of landslide slopes i n Dorset: P h i l . Trans. Royal Soc. London, A, 283, pp 605-631. C a l i f o r n i a Department of Water Resources, 1970, Middle Fork Eel River land-s l i d e investigation: C a l i f o r n i a Department of Water Resources, Memorandum Report, 120 pp. Campbell, A.P., 1966, Measurement of movement of an earthflow: S o i l and  Water, 2, pp 23-24. Church, B.N., 1980, Exploration for gold i n the Black Dome Mountain area (920 7E, 8W): i n Geological fieldwork, 1979, B.C. Minist r y of Energy, Mines and Petroleum Resources, Paper 1980-1, pp 52-54. 280 Church, B.N., 1982, The Black Dome Mountain g o l d - s i l v e r prospect (920 7E 8W) : in Geological fieldwork, 1981, B.C. Ministry of Energy, Mines and Petroleum Resources, Paper 1982-1 pp 106-108. Church, M. and Ryder, J.M., 1972, P a r a g l a c i a l sedimentation: a consideration of f l u v i a l processes conditioned by g l a c i a t i o n : Geol. Soc. Am. B u l l . , 83, pp 3059-3072. Clague, J . J . , 1980, Late Quaternary geology and geochronology of B r i t i s h Columbia. Part 1: Radiocarbon dates: Geological Survey of Canada Paper, 80-13, 28pp. Clague, J . J . , 1981, Late Quaternary geology and geochronology of B r i t i s h Columbia. Part 2: Summary and discussion of radiocarbon-dated Quaternary history: Geological Survey of Canada Papaer, 80-35, 41 pp. Cooksley, J . W. , 1964, Geophysical i n v e s t i g a t i o n of a landslide employing seismic exploration technique: Engineering and S o i l s Engineering 2nd. Sym-posium, Pocatello, Idaho. Cote, W.A, and Day, A.C., 1965, Anatomy and u l t r a s t r u c t u r e of reaction wood: in C e l l u l a r u l t r a s t r u c t u r e of woody plants, edited by Cote, W.A. and Day, A.C., Syracuse Un i v e r s i t y Press. Craig, D., 1981, Mudslide plug flow within channels: Engineering Geology, 17, pp 273-281. Craig, R.F., 1978, S o i l mechanics: Van Nostrand Reinhold Co. Ltd, Wokingham, England. 318p. Crandell, D.R. and Varnes, D.J., 1961, Movement of the Slumgullion earthflow near Lake C i t y , Colorado: USGS Prof. Paper, 424-B, pp 136-139. Cropper, J.C., 1984, Relationships among t r e e - r i n g width chronologies for Alaska and the Yukon. A r c t i c and Alpine Research, 16, pp 243-250 Cross, W., 1924, H i s t o r i c a l sketch of the landslides of G a l l i a r d cut, Ap-pendix A i n Report of the Committee of the National Academy of Sciences on Panama Canal S l i d e s : National Academy of Sciences Memoir 18, pp 23-43. Cruden, D.M., 1984, Landslide problems i n the Canadian C o r d i l l e r a : IVth  International Symposium on Landslides, Toronto, Ontario. Dawson, G.M., 1896, Report on the area of the Kamloops map sheet: B.C.  Geological Survey of Canada, Annual Report for 1894, Part B, pp 1-427. Deere, D.U. and Patton, F.D., 1971, Slope s t a b i l i t y i n residual s o i l s : 4th  Panamerican Conference on S o i l Mechanics and Foundation Engineering, Puerto Rico, pp 87-170. Drew, L.G., (ed), 1975, Tree-ring chronologies of western North America: Chronology series 1, Laboratory of t r e e - r i n g research, University of Arizona, Tucson. 281 D u f f e l l , S. and McTaggart, K.C., 1952, Ashcroft map area, B.C.: Geol. Surv.  