FOREST STAND CHARACTERISTICS AS INDICATORS OF HYDROGEOMORPHIC ACTIVITY ON FANS by DAVID JOHN WILFORD M.F., University of British Columbia, Vancouver, 1975 B.Sc.F, University of British Columbia, Vancouver, 1973 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF DOCTOR OF PHILOSOPHY In THE F A C U L T Y OF G R A D U A T E STUDIES (Faculty of Forestry) Department of Forest Resources Management We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA January 2003 © David John Wilford, 2003 In p r e s e n t i n g this thes is in part ial f u l f i lment o f t h e r e q u i r e m e n t s f o r a n a d v a n c e d d e g r e e at t h e Un ive rs i t y o f Br i t ish C o l u m b i a , I a g r e e that t h e Library shal l m a k e it f ree ly avai lable f o r r e f e r e n c e a n d s tudy . I fu r ther agree that p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f this thes is f o r scho la r l y p u r p o s e s m a y b e g r a n t e d by the h e a d o f m y d e p a r t m e n t o r by his o r he r representat i ves . It is u n d e r s t o o d that c o p y i n g o r p u b l i c a t i o n o f this thesis fo r f inancia l ga in shal l n o t b e a l l o w e d w i t h o u t m y w r i t t e n p e r m i s s i o n . T h e Un ivers i ty o f Br it ish C o l u m b i a V a n c o u v e r , C a n a d a DE-6 (2/88) A B S T R A C T Forested fans are common landforms in west central British Columbia. They can be subject to hydrogeomorphic processes ranging from debris flows to floods carrying bedload and woody debris. These processes represent hazards for forest management activities such as road construction and harvesting. Conversely, forest management activities can exacerbate the effect of natural hydrogeomorphic processes, increasing the level of disturbance on fan surfaces and in stream channels. This thesis presents the scientific basis for a hazard classification of fans for forest management. The classification is based on forest stand characteristics, airphoto and site features, and basic watershed attributes. Forest stands are used to determine the power of hydrogeomorphic events (floods, debris floods, and debris flows). High-power events clear swaths through a forest stand, which lead to the establishment of distinctive cohorts. Low-power events lead to the deposition of sediments under a forest canopy, which generally results in growth responses rather than mortality or removal of the forest stand. In cases where there is mortality, tree stems remain in situ rather than being cleared from the affected area. Site features of sediment deposits are used to determine the hydrogeomorphic process. Watershed attributes were determined for identifying hydrogeomorphic processes, power and disturbance extent level. The attributes are basic morphometric parameters such as watershed area and relief, and most can be measured directly from topographic maps without the use of geographic information systems (GIS). Dendroecological techniques were used to determine the number of events occurring on study fans. Regression equations were developed to predict the number of events in the past 50 years. The independent variables were watershed attributes that can be determined by a combination of topographic map measurements and GIS. An examination of 55 fans with forestry activities provides a comprehensive hazard perspective on key fan and watershed attributes, and on the influence of road construction and harvesting prescriptions in zones of fans that are subjective to active hydrogeomorphic processes. The hazard classification and the forestry activity review provide the first comprehensive basis for sustainable forest management on fans in British Columbia. ii T A B L E O F C O N T E N T S A B S T R A C T ii L I S T O F T A B L E S vi L I S T O F F I G U R E S viii A C K N O W L E D G E M E N T S x C H A P T E R 1. I N T R O D U C T I O N 1 1.1 STATEMENT OF THE ISSUE 1 1.2 STATEMENT OF OBJECTIVES 2 1.3 RESEARCH QUESTION 2 1.4 SCIENTIFIC AND OPERATIONAL IMPLICATIONS OF THE RESEARCH 3 1.5 THESIS OUTLINE 4 C H A P T E R 2. F A N S A N D H Y D R O G E O M O R P H I C P R O C E S S E S 6 2.1 INTRODUCTION 6 2.2 FAN DEFINITION 6 2.3 TYPES OF FANS 8 2.4 CONTEMPORARY FAN ACTIVITY 9 2.5 HYDROLOGIC AND GEOMORPHIC PROCESSES 11 2.5.1 Floods 11 2.5.2 Debris flows 14 2.5.3 Debris floods 20 2.5.4 Avulsions 21 2.5.5 Channel entrenchment 23 2.5.6 Groundwater hydrology of fans 25 2.6 CONCLUSIONS 26 C H A P T E R 3. D E S C R I P T I O N O F T H E S T U D Y A R E A 28 C H A P T E R 4. T H E S T U D Y F A N S : F O R E S T S T A N D S A N D H Y D R O G E O M O R P H I C P R O C E S S E S 31 4.1 INTRODUCTION 31 4.2 FOREST STANDS AND HYDROGEOMORPHIC PROCESSES 31 4.3 METHODS 38 4.4 RESULTS 48 4.5 DISCUSSION 51 iii 4.5.1 Morphological overview of study fans and their watersheds 56 4.5.2 Forest stands on fans as indicators of hydrogeomorphic activity 58 4.5.3 The hydrogeomorphic role of forests on fans 61 4.6 CONCLUSIONS 64 C H A P T E R 5. A N A L Y S I S O F W A T E R S H E D A T T R I B U T E S 66 5.1 INTRODUCTION 66 5.2 WATERSHED ATTRIBUTES 66 5.3 METHODS 69 5.4 RESULTS 70 5.4.1 Differentiating hydrogeomorphic processes 70 5.4.2 Differentiating floods 70 5.4.2 Differentiating debris floods 71 5.4.3 Differentiating debris flows 73 5.5 DISCUSSION 78 5.5.1 Differentiating hydrogeomorphic processes 78 5.5.2 Differentiating floods 79 5.5.3 Differentiating debris floods 80 5.5.4 Differentiating debris flows 81 5.6 CONCLUSIONS 81 C H A P T E R 6. F R E Q U E N C Y O F H Y D R O G E O M O R P H I C E V E N T S 83 6.1 INTRODUCTION 83 6.2 METHODS 84 6.3 RESULTS 91 6.3.1 Skeleton plots and summary of events 91 6.3.2 Regression equations 92 6.3.3 Details on ESASx polygons 92 6.3.4 Details on site-level debris flow watersheds 94 6.4 DISCUSSION 94 6.4.1 The equations and their variables 94 6.5 CONCLUSIONS 99 C H A P T E R 7. F O R E S T R Y A C T I V I T I E S O N F A N S 100 7.1 INTRODUCTION 100 7.2 METHODS 100 7.3 RESULTS 108 7.3.1 Overview 108 7.3.1.1 Pre-logging aerial photographic features 108 7.3.1.2 Characterizing events 110 iv 7.3.1.3 Overview of forestry activity relations and impacts Ill 7.3.2 Specific details 115 7.3.2.1 Roads and drainage structures 117 7.3.2.2 Logging 122 7.3.3 Impacts 124 7.4 DISCUSSION 125 7.4.1 Pre-logging aerial photographic features 125 7.4.2 Characterizing events 127 7.4.3 Forestry activities 128 7.4.3.1 Drainage structures 128 7.4.3.2 Riparian logging 131 7.4.3.3 Soils and mass wasting 135 7.4.3.4 Silvicultural systems 136 7.4.3.5 Utilization standards 137 7.4.3.6 Recognition of impacts 137 7.5 CONCLUSIONS 138 C H A P T E R 8. C O N C L U S I O N S 139 L I T E R A T U R E C I T E D 147 A P P E N D I X A . O P E R A T I O N A L N E T W O R K 158 A P P E N D I X B . S U M M A R Y O F D I S T U R B A N C E A G E N T S 159 A P P E N D I X C . A B B R E V I A T I O N S F O R T R E E S P E C I E S A N D E C O S Y S T E M S 161 A P P E N D I X D . A T T R I B U T E S O F T H E S T U D Y W A T E R S H E D S U S E D F O R T H E A N A L Y S I S O F V A R I A N C E 163 A P P E N D I X E . A N E X A M P L E S K E L E T O N P L O T A N D S U M M A R Y O F E V E N T S 172 A P P E N D I X F. S U M M A R Y O F E V E N T S O N S T U D Y F A N S 184 A P P E N D I X G. D A T A S E T S F O R R E G R E S S I O N A N A L Y S I S 191 A P P E N D I X H . D A T A S U M M A R Y O F F O R E S T R Y A C T I V I T I E S O N F A N S . . . 208 LIST O F T A B L E S Table 2.1. Watershed characteristics related to floods 13 Table 2.2. Classification and characteristics of flow processes and deposits 14 Table 4.1. Approximate ages associated with later developmental stages (west central British Columbia) 33 Table 4.2. Characteristic forest-stand signatures on fans and their watersheds caused by hydrogeomorphic processes 35 Table 4.3. A subsample of fans with elevated surfaces comparing forest cover on elevated and active fan surfaces 49 Table 4.4. The number of fans in the different forest stand-based classification categories 51 Table 4.5. Hydrogeomorphic, forest stand and site level information for the study fans 52 Table 4.6. Summary data on measured widths of disturbance in the forest canopy from hydrogeomorphic events 54 Table 4.7. Summary of hydrogeomorphic roles of forest stands on fans with riparian forests 55 Table 4.8. Summary of site features related to sediment deposition from hydrogeomorphic events on fans with riparian forests 55 Table 4.9. Stream channel gradients at the apex by hydrogeomorphic process 55 Table 4.10. Watershed size and relief associated with hydrogeomorphic process 56 Table 5.1. Watershed attributes 67 Table 5.2. Differentiating watershed attributes for floods, debris floods and debris flows and their associated P-values 71 Table 5.3. Effectiveness of class limits in identifying hydrogeomorphic processes 72 Table 5.4. Differentiating watershed attributes for the three categories of floods and their associated P-values 73 Table 5.5. Effectiveness of class limits in classifying flood watersheds 74 Table 5.6. Effectiveness of class limits in classifying low- and high-power debris flood watersheds 75 Table 5.7. Effectiveness of class limits in classifying high-power debris flood watersheds 76 Table 5.8. Effectiveness of class limits in classifying debris flow watersheds 77 vi Table 5.9. Class limits for the hydrogeomorphic processes 82 Table 5.10. Class limits for power and disturbance extent for the hydrogeomorphic processes 82 Table 6.1. Watershed attributes 89 Table 6.2. Regression equations to predict the total number of events during a 50-year period 92 Table 6.3. Details on the site-level debris flow watersheds 93 Table 7.1. Description of forestry activities 107 Table 7.2. Pre-logging aerial photographic features and relation to management issues 108 Table 7.3. Occurrence of characterizing events since forestry activities 110 Table 7.4. Impacts by category of the forest stand-based hazard classification 111 Table 7.5. Associated forestry activities and impacts by impact classes 112 Table 7.6. Forestry activities associated with PI impacts by category of the forest stand-based hazard classification .114 Table 7.7. Forestry activities associated with P2 impacts by category of the forest stand-based hazard classification 114 Table 7.8. PI impacts by category of the forest stand-based hazard classification 115 Table 7.9. P2 impacts by category of the forest stand-based hazard classification 115 Table 7.10. A summary of specific details on forestry prescriptions grouped according to their role in exacerbating hydrogeomorphic events 116 Table 7.11. Summary of roads that climb to streams in at least one direction 120 Table 7.12. Details of cases where climbing roads are associated PI and P2 impacts 121 Table 7.13. Details of cases where climbing roads did not cause impacts 121 Table 7.14 A forestry time frame for the significance of hydrogeomorphic disturbances on fans 134 Table 8.1 Forest practices and impacted features associated with hydrogeomorphic events 142 Table 8.2 A summary of hydrogeomorphically appropriate forestry prescriptions on fans.... 144 Table 8.3 A summary of forestry prescriptions that exacerbate hydrogeomorphic events on fans 145 vii LIST OF FIGURES Figure 2.1. A clearcut fan with a classic shape 6 Figure 2.2. Characteristic sorting of alluvial sediments 8 Figure 2.3. Poorly sorted sediments of a colluvial fan 9 Figure 2.4. Geomorphic and sedimentologic features of debris flows, debris floods and alluvial floods 16 Figure 2.5. An entrenched stream channel on a debris flow fan 23 Figure 3.1. Location map of the study area 28 Figure 4.1. An aerial photograph of a 100-year old cohort following a high-power stand level disturbance flood 40 Figure 4.2. A stereoscopic aerial photographic pair of a recent high-power stand level disturbance from a debris flow 41 Figure 4.3. A young cohort growing on the sediments of a high-power site level disturbance debris flow 42 Figure 4.4. Recently deposited sediments behind a log are forming a "log step" 42 Figure 4.5. A log retaining wall storing a considerable volume of sediments 43 Figure 4.6. A woody dyke composed of small woody debris along the edge of a channel 43 Figure 4.7. Roots provide reinforcement to the soil mass, delaying the erosional effects of broadcast flows 44 Figure 4.8. An example of a buried tree. Note the lack of butt flare 45 Figure 4.9. Tree holes are the result of rotting following deep burial of tree stems 46 Figure 4.10. Erosion of sediment deposited around trees can expose adventitious roots 46 Figure 4.11. Scars can provide accurate dates of hydrogeomorphic events 47 Figure 4.12. A "recent" deposit of sediments within the hydrogeomorphic riparian zone with limited organic accumulation and moss cover 47 Figure 4.13. A contrast between the high volume forest cover on the more active fan surface and the low-volume forest stand on the adjacent elevated fan surface 50 Figure 4.14. The avulsion process or creation of new channels is delayed by forest stands and the associated woody debris 64 Figure 5.1. Scattergram using Melton ratio and watershed length with class limits for the hydrogeomorphic processes 72 viii Figure 5.2. Bar graph of watershed area with class limits for the three flood categories 74 Figure 5.3. Scattergram of watershed relief versus commercial forest cover, and the class limit for differentiating low- and high-power debris flood watersheds 75 Figure 5.4. Scattergram of high-power debris floods with the class limit between site and stand level disturbance levels 76 Figure 5.5. Bar graph of commercial forest cover with a class limit for differentiating debris flow watersheds 77 Figure 6.1. A wedge with multiple scars 85 Figure 6.2. Traumatic resin canals in the late wood 86 Figure 6.3. Growth rings with compression wood 86 Figure 6.4. A magnified core from a spruce showing strong, abrupt growth change 87 Figure 6.5. ESAs are forested areas with steep, failing slopes 93 Figure 7.1. Multiple channels are present on this fan 102 Figure 7.2. High sediment loads are being transported by this stream 103 Figure 7.3. The stream channel disappears from view on the aerial photograph 104 Figure 7.4. A large debris jam in the main channel led to the formation of a second channel which initiates at an abrupt angle 105 Figure 7.5. Major sediment sources near the watershed mouth can provide direct delivery of sediments and debris to fans 106 Figure 7.6. The maximum width of this structure is 8.4 meters while the channel is 15 meters wide. The resulting change in channel geometry can lead to scour and subsequent downstream deposition 128 Figure 7.7. Excavation into the stream channel on this fan was not stabilized with rip-rap 130 Figure 7.8. Pre- and post-harvesting aerial photographs illustrating the effects of logging the hydrogeomorphic riparian zone 132 ix A C K N O W L E D G E M E N T S I am grateful to my initial supervisor, Professor Roy Sidle, for providing the opportunity, technical support and guidance to initiate this project. With the beginning of fieldwork Professor John Innes became my supervisor and embraced this role well beyond my expectations. I have benefited from Professor Innes' advice and guidance in all aspects, from fieldwork to completion of this thesis. Working under him has been a stimulating and rewarding experience. I am grateful to my committee, Karel Klinka, Michael Bovis, and Brent Ward, for their questions, reviews and guidance during this project. I gratefully acknowledge funding for this project from the BC Ministry of Forests and Forest Renewal BC. The need for operational forestry guidance on fans was identified by David Rebagliati and Ted Wilson of the Ministry of Forests (MOF), and Andy Witt of the Ministry of Water, Land and Air Protection. Andy has worked diligently to establish a policy framework for the results of this project. I appreciated the support of my colleagues in the MOF Research Group in Smithers: Allen Banner, David Coates, Karen McKeown, Marty Kranabetter, Phil LePage, Jim Pojar, Jim Schwab, and Doug Steventon. Matt Sakals was my very able research assistant. His technical knowledge, keen sense of exploration, and enthusiasm made him a joy to work with, and made the three years of very hard work fly by. I acknowledge the permission to use aerial photographs given by Damien Keating of West Fraser Mills Inc. (Skeena Sawmills Div.) and Rob Ziegler of Skeena Cellulose Inc. (Terrace). I thank my wife Katherine and sons Jordan and Michael for joining me at UBC and sharing in the joys and sacrifices this project presented. Finally, I dedicate this thesis to the memory of my grandfather H. H. Tate who instilled in me the desire to become a forester. x C H A P T E R 1. I N T R O D U C T I O N Fans are conical-shaped deposits of sediment formed where a stream emerges from the confines of a steep mountain channel. Deposition can be the result of historic or active alluvial or colluvial processes. A central feature of most fans is an unconfined stream channel or channels. Characteristically, fans are productive forest sites that offer low cost harvesting opportunities because of easy access and relatively gentle gradients. Roads frequently cross fans during the course of forest development. Unfortunately, harvesting and road building on fans can be problematic due to the presence of active geomorphic and hydrologic processes. This thesis explores the use of the composition and structure of forest stands on fans as indicators of the frequency, power and disturbance extent of contemporary geomorphic and hydrologic processes influencing fans. 1.1 STATEMENT OF THE ISSUE At present in British Columbia, foresters have very limited guidance regarding the identification of hydrogeomorphic hazards with regard to operational planning and practices on fans. There is very little mention of fans in the Forest Practices Code of BC (FPC). Also, there has been very limited hydrogeomorphic research related to forestry on fans. From this lack of attention, it might be assumed that hydrogeomorphic hazards on fans are not important issues. The opposite is the case. Across the province, roads and harvesting on fans have aggravated natural hydrogeomorphic processes. The results have been impacts to fish habitat, loss of drainage structures and roads, mortality in plantations and loss of property and improvements. These impacts stem from a failure to recognize active hydrologic and geomorphic processes and an inability to translate such information into appropriate sustainable forestry prescriptions. 1 1.2 STATEMENT OF OBJECTIVES The objectives of this project were to provide a scientific basis for forest management on fans by: 1) Determining the spatial and temporal characteristics of the hydrogeomorphic processes that influence fans; 2) Determining site and watershed factors that can be used to identify hydrogeomorphic processes that pose hazards to fan stability and forest land use; and, 3) Developing a compendium of forestry experience on fans that provides guidance for sustainable forest management practices. The project was undertaken in the Prince Rupert Forest Region of British Columbia, with emphasis in the Terrace through Houston area. It was recognized at the outset that any quantitative results would only be applicable to the study area, however the qualitative results would most likely be widely applicable to forested fans in general. 