Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation of the stability of the steepland forest soils in the coast mountains, southwest British… O'Loughlin, Colin Lockhart 1972

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

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

Item Metadata

Download

Media
831-UBC_1973_A1 O46.pdf [ 11.57MB ]
Metadata
JSON: 831-1.0101126.json
JSON-LD: 831-1.0101126-ld.json
RDF/XML (Pretty): 831-1.0101126-rdf.xml
RDF/JSON: 831-1.0101126-rdf.json
Turtle: 831-1.0101126-turtle.txt
N-Triples: 831-1.0101126-rdf-ntriples.txt
Original Record: 831-1.0101126-source.json
Full Text
831-1.0101126-fulltext.txt
Citation
831-1.0101126.ris

Full Text

(51MI AN- INVESTIGATION OF THE STABIL ITY OF THE STEEPLAND FOREST SOILS IN THE COAST MOUNTAINS, SOUTHWEST BRITISH COLUMBIA by COLIN LOCKHART O'LOUGHLIN M.Sc. ( G e o l . ) , U n i v e r s i t y o f C a n t e r b u r y ( N Z ) , 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY ( F o r e s t H y d r o l o g y ) i n t h e F a c u l t y o f F o r e s t r y We a c c e p t t h i s t h e s i s as c o n f o r m i n g t o t h e r e q u i r e d s t a n d a r d THE UNIVERSITY OF BRITISH COLUMBIA O c t o b e r 1972 In p re sent ing t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree a t the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r re fe rence and s tudy. I f u r t h e r agree t h a t permis s ion f o r ex ten s i ve copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l ga in s h a l l not be a l lowed wi thout my w r i t t e n pe rm i s s i on . Department o f Fo re s t r y  The U n i v e r s i t y o f B r i t i s h Columbia, Vancouver 8, Canada. ABSTRACT In B r i t i s h Columbia's Coast Mountains a large area of forested 2 steepland of probably more than 100 km , i s c l ea r f e l l ed annually by high-lead logging methods. Although the deforestation of steep slopes has been shown to ser iously accelerate mass wasting in Alaska, Oregon and elsewhere in the United States, the effects of c l e a r f e l l i n g on slope s t a b i l i t y are largely unknown for coastal B r i t i s h Columbia. The objective of th is study was to determine the extent and seriousness of lands! iding on undisturbed, forested slopes and on c l ea r f e l l ed slopes and elucidate some of the natural and human-caused factors which are detrimental to slope s t a b i l i t y in the Coast Range. 2 Within a study area of 640 km , 77 large landsl ides of the debris avalanche or debris s l ide type were discovered. Landslides not associated with roads were predominantly confined to long, uniform slopes of over 30° underlain by poorly-drained podsolic s o i l s . Slopes underlain by shallow, regosol ic s o i l s were r e l a t i v e l y res i s tant to mass wasting. Large landsl ides were more frequent on c l ea r f e l l ed areas than on undis-turbed slopes. Road construction, which was responsible for 14 large landsl ides and more than 100 smaller f a i l u r e s , appeared to be more d e t r i -mental to the s t a b i l i t y of the Coast Range slopes' than other a c t i v i t i e s carr ied on by man. A network of simple piezometers established in steep drainage depressions revealed that the piezometric surface within the s o i l mantle approached the ground surface during ra in storm or snow melt periods to cause marked increases in pore water pressures. Curv i l inear relat ionships between the piezometric head and the da i l y (24 hr.) r a i n f a l l indicated that da i l y r a i n f a l l s which exceeded approximately 120 mm caused complete i i saturation of most drainage depression s o i l s . Pore water pressures 2 larger than 800 kg/m were recorded at the base of the s o i l mantle. Such pressures s i g n i f i c an t l y reduce the e f fec t i ve normal stress acting on potential f a i l u r e planes at the so i l -basa l t i l l or soil-bedrock i n t e r -face and decrease the s t a b i l i t y of the slope. General observations of the growth habits of tree root systems in the steepland so i l s suggested that roots help bind the cohesionless so i l s into a coherent mantle and anchor the mantle to the substratum. Direct shear tests performed in the f i e l d indicated that the s o i l strength was l i nea r l y related to the bulk weight of roots in the s o i l . Where root networks were dense the s o i l ' s shear strength may be increased by several . 2 hundred kg/m compared to so i l s with few or no roots. Under saturated conditions the forest s o i l s ' shear strengths are largely derived from the apparent cohesion provided by the tree root network. Laboratory strength tests of tree roots showed that Douglas f i r and cedar tree roots deteriorate rapidly a f te r death of the parent t ree. Within 3 to 5 years a f te r cutt ing of the parent tree, small roots may lose over hal f of t he i r o r i g ina l tens i l e strength. Five small landslides were investigated with simple, approximate, s t a b i l i t y analyses encompassing a range of possible shear strength and s o i l water conditions. The results confirmed that the s t a b i l i t y of r o a d - f i l l slopes as well as natura l , forest s o i l slopes i s very sens i t ive to changes in the ground water conditions and that the s t a b i l i t y of the forest s o i l mantle on steeper slopes is very dependent on the reinforcement provided by roots. In the l i g h t of these findings i t i s apparent that current forest cutt ing and road bui ld ing practices on the steep slopes of B r i t i s h Columbia's Coast Mountains are not compatible with sensible mountainland management which fosters protection of the s o i l resource. i i i TABLE OF CONTENTS Page INTRODUCTION 1 CHAPTER 1 LANDSLIDES AND THEIR IMPORTANCE IN THE COAST 3 MOUNTAINS Introduction 3 Physical sett ing of the study area 3 F ie ld data co l l ec t i on 12 Use of ae r i a l photographs 13 Types of mass wasting 17 Signif icance of landsl ides in the study area 24 Relationship of landsl ides to s i t e factors 27 Ef fect of c learcutt ing on landsl ide occurrence 33 Road construction and landsl ides 38 Natural revegetation of landsl ide scars 43 Discussion and conclusions 43 CHAPTER 2 PIEZOMETRY STUDIES ON STEEP SLOPES 46 Introduction 46 Piezometric instrumentation 51 Results 55 Relationship between piezometric head, r a i n f a l l g 3 and subsurface flow Estimation of maximum pore water pressures 66 Pore water pressures and e f fec t i ve stresses 68 Conclusions 68 CHAPTER 3 THE IMPORTANCE OF TREE ROOTS TO SLOPE STABILITY 71 Introduction 71 Rooting habits of forest trees on the Coast Range 72 slopes A f i e l d study of the e f fect of tree roots on the 75 s o i l ' s resistance to shear Laboratory test ing of root strength go Summary and conclusions 104 i v Page CHAPTER 4 A DETAILED EXAMINATION OF SELECTED LANDSLIDES 106 Introduction 106 S t a b i l i t y analysis 107 Landslide 1 (road-cut slope) 110 Landslide 2 (road-cut slope) 114 Landslide 3 ( r o a d - f i l l slope) 117 Landslide 4 ( r o a d - f i l l slope) 121 Landslide 5 (natural slope) 124 Concluding remarks 127 BIBLIOGRAPHY 130 APPENDIX 1 DETAILS OF MAJOR STORM RAINFALLS AT SEYMOUR FALLS AND CLEVELAND DAM, SOUTHWESTERN BRITISH 138 COLUMBIA, 1959-1970 APPENDIX 2 TECHNIQUES FOR CALCULATING VOID RATIOS AND SATUR- ^ ATED UNIT WEIGHTS APPENDIX 3 DETAILS OF SLOPE ANALYSES 141 LIST OF TABLES Table Page 1 Some mean physical properties of steepland so i l s 9 and unweathered t i l l from the Coast Range, southwestern B.C. 2 Details of high a l t i tude color and color infrared 14 photography taken 16-7-71 over southwestern B.C. 3 Information obtainable from various types of 19 aer ia l photographs. 4 A summary of landsl ide survey data co l lected in 25 the B r i t i s h Columbia Coast Range, 1970. 5 Landslide number, area and quantity of debris moved 26 2 per km of the study area 6 Numbers of landsl ides before and af ter logging, Coast 34 Range, southwestern B.C. 7 Landslide densit ies and quantit ies of landsl ide debris produced within selected watershed areas in the Coast Range, southwestern B.C. 36 8 Landslide densit ies on forested and c l ea r f e l l ed 37 mountain slopes for regions in northwestern North America and northern Japan. 9 Maximum piezometric head and equivalent pore water pressure recorded at piezometer stations in the 56 Coast Range, southwestern B.C. 10 Regression equations describing relat ionships between piezometric head (Y) and da i l y r a i n f a l l (X), Coast 52 Range, southwestern B.C. 11 Saturated permeability coef f i c ient s of steepland 55 subsoi l s , Coast Range, southwestern B.C. 12 Results of d i rect shear tests on so i l s containing si roots. 13 14 15 16 17 18 19 20 21 22 R e s u l t s o f d i r e c t s h e a r t e s t s on s t e e p l a n d s o i l s w i t h no r o o t s . 2 D i a m e t e r (cm) and t e n s i l e s t r e n g t h (kg/cm ) o f r o o t s sampled f r o m l i v i n g t r e e s and f r o m stumps o f f o r m e r t r e e s . R e s u l t s o f a n a l y s e s o f v a r i a n c e and m u l t i p l e range t e s t s f o r Doug las f i r and w e s t e r n r e d c e d a r r o o t s t r e n g t h d a t a . T o t a l d e f o r m a t i o n A l (cm) a t t h e r u p t u r e p o i n t 2 and modulus o f e l a s t i c i t y E (kg/cm ) f o r c e d a r and Doug la s f i r r o o t s t e s t e d i n t e n s i o n . Means, ex t remes and s t a n d a r d d e v i a t i o n s o f d i a m e t e r s o f r o o t s g r ow ing i n s u b s o i l s be low 45 cm a t l a n d s l i d e h e a d s c a r p s . F a c t o r s o f s a f e t y (FS) c a l c u l a t e d by t h e o r d i n a r y method o f s l i c e s (OMS) and by t h e B i s h o p method (BMS) f o r a r o a d - c u t s l o p e a t t h e s i t e o f L a n d s l i d e 1. F a c t o r s o f s a f e t y (FS) c a l c u l a t e d by t h e o r d i n a r y method o f s l i c e s (OMS) and by t h e B i s h o p method (BMS) f o r a r o a d - c u t s l o p e a t t h e s i t e o f L a n d s l i d e 2. F a c t o r s o f s a f e t y c a l c u l a t e d by t h e o r d i n a r y method o f s l i c e s f o r a r o a d - f i l l s l o p e a t t h e s i t e o f L a n d s l i d e 3. F a c t o r s o f s a f e t y c a l c u l a t e d by t h e o r d i n a r y method o f s l i c e s f o r a r o a d - f i l l s l o p e a t t h e s i t e o f L a n d s l i d e 4 . F a c t o r s o f s a f e t y and t h e a p p a r e n t c o h e s i o n r e q u i r e d f o r s t a b i l i t y f o r a n a t u r a l s l o p e a t t h e s i t e o f L a n d s l i d e 5. Page 87 96 98 100 103 113 116 121 123 126 v i i LIST OF FIGURES Figure Page 1 Map of study area showing locat ion of 4 lands!ides. 2 Oblique color aer ia l photograph of study area. 7 3 Pa r t i c l e s ize d i s t r i bu t i on curves for forest ^ s o i l B horizons. 4 Stereogram of part of the lower Capilano catchment. 15 5 Color infrared stereogram of part of the middle -jg Seymour Val ley. 6 Small debris s l i de in upper Harvey Creek catchment. 21 7 An almost ve r t i ca l headscarp of a large debris avalanche in the middle Seymour catchment. 8 The upper part of a large debris avalanche in the 22 Seymour catchment. 9 A debris avalanche in second growth forest . 22 10 Top D i s t r ibut ion of Landslides with respect to slope 29 Bottom Longitudinal p ro f i l e s of selected lands l ides. 11 Top D i s t r ibut ion of landsl ides with respect to aspect 3 0  Bottom Dis t r ibut ion of landsl ides with respect to a l t i t ude . 12 Smooth-surfaced d i o r i t e bedrock forming the s l i d i ng surface near the head of a debris avalanche in the ^ upper Capilano catchment. 13 Road-caused debris avalanches resu l t ing from the collapse of r o a d - f i l l and s ide-cast materials on a 4 ^ 39° slope in the Howe Sound area. 14 Flow net and neutral stresses for a s loping, i s o -48 t r op i c , saturated so i l mantle rest ing on imper-meable substratum. F i g u r e 15 16 17 18 19 20 21 22 23 24 25 26 27 v i i i Page 54 57 T o p o g r a p h i c and s o i l c h a r a c t e r i s t i c s o f t h e s i t e s 52 i n s t r u m e n t e d w i t h p i e z o m e t e r s . D e t a i l s o f t h e two t y p e s o f p i e z o m e t e r s used i n t h e s t u d y . P i e z o m e t r i c head v a r i a t i o n s d u r i n g 1970, 1971 and 1972 a t s t a t i o n s C l - 4 and C l - 5 , upper C a p i l a n o c a t c h m e n t . P i e z o m e t r i c head r e s p o n s e s a t t h r e e s t a t i o n s i n t h e upper C a p i l a n o c a t c h m e n t d u r i n g a r a i n s t o r m 58 i n Sep tember , 1971 . R e l a t i o n s h i p s between p i e z o m e t r i c head and d a i l y gg r a i n f a l l . R e l a t i o n s h i p s between p i e z o m e t r i c head and d a i l y r a i n f a l 1 . R e l a t i o n s h i p s between p i e z o m e t r i c head and d a i l y r a i n f a l l . H y p o t h e t i c a l r e l a t i o n s h i p between r a i n f a l l , s u b -s u r f a c e seepage and p i e z o m e t r i c h e a d . E f f e c t i v e normal s t r e s s a c t i n g on a p o t e n t i a l s l i d e p l a n e i n t h e c a s e o f (a ) a m o i s t , s l o p i n g s o i l m a n t l e w i t h no seepage and (b) a s a t u r a t e d , s l o p i n g s o i l m a n t l e w i t h s eepage . The exposed r o o t s y s t em o f a r e c e n t l y o v e r t u r n e d -j^ w e s t e r n r e d c e d a r . 60 61 64 69 R e c e n t l y exposed w e s t e r n r ed c e d a r r o o t p e n e t r a t i n g an open j o i n t i n d i o r i t e b e d r o c k . D e t a i l s o f s h e a r b o x c o n s t r u c t i o n and i t s mode o f o p e r a t i o n i n t h e f i e l d . S hea r l o a d vs s h e a r d i s p l a c e m e n t c u r v e s f o r f o u r 82 d i r e c t s h e a r t e s t s on s t e e p l a n d s o i l s c o n t a i n i n g r o o t s . 74 78 Page 88 Results of d i rect shear tests on steepland so i l s plotted on a shear stress (T) vs normal stress (a) diagram. Instron Universal Testing Machine. 93 Details of pneumatic holding clamps. 93 Results of tens i l e strength tests on three Douglas f i r roots in d i f fe rent stages of deter io rat ion. 95 Diagram showing the change in ten s i l e strength of small Douglas f i r and cedar roots with change in the time elapsed since f e l l i n g of the parent t ree. Formation of tension cracks in f i l l materials near the outer edge of an abandoned logging road. Incipient lands l id ing along an abandoned logging road in the Howe Sound area. 97 109 109 Small, c i r c u l a r debris s l i de in a road-cut slope m (Landslide 1). Debris s l i de in a road-cut slope (Landslide 2), upper Magnesia Creek catchment. 115 Small, c i r c u l a r f a i l u r e in a 34° r o a d - f i l l slope 118 (Landslide 3), upper Harvey Creek catchment. Results of d i rect shear tests on r o a d - f i l l materials plotted on a shear stress (T) vs normal stress (a) diagram. Shallow, c i r c u l a r f a i l u r e in a r o a d - f i l l slope (Landslide 4) , upper Magnesia Creek catchment. Longitudinal p r o f i l e of a shallow debris s l i de on a 30°, c l e a r f e l l ed slope (Landslide 5), upper Harvey Creek catchment. X ACKNOWLEDGEMENTS I am indebted to Dr. B.C. Goodel l f o r adv ice and a s s i s t ance dur ing most phases o f t h i s p r o j e c t and to Dr. D.S. Laca te , Dr. A. Kozak, Dr. R.G. Campanella and Dr. 0. Slaymaker f o r c o n s t r u c t i v e comments and adv ice dur ing the p repara t ion of t h i s t h e s i s . I am a l s o indebted to the New Zealand Fores t Se rv i ce f o r g ran t i ng me study leave and f i n a n c i a l support to undertake t h i s p r o j e c t . A d d i t i o n a l f i n a n c i a l support was prov ided by the Department o f the Environment (NCWRR), by the Greater Vancouver Water D i s t r i c t and by the Van Dusen Graduate Fe l l owsh ip i n F o r e s t r y . 1 INTRODUCTION For over 50 years the forested steeplands of B r i t i s h Columbia's Coast Mountains have been managed pr imar i ly for the production of timber. T rad i t i ona l l y , the c learcutt ing of large areas of mature forests with high lead methods, often followed by broadcast burning of the cleared areas, is applied as a general si 1v icu l tura l system. Road bui ld ing usually accompanies c l ea r fe l 1 ing . Under th is type of management large areas of steep slopes, often several square kilometers in extent, may remain in an e s sent ia l l y unvegetated state f o r , perhaps, 5 years or more unt i l natural regrowth vegetation or a r t i f i c i a l l y planted conifer seedlings gradually reestabl i sh a mantle of plant cover. Within the Coastal Forest Zone of B r i t i s h Columbia approximately 2 360 km (90,000 acres) of forest are c l ea r f e l l ed annually (B.C. Forest Service, 1957). Although the records of areas c l earfel fled annually do not include information on land slope, i t i s certa in that a large acreage of steeplands is logged over each year. Possibly a t h i r d or more of the to ta l area cut over annually within the Coastal Forest Zone, some 120 to 140 km2 (30,000 to 40,000 acres) has slopes exceeding 20°. This estimation is regarded as being not unreasonable by Dr. Smith, Faculty of Forestry, Univers ity of B r i t i s h Columbia (personal communication). On steep, forested lands in southeast Alaska, in central Oregon and in northern Idaho, accelerated rates of mass wasting have been associated with c l e a r f e l l i n g (Bishop and Stevens 1964, Dyrness 1967 and Gonsior and Gardner 1971). However, in spite of the importance of c l e a r f e l l i n g on the steeplands of western B r i t i s h Columbia, the effects 2 on s o i l s t a b i l i t y in th i s region have not, to the present w r i t e r ' s knowledge, come under invest igat ion. In f a c t , the p o s s i b i l i t y that c l e a r f e l l i n g on steep slopes may resu l t in accelerated erosion, erosibn in th i s instance being used in i t s broadest sense, has not, in general, been given more than passing at tent ion. Golding (1968) points out that erosion is not generally recognised as a problem in B.C. . . . . " although, he also indicates that erosion on logged areas in the Province may be l o c a l l y severe. Jef f rey (1968) att r ibutes th i s lack of interest in the welfare of the s o i l resource to the fact that land management in western Canada i s so strongly devoted to wood production that other effects of logging, such as increased sediment y i e l d s , are not taken into consideration. This thesis examines the natural s t a b i l i t y of the forested te r ra in over a small portion of the Coast Mountains in southwestern B r i t i s h Columbia and the changes which occur in slope s t a b i l i t y when the forest vegetation is removed from steep slopes. The invest igat ion is wholly concerned with mass wasting, that i s , the downward movement of s o i l and rock materials under the influence of gravity without benefit of the contr ibut ing force of independent agencies such as flowing water or wind (Leopold et al_ 1964). The study should help to place the importance of mass wasting within the Coast Range environment in correct perspective and elucidate some of the natural and human-caused factors which are detrimental to the s t a b i l i t y of steep, forested slopes. 3 CHAPTER 1 LANDSLIDES AND THEIR IMPORTANCE IN THE COAST MOUNTAINS  Introduction During 1970 a combined f i e l d and ae r i a l photograph survey was completed of the types, numbers and volumes of landsl ides on undisturbed forested slopes and c l ea r f e l l ed slopes over a segment of the Coast 2 Mountains. The area investigated forms a 640 km , t r iangu lar block of steep mountain country ly ing between Howe Sound and Indian Arm north of the c i t y of Vancouver (Figure 1). A major question was whether or not the large scale c l e a r f e l l i n g operations, which are employed in the Coast Mountains, ser iously influence the s t a b i l i t y of steep slopes. This chapter presents physiographic information of the area and of indiv idual landsl ide s i te s to provide information on which to base the spec i f i c studies of Chapters 2, 3 and 4. Physical sett ing of the study area Geology. The area investigated forms part of the southern end of a large complex of plutonic and metamorphic rocks that characterises the B r i t i s h Columbian coastal geology. The bedrock is composed of many d i o r i t i c bodies of d i f f e r i n g age and s i ze , numerous i so lated masses of volcanic and sedimentary rocks var iously interpreted as roof pendants or remnants of older te r ra in and, in add i t ion, vast quantit ies of gneiss and migmatite (Armstrong 1965, Roddick 1965). The most common rock types are d i o r i t e and quartz d i o r i t e in which hornblend is Figure 1. Map of study area showing l o c a t i o n o f l a n d s l i d e s . 5 more abundant than b i o t i t e . Armstrong estimates that plutonic rocks underlie approximately 80 percent of the area in the v i c i n i t y of Howe Sound. Natural outcrops and roadcut exposures of the d i o r i t e rock generally display smooth, s u p e r f i c i a l l y weathered surfaces r e l a t i v e l y free of open j o i n t s , ind icat ing that the bedrock i s , for the most part, impermeable. For th i s reason deep seepage i s not considered to be an important part of the hydrology of the va l ley sides within the study area. Exceptions to th is do occur, however, where deeply weathered and fractured breccias, f ine grained sedimentary rocks, and low grade schists belonging to the Gambier Group (Roddick 1965) form the basement rock in the Howe Sound area. Pleistocene g lac ia l ice inundated most of the area covered by the survey except, possibly, the highest peaks above 1,800 meters (Geological Association of Canada 1958). At least three major g lac iat ions (Seymour, Semiamu and Vashon) affected the southern parts of the Coast Range while the existence of a fourth g lac iat ion (Sumas) of more l imited extent has been proposed by Armstrong (1956). The retreat ing ice l e f t widespread t i l l deposits of var iable th i ck -ness f lanking the val ley slopes. Exposures in Howe Sound, Capilano and Seymour val leys reveal t i l l thicknesses in excess of 4 m at a l t i tudes below 700 m but at higher a l t i tudes the t i l l i s thinner and haphazardly d i s t r ibu ted . Except where i t i s weathered, the t i l l i s grey in colour, highly compacted and impermeable to water, presumably a consequence of the great ice pressures which exerted themselves on the va l ley slopes during the g lac ia t ions . The t i l l possesses a hard consistency but tends to rupture along planes approximately para l le l to the slope surface when h i t with a maddock. Nevertheless, i t s high shear strength i s evidenced by i t s obvious s t a b i l i t y . 6 The presence of th i s unweathered t i l l substratum influences the s t a b i l i t y of the s o i l overburden. Not only does the t i l l surface form an impermeable bar r ie r to downward water flow, thereby encouraging saturation of the so i l mantle during storm periods, but i t also determines the general lower l i m i t of root penetration; only occasional ly do s o l i t a r y tree roots enter cracks in the compacted t i l l . Furthermore, the s o i l - t i l l interface is well defined and often smooth in longitudinal p r o f i l e , causing i t to act as a s l i de plane. Details of the f i l l ' s physical properties are presented in Table 1. The topography strongly re f l ec t s the youthfulness of the mountains and the effects of the g lac iat ions (Figure 2). The higher peaks, of maximum a l t i tude 2,025 meters, are connected by uneven ridges interrupted by deep transverse saddles and steep-walled narrow va l leys . The average r e l i e f approximates '1,200 meters. Soi Is. Large and often abrupt var iat ions occur in s o i l depths, p r o f i l e development, textures and drainage conditions from one part of the same slope to another. Lesko (1961) and Brooke (1966) described and c l a s s i f i e d the so i l s of the southern and eastern parts of the area according to the National Soi l Survey Committee's c l a s s i f i c a t i o n system for Canada. More recently a s o i l survey of the area by the B r i t i s h Columbia Department of Agr icu l ture, Soi l Survey Div is ion resulted in the recognition and mapping of more than twenty s o i l ser ies . However, for the purposes of the present landsl ide study, three broad, ea s i l y recog-niseable categories of steepland so i l s were distinguished on the basis of the i r drainage cha rac te r i s t i c s , parent material and p r o f i l e development. Category 1. Soi l s developing on moderately to well drained s i tes  over t i l l , colluvium or al luvium. These so i l s show good horizon develop-F igure 2. An ob l i q ue c o l o r a e r i a l photograph o f study a r ea . IA, Indian Arm; SV, Seymour V a l l e y ; CV, Cap i lano V a l l e y ; HS, Howe Sound. 8 ment with inc ip ient to substantial signs of podsol izat ion. L, F and H horizons together are usually less than 35 cm deep while the Ae horizon, i f present, is less than 5 cm th ick . The B horizon ranges in colour from l i gh t yel lowish brown to dark brown, normally possesses weak crumb or weak blocky st ructure, i s usually a gravel ly or bouldery sandy loam in texture and is f r i a b l e without any noticeable accumulation of c lay. B horizons rare ly exceed 90 cm in thickness. In terms of the Canada So i l Survey Committee's s o i l c l a s s i f i c a t i o n for Canada (Canada Department of Agr icu l ture, 1970) these so i l s range from Orthic Ferro Humic Podsols to Mini Ferro Humic Podsols. Category 2. So i l s developing on poorly drained s i tes over unweathered  t i l l or bedrock. Drainage depressions, seepage hollows and f l a t midslope benches are usually characterised by waterlogged so i l s lacking well marked e luv ia l and i l l u v i a l horizons. The L, F and H horizons usually total less than 20 cm deep and are underlain by a gleyed B horizon up to 100 cm deep. The B horizon i s very stoney and may contain boulders more than 30 cm in diameter. These s o i l s f a l l into the category of Gleyed Ferro Humic Podsols. Category 3. Shallow so i l s developing on bedrock. Thin dark organic -soils frequently occur on steep rocky slopes of over 35°. These s o i l s consist of a f ibrous, decomposed organic matter horizon less than 20 cm deep, sometimes over a th in f r i a b l e Ae horizon but often in d i rect contact with the bedrock. Occasionally the organic matter horizon rests on a stoney dark brown B horizon less than 30 cm deep. These s o i l s f a l l into the category of L i t h i c Regosols, and Fo l i so l s (Lewis and Lavkul ich, 1972). 