Can. Memoir, 262, 122 pp. Duford, J.M., 1976, Late Pleistocene and Holocene cirque g l a c i a t i o n s i n the Shuswap Highland area, B r i t i s h Columbia: M.Sc. t h e s i s , unpublished, Uni-v e r s i t y of Calgary, Calgary, Alberta. Duford, J.M. and Osborn, G.D., 1978, Holocene and l a t e Pleistocene cirque g l a c i a t i o n s i n the Shuswap highland, B r i t i s h Columbia: Can. J. Earth S c i . , 15, pp 865-873. Dunkerly, D.L., 1976, A study of long-term slope s t a b i l i t y i n the Sydney Basin, A u s t r a l i a : Eng. Geol., 10, pp 1-12. Eisbacher, G.H., 1979, F i r s t - o r d e r r e g i o n a l i s a t i o n of landslide character-i s t i c s i n the Canadian C o r d i l l e r a : Geoscience Canada, 6, pp 69-79. Erdtman, G., 1945, An introduction to p o l l e n a n a l y s i s ; Chromica Botanica Company, Waltham, Mass., U.S.A., 239 pp. Evans, S.G., 1982, The development of Big S l i d e , near Quesnel, B r i t i s h Columbia, between 1953 and 1982: Geoscience Canada, 9, pp 220-222. Ferguson, C.W., 1970, Concepts and techniques of dendrochronology: i n Berger, R., editor, S c i e n t i f i c methods i n medieval archaelogy, Berkeley, Uni v e r s i t y of C a l i f o r n i a Press. Fraser, G.D., Waldrop, H.A. and Hyden, H.J., 1969, Geology of the Gardiner Area, Park County, Montana: Geol. Surv. B u l l . 1277 118 pp. Freeze, R.A. and Cherry, J.A., 1979, Groundwater, Prentice-Hall Inc., Eaglewood C l i f f s , New Jersey, 07632. Freeze, R.A. and Witherspoon, P.A., 1967, T h e o r e t i c a l analysis of regional groundwater flow. 2. E f f e c t s of water-table configuration and subsurface permeability v a r i a t i o n : Wat. Res. Res., 3, pp 623-634. Freeze, R.A. and Witherspoon, P.A., 1968, T h e o r e t i c a l analysis of regional groundwater flow. 3. Q u a l i t a t i v e i n t e r p r e t a t i o n s : Wat. Res. Res. 4, pp 581-590. F r i t t s , H.C., 1976, Tree rings and climate: Academic Press, London. F r i t t s , H.C., Lofgren, G.R. and Gordon, G.A., 1979, Variations i n climate since 1602 as reconstructed from tree rings: Quaternay Research, 12, ppl8-46. Fukuoka, M., 1953, Landslides i n Japan: Proc. 3rd. Int. Conf. Mech. Fon.  Engng, Zurich, Volume 2, pp 234-238. Fulton, R.J., 1967, Deglaciation studies i n Kamloops region, an area of moderate r e l i e f , B r i t i s h Columbia: Geological Survey of Canada, B u l l , 154, 36 pp. Fulton, R.J.,1969, G l a c i a l lake h i s t o r y , southern i n t e r i o r plateau, B r i t i s h Columbia: Geol. Surv. Can. Paper 69-37, 14 pp. 282 Fulton, R.J., 1971, Radiocarbon geochronology of southern B r i t i s h Columbia: Geol. Surv. Canada Paper, 71-37, 28 pp. G a l l i e , T.M. I l l , 1983, Chemical denudation and hydrology near tree l i m i t ,  Coast Mountains, B r i t i s h Columbia. Ph.D. t h e s i s , U n i v e r s i t y of B.C., 262 pp. G i l , E. and Kotarba, A., 1977, Model of s l i d e slope evolution i n Flysch Mountains: Catena, 4, pp 233-248. G i l e , L.H. and Hawley, J.W., 1966, Periodic sedimentation and s o i l formation on an a l l u v i a l fan piedmont i n southern New Mexico: Proc. S o i l S c i . Soc.  Am., 30, pp 261-268. Goudie, A., 1984, The nature of the environment: an advanced phy s i c a l ge- ography: Blackwell, Oxford Grove, A.T., 1952, Account of a mudflow on Bredon H i l l , Worcestershire, A p r i l 1951: Proc. Geol. "Ass., 64, pp 10-13. Hadley, J.B., 1964, Landslides and re l a t e d phenomena accompanying the Hebgen Lake earthquake of August 17, 1959: USGS Prof. Paper 435-K. Hansen, H.P., 1947, Climate versus f i r e and s o i l as factors i n p o s t g l a c i a l forest succession i n the Puget Lowland of Washington: Am. J . S c i . , 245, pp 265-286. Hansen, B.S. and Easterbrook, D.J., 