1.3 RESEARCH QUESTION The research question addressed in this thesis is: To what extent do natural forest stands on a fan indicate the power, disturbance extent and frequency of natural, contemporary hydrogeomorphic processes influencing the fan? For clarification, the following definitions apply in this study: • "Natural" implies no or very limited human modification to a fan and its contributing watershed. • "Forest stand" is a spatially continuous group of trees and associated vegetation having similar structures and growing under similar soil and climatic conditions (Oliver and Larson 1996). • "Contemporary" is based on the maximum age of trees in the study area (500 years) but is generally less than 200 years. 2 • "Hydrogeomorphic" means a combination of hydrologic and geomorphic processes, such as debris flows, debris floods, floods and channel avulsions. 1.4 SCIENTIFIC AND OPERATIONAL IMPLICATIONS OF THE RESEARCH At present in British Columbia there is no coherent body of scientific or operational knowledge regarding forestry on fans. This study will begin laying a foundation for scientific investigation. The exploration of forestry practices on fans will also provide a framework for a coherent body of knowledge related to operational forestry on fans. Currently the FPC provides very limited guidance with regards to fans. The Gully Assessment Procedure (GAP) (Anon. 2001) uses three fan features to determine potential hazards on fans: number of channels, channel depth, and indications of past debris flows. However, the GAP does not provide guidance related to roads or harvesting on fans. The Channel Assessment Procedure (CAP) (Anon. 1998) refers to fans as geomorphically active landscape features, but provides no guidance for forestry activities. The current FPC is prescriptive, meaning that procedures are used to provide guidance for management to achieve desired results. In the case of forest hydrology, the Watershed Assessment Procedures (Anon. 1999a), GAP and CAP are employed to limit cumulative and site-specific watershed impacts. However, the current FPC is under major review with the objective of becoming results-based rather than prescriptive. It is therefore unlikely that a "Fan Assessment" procedure will be added to the FPC as a requirement for forest management. The challenge for this project is to provide clear evidence that the current approach to forest management on fans is costly in terms of financial losses to structures, roads and plantations as well as unnecessarily impacting fish habitat and other environmental attributes. The project must also provide a cost-effective method for operational forestry staff to determine hydrogeomorphic hazards on fans and provide guidance for sustainable forestry practices. 3 Fortunately operational foresters are recognizing the problems with forestry activities on fans. Preliminary work on the topic, including an extension note, management strategies, field training sessions and consulting, have been welcomed (Wilford 1998; Wilford 2002a). There is growing recognition that it is cost effective in both the short and long-term to identify hydrologic and geomorphic hazards on fans and plan forestry operations accordingly. As a testament to this increasing awareness, support for this project was obtained from several forest companies, the BC Ministry of Forests, Forest Renewal BC, Fisheries and Oceans Canada, and the BC Ministry of Water, Land and Air Protection. In applied science, extension of research results is essential if the knowledge is to become operational. Some view the need for extension as an admission that the process of research was flawed. By this they mean that end users should be involved from the beginning, so that extension is not necessary. Extension was built into the project at an early stage by involving operational resource people (Appendix A). However, it is not possible to include all possible end users, so extension products (a field training course and manual of management strategies) were developed and delivered during the research project. Six presentations were made at conferences during the project and two technical articles were written (Wilford et al. 2002b, 2002c). In addition, throughout the research project, efforts were made to incorporate site and watershed features that can be identified by operational forestry staff at very little cost. 1.5 THESIS OUTLINE Definitions of fans and hydrogeomorphic processes are presented in Chapter 2. Chapter 3 presents a description of the study area. A forest stand-based method of determining the power and disturbance extent of hydrogeomorphic processes is presented in Chapter 4 and the study fans are classified in an eight-category scheme. The classification scheme is explored statistically in Chapter 5 through the identification of unique watershed attributes for each category. Dendroecology was used to determine the frequency of hydrogeomorphic events on study fans. These data are analyzed using regression analysis in Chapter 6 to develop predictive equations 4 with watershed attributes as the independent variables. Forestry activities on fans are explored in Chapter 7. Specific pre-forestry aerial photographic features are identified that indicate high hydrogeomorphic hazards. The relationship between forestry activities and impacts are described. Chapter 8 summarizes the major findings of the study and presents recommendations for future research. 5 C H A P T E R 2 . FANS A N D H Y D R O G E O M O R P H I C PROCESSES 2.1 I N T R O D U C T I O N Fans are present in every climatic setting and have been the subject of considerable geomorphic research. While the focus of this project is on humid temperate fans, the broad geomorphic literature allows for a more complete appreciation of the processes affecting fans. Thus, the discussion that follows employs a broad range of climatic settings to define fans and the hydrogeomorphic processes that contribute to their formation and development. 2.2 F A N DEFINITION A fan is a landform whose surface forms a segment of a cone that radiates downslope from the point where a stream emerges from the confines of a mountain (Bull 1977). The plan view of a fan is generally that of a flattened cone, and the contours bow down-fan (Figure 2.1). Overall Figure 2.1. A clearcut fan with a classic flattened cone shape (Carrigan3). 6 radial profdes are concave and cross-fan profdes are convex. The point of emergence or top of a fan is referred to as the apex. Three different zones are recognized on a fan surface: proximal or close to the apex; medial or in the mid-position; and distal toward the outer edges of a fan (Mukerji 1990). Fans are composed of sediments that originate in a source-area watershed. Transport of sediments to the fan apex is by fluvial processes or debris flows, or a combination of the two. Interestingly, fans are commonly and incorrectly referred to as "alluvial fans" regardless of the sediment transport process (Lecce 1990). As a stream emerges from the confines of a mountain and enters the fan apex, conditions can change resulting in the deposition of sediment. The primary change is a lack of confinement that allows the stream to become wider and shallower, reducing stream competency (Bull 1977). Similarly, the lack of confinement allows debris flows to spread out, dissipating both energy and water. A secondary change, in some situations, is a decrease in channel gradient that again reduces the competency of the stream or the energy of the flowing debris. Impediments to flow may also be present on a fan surface. These may include vegetation, woody debris, boulders, or topographic irregularities. Slope gradients from the apex to the toe of fans can range from 2 to 20 degrees. Fans steeper than 20 degrees are referred to as cones or alluvial [sic] cones (Bull 1977). However, in British Columbia, cones are distinguished from fans if the gradient is greater than 15 degrees (Howes and Kenk 1997). For the purpose of this project, fan-like landforms with slopes in the 15 to 20 degree range were considered to be fans if a stream channel was present. With distance from the apex, sediments on a fan characteristically become finer. This is due to water transport and sorting of the sediments. An exception can occur in fans that have an entrenched stream channel. In these situations, coarse sediments are transported further down fan than on the original fan surface. 7 As discussed later, fans develop characteristic features depending on the hydrogeomorphic processes that are, or historically were generated from their contributing watershed, and on the geologic materials present in the contributing watershed. Characteristic fan features include size, gradient, texture of soils, degree of sorting, surface roughness, and vegetation. 2.3 T Y P E S OF F A N S Two major types of fans are recognized on the basis of the transporting flow and the type of sediment being deposited (Schumm et al. 1987). Fluvial fans are formed by water transport and characteristically have sorted sediments in discrete layers (Figure 2.2). The gradient of fluvial fans is generally less than 4 degrees (Jackson et al. 1987). Debris flow fans are formed primarily by mass wasting (debris flows and debris floods). Sediments are characteristically poorly sorted although there can be beds of sorted sediments as a result of alluvial action, particularly in distal areas (Figure 2.3). Debris flow fans have gradients ranging from 5 to 15 degrees. In British Figure 2.2. Characteristic sorting of alluvial sediments (Tableland). 8 Figure 2.3. Poorly sorted sediments of a colluvial fan (Skilokus). Columbia the terrain classification system uses the terms fluvial fans and colluvial fans, and bases the distinction on respective fan materials (Howes and Kenk 1997). A stream flowing from the contributing watershed is common to both types of fans. On fluvial fans, the stream is a true "alluvial stream" since its bed and banks are alluvial, or transported and deposited by fluvial processes. For clarification, "alluvial" and "alluvium" refer to the sediments deposited by a river or stream while " f luvial" refers to the process of running water (Whittow 1984). On colluvial fans, streams erode channels through debris flow deposits. 2.4 C O N T E M P O R A R Y F A N A C T I V I T Y There is considerable discussion in the literature regarding the temporal origins of fans. In some situations field evidence supports the position that fan origin is linked to late-glacial sedimentation episodes unrelated to modern conditions (Ryder 1971a, 1971b; Ritter et al. 1993). 9 In other situations, it is apparent that fans are actively growing (Beaty 1970); being dissected (Hunt and Mabey 1966); or in a steady-state, dynamic equilibrium with sediment supply and water (Denny 1965, 1967; Hooke 1968). From this we can conclude that there are three basic levels of fan activity: • Actively aggrading fans - These fans accrete through annual or periodic sediment deposition processes. While the present volume increments to the fan may be small compared to previous conditions (e.g., immediate post-glaciation), they are sufficient to modify the fan surface, including vegetation and topographic expression. • Degrading fans - These fans originated during previous periods of accretion. Currently the watersheds are sediment-supply limited and the stream channels are entrenching. Characteristically, the trenching begins at the apex and the situation is referred to as a "fan-headed trench" (see Section 6.4.5). Reworking of sediments on the fan by entrenching can lead to radial extension of the fan (Bull 1964a, 1964b). The reworking can be by fluvial erosion or mass wasting (e.g., debris slides from the channel banks leading to debris flows down the entrenched channel as described by Osterkamp and Hupp 1987). • Fans in dynamic equilibrium - The stream channel on these fans is able to transport sediments across the fan. The volume transported may be high or low, but it matches the transport capability of the stream channel ("steady state"). The fan size and volume is not appreciably aggrading or degrading and the fan surface is characterized by lack of disturbance by hydrogeomorphic processes. The level of fan activity is directly related to hydrogeomorphic processes and includes debris flows, debris floods, floods, snow avalanches, and channel erosion on the fan. As these processes are influenced by climate, over periods of time any individual fan could exhibit some degree of all three types of behaviour. 10 2.5 HYDROLOGIC AND GEOMORPHIC PROCESSES In very basic terms, fans are influenced by the supply of water, sediment, and where appropriate, woody debris. The influence is related to both the total amount as well as the relative ratios of water and sediment/debris ("sediment" includes bed and traction load, and texture of sediment is an important variable). High ratios of sediment to water will lead to aggradation on the fan while low ratios lead to degradation (or entrenchment of the channel). The primary hydrologic and geomorphic processes that influence the supply of water and sediments to fans originate in the contributing watershed: floods, debris floods, debris flows, and snow avalanches (although snow avalanches are only addressed to a limited extent in this study). On the fan itself, stream channels, overbank flows, high stage distributaries, channel avulsions and channel entrenchment are processes that lead to the redistribution of sediments and water. 2.5.1 Floods Floods are discharge events that modify channel morphology and usually lead to disturbance well beyond the established stream channel. There are two basic types of floods: hydrologic and hydrogeomorphic. Rainfall and/or snowmelt generate hydrologic floods. Hydrogeomorphic floods are caused by a rapid release of water such as moraine, snow avalanche or debris flow dam ruptures, glacial lake outbursts, beaver dam failures and downstream dilution of debris flows and debris floods. A hydrogeomorphic flood can exceed a design streamflow flood (e.g., 100 year return period) from a watershed by a factor of up to 100 (Jakob and Jordan 2001). Actively growing fans are influenced by hydrologic floods, and may be subjected to hydrogeomorphic floods on a periodic basis. Stream channels strive to establish equilibrium between the dominant discharge and load by adjusting the hydraulic variables (e.g., channel width and depth, velocity, roughness, and water slope). The "dominant discharge" has been defined as the flow at which most of the sediments are transported leading to a characteristic morphology (Wolman and Miller 1960). Dominant 11 discharge has been determined to be bank-full discharge with a recurrence interval of 1.1 to 2.0 years (Wolman and Miller 1960; Kilpatrick and Barnes 1964; Dury 1973). This means that on average in a stream channel that is in a "quasi-equibrium state" (Leopold and Maddock 1953), the stream leaves its channel once every two to three years. However, a closer inspection of bank-full discharge by Williams (1978) indicates that there is not a common recurrence frequency, and the range can be from 1.01 to 32 years. In addition, the morphology of small streams can be influenced by large woody debris (see Section 2.6) to the extent that it is the wood rather than flood flows that create the "characteristic morphologies" (Gomi et al. 2002). One of the characteristic morphologies associated with large woody debris is a wide, shallow channel, which can be subject to very regular over-bank flows in some reaches (e.g., several times a year). "Flooding" in alluvial streams is defined as the situation where the stream channel no longer confines flow, resulting in over bank discharge of water and sediment. Given the previous discussion, it follows that overbank floods occur on most streams on a relatively frequent basis. Several factors are at work on actively growing alluvial fans that favour flooding on an even more regular basis. Actively growing fans are characterized by annual or periodic delivery of sediments and debris to the fan. This represents a situation of "oversupply" relative to flow discharge. As a result, there is either localized or widespread aggradation of sediments and/or debris jams in the stream channel. Thus, on a periodic basis the stream is in a disturbed state. A period of time is required for a stream channel to recover from a major disturbance and regain its equilibrium and thus characteristic morphology (Wolman and Gerson 1978). During this period all or part of the stream on the fan will be shallower and more prone to overbank flows, subjecting the fan to either localized or widespread flooding. The extent of flooding is worthy of further exploration. The techniques developed for floodplain hazard mapping do not apply to fans due to the unconfined nature of the stream channel. Normal backwater profile computational procedures are not applicable (Kellerhals and Church 1990). In addition, when flooding begins, there is a possibility that the whole or major portions of the 12 channel may avulse or shift laterally (see Section 2.4.4). Avulsions are usually much more severe hazards than normal flooding on fans (Kellerhals and Church 1990). Climatic events and watershed characteristics influence floods (Table 2.1). The magnitude of hydrologic floods is related to the severity of the climatic event and antecedent conditions. Hydrogeomorphic floods are generally watershed specific due to a range of factors such as achieving a threshold of a triggering mechanism for a debris flow with subsequent dam formation and rapid breaching. Because of the need for a triggering mechanism, there is generally a poorer correlation between hydrogeomorphic floods and climatic events. The characteristic flood deposits on fans are bars, sheets, and splays (Table 2.2 and Figure 2.4). Textures of the sediments can range from silt through cobbles. Sorting may be absent (massive) or present (vertical and horizontal with cross stratification). The overall relief of flood deposits is T a b l e 2.1. Watershed characteristics related to floods. F l o o d type W a t e r s h e d c h a r a c t e r i s t i c s f a v o u r i n g floods C l i m a t i c c o n d i t i o n s H y d r o l o g i c floods Storms track (move) down the basin. Basin Morphometry - small area, high drainage density, high basin magnitude (Shreve 1966, 1967 - number of first order streams), high drainage density, high relief, high ruggedness number (Melton 1957 - basin refieCsquare root of basin area), equidimensional basin shape. Soils - shallow or rapidly drained - either as subsurface stormflow or overland flow (Kirkby and Chorley 1967). Compacted soils such as roads and landings. Vegetation - sparse/limited in extent. Bedrock - very taw hydraulic conductivity. Rain-on-snow - rain on a melting snowpack with strong winds. Rain - Intense short duration rain with wet antecedent conditions Rain - Long duration, moderate intensity rain. Solar with advection - Warm windy conditions with sun on a melting snowpack H y d r o g e o m o r p h i c floods Lakes - Presence of lakes damned by glaciers, beavers, or moraines. Dams - High probability or history of dams being formed by mass wasting into confined valleys. History - of debris floods or debris flows. Same as above but a triggering mechanism is required such as a dam rupture of debris flow. 13 low unless log steps have formed behind woody debris (Section 2.6). Evidence on a fan surface of a hydrogeomorphic flood is variable, ranging from debris flow depositional features (Section 2.4.2) to extensive water flood features. 