9 Table 1 presents a summary of some of the B horizon's mean physical properties based on twenty four large bulk samples taken from s o i l s belonging to categories 1 and 2. Detai ls of the sampling, which was carr ied out in conjunction with a series of f i e l d s o i l strength t e s t s , are given on page 79. P a r t i c l e s ize d i s t r i b u t i o n curves based on sieve and hydrometer analyses of the same samples are shown in Figure 3. Table 1 Some Mean Physical Properties of Steepland Soi ls and Unweathered  T i l l from the Coast Range, southwestern B.C. Property So i l B horizon Unweathered t i l l Dry unit weight ( Y d ) kg/m3 1,110 2,170 Moist unit weight ( Y J kg/m3 1 ,480 2,347 Saturated unit weight ( Y ) kg/m 1,660 2,347 Void r a t i o (e) 1.21 0.26 Percent gravel 60.10 67.30 Percent sand 30.00 25.30 Percent s i l t 9.00 7.30 Percent clay 0.90 0.10 The p a r t i c l e s i ze d i s t r i b u t i o n for the s o i l and the unweathered t i l l are s i m i l a r , suggesting that the s o i l B horizons are derived from the weathering of ablat ion t i l l which i s more permeable and of lower unit weight than the basal t i l l . Vegetation. Most slopes support a dense cover of coniferous forest except on the most precipitous b l u f f and bedrock areas. Western red cedar (Thuja p i i c a t a Donn), western hem!ock (Tsuga heterophyl1 a (Rafn.) 10 Gravel Sand S i l t Clay 100 90 80 70 60-50 AC 30 20 10 \ \ \ \ \ \ \ \ \ \ 4 \ . \ \ \ ^ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ s \ t " " " " "— "•^•^•^^jr* — ~~ sz cn > X3 O D_ 0 100 10 1.0 o Diameter in m m o.oi 0001 Figure 3. P a r t i c l e s i z e d i s t r i b u t i o n curves f o r f o r e s t s o i l B ho r i z on s . S o l i d curves represent l i m i t curves-between which the measured d i s t r i b u t i o n s f o r i n d i v i d u a l samples f a l l . The approximate mean curve i s shown as a broken l i n e . 11 Sarg.) and Douglas f i r (Pseudotsuga menziezii (Mayr.) Franco) are the dominant canopy formers on the lower slopes. Old growth mixtures of these species a t ta in merchantable volumes of 30,000 board feet or more per acre on favourable s i t e s . At approximately 900 m the increasing importance of mountain hemlock (Tsuga mertensiana (Bong.) Carr) , yellow cedar (Chaemaecyparis nootkatensis (D. Don) Spach) and amabilis f i r (Abies amabilis (Dougl.) Forb.), accompanied by a reduction in the occurrence of western red eedar and western hemlock and the el iminat ion of Douglas f i r , mark the lower l im i t s of the 'Mountain Hemlock Zone' (Krajina 1965). The combination of steep slopes, weak-structured so i l s and an impermeable t i l l or bedrock substratum would presumably create ideal conditions for large scale mass wasting were i t not for the vegetation cover whose roots help strengthen the so i l and anchor the s o i l to the slopes. The importance of roots to the s t a b i l i t y of slopes w i l l be discussed in Chapter 3. Commercial logging a c t i v i t i e s within the study area were i n i t i a t e d during the l a t t e r part of the la s t century on the lower slopes close to Burrard In let . Logging on the steeper slopes of the lower parts of Capilano and Seymour Valleys began at about the time of World War 1. Since that time c l e a r f e l l i n g has been carr ied out in most of the major watersheds t r ibutary to Howe Sound, the middle and upper reaches of the Capilano and Seymour watersheds and on the higher slopes overlooking Burrard In let . In the past, as at the present time, timber harvesting on the steep slopes has been carr ied out with cable systems (high lead systems). Downhill yarding to an accessible landing area is a common pract ice. 12 Individual c l ea r f e l l ed areas, which range in s ize from less than 0.04 km 2 to more than 4 km , seldom extend above a l t i tudes of 1,200 m. Areas cleared before 1950 support pole stands of western red cedar, western hemlock and Douglas f i r while red alder (Alnus rubra Bong.) i s usually dominant along old railway grades, abandoned roads and on old landsl ide scars. Climate. The study area possesses a coo l , moist, mountain climate notable for i t s heavy winter snowfal ls. Seymour Fa l l s in the middle of Seymour Val ley (elevation 229 m) receives an average annual p rec ip i ta t ion of 3,700 mm, 80 percent of which f a l l s between September and March inc lus ive (Department of Agr icu l ture, 1963). The maximum measured tota l p rec ip i ta t i on for a 24 hour period is 252 mm but generally r a i n f a l l inten-s i t i e s are not pa r t i cu l a r l y high. At Seymour Fa l l s the maximum 5 minute, 15 minute and 1 hour i n tens i t i e s based on 11 years of records are 61 mm, 48 mm and 31 mm per hour respect ive ly. Appendix 1 shows a summary of r a i n f a l l in tens i ty data for 26 storms including a l l the storms with r a i n -f a l l to ta l s exceeding 125 mm. F ie ld data co l l e c t i on The f i e l d survey did not attempt to locate and record the deta i l s of a l l the mass wastage features within the study area. Rather, most of the c l ea r f e l l ed slopes were v i s i t ed as well as those slopes with reason-able access (within 2 hours climbing time from the nearest road) where landsl ides were known to occur. At each landsl ide s i t e the mean width, mean depth and length of the scar, the mean slope along each 15 m section of the scar, the a l t i tude and aspect of the scar head and the mean s o i l 13 depth to impermeable bedrock or unweathered t i l l were measured. Mean depth was the most d i f f i c u l t landsl ide character i s t i c to determine and was based on the mean depth of the scarps along the l a te ra l margins of each scar. The mean depths of several avalanches with i r regu la r or i l l - d e f i n e d l a te ra l scarps and uneven f a i l u r e surfaces, were simply judged from the surface morphology of the s l i de scar. In add i t ion, the character of the s o i l materials, the surface and subsurface hydrological condit ions, the nature of the vegetation and the microtopography of the slope surrounding each scar head were recorded. The survey was l imi ted to those predominantly unvegetated scars with surface areas larger than 2 approximately 90 m . Approximately 70 percent of the major landsl ides within the study area were v i s i t ed and measured. Use of aer ia l photographs T rad i t i ona l l y , the use of aer ia l photographs has played an important role in the study of lands l ides. According to Liang (1952) one of the main advantages in using aer ia l photographs to invest igate slope f a i l u re s is that the ent i re area of movement i s usually apparent at a glance whereas, on the ground, man is dwarfed by natural topographic features and perspection of the f a i l e d area i s d i f f i c u l t . The study was s i g n i f i c an t l y aided by a stereoscopic examination of several hundred a i r photos supplied by the A i r D iv i s ion, Department of Survey Services, V i c t o r i a , B.C. Five series of approximately 1:30,000 black and white panchromatic photographs taken in 1939, 1952, 1966, 1967 and 1969 respect ively, and three series of 1:15,000 black and white photos taken in 1957, 1963 and 1968 enabled the locations and approximate age of many of the larger landsl ides to be determined. The study of 14 sequential photographs of the same slope f a i l u re s permitted the assess-ment of scar s i ze , scar form, and scar vegetation changes a f te r the i n i t i a l s l i d i n g . Landslides were recogniseable on black and white photo-pairs by the sharp l i ne of break at the head scarp, the l i g h t coloured appearance of the landsl ide scar which, in the case of many small lands l ides, appeared as a th in white l i ne broadening towards i t s lower l i m i t s , and, in the case of older revegetated scars, by the presence of uniform l i g h t grey s t r ips representing alder and young conifer growth (Figure 4). Very dul l tone contrasts or poor sharpness of image or presence of a snow cover re s t r i c ted the usefulness of several sets of older photos but generally the 1:15,000 photos provided more information than the 1:30,000 p r in t s . Since the completion of the or ig ina l landsl ide survey, very high qual i ty colour and colour infrared aer ia l photographs taken at 12,000 m a l t i tude from a CF-100 reconnaissance a i r c r a f t have been made ava i lab le . Details of th i s photography are given in Table 2. Table 2 Details of High A l t i tude Colour and Colour Infrared Photography taken 16-7-71 over Southwestern B r i t i s h Columbia Film Lens Focal Length 225 mm 88.9 mm Aerochrome 2445 F i l t e r F 1 ^ i n g  M l t e r Height 12,200 m Shutter Speed 4 0 0 s e c * Type and Scale of P r in t Contact 22.5 x 22.5cm. Scale 1:140,000 70 mm 76.2 mm Wratten 12,200 m Aerochrome No. 12 I.R. Yellow nr>&eC . Enlargement 22.5 x 22.5 cm Scale 1: 49,000 Figure 4. Stereogram of part of the lower Capilano catchment. A, debris avalanches on a steep, forested slope; B, l i gh t -co lo red alder and maple marks locations of skid t r a i l s and railway grades on slope c l ea r f e l l ed 34 years previously; C, boundary between c l e a r f e l l e d and forested slope; D, o l d , part ly revegetated debris s l ides on a slope c l ea r f e l l ed 32 years previously. Scale 1:15,000. 16 On colour aer ia l photos recent large landsl ide scars appear as l i gh t brown areas which contrast markedly with the dark green conifer fo l iage of undisturbed slopes or the l i g h t green regrowth on old c learcut areas. However, the small scale of photography prohibited detection of 2 small f a i l u re s less than 1,000 m in area. Nor could landsl ides be readi ly distinguished from mining excavation features in Britannia and Furry Creek va l leys . Colour photos were superior to black and white photos for studying the gross features of the landscape, the nature of the vegetation and the broad drainage character i s t i c s of spec i f i c slopes. The colour oblique shown in Figure 2 provides an example of the colour tones and image de f i n i t i on depicted on ve r t i ca l colour p r in t s . Colour infrared aer ia l photography offers many advantages over colour and panchromatic black and white photography (Mintzer 1968, Poole 1969). Colour infrared not only enhances the contrast of the te r ra in and emphasises gul ly shapes, drainage depressions, and erosional features, but i t also creates strong contrasts between the darker coloured conifer growth and deciduous vegetation and between vegetation of d i f fe rent ages on c l ea r f e l l ed areas. Colour infrared f i lm u t i l i z e s three image layers sens it ized to green, red and infrared instead of blue, green and red in normal colour f i l m . Blue is witheld with a yellow f i l t e r . The v i s i b l e l i g h t component together with the infrared component produces a modified colour rendit ion of the subject photographed. Depending on the proportions of green, red and infrared ref lected or transmitted by the subjects, a great var iety of colours can be produced. For instance, healthy deciduous green fo l iage (alders, birches and maples) appear red, conifers appear dark blue to dark purple while exposed bedrock may range from greyish brown to l i g h t blue. Where vegetation i s th in or lack ing, poorly drained slopes 17 and gul ly seepage areas are sometimes recognisable by the i r dark tones which contrast with l ight-coloured dry slopes. Recently exposed s o i l or t i l l surfaces are coloured white in dryer s i tuat ions or l i g h t greyish blue on moist s i t e s , causing landslides to be eas i l y i d e n t i f i e d . Several narrow erosion features which were p a r t i a l l y masked by the tree canopy and not dist inguishable on panchromatic black and white photographs were recognisable on colour infrared photos. The colour infrared prints did present some disadvantages, the most important being that snow-f i l led hollows and snow-chutes were d i f f i c u l t to d ist inguish from recent landsl ide scars. The stereogram in Figure 5 emphasises some of the features discussed above. A summary of the landsl ide information obtainable from each of the types of aer ia l photographs is presented in Table 3. Types of mass wasting During the survey, landsl ides were not witnessed in act ion. Therefore, the types and rates of processes responsible for a par t i cu la r slope scar or debris accumulation could only be judged from the ex i s t ing form of the feature. Furthermore, interpretat ions were sometimes confounded because f l u v i a l erosion or continued mass movements had strongly modified the o r i g ina l form of the mass movement scars. Nevertheless, it.was possible to assign most of the fa i l u re s investigated to one of the types of landsl ides discussed below. Adopting the well known landsl ide c l a s s i f i c a t i o n scheme presented by Varnes (1958), most of the mass wastage features within the study area belong in one of three categories (or combinations of these categories) 1 3 Figure 5. Color inf rared stereogram of part of the middle Seymour Val ley. A, large debris avalanche on a steep slope covered with cedar-hemlock fo res t ; B, discontinuous groves of alder-maple forest are eas i l y d ist inguished by t h e i r l i gh t pink co lo r ; C, recently c l e a r -f e l l e d area. Scale, 1:49,000. 19 Table 3 Information Obtainable from Various Types of Aer ia l Photographs Photography Information Obtainable Panchromatic B & W 1:30,000 Panchromatic B & W 1:15,000 Colour (Aerochrome) 1:140,000 Colour Infrared 1:49,000 Gross features of the landscape, slope forms, drainage patterns and broad vegetation features i d e n t i f i a b l e . Land-s l ides over 500 in area recognisable. Most landsl ides over 200 in area i d e n t i f i a b l e . Type of s o i l materials and morphology of large scars recognisable. Gross features of landscape, slope forms, broad vegetation features and drainage patterns i d e n t i f i a b l e . Landslides over 1,000 m^  in area recognisable. 2 Most landsl ides over 500 m in area i d e n t i f i a b l e . On bare areas s o i l d ra in -age conditions and amount of bedrock assessable. Broad vegetation features i d e n t i f i a b l e . Interpretations in Table 3 based on stereoscopic examination with and Old Delft Scanning Stereoscope, Model 0DSS I I I. which form a t rans i t i ona l series of landsl ide types from r e l a t i v e l y dry debris si ides to debris avalanches to very wet, mobi1e debris flows. Despite the plea by Rapp (1963) to reserve the term 'avalanche' only for mass movements of snow the term has been widely used in North America to describe mass wasting of s o i l s and rock (Bishop and Stevens 1964, Swanston 1969, 1970, Dyrness 1967 and Gray 1969). For lack of a better expression the term 'debris avalanche' i s adopted in th i s thesis to describe a very important type of slope f a i l u r e . As the d i s t i nc t ions between debris s l i de s , debris avalanches and debris flows are vague and 20 the terminology is often confusing, the three main landsl ide types recognisable in the study area are described in the fol lowing paragraphs. Debris s l i de s . Debris s l ides form when part or a l l of the vegetation cover, humus layers, weathered so i l mantle, and occasional ly the under-ly ing substratum are dislodged and s l i de downslope. The resu l t ing scar possesses an obvious s l i d i n g plane which is often formed by the upper surface of the unweathered t i l l or bedrock (Figure 6). The longitudinal p ro f i l e s of large debris s l i de scars were usually l i nea r to gently concave upwards but many of the smaller s l ides along stream sides and on road-cut and r o a d - f i l l slopes appeared to be rotat ional f a i l u re s possessing spoon-shaped scars s imi la r to the features described as 'slumps' by Sharpe (1938) and Varnes (1958). In plan view the scars ranged in shape from broad oval features to long narrow trenches. Most debris s l ides possessed well marked head scarps while the lower parts of several scars were covered with detached but par t ly intact blocks of so i l and tree root masses, ind icat ing that movement occurred in a s l i d i ng manner rather than by f lowing. Debris s l ides ranged from 15 m to 150 m in length, averaging approximately 80 m. Debris avalanches. Debris avalanches, recognisable by the i r long narrow V or crescent shaped scars sometimes over 600 m in length, were the largest mass wastage features on undisturbed and c l ea r f e l l ed slopes. The scars cha r ac te r i s t i c a l l y tapered towards t he i r heads where ver t i ca l scarps usually marked the s i te s of the i n i t i a l f a i l u re s (Figure 7). Many debris avalanches began in shallow drainage depressions or other wet s i tes (Figure 8 ) , suggesting that groundwater played an important role in the i r release. At the s i tes of i n i t i a l f a i l u r e the basal s l i d i n g plane was often the upper surface of the unweathered t i l l although, occas ional ly, F igure 6. Small deb r i s s l i d e i n upper Harvey Creek catchment. The s l i d e occur red on a 34° s lope and exposed an unweathered, basal t i l l . ; , F i gure 7. An almost v e r t i c a l headscarp of a l a r g e deb r i s avalanche i n the middle Seymour catchment. Figure 8. The upper part of a large debris avalanche in the Seymour catchment. A s izeable stream flows down the centre of the scar during wet weather. Figure 9. A debris avalanche in second growth fo res t . Near the base of the slope the scar narrows and develops a sinuous course. 23 smooth d i o r i t e bedrock surfaces formed the debris avalanche f l o o r . Most debris avalanches were less than 2 m deep except for those formed on slopes underlain by deep co l l u v i a l gravels or coarse alluvium. Apparently debris mobi l izat ion begins when large intact masses of s o i l break away from the slope leaving a well defined headscarp. These masses transform into debris avalanches when stresses cause breakdown in the s o i l structure as explained by Swanston (1967). Often the landsl ide track narrowed near i t s lower l i m i t to form a sinuous channel bordered by raised levees (Figure 9) , demonstrating that, as they pass downslope, debris avalanches may increase the i r water content and transform into true debris flows. Hence Swanston (1969) refers to 'debris avalanche-debris flow combinations' while Dyrness (1967) c l a s s i f i e s s im i la r composit mass wasting as 'debris avalanches with earthflows!. Debris flow formation i s discussed by Johnson and Rahn (1970) who note that the t ran s i t i on from lands l id ing to channelized flow of unsorted debris results from the' d i l u t i on by water of s l i de masses. High rates of movement during f a i l u r e were indicated by the frequent occurrence of gravel and boulders lodged in the bark and wood of tree trunks growing on the margins of avalanche scars. In contrast to debris s l i de mater ia ls , the transported debris avalanche materials were completely disrupted as the i r movement i s more akin to a true flow process rather than a s l i d i n g process. The debris avalanches measured in the f i e l d ranged in length from 2 38 m to 610 m while the i r mean area and mean volume* were 4,300 m and 3 4,200 m respect ively. These were r e l a t i v e l y small features compared to Mean volume was calculated as the product of scar length, mean scar width and mean scar depth. 24 the debris avalanches and related types of mass wasting described by Wentworth (1943) in Hawaii, Flaccus (1958) in the Appalachian Mountains, Simonett (1967) in New Guinea, Rapp (1963) in Norway, Dyrness (1967) in the Cascade Mountains and Sheng (1966) in Taiwan. Debris flows. The mass wastage features pos i t i ve ly i den t i f i ed as debris flows were s i ted along the courses of steep permanent or ephemeral t r ibutary stream channels. Their points of o r ig in were not well defined except in those cases where debris s l ides or debris avalanches entered stream channels and transformed into wet debris flows. Presumably, many debris flows i n i t i a t e when temporary log dams holding back rock debris and water collapse during storm periods, releasing a saturated mobile s lu r ry . The debris flows which resu l t gather addit ional materials as they pass down the stream beds leaving conspicuous t r iml ines on stream gul ly s ides, scouring channel beds, depositing l a te ra l levees of coarse rock materials and sometimes overtopping the stream banks to form lobes or cones of debris on the lower slopes. S imi lar features in southeast Alaska have been described by Bishop and Stevens (1964) and Swanston (1969, 1971) as 'debris t o r ren t s . ' The present study is pr imar i ly concerned with debris s l ides and debris avalanches while other types of mass wasting recognisable in the area, including stream channel debris f lows, rock f a l l s , talus s l i de s , rock and s o i l transport by snow avalanches, and s o i l creep, were not invest igated. The term ' l and s l i de ' i s used in a general sense to include debris s l ides and debris avalanches. S ignif icance of landsl ides in the study area Seventy-seven large, recent, debris avalanches and debris s l ides were recorded in the study area; forty-n ine of these were investigated "ABLE 4 A SUMMARY OF LANDSLIDE SURVEY DATA COLLECTED IN THE BRITISH COLUMBIA COAST RANGE 1970 LANDSLIDE LANDSLIDE SITE VEGETATION SLOPE ALTITUDE ASPECT SOIL SOIL TYPE OF AREA OF VOLUME OF APPROXIMATE IDENTIFICATION TYPE TYPE* COVER* (DEGREES) (METERS) ! CATEGORY SERIES* BEDROCK* LANDSLIDE (METER 2) LANDSLIDE (METER 3 ) AGE (YEARS) 31 D. sl i ' d e " OS CF15 35 655 ssw 1 F sed 3,430 1, 130 3 - 5 •CI D. ' s l i d e OS CF12 33 488 s a 1 Dior 190 110 - - 2 - 3 -C2 D. 6 l i d s DD CF15 35 503 NW 1 D i o r 90 60 3 - 5 FC3 D. a v a l . OS CF15 35 533 liNW I D i o r 1,020 4,670 12 -18 FC4 D. a v a l . DD CF20 35 524 SSE 1 D i o r 930 4,390 12 -18 HSI D. s l i d e OS CF12 48 884 WSW 1 GE F sed 1,210 3,540 3 - 5 HS2 D. a v a l . DD F 45 1,052 wsw 11 WH Di o r 2,790 5,950 3 -10 HS3 D. a v a l . R? CF10 ' 39 1,036 SE 1 GE Di o r 1,950 1,780 3 -10 HS*. D. a v a l . • RF CF10 37 1,052 SW 1 GE F sed 2,320 2,830 3 -10 HS5 D. s l i d e OS CF30 43 524 WSW 1 GE Congl 460 140 2 - 3 KS6 D. a v a l . RF CF S 41 951 ssw 1 GE F sed 1,860 2,830 3 -10 H57 D. a v a l . RF CF10 37 1,021 . su 11 WH Schl 2,690 3,260 3 -10 HS8 D. a v a l . -RF CF10 36 .1.036 SW 11 WH F sed 2,320 2,830 3 -10 H59 D. a v a l . RF CF 8 • 36 1,030 SSE • 1 CE F sed 2,510 3, 110 2 - 3 KSIO D. a v a l . RF CF 8 36 1, 128 S-SW 1 GE F sed 2,320 2,120 3 -10 KS1I D. a v a l . OS "CF10 37 1.036 WNW 1 PA Dior 3,900 9,510 3 -10 HSI 2 D. a v a l . . RC CF10 42 1,036 s 1 PA D i o r 1,300 3,170 3 -10 H3I3 D. a v a l . kC F 45 762 tsw I PS D i o r 65,030 39,630 18 -30 KSI4 D. s l i d e DD CF 5 30 1,067 i.SW 11 WH Di o r 930 570 3 - 5 HS 15 D. s l i d e DD CF 5 35 1,073 SSE 11 WH Di o r 370 230 3 - 5 CI D. a v a l . DD FF 38 411 INE 11 SN D i o r 2,230 1,390 1 - 2 C2 D. a v a l . DD F 34 274 ESE 11 PA D i o r 740 3,570 12 -18 C3 D. a v a l . DD F 37 960 NE 11 WH Di o r 1,860 6,510 C i D. a v a l . OS F 31 808 SSE 1 PA D i o r 1,490 3,170 12 -18 C5 •D. s l i d e OS CF30 35 766 SSE 1 S D i o r 370 230 18 -30 C6 D. s l i d e DD CF30 31 329 ESE 11 SN D i o r 90 60 2 - 5 C7 D. s l i d e 03 F 34 543 NNW 1 S Di o r 560 230 ca D. s l i d e OS CF20 37 469 W 1 SN D i o r 190 60 C9 D. a v a l . OS F 33 817 SSE 11 BW Dior 1,670 1,020 2 - 5 CIO D. a v a l . DD F 32 762 SSE 11 S Di o r 5,020 3,060 12 - i a C I I D. a v a l . DD CF.30 34 375 SW 1 SN Dior 460 280 3 - 5 CI2 D. a v a l . DD F 32 1,017 SSE 11 S D i o r 4,000 3,680 12 -18 CI3 D. a v a l . DD CF 4 39 463 SW 1 CE D i o r 560 340 k - 1 C H D. a v a l . OS FF 35 518 ENE 11 SN D i o r 2,600 2,380 2 - 3 CI5 D. a v a l . DD F 30 503 SW 1 SN Dior 1,860 850 3 -12 cie D. s l i d e OS F 36 747 SE 1 CE Dior 280 2S0 CI7 D. s l i d e DD CF30 41 336 SE 11 WH Di o r 460 230 7 -12 c:8 D. ava1. DD F 29 1,237 NNE 11 WH Dior 1,390 280 Ct9 D. a v a l . DD F 34 917 WSW 11 WH Di o r 1,770 850 4 - 7 SI D. s l i d e DD F 37 625 ENE 11 WH D i o r 190 170 1 - 2 S2 D. a v a l . DD F 36 732 SW 11 WH Congl 2,600 1,960 \ - I S3 D. s l i d e OS F . 37 655 SE 111 DT D i o r 4,830 1,470 3 -12 S4 D. a v a l . DD F 36 1,036 SE 11 WH Di o r 8,450 10,300 3 - 7 SS D. a v a l . DD F 37 1,036 SE 11 WH D i o r 6,970 8,490 3 -12 S6 D. a v a l . DD F 33 732 SW 11 S D i o r 1,860 850 S7 D. s l i d e DD F 38 1,265 MfE 11 CE Dior 460 650 S3 D. s l i d e DD F 24 1,250 NNE 11 GE D i o r 460 420 S9 D. a v a l . DD . F 31 347 SSE 11 WH Di o r 650 400 3 - 5 CIO D. a v a l . DD F 38 1,036 SE 11 WH D i o r 840 510 Means 36 794 Means 3,113 2,972 CS Open s l o p e . DD Drainage d e p r e s s i o n . ? V i r g i n f o r e s t . RF F i l l s l o p e or slop e formed of s i d e - c a s t m a t e r i a l s . RC Cut-elope above roadway. FF F i r e - k i l l e d f o r e s t . CF C l e a r f e l l e d a r e a . (Numbers i n d i c a t e approximate time I n ycar a str.ee c l e a r f e l l i n g } . £ Based on s o i l s e r i e s nap (1:50,000) produced by B.C. Dept. of A g r i . BW B u r w e l l S e r i e s . GE Colden Ears S e r i e s . PN Paton S e r i e s . WH Whonnock S e r i e s . DT Dennet S e r i e s . SN Strachen S e r i e s . CE Cannel S e r i e s . S Sayres S e r i e s . | F sed F i n e g r a i n e d sedimentary rock. D i o r D i o r i t e or r e l a t e d p l u t o n i c r o c k . S c h i S c h i s t . Congl Conglomerate. ro cn 26 in the f i e l d while the remaining twenty-eight were i den t i f i ed on 1968 and 1969 aer ia l photographs. Their d i s t r i bu t i on is shown in Figure 1 and a summary of each lands l ide ' s measured character i s t i c s appears in Table 4. The importance of lands l id ing in terms of landsl ide numbers, area of landsl ides and weight of landsl ide debris transported per square kilometer can be gauged from Table 5. As the mean values of measured debris avalanche areas and volumes were assigned to the twenty-eight debris avalanches not measured in the f i e l d , while a moist unit weight of 3 1,480 kg/m was used for converting volumes to weights, the two righthand entries in Table 5 are no better than rough approximations. Table 5 Landslide number, area and quantity of debris moved per  square kilometer of the study area Mean number of land- Mean area of land- ^ L ^ ^ f l^t s l ides per km? s l i de s , m? per km2 5 1 ° V e 0.12 427 612 Compared to other forested mountain regions for which data are avai lable (Katsumi 1965, Sheng 1966, Simonett 1967, and Dyrness 1967), landsl ides in the Coast Range study area are less frequent and influence a smaller percentage of the tota l area, ind icat ing that the s o i l mantle i s r e l a t i v e l y stable. Nevertheless, landsl ides are an important feature of the environment of some steep slopes. 