1974, Stratigraphy and palynology of lat e Quaternary sediments i n the Puget lowland, Washington: B u l l . Geol. Soc. Am., 85, pp 587-602. Harden, D.R., Janda, R.J. and Nolan, K.M., 1975, Mass movement and storms i n the drainage basin of Redwood Creek, Humboldt County, C a l i f o r n i a - a progress report: USGS Open F i l e Report, 78-486, 161 pp. Hawkins, A.B., 1973, The geology and slopes of the B r i s t o l region: Q. J l . Engng. Geol.., 6, np 185-205. Hays, W.L. and Winkler, L., 1970, S t a t i s t i c s : p r o b a b i l i t y , inference and  decision, New York: Holt, Rinehart and Winston. Hazell, S.D., 1978, Two lat e Quaternary p o l l e n diagrams from southeast B r i t i s h Columbia: Am. Quat. Ass. Abstracts, 5th b i e n n i a l conference, Edmonton, Alberta. Hebda, R.J., 1982, P o s t g l a c i a l h i s t o r y of grasslands of southern B r i t i s h Columbia and adjacent regions: i n Grassland Ecology and C l a s s i f i c a t i o n ; Symposium Proceedings. B r i t i s h Columbia Ministry of Forests. Edited by: A.C. Nicholson, A. McLean, and T.E. Baker, pp 157-191. Hebda, R.J. and Mathewes, R.W., 1984, Holocene h i s t o r y of cedar and native Indian cultures of the North P a c i f i c coast: Science, 225, pp 711-712. Heginbottom, J.A., 1972, S u r f i c i a l geology of Taseko Lakes map area (92-0) B r i t i s h Columbia: Geol. Surv. Canada Paper 72-14, pp 9. 283 Heusser, C.J., 1960, Late Pleistocene environments of P a c i f i c North America: Am. Geogr. Soc. Spec. Pub., 35, 308 pp. Heusser, C.J., 1973, Environmental sequence following the Fraser advance of the Juan de Fuca Lake, Washington: Quaternary Reasearch, 8, pp 284-306. Heusser, C.J., 1974, Quaternary vegetation, climate and g l a c i a t i o n of the Hoh River Valley, Washington: B u l l . Geol. Soc. Am., 85, pp 1547-1560. Heusser, C.J., 1977, Quaternary palynology of the P a c i f i c slope of Washington: Quaternary Research, 8, pp 282-306. Hitchcock, C.L. and Cronquist, A., 1973, F l o r a of the P a c i f i c northwest: Seattle, University of Washington Press. Hodge, A.L. and Freeze, R.A., 1977, Groundwater flow systems and slope s t a -b i l i t y : Can. Geotech. J . , 14, pp 466-476. Holland, S.S., 1964, Landforms of B r i t i s h Columbia, B.C. Dept. of Mines and  Pet. Res. B u l l . , 48, 138 pp. Hosie, R.C, 1969, Native trees of Canada, 7th e d i t i o n , Queen's Printer for Canada, Ottawa. Howe, E., 1909, Landslides i n the San Juan Mountains, Colorado: U.S. Geol.  Surv. Prof. Paper, 67, 58 pp. Huber, B., 1970, Dendrochronology: i n Dendrochronology i n Europe, edited by J. Fletcher. Hutchinson, J.N., 1970, A coastal mudflow on the London Clay C l i f f s at Beltinge, North Kent: Geotechnique, 20, pp 412-438. Hutchinson, J.N., 1973, The response of London clay c l i f f s to d i f f e r i n g rates of toe erosion: Geologia Applicata e Idrogeologia, B a r i , 8, 221-239. Hutchinson, J.N. and Bbandari, R.K., 1971, Undrained loading, a fundamental mechanism of mudflows and other mass movements: Geotechnique, 21, pp 353-358. Hutchinson, J.N., P r i o r , D.B. and Stephens, N., 1974, P o t e n t i a l l y dangerous surges i n an Antrim mudslide: Q. J l . Engng. Geol., 7, pp 363-376. Iverson, R.M. and Major, J . J . , 1987, R a i n f a l l , groundwater flow and seasonal movement at Minor Creek landslide, northeastern C a l i f o r n i a : Geol. Soc. Am.  