2.5.2 Debris flows Debris flows are rapid mass movements in which soil, rock, water and organic debris move together as a single viscoplastic body (Johnson 1970; Iverson 1997) (Table 2.2). This study Table 2.2. Classification and characteristics of flow processes and deposits (after: Costa 1984, 1988; Pierson and Costa 1987; Smith 1986; Wells and Harvey 1987; VanDine 1985; Hungr et al. 2001). Characteristics Flood Debris flood Debris flow 1. Flow Fbw type Fully turbulent Partly turbulent (at afl times, however high sediment bad dampens small eddies) to laminar Laminar at time of deposition but may be turbulent on steep sbpes Relative peak discharge 1 1-2 5-40 Sediment concentration 1-40% by weight 0.4-20% by volume 40-70% by weight 20-47% by volume 70-90% by weight 47-77% by volume Sediment support mechanism Turbulence, electrostatic forces Turbulence, grain dispersive pressure, buoyancy. Matrix strength (cohesbn and structural support), grain dispers. press., buoyancy. Sediment concentration profile Non-uniform (Solids and water are separate components of flow) Non-uniform to uniform (Solids &water are sep. components of flow) Uniform (Solids & water move as a single viscoplastic body) Bulk density 1.01-1.33 g/cc 1.33-1.80 g/cc 1.80-2.30 g/cc Shear strength 0-100 dyne/cm2 100-400 dyne/cm2 >400 dyne/cm2 Fluid type Newtonian Newtonian to Non-Newtonian Non-Newtonian Viscoplastic (coubmb-viscous and Bingham-plastic models) Viscosity 0.01-20 poise 20->200 poise » 2 0 0 poise Fall vebcity (% of clear water) 100-33 33-0 0 14 Table 2.2. Continued. Characteristics Flood Debris flood Debris flow 2. Deposits Mode of deposition Grain-by-grain, dominated by traction processes Rapid grain-by-grain aggradation from both suspension and traction En masse Stratification Massive or horizontal stratification (with cross stratification) None or horizontal stratification None Grading Variable: as a result of sequential processes rather than a single process Frequently distribution normal graded (coarse on bottom, fine on top) None; reverse; reverse to normal, coarse-tail normal Sediment characteristics and texture Clast-supported with an open framework or distinctly finer grained matrix of infiltrated sand; rounded clasts; wide range of particle sizes; sorting from front to tail; bmax -< 100mm to >200mm Ckst-supported, with predominantly coarse sand, moderate to poorly sorted, bmax typically < 180mm Matrix-supported; rarely clast-supported; very poor to extremely poor sorting; extreme range of particle sizes; bmax 60-230 mm, may contain megac lasts -4000mm Clast long-axis (A) orientation; imbrication Always perpendicular to flow; usually well imbricated. Large cobble to boulder; usually perpendicular to flow. Pebbles to small cobbles; usually parallel to flow. Weak imbric. and collapse packing. Variable, based on location within flow; parallel to flow is most prominent; weak to no imbrication. Landforms and deposits Bars, fans, sheets, splays, channels have large width-to-depth ratio Similar to water flood Marginal levees, term, bbes, trapezoidal to U-shaped channel focused on channelized or valley-confined debris flows (Innes 1983). Sediment entrainment is irreversible; water and solids move at the same velocity. Solids may constitute 70-90% by weight of the flow mass. In forested terrain, large organic debris can constitute up to 50% of a debris flow (Swanston and Swanson 1976). Debris flows are powerful events that exert enormous forces on objects in their flow path. The impact forces can reach several thousand newtons per square metre (Takahashi 1981). The 15 Figure 2.4. Geomorphic and sedimentologic features of debris flows (DI and D2), debris floods (Tl) and alluvial floods (SI, S2 and S3) (Wells and Harvey 1987). 16 magnitude of the impact force is proportional to the mean velocity squared (Watanabe and Ikeya 1981; Costa 1984). This is why debris flows can be so destructive when rapidly moving down a channel or across a fan, and yet, due to low velocities near the edges of the flow, can "wrap" themselves around small vegetation and fail to scar tree bark or damage vehicles (Costa 1984; VanDine 1985). Debris flows also exert significant shear stress, causing scour in the stream channel (Costa 1984). Evidence of debris flow scour in a stream channel can remain for decades: a simplified channel lacking cross-channel debris, scour to bedrock, deciduous riparian vegetation, abrasion of trees (particularly along channel bends) and debris jams (in confined channel reaches, on the floodplain or on the fan). However, in humid temperate forests, the forest cover may obscure aerial photographic identification of all or some of the components of a small- and moderate-sized debris flows: the initiation zone, transport zone and deposition zones (Wilford and Schwab 1982; Sterling 1994). Thus, forest cover may significantly bias regional aerial photographic inventories of debris flows. Valley-confined debris flows in humid temperate forests characteristically have three forms of initiation (Takahashi 1981). The most common in BC is an open slope failure (debris slide or debris avalanche) that enters the stream channel, gains moisture and proceeds down channel as a debris flow (Sidle et al. 1985; Jordan 1994). The two other forms of initiation occur during a major runoff event with the mobilization of sediment and debris stored in the stream channel, or the rupture of a debris jam. All three forms of initiation are stochastic; however the latter two have the deterministic component of the build-up of material in the channel. While climatic events are essential (Septer and Schwab 1995) other conditions must be present for either an open slope failure or mobilization of channel sediments (Dagg 1987; Miles and Kellerhals 1981). The key features are unstable sites that lead into gullies and sufficient materials in a gully to release as a debris flow. 17 Central to debris flow initiation is erosion-prone terrain, including both bedrock and surficial materials. This terrain must be "connected" to gullies, meaning that the sites should be either in the gully (headwalls or sidewalls) or situated in a position that would allow an open slope failure to enter a gully. High rates of mass wasting are associated with highly jointed and strongly fissured volcanic bedrock (Eisbacher and Clague 1984; Evans 1984; Schwab 1983). Mass wasting is also associated with competent plutonic and metamorphic rocks; however, the size of failures are generally smaller because it is the shallow surficial materials on steep slopes that fail (Innes 1982; Schwab 1983). Surficial materials subject to mass wasting or sediment supply to channels include thick, wet tills and unvegetated materials (colluvial, fluvial, and morainal) (Bovis and Dagg 1988). As noted previously, climatic events are essential for debris flow initiation. The events are required to build up sufficient pore water pressure on unstable slopes or lead to significant streamflows in gullies. The events may include rain, rain-on-snow and snowmelt, but can also include non-climatic events such as dam ruptures (Clague et al. 1985). Caine (1980) presents a threshold relation for predicting shallow, rapid failures using intensity and duration of rainfall. Church and Miles (1987) explored the meteorologic conditions associated with debris flows in southwestern British Columbia and concluded that Caine's method did not apply. Key reasons included antecedent weather (soil moisture), snowpacks and difficulties in extrapolating precipitation data from valley bottom stations to remote, high elevation debris flow initiation sites (a common challenge in British Columbia). However, Sidle et al. (1985) used a range of case studies to explore the basic assumptions of Caine's threshold approach and concluded that it can be a useful general predictor for the occurrence of shallow rapid landslides. In addition to the presence of erosion-prone terrain, several other watershed characteristics are associated with debris flows. Jackson et al. (1987) used Melton's ruggedness number to identify debris flow prone watersheds (R = H A 0 5 , where H is basin relief and A is basin area) (Melton 1965). They found that ruggedness numbers greater than 0.25 to 0.3 are associated with debris 18 flow prone watersheds. VanDine (1985) observed that channels must have gradients of at least 25° for debris flow initiation and that debris flow deposition begins at approximately 15°. Beyond a certain watershed size, it appears that the amount of water supplied by a watershed leads to dilution of debris flows and the evolution into debris floods or water floods. This size is variable due to climate, geomorphic characteristics, surficial materials and geology; however it appears to be in the range of 0.7 to 10 square kilometres (Thurber Consultants 1983; VanDine 1985). Church and Miles (1987) identified that direct exposure of a watershed to storms is a key to explaining concentrations of debris flows in a landscape. Deposition of debris flows occurs due to a decrease in gradient or a widening of the channel or both (VanDine 1985), or at abrupt tributary junction angles (Benda and Cundy 1990). Deposition can occur in low gradient channel reaches, characteristically as debris jams in temperate forest situations (Hogan et al. 1998). The deposition of debris flows on fans results in a range of specific features or signatures. Foremost is fan slope, which is characteristically > 4°. Gradient is a key factor in differentiating fluvial versus debris flow fans (Jackson et al. 1987). Sediments comprising a debris flow fan are poorly sorted, although different depositional events may give the appearance of stratification. Several factors may cause some degree of sorting. Finer material is commonly transported down the creek and deposited after the initial debris flow. This finer material is referred to as "afterflow" (VanDine 1985). Fluvial action after an event can lead to a reworking and sorting of sediments. Such reworking is particularly evident in distal areas of a debris flow fan. Debris flows create distinctive features on the fan surface (Figure 2.4). As a debris flow progresses down a fan, dispersive forces and differential particle velocities cause migration of large particles to the margins of the flow (Suwa 1988). Lateral areas of the flow mass are pushed to the sides and sheared from the mass as the rigid plug passes through the middle of the flow, leaving levees which are commonly studded with large boulders (Sharp 1942). When debris flows stop, the strength of the material or concentrations of coarse clasts at the margins of the 19 flow allows for the formation of steep fronts and sides, creating terminal lobes (Costa 1984). The height and steepness of the terminal lobes depends on the nature of the sediments and the water content of the debris flow (Wells and Harvey 1987). The characteristics of the terminal lobe can also be strongly influenced by large woody debris (Swanston and Swanson 1976). The lateral and terminal lobes of debris flows create a stepped or segmented fan profile. A characteristic of levees and lobes is a uniform distribution of sizes up through boulders in a matrix of fine-grained sediments (diamicton) (Costa 1984). Scattered large boulders on the fan surface have also been used as evidence of debris flows (Jackson et al. 1987) although rockfalls should be ruled out as a source. 2.5.3 Debris floods There is a continuum from water floods through debris flows, with debris floods being an intermediary (these events are also referred to as hyperconcentrated flows) (Table 2.2). Where debris flows have been observed closely, some evolve into debris floods as they proceed downstream due to addition of water from tributaries (Pierson and Scott 1985). The condition of irreversible sediment entrainment separates debris flows from water floods and debris floods. The latter two can vary their sediment load readily by deposition and erosion, but a debris flow cannot selectively deposit any but the coarsest particles. Thus, a debris flow cannot be converted into a water or a debris flood by deposition, only by the addition of water (Hooke 1967). The term "debris flood" emphasizes rapid transport and deposition of sediment during high sediment and water discharge events (Hungr et al. 2001). There is a subtle difference between debris flood deposits and water flood deposits, primarily in the orientation of pebbles to small cobbles, lack of cross stratification, grading of sediments and nature of the gravel framework (Smith 1986). In addition, Smith (1986) notes that debris flood deposits are characteristically deeper than water flood deposits. Unlike debris flows, deposition is not en masse. 20 The reason for the similarity in deposits between debris floods and water floods is that most transport of coarse sediment occurs during periods of high discharge. However, not all floods involve high concentrations of sediment. Debris floods should be viewed as an extraordinary form of flood in which an extremely large volume of sediment with a wide range in grain size is moved and deposited in a short period of time (Smith 1986). During the event, the high sediment concentrations lead to additional sediment support mechanisms (buoyancy, grain dispersive pressure) relative to water floods (Table 2.2). There are several reasons why identification of debris floods is important in this study. The first is the extent and depth of materials deposited on a fan surface. The extent can be similar to that of water floods; however the depth can be considerably greater. Conversely, the extent can be much greater but generally not as deep as debris flow deposits. The second reason has to do with the possibility that debris floods, because of their lower sediment concentration, could move down a channel beyond the point of deposition for debris flows. As the debris flood proceeds down the channel and across a fan, the resulting impacts to forest cover could be high (stand clearing) or low (spreading sediment under the canopy). Low-power events may not be characteristic of debris flows. 2.5.4 Avulsions Channel avulsions are the sudden lateral shifting in position of stream channels (Allen 1965). Avulsions can be common processes on fans. Even if avulsions are uncommon on a fan, several features of fans and streams on fans tend to favour their occurrence. The position of a stream on a fan with regards to cross-sectional elevation is a key feature for channel stability. Characteristically, the stream is in a high position. This can be related to gradual aggradation of the fan surface from over-bank deposition of fine sediments (levee deposits close to the channel, thinning in depth with distance from the channel) and bedload 21 deposition in the stream channel (Kellerhals and Church 1990). A factor that could aid in levee formation is dense riparian vegetation (Section 4.5). Stream channels on fans are unconfined, meaning that their banks are composed of unconsolidated material. Being unconsolidated, they are prone to erosion by the stream, and hence avulsions. In situations where the stream is not deeply incised into the fan, woody debris from blowdown or small debris jams can easily block the channel. Debris flow deposits or bedload deposition can also form channel blocks, even in incised channel reaches. These blockages in the channel can set the stage for avulsions. The continuity of channel incision is critical. It is possible to have a stream on a fan that is incised in some reaches of the fan, and unconfined in other reaches. Channel avulsion is, in essence, a process waiting to happen and a shallow channel exacerbates this possibility. As discussed in Section 2.4.1, an alluvial stream experiences over-bank flow on a relatively frequent basis. Once the flow leaves the main channel there is a reduction in stream competence or ability to carry bedload. If this loss in competence becomes significant, bedload is deposited in the main channel and plugging of the channel occurs (Kellerhals and Church 1990). As a result, more flow goes over-bank, and what began as a benign (usually depositional) process of over-bank flow can quickly lead to the erosion of a new channel. With an avulsion, a new channel is generally created across the fan surface. The exact location of an avulsion and the subsequent new stream course is difficult to determine prior to an avulsion, although several factors should be considered. Shallower stream reaches appear to be more prone to avulsions, although entrenched reaches can have an avulsion given the deposition of significant bedload. Once an avulsion occurs, stream flow will follow the line of least resistance - topographic lows (e.g., old channels). A confounding factor is that any obstruction to flow (e.g., large woody debris) could lead to a re-routing of the new channel. Development of a new 22 channel can be determined by a series of random events and the exact route is often quite unpredictable (Kellerhals and Church 1990). 2.5.5 Channel entrenchment Channel entrenching is the process by which the channel on a fan incises into the fan surface (Figure 2.5). Characteristically this process occurs at or near the fan apex and the descriptive term is "fan-headed trenching" (Bull 1964a, 1964b). Trenching can be significant with incision up to 50 metres below the fan surface (Bull 1977). The point at which the trench emerges on the fan surface is referred to as the "intersection point" (Hooke 1967). There are two schools of thought regarding the causes of trenching on alluvial fans (Weaver and Schumm 1974). External forces such as altered base levels, tectonics and climate change are the classical reasons. However, it has been shown through laboratory experiments and field Figure 2.5. A n entrenched stream channel on a debris flow fan (Wan). 