27 Th i r t y - s i x debris avalanches and debris s l i de s , representing 47 percent of a l l landsl ides in the study area, ran out into t r ibutary stream beds or main stream channels. Generally, t r ibutary catchments with large landsl ides descending into t he i r stream channels possessed recently aggraded stream beds in the i r lower reaches. For instance, 3 3 large debris avalanches had deposited approximately 10,400 m of debris into the main stream channel of Strachen Creek which i s a steep bedrock gorge to within a few hundred meters of i t s confluence with S isters Creek. Most of the landsl ide materials had passed down the gorge to the lower reaches of S isters Creek, and possibly as far as the Cleveland Resevoir. The Coast Range landsl ides have an addit ional hydrological s i g n i f -icance because they form part of the surface drainage network. Most large landsl ide scars possessed ephemeral streams down the length of the i r longitudinal axes and, in many cases, the streams originated at the land-s l i de headscarp as a subsurface seepage. In a l l cases invest igated, however, aer ia l photographs showed that discernable surface drainage features did not ex i s t on these s i tes pr ior to the time of f a i l u r e . Furthermore, f i e l d inspections showed that surface drainage features were often absent (four notable exceptions ex i s t ) upslope from the headscarps of lands l ides. Presumably, the pre-landsl ide downslope drainage was f a c i l i t a t e d pr imari ly by subsurface flow and the occurrence of a land-s l i d e , therefore, represents an addit ion to the surface drainage network. Relationship of landsl ides to s i t e factors Microtopoqraphy. Thirty-one of the forty-nine landsl ides investigated in the f i e l d i n i t i a t e d in drainage depressions, small seepage hollows or on poorly drained parts of open slopes. Many of these s i tes were underlain by so i l s belonging to Category 2 which exhibited some degree of mottl ing 28 in the B horizon, a condition which normally, but not always, r e f l ec t s poor drainage conditions and high water tables for long periods during the year (Canada Department of Agr icu l ture, 1970). Hack and Goodlett (1960) noted that debris avalanches on the steep slopes of the Central Appalachian Mountains commonly or ig inate in hollows although they do not mention the drainage conditions at the s i tes of landsl ide release. Debris avalanches frequently occur within l i near seepage depressions on steep slopes in southeastern Alaska (Swanston 1969) which, combined with the fact that depression so i l s often displayed evidence of former i n s t a b i l i t y , led Swanston to conclude that depressions enlarge themselves by lands l id ing. Large l i near drainage depressions occur on uniform slopes which are eas i l y recognised on 1:15,000 aer ia l photographs. On the other hand many slopes have 'stepped' longitudinal p ro f i l e s consist ing of short slope facets separated by bedrock outcrops and ve r t i ca l rock wal l s . The discontinuous s o i l mantle on these 'broken' slopes i s not susceptable to lands l id ing . Slope. Most large landsl ides originated on s i tes which f a l l within a r e l a t i v e l y narrow range of slope steepness. Figure 10 (top) shows that over 75 percent of the measured landsl ides occurred on slopes ranging from 31° to 39° inc lu s i ve . Sites with surface slopes greater than 40° are usually extremely rocky and underlain by thin regosol ic s o i l s (Category 3 s o i l s ) not prone to large scale f a i l u r e . Aspect. Approximately 75 percent of the landsl ides investigated possessed aspects with a southerly component, SSE being the most frequent p aspect, accounting for 20 percent (Figure 11 (top)). A chi squared ( x ) test was used to determine whether or not the d i s t r i bu t i on of landsl ides with respect to aspect- d i f fe red s i g n i f i c an t l y from the d i s t r i bu t i on 29 8| Figure 10. Top. D i s t r ibut ion of landsl ides with respect to s lope. Bottom. Longitudinal p ro f i l e s of selected lands l ides . Figure 11. Top. D i s t r ibut ion of landsl ides with respect to aspect. Bottom. D i s t r ibut ion of landsl ides with respect to a l t i t u d e . 31 which would be expected i f aspect exercised no influence on lands l ide 2 2 locat ions. The test yielded a x equal to 45.3 (tabulated x at 0.01 level of s ign i f icance and 15 degrees of freedom is equal to 30.6) ind icat ing that aspect s i g n i f i c an t l y influences the locat ion of slope f a i l u r e s . However, th i s s ign i f icance is based on the premise that slopes of a l l aspects are equally represented. There i s a preferred or ientat ion of slopes towards the south in the Capilano and Seymour catchments (Slaymaker, personal communication) but a sample of 134 points obtained by laying a dot gr id over a 1:50,000 topographic map of the study area, showed that th i s southerly bias was very weak for the to ta l study area. S ixty-four points f e l l on slopes with a southerly component, f i f t y - n i n e points f e l l on slopes with a northerly component and eleven, points could not be c l a s s i f i e d . There i s one probable reason why landsl ide or ientat ion i s biased towards a southerly d i r ec t i on . In Furry Creek, Strachen Creek and Harvey Creek, in most of the large t r ibutary catchments in the upper Capilano Val ley, and in two t r ibutary catchments draining the western slopes of the middle Seymour Val ley, the north-facing slopes are rocky and broken, a condition which discourages landsl ide formation, while south-facing slopes are r e l a t i v e l y uniform and underlain by an extensive unweathered t i l l substratum. These physiographic differences presumably resu l t from the general North to South flow d i rect ion of the Pleistocene ice sheets which inundated the Coast Mountains (Armstrong and Brown 1954). If the landsl ides associated with roads were neglected,the percentage of fa i lu res with a southerly component in the i r or ientat ion remained at approximately 75 percent. 32 A l t i t ude . The d i s t r i bu t i on of landsl ides with respect to a l t i tude i s depicted in Figure 11 (bottom). The fact that landsl ides were most numerous between a l t i tudes of 1,000 and 1,100 m i s seen to coincide with the fact of numerous road associated f a i l u re s within th i s range of a l t i t ude . If road-caused landsl ides are discounted, then the apparent 2 influence of a l t i tude becomes questionable. A x test showed that the d i s t r i bu t i on of lands l ides, except for road-caused f a i l u r e s , was not s i g n i f i c an t l y d i f fe rent at the 0.05 level of s ign i f icance from the uniform d i s t r i bu t i on formed by assuming a l l a l t i t ude classes contained 2 equal numbers of lands l ides. The calculated x with 10 degrees of freedom 2 was equal to 11.20 compared to 18.31, the tabulated x for 10 degrees of freedom at the 0.05 level of s i gn i f i cance. Bedrock geology. The overal l uniformity of the d i o r i t e bedrock over a large portion of the study area would tend to rule against bedrock type as a cause of spat ia l v a r i a b i l i t y in landsl ide numbers. Moreover, the widespread covering of compacted t i l l , the upper surface of which often forms the basal f a i l u r e plane, reduces the influence of the bedrock on lands l id ing. Nevertheless, the f i e l d survey data showed that bedrock differences may exercise some control over large slope fa i l u re s related to road construction. On slopes underlain by sedimentary and metamorphic rocks belonging to the Gambier Series (Armstrong 1965), road construction had caused s ix major debris avalanches but on slopes underlain by d i o r i t e and related igneous rocks, the s i tuat ion on approximately 90 percent of the logged-over slopes v i s i t e d , there were only eight major f a i l u re s related to roads. However, the ef fect of bedrock type on road-caused debris avalanches may be more apparent than r e a l . The s ix landsl ides on the Gambier Series originated in road f i l l or sidecast materials and 33 steep slopes ( a l l f a i l u re s originated on slopes steeper than 35 ) combined with poor drainage conditions seemed to be the main factors contr ibuting to the slope f a i l u r e s . Of greater importance than the bedrock composition is the nature of the bedrock's upper surface. I f the s o i l mantle rests d i r e c t l y on bedrock, then smooth uniform bedrock surfaces para l le l to the slope surface favour s l i d i n g as in the cases of debris s l ides C17 (Figure 12) and S3. So i l s . The poorly drained s o i l s , belonging to Category 2 were associated with 25 lands l ides, while the so i l s at Category 1 subject to only moderately good drainage but dist inguishable from Category 2 so i l s by the lack of well defined mottl ing in t he i r B horizon, accounted for f i f t een slope f a i l u r e s . Only one debris s l i d e , S3, had formed in the shallow Category 3 s o i l s . The close association between landsl ides and potent ia l l y very wet s o i l s , supports a concept that pore water pressures from high water tables i s a pr inc ipa l factor leading to slope f a i l u r e . An attempt was made to place the f a i l e d so i l s in the s o i l series categories mapped by the Soi l Survey D iv i s ion, B.C. Department of Agr iculture (see Table 4). Although th i s placement should be regarded as tentat ive i t does appear that the Golden Ears, Whonnock, Pal isade, and Strachen s o i l series are most subject to mass wasting. Effect of c learcutt ing on landsl ide occurrence Croft and Adams (1950), Bishop and Stevens (1964), Dyrness (1967), Gray (1969) and Fujiwara (1970) have f i rmly established that the removal of forests may ser iously increase the incidence of lands l id ing on steep slopes. The protective role of a forest cover is part ly manifested through the mechanical strengthening of the so i l by tree root systems 34 (Turmanina 1965, Endo and Tsurata 1969, Gray 1969, and Swanston 1969, 1970) and part ly through the modif ication of s o i l moisture d i s t r ibut ions and possibly pore water pressures (Bethlamy 1962, Patr ic et al 1965, Hall in 1967 and Gray 1969). Although the Coast Range landsl ide survey revealed an increase in the density of landsl ides on c l e a r f e l l ed areas, the extremely e r r a t i c d i s t r i bu t i on of slope fa i lu res on c l e a r f e l l ed and undisturbed slopes, combined with the lack of information suitable for accurately dating most landsl ide events, complicated the assessment of s t a b i l i t y changes fol lowing logging. Table 6 indicates the broad changes in landsl ide frequencies a f te r logging on areas c l ea r f e l l ed between 1957 and 1968. The pre-logging information was obtained from 1957, 1963 and 1966 aer ia l photographs. Table 6 Numbers of landsl ides Before and After Logging, Coast Range, Southwestern B r i t i s h Columbia Approx. area c l e a r f e l l ed between 1957 and 1968 Number of landsl ides 1957 1970 Howe Sound Drainages 13.7 km2 3 11 Capilano Catchment 0 3 Seymour Catchment 0 0* * Two small s l ides occur on a c l e a r f e l l ed slope in the middle Seymour catchment. 35 Eight of the 14 landsl ides recorded in 1970 were pr imar i ly caused by road construction. In order to invest igate the effects of forest removal on landsl ide occurrences in more d e t a i l , the landsl ide densit ies and the quantit ies of landsl ide debris transported per square kilometer on c l ea r f e l l ed slopes and undisturbed slopes were determined in ten small watershed areas. Estimates of landsl ide volumes, and consequently weights of debris moved, were mainly based on the data presented in Table 4 but, in add i t ion, rough f i e l d estimates of the volumes of small landslides (not included in Table 4) on forested slopes in Furry Creek catchment (2), Br itannia Creek catchment (1), Lynn Creek catchment (4), Magnesia Creek catchment (1) and Harvey Creek catchment (1) and on c l ea r f e l l ed slopes in the middle Seymour catchment (2) were included in the ca lcu lat ions . Within each subcatchment the broad physical form of the forested slopes used in the analysis was s im i l a r to that of the c l ea r f e l l ed slopes except in Magnesia Creek and the middle Seymour catchment where the forested slopes were generally steeper and at higher elevations than the logged slopes. Areas were measured from enlargements of 1:50,000 topographic maps and from 1:15,000 aer ia l photographs with a polar planimeter. The results are presented in Table 7. A t test was employed to determine i f the mean landsl ide density and the mean quantity of landsl ide debris produced per square kilometer for c l ea r f e l l ed slopes were s i g n i f i c an t l y d i f fe rent from the equivalent means for undisturbed slopes. The data were f i r s t tested by assuming that population variances were homogeneous, an assumption which F tests showed to be i n v a l i d . This necessitated the use of a t 1 test (Walpole 1968, p. 230) designed to take into account heterogeneous variances. In the f i r s t set of tests a l l landsl ides were considered, in the second 36 set only those fa i l u re s not caused by road construction were included. Table 7 Landslide Densities and Quantities of Landslide Debris Produced  Within Selected Watershed Areas in the Coast Range, Southwestern B r i t i s h Columbia Catchment No. of landsl ides Tons of landsl ide debris per km2 per knr C lear fe l led Forest C lear fe l led Forest A B C D E F Br itannia 0.32 0.32 0.05 539 539 293 Furry 0.75 0.30 0.05 2,227 212 180 Brunswick 6.25 2.50 0.19 20,573 6,808 1,639 Magnesia 1.43 0.48 0.30 6,484 56 1,933 Harvey 1.11 0.56 0.19 3,075 469 1 ,227 Strachen 2.50 2.50 0.56 6,290 6,290 2,716 Sisters 0.64 0.64 0.16 126 126 81 U Capilano 0.95 0.95 0.09 1,917 1,917 412 M Seymour 0.59 0.59 0.22 252 252 823 Lynn 0.42 0.42 0.04 1,890 1,890 240 Means 1.50 0.93 0.19 4,337 1 ,856 954 A and D A l l landsl ides considered B and E Road-caused landsl ides not considered The test results are tabulated below: A vs C t ' = 2.31* with 9 degrees B vs C t ' = 2.71* with 10 degrees D vs F t ' = 1.73 with 9 degrees E vs F t ' = 1.05 with 11 degrees of freedom. (Tabulated t.05 = 2.26) of freedom. (Tabulated " = 2.23) of freedom. (Tabulated " = 2.26) of freedom. (Tabulated " =2.20) 37 The results of the analysis indicate that the density of landsl ides was s i g n i f i c an t l y higher, at the 5 percent level of s i gn i f i cance, on c l ea r f e l l ed slopes compared to forested slopes. However, th i s s t a t i s t i c a l s ign i f icance may have l i t t l e real meaning, as evidence that the c l e a r f e l l ed slopes and the forested slopes were comparable was l imited to s i m i l a r i t i e s in gross slope morphology. The mean landsl ide densit ies for logged and forested slopes presented in Table 7 are s im i la r to those recorded by Dyrness (1967) in the H.J. Andrews Experimental Forest, Oregon but are much smaller than the densit ies presented by Fujiwara (1970) for two unstable mountain areas in northern Japan (Table 8). Table 8 Landslide Densities on Forested and C lear fe l led Mountain Slopes  for Regions in Northwestern North America and Northern Japan Source of information O'Loughlin 1972 Dyrness 1967 Fujiwara 1970 Region Coast Range B.C. . Cascades, Oregon Nth.Japan a) B i ra to r i b) Urahoro Number of landsl ides per km Forest Logged A B 0.2 0.1 9.0 3.0 1.5 1.0 80.0 25.0 Ratio of B:A 7.5 10.0 8.9 8.3 Although the natural resistance to mass wasting in the undisturbed conditon varies widely from region to region, the number of landsl ides per unit area increases by a s imi la r factor a f ter logging, implying that the re l a t i ve effects of deforestation on the s o i l ' s resistance to f a i l u r e do not vary 38 greatly over a wide range of geolog ica l , s o i l , topographical and c l imat ic conditions. The importance of lands l id ing in terms of numbers of events is not as meaningful as the volumes or weights of landsl ide debris transported. The t ' tests showed that the quantit ies of landsl ide materials on c l ea r -f e l l ed slopes were not s i g n i f i c an t l y d i f fe rent from the quantit ies moved on undisturbed slopes, a consequence of the large var iat ion in debris weights from catchment area to catchment area on the same class of slope and the larger mean s ize of landsl ides on forested slopes. Nevertheless, the mean weight of debris moved per unit area on logged slopes was approx-imately 4.5 times larger than the equivalent mean weight on forested slopes. More than two thirds of th i s debris increase on cleared areas is a t t r ibutab le to road-caused fa i l u re s (Table 7). Road construction and landsl ides Road construction has been i den t i f i ed as a major cause of mass wasting on steep slopes in many d i f fe rent environments (Fredriksen 1963, 1970, Dyrness 1964, 1967, Sheng 1966, Jeff rey 1968, Gonsior and Gardner 1971, Lantz 1971 and Nobel and Lundeen 1971). Road construction i s more detrimental to the s t a b i l i t y of the mountain slopes within the study area than any other a c t i v i t y carr ied on by man. Approximately 260 km of access roads and logging roads, including old disused roads that are now in various states of deter iorat ion and recolonizat ion by vegetation, ex i s t in the study area. Road lengths were measured on 1:15,000 aer i a l photographs and enlargements of 1:50,000 topographic maps. Of these about 180 km traverse moderately steep to steep slopes, and thereby create favourable s i tuat ions for the development of fa i l u re s on the disturbed areas adjacent to the road r ight of way. 39 Fourteen landsl ides of the seventy-seven recorded were d i r e c t l y related to road construction while possibly three addit ional f a i l u re s (HS1, HS14 and HS15) were associated with road construction less than 200 m upslope from the i r points of o r i g i n . Of the landsl ides d i r e c t l y related to road formation, seven originated in sidecast or road f i l l materials at or below the level of the road surface, f i ve i n i t i a t e d in natural s o i l s and colluvium immediately below the outer edge of the road, and two resulted from the collapse of s o i l and co l l u v i a l materials on steep, cut slopes. However, the constraints on the survey resulted in strong under-estimation of the importance of road construction as a primary cause of mass wasting. If the minimum landsl ide s ize for inclus ion in the survey 2 2 had been set at 10 m instead of 90 m , then more than 100 additional debris avalanches and debris s l ides would have been included and most of these would have been road-caused events. For example, along an approx-imately 4 km section of an abandoned logging road in Furry Creek va l ley at 590 m above sea - leve l , twelve small landsl ides occurred on cut and f i l l slopes. Although the quantit ies of materials displaced by such fa i l u re s are small compared to those moved by many of the large debris avalanches, secondary f l u v i a l erosion resu l t ing from blockage of roadside ditches i s common. Four factors appear to be most s i gn i f i can t in the formation of land-s l ides related to logging roads; (1) the natural steepness of the t e r r a i n , (2) the amount of s idecasting of s o i l and rock materials onto the slopes below the road-l ine during road construction, (3) the drainage condit ions, (4) the rate at which disused roads become recolonized by vegetation. These factors are b r i e f l y discussed below. Figure 12. Smooth-surfaced d i o r i t e bedrock forming the s l i d i n g surface near the head of a debris avalanche in the upper Capilano catchment. Figure 13. Road-caused debris avalanches resu l t ing from the collapse of r o a d - f i l l and s ide-cast materials on a 39° slope in the Howe Sound area. 41 p Natural slope steepness. Large landsl ides (over 90 m in area) related to roads did not occur on val ley sides with slopes less than 30°, while ten debris avalanches had formed on areas sloping at 36° or more. Road-caused landsl ides have not occurred in the Capilano and Seymour catch-ments where most of the access roads and logging roads are located on the j val ley f loors or on the lower va l ley sides with slopes less than 35°. Steep slopes generally require larger road cuts and the placement of larger quantit ies of f i l l and sidecast materials than gentle slopes. Sidecasting. Fai lures in the sidecast and f i l l materials most frequent-ly originated at or near the outer road surface edge, producing long narrow scars up to 3 m in depth (Figure 13). I n s t ab i l i t y p r i n c i pa l l y derives from the loose, uncompacted condition of the sidecast mantle combined with saturated conditions resu l t ing from the concentration of road surface runoff onto the s idecastings, pa r t i cu l a r l y along the incurved sections of o roads. Fai lures associated with s imi la r conditions in the Idaho Bathol ith region received detai led study by Gonsior and Gardner (1971). Swanston (1971) indicates that sidecasting and road f i l l s may encourage slope fa i lu res by overloading the slope below the roadcut and by obstructing upslope s o i l drainage. Drainage condit ions. At least eight road-caused debris avalanches were d i r e c t l y associated with excessive seepage condit ions, a consequence of inadequate road drainage f a c i l i t i e s . In suf f i c ient roadside ditch capac i t ies , often resu l t ing from cut-slope f a i l u r e s , and a lack of func-tionable cu lver t s , caused road surface ponding along f l a t sect ions, or uncontrolled surface runoff and gul ly ing on steep sections, of abandoned logging roads in Furry Creek, Brunswick Creek, Magnesia Creek and Harvey Creek catchments. Such conditions lead to saturated slope mantles and a weakened resistance to mass wasting. The plugging of roadside drains 42 by compacted snow during the spring thaw may also be a primary cause of road surface gul ly ing and, i n d i r e c t l y , a cause of mass wasting at higher a l t i tudes (Goodell, personal communication). Reyegetation of disused roads. It was found that , at a l t i tudes below 1,000 m, abandoned logging roads are usually invaded by red alder seedlings within 2 to 3 years a f te r the cessation of logging. This growth is obviously helping to protect road surfaces against gu l ly ing and the root systems are presumably adding strength to the road f i l l s and sidecastings. However, above a l t i tudes of 1,000 m natural revegetation of the roads is a slow process. At these higher a l t i tudes obsolete roadbeds and f i l l slopes may remain devoid of vegetation for many years. Disused roads in upper Magnesia Creek, for instance, remain e s sent ia l l y unvege-tated (alder seedlings occur occasional ly in wetter roadside ditches) although logging terminated here approximately 5 years ago. Kochenderfer (1970) emphasised the importance of road care a f te r logging in order to condition abandoned roads against erosion. Practices such as the removal of wooden cu lver t s , construction of waterbars, smoothing and outsloping of road surfaces and the a r t i f i c i a l revegetation of steep roadbeds and unstable f i l l slopes a f te r the completion of logging have been largely ignored in the study area. Employment of these techniques, better provision for drainage and, most important of a l l , much improved road planning and route se lect ion (Larse, 1971), are urgent requirements to ensure that timber harvesting on the steep Coast Range slopes atta ins some degree of compat ib i l i ty with sensible watershed management. 43 Natural revegetation of landsl ide scars Red alder usually colonizes landsl ide scars within 3 to 4 years a f ter the time of f a i l u r e . On exposed, compacted t i l l s i tes alder may remain the only coloniz ing vegetation for 10 years or more but on uncon-sol idated co l l u v i a l deposits or rocky debris accumulations, cedar and hemlock seedlings often establ i sh themselves amongst young alder growth before the scars are 4 years o ld . Two landsl ides in the lower Capilano watershed and one in the lower Lynn Creek catchment, a l l of which appeared to be fresh and unvegetated in 1939, are completely revegetated with young coni fer forest but are s t i l l recognisable on 1968 1:15,000 aer ia l photo-graphs. On the other hand, several scars which originated between 1952 and 1957 remained es sent ia l l y unvegetated in 1970 (landsl ides FC4, HS1, C10 and C12). This, combined with the observation that the vegetation on several ancient landsl ide scars has f a i l ed to develop beyond a low shrub-stunted coni fer growth stage, suggests that snow avalanches, which recur on many debris avalanche_sites each winter, retard the natural reveg-etat ion process. Discussion and conclusions The combination of steep slopes, shallow, weak-structured so i l mantles, heavy seasonal r a i n f a l l s , and the presence of an impermeable t i l l substratum or impermeable smooth-surfaced igneous bedrock in d i rect contact with the s o i l , predisposes certa in undisturbed slopes in the Coast Range to catastrophic f a i l u r e s . However, on the densely forested va l ley s ides, s o i l mass wasting occurs at a subdued rate compared with other forested mountain regions for which data are ava i lab le . Consequently i t would be misleading to claim that l ands l id ing , under undisturbed 44 condit ions, i s of great economic importance in the study area. On the other hand, man's a c t i v i t i e s can substant ia l ly increase the importance of mass wasting to the point where landsl ides may be a serious threat to the permanence of the so i l resource of certain steep slopes. Measure-ments indicate that the average quantit ies of s o i l and rock debris d i s -lodged by landsl ides on the more unstable c l ea r f e l l ed slopes with roads are more than 10 times larger than the average quantit ies transported on comparable undisturbed slopes (Table 7). As f l u v i a l surface erosion, small scale mass movements and f l u v i a l gu l ly ing are not considered in these estimates the increases in to ta l erosion resu l t ing from logging and road construction i s , presumably, much greater. Gonsior and Gardner (1971), for instance, c i t e sediment production increases of more than a thousandfold a f te r roadbuilding and c l e a r f e l l i n g in steep forested water-sheds in Idaho. Fredriksen (1970) found that to ta l stream-carried sediment y i e l d from a small watershed in western Oregon a f te r the completion of roadbuilding and patchwork logging-, was approximately 100 times greater than the sediment y i e l d from a nearby undisturbed watershed. Although rather dramatic increases in mass wasting fol lowing timber harvesting on some areas o f the steep Coast Range slopes are indicated by Table 7, i t i s apparent that a^great d i ver s i t y exists in the natural s t a b i l i t y of steep areas. The survey established several general points d i r e c t l y related to th i s variable s t a b i l i t y . (1) Long uniform slopes are more susceptible to lands l id ing than short, broken, stepped, slopes with large amounts of exposed bedrock. (2) Poorly drained hollows, l i nea r drainage depressions, and wet seepage zones are more l i k e l y to be landsl ide i n i t i a t i o n points than well drained convex interf luves separating drainage depressions. 