B u l l . , 99, pp 579-594. Jeletzky, J.A. and Tipper, H.W., 1978, Upper J u r a s s i c and Cretaceous rocks of Taseko Lakes map area and t h e i r bearing on the geological h i s t o r y of southwestern B r i t i s h Columbia: Geol. Surv. Canada Paper, 67-54. Johnson, A.M., 1970, Physical Processes i n Geology: San Francisco, Freeman, Cooper, 577p. Jozsa, L.A., Oguss, E., Bramhall, P.A. and Johnson, S.G., 1984, Dendroclimatological analysis of white spruce from 5 t r e e - r i n g s i t e s i n the 284 Yukon t e r r i t o r y : Environment Canada Nat. Hydrology Res. Inst. Snow and i c e d i v i s i o n . Kapp, R.O., 1969, How to know po l l e n and spores: W.C. Brown and Co., Dubuque, Iowa. Keefer, D.K., 1977, Earthflow: PhD t h e s i s , Stanford University, Stanford, C a l i f o r n i a . Keefer, D.K. and Johnson, A.K., 1983, Earth flows: morphology, mobilization, and movement: USGS Prof. Paper, 1264, 56 pp. Kelsey, H.M., 1977, Landsliding, channel changes, sediment y i e l d and land use i n the Van Duzen River Basin, north coastal C a l i f o r n i a , 1941-1975: PhD  t h e s i s , Santa Cruz, University of C a l i f o r n i a , Santa Cruz. Kelsey, H.M., 1978, Earthflows i n Franciscan melange, Van Duzen River Basin, C a l i f o r n i a : Geology, 6, pp 361-364. Kenney, T.C., 1967, The influence of mineral composition on the r e s i d u a l strength of natural s o i l s : Geotechnical Conference, Oslo, Norway, 1967, Proceedings, v 1, pp 123-129. King, M. , 1980, Palynological and macrofossil analysis of lake sediments from the L i l l o o e t area, B r i t i s h Columbia: M.Sc t h e s i s , Simon Fraser University. Krajina, V.J., 1969, Ecology of trees i n B r i t i s h Columbia: Ecology of  Western North America, 2, pp 1-146. Kuivenen, K.C. and Lawson, M.P., 1982, Dendroclimatological analysis of b i r c h i n South Greenland. A r c t i c and Alpine Research, 14, pp 243-250 Lambe, T.W. and Whitman, R.V., 1951, S o i l t e s t i n g for engineers: John Wiley and Sons, New York, 165p. Lamarch, V.C., J r . and Mooney, H.A., 1967, Altithermal timberline advance i n the western United States, Nature, 213, 980-982. Lang, W.D. , 1943, Geological notes: Proc. Dorset nat. Hist. Archaeol. S o c , 64, 129-130. Lang, W.D. , 1944, Geological notes: Proc Dorset nat. H i s t . Archael. S o c , 65, 147-149. Lang, W.B., 1945, Report on Dorset natural h i s t o r y for 1944 (except b i r d s ) : Proc. Dorset nat. Hist, archaeol. S o c , 66, pp 127-129. Leech, G.B., 1953, Geology and mineral deposits of the Shulaps Range: B u l l .  B r i t . Col. Dept. Mines, 32. Liese, W. , 1978, Bruno Huber: Pioneer of European dendrochronology: i n Dendrochronology i n Europe edited by J . Fletcher. Lowden, J.A. and Blake, W. , 1976, Geological Survey of Canada radiocarbon dates XVI: Geol. Surv. Canada Paper 76-7, 21 pp. 285 Luckman, B.H., Harding, K.A. and Hamilton, J.P., 1987, Recent g l a c i a l ad-vances i n the Premier Range, B r i t i s h Columbia: Can. J. Ear. S c i . , 24, pp. 1149-1161. Lowden, J.A. and Blake, W. , 1978, Geological Survey of Canada radiocarbon dates XVIII: Geol Surv. Canada Paper, 78-7, 20pp. Lyons, CP., 1965, Trees, shrubs and flowers to know i n B r i t i s h Columbia. Toronto, Dent, 194p. Mack, R.N., Bryant, V.M. and F r y x e l l , R. , 1976, Pollen sequence from the Columbia r i v e r basin, Washington: reappraisal of p o s t g l a c i a l vegetation: The American Midland N a t u r a l i s t , 95, pp 390-397. Mack, R.W., Rutter, N.W., Bryant, V.M. and Valastro, S., 1978a, Late Quaternary pollen record from Big