23 investigations that a cyclical pattern of fan evolution could also be a cause (Weaver 1984). The pattern involves cycles of apex sedimentation, increase in slope, incision, lateral channel migration and backfilling. Channel entrenchment on fans has both geomorphic and ecological implications. A fan-headed trench channels flow (water, debris flows and debris floods) out of the contributing watershed, transmitting materials down-fan to the end of the trench before deposition occurs. In addition, the banks of the entrenched channel can experience mass wasting leading to channel plugging and on-fan initiated debris flows (Osterkamp and Hupp 1987). The result of the transmission of sediments is an elongation of the fan, with coarser sediments being transported further down-fan and being deposited over finer sediments. Trenching can have a significant effect on the water relations on the upper portions of a fan. Broadcast flows over the fan surface become rare and the groundwater table is lowered. As a result, forest site productivity can decline significantly, particularly if the fan is composed of coarse textured materials (see Table 4.3 and Section 4.5.1). I have observed trenching that began some distance below the fan apex with the knickpoint being either bedrock or large boulders or, in some cases, woody debris. While these observations may be "special cases", it is my perception that the point of trench initiation can have significant implications for fan surface stability with regard to hydrogeomorphic processes. For example, if a trench begins some distance below the fan apex, a debris flow may deposit material on the upper fan segment causing a channel avulsion. The resulting disturbance would thus influence the upper fan surface. If the trench begins at the fan apex, the debris flow would continue down the channel, depositing material at the intersection point, leaving the upper fan unaffected. A further interesting point is the role of woody debris as a hydraulic control structure in forested streams (Heede 1972; Marston 1982; Gomi et al. 2002). The extent of trenching on forested fans due to these structures could be limited relative to non-forested fans in more arid regions. A parallel consideration is the role of large boulders, a common element on some debris flow fans. 24 It is possible that boulders provide stable hydraulic structures in fan streams and thus limit the degree of channel incision (Church in prep.). 2.5.6 Groundwater hydrology of fans Given the preceding discussion on the hydrogeomorphic processes that form and influence fans, it is apparent that fans can have a range of sediment types, and any individual fan may have a complex interbedding (vertically and horizontally) of sorted and unsorted sediments with a wide range of textures. The net result for groundwater is an aniostropic and heterogeneous aquifer with complex flow patterns (Freeze and Cherry 1979; Creuze des Chatelliers et al. 1994). This section focuses on some of this complexity and other features of fans related to groundwater movement. Fans are moderately porous landforms and as a result streams that flow over fans characteristically lose water through their bed and banks (influent streams) as well as on the fan surface during over-bank flooding (Bull 1977). The movement of groundwater in a fan is determined by the hydraulic conductivity of different layers, although the general flux pattern is downslope and radiating outwards. Since fans characteristically show rapid down-fan decreases in grain size (Nilsen 1982), there tends to be a decrease in hydraulic conductivity in this direction. An exception to this decrease in conductivity can occur due to a higher prevalence of water sorting in the medial to distal fan locations found on debris flood and debris flow fans (Nilsen 1982). The interbedding of different sediment textures can lead to perched or confined aquifers, with the possibility that both aquifer types are present on any given fan. The implications of this situation are that some areas of the fan may have very limited groundwater while other areas could have seepage zones or streamflow from return groundwater flow. 25 Trenching, either as a natural process or related to land use activities, can have significant groundwater implications on a fan (Creuze des Chatelliers et al. 1994). Given the porous nature of fans, a trench serves to either drain groundwater and/or reduce recharge. As noted in Section 2.4.5, the implications of a lower groundwater table for forest growth can be significant. With less surface water losses to groundwater, the implications are that more water in the channel will facilitate greater sediment transport (including erosion) in the channel (Bull 1977). Reduced groundwater recharge through natural channel trenching can affect both the total amount of groundwater moving through the fan as well as locations of discharge (road ditchlines can have the same effect, although generally the ditchlines are not as deep as natural channel trenches). This can have consequences for low flow in the fan stream as well as for groundwater-fed surface water bodies along the toe of the fan. These water bodies are characteristically high-valued fish habitat due to their moderate winter water temperatures and relatively constant flows (Dave Gordon, Fisheries Biologist, Triton Env. Consultants, Terrace B.C., pers. com., 2000). Situations have developed in the Terrace area where groundwater-fed surface water bodies along the toe of a fan have lost significant flow (and thus fish habitat) due to stream channel entrenchment on the fan. Given that the principal source of groundwater recharge on a fan is the stream, annual variations in streamflow can have a significant influence on the volume of groundwater both within and between years. 2.6 CONCLUSIONS There is a large body of knowledge regarding fans and hydrogeomorphic processes. This knowledge has been applied to individual fans to identify hydrogeomorphic hazards (Thurber Consultants 1983) and in a general way through the identification of characteristic site features (Kellerhals and Church 1990). There are three significant gaps in the knowledge related to conducting forestry activities on fans. There is a need to identify hydrogeomorphic hazards on fans at the landscape level for forestry planning (i.e., a classification scheme for fans based on their watersheds). Site specific features need to identified that will give guidance to foresters 26 regarding zones on fans that require special management attention (i.e., a classification of fans based on forest stands and site features). It is necessary to have a coherent body of operational knowledge regarding appropriate forestry activities on fans given the observed variations in geomorphic and hydrologic attributes. These gaps are the specific focus of this thesis. 27 C H A P T E R 3. DESCRIPTION OF T H E S T U D Y A R E A The study area is in west central British Columbia, within the Prince Rupert Forest Region (Figure 3.1). Study fans lie across a broad geographic area, between 53° 46' and 55° 43' north latitude and 126° and 129° 10' west longitude. Figure 3.1. Location map of the study area. The western fans are within the Coastal Western Hemlock (CWH) biogeoclimatic zone and their watersheds have Mountain Hemlock (MH) and Alpine Tundra (AT) at higher elevations (Banner 28 et al. 1993). The central fans are within the Interior Cedar Hemlock (ICH) biogeoclimatic zone and their watersheds have Englemann Spruce - Subalpine Fir (ESSF) and AT at higher elevations. The eastern fans are within the Sub-Boreal Spruce biogeoclimatic zone (SBS) and their watersheds have ESSF and AT at higher elevations. Some of the eastern fans are totally within the ESSF. The study area lies within the Western and Interior Systems of the Canadian Cordillera (Holland 1964). The Kitimat Ranges are within the Coast Mountains of the Western System, and consist of granitic mountains, characteristically round-topped and domed because they were overridden by large Pleistocene ice sheets. The Interior System includes the Skeena Mountains, Nass Basin, Hazelton Mountains and the Nechako Plateau. This system is underlain chiefly by volcanic and sedimentary rocks and overall is less rocky and rugged than the Western System. The study area was last glaciated during the Fraser glaciation with ice retreat completed between 10,700 and 9,300 years BP (Alley and Young 1978; Clague 1984). The legacy of the glaciation is extensive morainal and glacial-fluvial deposits that dominate the landscape, masking much of the underlying bedrock (Runka 1972). Fans are a post-glacial feature in the study area reflecting paraglacial (Ryder 1971a, 1971b) and contemporary conditions (as discussed in this thesis). Long-term climatic stations are limited within the study area to valley bottom locations. Extrapolation of records to study watersheds can be problematic due to mountainous terrain. However, this information can provide an indication of extreme events even though the absolute amounts of precipitation may be uncertain. Septer and Schwab (1995) compiled a record of floods for the Prince Rupert Forest Region. The records include data from long-term stations plus newspaper reports and records from the Hudson's Bay Company. A series of stream gauging stations are operated by the Water Survey of Canada in the study area. Most gauged watersheds are very large compared to the study watersheds, making unit runoff calculations and even event dating problematic. However, the hydrometric data is useful in 29 describing the principal runoff regimes. Characteristically the western and central portions of the study area experience biannual peakflows. Spring snowmelt provides the highest volume of runoff, and in some years the highest peakflows. Fall rain or rain-on-snow events can produce significant peakflows as well as erosion events (debris avalanches and debris flows). The same biannual peakflows occur in the eastern portion of the study area, although in general the spring snowmelt peaks are significantly larger than the fall peakflows. 30 C H A P T E R 4. T H E S T U D Y F A N S : F O R E S T S T A N D S A N D H Y D R O G E O M O R P H I C P R O C E S S E S 4.1 INTRODUCTION This chapter presents an overview of the selection and characteristics of the study fans. The primary focus is to present a forest stand-based method for determining the power and disturbance extent of hydrogeomorphic processes, and documentation of the hydrogeomorphic role of forest stands. 4.2 FOREST STANDS AND HYDROGEOMORPHIC PROCESSES A forest stand is a spatially continuous group of trees and associated vegetation having a similar structure and growing under similar soil and climatic conditions (Oliver and Larson 1996). Structure is the physical and temporal distribution of trees and other plants in a stand. The distribution can be described by: species; vertical or horizontal spatial patterns; size of living and/ or dead plants or their parts, including crown volume, leaf area, and stem cross section; plant ages; or a combination of these features. Forest stands are influenced by biotic and abiotic factors including insects, pathogens, mammals, fire, floods, mass wasting, wind and snow. When these factors make growing space available for plants, they are considered disturbances. Disturbances are relatively discrete events that disrupt the stand structure and/or change resource availability or the physical environment (Pickett and White 1985). Disturbances can be roughly divided into two types based on the amount of overstory forest removed (Oliver 1981). Those that remove or kill all the existing trees above the forest floor vegetation are referred to here as major disturbances or "stand-replacing disturbances". Those that leave some of the pre-disturbance trees alive are referred to here as minor disturbances. Hydrogeomorphic processes produce a range of major and minor disturbances to forests on fans. High-velocity debris flows remove forest stands while low-velocity debris flows may have 31 limited effects even on shrubs (Costa 1984). Hydrogeomorphic processes may produce major disturbances along a stream channel, resulting in a corridor parallel to the stream with a distinctive forest stand (Swanson et al. 1998; Gomi et al. 2001). These processes may also produce minor disturbances on the fan surface through deposition of sediments and short-term flooding. However, if the deposits are sufficiently deep or flooding is prolonged, mortality of a stand would result, and they would be considered major disturbances. Two basic types of forest stands are created by disturbances: single-cohort and multicohort (Oliver and Larson 1996). A single major disturbance is a stand-replacing event that leads to the establishment of a "single cohort" or group of trees (generally referred to as an "even-aged stand" by foresters). Another disturbance may occur after establishment, but a new cohort may not develop if the residual trees reoccupy the released growing space and exclude a new cohort from initiating or surviving. If the single-cohort stand continues in time without a significant disturbance, it will progress through a series of developmental stages (Oliver and Larson 1996): • Stand initiation stage when plants become established on the disturbed site. • Stem exclusion stage when competition results in the expression of dominance and some trees/plants die out. • Understory reinitiation stage when individual trees in the overstory begin to die, opening the canopy and allowing more light to reach the forest floor. Foresters have generally recognized this as a time when the overstory "loses its grip on the site" (Table 4.1). During this stage new tree/plant species appear in the forest floor that are capable of living under low light intensity. • Old-growth stage when the trees that invaded immediately after a disturbance have all died, and the stand consists entirely of trees that grew upward through the deteriorating overstory. Stands that have relict trees dating to the original stand scattered amongst young trees that have grown into the overstory are referred to as transition old growth. 32 The time taken to proceed through the stages is dependent upon the tree species involved in the original stand and the climate and quality of the site, but can range from 100 to 700 years in the study area (Table 4.1). Table 4.1. Approximate ages associated with later developmental stages (west central British Columbia). (Source: Fowells 1965; Burns and Honkala 1990; A. Banner, Research Forest Ecologist, MOF, Smithers, pers. com. 2000.) Later understory Old-growth reinitiation/Early transition stage (years) old-growth (years) Aspen (Populus tremuloides) 80 100 Red alder (Alnus rubra) 80 100 Paper birch (Betula papyrifera) 100 125 Cottonwood {Populus balsamifera ssp. trichocarpa) 120 150 Lodgepole pine (Pinus contorta var. latifolia) 200 300 White spruce (Picea glauca) 250 350 Subalpine fir (Abies lasiocarpa) 300 350 Engelmann spruce (Picea engelmannii) 300 500 Amabalis fir (Abies amabilis) 350 500 Western hemlock (Tsuga heterophylla) 400 500 Sitka spruce (Picea sitchensis) 400 600 Western redcedar (Thuja plkata) 500 700 If a minor disturbance influences a single-cohort stand, surviving trees expand and newly initiating ones invade and expand. This is termed a multicohort stand, or a stand comprised of trees from two or more disturbances (generally referred to as an uneven-aged stand by foresters, but more accurately should be called a multi-aged stand) (Oliver and Larson 1996; O'Hara 1998). At some time following the disturbance, all growing space is reoccupied by existing trees and 33 newly invading trees are excluded and some existing trees are eliminated or are severely suppressed through competition. In this way, multicohort stands have stand initiation and stem exclusion stages. Barring another minor disturbance, the stand will progress through to the old-growth stage. As previously discussed, there are many biotic and abiotic factors that cause disturbance. It is generally possible to discern the causes of disturbance from field evidence (Appendix B). As noted in Chapter 2, hydrogeomorphic processes produce a range of characteristic field evidence on the fan surface. In addition, evidence of these processes may also be present within the contributing watershed. Table 4.2 outlines characteristic forest stand signatures of hydrogeomorphic processes. Vegetation on fans can influence deposition and erosion of water flood events. Once water flows overbank, velocities decrease and deposition occurs in the form of levees (Costa 1984). Coarse sediments are typically deposited close to the stream while finer sediments are transported further. This differential deposition of sediments leads to a notable aggradation along stream channels. Riparian vegetation reduces water velocities because it is an obstruction to flow (Sato 1991). Vegetation increases Manning's "roughness coefficient" ("n") of a channel by an order of magnitude, reducing velocity by an order of magnitude and thereby significantly reducing sediment transport capability (Ogrosky and Mockus 1964). As a result, on vegetated fans, it is possible that natural levees are somewhat higher than would be expected on arid fans. This could lead to higher cross-sectional positions of streams on vegetated fans since any aggradation of the channel would be compensated for by an increase in levee height. The result would be potentially fewer channel avulsions in a given time period, however once an avulsion occurs the new channel gradient could be expected to be steeper and erosion potentially more significant than would occur without riparian vegetation influence. 34 Table 4.2. Characteristic forest-stand signatures on fans and their watersheds caused by hydrogeomorphic processes (based on: Sigafoos 1964; Hupp 1983; Osterkamp and Hupp 1987; and unpublished personal field notes). H y d r o g e o -m o r p h i c p r o c e s s F o r e s t s t a n d s i g n a t u r e o n f a n F o r e s t s t a n d s i g n a t u r e i n w a t e r s h e d L o w - p o w e r Floods and debris floods On fan areas that do not sustain probnged flooding or deep burial by sediment: m u l t i c o h o r t - species mix can be wide, but is characteristically a dominant older coniferous stand cohort with one or more younger cohorts of deciduous or conifer species. Structural evidence of original stand is present (snags and down woody debris). On fen areas that sustain probnged flooding or deep burial by sediments: can range from bare sediments to m u l t i c o h o r t stand (primarily deciduous with a minor coniferous component). Structural evidence of the original stand is present as snags or downed woody debris. Tree bowls and understory vegetation stems are characteristically buried by fine sediments. Understory vegetation can range from absent (bare soil) to what would be expected in a given stand stage. Riparian forests abng upstream channels may have a significant deciduous component (s ing le-cohort) . There may be vegetative evidence of mass wasting including vertical hillsbpe strips ranging in devetapment from devoid of vegetation to s i n g l e - c o h o r t stands. H i g h - p o w e r Fbods, debris floods and debris flows Signature stand type: 1. S i n g l e - c o h o r t - The original stand has been removed with no evidence of the previous forest stand (except in piles). The present forest stand may range from stand initiation stage to early understory reinitiation stage. A multicohort stand may be found abng the margins of the single-cohort stand. 