45 (3) Mountainsides with slopes greater than 30° but less than 40° are most l i k e l y to suf fer landsl ides. (4) C l ea r fe l l i n g and road construction increase vu lne rab i l i t y to the physiographic factors of points (1), (2) and (3). Although general information of the sort obtained in the survey provides guidelines for f i e l d project planning, c r i t e r i a to be employed in making (a) r e l i ab l e assessments of the landsl ide potential of natural and a r t i f i c i a l slopes and (b) meaningful predictions of slope s t a b i l i t y changes a f te r logging or roadbuilding should, i d e a l l y , be based on understanding of the basic processes and mechanisms involved in landsl ide i n i t i a t i o n . To gain such understanding, three detai led studies were made of the hydrological and strength character i s t i c s of representative slope materials and of some of the processes which render them susceptible to f a i l u r e . 46 CHAPTER 2  PIEZOMETRY STUDIES ON STEEP SLOPES Introduction On the steep, soi l-covered slopes of the Coast Mountains, downslope drainage is predominantly by subsurface flow (Chamberlin, 1972). The p ro l i f e ra t i on of springs and seepages from the s o i l mantle, pa r t i cu l a r l y at the soil-impermeable t i l l interface as exposed along road cutslopes and at the headscarps of lands l ides, suggests that the basal s o i l horizons are susceptible to saturation and that th i s condition may pers i s t for long periods of the year in drainage depressions and in other preferred sub-surface drainageway zones. It i s well known that saturated s o i l mantles favour mass wasting, a fact borne out by the common association of land-s l ides with wet slope s i t e s . According to Cedergren (1967), s o i l water lowers s t a b i l i t y and contributes to slope fa i l u re s by:-(a) causing s o i l par t ic les to migrate to escape e x i t s , result ing in piping and erosional f a i l u r e s . (b) reducing or el iminating cohesive strength. (c) increasing neutral pore water pressures and thereby reducing e f fect i ve stresses and shear strength. (d) producing hor izonta l l y inc l ined seepage forces which increase downslope tangential forces on so i l masses and the p o s s i b i l i t y of f a i l u r e . (e) lubr icat ing f a i l u r e planes a f te r small i n i t i a l movements occur. (f) supplying an excess of f l u i d that becomes trapped in s o i l pores during earthquake or other severe shocks and promotes l i q u i -fact ion f a i l u r e s . 47 During heavy rainstorms the free water surface (phreatic surface) in the shallow, steepland so i l s approaches the s o i l surface and accen-tuates some or a l l of the mechanisms described by Cedergren. Possibly a l l of these mechanisms influence the i n i t i a t i o n of f a i l u re s in the study area but i t i s l i k e l y that the ef fect of s o i l water on the e f fec t i ve stresses in the s o i l mantle and the influence of seepage pressures are the factors of most importance. The pore water pressure in the s o i l i s related to the piezometric head (the height to which water r ises in a piezometer) by the expression 2 2 u = pore water pressure, kg/m ( lbs/ f t ) hp = piezometric head, m ( f t . ) 3 3 v, = unit weight of water, kg/m ( lbs/f t ) w The development of pos i t ive pore water pressures affects the stab-i l i t y of a sloping s o i l mass by reducing the e f fec t i ve stress acting in a d i rect ion perpendicular to the potential f a i l u r e surface. The empirical Coulomb's Law which can be written as s = c ' + a* tan 0' — - (2) 2 2 s = s o i l shear strength, kg/m ( lbs/ f t ) 2 2 c ' = e f fec t i ve so i l cohesion, kg/m ( lb s/ f t ) 2 2 a' = e f fec t i ve normal stress, kg/m ( lbs/ f t ) 0' = e f fec t i ve internal f r i c t i o n angle, degrees indicates that the s o i l shear strength i s d i r e c t l y dependent on the magnitude of the e f fec t i ve normal stress a ' = a - u — — - — (3) a = tota l normal s t ress , kg/m ( lbs/f t ) 48 In order to develop th i s reasoning further, consider the diagram in Figure 14 showing part of a s lop ing, saturated, cohesionless s o i l through which water i s seeping downslope. The s o i l rests upon an imper-meable t i l l layer and thus represents a s i tuat ion which is not uncommon in the study area. The free water surface coincides with the s o i l surface Figure 14. Flow net and neutral stresses for a s lop ing, i s o t rop i c , saturated so i l mantle rest ing on impermeable substratum. which i s pa ra l l e l to the potential s l i de plane at the t i l l surface. It is assumed that the so i l i s i sot rop ic or that the maximum s o i l permeability is in the d i rect ion para l le l to the slope, permitting construction of a simple flow net. The flow l ines (the phreatic surface) and ^ are pa ra l l e l to the s o i l surface while the equipotential l ines 0n and 0 n +] are normal to i t . This highly ideal i sed flow s i tuat ion is exactly the 49 same as that assumed by Taylor (1948) in his s t a b i l i t y analysis of an i n f i n i t e , cohesionless slope with seepage occurr ing. The pore water pressures operating on the s o i l ' s basal plane, delineated by a n and a ^ i , are shown as u = u ,, and are equivalent to the ve r t i ca l Ii = n+l n n+l n h n +-| mul t ip l ied by the unit weight of water. The to ta l u p l i f t force u acting on a^ - a n + ^ is U = u n > A l - (4 ) assuming unit length in the th i rd dimension. Taylor shows that the e f fec t i ve normal stress a' at depth z i s a' = (YS " Y w ) . z - co s 2 a (5a) = Yk« z « c os a (5b) 3 YS = saturated so i l unit weight, kg/m 3 Y^  = buoyant s o i l unit weight, kg/m and that the tangential stress T, acting para l le l to the slope on the potential f a i l u r e plane is represented by T = Y s -z . s ina . cosa (6) The ef fect of seepage forces applied by the moving water to the s o i l skeleton through f r i c t i o n a l drag are included in the computation of the tangential stress T. The seepage forces cause the tangential stress to be proportional to the saturated unit weight of the s o i l whereas the e f fec t i ve normal stress a ' i s proportional to the buoyant unit weight. It follows from consideration of the Mohr-Coulomb c r i t e r i a for s o i l strength (Lambe and Whitman, 1969) that, for cohesionless mater ia l s , -i = tan 0' - - (7a) Y = — tana - (7b) Y b and that the steepest possible stable slope a c r i s 50 a c r = tan tan 0 ' ) (8) As i s t y p i c a l l y about ha l f Ys for many cohesionless mater ia ls , the saturation of a s o i l mantle accompanied by downslope seepage reduces the maximum stable slope (dec l i v i t y ) to approximately one half that for an unsaturated cohesionless s o i l mantle. Furthermore, the safety factor FS for a saturated cohesionless slope with seepage is defined as F S = — tan 0' (9) Y S tana which y ie lds a value approximately one hal f that for a slope with no saturation and downslope subsurface flow. Two aspects of the groundwater hydrology of the steep slopes in the Coast Range appeared to warrant invest igat ion: (a) establishment of the re lat ionsh ip ( i f i t ex i sts ) between the amount of r a i n f a l l received and the piezometric head at the base of the s o i l mantle. (b) ascertainment of the maximum pore water pressures which develop in the steepland so i l s on varying topographies, pa r t i cu l a r l y in l ands l ide-susceptible drainage depressions. Both aspects are int imately t i ed into the estimation of in s i t u e f fect i ve stresses. I f a defineable re lat ionsh ip ex ists between r a i n f a l l and piezometric head r i s e , then i t should be possible to predict the groundwater conditions i f given measured r a i n f a l l quant i t ies . Moreover, i f the maximum pore water pressures that are l i k e l y to develop during extreme storm periods could be predicted, then th i s information could be used to quant i tat ive ly define the most c r i t i c a l s t a b i l i t y conditions of potent ia l l y unstable t e r r a i n . To investigate the rainfall-groundwater conditions on steep slopes, r a i n f a l l and piezometric measurements were 51 undertaken on representative s i t e s . Piezometric instrumentation Four s i t e s , three of which are considered to be potent ia l l y unstable, were selected for the piezometry study. At each s i t e a l i ne of piezometers, each recording the maximum piezometric head or maximum pressure head at the soi1-impermeable t i l l or soil-bedrock in ter face, was establ ished. S ite SI i s a very shallow, l i near depression located on an uneven poorly-drained slope in the middle Seymour catchment. The so i l s in the depression are shallow, stoney and rest on an unweathered t i l l mantle. S ite S2 consists of a narrow, steep depression which originates in a gentle, forested hollow and extends downslope onto a steep c l ea r f e l l ed slope. The upper part of the depression i s underlain by shallow so i l s over bedrock but at lower elevations the s o i l s are deeper and rest on com-pacted t i l l . A small debris s l i de has recently formed in the upper part of the depression while a dry rocky channel extends along the bottom of the depression's lower section but at no time was surface runoff observed at th is s i t e . S i te S3 is located on a very well drained, convex slope which forms an in ter f luve between two smal l , t r ibutary stream va l leys . The r e l a t i v e l y deep so i l s on th i s slope are underlain by coarse colluviurn and compacted t i l l ; The s i t e was c l e a r f e l l ed and burnt approximately 5 years ago. S ite Cl i s a steep, poorly drained depression underlain by stoney s o i l s which show evidence of former i n s t a b i l i t y . The depression i s morphologically s im i l a r to an adjacent depression in which landsl ide Cl3 occurred in 1969. S i te Cl i s , therefore, regarded by the present wr i ter as a l i k e l y future landsl ide locat ion. The lower parts of the s i t e are drained by a smal l , ephemeral stream channel. Figure 15. Topographic and s o i l character i s t i c s of the s i te s instrumented with piezometers. 53 Details of the instrumented s i tes and the locations of the piezo-meters are shown in Figure 15. Two types of piezometers were i n s t a l l e d . Both types of instruments are considered to be simple piezometers rather than observation we l l s , in that they respond to water pressures at spec i f i c locations within the s o i l mantle. Figure 16 i l l u s t r a t e s the deta i l s of these instruments. The larger diameter piezometer tubes employ powdered cork which adheres to a steel rod to indicate the maximum piezometric head r i s e . A foam p l a s t i c r i s e r ins ide a transparent, a c r y l i c tube performs a s im i l a r function in the smaller diameter piezometers. Preliminary tests indicated that both types of instruments recorded r e l i a b l y in saturated sand. The tubes were inserted in augered holes which penetrated the s o i l mantle to the upper l im i t s of the unweathered t i l l or the bedrock. Coarse sand was poured into each hole so that the basal portion of each tube, when forced into the sand, was surrounded by a permeable medium. Each tube was sealed in place with a r e l a t i v e l y impermeable layer of c lay. Ten cork type piezometers were established in September 1970 and an addit ional 12 double-tube piezometers were i n s ta l l ed in August 1971. The double-tube piezometers were easier to insert and read in the f i e l d than the larger tubes but the use of the th in -wal led, aluminium conduit allowed bending by snowcreep at the ground surface. Ra infa l l was measured with 2 automatic, 12 inch capacity gauges located in the upper Capilano catchment at s i t e Cl and in the upper Seymour Valley approximately 2 km WNW of s i t e SI respect ive ly. The piezometers were v i s i t ed as soon as possible a f te r each major storm. 54 rubber stopper -soil surface 8 inch steel pipe 2 inches i.d. ' w 3_2 inch aluminium conduit 5 inch i.d. o 5 inch diameter steel rod 16 inch acrylic recording tube i inch i.d.-clay layer foam plastic riser --jr^-ltf*powdered cork nylon gauze cover water intake holes uniform sand layer — unweathered till Figure 16. Details of the two types of piezometers used in the study, (a) As water level r i ses in tube powdered cork i s carr ied upwards. After reaching point of maximum r i se water descends leaving cork adhered t6 steel rod which can. be withdrawn from tube for measurement of piezometric head, (b) Rising water level in ijiner tube carr ies buoyant, foam-plastic r i s e r upwards. The water f i l m between f l oa t and a c r y l i c tube permits r i s e r to remain at point of maximum water r i s e which i s measured by removing inner tube. 55 Results Heavy r a i n f a l l was found to cause sharp increases in the piezometric head at most piezometer s tat ions. Table 9 l i s t s the maximum recorded r i se at each stat ion and the equivalent pore water pressures. The average maximum piezometric head for a l l stations within drainage dep-ressions was 585 mm whereas that for the stations on the well drained, convex slope was only 103 mm. Figure 17 i l l u s t r a t e s the seasonal trends in piezometric head at two stations during 1970, 1971 and 1972. Groundwater levels in drainage depressions remained low during the winter months (January to Apr i l inc lus ive) but with the onset of snowmelt in A p r i l , levels rose and thereafter f luctuated about a r e l a t i v e l y high average level unt i l l a te November when colder weather and a change from ra in to snow caused decl ine. However, on the well drained, convex slope detectable r i ses in the pressure head only occurred during the largest storms. The groundwater level in drainage depressions responds in a very sens i t ive manner to r a i n f a l l as shown in Figure 18. Within one hour a f te r the commencement of heavy rain of the intense but b r ie f September storm, the level continued to r i se at a rate of approximately 60 mm per hour. Unfortunately, the time of maximum r i se was not determined but was presumed to coincide with the cessation of heavy r a i n f a l l at 8 p.m. on the 10th September. The relat ionships between piezometric head and accumulated da i l y r a i n f a l l (24 hour tota l r a i n f a l l ) at 19 piezometer s i tes are shown in Figures 19, 20 and 21. The calculated curves which best describe these data (Table 10) are expressions of the polynomial Y = b Q + b 1 1 + b 2 X - (10) 56 Table 9 Maximum Piezometric Head and Equivalent Pore Water Pressure  Recorded at Piezometer Stations in the Coast Range, Southwestern B r i t i s h Columbia Piezometer Maximum piezometric head Equivalent pore water pressure mm inches kg/m2 l b s / f t 2 SI -1 737 29.0 737 150.8 SI-2 457 18.0 457 93.6 SI-3 886 34.9 886 181.5 SI-4 432 17.0 432 88.4 SI-5 711 28.0 711 145.6 Sl-6 732 28.8 732 149.8 S2-1 356 14.0 356 72.8 S2-2 737 29.0 737 150.8 S2-3 406 16.0 406 83.2 S2-4 Did not function cor rect ly S2-5 853 33.6 853 174.7 S3-1 132 5.2 132 27.0 S3-2 81 3.2 81 16.6 S3-3 97 3.8 97 19.8 Cl-1 Did not function cor rect ly Cl -2 292 11.5 292 59.8 Cl-3 597 23.5 597 122.2 Cl-4 404 15.9 404 82.7 Cl-5 615 24.2 615 125.8 C l -6 * 813 32.0 813 166.4 C l -7 * 508 20.0 508 104.0 C l -8 * 409 16.1 409 83.7 k Maximum piezometric heads and pore water pressures a f te r the rapid ablation and disappearance of a 1.7 meter deep snowpack are l i s t e d be.l ow. Cl-6 851 33.5 851 174.2 Cl-7 889 35.0 889 182.0 Cl-8 1,029 40.5 1,029 210.6 Figure 17. Piezometric head variat ions during 1970, 1971 and 1972 at stations Cl-4 and C l -5 , upper Capilano catchment. 58 Figure 19. Relationships between piezometric head and da i l y r a i n f a l l Figure 20. Relationships between piezometric head and da i ly r a i n f a l l . 61 Figure 21. Relationships between piezometric head and da i l y r a i n f a l l . 62 Y = piezometric head in mil l imeters X = accumulated da i l y r a i n f a l l in mil l imeters Table 10 Regression Equations Describing Relationships Between Piezometric  Head (Y) and Daily Ra infa l l (X), Coast Range, Southwestern B r i t i s h Columbia 2 Piezometer Equation of best f i t R * SE^** SI-1 Y = 432.5 - 2 , 172 . l l + 4.3X .91 51.7 13 Sl-2 Y = 388.7 - 2,017.9j + 0.7X .79 44.5 13 SI-3 Y = 537.5 *** 3 3 37 5 • 3 ^  + 5.8X .89 80.8 15 SI-4 Y = 81.6 - 590.7j + 4.3X .86 46.7 15 SI-5 Y = 641.6 - 2,608.0^ + 1.5X .92 34.8 14 Sl-6 Y = 753.7 — 3) 557 • 3"^ " + 0.2X .94 33.9 13 S2-1 Y = 242.0 S2-2 Y = 189.9 S2-3 Y = 333.6 + 2.IX .85 36.2 10 + 7.3X .93 53.0 12 + 7.2X .96 45.9 12 Cl-2 Y = 342.3 - 1,591.3^ + 0.5X .98 11.0 10 Cl-3 Y = 289.4 - 1,225.81 + 3.9X .96 35.0 10 Cl-4 Y = 220.0 - 705.2^- + 2.OX .87 32.0 21 Cl-5 Y = 377.8 - 1,506.91 + 3.OX .92 41.1 21 Cl-6 Y = 342.8 - 1,006.6^ + 4.4X .88 59.4 21 Cl-7 Y = 274.6 - 1 ,225.2^-+ 2.7X .97 27.6 9 Cl-8 Y = 33.6 - 48.2^ + 4.IX .97 28.7 6 2 *R = mult iple coe f f i c ient of determination **SE^ = standard error of estimate The curves vary considerably from piezometer to piezometer, presumably a re f l ec t i on of spat ia l differences in s o i l physical prop-e r t i e s , source area of subsurface flow and configuration of the t i l l or bedrock-soil inter face. Relationship between piezometric head, r a i n f a l l and subsurface flow A r i se in the free water surface and, consequently, in the piezometric head w i l l only occur i f the supply of water from the s o i l surface and from interf low and excessive to any transpi rat ion al lo s s , exceeds the seepage a b i l i t y of the s o i l . The piezometer study, combined with the findings of Chamberlin (1972), indicated that, during heavy r a i n f a l l , water passes readi ly through the so i l mantle and moves over the impermeable t i l l or bedrock in a shallow, saturated layer at the base of the B horizon. Flow is directed towards the drainage depressions from adjacent slopes. The concentration of subsurface water in the depressions causes a rapid increase in the thickness of the saturated zone and a concomitant increase in the s o i l ' s a b i l i t y to conduct water downslope. Swanston (1967a) proposed that the relat ionships between increments of r a i n f a l l ( A P ) , the downslope subsurface flow rate (AQ) and changes in piezometric head (Ah ) could be described by the simple model depicted in Figure 22. During a rainstorm, small i n i t i a l values of Q and a large volume of water moving downslope cause hp to increase rap id ly . As hp increases, Q also becomes larger which, u l t imate ly , leads to a condition of equi l ibr ium between r a i n f a l l , seepage and piezometric head. The nature of th i s re lat ionsh ip depends on the so i l mantle's a b i l i t y to conduct water downslope. Spatial v a r i a b i l i t y in s o i l permeab-i l i t y may p a r t i a l l y account for the variable response of the piezometric head between piezometer stations during the same storm. The saturated 64 AP Figure 22. Hypothetical re lat ionsh ip between r a i n f a l l , subsurface seepage and piezometric head. (After Swanston 1967a). permeabil it ies of eight large s o i l samples taken from the lower B horizon near piezometers SI-2 (samples 1 and 7) , SI-3 (samples 2, 3 and 8) and Cl-5 (samples 4, 5 and 6) were determined in the laboratory by a constant head method s im i la r to the method described by Klute (1966). C i rcu lar disks of s o i l , approximately 14 cm in diameter and 10 cm long, were ca re fu l l y trimmed from large, intact s o i l blocks and placed in 16 cm long, a c r y l i c cyl inders f i t t e d with a perforated plate at one end. The sides of the disks were sealed to the a c r y l i c cy l inder walls with a sealing compound. The samples were selected so that 6 tests deter-mined the permeability para l le l to the natural slope in the f i e l d and 2 tests provided estimates of the ve r t i ca l permeabil ity. Decayed root channels as well as undecayed roots were interspersed within the so i l disks which were soaked in water for several hours pr ior to te s t ing . The test results are presented in Table 11. Table 11 Saturated Permeability Coeff ic ients of Steepland Subsoils, Coast Range, Southwestern B r i t i s h Columbia Sample Orientation Permeability Coeff ic ient cm/sec 1 para l le l to slope 0.0058 2 0.0056 3 " " " 0.0051 4 " " " 0.0055 5 0.0007 6 " " " 0.0010 7 ve r t i ca l 0.0005 8 " 0.0030 9* " 0.00002 * Sample of unweathered, compacted t i l l . The mean downslope permeability was 0.0039 cm/sec (standard deviation = 0.0024 cm/sec) equivalent to 337 cm/day which corresponds to a rapid class of permeability according to O'Neal 1952 (in Klute 1965). The t i l l , by comparison, has a very slow permeabil ity. These results imply that the B horizon i s capable of transmitt ing large volumes of water downslope and that substantial changes in permeability occur from one part of a slope to another. In a recent study, Aubertin (1971) demonstrated that macropores, including those resu l t ing from the decay of tree roots, greatly increase 66 the permeability of forest so i l s by permitting rapid transmittance of water along paths of low resistance. Variations in the cont inuity and density of root channels within the B horizon of drainage depression so i l s probably contribute to var iable permeability and to uneven piezo-metric surfaces. Estimation of maximim pore water pressures During the period of observation, the maximum recorded da i l y r a i n -f a l l s in the upper Seymour catchment and the upper Capilano catchment were 74 mm and 86 mm respect ively. Many of the curves in Figures 19, 20 and 21 suggest that da i l y r a i n f a l l s of 130 to 150 mm would cause the piezometric head to r i se to the s o i l surface. According to the r a i n f a l l records at Seymour F a l l s , r a i n f a l l events of these magnitudes have an expected recurrence interval of approximately 2 years. As the pressure head shown in Figure 19, 20 and 21 is probably less than the height of the phreatic surface above the impermeable substratum under conditions of downslope seepage, i t can safely be assumed that most drainage depression so i l s saturate completely during very heavy rainstorms of th i s magnitude. The presence of ephemeral surface runoff channels and inc ip ient channels in many depressions substantiates th is assumption. A b r ie f reconsideration of Figure 14 permits a comparison of the calculated pore pressures, assuming complete saturation and seepage para l le l to the slope, with the pressures measured in the f i e l d . If values of 90 cm and 35° are assigned to z and a respect ive ly, then the pore water pressure u at the base of the s o i l mantle in Figure 14 is 2 U = Z . y W . C O S A " (11 ) = 604 kg/m2 67 which corresponds to a piezometric head of 604mm. However, th i s pressure was exceeded at eight piezometer stations with comparable slopes (S l -1 , S l -3 , S l - 5 , S l -6 , S2-2, S2-5, Cl-5 and Cl-6) even though the s o i l mantle at four of these s i tes was less than 90 cm deep, ind icat ing that pore pressures are larger than the pressure given by equation 11 and may approach the hydrostatic pressure. In Figure 14 the hydrostatic pressure 2 = 900 kg/m for z = 90 cm. The curves in Figures 19, 20 and 21 indicate that pore pressures of th i s magnitude can be expected in some drainage depression s i tes when da i l y r a i n f a l l s exceed 120 mm. The maximum recorded piezometric heads at three stations occurred during the 1972 snowmelt period (Table 9). At the beginning of March 1972 the snowpack at s i t e C-l was approximately 1.5 m deep but when the s i t e was rev i s i ted on the 3rd May the snowpack had almost disappeared. During A p r i l , ra in on rapidly melting snow caused the piezometric head levels at stations C l -6, .C l -7 and Cl-8 to r i se to heights above the ground surface. Poss ibly, at some stage during the snow ablation period, the s o i l mantle and the lower part of the snow pack saturated to produce large piezometric heads. Klock (1972) discovered that the water retention capab i l i t y of a sandy so i l was increased and the permeability was decreased at temperatures close to 0°C. Low temperature snowmelt water may cause a decrease in the conductiv ity of the Coast Range steepland so i l s which promotes a buildup of pressure head in the so i l mantle. The piezometric data provides no ind icat ion that c l e a r f e l l i n g promotes increased s o i l pore water pressures although the small numbers of piezometers involved in the study precluded s t a t i s t i c a l test ing of t h i s . The average maximum pore water pressure recorded at the four stations established in forested drainage depressions (C l -7, C l -8 , Sl-6 2 2 and S2-5) was 626 kg/m compared with 573 kg/m for the th i rteen 68 staions i n s t a l l ed in c l ea r f e l l ed drainage depressions. Pore pressures and e f fec t i ve stresses The piezometer study established that pore water pressures of 800 2 kg/m or more may develop in the basal regions of the s o i l mantle in drainage depressions. The magnitude of the e f fec t i ve stress changes which accompany such pressure r i ses can be gauged from Figure 23 which compares a moist, sloping s o i l mantle with a saturated so i l mantle with seepage. Typical f i e l d values of ym» Ys» <*> u, and z are used for computing the e f fec t i ve normal stresses acting on the potential f a i l u r e 2 plane. A pore water pressure increase of 800 kg/m at the base of the s o i l mantle causes a 76 percent decrease in the e f fec t i ve normal stress a ' which, as shown by equation 2, reduces the s o i l ' s resistance to shear. Under moist s o i l conditions Chamberlin (1972) measured pore water 2 tensions (negative pore water pressures) of approximately 200 kg/m ( 20 cm water) near the base of the forest so i l mantle. Negative pressures raise the normal e f fec t i ve stress. Consequently, the value for the normal e f fec t i ve stress fo r the moist condition shown in Figure 23 is a conservative estimate as pore water tensions were not taken into consideration. The effects of pore pressures on slope s t a b i l i t y w i l l be examined in more deta i l in Chapter 4. Conclusions (1) Heavy r a i n f a l l causes a rapid r i s e in the piezometric head measured at the base of the so i l mantle in drainage depression s i te s . However, on well drained slopes measureable piezometric heads only develop during very large storms. 