Meadow, Pend O r e i l l e county, Washington: Ecology, 59, pp 956-966. Mack, R.N., Rutter, N.W. and Valastro, S., 1978b, Later Quaternary po l l e n record from the Sanpoil River Valley, Washington: Can. J . Bot., 56, pp 1642-1650. Mack, R.N., Rutter, N.W. and Valastro, S., (1979), Holocene vegetation h i s -tory of the Okanagan v a l l e y , Washington: Quat. Res. 12, pp 212-225. Mathewes, R.J., 1973, A pal y n o l o g i c a l study of p o s t g l a c i a l vegetation changes i n the University Research Forest, southwestern B r i t i s h Columbia.: Can. J. Bot., 51, pp 2085-2103. Mathewes, R.W., 1984, Paleobotanical evidence f o r c l i m a t i c change i n south-ern B r i t i s h Columbia during l a t e - g l a c i a l and Holocene time: Syllogeus, 55, pp 397-422. Mathewes, R.W., 1984, Holocene h i s t o r y of cedar and native Indian cultures of the North American P a c i f i c coast: Science, 225, pp 711-712. Mathewes, R.W. and Heusser, L.E., 1981. A 12,000 year palynological record of temperature and p r e c i p i t a t i o n trends i n southwestern B r i t i s h Columbia: Can. J. Botany, 59, pp 707-710. Mathewes, R.W. and Rouse, G.E., 1975, Palynology and paleoecology of Po s t g l a c i a l Sediments from the Lower Fraser River Canyon of B r i t i s h Columbia: Can. J. Earth S c i . , 12, 745-756. Mathewes, R.W. and Westgate, J.A., 1980, Bridge River tephra: revised d i s -t r i b u t i o n and s i g n i f i c a n c e for detecting o l d carbon errors i n radiocarbon dates of limnic sediments i n southern B r i t i s h Columbia: Can. J. Earth S c i , 17, pp 1454-1461. Mathews, W.H., 1951, H i s t o r i c and p r e h i s t o r i c fluctuations of alpine f l a c i e r s i n the Mount G a r i b a l d i map-area, southwestern B r i t i s h Columbia: J. Geology, 59, pp 357-380. Mathews, W.H., 1968, Geomorphology, Southwestern B r i t i s h Columbia i n Guide-book for Geological F i e l d t r i p s i n southwest B r i t i s h Columbia, p 18-24, De-partment of Geology, Uni v e r s i t y of B.C. Report no.6. 286 Mathews, W.H. and Rouse, G.E., 1984, The Gang Ranch-Big Bar area, south-ce n t r a l B r i t i s h Columbia: stratigraphy, geochronology, and palynology of the Tertary beds and t h e i r r e l a t i o n s h i p to the Fraser Fault: Can. J. Earth S c i . , 21, 1132-1144. McGreal, W.S. and Craig, D., 1977, Mass-movement a c t i v i t y : an i l l u s t r a t i o n of d i f f e r i n g responses to groundwater conditions from two s i t e s i n Northern Ireland: I r i s h Geogr., 10, pp 28-35. McSaveney, M.J., 1978, Sherman G l a c i e r rock avalanche, Alaska, i n Voight, B., ed., Rockslides and avalanches. E l s e v i e r , Amsterdam, v o l 1, pp 197-258. Monger, J.W.H., 1982, Geology of Ashcroft map area, southwestern B r i t i s h Columbia: i n Current Research, Part A, Geological Survey of Canada, Paper 82-1A, pp 293-297. Monger, J.W.H. and McMillan, W.J., 1984, Bedrock geology of Ashcroft (92-1) map area, Geol. Surv. Canada, Open f i l e 980. Moorman, R.F., 1939, Notes oh the p r i n c i p a l 'mud g l a c i e r ' at Hamstead: Proc.  I s l e of Wight Nat. H i s t . Archaeol. S o c , 3, p 417. Mott, R.J. and Jackson, L.E., 1982, An 18,000 year p a l y n o l o g i c a l record from the south Alberta segment of the c l a s s i c a l Wisconsinan " i c e - f r e e Corridor": Can. J. Earth S c i . , 19, pp 504-513. Nadler, D.E., 1984, Groundwater hydrology and slope movement at Pa v i l i o n , B r i t i s h Columbia: unpublished M.S