2. M u l t i c o h o r t - The original single-cohort or multicohort stand has the appearance of being thinned (e.g., bgged but with no stumps) with the younger stand members established on sediments from an event. Structural evidence of original stand is not present or is in piles of woody debris. Trees growing on the margins of the disturbed area are characteristically buried by sediments. Understory vegetation can range from absent to what would be expected in a given stand stage. There may be vegetative evidence of mass wasting including vertical hillsbpe strips ranging in devebpment from bare soil to s i n g l e -c o h o r t stands. Riparian forests abng upstream channels range in devebpment from bare ground to s i n g l e - c o h o r t stands. N o h y d r o -g e o m o r p h i c p r o c e s s e s Barring other stand disturbance agents, old conifer stands: S i n g l e or m u l t i c o h o r t stand of conifers in the late understory reinitiation, early transition old growth or old growth stage. No or very limited stand indications of mass wasting. Riparian forests abng the upstream channel may have a minor deciduous tree component ( m u l t i -c o h o r t ) . 35 Stream channels are also influenced in several ways by riparian vegetation. Reduced streambank erosion has been attributed to roots (Smith 1976). This could tend to make channels deeper rather than wider. However, another influence of riparian vegetation is the input of large woody debris (LWD). This material can have several influences on stream channels depending on its orientation relative to flow direction. When LWD is at a right angle to the flow it serves to make streams wider and shallower by obstructing bedload movement and by limiting down-cutting (and potentially trenching) by acting as hydraulic drop structures (Keller and Swanson 1979; Thomson 1991). Shallower streams are more prone to over-bank flows and broadcasting sediments and debris during hydrogeomorphic events. When LWD is oriented parallel to flow it may either play no functional hydraulic role or may serve to limit streambank erosion, particularly on bends. As noted in Chapter 2, the deposition of debris flows occurs due to a decrease in gradient or a widening of the channel, or both. The role of trees in increasing flow resistance appears to be overlooked in much of the literature. However, Irasawa et al. (1991) conducted hydraulic model experiments to evaluate resistance of trees to debris flow impact and shear forces. They determined that forests reduced the velocity and concentration of debris flows. The result was significantly enhanced sediment deposition (doubled) compared to unvegetated areas. A key factor was inter-tree spacing or the density of forest stands. Trees may also play a role along the lateral edges of a debris flow where the velocity and hence impact force is weak. The role may enhance levee formation, in a manner similar to water floods. The root system on forested fans serve to provide a "reinforced fabric" that has been observed by the author to limit the extent of fluvial down cutting into the fan surface by running water. This appears to limit the opportunities for the development of new stream channels through either avulsions or overbank flows where there invariably is some degree of flow concentration. A similar reinforcing of soils by roots has been discussed related to slope stability (Sidle et al. 1985; Sidle 1991). However, there could be a significant difference with regard to the longevity of this 36 reinforcement on fans following tree death or logging. On slopes, it is the fine roots that decay and cause a reduction in the factor of safety within 5 to 15 years while new roots are becoming established. On fans, it appears as though the larger roots that are providing the majority of resistance to erosion. While these roots may remain functional for longer periods than fine roots, they would also take longer to re-establish. Large woody debris laying on a fan surface can cause the deposition of sediments carried by debris floods and floods (Wilford 1984). These terraces form a stepped or segmented fan profile at the micro-scale that is of totally different origin than the macro-scale segmented profile on arid fans. Specifically, these fan surface features along with woody debris in stream channels may lead to flood fans that are steeper than would be expected without woody debris (e.g., the 4° slopes identified by Jackson et al. 1987). Dense riparian vegetation has been observed to both contain streamflows and limit re-entry of diverted water back into the channel (T. Lisle, Research Hydrologist, USDA Forest Service, pers. com. 2000; D. Wilford, field observations). The process involves organic debris jamming against riparian vegetation, effectively forming a dyke. Streams contained within these dykes would have reduced over-bank flows, effectively increasing stream velocity and thus sediment scour and transport capability. Streams outside the dykes could lead to either greater deposition or potentially erosion (avulsions) on the fan surface. Sato (1991) describes a fan in Japan that was prone to overbank flows. To protect farmland and residential areas, "flood-control forests" were planted along the watercourse in 1905. Over time some of the forests were removed and homes built. In 1990, a highly destructive flash flood with mobile logs and sand caused severe damage, primarily in reaches without flood-control forests. Sato notes that the flood-control forests along the river prevented overflow of drifted logs and led to reduced water flow and sediment deposition. It was also apparent that avulsions did not occur in the forested reaches. 37 4.3 METHODS The choice of study fans followed consultation with staff from forest licencees and government agencies (Fisheries and Oceans Canada, BC Ministry of Forests, BC Ministry of Water, Land and Air Protection) who identified undisturbed fans in forest development plans and fans with forestry activities. Aerial photographs and field reviews were undertaken during selection of the study fans. Study fans were selected based on the following criteria: • Reasonable access (vehicle preferred but helicopters were used for several fans). • Providing a reasonable cross section of the hydrogeomorphic processes, power and disturbance extent combinations across the study area. • Providing a mix of fans that were undisturbed and that had forestry activities. • No or very limited forestry activities in the contributing watershed. • Availability of historic information regarding the forestry activities. • Forestry is the only land use activity on a fan (e.g., fans with a complex history of log-ging, gas pipelines, highways and other activities obscure the objective of focusing on forest land use practices). Fans were characterized based on topographic map measurements, aerial photographic interpretation, and field observations and measurements. Attributes noted included shape, area, degree of dissection, channel entrenchment and the presence of elevated fan surfaces. From one to seven days was spent in the field describing each fan. Evidence of hydrogeomorphic processes was described according to the classification presented in Chapter 2. Sediment deposits were used to differentiate floods, debris floods and debris flows. The slope measurements were made using a Suunto clinometer at the apex of fans. The power of hydrogeomorphic processes was determined in the field based on the character of forest stands. Two characterizing stand types were recognized. High-power events cleared swaths through a forest stand. The single cohort forest stands established in the swaths were described (ranging from absent to single cohorts 100-years old). Low-power events spread 38 sediment under a forest stand and did not result in the physical removal of trees. The resulting multi-cohort stands were described (ranging from absent to co-dominants). Description in both cases involved determining ages of the trees and dimensions of the cohorts. Stand volumes of the older cohorts were determined with variable plot (prism) cruising. The width of swaths cleared through a forest by high-power hydrogeomorphic events was measured in the field. Two classes of disturbance extent were established based on the width of cleared swaths flanking the fluvial channel. Disturbances > 20 metres wide were termed Stand Level (Figure 4.1 and 4.2). Disturbances < 20 metres wide were termed Site Level (Figure 4.3). The width criterion was established based on the ability to perceive disturbances to the forest cover on 1:20,000 aerial photographs. Forest cover maps (1:20,000) of fans were used as base maps for fieldwork. Observations were made on the influence of forest stands on hydrogeomorphic events, primarily sediment storage and maintenance of stream channel location. Sediment storage was classified as "log steps" where sediments were stored up-slope of downed logs (Figure 4.4). "Log retaining walls" was the term used to describe the situation where multiple horizontal logs (in total greater than 1 metre in height) supported by trees were storing sediments (Figure 4.5). "Woody dykes" was the term used to describe situations where trees supported individual logs or woody debris (in total less than 1-metre in height) along the edge of the channel (Figure 4.6). "Alluvial levee enhancers" described the situation where trees provided channel roughness that reduced water velocities, resulting in obvious enhancement of alluvial levee formation along banks. "Soil reinforcement" described situations where roots were protecting the soil mass from the channelled erosion of avulsions (Figure 4.7). Site features related to forest stands were described in the field: buried tree butts, tree holes, exposed adventitious roots, scars, and recent sediment deposits. Typically trees with no butt flares had been buried (Figure 4.8). Tree holes are cylindrical holes in the ground formed where 39 Figure 4.1. An aerial photograph of a 100-year old cohort following a high-power stand level disturbance flood (Tsezakwa). Photo 30BCC96053 #096 taken in 1996 at a scale of 1: ~ 11 000. Source: BC Ministry of Sustainable Resources Manage-ment (MSRM), Victoria, BC. 4 0 Figure 4.2. A stereoscopic aerial photographic pair of a recent high-power stand level disturbance from a debris flow (Kitsl). Photos BC 7781 #82 and 83 at a scale of 1: -19 000, and taken in 1977. Source: MSRM. 41 Figure 4.3. A young cohort growing on the sediments of a high-power site level disturbance debris flow (Trapline). Figure 4.4. Recently deposited sediments behind a log are forming a "log step" (Gosnell4). 42 43 Figure 4.7. Roots provide reinforcement to the soil mass, delaying the erosional effects of broadcast flows (Carrigan2). trees are buried and then rot (Figure 4.9). Erosion of sediment deposited around trees can expose adventitious roots (Figure 4.10). Scars on the upstream and lateral sides of trees indicate damage by events (other sources of damage must be considered) (Figure 4.11). Sediment deposits beyond the channel that are unvegetated or have limited accumulation of organic matter (LFH) are considered to be "recent deposits" (within approximately the past 25 years) (Figure 4.12). Dendroecology (also referred to as dendrochronology) techniques were applied to determine dates (years) of disturbances. Disks were taken from young cohorts to establish ages. Increment cores were taken from older trees. Wedges were cut from trees to establish dates of scars. The number of samples ranged from 20 to 40 per fan. Details on the methods and results are presented in Chapter 6. 44 Figure 4.8. An example of a buried tree. Note the lack of butt flare. (Shedin) Descriptions of forestry activities included characteristics of roads, drainage structures, riparian reserves, and logged areas. Details on methods and results are presented in Chapter 7. Reconnaissance helicopter flights were undertaken in watersheds to supplement aerial photographic interpretation and GIS attribute investigations (Chapter 5). There were several objectives related to the flights. Close inspection of sediment sources allowed an identification of which sources were producing sediments that could be transported by the stream or by 45 Figure 4.9. Tree holes are the result of rotting following deep burial of tree stems. This hole is two metres deep (Kits 1). Figure 4.11. Scars can provide accurate dates of hydrogeomorphic events. The scar on this tree pre-dates the fine textured sediments deposited on the tree by a recent event (Wan). hydrogeomorphic events. Riparian vegetation in the watersheds was inspected for evidence of hydrogeomorphic activity (narrow cohorts that were not identified on forest cover maps). 4.4 RESULTS The selection process resulted in 65 fans for the study. Ten fans had no forestry activities; of these nine were forested and one maintained in an early serai stage by snow avalanches. Forestry activities were present on 55 fans, ranging from a single road crossing to complete clearcutting of the forest stand (in two cases). Forestry activities are described in Chapter 7. Dendroecology sampling was undertaken on 58 fans (Chapter 6). Dendroecologic evidence of hydrogeomorphic activity within the past 100 years was present on all fans. Evidence of at least one characterizing event within the past 50 years was present on 55 fans (95%), and multiple events on 53 fans (91%). Fan shape ranged from the classic conical to oblong and confined. Fan areas ranged from 0.05 to 6 km2 and the associated watersheds range from 0.21 to 99 km2. The degree of stream channel confinement ranged from unconfined to entrenched for a major portion of the fan. Fan surface cross sections ranged from smooth to dissected. Elevated, inactive fan surfaces were observed on 13 fans and ranged from a major portion of the present fan landform to minor remnants along the adjacent hillslopes, elevated 2 to 10 m above the active fan surface. Forest site productivity on the elevated surfaces was significantly less than on the lower surfaces in nine cases (Table 4.3) (Figure 4.13). Forest productivity was similar in two cases and logging on two fans precluded observations. Hydrogeomorphic process, power and disturbance extent information for the 65 study fans is summarized in Table 4.4. Hydrogeomorphic, forest stand and site level information for the study fans is presented in Table 4.5. Disturbance extent class and width of disturbance are used in the classification of high-power events and are not relevant for the classification of low-power events. The widths of disturbance measurements are made in the natural or pre-logging condition (summarized in Table 4.6). The width of disturbance is the total distance cleared on 48 both banks, beyond the normal ("two-year return period") fluvial channel. Within the disturbed area there was a wide range of successional development from recently deposited sediments to 100-year old cohorts. The hydrogeomorphic role of stands was based on field observations; thus logged fans with no riparian forests do not have information (i.e., data from only 61 fans are presented in Table 4.7). Bank reinforcement by roots is not included because it was a common feature on all 61 fans because the immediate riparian forest was retained. Site level information related to sediment deposition from hydrogeomorphic events is summarized in Table 4.8. The slope of the fan surface at the apex of each study fan is presented in Table 4.5 and the data are summarized in Table 4.9. An overview of watershed size and relief associated with the hydrogeomorphic processes is presented in Table 4.10. A detailed analysis of watershed morphometric measurements and attributes is presented in Chapter 5. Table 4.3. A subsample of fans with elevated surfaces comparing forest cover on elevated and active fan surfaces (ecosystem and species abbreviations are presented in Appendix C). Forest cover Elevated fan surface Active fan surface Fan Ecosystem Tree species Av. height Volume Ecosystem Tree species Av. height Volume Gosnell6 ESSFmc 02 Pa Hm Bl 11m 76rrr7ha ESSFmc 05 Bl 26m 703m7ha Gosnell7 ESSFmc 02 Bl Pa Hm 10m 183rrr7ha ESSFmc 06 Bl 21m 810rrr7ha CP095-2 SBSmc2 01C PI 19m 260m7ha SBSmc2 05 S PI 26m 707mVha Sibola SBSmc2 02 PI 12m 116m7ha SBSmc2 06 Bl PI S 29m 770m3/ha 49 50 4.5 DISCUSSION The selection criteria and time available for field investigations (12 months during a two-year period) yielded 65 study fans. The objective of having a reasonable representation of hydrogeomorphic processes was achieved, with 16 flood fans (25%), 36 debris flood fans (55%) and 13 debris flow fans (20%). The representation in the different power and disturbance classes was reasonable, although only two high-power site level disturbance flood fans were sampled. Achieving an equal sample size for each process, power and disturbance extent category was difficult. All fans required field investigations to determine the category because predictive equations based on watershed attributes were unavailable (developing these equations was one of the objectives of the study). Given that a field investigation of a fan took at least one day, limitations on field time did not allow many inspected fans to be discarded from the study. The representation of fans by land use was good, with 10 natural fans and 55 fans with forestry activity ranging from a single road to complete clearcut of the forest stand. The discussion begins with a brief overview of the morphology of the fans and their watersheds and then focuses on the forest stand-based approach to describing hydrogeomorphic processes and the hydrogeomorphic role of forests on fans. Table 4.4. The number of fans in the different forest stand-based classification categories. Hydroge omorphic process High-power Stand level Site level Low-power Totals Floods Debris floods Debris flows Totals 7 18 8 33 2 13 5 20 7 5 12 16 36 13 65 51 Table 4.5. Hydrogeomorphic, forest stand and site level information for the study fans. Abbreviations: HG is hydrogeomorphic, LS is log step, LW is log retaining wall, LE is alluvial levee enhancement, WD is woody dyke, SR is soil reinforcement by roots, BT is buried tree butts, D is recent sediment deposit, EAR is exposed adventitious roots, S is scar, and TH is tree hole. Stand type Power class Dist extent class H G process Fan Disturb, width (m) Present H G role of stands Site features Slope apex (°) Alice 175 Log'd,LS,LE,WD,SR BT, D, EAR, S 2.5 Dasque 165 Logged (BT) 2.5 Flood Hunter 120 LS, LE BT, D 2 Sinclair 33 LS, LE, WD BT, EAR, S, D 2 Tszkwa 211 LS, WD, LE BT, S, D 3 Winfield 33 LS, LE, WD BT, EAR, S, D 2 Middle 165 LS, WD, LE BT, S, D 1.5 16Wmse 25 LS, LE / fogged BT, S 13 16Wmsw 20 LS BT, S 10 Big_Wdn2 27 LS, LE, WD BT, S 13 Crnberry 100 Logged/ LS, LE,WD BT 4 Compass 33 LS, LE, SR, WD BT, TH, S, D 3 Single cohort High power Stand level Skilokus Logged/ LS,LE,WD BT, S, D 3.5 Gosnell 8 24 LS, LE, WD BT, S, D 4.5 Ktwancool 50 LS, LE, WD BT, D, S 4.5 Debris flood 