69 Figure 23. Ef fect ive normal stress acting on a potential s l i de plane in the case of (a) a moist, sloping s o i l mantle with no seepage and (b) a saturated, sloping s o i l mantle with seepage. 70 (2) At most piezometric stations a well defined, cu rv i l i nea r re lat ionsh ip ex ists between the da i l y r a i n f a l l received and the piezo-metric head. Such a re lat ionship permits the estimation of pore water pressures i f r a i n f a l l i s measured. 2 (3) In many drainage depressions, pore water pressures of 800 kg/m or more can be expected in the lower B horizon during intense rainstorms. Pressures of th i s magnitude result in substantial reductions in the e f fect i ve normal stress acting on potential f a i l u r e planes at the base of the B horizon which, in turn, decrease the slope s t a b i l i t y . 71 CHAPTER 3 THE IMPORTANCE OF TREE ROOTS TO SLOPE STABILITY  Introduction Although the role played by tree roots in the s t a b i l i z a t i o n of steep slopes has been conjectured by several investigators (Wentworth 1943, White 1949, Penck 1953, Bishop and Stevens 1964, Swanston 1969, Gray 1969, Gonsior and Gardner 1971, Fujiwara 1970, and Rice and Foggin 1971) very l i t t l e quantitat ive work has been accomplished which indicates to what degree and by what mechanisms the tree root network enhances the s o i l ' s resistance to f a i l u r e . Less i s known about the rap id i ty at which the added mechanical strength provided by roots, disappears a f te r a forest stand i s c l e a r f e l l e d . There is l i t t l e doubt that the s t a b i l i z i n g functions of root systems are very important, pa r t i cu l a r l y where s o i l mantles are shallow and mass wasting occurs within the rooting zone as i s the case in the Coast Mountains study area. Where the tree roots penetrate the s o i l down to the compacted t i l l or substratum, root systems may not only add mechanical strength to the so i l mantle but also modify the stress d i s t r i bu t i on within the s o i l mass by t ransferr ing surcharge loads to the substratum. Furthermore, i f roots penetrate the basal t i l l or cracks in the bedrock then the root network helps to attach the s o i l mantle to the underlying substratum. On the other hand, where trees are blown over or overturned by grav i ty, tree roots are responsible for loca l disturbances of the s o i l mantle (Lutz and Gfiswold 1939, Stephens 1956 and Stone 1969) which may contribute to larger, damaging slope fa i l u re s (Wright and M i l l e r 1952, Schweinfurth 1967, Jackson 1966, and Swanston 1969). However, these 72 adverse effects are generally considered to be i n s i gn i f i c an t , at least in temperate regions, compared to the s t a b i l i z i n g functions performed by roots (Gray 1969). This c lea r l y i s the case in the B r i t i s h Columbia Coast Range. In f a c t , in a region where c l e a r f e l l i n g i s associated with an increased frequency of lands l ides , i t could be argued that the strength reinforcement provided by roots, or lack of i t , i s the pr inc ipa l factor cont ro l l i ng s o i l s t a b i l i t y . In th i s chapter an attempt i s made to assess the importance of tree root systems to the shearing resistance of the so i l and the rap id i ty at which the additional strength provided by roots decreases a f te r c lea r -f e l l i n g . Rooting habits of forest trees on the Coast Range slopes The root development of forest trees i s a most important consider-ation in an evaluation of the effects of forest vegetation on slope s t a b i l i t y . Deep rooted species are more l i k e l y to o f fe r greater protection against s o i l f a i l u r e than shallow rooted species and widely-spreading root systems which intermingle with adjacent root systems serve a greater protective function than re s t r i c ted root systems which form a dense, i so lated ba l l at the base of the tree stem (Kostler et_ al_, 1968). Although a spec i f i c study of root development was not conducted in the present study, general information on rooting habits was provided by exposed root systems at the margins of lands l ides, along fresh road cuts and at the s i tes of wind thrown trees. The rooting habits of the major tree species in the study area are strongly influenced by the physical character i s t i c s of the steepland s o i l s , pa r t i cu l a r l y t he i r depth. On shallow so i l s i tes where basal t i l l 73 or bedrock occurs within 50 centimeters of the ground surface, western red cedar, western hemlock and mountain hemlock develop p la te l i ke root systems which display vigorous l a te ra l root development (Figure 24). The western cedar root system shown in Figure 24 was exposed when the parent tree (98 centimeters dbh) was blown over a few months before the photograph was taken. The oval root p late, which measured 8.5 meters by 5.0 meters, consisted of a dense radia l system of large l a te ra l roots up to 25 centimeters in diameter which gave r i se to smaller second-ary, t e r t i a r y and quarternary roots. The deepest l a te ra l roots forming the basal surface of the root plate had extended to a depth of 30 c e n t i -meters where they had been t i g h t l y pressed against the smooth surface of the d i o r i t e bedrock. Roots up to 4 centimeters thick had penetrated a large, open j o i n t in the d i o r i t e (Figure 25). S imi lar p l a t e - l i k e root systems were observed on overturned western and mountain hemlock trees growing on shallow so i l s over bedrock. Large Douglas f i r trees growing on shallow, rocky so i l s tend to develop massive l a te ra l roots which may extend more than 5 meters from the root stock. Lloyd e_t al_ (1956) point out that p l a te - l i k e root systems are just as charac te r i s t i c of Douglas f i r as they are of other species where s o i l s are shallow. On deeper so i l s the root systems of Douglas f i r and cedar extend into the lower B horizon. A part ly exposed Douglas f i r root system attached to a stump 140 centimeters in diameter, extended 96 centimeters in depth to the upper l im i t s of a compacted t i l l substratum on a steep, c l ea r f e l l ed slope in the middle Seymour catchment. An excavation under the stump showed that the root system consisted of at least three large, cent ra l l y - s i tua ted , ve r t i ca l roots up to 30 centimeters in diameter and many large l a te ra l s with numerous branches mainly concentrated in the 74 Figure 24. The exposed root system of a recently overturned western red cedar. The lower surface of the p l a t e - l i k e root system had developed in d i rec t contact with smooth bedrock. Figure 25. Recently exposed western red cedar root penetrating an open j o i n t in d i o r i t e bedrock (a). 75 upper 50 centimeters of s o i l . Smaller secondary and t e r t i a r y roots, many derived from the central ve r t i ca l s inkers, were common in the lower B horizon but did not enter the t i l l substratum. Although Douglas f i r i s c ha rac te r i s t i c a l l y a deep rooting species (McMinn 1963) th i s a t t r ibute is not manifested in the steepland so i l s where basal t i l l or bedrock forms an impedence to roots. Freshly-exposed, unweathered t i l l surfaces in the bottoms of land-s l i de scars often exh ib i t imprints of roots which followed tenuous courses across the t i l l ' s surface, but, except for small roots embedded in the top few centimeters of t i l l , very few larger roots were observed to penetrate the tough, unweathered t i l l matrix. Schlots et al_ (1956) noted that Douglas f i r root s - fa i l ed to penetrate compacted Vashon t i l l in western Washington but Patr ic and Swanston (1968) discovered that cedar roots f ree ly entered s im i l a r compacted t i l l in Alaska. Small cedar or mountain hemlock roots did enter cracks in fractured compacted t i l l in the upper regions of landsl ides CIO and C12. In summary, the roots of mature Douglas f i r , cedar and western hemlock u t i l i z e most of the so i l volume on the steep slopes underlain by shallow so i l s and, although they do not generally penetrate the compacted t i l l matrix, roots do extend through cracks and jo in t s in the t i l l or the bedrock substratum. Presumably, root systems contribute s i g n i f i c an t l y to the s t a b i l i t y of the forested slopes by binding the s o i l mass together and by helping to anchor the s o i l mantle to the substratum. A f i e l d study of the ef fect of tree roots on the s o i l ' s resistance to  shear The influence of tree, roots on the shearing strength of so i l s has, to the present w r i t e r ' s knowledge, only been measured d i r e c t l y in Japan. 76 Endo and Tsuruta (1968) conducted a number of f i e l d shear tests on a uniform nursery s o i l containing varying quantit ies of l i v i n g alder roots. Using a large shear-box device equipped with a gauge or recorder to measure deformation, they found that s o i l shear strength increased with the bulk weight of roots in the s o i l and that the re lat ionsh ip could be adequately described by a l i nea r equation with the form S r = a(R + b) - (12) 2 = increased s o i l shear strength due to roots, kg/m 3 R = bulk weight of fresh roots, g/m a and b = empirical constants A s im i la r study by Takahasi (1968) indicated that the apparent cohesion of a f i e l d s o i l depended on the quantity of white birch roots in the s o i l according to the formula C = a - bD + cW — (13) 2 C = apparent so i l cohesion, kg/m D = depth of shear plane, cm 2 W = weight of birch roots, kg/m (weight of roots per volume of s o i l , D x 1m x lm) a, b, c = empirical constants The d i f f i c u l t i e s involved in measuring the component of s o i l strength provided by the tree root network are compounded under conditions of s o i l heterogeneity. In f ac t , the determination of meaningful shear strength parameters of extremely variable bouldery and gravel ly so i l s i s a d i f f i c u l t i f not impossible task. The problems were minimized by the Japanese workers by studying r e l a t i v e l y homogeneous so i l s on f l a t , uniform s i t e s . However, extrapolation of t he i r results to heterogeneous mountain s o i l s would be of dubious value. Accordingly, the mountain s o i l s of the 77 Coast Range were investigated d i r e c t l y in an attempt to determine the re lat ionsh ip between the tree root network and the steepland s o i l s ' resistance to shear. Methods. Direct shear tests were conducted on the subsoils of two f resh ly c l ea r f e l l ed slopes formerly occupied by mixtures of old growth Douglas f i r , cedar and hemlock and on the subsoils of a forested slope supporting ah old-growth cedar-western hemlock stand. The so i l s at the test s i tes were representative of s o i l Categories 1 and 2, a l l possessing yellowish-brown, gravel ly sandy-loam B horizons generally less than 100 cm deep and rest ing on compacted basal t i l l or d i o r i t e bedrock. A portable, steel shear-box, 30.5 cm x 30.5 cm x 15.3 cm (1 f t . x 1 f t . x 0.5 f t . ) in s ize was employed (Figure 26). The shearing force was provided by a small hand winch while the magnitude of the shearing force was recorded with a set of high qua l i ty scales of capacity 90.7 kg (200 lbs) and s e n s i t i v i t y of 0.22 kg (0.5 l b s ) . At each test s i t e an excavation was dug to i so la te a r e l a t i v e l y undisturbed block or column of subsoil approximately the same s ize as the shear-box. Each column was ca re fu l l y trimmed with a knife so that the shear-box f i t t e d snugly over the column which supported the f u l l weight of the box and any add i t -ional load applied to the top of the box. The blocks were sheared at t he i r bases and the shear planes normally sloped towards the val ley bottom at angles between 5° and 10°. The shearing force was applied in the downslope d i rec t i on . Attempts to conduct the tests at constant displacement rates were unsuccessful. Rather, the shearing force was increased by 4.5 kg (10 lbs) increments each 60 seconds up to the time when displacement rates increased to approximately 0.5 cm per minute. Thereafter, 2.3 kg (5 lbs) 78 E o in 6 rO f K Anchor Winch •30-5 cm-Scales 3—•f^t* -30-5 cm-Shear box constructed of 1/8" steel plate-Shear box Figure 26. Detai ls of shearbox construction .and i t s mode of operation in the f i e l d . 79 shear load increments were applied every 60 seconds un t i l sudden or complete f a i l u r e occurred. The maximum shear strength of the so i l was taken to be the maximum shear force recorded on the scales divided by the area of the so i l column. During each test the shear displacement was measured with a displacement gauge. Tests in which the so i l columns crumbled instead of f a i l i n g along a reasonably wel l-def ined shear plane were regarded as unsatisfactory and were not taken into account. Approximately 40 so i l columns were cut but only 24 successful tests were completed. At the completion of each test a 2,500 cc, r e l a t i v e l y undisturbed s o i l sample was retr ieved from the s o i l immediately adjacent to the f a i l u r e plane and taken to the laboratory for pa r t i c l e s ize analysis and determinations of moisture content, unit weight and root weight content. The technique adopted for co l l e c t i ng the so i l samples was rather laborious but e f f ec t i ve . A roughly c i r c u l a r s o i l column, approximately 20 cm in diameter and 15 cm t a l l , was i so lated at each test s i t e and a large, c i r c u l a r , metal container, 15 cm in diameter and 14.5 cm deep, was placed, open end down, on top of the column. The sides of the column were ca re fu l l y trimmed with a knife and protruding roots were cut with a secateur so that the container could be worked down over the column without ser iously disturbing the s o i l . After the container was f u l l y occupied the column was cut f lush with the open end of the container which was then sealed with a l i d . Samples were weighed in the undisturbed s tate, then broken up and the roots separated by hand and weighed and, f i n a l l y , s o i l and roots were ovendried at 100° and reweighed. Pa r t i c l e s ize d i s t r ibut ions were 80 measured by dry s ieving and, for par t i c les passing a No. 200 s ieve, by hydrometer analysis according to the methods described by Lambe(1951). The mean diameter and weight of par t i c les retained on 5 cm (2 in) sieve were measured separately. Five addit ional samples of steepland subsoils from the 40 to 60 cm depth zone were col lected in the same manner as described above for det-ermination of the saturated unit weight (y ) and the void ra t io (e). Two smal l , undisturbed samples were also taken from the unweathered t i l l sub-stratum for s im i la r determinations. Methods and results are presented in Appendix 2. A l l the s o i l columns tested contained roots, mostly less than 2 cm in diameter, and, as the shear plane depths were s ituated between 40 and 60 cm below the or ig ina l ground surface, most of the roots were presumed to be tree roots. Results and discussion. The deta i l s of the so i l s tested and the test results are summarized in Table 12. At the test loads employed (normal loads on the shear plane), shear strengths ranged from 208 kg/m to 415 kg/m2. So i l columns which did not contain large boulders or large roots over 4 cm in diameter generally f a i l e d along a well defined plane. As the columns displaced, small roots crossing the shear plane were usually sheared of f d i r e c t l y but larger roots over 0.5 cm in diameter tended to pul l out of the s o i l block or f a i l in tension. Typical shearing force-displacement curves are shown in Figure 27. So i l s with many roots (curves 4 and 15) attained the i r peak shear force at a displacement of 2.6 cm and 2.1 cm respect ive ly, but the curves for Table 12. Results o f D i r e c t Shear Tests on S o i l s Containing Roots Normal Normal Shear Shear Water Dry u n i t % % % % Root f, l o a d s t r e s s l o a d s t r e s s content weight Gravel Sand S i l t Clay content kg- kg/m2 kg kg/m 2 g/g kg/m3 g/m3 1 29.4 316.5 31.3 336.9 44.9 1,090 51.8 27.6 19.0 1.6 2; 000 2 29.1 313.2 31.5 339.3 35.7 1,037 61.6 24.2 13.9 0.3 1,300 3 29.4 316.5 20.4 219.7 36.6 1,207 67.4 16.2 16.3 0.1 2,700 4 29.1 313.2 37.2 400.3 38.1 965 41.1 47.2 11.3 0.4 9,300 5 29.4 316.5 19.5 209.9 39.6 878 51.2 40.7 7.5 0.6 400 6 29.1 313.2 22.5 241.7 48.6 910 60.0 34.2 4.9 0.9 1,300 7 29.4 316.5 19.5 209.9 33.7 917 61.5 31.1 6.7 0.7 400 8 29.2 314.3 29.9 322.2 21.3 1,377 61.9 27.4 9.2 1.5 1,100 9 29.4 316.5 29.0 312.4 20.5 1,391 61.9 27.4 9.2 1.5 700 10 29.1 313.2 26.8 288.0 33.4 1,066 59.8 31.4 7.1 1.7 3,700 11 29.1 313.2 22.2 239.2 20.6 1,112 64.2 27.6 6.2 2.0 3,300 12 29.2 314.3 22.0 236.8 25.4 1,192 60.4 29.9 8.1 1.6 1,300 13 29.1 313.2 19.3 207.5 23.6 1,341 63.1 30.3 5.3 1.3 2,000 14 29.1 313.2 20.9 224.6 18.2 1,388 69.4 26.0 3.3 1.3 400 15 29.1 313.2 36.3 390.6 30.7 1,256 69.5 20.1 10.2 0.2 7,100 16 29.2 314.3 27.2 292.9 48.6 936 50.1 35.1 12.2 2.6 2,700 17 29.2 314.3 20.0 214.8 36.1 1,074 59.1 30.8 9.1 1.0 700 18 29.4 316.5 21.8 234.3 25.3 1,066 60.9 29.4 9.7 0.1 400 19 29.4 316.5 20.9 224.6 33.5 1,066 60.9 29.4 9.7 0.1 400 20 29.2 314.3 38.6 415.0 49.8 918 70.2 22.0 6.4 1.4 8,000 21 29.4 316.5 34.0 366.2 38.8 1,045 49.4 41.2 9.4 0.1 5,300 22 29.4 316.5 34.0 366.2 31.0 1,163 65.8 27.4 6.7 0.1 2,700 23 29.2 314.3 20.4 219.7 34.6 1,149 66.5 26.0 7.0 0.5 2,700 24 29.2 314.3 21.8 234.3 36.1 1,098 56.5 34.9 8.6 0.1 400 26.1 281.1 33.5 1,107 60.1 30.0 9.0 0.9 2,513 "Includes weight o f s o i l column above shear plane plus weight o f shear box adjusted f o r slope according t o the r e l a t i o n N = W cosa. 50 4 0 E o £ 30 O O _ l o x: in 20 10 0 Curves 4 and 15 for soils with many roots Curves 3 and 7 for soils with few roots ±__L 15 0 0 10 2 0 3-0 Shear Displacement,centimeters Figure 27. Shear load vs shear displacement curves for four d i rect shear tests on steepland soi l s - containing roots. 83 so i l s with few roots (curves 3 and 7) show that the peak shear forces coincide with r e l a t i v e l y small displacements. Possibly roots impart an addit ional e l a s t i c i t y to the s o i l which enables i t to undergo considerable s t ra in without f a i l i n g completely. Because of the inherent crudity of the test ing procedure (the short-comings of d i rect shear tests discussed by Sowers and Sowers 1970, Lambe (1951) and Sowers (1963) were added to by the poor experimental control that existed during test ing in the f i e l d ) , the results of the experiment provide no more than a broad ind icat ion of the effects of the root network and the physical properties of the s o i l on i t s resistance to shear. Table 12 indicates that the s o i l columns varied considerably in texture, moisture content and dry unit weight as well as in root content. In order to ascertain the re l a t i ve influence that these variables exercise on the s o i l ' s shear resistance the data were subjected to a mult iple regression analys is . The model adopted had the form Y = b Q + b 1X ] + b 2 X 2 +• b y X y (14) 2 Y = s o i l shear res istance, kg/m 3 3 X-| = s o i l moisture content, cm /cm 3 X 2 = dry unit weight, kg/m X^ = percent gravel by weight X^ = percent sand by weight X 5 = percent s i l t by weight Xg = percent clay by weight 3 Xy = fresh root weight, g/m of s o i l To what extent the data v io late the basic assumptions of regression and just how ser iously the ex i s t ing v io lat ions a f fect the analysis are not known by the present wr i te r . The most serious inherent weakness of 84 the analysis may derive from a lack of independency among the predictor var iab les , pa r t i cu l a r l y in the cases of Xg and which, according to the corre lat ion matrix for these data, are negatively correlated (r^ 4 = 0 . 8 ) . Beginning with the tota l model (equation 3) a stepwise el iminat ion procedure was adopted in which the independent variables were eliminated one by one, el iminat ion at each stage depending on that variable d i sp lay-ing the least s i gn i f i can t pa r t i a l F value. The to ta l regression model p accounts for 70 percent of the var iat ion in s o i l shear strength (R = 0.70, n = 24). Successive el imination of X^, X^, Xg, Xg and X^ reduced the regression model to Y = b Q + b 5 X 5 + b 7 X ? -r (15) 2 which explains 60 percent of the var iat ion in strength (R = 0.60, n = 24) and i s s i gn i f i can t at the 0.01 level (F 2 20 = 16.0). Throughout the el iminat ion procedure, X-, remained the most s i gn i f i can t var iab le, exemplifying the importance of roots to the steepland s o i l s ' strength. The simple regression of s o i l shear resistance on the s o i l root content (X 7) accounted for 56 percent of the td ta l shear resistance var iat ion 2 (r = 0.56, n = 24). The equation i s Y = 229.4 + 0.02 X ? - - - - - -r (16) and is s i gn i f i can t at the 0.01 level (F^ 2 2 = 2 8 «1 )« This re lat ionsh ip applied s p e c i f i c a l l y to the conditions of normal loading during the 2 tests (approximately 315 kg/m ). The la s t term in equationl6 provides a f i r s t approximation of the s o i l ' s apparent cohesion due to the binding effects of the roots. The re lat ionship can be represented by C = 0.02 R - - (17) a 85 C = apparent cohesion, kg/m a 3 R = root weight, g/m However, as the shear tests are influenced only by the ve r t i ca l roots crossing the potential shear plane at the base of the columns, th i s empirical estimate of the s o i l ' s apparent cohesion is probably very conservative. In the undisturbed s o i l mantle, l a te ra l roots as well as ve r t i ca l roots intermingle and may sometimes graft together (Eis 1972) to help bind the s o i l into a coherent layer. Furthermore, columns could not be successful ly cut in so i l s containing roots with diameters larger than 3 to 4 centimeters. Therefore the shear tests measured only the strengthening e f fect of the smaller tree roots. A second series of d i rect shear tests were made in the f i e l d to establ i sh an estimate of the angle of internal f r i c t i o n of the steep-land subsoi l s . Nine d i rect shear tests were conducted on a g rave l ly -sandy-loam B horizon developed over compacted t i l l on a 20 degree slope in the middle Seymour catchment. The slope had been c l ea r f e l l ed 6 years previously. The subsoils selected for the tests contained no l i v i n g roots except for occasional fireweed (Epilobium angustifolium) roots less than 1 mi l l imeter in diameter and very few smal l , decayed roots. At three s i tes large, undisturbed blocks of subsoils were i so lated by careful excavation and then each block was divided into several s o i l columns separated by narrow trenches. The methods used remained the same as those used to test so i l s with roots but the loading conditions normal to the potential shear plane were var ied. Three successful tes t s , each with a d i f fe rent loading condit ion, were completed at each s i t e . The depths of the shearing planes ranged between 40 and 60 cm below the or ig ina l ground surface. At the time of tes t ing the so i l s were 86 extremely moist but not saturated. The results of the tests are pres-ented in Table 13 and shown graphical ly in Figure 28. Because only the normal stresses {a) and shear stresses (T) on a s ingle plane are known in a d i rect shear t e s t , i t i s necessary to assume that the stresses at f a i l u r e are in the ra t i o — = tan 0 (Lambe and a Whitman 1969) in the case of cohesionless s o i l s . The low clay and s i l t content of the steepland subsoils suggests that these materials are es sent ia l l y cohesionless. The curves of T vs a f i t t e d to the three sets of test results (Figure 28) produce f r i c t i o n angles of 40, 41 and 34 degrees respect ively. If values of the apparent cohesion estimated from equation 17, an internal f r i c t i o n angle of 40 degrees, and e f fec t i ve normal stress values s im i l a r to those which can be expected at the basal regions of a typ ica l sloping s o i l mantle in the study area, are inserted into Coulomb's shear strength equation, then i t i s possible to assess the importance of roots to the tota l shear resistance of the steepland so i l s in quantitat ive terms.. For a moderately dense tree root network (R = 8,000 g/m ) and an unsaturated, 90 centimeter deep s o i l s loping at 35 degrees (a 1 = 900 kg/m , see Figure 23) the shear resistance at the basal plane i s s = 160 + 900 . 0.839 =915-kg/m2 and the root network provides approximately 18 percent of the tota l shear strength. However, under saturated s o i l conditions the e f fec t i ve 2 normal stress reduces to 200 kg/m (Figure 23) and the shear resistance at the basal plane is s = 160 +200 . 0.839 =328 kg/m2 In this case the root network provides 49 percent of the tota l shear strength. Table 13 Results of Direct Shear Tests on Steepland Soi l s with no roots S ite One S i te Two S i te Three Normal Normal load stress kg kg/m2 Maximum Maximum shear shear load stress kg kg/m2 Maximum Maximum shear shear load stress kg kg/m2 Maximum shear load kg Maximum shear stress kg/m2 29.5 318 21.8 235 25.3 272 30.7 330 48.0 517 33.6 362 32.3 348 41.3 445 81.5 877 63.2 680 58.1 625 72.3 778 Dry unit weight kg/m3 1,180 1,055 1,094 Moisture content cm3/cm3 38.6 35.0 27.8 % gravel 62.0 53.8 49.9 % sand 27.3 36.4 39.1 % s i l t 8.5 8.0 11.2 % clay 2.2 1.8 0.8 With the aid of an i n f i n i t e slope analysis (Taylor 1948, Lambeand Whitman 1969) the safety factors for the two conditions of slope draina can be computed and used to check whether or not the estimates of shear strength are reasonable. The safety factor FS for a slope is defined by Lambeand Whitman as Fs = avai lable shear strength shear stress required for equi l ibr ium 1,000 Figure 28. Results of d i rect shear tests on .steepland so i l s plotted on a shear stress (T) vs normal stress (a) diagram. 