25 Ukit 33 LS, LE, WD BT, S, D, EAR 11 39 Ukit 20 LS, LE, WD, RW BT, S, D 18 3Copper 1 20 LS, LE BT, EAR, S, D 18 8McDnll 3 30 WD, LE BT 6 8McDnU 1 20 LS, LE BT, S, D 5 20 LS, LE, SR, WD BT, S, D 10 Sibola 40 LS, LE BT, S, D 4.5 Tettack 20 LS, WD, LE BT, S, D 9.5 PowerBne 23 LS, LE, RW BT, TH, S, D 9.5 W_Bottm 100 LS, LE, WD BT, EAR, S, D 4.5 52 Table 4.5. Continued Stand type Power class Dist extent class H G process Fan Disturb, width (m) Present H G role of stands Site features Slope apex (°) Big_Wdnl 24 LS, LE BT, D, S 7 Fernando 24 LS, SR, LE BT, S, D, EAR 19 Kits 1 30 LS, LE, RW, SR BT, TH, S, D 12 Stand Debris Kits 2 20 LS, LE BT, S, D 8 level flow Kits 4 20 LS, LE, WD BT, S, D 8 Rico 40 SR, L E BT, S, D, EAR 16.5 Wan 25 LS, LE, SR, RW BT, S, D 10.5 Z_Cascad 22 Early serai - 12 Flood Shelford 10 LE D 2 Ailport 15 LE, LS BT, D 1 8McDnU2 5 LS, LE D, BT 12 Herb 18 LS, LE, RW BT, S, D 1.75 Luno 15 LS, LE, SR, WD BT, S, EAR, D 4 Single High OMcDnll 15 LS, LE, WD BT, D, S 6 cohort power 18SKit 10 LS, LE, WD BT, S, D 3 19SKit 10 LS, WD BT, D 3 Debris flood Gosnell 7 10 LS, LE BT, TH, S, D 10 Site Mill 6 LS BT, EAR, D, S 10 level Miller 8 LS, LE, WD BT, D, S, EAR 3 22 Shedin 6 LS, RW, SR BT, S, D, EAR 7 Gosnell 1 8 LS, LE BT, EAR, S, D 9 Gosnell 4 15 LS, LE, SR BT, EAR, S, D 4.5 Gosnell 6 12 LS, LE, WD BT, S, D 7 3 Copper2 10 LS BT, S, D 14.5 Carrigan 1 6 LS, RW, SR BT, S, D 8 Debris flow Kits 3 6 LS, RW, SR, LE BT, S, D 11 Legate 15 LS, LE, WD BT,TH,EAR,S,D 22 Trapln 8 LS, LE BT, S, D 13 53 Table 4.5. Continued Stand type Power class DisL extent class H G process Fan Disturb, width (m) Present H G role of stands Site features Slope apex (°) 3 D LS, LE, SR BT, S, D 2 CP095-1 WD D 4 CP095-2 LS, WD BT,D 10.5 Flood M3 LS, LE BT, D 2 McKndrkl Logged - 3 Multi- Low N/A Newcmbl LS BT 6.5 cohort power Canyon N/A Logged/ LS, WD BT, S 3 Debris flood Carrigan2 LS, LE, SR BT, S, D 12 SprCmp LS, SR, LE, WD BT, D 12 Tableland LS, LE, WD BT, D, EAR 4.5 Carrigan3 Logged (BT) 11 Poplar Logged/ LS, L E BT, D, S 6 Table 4.6. Summary data on measured widths of disturbance in the forest canopy from hydrogeomorphic events. Disturbance extent class Measured disturbance width Minimum Maximum Mean Stand level 20 m 211 m 52.5 m Site level 5 m 18 m 10.4 m 54 Table 4.7. Summary of hydrogeomorphic roles of forest stands on fans with riparian forests (total number of fans is 61 due to clearcutting of riparian forests on 4 fans). Hydrogeornorphic role Number of fans As a percent of fans with riparian forests (61) Log steps 57 93% Alluvial levee enhancers 51 84% Woody dykes 30 49% Soil reinforcement 15 24% Log retaining walls 8 13% Table 4.8. Summary of site features related to sediment deposition from hydrogeomorphic events on fans with riparian forests. Site feature Number of fans As a percent of fans with riparian forests (61) Buried trees 59 97% Recent sediment splays 54 88% Scars 48 79% Exposed adventitious roots 16 26% Tree holes 5 8% Table 4.9. Stream channel gradients at the apex by hydrogeomorphic process. Hydrogeomorphic High-power P r o c e s s Stand level Site level Low-power Overall Floods Mean channel gradient 2.2° 1.5° 4.4° 3.0° Range in gradients 1.5° - 3.0° 1.0° - 2.0° 2.0° - 10.5° 1.0° - 10.5° Number of fans 7 2 7 16 Debris floods Mean channel gradient 8.4° 6.2° 9.1° 7.7° Range in gradients 3 . 0 ° - 18.0° 1.75° - 10.0° 4.5° - 12° 1 .75°- 18° Number of fans 18 13 5 36 Debris flows Mean channel gradient 11.6° 13.7° N/A 12.4° Range in gradients 7 . 0 ° - 16.5° 8.0° - 22.0° 7.0° - 22.0° Number of fans 8 5 13 55 Table 4.10. Watershed size and relief associated with hydrogeomorphic process. Hydrogeomorphic High-power Low-power Overall process Stand level Site level Floods Mean area Range Mean relief Range Number of fens 64 km2 39 - 99 km2 1.6 km 1.1 - 2.1 km 7 28.4 km2 19 -37 km2 0.48 km 0.46 - 0.50 km 2 6.2 km2 1.4 - 9.6 km2 0.7 km 0.4 - 1.1 km 7 34 km2 1.4 - 99 km2 1.1 km 0.4 - 2.1 km 16 Debris floods Mean area Range Mean relief Range Number of fens 7.1 km2 0.7 - 18 km2 1.3 km 0.9 - 1.7 km 18 8.5 km2 1.4 - 31 km2 1.2 km 0.8 - 1.7 km 13 2.5 km2 1.3 - 4.2 km2 0.7 km 0.5 - 0.8 km 5 7.0 km2 0.7 - 31 km2 1.2 km 0.5 - 1.7 km 36 Debris flows Mean area Range Mean relief Range Number of fens 1.5 km2 0.25 - 4.1 km2 1.0 km 0.7 - 1.4 km 8 0.9 km2 0.7 - 1.4 km2 0.9 km 0.5 - 1.2 km 5 N/A 1.3 km2 0.25 - 4.1 km2 0.9 km 0.5 - 1.4 km 13 4.5.1 Morphological overview of study fans and their watersheds The size of fans ranged from 0.05 to 6 km2. Determination of fan area was not an integral measurement in this project. Delineation of the forested fans can be a challenge due to forest cover that obscures the landform boundary where there are subtle changes in slope, particularly along less active margins. In many cases, detailed fieldwork would be required to locate boundaries. Most fans were located in narrow valleys with main valley streams actively truncating the deposits, thereby reducing their size. Determination of fan size can be an important geomorphic measurement related to bedrock geology and weathering processes (Hooke and Rohrer 1977); however, it is a topic that is marginal to this study. As a result, areas were only measured to determine the upper and lower size limits for the study fans. Associated watersheds ranged from 0.21 to 99 km2. Debris flows were generated in the smallest watersheds, with an average size of 1.3 km2 and a range of 0.25 to 4.1 km2. This is within the 56 size range identified in the literature (VanDine 1985). Debris floods were generated in larger watersheds, with an average of 7 km2 and a range of 0.7 to 31 km2. Floods were generated in the largest watersheds, with an average of 34 km2 and a range of 1.4 to 99 km2. In general, the study fans had the classic conical shape. Some fans were confined laterally by valley walls or bedrock knolls for some distance from the apex and had a shape more akin to a steep flood plain. Oblong fans were associated with channel entrenchment. Entrenchment in one case was due to debris flow levees that were at least as old as the forest stand (greater than 250 years). The degree of stream channel confinement on study fans ranged from unconfined for the length of the fan to entrenched for a major portion of the fan. Stream channel gradient at the apex ranged from 1 ° to 22°. Flood fans had an average stream channel gradient at the apex of 3.0°, with a range of 1.0° to 10.5°. In two cases the gradient at the apex of flood fans was greater than the 4° limit reported by Jackson et al. (1987). In one case the channel was entrenched with a gradient of 10.5° and did not reflect the gradient of the fan at the apex, which was 5.5°. In both cases it is likely that the contemporary hydrogeomorphic process (i.e., floods) did not reflect the mode of origin of the fan (i.e., debris floods). Debris flood fans had an average stream channel gradient at the apex of 7.7°, with a range of 1.75° to 18°. In seven cases the gradient at the apex of debris flood fans was < 4°. This was most likely due to sediment wedges in the channels that reduced the channel gradient. This was most apparent in one channel where a large debris jam just below the apex resulted in a significant sediment deposit with a gradient of 1.75°. Debris flow fans had an average stream channel gradient at the apex of 12.4°, with a range of 7° to 22°. Fan surface cross-sections ranged from smooth to dissected. Characteristically, the larger fans had relatively smooth, convex cross-sections while the smaller fans had dissected cross-sections with a generally convex shape. Field identification of forested fans, particularly small fans, can be problematic without aerial photographs. The general impression resulting from a traverse of 57 lower slopes is often a series of streams in a dissected landscape rather than a series of coalescing fans. Elevated and inactive fan surfaces were observed on 13 fans. It is possible that these are early Holocene or paraglacial fan surfaces (Ryder 1971a, 1971b), however dates of the surfaces were not established. The elevated surfaces ranged from a major portion of the present fan landform to minor remnants along the adjacent hillslopes, elevated 2 to 10 m above the active fan surface. Forest ecosystems and stand volume differences between the elevated and active fans surfaces on nine fans were in stark contrast (Table 4.3). Stand volumes were from three to nine times higher on the lower, more active fan surfaces than the elevated fan surfaces. The differences were most likely due to the coarse-textured soils on the elevated surfaces, entrenchment of the stream relative to the elevated surface that would have resulted in a much lower water table, and lack of ecosystem disturbance by hydrogeomorphic processes (i.e., the periodic broadcasting of sediments and inundation characteristic of active zones on fans). The dendroecological investigations on 58 study fans indicated that 57 (98%) showed evidence of characterizing hydrogeomorphic events within the past 100 years and 55 (95%) had evidence of multiple events in the past 50 years. It is clear that hydrogeomorphic events on the study fans are frequent occurrences rather than rare events. Sediment mass balances were not determined for the study fans, so it is not possible to determine whether the fans were in a steady state, aggrading or degrading. However, the evidence indicates that there are definite zones of hydrogeomorphic activity and that forest cover can be used to identify or delineate these zones. 4.5.2 Forest stands on fans as indicators of hydrogeomorphic activity Forest cover has been used to identify where hydrogeomorphic events have occurred on fans (e.g., Kellerhalls and Church 1990), with a focus primarily on high-power or stand clearing events. The focus of this study was to extend the observations to low-power events and to place 58 the whole forest stand-hydrogeomorphic event interaction into an applied classification for forestry. Presently in British Columbia fans are studied individually whenever problems develop (e.g., Thurber Consultants 1983). There is very limited awareness of the hydrogeomorphic hazards on fans and the consequences of "standard" forestry prescriptions (Chapter 7). In addition, there is very limited guidance available to foresters who are planning activities at the landscape level with potentially many fans. The solution is a classification scheme that uses aerial photographs and fieldwork, focusing on forest stands and site features. Such a scheme has been presented in this chapter. With limited effort, all 65 fans were classified using forest stands on fans to determine the power and disturbance extent of hydrogeomorphic events. Two classes of power are based on the ability of events to clear the original forest. Two classes of disturbance extent are applied to the high-power class. Only high-power stand-level events can be identified on aerial photographs. Site level high-power and low-power events require fieldwork to classify past events, although information related to hazards can be garnered from aerial photographs (Chapter 7). Since most study fans have experienced at least one event in the past 50 years (Appendix F), from a forestry perspective it is prudent to conclude that all fans have a high probability of hydrogeomorphic events and require field inspection. Hydrogeomorphic events with the greatest effect on forest cover (highest power and greatest disturbance) are referred to as characterizing events and are used to classify a fan. In some cases the evidence was extensive across a fan surface, but in general the evidence was limited in spatial extent. The zone of historic disturbance from hydrogeomorphic events is generally evident in the case of high-power events, and is present for the observant eye in the case of low-power events. Buried trees were a common feature in zones of sediment accumulation. The lack of butt flares in trees is a notable feature, although it was apparent that in some areas of the Kispiox River watershed western hemlock does not have butt flare. The critical burial depth and sediment texture that causes mortality has not been determined, but other work indicates that 1.6 to 1.9 m 59 may be the maximum burial depth in one event for spruce to survive (Strunk 1997). Scars on tree stems indicate abrasion by either sediment or debris (Figure 4.11). This damage also indicates that the stems were offering resistance to transport. Cylindrical holes up to 2 m deep were found in the ground on five fans (Figure 4.9). It was apparent that these were "tree holes", the result of sediment deposition around the stem, and the long-term result of tree burial. These features can be elusive and are most likely confined to zones that are not actively aggrading (or they would fill in). Adventitious roots are formed when the stems of trees are buried (Figure 4.10). With subsequent erosion of sediments these roots can become exposed. All conifer and deciduous species in the study area were observed to develop adventitious roots after burial by sediments. Exposed adventitious roots were found on a quarter of the study fans and are a key feature for identifying zones of sediment accumulation and subsequent erosion. Sediment deposits within the forests that were unvegetated or had limited accumulation of organic matter and moss cover were considered to be "recent" deposits (Figure 4.12). These features were observed on 88% of the study fans. The value in identifying specific zones on fans with hazards should be apparent and has a parallel to steep slopes. During the past three decades in BC, methods have been developed for identifying landslide prone terrain and, in particular, sites that are either actively failing or commonly fail after forestry activities (Howes and Swanston 1994; Anon. 1999). Terrain stability mapping was embraced rather than using arbitrary slope angles to identify hazards. The result is that only unstable areas rather than whole hillslopes are removed from the productive forest land base for allowable annual cut determination. The parallel with fans is that we can either identify the landform and apply a high reduction factor, or we can apply a classification scheme that identifies specific hazard zones. Forest cover on a fan by itself does not provide evidence of specific hydrogeomorphic process influencing a fan (i.e., flood, vs. debris flow). Stand and site level high-power signatures in the forest canopy were found for floods, debris floods and debris flows. However, no forest stand 60 evidence was found for low-power debris flows. Fieldwork is required to identify hydrogeomorphic processes from the characteristics of the sediment deposits. Dendroecology is an established science that uses tree ring measurements and morphological changes in tree cells for dating events that influence tree establishment and growth. Hydrogeomorphic events are one such influence (Schweingruber 1996). This study used established techniques to determine the occurrence of hydrogeomorphic events on the study fans (Chapter 6). Essentially all study fans have experienced characterizing events in the past 100 years and the majority (95%) have experienced at least one event in the past 50 years. Given the frequency of events, successional development did not obscure the characterizing stand types. 4.5.3 The hydrogeomorphic role of forests on fans Riparian zone forests provide many physical and biological functions, including nutrient inputs, bank stabilization, large woody debris inputs, shade, and deposition zones for sediment from upland sources (Beschta et al. 1987; Bisson et al. 1987; Bjornn and Reiser 1991; Murphy and Meehan 1991; Naiman et al. 2000). A function that has been inadequately described in the literature is the role of riparian forests with regard to hydrogeomorphic processes beyond or outside the stream channel. In this section a case is presented for the recognition of this role. To strengthen the concept of this role, a new term is suggested - the hydrogeomorphic riparian zone. Stream channels are common features on both alluvial and colluvial fans. Stream channels are linear clearings through the forest cover, and as such are frequently, although not always, the avenue for debris flows and debris floods. While the channels lack lateral and vertical confinement by bedrock, a degree of confinement and hydraulic structure may be provided by large woody debris, levees (debris flow and fluvial), and large clasts transported by previous fluvial or mass-wasting events (Keller and Swanson 1979; Thomson 1991). Where vertical confinement is absent, erosion by fluvial action and mass wasting events can lead to channel 61 entrenchment into the fan surface and thereby provide channel confinement, although generally only on portions of a fan. Evidence of lateral confinement for streams and hydrogeomorphic events was observed where woody debris became trapped against riparian trees, forming dykes. Eight large dykes, one to two metres in height, were found on debris flow and debris flood fans. In the absence of these structures, sediments would have been spread for a considerable distance across the fan surface. Small woody dykes, less than one metre in height, were found on 30 (46%) of the study fans. These dykes appear to have several functions. By containing the stream, stream power is maintained and sediments are carried further down-fan. By containing the stream they also reduce or eliminate erosional or depositional impacts to the fan surface. Contemporary hydrogeomorphic activity can subject stream channels to periodic pulses of sediment. Reaches that are deeply entrenched may transport this material, although Osterkamp and Hupp (1987) describe a channel that was periodically subject to mass wasting events that led to cycles of six metre deep entrenchment and re-filling. Poorly-confined reaches may tend to move the sediments laterally out of the channel. In the case of forested fans, this leads to forest/ sediment interactions. Forests influence sediment transport, sediment storage and the establishment of new channels. When streams leave their channels, the flow generally becomes more shallow (dispersed and unconfined) than within the channel, velocities decrease, and sediment deposition results. Deposits are generally deepest along the channel bank and are referred to as levees. Riparian vegetation presents an obstruction to streamflow, reducing further velocities and enhancing sediment deposition adjacent to channels. Evidence of situations where riparian forests were playing a role in enhancing fluvial levee development was found on 51 fans (84% of the study fans), spanning all hydrogeomorphic processes. 62 When sediments are transported across a fan surface, obstructions to flow are encountered that lead to deposition. Aspects of the roles played by forests in sediment storage have been discussed in the literature. Woody debris on the forest floor has been referred to as cross-slope obstructions that lead to sediment storage (Haupt 1959; Wilford 1984). The resulting "log steps" were very common on the study fans (93% of the study fans) (Figure 4.4). Tree stems appear to increase sediment storage of debris-flow events, leading to an increase in slope angles above what would be expected without forests (Irasawa et al. 1991). Detailed slope angle measurements were not made in this study, although buried trees were very common on the study fans (97% of the study fans) (Figure 4.8). When streams leave their established channels and enter a forest a series of factors hinder the rapid establishment of a new channel. These factors are recognized in the literature as root reinforcement of soils and Manning's channel roughness factor, although little information is available specifically on forested fans subject to channel avulsions (Ogrosky and Mockus 1964; Smith 1976). Soil reinforcement by roots was observed on 15 fans where stream channels were flowing through forested areas. It appeared that the reinforcement was effective, but limited to a degree as all sites had evidence of water scouring of sediments below the roots, exposing root systems (Figure 4.10). In most cases of channel avulsions, considerable amounts of sediment were being stored behind woody debris and it was apparent that the process of channel establishment was being hindered (Figure 4.14). The evidence described in this study suggests that forests are playing an important hydrogeomorphic role. This function can extend to at least 100 m from the channels on some fans. Since the current BC Forest Practices Code guidelines for riparian zones call for 30 m reserves on fish streams and streams directly tributary to fish streams (Anon. 1995a), it is important to expand the concept of riparian zones. One way to achieve this is to define a new term - the hydrogeomorphic riparian zone. This zone can be delineated based on site features described in this section. From an operational perspective, delineation of the zone is more time 63 Figure 4.14. The avulsion process or creation of new channels is delayed by forest stands and the associated woody debris (Herb). consuming than simply laying out a 30 m reserve. However, if forest management is to be "sustainable", it is critical to maintain the forest influence in active hydrogeomorphic zones. 4 . 6 CONCLUSIONS A forest stand-based method of determining the power and disturbance extent of hydrogeomorphic processes was applied to the study fans. The classification scheme is relatively simple and provides an indication of hydrogeomorphic hazards for forestry activities. Given the frequency of events, it is clear that hazards exist on all of the study fans. For sustainable forest management it is therefore prudent to apply the scheme to all fans. A case is presented for the recognition of hydrogeomorphic riparian zones on fans; namely areas where forest stands are playing a role in sediment accumulation, erosion control, and mainte-64 nance of stream channel location. Site features are identified to enable delineation of this zone. Maintenance of this forest influence (Kittredge 1948) is directly related to two criteria for sus-tainable forest management (Montreal Process Working Group 1999): (1) maintenance of forest ecosystem health and vitality, and (2) conservation and maintenance of soil and water resources. 65 C H A P T E R 5. A N A L Y S I S O F W A T E R S H E D A T T R I B U T E S 5.1 INTRODUCTION This chapter presents an overview of the watershed attributes and examines them in the context of the forest stand based classification scheme for hydrogeomorphic events (Chapter 4). The specific focus is to identify watershed attributes that characterize the different classes within the scheme. The forest stand based classification scheme has eight categories. Debris flows are associated with two categories: high-power stand and high-power site disturbance levels (referred to as stand and site level debris flows). Floods and debris floods are each associated with three categories: high-power stand and high-power site disturbance levels, and low-power (referred to as stand, site and low-power floods and debris floods). In this chapter the following hypothesis is tested: if the classes are truly unique then the contributing watersheds should also be unique, and it should be possible to identify characterizing watershed attributes. 5.2 WATERSHED ATTRIBUTES Sixteen biophysical watershed attributes were selected based on two general criteria: their influence on peak flow generation and the production of sediment (Table 5.1); but also considering ease of measurement by operational foresters (i.e., topographic and forest cover map attributes). Forestry activities were not present in most watersheds, and those with activities had very limited areas occupied by roads or logging and associated erosion. For this reason, measurements such as road density and extent of forest harvesting were not used as watershed attributes. Six attributes were selected that are related to peak flow generation. Watershed area directly influences the amount of precipitation available for streamflow (Murphey et al. 1977). The length of channels and drainage density in a watershed are directly related to the routing of water (Carlston 1963; Patton and Baker 1976). Watershed shape is a measure of efficiency for the 66 Table 5.1. Watershed attributes. H y d r o g e o m o r p h i c c a t e g o r y W a t e r s h e d at tr ibute A b b r e v -i a t i o n D e s c r i p t i o n U n i t s Relief Relief The difference between the highest and lowest points in a watershed. km Environmentally sensitive areas for soil stability ESA ESAs are forest cover map attributes that are identified by forest classifiers or terrain specialists. Map polygons that contain the initiation sites for natural mass wasting are labeled as ESA. ESAs are expressed as a percent of the total watershed area. % S e d i m e n t p r o d u c t i o n Environmentally sensitive areas for soil stability and other factors ESASx ESASxs are forest cover map attributes that are identified by forest classifiers or terrain specialists. These map polygons contain initiation sites for natural mass wasting and other factors that are sensitive to forestry activities (e.g., wildlife habitat, reforestation issues such as high moisture levels, visual or landscape retention objectives). ESASxs are expressed as a percent of the total watershed area and for the purpose of this study include the extent of ESAs. % Commercial forest cover Comm The percent of a watershed with commercial forest cover, defined as areas of mature and immature forest, and areas that are not satisfactorily restocked as a result of tagging or natural disturbances (e.g., wildfire). % Extent of terrain greater than 30° G30 The percent of the watershed that is greater than 30°. % Extent of terrain greater than 35° G35 The percent of the watershed that is greater than 35°. % Extent of terrain greater than 40° G40 The percent of the watershed that is greater than 40°. % Extent of terrain between 30° and 40° B3040 The percent of the watershed that is between 30° and 40°. % 67 Table 5.1. Continued. Hydro geomorphic category Watershed attribute Abbrev-iation Description Units Area Area Topographically defined area of the watershed. km2 Watershed length Length The straight-line length from the fan apex to the most distant point on the watershed boundary. km Shape Shape Watershed area divided by watershed length squared. knfkm2 Peak flow generation Length of channels Channels The total length of stream channels identified on TRIM maps km Drainage density DrainDen The total length of stream channels divided by watershed area. knvkm2 Hypsometric integral Hypso The hypsometric curve is a pbt of the relative area (a/A) of a watershed that is above a relative height (h/H). The hypsometric integral is the area under the curve. %/% Ratios Melton ratio Melton Watershed relief divided by the square root of watershed area. km/km Relief ratio ReHefRatio Watershed relief divided by watershed length. km/km delivery of water to the mouth of the watershed, with equidimensional watersheds (round) having a higher flood potential than elongated watersheds. The hypsometric integral (Strahler 1952) relates elevation to watershed area. It can be used as an indicator of a range of factors from snow accumulation and melt to overall channel steepness. Eight attributes are related to the production of sediment. Relief is the elevational difference in a watershed, and when combined with watershed length or area, can be used as a measure of watershed steepness. When comparisons between watersheds are undertaken in similar geomorphic settings, such as the present study area, relief can provide a relative comparison of watershed steepness. General slope steepness is related to both peakflows (Patton 1988) and the potential for sediment production. Slope stability mapping was not available for the study watersheds, so a series of watershed attributes were select as surrogates. Four slope gradient 68 classes were selected to highlight potential initiation zones of mass movements. Environmentally sensitive areas (ESAs) are an attribute on forest cover maps (Anon. 1992). These polygons contain the initiation sites for natural mass wasting. The percent of a watershed with commercial forest cover was selected due to the role of forests in moderating runoff and enhancing slope stability (Sidle et al. 1985; Hetherington 1987). Two ratios that integrate watershed size and relief were used as watershed attributes. The Melton ratio (ruggedness number) is a measure of relative relief (relief divided by the square root of watershed area) (Melton 1957). The Melton ratio is related to flooding and debris flow potential (Patton and Baker 1976; Jackson et al. 1987). The second ratio (relief ratio) is watershed relief divided by watershed length (Strahler 1958). Along with other factors such as the extent of exposed bedrock and vegetative cover, watersheds with high peak flows are characteristically short and have high relief (i.e., a high relief ratio) (Costa 1988). 5.3 METHODS Each of the 65 study fans was classified in the field based on the scheme presented in Chapter 4. Watershed boundaries were established along the topographic height of land using TRIM (Terrain Resource Information Management) coverages and GIS (Geographic Information Systems). The lowest point in a watershed was taken to be the apex of the fan (i.e., fans were not included in the watersheds). Overlays were made using TRIM, forest cover, and digital elevation models. From these overlays, the 16 watershed attributes were derived (Appendix D). Identification of watershed attributes that characterize the eight categories of the fan classification scheme was a four-step process: statistical selection of attributes, plotting of attribute data, and exploration of outliers. Version 8.2 of the Statistical Analysis System (SAS) was used for the one-way analysis of variance (ANOVA) to identify if watershed attribute means were significantly different. If differences were detected, Bonferroni multiple comparisons were conducted to determine which categories had different means for the identified attributes. The 69 maximum acceptable a was 0.1. When several attributes were identified, selection of an attribute to characterize a category was based on operational forestry (practical) considerations. For example if two attributes were statistically significant for differentiating a category and one required a topographic map to determine while the other required GIS, the topographic map based attribute was selected. Since ANOVA is a test to compare sample means rather than sample distributions, data were plotted to determine class limits. When attribute data for each category were clustered with no overlap, class limits were easily established. Where overlap occurred, details of the outlier watersheds and their fans were explored. The basic assumptions of ANOVA are: random sampling, and the error terms are independent and normally distributed with a zero mean and equal variance (Sokal and Rohlf 1995). While the selection of watersheds was random, tests to confirm the error term assumptions were not undertaken for several reasons. Sample sizes in general are small, which would result in acceptance of the assumptions even if not true. ANOVA results were just one of several factors used in the attribute selection process. 5.4 RESULTS 5.4.1 Differentiating hydrogeomorphic processes Table 5.2 presents the results from the ANOVA results for differentiating floods, debris floods and debris flows. Figure 5.1 presents a scattergram of Melton ratio versus watershed length and class limits for differentiating hydrogeomorphic processes. Class limits were established by maximizing the number of correctly classified watersheds in each of the three groups: 88% for floods, 83% for debris floods, and 92% for debris flows (Table 5.3). 5.4.2 Differentiating floods There are three categories of floods: low-power, site level, and stand level. The ANOVA identified five watershed attributes with significantly different means for the three categories (Table 5.4). Watershed area was selected as the attribute to differentiate the three flood 70 Table 5.2. Differentiating watershed attributes for floods, debris floods and debris flows and their associated P-values. (See Table 5.1 for an explanation of abbrevia-Debris floods Debris flows Area <0.0001 Area <0.0001 Length <0.0001 Length <0.0001 Channels <0.0001 Channels <0.0001 Melton <0.0001 Floods Melton <0.0001 ReliefRatio <0.0001 ReliefRatio <0.0001 DrainDen 0.0037 B3040 0.0020 G30 <0.0001 Comm 0.035 G35 0.0001 G40 <0.0001 B3040 <0.0001 ESASx 0.0279 Length 0.0375 Melton <0.0001 ReliefRatio <0.0001 Debris floods N/A DrainDen 00036 G30 0.0005 G35 <0.0001 G40 <0.0001 B3040 0.0238 categories. Class limits were established to maximize the number of correctly classified watersheds in each of the three groups, and all flood watersheds were correctly differentiated (Figure 5.2 and Table 5.5). 5.4.2 Differentiating debris floods There are two power categories of debris floods: low-power and high-power. ANOVA determined that two watershed attributes had significantly different means for the two power categories: watershed relief with a P-value of 0.0065, and percent commercial forest in a watershed with a P-value of 0.0012. Figure 5.3 presents a scattergram of watershed relief versus commercial forest cover. Class limits were established to maximize the number of correctly classified watersheds in each of the two groups, and all debris flood watersheds were correctly differentiated (Figure 5.3 and Table 5.6). 71 20 18 16 14 r 12 •«-» Q) x: (0 Q) 8 (0 6 4 2 0 F l o o d s • Floods • Debris floods A Debris flows • • • • • • •: Melton = 0.30 D e b r i s f l o o d s Deb r i s f l o w s length = 2.7km 0.2 0.4 0.6 0.8 1 M e l t o n rat io 1.2 1.4 Figure 5.1. Scattergram using Melton ratio and watershed length with class limits for the hydrogeomorphic processes. Table 5.3. Effectiveness of class limits in identifying hydrogeomorphic processes. Process Sample size Correct Incorrect Flood 16 14 (88%) Debris flood 36 30 (83%) Debris flow 13 12(92%) 2 - as debns floods 2 - as floods 4 - as debris flows 1 - as debris flood 72 Table 5.4. Differentiating watershed attributes for the three categories of floods and their associated P-values. (See Table 5.1 for an explanation of abbreviations.) Site level floods Stand level floods Area 0.0771 Length 0.0242 Channels 0.0121 Area <0.0001 Low-power floods Relief <0.0001 Length <0.0001 Channels <0.0001 Area 0.0002 Site level floods N/A Relief 0.0001 Channels 0.0295 Hypso 0.0415 No significant watershed attributes were found using the ANOVA procedure to differentiate site-and stand-level disturbance high-power debris flood watersheds. The ANOVA procedure was run again after removing five watersheds that were smaller than 2 km2 (3 stand level and 2 site level watersheds) from the data set. This analysis identified three watershed attributes with significantly different means: relief ratio with a P-value of 0.0526, percent of watershed greater than 35° (G35) with a P-value of 0.0710, and percent of watershed greater than 30° (G30) with a P-value of 0.0808. Figure 5.4 presents a scattergram of the high-power debris flood watersheds using the two most significant attributes, relief ratio and G35 (Figure 5.4). A class limit was established to maximize the number of correctly classified watersheds in each category: 77% for site level debris floods and 89% for stand level debris floods (Table 5.7). The scattergram includes the five small watersheds, two of which are misclassified. 5.4.3 Differentiating debris flows There are two categories of debris flows: high-power site level disturbance and high-power stand level disturbance (referred to as site level and stand level debris flows). The ANOVA procedure identified one watershed attribute with significantly different means for the two categories: percent commercial forest in a watershed with a P-value of 0.0178. A class limit was established to maximize the number of correctly classified watersheds in each category: 80% for site level debris flows and 100% for stand level debris flows (Figure 5.5 and Table 5.8). 