89 and, in the case of the hypothetical slopes shown in Figure 23, Chapter 2, i s evaluated at the soil-unweathered t i l l i n ter face. For the unsaturated condition FS = C + a 1 tan 0 Q 1 C  a = 915 Y m . z . s ina . cosa 626 1.46 and for the saturated condition FS = C a + a ' tan 0 = ^ = Q A 7 Y s . z . s ina . cosa 702 Theoret ica l ly , a slope with a safety factor less than 1.0 cannot remain in a stable condition and must f a i l . As poorly drained so i l s subject to periodic saturation ex i s t on slopes of 35 degrees or steeper in the study area, the very low FS calculated for the saturated condition presumably derives p r i n c i pa l l y from an inadequate estimate of the apparent cohesion C due to the tree roots. To ra ise the safety factor to a a 2 value of 1.0 the apparent cohesion must be increased to 540 kg/m which, judging by the estimates of cohesion due to tree roots made by Endo and Tsuruta (1968) and Takahasi (1968), i s a more r e a l i s t i c value than the estimate provided by equation 17. Then, at saturat ion, the tree network accounts for 71 percent of the s o i l ' s shear strength. Although the estimates of the apparent cohesion and the s o i l ' s f r i c t i o n angle are extremely crude, these calculat ions serve to show that the s t a b i l i t y of saturation-susceptible so i l s of drainage depressions and of other poorly-drained, sloping s i tes may be largely dependent on the cohesion derived from the root network. Loss of th i s root strength-ening e f fect through c l e a r f e l l i n g and subsequent decay of the root network w i l l , according to the analyses presented above, predispose such so i l s 90 to imminent f a i l u r e during heavy rainstorm periods when the s o i l mantle attains a condition of par t ia l or complete saturat ion. Laboratory test ing of root strength The deter iorat ion of tree roots accompanied by a decrease in t he i r strength is believed to be the most important means by which c l e a r f e l l i n g reduces the s t a b i l i t y of steep, forested slopes (Croft and Adams 1950, Bishop and Stevens 1964, and Fujiwara 1970). However, very few attempts have been made to measure the strength of roots d i r e c t l y and less i s known about the rate at which th i s strength is l o s t a f te r the death of the parent t ree. Strength measurements of root wood of various hardwood and coni fer tree species reported by Fengel in Brown et_ al_ (1952) indicate that root wood is mechanically weaker than stem wood. A deta i led study of the strength of l i v i n g tree roots was conducted by Turmanina (1965). He tested the roots, sampled from several species of hardwood and con i fers , in tension in order to d ist inguish those species with the strongest root systems and, therefore, with the most desirable qua l i t i e s fo r re inforc ing so i l s against lands l ides. The condition of partly-exposed roots at the headscarps and l a te ra l scarps of many landsl ide scars in the Coast Range study area suggests that a high percentage of the larger, structual roots,as well as the smaller roots at the margins of landsl ides, f a i l in tension. Broken roots often extend some distance out from the head and l a te ra l scarps (Figure 7), ind icat ing that they were subjected to considerable pu l l i ng forces before they f i n a l l y ruptured. Shear fa i l u re s as well as tension fa i l u re s were discovered at most landsl ide s i tes but, general ly, a larger number of roots appear to f a i l in tension. During the shear box 91 experiment i t was also noted that many roots f a i l e d in tension. Therefore, the tens i le strength of tree roots appears to be a very important mechanical property contr ibuting to the s t a b i l i t y of slopes. The purpose of the research reported in th i s section was to determine the tens i l e strength of tree roots sampled from l i v i n g trees and i den t i f y the rate at which the root strength deteriorates a f te r the parent trees are f e l l e d . Methods. Roots were taken from mature, l i v i n g , Douglas f i r , western red cedar and western hemlock trees and from Douglas f i r and cedar stumps on slopes c l ea r f e l l ed at various times in the past (Table 14). Only roots growing in the B horizon of poorly drained so i l s in drainage depressions and on other wet s i tes were co l lec ted. At each s i t e , roots were taken from two or three trees or stumps of the same species. In order to ensure that i n te r - t ree root grafts did not influence the results of the strength tests on roots from stumps, only those stumps located well away from l i v i n g trees at the margins of c learcut areas, were sampled. . Eis (1972) has shown that i n te r - t ree root grafts have permitted Douglas f i r stumps to remain a l i ve and to continue to grow for many years a f te r the parent trees were f e l l e d . The root samples, which ranged from 1 mi l l imeter to 2 centimeters in diameter, were excavated with a hand shovel, traced to the i r respective parent tree or stump, l abe l l ed , sealed in p l a s t i c bags and taken to the laboratory for tes t ing . Uniform, stra ight sections were selected from the root samples, trimmed to a length of 25 centimeters and debarked. Approximately 5 to 10 samples were prepared in th i s manner at one time and then promptly tested in tension to avoid excessive moisture loss from the root t i s sues. Wangaard (1950) pointed out that, 92 above the f i b re saturation point (25 to 30 percent water content by weight for most woods), the strength properties of wood are l i t t l e affected by changes in moisture content. However, i f the wood water content is permitted to f a l l below the f i b re saturation point, then marked increases in strength may occur. Root test ing was performed with a Floor Model TT-CML Instron Uni-versal Testing Machine equipped with an Instron 5,000 kilogram capacity, revers ible loadcel l (No. 2511.304). Type 3D pneumatic-hydraulic clamps (Figures 29 and 30) with f l a t , non-serrated jaw faces 5 centimeters square were employed to gr ip the root ends. The load speed and recorder chart speed were set at 2 cm per minute for a l l tension te s t s . The determination of a sat i s factory technique for gripping the root sample ends without permitting slippage or damage to the wood tissues posed the most serious experimental problem. Preliminary exper-iments showed that tension tests could be conducted s a t i s f a c t o r i l y without damage to the root ends i f short segments of p l a s t i c tubing about 4 cm long were f i t t e d over the root ends and attached with epoxy-resin glue. A clamping force of approximately 300 kilograms at the jaw faces was s u f f i c i en t to prevent slippage up to tensions of 100 kilograms. Further experimentation showed that gluing was not necessary i f a clamping force of 550 to 600 kilograms was applied to the root ends protected by p l a s t i c sheathing. Under these conditions tens i l e forces of 100 kilograms or more could be applied to the roots without slippage and with only minor crushing of the root ends. At the beginning of each test the upper and lower sets of jaws were set at 18 centimeters apart; th i s represented the length of the root section strained in tension. Each test was continued unt i l the Figure 29. Instron Universal Testing Machine showing recorder housing ( far l e f t ) , holding clamps and control panel ( far r i g h t ) . Figure 30. Details of the pneumatic holding clamps (Type 3D). 94 root, which was strained at a constant rate of 2 centimeters per minute, ruptured. If the root fractured at the jaw edges, then the results of that test were not considered. After the completion of each t e s t , the mean diameter of the root was measured with a d ia l gauge near the point of breakage and the equivalent root cross-sectional area was computed. Test results were automatically recorded as a plot of ten s i l e load in kilograms against time in minutes (Figure 31) and from these data the ultimate ten s i l e load in kilograms and the tota l longitudinal deformation ( A l ) in centimeters were determined. Ultimate tens i l e loads were conver-ted into maximum tens i l e strengths in kilograms per square centimeter. Approximately 200 tests were completed, including a l imi ted number oh l i v i n g roots of red alder and fireweed (Epilobium angustifol ium), a vigorous colonizer of c learcut areas in the Coast Range. Results and discussion. Table 14 presents a summary of the test re su l t s . Regressions of tens i l e strength on mean root diameter were calculated for the 12 groups of cedar and Douglas f i r roots. A l l the relat ionships were nons ignif icant, ind icat ing that, within the rather narrow range of root diameters sampled, diameter does not s i g n i f i c a n t l y influence the ten s i l e strength. The ten s i l e strength of cedar and Douglas f i r roots declined rap id ly with increasing time after f e l l i n g of the parent trees. Douglas f i r roots sampled from stumps which had been cut.3 years previously, possessed less than half the mean strength of roots taken from l i v i n g trees. Cedar roots displayed a slower decline in strength up to 5 years a f te r cutt ing of the parent tree at which time approximately half t he i r o r ig ina l mean tens i l e strength had been l o s t . These trends are shown in Figure 32. Only l i v i n g roots of western hemlock were tested. L iv ing roots of red 4-0 CD 30 C e 2-0 CD E 1-0 00 Root from living Douglas fir Tangent line 10 15 20 25 Load, kilograms RP CS area =0-059 c m 2 Load at RP = 35-2 kg Maximum tensile strength =593-0 kg , Deformation at RP = 2-5cm / c m ' E = 1317 k g / c m 2 30 35 CO CD 3 C CD E Root from Douglas fir stump cut 3 years before test CO CD C CD £ CS area = 0-116 c m 2 Load at RP = 22-5 kg Tangent line yf Maximum tensile strength = 194-8 k g / c m 2 Deformation at RP = 0-8 cm E = 6840 k g / c m 2 ± 5 10 15 20 .. 25 Load, kilograms Root from Douglas fir stump cut 10 years before test 30 2-0 1-0 o-o*=»=»——J' ± CS area = 0-245 c m 2 Load at RP = 9-0 kg Maximum tensile strength = 36-7 k g / c m 2 Deformation at RP =0-2 cm 5 10 15 Load,kilograms Figure 31. Results of t en s i l e strength tests on three Douglas f i r roots in d i f fe rent states of deter io rat ion. RP = rupture point; PL = proportional l i m i t . 96 2 Table 14. Diameter- (cm) and t e n s i l e s t r e n g t h (kg/cm ) o f r o o t s sampled  from l i v i n g t r e e s and from stumps o f former t r e e s (1) Western r e d cedar Living I t r e e P e r i o d elapsed s i n c e stump was cut (years) 2 NB 3 B 5 B 6 B 8 B Diam TS Diam TS Diam TS Diam TS Diam TS Diam TS .190 485.4 .175 398.0 .173 47,3.5 .521 351.3 .597 130.4 .368 49.0 .104 337.2 .122 473.8 .249 4d3.0 .368 341.0 .919 35.4 .485 30.9 .246 393.3 .099 451.1 .213 372.0 .406 216.9 .552 173.9 .358 49.6 .157 510.9 .310 507.6 .231 362.4 .490 317.5 .752 71.4 .721 70.5 .145 854.0 .274 522.3 .318 364.2 .267 207.2 .528 113.5 .663 37.3 .155 812.1 .079 496.0 .335 306.8 .112 237.6 .411 26.3 .638 25.7 .157 677.2 .368 142.8 .284 325.6 .140 515.5 .488 41.7 .453 69.5 .152 454.4 .478 351.2 .277 336.6 .203 198.4 .429 75.3 .482 74.5 .175 389.5 .188 562.3 .300 268.8 .445 269.3 .267 97.8 .669 38.7 .185 502.2 .201 480.8 .518 411.4 .419 136.6 .305 18.1 .485 56.2 .135 562.4 .185 442.8 .490 185.2 .337 266.0 .272 . 8 9 . 8 .238 81.8 .104 834.5 .145 435.2 .523 235.5 .368 429.2 .272 64.0 .372 57.7 .086 396.1 .254 217.7 .503 259.1 .411 363.9 .348 94.7 .364 28.7 .132 450.2 .368 356.9 .513 198.1 .277 169.4 .333 87.5 .745 35.9 .079 870.0 .272 551.1 .511 220.9 .298 131.2 .374 76.4 .521 24.5 .261 585.5 .310 450.4 .254 349.2 .165 287.5 .137 189.5 .330 43.5 569.7 427.5 317.0 277.4 86.6 48.4 (2) Douglas f i r L i v i n g t r e e P e r i o d elapsed s i n c e stump was cut (years) 2 NB 3 B 5 B 8 B 10 B Diam TS Diam TS Diam TS Diam TS Diam TS Diam TS .163 687.7 .292 402.3 .140 180.8 .495 124.9 .510 20.3 1.242 43.2 .206 432.1 .191 602.3 .384 194.8 .246 207.1 .393 52.1 .724 29.1 .203 511.4 .564 416.5 .279 228.4 .376 118.9 .402 7.2 .533 20.6 .533 519.5 .572 412.8 .157 253.2 .470 123.9 .510 13.9 .965 6.4 .274 59 3.0 .325 550.2 .277 181.7 .500 76.7 .584 33.3 .559 36.7 .208 453.2 .312 412.9 .351 124.0 .538 57.9 .367 44.1 .605 19.2 .147 590.2 .231 441.1 .203 238.7 .620 115.2 .421 37.8 .457 21.7 .231 443.5 .259 398.7 .165 234.8 .478 139.4 .693 20.8 .737 30.7 .384 457.2 .501 439.4 .288 204.7 .455 130.4 .450 26.6 .582 25.2 .345 552.6 .372 468.8 .129 191.2 .531 150.0 .718 46.0 .990 26.4 .140 728.2 .249 293.4 .180 220.8 .511 134.5 .936 50.8 .731 15.7 .127 966.6 .357 482.7 .341 103.8 .762 126.5 .580 30.8 1.140 30.6 .152 961.1 .441 518.9 .258 267.8 .447 169.0 .386 46.5 .647 35.2 .229 652.5 .208 442.2 .435 288.9 .447 156.3 .647 57.9 .850 16.8 .114 889.3 .419 592.1 .608 119.4 .221 183.8 .671 40.8 .858 17.9 629.2 458.3 202.2 134.3 35.3 25.2 Period separating time of cutting of parent tree and time of root strength test (Years ) Figure 32. Diagram showing the change in tens i l e strength of small Douglas f i r and cedar roots with change in the time elapsed since f e l l i n g of the parent t ree. 98 alder and fireweed were considerably weaker than l i v i n g conifer roots. The root strength data fo r Douglas f i r and western red cedar were investigated with analyses of variance and Duncan's mult iple range tests in order to determine whether or not the differences in mean root strength at various times a f ter c l e a r f e l l i n g were s i gn i f i c an t . The test results are summarized in Table 15. Table 15 Results of Analyses of Variance and Mult ip le Range Tests  for Douglas f i r and Western Red Cedar Root Strength Data (1) Douglas f i r One way analysis of variance yielded an = 119.48** A Duncan's mult iple range test on the root strength means of the s i x root classes produced the fol lowing results at the 0.05 level L iv ing Period elapsed since f e l l i n g of parent tree (yrs) tree 2 3 5 8 10 (2) Western red cedar One way analysis of variance y ielded an F^5 g Q j = 56.41** A Duncan's mult ip le range test on the root strength means of the s i x root classes produced the fol lowing results at the 0.05 level L iv ing Period elapsed since f e l l i n g of parent tree (yrs) tree 2 3 5 6 8 99 Both ANOVA tests were highly s i g n i f i c an t . The Mult ip le Range tests indicated that, within the same species, the mean tens i l e strength decreased s i g n i f i c an t l y from one class of roots to the next class (pro-gressing from l e f t to r ight) except in those cases where adjacent classes are underlined. From the traces of tens i le load against time i t was possible to invest igate the e l a s t i c behaviour of roots in tension. The load-time curves for roots sampled from l i v i n g trees (Figure 31) e s sent ia l l y consist of an i n i t i a l s tra ight portion within the e l a s t i c range and a curved portion which l i e s in the p l a s t i c range. An i n i t i a l curvature i s sometimes apparent but was disregarded. That part of a curve separating the e l a s t i c range from the p la s t i c range is ca l led the proportional l i m i t and is delineated where a l i n e , tangent to the curve in the e l a s t i c range, deviates from the curve (Brown et_ al_ 1952). The modulus of e l a s t i c i t y represented by r _ T . 1 E " S. _ — - (18) A A l 2 E = modulus of e l a s t i c i t y , kg/cm Tp = tens i l e load at proportional l i m i t , kg 2 A = cross sectional area of root specimen, cm 1 = gage length of specimen = 18 cm A l = deformation at proportional l i m i t , cm was determined for cedar and Douglas f i r roots taken from l i v i n g trees and from stumps cut 2 years and 3 years pr ior to te s t i ng . Roots sampled from older stumps generally produced load-time curves which did not possess defineable proportional l i m i t s . The tota l deformation at the rupture point Table 16 Total deformation A l (cm) at the Rupture Point and Modulus of p E l a s t i c i t y (E) kg/cm for Cedar and Douglas F i r Roots Tested in Tension Liv ing tree A l (1) Cedar Time elapsed since stump was cut (years) 2 3 5 6 8 A l A l A l A l A l 2.0 2.0 1.5 1.3 1.3 2.6 1.3 2.2 2.8 2.8 3.4 3.7 1.7 2.8 2.2 2.9 33,300 5,010 5,105 12,630 17,640 7,756 9,376 6,428 7,515 3,228 8,550 8,460 15,030 6,000 2,000 5,618 2.4 2.1 1.7 0.9 2.4 2.2 0.9 1.1 2.4 2.3 1.0 1.5 2.0 1.5 2.0 2.4 5,348 7,515 11,655 15,840 12,600 18,000 4,260 9,270 10,680 12,690 13,320 42,300 1,959 846 1,089 798 3.2 2.2 3.0 1.1 0.9 2.9 3.6 3.0 3.1 1.6 2.0 1.1 2.8 1.4 1.6 1.1 6,264 6,680 3,420 8,550 7,515 6,840 4,110 6,345 4,980 11,610 7,020 3,394 7,020 31,750 5,220 7,853 1.0 1.3 0.6 1.5 0.5 1.2 1.1 0.4 0.7 0.2 0.9 1.2 0.4 0.3 0.8 1.2 0.5 0.6 1.0 0.4 0.6 0.1 0.2 0.5 0.5 0.2 0.3 0.2 0.4 1.4 0.3 0.4 0.5 0.3 0.9 0.4 0.1 0.2 0.1 1.0 0.4 0.2 0.6 0.8 0.1 0.2 0.5 0.2 2.3 9,605 1.8 10,511 2.2 8,036 0.8 0.5 0.4 (2) Douglas f i r L iv ing tree Time elapsed since stump was cut (years) 2 3 5 8 10 A l E A l E A l E A l A l A l 3.0 2,004 1.0 12,060 - 1.1 4,500 0.9 0.5 0.5 1.5 16,596 2.4 3,870 0.8 7,380 0.9 0.8 0.8 2.7 25,020 0.6 7,920 0.9 11,790 0.8 0.4 0.2 2.5 1,317 1.1 16,110 0.8 6,624 0.8 0.2 0.1 3.1 9,818 2.5 10,305 0.6 5,265 0.6 0.2 0.2 1.7 9,652 1.7 5,683 0.8 3,690 0.7 0.5 0.2 2.7 4,635 1.6 11,400 0.6 9,360 0.8 0.9 0.1 2.0 3,750 1.8 11,880 0.8 6,840 0.9 1.2 0.4 1.3 4,860 1.7 6,250 1.7 8,860 0.8 0.4 0.8 2.5 11,430 1.3 8,742 0.9 10,420 1.2 1.3 0.2 1.5 7,950 2.2 6,810 0.8 5,204 0.7 1.2 0.9 2.5 7,290 2.0 6,645 1.9 3,418 0.7 0.6 0.1 1.4 6,840 1.8 10,130 0.5 8,610 '1.1 0.8 0.3 1.7 8,460 0.6 4,200 0,8 9,730 1.1 0.4 0.7 1.8 11,475 1.5 5.810 0,9 7,110 0 f 9 0.6 I.T 101 was also estimated for a l l te s t s . The results appear in Table 16. The mean deformation at the rupture point decreased from 2.3 and 2.1 centimeters for roots taken from l i v i n g cedar and Douglas f i r , respect ively, to 0.8 and 0.9 centimeters for roots sampled from cedar and Douglas f i r stumps cut 5 years previously. The modulus of e la s -t i c i t y , however, did not show a s im i l a r , well-marked trend as the roots deter iorated, suggesting that the s t i f fnes s or r i g i d i t y of the wood does not a l t e r during the early stages of decay. According to Kennedy (1958) e l a s t i c i t y is one of the mechanical properties of wood least affected by decay, at least in i t s early stages. The a b i l i t y of l i v i n g roots to elongate without rupturing in response to tens i le stresses may permit the so i l mantle on steep slopes to undergo small scale d i f f e r e n t i a l movements or creep without serious loss of strength. The rapid strength deter iorat ion of the smaller tree roots, such as tested in th i s study, i s presumably the consequence of fungal decay, although deta i led microscopic examinations of root wood were not made. According to Wallis and Ginns (1968), the stumps of most coniferous species in Coastal B r i t i s h Columbia are susceptible to Fomes annosus spore i n fec t i on . Following i n fec t i on , the stumps and roots of Douglas f i r and of western hemlock,in pa r t i cu l a r , are rapidly invaded by the fungus and often suf fer advanced root decay within 2 to 3 years. However, the rapid root deter iorat ion rates measured in the experiment reported above are un l ike ly to be representative of the whole tree root system because the larger structual roots of Douglas f i r and cedar may remain reasonably sound and intact for many years a f te r the parent trees are cut down. For instance, old cedar stumps in the lower Capilano val ley 102 on areas c l ea r f e l l ed in the 1930's, have retained t he i r roots of diameters greater than approximately 15 cm. Although these roots are in an advanced state of decay, some reta in central cores of r e l a t i v e l y sound wood. The smaller members of these old root systems have decayed and disappeared, probably many years ago. McMinn (1963) also noted that many of the massive roots of a mature Douglas f i r tree f e l l e d more than 25 years previously were only p a r t i a l l y decayed. The du rab i l i t y of large roots raises the question as to whether or not the smaller roots less than, say, 3 centimeters in diameter, are very important to slope s t a b i l i t y . The evidence provided by measurements of root sizes at the lands l ide s i tes suggests that they are. At the head-scarps of three landsl ides (S4, CIO and C12 in Table 4) the diameters of 150 exposed, broken roots growing in s o i l or t i l l below a depth of approximately 45 centimeters were measured with a pair of ca l i pe r s . The mean diameters were 1.4 cm, 0.7 cm and 1.1 cm (Table 17), emphasising that small roots const itute a major part of the root network in the lower B horizon. This confirms the findings of McMinn (1963) who discovered that large Douglas f i r roots tend to p ro l i f e ra te when they extend into stoney or bouldery subsoils and that large roots are mainly confined to the upper s o i l layers. In very shallow s o i l s , however, large roots extend to the base of the B horizon and may be one of the main reasons why so i l s less than 40 centimeters deep are very res i s tant to large scale mass wasting. The changes which occur in root strength and, consequently, in the s o i l ' s resistance to f a i l u r e a f te r c l e a r f e l l i n g have several management impl icat ions. I f the decaying root network in the s o i l of a steep, c l e a r f e l l ed slope i s quickly reinforced with a vigorously expanding, replacement network from natural forest regrowth or a r t i f i c i a l l y planted 103 Table 17 Means, extremes and standard deviations of diameters of roots  growing in subsoils below 45 centimeters at landsl ide headscarps Landslide No.of roots Mean Max Min Standard No. measured diameter diameter diameter deviation cm cm cm cm S4 50 1.4 8.8 0.1 1.9 CIO 50 0.7 3.2 0.1 0.8 C12 50 1.1 5.2 0.1 1.1 con i fers , then i t i s conceivable that the reduction in slope s t a b i l i t y brought about by c l e a r f e l l i n g w i l l be minimal. The root study reported above indicates that a substantial loss of root strength occurs within 3 to 5 years a f te r c l e a r f e l l i n g . Such a period provides very l i t t l e time for the regrowth vegetation to develop an e f fec t i ve root network. On the g rave l ly , sandy-loam, steepland so i l s of the Coast Range, root devel -opment of Douglas f i r and hemlock seedlings i s slow. On c l ea r f e l l ed areas in the Seymour catchment, Douglas f i r sapl ings, planted approximately 5 years previously, possess root systems which are predominantly confined to the upper 40 centimeters of s o i l . Pruning of taproots in the nursery may account for these r e l a t i v e l y shallow root systems. A l l Douglas f i r planting stock used in the study area receives two and usually three separate root prunings while being raised in the nursery (Mr. Armit, B.C. Forest Service nursery, Surrey; personal communication). McMinn (1963) points out that an aggressive taproot i s l i k e l y to be a desirable survival cha rac te r i s t i c , pa r t i cu l a r l y on low f e r t i l i t y s i tes or on dry 104 s i t e s . An aggressive taproot which imparts strength to the subsoil would also be a desirable charac te r i s t i c on steep, unstable s i t e s . Therefore, the use of unwrenched, containerised, Douglas f i r seedlings rather than open-grown, root-pruned seedlings for revegetating steep, potent ia l l y unstable slopes would provide some advantages. I f the decay rate of the root systems of c l e a r f e l l ed trees could be retarded unt i l the young replacement root network was well established throughout the s o i l mantle, then the impact of the decay ef fect on the s t a b i l i t y of sloping so i l s might be reduced considerably. The appl icat ion of borax to stumps (Wallis and Ginns 1968) immediately a f te r f e l l i n g prevents fungal in fect ion and i s , perhaps, the most obvious and most pract ica l means of slowing the process of root deter io rat ion. For pract ica l purposes th i s treatment could be re s t r i c ted to steep drainage depression s i tes and to areas over 35° in slope. The exceptional a b i l i t y of red alder to colonize and grow rap id ly on wet, cutover slopes and exposed surfaces of t i l l and subsoi l s , at the same time producing a widely spreading root system of slender l a t e r a l s , makes th i s species potent ia l l y valuable for s t a b i l i z i n g steepland s o i l s . The establishment of alder plants from nursery raised seedlings or from cuttings on potent ia l l y unstable, c l e a r f e l l ed slopes, a pract ice which is common in parts of Europe and in New Zealand, could prove to be an e f fec t i ve counter-landsl ide measure in the Coast Range of B r i t i s h Columbia. Summary and conclusions (1) General observations of the growth habits of tree roots in the Coast Range steepland so i l s suggest that roots help bind the noncohesive 105 so i l s into a coherent mantle and anchor the s o i l mantle to the sub-stratum. (2) Direct shear tests conducted in the f i e l d provide addit ional evidence that the tree roots do, indeed, provide a cohesive e f fect which may, in so i l s containing high root dens i t ies , increase the s o i l shear strength by several hundred kilograms per square meter. Under saturated so i l condit ions, the s o i l ' s shear strength may be largely derived from the apparent cohesion provided by the tree root network. (3) Laboratory strength tests of tree roots from l i v i n g trees and from stumps in various stages of decay show that a f te r c l e a r f e l l i n g , the smaller roots less than 3 centimeters in diameter, which const itute a large proportion of the root network in the lower subsoi l s , rap id ly lose t he i r strength. Within 3 to 5 years a f te r c l e a r f e l l i n g the small roots of Douglas f i r and cedar stumps may lose more than hal f t he i r o r ig ina l ten s i l e strength. (4) Several simple management techniques could be employed to help reduce the impact of c l e a r f e l l i n g on the s t a b i l i t y of steep slopes. These include the planting of potent ia l l y deep-rooting, unwrenched, contain-er ized Douglas f i r seedlings on steep, c l e a r f e l l ed areas, treatment of f resh ly-cut stumps with borax, and the a r t i f i c i a l establishment of red alder seedlings or cuttings on potent ia l l y unstable, wet areas. 106 CHAPTER 4 A DETAILED EXAMINATION OF SELECTED LANDSLIDES  Introduction The evidence presented in the e a r l i e r chapters indicates that the so i l water conditons and the density and condition of the tree root network are two very important factors inf luencing the s t a b i l i t y of forest s o i l s . However, the large var iat ion in the s i ze , shape, topog-raphy and subsurface s i t e conditions among d i f fe rent lands l ides, suggests that each f a i l u r e results from a unique combination of movement-promoting and movement-resisting forces. Moreover, several large debris avalanches and debris s l ides have formed on well drained forested s i tes where excessive pore water pressures and lack of a strong, well developed root network are un l ike ly to have been major causes of f a i l u r e . On the other hand, there ex i s t steep, poorly-drained seepage depressions on c l e a r f e l l ed slopes within the study area which appear to be l i k e l y landsl ide areas but, nevertheless, show no evidence of recent mass wasting. Such anomalies indicate that general rules explaining the cause of landsl ides may lead to many misinterpretations in landsl ide studies. Probably most landsl ides resu l t from a combination of many causes with one factor being f i n a l l y dominant as suggested by Baker (1953). This i s why attempts to assess the re la t i ve s t a b i l i t y of slopes and to predict the locat ion of future landsl ides are often unsuccessful. It follows that there are many advantages in undertaking a deta i led invest igat ion of a sample of ca re fu l l y selected indiv idual landslides which are, at least pa r t l y , amenable to quant itat ive study. Although i t i s normally impossible to accurately define the stress f i e l d and the subsurface conditons at f a i l u r e , careful examination of the f a i l e d s i tes 107 combined with an analysis of the estimated forces inf luencing the slope at time of f a i l u r e often permit an understanding of the pr inc ipa l causal factors. This chapter describes the nature and results of the quantitat ive analyses of f i ve selected lands l ides. S t a b i l i t y analysis Simple, approximate, s t a b i l i t y analyses based on the theory of p l a s t i c equi l ibr ium were used to invest igate the selected lands l ides. Most, i f not a l l , of the larger landsl ides in the study area are not amenable to quant itat ive analyses because t h e i r topography has been modified by erosion since f a i l u re and because the shearing strength char-ac te r i s t i c s of the materials involved cannot be determined with any degree of accuracy. Even in the cases of the selected lands l ides, where so i l s were r e l a t i v e l y homogeneous and reasonably accurate estimations of the dimensions, volumes, and weights of the displaced so i l s could be made, the strength of the so i l s i s probably so variable and the number of d i f fe rent forces acting i s so large that the theoret ical studies provide l i t t l e more than a broad indicat ion of what i s l i k e l y to have occurred. The ordinary method of s l i ce s and the Bishop method of s l i c e s , assuming c i r c u l a r surfaces of f a i l u r e , were applied to four fa i lu res associated with roads. In add i t ion, an i n f i n i t e slope analysis was used to invest igate a f i f t h landsl ide on a steep, c l e a r f e l l ed slope. Appendix 3 b r i e f l y describes each analysis method. Detailed descr ipt ions, app l i c -ations and l im i ta t ions of the method of s l i ce s are presented by Bishop (1955), Lambeand Whitman (1969), Whitman and Bailey (1967), Wu (1966) and Zaruba and Mencl (1969). Lambeand Whitman (1969) and Taylor (1948) out l ine the theory and appl ications of i n f i n i t e slope analyses. 108 Before a r e l i ab l e s t a b i l i t y analysis can be conducted several fundamental pre-requis ites must be met: (1) The geometry of the landsl ide including the scar topography must be accurately described. Only r e l a t i v e l y fresh landsl ides which showed no evidence of pos t - fa i lu re erosion were selected for analys is . In add i t ion, the selected landsl ides possessed smal l , well defined scars which permitted reasonably accurate ' l eve l and s t a f f surveys to be made of t he i r dimensions and shape. (2) The s o i l propert ies, pa r t i cu l a r l y the shearing strength and the unit weight, must be known. Unit weights were estimated from undisturbed bulk samples and, in the case of road f i l l mater ia ls, by a sand replacement method. The strength parameters were based on f i e l d shear tes t s . To compensate for the inaccuracies in the strength test re su l t s , several analyses were made of each f a i l u r e using d i f ferent shear strength parameter values which embraced the range of measured values. Although there was no way of gauging which values, i f any, were appropriate, repet i t i ve analyses provided information on the s e n s i t i v i t y of the slope s t a b i l i t y to changes in the s o i l shear strength. (3) The external loads, i f any, should be cor rect ly defined. The weight of the vegetation was not considered to be important in any of the fa i lu res invest igated. However, the weight of the snow on the slopes may have contributed to f a i l u r e , espec ia l ly in the case of the debris avalanches in road f i l l mater ials. A heavy snowpack on the f l a t crown portions of the landsl ides would greatly add to the downslope shearing forces, but, because these loads were impossible to def ine, they were not considered in the analyses. (4) The groundwater or seepage conditions at the time of f a i l u r e must 109 Figure 33. Formation of tension cracks in f i l l materials near the outer edge of an abandoned logging road. Figure 34. Incipient lands l id ing along an abandoned logging road in the Howe Sound area. 110 be cor rect ly interpreted. The most c r i t i c a l condit ion, that of complete saturation of the s o i l mantle with the top, flow l i ne coincident with the s o i l surface, was assumed for 3 landsl ides on poorly drained slopes. Piezometer measurements on poorly drained forest s o i l s i tes showed that the assumption of seepage pa ra l l e l to the slope may cause serious under-estimates of the true pore water pressures (Chapter 2) but on slopes underlain by r e l a t i v e l y uniform road f i l l materials th i s assumption i s probably v a l i d . P a r t i a l l y saturated conditions were assumed for 2 land-s l ides on well drained slopes. (5) The s o i l materials at f a i l u r e should behave in a r i g i d - p l a s t i c manner. In other words, the i n i t i a l l y displaced so i l mass should move as one or more d i sc rete, so l i d units. Within the study area the formation of tension cracks (Figure 33) and slump steps (Figure 34) along road r i gh t -of-ways and occasionally on steep, unstable slopes indicate inc ip ient lands l id ing of d i sc rete , coherent blocks of s o i l or road f i l l mater ials. The greatest uncertainties in the s t a b i l i t y analyses ar ise from the se lect ion of the strength parameters and pore water pressures. Landslide 1 (road-cut slope in upper Magnesia Creek catchment). The collapse or slumping of road-cut slopes represents one of the commonest types of mass wasting features in the study area, although 2 the ind iv idual f a i l u re s are rarely more than 100 m in area. Landslide 3 1 was a small debris s l i de involving the f a i l u r e of approximately 6 m of forest s o i l and colluvium from a steep, cut slope of 63 degrees onto an abandoned logging road in upper Magnesia Creek. The slope had been cut about the same time that the general area had been c l e a r f e l l e d , approximately 5 to 7 years p r io r to the time of f a i l u r e . At the time of the landsl ide survey (July 1970) the scar surface l l i Scale 0 1 2 meters Figure 35. Small, c i r c u l a r debris s l i de in a road-cut slope (Landslide 1 ) . Broken l i ne shows the most c r i t i c a l ant ic ipated ground water table condit ions. The heavy l i ne indicates the surveyed surface. Area (a) represents the accumulation of f a i l e d debris on the road surface. 112 was fresh and had, presumably, been exposed no more than 2 to 3 months. The scar formed a toe f a i l u r e with a concave longitudinal p r o f i l e approximately c i r c u l a r in shape (Figure 35). The s o i l consisted of a dark-brown, gravelly-sandy-loam 130 cm deep which gradually passed into a 50 cm thick layer of grey-brown, gravel ly colluvium containing boulders up to 20 cm in diameter. Dead tree roots f ree ly penetrated the s o i l down to the c o l l u v i a l gravels. Ruptured roots extended out from the upper part of the f a i l u r e surface and the absence of gleying or mottling in the subsoils suggested that saturation and excessive seepage were not a problem. Two r e l a t i v e l y undisturbed s o i l samples taken from a rdepth of approximately 100 cm 3 produced a mean moist unit weight of 1,450 kg/m and a mean saturated 3 unit weight of 1,680 kg/m . Shear strength tests were not made of the s o i l s in the v i c i n i t y of Landslide 1 or Landslide 2. Rather, a number of s t a b i l i t y analyses were conducted using d i f fe rent 0 values ranging in the v i c i n i t y of the 0 values, for s im i l a r forest s o i l s determined by d i rect shear tests in the middle Seymour catchment (Figure 28). On a s im i l a r dry s i t e in the middle Seymour catchment, piezometers S3-1, S3-2 and S3-3 (Figure 15, Chapter 2) showed that, during wet periods, the s o i l saturated in a th in layer immediately above the impermeable t i l l or bedrock substratum. The most c r i t i c a l seepage condition assumed in the analyses of Landslides 1 and 2 was a 30 cm saturated zone immediately above the bedrock with a drawdown ef fect near the landsl ide toe (Figure 35 and Figure 36). Pore water pressures were estimated with the re lat ionship 2 = n,.Yw C 0 S <* 1 W'W where u. = pore water pressure at the f a i l u r e surface at the mid point of s i ice i . h,, = ve r t i ca l distance between the f a i l u r e plane and the w r water table at the mid point of s l i c e i . S t a b i l i t y analyses were made with the .ordinary method of s l i ce s and with the Bishop method for the p a r t i a l l y saturated condition and for an unsaturated condition with no seepage. Results are shown in Table 18. Table 18 Factors of Safety (FS) Calculated by the Ordinary Method  of S l ices (QMS) and by the Bishop Method (BMS) for a  road-cut slope at the s i t e of Landslide 1 Seepage condition Method Moist, no seepage OMS P a r t i a l l y saturated OMS P a r t i a l l y saturated BMS Factor of Safety 0 35° 36° 37° 38° 39° 40° 0.47 0.49 0.51 0.52 0.54 0.56 0.37 0.39 0.41 0.42 0.44 0.45 0.38 0.41 0.4.3 0.45 0.46 0.47 The calculated factors of safety f a l l well below unity, ind icat ing that the so i l possessed addit ional strength derived from the tree root network and, possibly, from some other factor not recognised in the o 9 n analys is. An apparent cohesion of 250 kg/m and 200 kg/m for 0 = 35 and 40°, respect ive ly, i s s u f f i c i en t to raise the factor of safety to a value of 1.0 for the moist condition with no seepage. Under p a r t i a l l y 2 saturated conditions with seepage, apparent cohesions of 300 kg/m and 260 kg/m2 (0 = 35° and 40° respectively) are required to raise the 114 factor of safety to 1.0. As the only roots in the v i c i n i t y of the shear plane were dead tree roots, i t i s probable that root deter iorat ion a f te r c l e a r f e l l i n g was an important factor promoting the debris s l i de . However, the most obvious cause of f a i l u r e was the steepness of the road cut. For cuts in cohesionless sands Smith and Cedergren (1963) recommend that the slope should be no steeper than 1%:1 or approximately 33°. For reasons of economy, logging road .cuts in the study area usually slope at angles greater than 45° and, consequently, are subject to numerous f a i l u r e s . Landslide 2 (road-cut slope in upper Magnesia Creek catchment) Landslide 2 was a debris s l i de with a shallow, oval scar on a road-cut slope of 49° about 20 meters north of Landslide 1. The s o i l and drainage conditions were very s im i l a r to those at the Landslide 1 s i t e although the co l l u v i a l gravel layer at the base of the so i l B horizon contained more large boulders than the.same layer at the former s i t e . The f a i l u r e surface had cut through the rooting zone and dead tree roots extended from the upper and middle parts of the scar surface. Figure 36 presents the longitudinal section of the scar. The displaced debr is , 3 consist ing of 7.5 m of part ly intact blocks of s o i l , dis integrated s o i l and colluvium and two large tree stumps, had accumulated on the road surface and plugged a road side drainage d i t ch . S t a b i l i t y analyses were made under assumptions of a moist condition with no downslope seepage and a saturated so i l layer 30 cm thick immediately above the impermeable bedrock. The results appear in Table 19. As was the case with the f i r s t landsl ide examined, the low factors of safety indicate that the s t a b i l i t y of the cut slope was largely dependent on the strengthening e f fect of the tree root systems. For the 115 Figure 36. Debris s l i de in a road-cut slope (Landslide 2), upper Magnesia Creek catchment. The most c r i t i c a l ant ic ipated ground water table condition i s .depicted by the broken l i n e . 116 Table 19 Factors of Safety (FS) Calculated by the Ordinary  Method of S l ices (OMS) and by the Bishop Method (BMS) for  a Road-cut slope at the s i t e of Landslide 2 Seepage condition Method Factor of Safety 0 35° 36° 37° 38° 39° 40° Moist, no seepage OMS 0.73 0.76 0.79 0.81 0.84 0.87 P a r t i a l l y saturated OMS 0.59 0.61 0.64 0.66 0.68 0.71 P a r t i a l l y saturated BMS 0.60 0.62 0.64 0.67 0.69 0.72 part ly saturated condit ion, apparent cohesions of 240 kg/m (0 = 35 ) and 160 kg/m (0 = 40 ) are required to raise the factors of safety to 1.0. Between July 1970 and July 1972, f i ve addit ional small debris s l ides formed on cut slopes near Landslides 1 and 2. P r io r to 1970 only two events had occurred on these slopes over a 5 to 7 year period since the cut slopes were formed. This increased slumping a c t i v i t y a f te r 1970 suggests that root strength deter iorat ion reached a c r i t i c a l stage approximately 5 to 7 years a f te r c learcut t ing . However, the 1970-1971 and 1971-1972 winters produced above average snowpacks within the study area and possibly caused very wet and unstable s o i l conditions during the snowmelt period. Landslide 3 ( r o a d - f i l l slope in upper Harvey Creek catchment) Some of the most damaging landsl ides in the study area or ig inate when r o a d - f i l l materials collapse to form debris avalanches. The largest and most damaging examples are located along abandoned logging roads which have f a l l e n into d i s repa i r (Figure 13). Landslide 3 was a debris avalanche which originated when a portion of a 34° f i l l slope collapsed at the outer edge of an abandoned logging road in upper Harvey Creek catchment. The upper part of the f a i l u r e scar formed a spoon-shaped hollow in poorly compacted, gravelly-sand, f i l l materials but downslope the scar narrowed to form a shallow trench, ind icat ing that the debris avalanche had become a debris flow below the i n i t i a l f a i l u r e zone. The condition of uprooted vegetation near the foot of the scar indicated that the debris avalanche occurred no more than 3 to 4 months pr ior to i t s examination. Near the headscarp, approximately 180 cm of gravelly-sand f i l l rested on d i o r i t e bedrock but 118 0 1 2 3 j meters j Figure 37. Small, c i r c u l a r f a i l u r e in a 34 r o a d - f i l l slope (Landslide 3), upper Harvey Creek catchment. The f i l l consisted of gravels (68%), sands (28%), s i l t s (4%) and clay (-) and was loosely compacted (e = 0.48). downslope the f i l l overburden thinned to approximately 100 cm. Boulders up to 1 m in diameter were embedded in the f i l l . There were no tree roots in these materials. The curved, f a i l u r e surface was roughly c i r c u l a r in longitudinal section with an estimated radius of 17.7 m (Figure 37). The unit weight of the f i l l materials was determined by the sand replacement method (Capper and Cassie 1963). A hole, approximately 20 cm x 20 cm x 10 cm, was dug in the f i l l and the excavated materials were co l lected in p l a s t i c bags fo r water content and pa r t i c l e s ize d i s t r i bu t i on determinations. The volume of the hole was measured by f i l l i n g i t with dry, uniform sand of known dry, unit weight. Measure-ments made at two locations near the scar head produced a mean dry unit 3 3 weight of 1 ,730 kg/m and a mean saturated unit, weight of 2,130 kg/m . Direct shear tests by the same shear box equipment and methods described in Chapter 3, were made on excavated columns of f i l l materials near the head of Landslide 3. Two sets of tests produced f r i c t i o n angles of 38° and 41° respectively (Figure 38). However, these results may be very misleading because the cohesionless gravels and sands tended to crumble within the shear box af ter small test displacements rather than f a i l along a d i s t i n c t i v e shear plane. The s i t e was very poorly drained. Surface water ponded on the uneven road surface at the landsl ide head during wet periods and i t was assumed in the s t a b i l i t y analyses that the f i l l materials attained complete saturation and that seepage occurred pa ra l l e l to the slope. The ordinary method of s l i ce s was used to calculate the factors of safety for the slope under dry and saturated conditons (Table 20). Figure 38. Results of d i rect shear tests on r o a d - f i l l materials plotted on a shear stress (T) vs normal stress (a) diagram. Table 20 Factors of Safety Calculated by the Ordinary Method of S l ices for a Road- f i l l slope at the s i t e of Landslide 3 Seepage condition Factor of Safety 39° 40° 41° 42° 0 38 o Dry 1.46 1.52 1.57 1.63 1.69 Saturated 0.84 0.88 0.91 0.94 0.97 The s t a b i l i t y of the dry slope appears to be adequate. However, at saturat ion, the s t a b i l i t y of the slope i s reduced to a c r i t i c a l state as indicated by the factors of safety which are less than 1.0. Landslide 4 ( r o a d - f i l l slope in upper Magnesia Creek catchment) Landslide 4 originated when part of a loose, cohesionless f i l l slope collapsed at the outer edge of an abandoned logging road in upper Magnesia Creek. The resu l t ing scar formed a shallow,curved hollow with an approx-imately c i r c u l a r longitudinal p r o f i l e (Figure 39). The gravelly-sand 3 f i l l had a dry unit weight of 1,670 kg/m and a saturated unit weight of 3 2,030 kg/m . At the road edge the f i l l was about 150 cm thick and rested on weathered greywacke and low-grade schist bedrock. A small seepage discharged from the cut slope face immediately above the landsl ide s i t e and, as there was no roadside dra in , the upslope drainage passed through or across the road bed gravels onto or through 122 Figure 39. Shallow, c i r c u l a r f a i l u r e in a r o a d - f i l l slope (Landslide 4) , upper Magnesia Creek catchment. The f i l l was loosely compacted (e = 0.62) and was composed of gravels (53%), sands (38%), s i l t s (8%) and clays (1%). the f i l l slope below the road. The road surface was deeply rutted by smal l , surface runoff channels. The ordinary method of s l i ce s was used to analyse th i s landsl ide and i t was assumed that the slope would be subject to complete satur-ation during heavy storms. The shear strength of the f i l l materials were not measured at th is s i t e (the landsl ide was not accessible by vehicle) but i t was assumed that the e f fect i ve 0 value or values for these materials lay within the range of 0 used in analyses of Landslide 3. Table 21 Factors of Safety Calculated by the Ordinary Method  of S l ices for a Road- f i l l slope at the s i t e of Landslide 4 Seepage condition Factor of Safety 0 38° 39° 40° 41° 42° Dry 1.02 1.06 1.10 1.14 1.18 Saturated 0.54 0.56 0.58 0.60 0.62 The results of these analyses show that the f i l l s lope, even in a dry condit ion, did not possess a wide margin of safety against f a i l i n g . Therefore, i t seems un l ike ly that the slope attained a state of complete saturation before s l i d i n g . In f ac t , the calculated factors of safety suggest that a very small r i se in the pore water pressures, resu l t ing from the saturation of the basal regions of the f i l l , would have been su f f i c i en t to jeopardize the s t a b i l i t y of the slope. 1 2 4 Although the poor drainage condition at the head of th i s landsl ide was a pr inc ipa l cause of f a i l u r e , the steepness of the f i l l slope (41°) was also instrumental in bringing about movement. In the Howe Sound area lands l id ing of the r o a d - f i l l s i s mainly confined to slopes over 33 to 36 degrees. Downslope from the i n i t i a l f a i l u r e zone many debris avalanches, including those or ig inat ing in r o a d - f i l l s , appear to revert to debris flows which continue to the val ley bottoms. Gonsior and Gardner (1971) point out that l i q u i fact ion fa i lu res commonly occur in poorly compacted, saturated f i l l slopes in northern Idaho. Poss ibly, shear f a i l u re s in the loose, saturated f i l l slopes within the study area, are accompanied by a volume decrease which tr iggers 1 iqu i fac t ion . I f the f i l l s had received some compaction during road construct ion, then the frequency of lands l id ing might have been reduced. Landslide 5 (natural slope in upper Harvey Creek catchment) Landslide 5 was a small debris s l i de (HS 14 in Table 4, Chapter 1) which occurred in a shallow drainage hollow on a 30° c l ea r f e l l ed slope in upper Harvey Creek catchment. Approximately 570 m of s o i l had s l i d downslope into a small stream gul ly to expose a f l a t , even, unweathered t i l l surface (Figure 40). A s l i g h t l y overhanging head wall and ve r t i ca l l a te ra l scarps outl ined the f a i l u r e scar which broadened towards i t s toe to form a narrow, fan-shaped feature. Water seeped from the base of the headwal1. The so i l s in the v i c i n i t y of the landsl ide ranged between 50 and 80 cm in depth, were poorly drained and possessed stoney, mottled B horizons. Cedar and hemlock roots f ree ly penetrated the subsoils down to the s o i l - t i l l interface and small roots extended into the t i l l . The Scale 0 ~~5™"~ 10 meters Figure 40. Longitudinal p r o f i l e of a shallow debris- s l i de on a 30°, c l e a r f e l l ed slope in upper Harvey Creek catchment (Landslide 5). The r e c t i l i n e a r f a i l u r e surface coincided with the s o i l - t i l l in ter face. so i l B horizon contained scattered boulders of weathered d i o r i t e up to approximately 30 cm in diameter. Two undisturbed samples of the 3 B horizon gave a mean moist unit weight of 1 ,550 kg/m and a saturated 3 unit weight of 1,790 kg/m . Soi l shear tests were not conducted at th is s i t e . Factors of safety and the apparent cohesion required for s t a b i l i t y were calculated for d i f fe rent s o i l strength and so i l water conditions by an i n f i n i t e slope analysis (Table 22). Table 22 Factors of Safety and the Apparent Cohesion Required  for S t a b i l i t y for a Natural Slope at the s i t e of Landslide 5 Seepage condition Factor of Safety 0 35° 36° 37° 38° 39° 40° 1.21 1.26 1.31 1.35 1.40 1.45 0.53 0.56 0.58 0.60 0.62 0.64 Apparent Cohesion Required fo r S t a b i l i t y kg/m2 216 206 197 187 177 167 The results of the analysis indicate that, even without the benefit of a s t a b i l i z i n g root system, the slope would not have f a i l e d as long as the s o i l remained in an unsaturated condit ion. However, the topography and the s o i l drainage condition at th i s s i t e leaves l i t t l e doubt that Moist, no seepage Saturated, seepage para l le l to slope Saturated, seepage para l le l to slope 127 the slope did saturate during wet periods. Under these circumstances i t i s obvious from the very low calculated factors of safety, that the s t a b i l i t y of the slope depended heavily on the addit ional strength imparted by the tree root network. The deter iorat ion of the root net-work fol lowing c l e a r f e l l i n g appears to be the dominant causal factor in the release of th i s debris s l i d e . Concluding remarks Although the analyses described in the preceding section are fraught with uncertainties they, nevertheless, r a t i f y many of the conclusions outl ined in the e a r l i e r chapters. In pa r t i cu l a r , the analyses of fa i l u re s in the natura l , steepland so i l s (Landslide 1, 2 and 5) af f i rm that, under the most adverse s o i l water conditions l i k e l y to have occurred at these s i t e s , the slopes depended to a large degree on the tree root network for t he i r s t a b i l i t y . There i s , however, a strong p o s s i b i l i t y that so i l s with a high content of rocks may derive addit ional resistance to shear from the inter lock ing or keying effects of large boulders in close contact with each other. As the support provided by the tree roots i s gradually l o s t , roadcuts in the natural s o i l s with slopes greater than 35° become very susceptible to shallow f a i l u r e s . The a r t i f i c i a l estab-lishment of coni fer seedlings on the steep, cut faces and on the natural slopes at the crown of the cuts, soon a f te r road construct ion, would possibly prevent a great number of these smal l , shallow, debris s l i de s . The highly unstable road f i l l s on steep slopes const itute the most serious mass wasting problem in the study area. The loose structure of the f i l l s examined and the i r high void rat ios suggest that these materials received very l i t t l e or no compaction during road construct ion, presumably ,128 causing the f i l l s to have lower f r i c t i o n angles than equivalent, compacted materials and possibly rendering very loose f i l l s and s ide-castings susceptible to 1 i qu i f i ca t i on f a i l u r e s . Adverse road drainage conditions are, however, the most obvious cause of the large mass movements associated with roads. The analyses of Landslides 3 and 4 show that the s t a b i l i t i e s of the steeper f i l l slopes are extremely sens i t ive to water table r i se s . Although post-logging treatments such as water bar construction and outsloping, smoothing, and a r t i f i c i a l l y revegetating road surfaces might have preven-ted some lands l ides, there are logging roads above 800 m a l t i tude within the study area which traverse slopes so steep that such treatments would have probably been largely i ne f fec t i ve . On these high, steep slopes mass wasting must be regarded as an inev i tab le consequence of road construction. There i s no doubt in the w r i t e r ' s mind that current forest removal and road bui ld ing practices in many parts of the study area are at complete variance with sensible mountain-land management. I f road bui ld ing and timber removal threaten to destroy the s o i l resource, increase the sediment load of streams and r ivers and generally degrade the mountain environment, then, regardless of the conditions of land tenure, the standing timber values, and the magnitude of the potential p ro f i t s to be gained, road bui ld ing and logging cannot possibly be com-pat ib le with long term environmental goals. There i s a real and urgent need to define those mountain forest areas which can be managed for timber production without degrading the s o i l resource and hydrologic functions of those areas and those forests which should be l e f t in tact to f u l f i l a protective function. By way of example, i t i s possible to recognise and designate certa in areas of mountain forests in the study area, as being unsuitable for management by t r a d i t i o n a l , c l e a r f e l l i n g methods. C r i t e r i a of such areas are as fo l lows: (1) Forested slopes steeper than 35°. Mass wasting, pa r t i cu l a r l y debris avalanches associated with roads, becomes very troublesome on slopes steeper than 35°. In the past, slopes as steep as 50° have been c l e a r f e l l e d . (2) Forests above a l t i tudes of 1,000 meters. Although high level benches and other areas with stable te r ra in do ex i s t at a l t i tudes above 1,000 meters, construction of 'switch-back' road systems to provide access to these areas creates a serious mass wasting problem. (3) Forests adjacent to mountain torrents. (Riparian s t r ip s of fo res t , perhaps 200 to 300 feet wide, would provide protection against the accumulation of logging debris in the stream channels and would help maintain the s t a b i l i t y of the stream banks.) (4) Forests growing on poorly drained slopes and in large dra in -age depressions subject to periodic saturat ion. Forest areas included in (1), (2) and (3) can be recognised and delineated on 1:30,000 and possibly smaller scale aer ia l photographs but shallow drainage depressions and slopes with poorly drained so i l s are d i f f i c u l t to interpret on photographs and the i r i d en t i f i c a t i on nor-mally requires a ground inspection. Approximately 30-40 percent of the tota l area c l ea r f e l l ed within the study area over the la s t 10 years would probably be c l a s s i f i e d as unsuitable for timber production i f th i s forest c l a s s i f i c a t i o n scheme was adopted. BIBLIOGRAPHY Armstrong, J .E . , 1956. S u r f i c i a l geology of the Vancouver area, B r i t i s h Columbia. Geol. Surv. Can. Paper 55-40, 16 pp. , 1965. Geology, Vancouver North, B r i t i s h Columbia. Geol. Surv. Can. Map 1152A. Armstrong, J.E. and Brown, W.L., 1954. Late Wisconsin marine d r i f t and associated sediments of the Lower Fraser Val ley, B r i t i s h Columbia, Canada. B u l l . Geol. Soc. Amer., 65, p. 349-364. Aubertin, G.M., 1971. Nature and extent of macropores in forest so i l s and the i r influence on subsurface water movement, USDA Forest Serv. Res. Paper NE-192, 33 pp. Baker, R.F., 1953. Analysis of correct ive actions for highway land-s l i de s . Proc. Amer. Soc. Civ. Eng., 79, 25 pp. Baker, R.F. and Gray, H., 1960. Design of foundations, embankments and cut slopes. In 'Highway engineering handbook'. McGraw H i l l Book Co., p. 11-1-11-80. Barr, D.J. and Swanston, D.N., 1970. Measurement of creep in a shallow, slide-prone t i l l s o i l . Amer. J . S c i . , 269, p. 467-480. Bethlamy, N., 1962. F i r s t year effects of timber removal on s o i l moisture. B u l l . Int. Assoc. S c i , Hydrol., 7, p. 34-38. Bishop, A.W., 1955. The use of the s l i p c i r c l e in the s t a b i l i t y analysis of slopes. Geotechnique, 5, p. 7-17. Bishop, D.M. and Stevens, M.E., 1964. Landslides on logged areas in south-east Alaska. USDA Forest Serv. Res. Paper N0R-1, 18 pp. B r i t i s h Columbia Forest Service, 1957. Continuous forest inventory of B r i t i s h Columbia. Dept. of Lands and Forests. 223 pp. Brooke, R.C. 1966. Vegetation-environment relat ionships of subalpine mountain hemlock zone ecosystems. Ph.D. Thesis, University of B r i t i s h Jolumbia, 225 pp. 131 Brown, H.P., Pashin, A.S. and Forsa ith, C . C , 1952. Textbook of wood technology, Volume 2. McMillan and Sons Inc.* 783pp. Canada Department of Agr icu l ture, 1970. The system of so i l c l a s s i f i c a t i o n for Canada. Queen's Pr inter for Canada, Ottawa. 249 pp. Capper, P.L. and Cassie, W.F., 1963. The mechanics of engineering s o i l s . Spon Publishing Co., London, p. 250-252. Carson, M.A.,. 1971. An appl icat ion of the concept of threshold slopes to the Laramie Mountains, Wyoming. In 'Slope form and process. ' Inst. B r i t . Geog. Spec. Pub. 3, 175 pp. Cedergren, H.R., 1967. Seepage, drainage and flow nets. John Wiley and Sons Inc., New York. Chamberlin, T.W., 1972. Interflow in the mountainous forest so i l s of coastal B r i t i s h Columbia. In 'Mountain geomorphology'. B.C. Geog. Ser ies, No. 14, p. 121-127. Croft , A.R. and Adams, J .A., 1950. Landslides and sedimentation in the North Fork of the Ogden River, May 1949. USDA Forest Serv. Res. Paper INT-21. Department of Agr icu l ture, 1963. Climate of B.C., tables of temperature, p rec ip i ta t ion and sunshine. Annual Report for 1963. Dyrness, C.T., 1967. Mass s o i l movements in the H.J. Andrews Experimental Forest. USDA Forest Serv. Res. Paper PNW-42, 12 pp. E i s , S., 1972. Root grafts and the i r s i l v i c u l t u r a l impl icat ions. Can. J . of Forest Res., 2, p. 111-120. Endo, T. and Tsuruta, T., T969. The effect of tree roots on the shearing strength of s o i l . Annual Report, Forest Exp. Sta. , Hokkaido, p. 167-182. Flaccus, E., 1958. White Mountain lands l ides. Appalachia, 32, p. 175-191. Fredriksen, R.L., 1965. Christmas sto\rm damage on the H.J. Andrews Exper-imental Forest. USDA Forest Serv. Res. Note PNW-29, 11 pp. , 1970. Erosion and sedimentation fol lowing road construction and timber harvest on unstable so i l s in three small western Oregon watersheds. USDA Forest Serv. Res. Paper PNW-104, 15 pp. 132 Fujiwara, K., 1970. A study on the landsl ide by aer ia l photographs. Res. B u l l , of the College Exper. Forests, Hokkaido Univ., Japan, 27, p. 297-346. Gaiser, R.N., 1952. Root channels and roots in forest s o i l s . Soi l S c i . Soc. Amer. P r o c , 16, p. 62-65. Geological Association of Canada, 1958. Glac ia l Map of Canada. Golding, D.L., 1968. Regulation of water y i e l d and qual i ty in B.C. through forest management. Unpubl. Ph.D. Thesis, Univ. of B r i t i s h Columbia, 406 pp. Gonsior, M.J. and Gardner, R.B., 1971. Investigations of slope fa i l u re s in the Idaho Batho l i th . USDA Forest Serv. Res. Paper INT-97, 34 pp. Goodell, B.C., 1971. Forestry practices and water. Paper presented to the Annual Meeting of Inter ior Tree Farm Foresters, Kelowna, B r i t i s h Columbia, 8 pp. Gray, D.H., 1969. Effects of forest cutt ing on the s t a b i l i t y of natural slopes. Univ., of Michigan, College of Engineering, Progress Report, 67 pp. Hack, J.T. and Goodlett, J . C , 1960. Geomorphology and forest ecology of a mountain region in the central Appalachians. US Geol. Surv. Prof. Paper, 347, 66 pp. Haefe l i , R., 1965. Creep and progressive f a i l u r e in snow, s o i l , rock and i ce . Proc. of the Sixth Int. Conference on Soi l Mechanics and Foundation Engineering, p. 134-148. Hall i n , W.E., 1967. Soi l moisture and temperature trends in cutover and adjacent old growth Douglas f i r timber. USDA Forest Serv. Res. Note PNW-56, 11 pp. Holtz, W.G. and Gibbs, H.J., 1956. Shear strength of pervious gravel ly s o i l s . Proc. Amer. Soc. Civ. Eng., Paper No. 867. Hough, B.K., 1957. Basic so i l s engineering. Ronald Press, New York, 212 pp. Jackson, R.J., 1966. S l ips in re lat ion to r a i n f a l l and so i l cha rac te r i s t i c s . J . of Hydrol. (NZ), 5, p. 45-53. Jef f rey , W.W. 1968. Forest harvesting and water management. The Forestry Chronicle, 44, p.5-12. 133 Je f f rey , W.W. and Goodell, B.C. 1970. Land management in municipal watersheds. J . Amer. Water Works Assoc., 62, p. 380-385. Johnson, A.M. and Rahn, P.H., 1970. Mobi l izat ion of debris flows. Ze i t s ch r i f t f u r Geomorphologie, Supplementband 9, p. 168-186. Jones, F.O., Embody, D.R. and Peterson, W.L., 1961. Landslides along the Columbia River Val ley, Northeastern Washington. US Geol. Surv. Prof. Paper 367, 98 pp. Katsumi, S., 1965. Study of landsl ides in afforested land on Tert iary in Hokkaido. Annual Report of the Hokkaido Branch, Japan Government Forest Exper. S ta . , p. 186-188. Kennedy, R.W., 1958. Strength retention in wood decayed to small weight losses. Forest Products J . , 8, p. 308-314. Klock, G.O., 1972. Snowmelt temperature influence on i n f i l t r a t i o n and s o i l water retent ion. J . of Soi l and Water Conservation, 27, p. 12-13. Klute, A., 1966. Laboratory measurement of hydraulic conductiv ity of sat-urated s o i l . In 'B lack, C A . (Ed) 1966. Methods of s o i l ana ly s i s . ' Amer. Soc. of Agron. Inc., Madison, Wisconsin, p. 210-219. Kochenderfer, J .N . , 1970. Erosion control on logging roads in the Appalachians. USDA Forest Serv. Res. Paper NE-158, 28 pp. Kost ler, J .N . , Bruckner, E. and B ibe l r i e the r , H., 1968. Die Wurzeln der Waldbaume. Verlag Paul Parey, B e r l i n , 282 pp. Kraj ina, V . J . , 1965. Biogeoclimatic zones and c l a s s i f i c a t i o n of B r i t i s h Columbia. Ecology of Western North America, 1, p. 1-17. Lambe, T.W., 1951. Soi l test ing for engineers. John Wiley and Sons Inc., New York, 165 pp. Lamb, T.W. and Whitman, R.V., 1969. So i l mechanics. John Wiley and Sons Inc., New York, 553 pp. Lantz, R.L., 1971. Prevention and control of erosion and stream sediment-ation from forest roads. Proc. of Symposium on Forest Land Uses and Stream Environment, Oregon State Univ., Co r va l l i s , p. 76-83. Leopold, L.B., Wolman, M.G. and M i l l e r , J .P . , 1964. F luv ia l processes in geomorphology. W.H. Freeman and Co., San Fransisco, 522 pp. 134 Lesko, G.L. 1961. Ecological study of so i l s in the coastal western hemlock zone. M.Sc. Thesis, Univers ity of B r i t i s h Columbia, 141 pp. Lewis, T. and Lavkul ich, L.M., 1972. Soi l f o l i s o l s in the Vancouver area, B r i t i s h Columbia. Can. J . of So i l S c i , 52, p. 91-98. Liang, T., 1952. 'Landsl ides ' - an aer ia l photographic study. Unpubl. Ph.D. Thesis, Cornell Univers i ty, 274 pp. Lloyd, W.J., Schlots, F.E. and Deardorff, C.E., 1956. Forest management practices as related to and influenced by forest s o i l differences in Western Washington. Soi l S c i . Soc. Amer. Proc. 20, p. 105-107. Lutz, H.J. and Griswold, F.S., 1939. The influence of tree roots on so i l morphology. Amer. J . S c i . 237, p. 389-400. MacFarlane, I.C. (Ed.), 1969. Muskeg engineering handbook. Univ. of Toronto Press, 297 pp. McMinn, R.G., 1963. Character ist ics of Douglas f i r root systems. Can. J of Botany, 41, p. 105-123. Mi In, W.G., 1967. Earthquake epicentres and s t ra in release in Canada. Can. J . Earth S c i . , 4, p. 797-814. Mintzer, O.W., 1969. Applications of color aer ia l photography in engineering s o i l studies. Proc. of Seminar on New Horizons in Color Aer ia l Photo-graphy. Amer. Soc. Photogram. Pub., p. 145-164. Morton, J.W., 1970. Capilano - the story of a r i ve r . McClelland and Stewart L td . , Toronto, 184 pp. Nobel, E.L. and Lundeen, L . J . , 1971. Analysis of r ehab i l i t a t i on treatment a l ternat ives for sediment contro l . Proc. of Symposium on Forest Land Uses and Stream Environment, Oregon State Univ., Co r v a l l i s , p. 86-96. Paeth, R.C., Harward, M.E., Knox, E.G. and Dyrness, C.T., 1971. Factors a f fect ing mass movement of four so i l s in the western Cascades of Oregon. Soi l Sc i . Soc. Amer. P r o c , 35, p. 943-947. Parizek, E.J. and Woodruff, J . F . , 1957. Mass wasting and deformation of trees. Amer. J . S c i . , 255, p. 63-70. P a t r i c , J .H . , Douglas, J.E. and Hewlett, J.D., 1965. Soi l water absorp-t ion by mountain and piedmont forests . Soi l S c i . Soc. Amer. P r o c , 29, p. 303-308. 135 Pa t r i c , J.H. and Swanston, D.N., 1968. Hydrology of a sl ide-prone g lac ia l t i l l s o i l in southeast Alaska. J . Forestry, 16, p. 62-66. Penck, W., 1953. Morphological analysis of land forms. MacMillan and Co. L td . , London, 429 pp. Poole, D.H., 1969. Slope f a i l u r e forms; the i r i d e n t i f i c a t i o n , character-i s t i c s and d i s t r i bu t i on as depicted by selected remote sensor returns. Proc. Sixth Int. Symposium on Remote Sensing of the Environment, Vol. 11, Ann Arbor, Michigan, p. 927-965. Rapp, A., 1960. Recent development of mountain slopes in Karkevagge and surroundings, Northern Scandinavia. Geografiska Annaler, XLII, p. 73-200. , 1963. The debris s l ides at U lva lda l , Western Norway. An example of catastrophic slope processes in Scandinavia. Nachr. Wiss. Akad. Gottingen, 13, p. 195-210. Rice, R.M. and Foggin, G.T., 1971. Effects of high intens i ty storms on s o i l slippage on mountainous watersheds in Southern Ca l i f o rn i a . Water Resour. Res., 7, p. 1485-1496. Roddick, J.A., 1965. Vancouver, North Coquitlam and P i t t Lake map areas, B r i t i s h Columbia: with spec ia l emphasis on the evolution of the plutonic rocks. Geol. Surv. Can., Memoir 335, 276 pp. Schlots, F.E., Lloyd, W.J. and Deardorff, C.E., 1956. Soi l character-i s t i c s which af fect root penetration and timber s i t e qua l i ty of Douglas f i r in Western Washington. Soi l S c i . Soc. Amer. P r o c , 20, p. 101-105. Schweinfurth, U., 1967. liber eine besondere Form der Hangabtragung im neuseelclndishchen Fjordland. Ze i t . fur Geomorph., 11, p. 140-149. Sharpe, C.F.S., 1938. Landslides and related phenomena: a study of mass movements of s o i l and rock. Columbia Univ. Press, New York, 137 pp. Sheng, T.C., 1966. Landslide c l a s s i f i c a t i o n and studies of Taiwan. Chinese-Amer. Jo int Comm. on Rural Reconstruction, Forestry Series 10, 96 pp. Simonett, D.S., 1967. Landslide d i s t r i bu t i on and earthquakes in the Bewani and T o r r i c e l l i Mountains, New Guinea. In 'Landform studies from 136 Austra l ia and New Guinea'. Austral ian Nat. Univ. Press, Canberra, p. 64-75. Smith, T.W. and Cedergren, H.R., 1963. Cut slope design and lands l ides. In ' F i e l d test ing of s o i l s . A.S.T.M. Spec. Pub. 322, p. 135-158.' Sowers, G.F., 1963. Strength test ing of s o i l s . In 'Laboratory shear test ing of s o i l s . Spec. Tech. Pub. 361, A.S.T.M. p. 3-21. Sowers, G.B. and Sowers G.F., 1970. Introductory so i l mechanics and foundations. MacMillan Co., New York, 556 pp. Stephens, E.P., 1956. The uprooting of trees: a forest process. Soi l S c i . Soc. Amer. P r o c , 20, p. 113-116. Stone, E.C., 1969. An analysis of the buffers and watershed management required to preserve the redwood forest and associated streams in the Redwood National Park. Report produced for US Nat. Park Serv. Swanston, D.N., 1967a. Geology and slope f a i l u r e in the Maybeso Val ley, Prince of Wales Island, Alaska. Unpubl. Ph.D. Thesis, Michigan State Univ., 206 pp. , 1967b. So i l water piezometry in a southeast Alaska landsl ide area. USDA Forest Serv. Res. Note PNW-68, 17 pp. , 1969. Mass wasting in coastal Alaska. USDA Forest Serv. Res. Paper PNW-83, 15 pp. , 1970. Mechanics of debris avalanching in shallow t i l l s o i l s of southeast Alaska. USDA Forest Serv. Res. Paper PNW-103 17 pp. , 1971. Pr inc ipa l mass movement processes influenced by logging, road bui ld ing and f i r e . Proc. of Symposium on Forest Land Uses and Stream Environment, Oregon State Univ., Co r va l l i s , p. 29-39. Takahasi, T., 1968. Studies of the forest facu l t ie s to prevent lands l ides. B u l l , of the Faculty of A g r i c , Shizuoka Univ., 18, p. 85-101. Taylor, D.W., 1948. Fundamentals of s o i l mechanics. John Wiley and Sons Inc., New York, p. 408-438. Terzaghi, K., 1950. Mechanism of lands l ides. In 'Appl icat ion of geology to engineering p rac t i ce . ' Geol. Soc Amer. Berkey V o l . , p. 83-123. 137 Turmanina, V., 1965. The strength of tree roots. B j u l l . Mosk. Obsc. Ispyt. P r i r . (Otd. B i o l . ) , 70, p. 36-45. Varnes, D.J., 1958. Landslide types and processes. In 'Landslides and engineering p rac t i ce . ' National Res. Council Pub. 544, Highway Res. Board Spec. Report 29, p. 20-47. Wa l l i s , G.W. and Ginns, J .H . , 1968. Annosus root rot in Douglas f i r and western hemlock in B r i t i s h Columbia. Dept. of Fisheries and Forestry, Forest Pest Leaf let , 15, 7 pp. Walpole, R.E., 1968. Introduction to s t a t i s t i c s . MacMillan Co., New York, 365 pp. Wangaard, F.F., 1950. The mechanical properties of wood. John Wiley and Sons Inc., New York, 377 pp. Wentworth, C.K., 1943. Soi l avalanches on Oahu, Hawaii. Geol. Soc. Amer. B u l l . , 54, p. 53-64. White, S.E., 1949. Processes of erosion on steep slopes of Oahu, Hawaii. Amer. J . S c i . , 247, p. 168-186. Whitham, K., M i ln , W.G. and Smith, W.E., 1970. The new seismic zoning map for Canada, 1970 Ed i t ion . Dept. Energy, Mines and Resources, Ottawa Pub., 8 pp. Whitman, R.V. and Ba i ley, W.A., 1967. Use of computers for slope s t a b i l i t y analys i s . Proc. Amer. Soc. Civ. Eng., 93, SM4, p. 475-498. Wil l iams, G.P. and Guy, H.P., 1971. Debris avalanches - a geomorphic hazard. In 'Environmental geomorphology.' Pub. in Geomorph. State Univ. of New York, p. 25 Wright, A.C.S. and M i l l e r , R.B., 1952. So i l s of southwest Fiordland. N.Z. So i l Bur. B u l l . , 7, 36 pp. Wu, T.H., 1966. Soi l mechanics. Al lyn and Bacon Inc., Boston, 431 pp. Zaruba, Q. and Mencl, V., 1969. Landslides and t he i r cont ro l . E l serv ier Pub. Co. Inc., New York, 205 pp. 138 Appendix 1 Details of major storm r a i n f a l l s at Seymour Fa l l s and Cleveland Dam, southwestern B r i t i s h Columbia, 1959-1970. The fol lowing r a i n f a l l information was extracted from the precip-i t a t i o n records held at the Greater Vancouver Water D i s t r i c t Of f i ce , Vancouver. Maximum 5 minute, 15 minutes and 1 hour i n tens i t i e s ( m i l l i -meters per hour) as well as the 24 hour tota l (mil l imeters) are shown. (a) Seymour Fa l l s n . Intensit ies (mm/hour) 24 hour to ta l 5 minutes 15 minutes 1 hour (mm) 28-4-59 10.7 10.2 8.9 127.3 19-11-60 30.5 24.4 15.0 101.6 14,15-1-61 33.5 28.5 25.4 365.8 30-8-61 18.3 17.3 9.4 178.8 2-1-62 13.2 11.2 10.7 118.1 19-11-62 18.3 16.8 13.2 146.1 30-12-62 21.3 14.2 10.9 105.7 13-9-63 61.0 47.8 15.2 37.6 21-10-63 25.9 19.3 15.2 148.3 25-11-63 21.3 17.3 11.7 155.7 22-12-63 27.4 24.4 20.6 116.8 24-12-63 21.3 19.3 17.8 244.1 30-11-64 18.3 13.2 11.4 163.8 5-10-65 51.8 43.7 31.0 169.4 15-12-66 15.2 12.2 10.2 127.8 17-12-66 15.2 11.7 9.7 145.8 1-10-67 27.4 20.3 16.5 140.0 10-10-67 30.5 25.4 21.6 110.7 10-12-67 30.5 24.4 18.8 134.6 22-12-67 18.3 14.2 12.2 121.9 18-1-68 16.8 14.2 10.4 119.9 17-9-68 106.2 22-9-69 30.5 20.3 9.9 112.5 5-4-70 15.2 13.2 9.9 114.1 (b) Capilano Dam 30-11-64 124.2 15-12-66 93.7 1-10-67 51.3 7-10-67 116.1 10-10-67 48.0 10-12-67 41.7 22-12-67 71.9 18-1-68 157.5 17-9-68 124.5 22-9-69 75.9 Appendix 2 (a) Determination of void rat ios for steepland subsoils Each undisturbed, 2,500 cm sample of s o i l was oven-dried at 105° C and weighed. An approximately 150 g sub-sample of the dried subsoil was used for determination of the spec i f i c gravity of the s o i l so l ids by the water immersion method described by Lambe (1951). The void r a t i o is given as e = void ra t i o G g = spec i f i c gravity of the so l ids 3 Yw = unit weight of water g/cm 3 V = to ta l volume of sample cm Wg = weight of the so i l so l ids g The derivat ion of th is formula is given in Lambe (1951) p. 154. (b) Determination of void rat ios for compacted t i l l 3 I r regular, approximately 400 cm intact samples of cement-like t i l l were oven-dried, weighed, coated with a th in layer of parafin wax and immersed in water to obtain the i r volume. The spec i f i c gravity of the sol ids and the void ra t io were obtained in the same manner as for the subsoi l s . 140 (c) Determination of the saturated unit weight fo r subsoil and t i l l The saturated unit weight Ys fo r both types of material was obtained from (d) Results Subsoils. T i l l . Y s = {r^-r^ Yw Sample Spec i f ic gravity Saturated unit Void ra t i o of sol ids weight kg/m3 551 2.4627 1,650 1.25 552 2.4473 1,690 1.09 553 2.4584 1,750 0.93 554 2.4910 1,620 1.41 555 2.4631 1,610 1.38 2.4613 1,660 1.21 TI 2.6794 2,370 0.23 T2 2.6799 2,300 0.29 2.6797 2,340 0.26 141 Appendix 3 (1) Slope analysis by the ordinary method of s l i ce s The free body cut from a slope by a c i r c u l a r f a i l u r e surface ((a) in above f igure) i s divided up into several s l i ce s or segments and the equi l ibr ium of each of these s l i ce s is considered. The unknown forces acting on each sli<?e include the weight of the s l i c e W ,^ the normal force N.j and the shear force T. acting on the f a i l u r e plane, the normal forces F.. and F^ + n and the shear forces t^ and t 1 - + n pressures IK acting on the f a i l u r e arc. In the ordinary method of s l i ce s i t i s assumed that the forces acting on the sides of a l l s l i ce s have zero resultant in the d i rect ion normal to the f a i l u r e arc for that s l i c e . The factor of safety for the free body i s defined in terms of moments about the centre of the f a i l u r e arc. FS = moment of shear strength along f a i l u r e arc moment of weight of free body i=n i=n _ R ( s c i . + tan 0 z N.) , n " i=l 1 jsQ 1 u ; RF"^ . s i n a . Where R = radius of the f a i l u r e arc n = number of si ices In the ordinary method of s l i ce s N i = W. cosa . - U. — — - — (2a) = W. cosa . - u i l i - (2b) Combining equations (1) and (2b) and cancel l ing the R's i=n • z [cl,- + (W. cosa. - u. l.) tan 0] FS = i=l 1 1 1 1 1 (3) i=n z W. s ina. i=l 1 1 This approximate so lut ion, which neglects the forces on the sides of the s l i c e s , y ie lds a safety factor which i s generally 5 to 15 percent but may be 40 to 50 percent below the values obtained by more rigorous methods, pa r t i cu l a r l y i f the pore pressures are large and the curvature of the f a i l u r e arc is high. (2) Bishop method of s l i ces Bishop (1955) refined the computation of the factor of safety by considering the forces acting on the s l i c e sides and assuming they have 143 zero resultant in the ve r t i ca l d i r ec t i on . Under these circumstances Bishop showed that W1 - u i l i coso 1 - C l i S i n a i N, = FS — — — — - ( 4 ) 1 cosa^ + (tan 0 sina^) FS Combining equations (1) and (4) i = n 1 r j i [ c l , cos., • (H , - U 1 1 , cosc.)tan H ] C O S a . t ( t a n 0 FS = — F S i = n . s , W. s ina. i = l 1 1 As FS appears on both sides of equation (5), an i t e r a t i v e procedure must be employed to assess FS but convergence of t r i a l s i s very rap id. Often the FS obtained by the ordinary method of s l i ce s i s used as the i n i t i a l j-t r i a l value. To aid the ca lcu lat ions , charts can be used to evaluate the function cosa^ + (tan 0 sina^) FS Generally, the factor of safety produced by the Bishop method is 1 to 10 percent higher than the factor of safety given by the ordinary method of s l i c e s . D i f f i c u l t i e s in using equation (5) can be encountered i f the pore water pressures involved in the problem are large and the FS is less than unity. (3) I n f i n i t e slope analysis According to Taylor (1948) an i n f i n i t e slope designates " — a constant slope of unlimited extent which has constant conditions and constant s o i l properties at any given distance below the surface of the s lope." Such ideal conditions are rare ly shown by natural slopes but, i f the thickness of the s o i l mantle or mantle of potent ia l l y un-stable slope materials i s small compared to the height of the slope, then the slope may be ca l led i n f i n i t e . Under these circumstances the f a i l u r e plane i s pa ra l le l to the slope. Although the mineral portions of the steepland forest so i l s are predominently cohesionless, fo r pract ica l purposes these s o i l s can be considered cohesive because of the effects imparted by the tree root network. The analysis of cohesive, i n f i n i t e slopes is outl ined below. ( i ) Analysis of unsaturated i n f i n i t e slopes with no seepage In the diagram above showing part of a dry, i n f i n i t e slope the stress components on the base plane at depth z are T = y^Z Sina COSa a = y,z COS 2a Subst itut ion of these expressions into the f a i l u r e law T y ie lds the equation c 2 = cos a (tana - tan 0) c + CTtan 0 Y d (6) and rearrangement of the terms gives the cohesion needed for s t a b i l i t y p c = YD cos a (tana - tan 0) (7) ( i i ) Analysis of saturated i n f i n i t e slopes with seepage If the s o i l i s saturated with the free water surface at the ground surface and seepage occurs para l le l to the slope, then the stresses acting on the base plane are T = Y SZ cosa sina , 2 0" = Y^Z COS a Subst itut ing these expressions into the f a i l u r e law T = c + atan 0 gi ves (8) —=• = cos2 a (tana - — tan 0) Y „ Z X Y ' s r s and the cohesion required for s t a b i l i t y is c = Y C Z C O S a (tana - : — tan 0) 5 YS (9) The factor of safety (FS) for an i n f i n i t e , cohesive slope with seepage para l l e l to the slope is FS = avai lable shear strength  shear stress required for equi l ibr ium 2 C + y,Z cos a tan 0 - l b ( 1 0 ) YgZ Sina COSa For cohesionless materials the FS reduces to FS = Y b tan 0 YS tan a (11) Sample C a l c u l a t i o n s o f Factors o f Safety (FS) f o r a Road-cut Slope a t the s i t e o f Landslide 1 Ordinary Method o f S l i c e s Bishop Method o f S l i c e s S l i c e Area a. l 1. i u. x u . l . i i W. i (W\ cosaj_- W. s i n a • (W±- 1 a x b a x b u . l . ) t a n 0 l l i l u.l.cosa.) i l l tan 0 cosa^ + (tan0.sina i/FS) 2 m o m kg/m kg kg kg kg kg FS=.38 FS=.40 FS=.38 FS= .40 1 0.121 42 0.549 122.1 67.0 190.7 51.9 127.6 98.6 0.506 0.523 49.9 51 .6 2 0.204 46 0.457 182.8 83.5 319.3 96.9 229.6 182.9 0.495 0.512 90.5 93 .6 3 0.251 50 0.457 243.7 111.4 387.0 96.2 296.4 220.8 0.487 0.504 107.5 111 .3 4 0.307 54 0.488 122.1 59.6 456.0 146.0 368.9 294.7 0.481 0.499 141.8 147 .1 5 0.372 58 0.610 - - 539.4 200.1 457.4 377.6 0.478 0.497 180.5 187 .7 6 0.279 66 0.762 - - 404.6 115.3 369.8 283.2 0.478 0.498 135.4 141 .0 7 0.102 72 0.945 - - 147.9 32.0 140.7 103.5 0.485 0.507 50.2 52 .5 738.4 1,990.4 755.8 784 .8 FS 738 .4 0.371 1st t r y : FS 784.8 = 0.394 1990 .4 1990.4 Y m = 1,450 kg/m 2nd t r y : FS 755.8 = 0.380 TR 1,680 kg/m 3 0 = 35° 1990.4 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0101126/manifest

Comment

Related Items