73 in T3 O JC CO L. 0) (0 3 2 o V -0 n £ 3 15 krn^ 38 km' Low-power High-power site level disturbance • Low-power • High-power site level disturbance • High-power stand level disturbance High-power stand level disturbance o m o m o i r t o i n o L n o m o L O O L o o i o o m o • i - i - t M C M c o c O ' ^ - ^ t i n m c o c o h - r ^ - c o o o c n c D O Watershed area (square km) Figure 5.2. Bar graph of watershed area with class limits for the three flood categories (5 km 2 classes). Table 5.5. Effectiveness of class limits in classifying flood watersheds. Flood category Sample size Correct Incorrect Low-power 7 ( 1 0 0 % ) High-power site level disturbance 2 ( 1 0 0 % ) High-power stand level disturbance 7 ( 1 0 0 % ) 74 100 90 80 73 0) ra 5 ro ' o k_ E £ o o 70 •K 60 50 40 30 20 10 o I • High-power debris flood / o Low-power debris flood / Low-power debris floods C o m m > -109 + 182 (Relief) O O / / / / / • • • • • y = -109 + 182x / High-power debris floods + C o m m < -109 + 182(Relief) 0.2 0.4 0.6 0.8 1 1.2 W a t e r s h e d r e l i e f ( k m ) 1.4 1.6 1.8 Figure 5.3. Scattergram of watershed relief versus commercial forest cover, and the class limit for differentiating low- and high-power debris flood watersheds. Table 5.6. Effectiveness of class limits in classifying low- and high-power debris flood watersheds. Debris flood category Sample size Correct Incorrect Low-power 5 5 (100%) 0 High-power 31 31 (100%) 0 75 60 50 40 73 £ w i— 0) +-> (0 3 ~^ o 30 CO O 20 10 O O Stand level • Site level O Stand level debris flood G35 > 34.6 - 68.4(ReliefRatio) O O • V o Site level debris flood G35 < 34.6 - 68.4(Relief Ratio) • V o • 4 n V O ^ = -34.6 - 68.4x ^ % 0.1 0.2 0.3 Relief ratio 0.4 0.5 0.6 Figure 5.4. Scattergram of high-power debris floods with the class limit between site and stand level disturbance levels. Table 5.7. Effectiveness of class limits in classifying high-power debris flood watersheds. Debris flood category Sample size Correct Incorrect Site level 13 10 (77%) 3 - as stand level Stand level 18 16 (89%) 2 - as site level 76 4 i • Site level • Stand level 45% Commercial forest cover S t a n d l e v e l d i s t u r b a n c e h i g h -p o w e r d e b r i s f l o w s S i te l e v e l d i s tu rbance h i g h - p o w e r d e b r i s f l o w s 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 C o m m e r c i a l forest c o v e r (%of w a t e r s h e d ) Figure 5.5. Bar graph of commercial forest cover with a class limit for differentiating debris flow watersheds. Table 5.8. Effectiveness of class limits in classifying debris flow watersheds. Debris flow category Sample size Correct Incorrect Site level 5 4 (80%) 1 - as stand level Stand level 8 8 (100%) 77 5.5 DISCUSSION 5.5.1 Differentiating hydrogeomorphic processes The Melton ratio has been recognized as a method of differentiating flood and debris flow watersheds. Jackson et al. (1987) found that watersheds in the Canadian Rocky Mountains that produced debris flows had Melton ratios of greater than 0.3 and watersheds producing floods had values below 0.3. Bovis and Jakob (1999) found that debris flow watersheds in Coast Mountains of southwest British Columbia had a minimum ratio of 0.53. These values are similar to those found in this study: flood producing watersheds can be characterized with a Melton ratio of less than 0.3 and debris flow watersheds with a ratio of more than 0.6. Literature reports for the Melton ratio range for debris floods were not found. This study found that debris flood watersheds were primarily in the 0.3 to 0.6 range, but extending to 1.13 in watersheds longer than 2.7 km. The class limits using the Melton ratio correctly identified 14 (88%) of the 16 field classified flood fans. Both misidentified watersheds could have been the result of misclassification in the field. One was the smallest "flood" watershed (1.4 km2) (Newcmbl). The watershed and stream channel appeared to be very stable aside from a mid-fan reach that had a recent, large accumulation of bedload. No evidence of other contemporary hydrogeomorphic activity was found. On the second fan (M3), deposits were very challenging to classify (i.e., flood versus debris flood deposit). Field classification of hydrogeomorphic processes on forested fans can be challenging due to the influence of forests. Clast orientation can be influenced by turbulence around stems and downed woody debris. Classic signatures can be obscured or enhanced by woody debris and trees. It is usually difficult to "stand back" and view the deposit due to the trees. Also, as with fans lacking forest cover, post-event fluvial reworking of sediments can change the orientation of clasts. While this reworking also occurs on fans without forest cover, the direction of water flow for the reworking can be significantly modified by the presence of trees and woody debris. The only way to be 100% sure of orientation is to excavate pits and 78 undertake detailed sedimentologic descriptions. While this detailed work was beyond the synoptic scope of the present study, the field identification of hydrogeomorphic processes is considered to be reasonably accurate. The class limits using the Melton ratio and watershed length correctly identified 12 (92%) of the 13 field classified debris flow fans. The limits placed one as a debris flood (Big Wdnl). As discussed previously, this could be the result of misclassification in the field. The class limits using the Melton ratio and watershed length correctly identified 30 of the 36 (83%) field classified debris flood fans. Two were identified as flood fans, and could be the result of misclassification in the field. Four were identified as debris flow fans. Three of these watersheds have snow avalanches that influence a major portion of the stream channel directly above the fans. It is possible that the snow avalanches in these watersheds are distributing sediments more uniformly along the channels, reducing the potential for debris flows and enhancing the potential for debris floods. It is possible that the fourth "debris flow" watershed was misclassified in the field although this is unlikely as the sedimentologic signatures of debris flows were not present. 5.5.2 Differentiating floods Watershed area was used to differentiate the three classes of flood fans. The class limits correctly identified all sampled fans. The importance of this attribute is logical given that floods with the power to clear wide swaths through forests required a substantial watershed area, while low-power floods only require small watersheds. Intermediate watersheds would be expected to produce events with a more limited, but still powerful influence on forest stands. The utility of selecting watershed area is that in the classification process it has already been determined for the Melton ratio. 79 5.5.3 Differentiating debris floods Watershed relief and the percent of commercial forest cover were used to differentiate low- and high-power debris flood watersheds. The class limits correctly identified all sampled fans. The importance of these attributes are logical since erosion potential is related to watershed relief but is also moderated by forest cover. The percent of watershed with commercial forest cover has not been identified as a common watershed attribute in the debris flow literature. However, it is reasonable that it is a significant attribute in the study region. Forest stands have a generally positive role in the maintenance of slope stability through the effects of roots on soil cohesion (Sidle et al. 1985). Forests play a role in moderating peak flows by reducing snowpacks and delaying melt, intercepting and evaporating precipitation, and enhancing soil infiltration capacity (Hetherington 1987). Thus, as the percent of forest cover in a watershed increases, it follows that incidence of high-power debris floods should decline and low-power debris floods become more common. High-power site and stand disturbance level debris flood watersheds were differentiated based on relief ratio and the percent of the watershed greater than 35°. The class limits correctly identified 16 (89%) of the 18 stand level disturbance debris flood fans. Both misidentified watersheds have low percentages of steep terrain (G35) but this terrain is directly connected to the streams (Compass and Sibola). The class limits correctly identified 10 (77%) of the 13 site level disturbance debris flood fans. One of the misidentified watersheds has snow avalanche activity that influences a major portion of the stream channel directly above the fan (Mill). As previously described, this may reduce the potential for stand level disturbance debris floods. The second misidentified watershed is small and steep with limited sediment sources, perhaps limiting the level of sediment carried in debris floods and thus the level of disturbance to forest cover (8McDnll2). The third misidentified watershed has a high percentage of steep terrain that is drained by a high density of small (ephemeral) tributaries (19Skit). It is possible that the tributaries do not have the capacity to deliver sediments to the main channel, thus limiting the volume and power of debris floods. 80 5.5.4 Differentiating debris flows The percent of watershed with commercial forest cover was used to differentiate the two high-power debris flow classes. As previously discussed, given the role of forests in maintaining slope stability and moderating runoff, it is logical that as the percent of forest cover in a watershed increases, the incidence of high-power stand level disturbance debris flows should decline and high-power site level disturbance debris flows become more common. The class limits correctly identified all eight stand level disturbance fans. The class limits correctly identified four (80%) of the five site level disturbance fans. The one misidentified watershed (Kits3) is one of several sampled along a hillside unit. The adjacent watersheds are larger and have high-power stand level disturbance debris flows. This is the smallest watershed (0.21 km2) in the study and consists primarily of steep rock bluffs, talus slopes and alpine forests. It is possible that the sediment clasts are generally too large to transport given the amount of water available. The result is that the watershed produces high-power site level disturbance debris flows. 5.6 CONCLUSIONS Watershed attributes for the differentiation of all eight categories in the forest stand classification scheme were determined using ANOVA. Class limits were selected by plotting the attribute data and maximizing the number of watersheds in the correct categories. Compared to field classification, prediction success ranged from 77% to 100%. Attributes for predicting hydrogeomorphic processes and six categories can be determined manually from topographic and forest cover maps. Two categories require GIS support to determine attributes. Table 5.9 presents the characterizing watershed attributes and class limits for the hydrogeomorphic processes. Table 5.10 presents the eight categories with characterizing watershed attributes and class limits. The results confirm that the Melton ratio is useful for differentiating flood and debris flow watersheds (Jackson et al. 1987; Bovis and Jakob 1999). However this study provides new evidence that the ratio, in conjunction with watershed length, can be used to differentiate debris 81 flood watersheds. This study also provides new evidence that watershed attributes can be used with a high degree of confidence to predict the power and disturbance extent of hydrogeomorphic events on forested fans. Table 5.9. Class limits for the hydrogeomorphic processes. Hydrogeomorphic process Watershed attribute Class limits Floods Melton ratio < 0.30 ^ ,_ . „ » „ , . . . . . Melton: 0.30 to 0.6 Debris floods Melton ratio and watershed length . . . . , . . „ _ . 6 When Melton > 0.6, Length >2.7 km Debris flows Melton ratio and watershed length Melton > 0.6 and Length < 2.7 km Table 5.10. Class limits for power and disturbance extent for the hydrogeomorphic proc-esses. (See Table 5.1 for an explanation of abbreviations.) Hydrogeomorphic Watershed _ Watershed Dist. „ . ,. .. J " Power ^ * ^ . Class limits process attribute attribute extent Low-power < 15 km2 Floods Area Site 15-38 km2 High-power Stand > 38 km2 Low-power Comm >-109 + 182(Relief) Debris floods R e l i e f R a t i o Site G35 < 34.6 - 68.4(ReliefRatio) v - u i i i n i High-power andOJS S t a n d G35 > 34.6 - 68.4(ReliefRatio) „ K . n r High-power S i t e > 4 5 % Debns flows Comm Stand < 45% 8 2 C H A P T E R 6. F R E Q U E N C Y OF H Y D R O G E O M O R P H I C E V E N T S Tree-ring analysis was conducted on the study fans to enable the determination of past hydrogeomorphic events. Approximately 1050 samples were taken from living trees, including increment cores of stems, stem wedges stems at scars and disks of small stems. This chapter presents the results that identify the number of characterizing hydrogeomorphic events occurring on the study fans during the past 50 years. Characterizing events are defined as the highest-power category of events influencing the forest stand on a fan and are used to place a fan into one of eight categories (Chapter 4). Regression equations used to predict the number of events that have occurred in the last 50 years for each category were developed using watershed attributes as the independent variables. 6.1 INTRODUCTION Trees respond to environmental conditions by increasing or decreasing their rate of radial growth. Additionally, associated morphological changes in the wood structure as well as damage to the cambium can occur. The study of tree response to environmental conditions is called dendroecology (Schweingruber 1996). When the change in conditions is widespread, as caused by climate, entire stands can be affected; this is referred to as an exogenous influence (Cook and Kairiukstis 1990). When the change in conditions is localized (endogenous influence), such as through burial by sediments, only a limited number of trees are affected (Cook and Kairiukstis 1990). Hydrogeomorphic events can have a significant endogenous influence on trees. The application of dendroecology to the dating of hydrogeomorphic events is rather recent (Alestalo 1971), although it has become a relatively common investigative tool primarily because other methods of dating events are generally unavailable (Jakob 1996; Yoshida et al. 1997). Dating of scars and tree establishment can provide the year, or approximate year of hydrogeomorphic events. Abrupt changes in growth can be associated with tree burial or removal of adjacent competing vegetation (Strunk 1997). Information derived from tree-ring width, tree scars, morphological changes, and 8 3 dates of establishment can be portrayed in skeleton plots (Schweingruber et al. 1990). These plots include time (in years) on the x-axis and growth response on the y-axis. Abrupt growth response is recorded based on an increase or decrease from previous years. Annotations are made for scars and physiological changes. Each tree from a study area is plotted on a line, starting with the end of the core or date of establishment and progressing in time until death or time of sampling. When all trees are plotted for a study area it is possible to determine the history of changing conditions affecting trees. Separating exogenous from endogenous influences generally requires that additional samples be taken beyond the specific impact or study area. Dendroecology can provide moderately accurate estimates of the years in which events occur, and in some cases the season (Strunk 1997). Error can be introduced through missing rings and very narrow growth rings. Such errors are usually removed using routine data analysis such as cross-dating. However, for the purpose of this project, exact dates are not required; rather a reasonable estimate of the total number of hydrogeomorphic events during the past 50 years was sought. 6.2 METHODS Following the reconnaissance of a fan, tree-ring sampling was undertaken on 59 of the 65 study fans to determine the ages of cohorts and scars, as well as the dates of abrupt growth change due to burial (negative change) or clearing (positive change). It was necessary to determine whether cohorts were growing on sediments related to a hydrogeomorphic event (i.e., there had to be a connection to the channel). The potential cause of a scar was also explored. If the scar faced upstream or laterally to a stream and there was a high probability it was related to a hydrogeomorphic event, it was sampled. Scars were not sampled if observations indicated that the cause was wildlife or windthrow damage. Tree cores were taken from as low on a tree as possible because growth effects from a burial diminish with height (Strunk 1997). Increment cores were glued to wooden mounts and, along 84 with wedges and disks, were sanded with up to 400-grit paper on a belt sander. Samples were then dated using a 30x zoom microscope. Skeleton plots were constructed that identified a range of features. The date of establishment was determined by adding a sampling height correction to the date of the pith (Thrower et al. 1994; Nussbaum 1996). The error associated with the correction factor was considered to be less than five years. Wood anatomical features caused by physiological stress included scars, traumatic resin canals, and compression wood (Figures 6.1, 6.2, and 6.3). Abrupt positive and negative growth changes were noted, as was the duration of the growth effect (Figure 6.4). Three classes of growth reductions were identified: slight (40 to 55%), moderate (56 to 70%), and strong (>71%). Three classes of growth increases were identified: slight (50 to 100%), moderate (101 to 200%), and strong (>201%). Figure 6.1. A wedge with multiple scars. 85 86 Figure 6.4. A magnified core from a spruce showing strong, abrupt growth change in 1904 resulting from deep sediment burial associated with a high-power stand distur-bance level flood. Growth returned to normal and then greater than normal, most likely due to the establishment of adventitious roots and more growing space (a 200m wide cleared swath adjacent to the tree). For the purpose of event analysis, only dendroecology data for the past 50 years were used. While the data provide information going back to the 1600s on some fans, the most complete coverage and reliable data are for the past 50 years. The key data relate to the dating of scars and cohorts, both of which can present challenges beyond 50 years. Wedges were cut from trees using a hand saw and this limited the size of tree for sampling. Generally scars older than 50 years had significant callous tissue that precluded sampling. Dating events using cohorts can be challenging because of the increased probability that cohorts are removed by a subsequent high-power event (Luckman 1992). For this reason, a 50-year time frame was considered to limit but not eliminate the error associated with absent cohorts. This time frame is well within the design period for drainage structures and rotation lengths for managed forests in the study area. Fans were classified based on hydrogeomorphic process, power and disturbance extent (Chapter 4). Attention was paid in the field descriptions and skeleton plot interpretations to include only characterizing events. These are the events that produce the forest stand signatures for the eight categories of fans. It is possible that on high-power stand-level disturbance fans, site-level disturbance events have also occurred in the past 50 years. It is also likely that on many high-power fans, low-power events have occurred. 87 The dates of certain and probable hydrogeomorphic events were established for each fan. Certain events were indicated where cohorts were established or where scars were present together with abrupt growth changes and wood anatomical features. Probable events were indicated where only growth and wood anatomical features were noted. Thirty-one watershed attributes were selected based on their influence on peak flow generation and the production of sediment (Table 6.1). Six attributes were selected that are related to peak flow generation: watershed area, watershed length, length of channels, drainage density, watershed shape, and hypsometric integral (Section 5.2). Twenty attributes are related to the production of sediment. Eight attributes were described in Section 5.2: relief, environmentally sensitive areas for soils (ESASx) as identified on forest cover maps, extent of commercial forest cover, and four slope gradient classes. In addition, 12 slope gradient classes were selected to provide information on slopes within a given distance from stream channels; these were identified from TRIM. Three attributes were related to watershed elevations: minimum, maximum and mean. The mean elevation was determined through GIS analysis, and was the weighted mean of elevation for 625 m2 grid cells for an entire watershed. Two ratios that integrate watershed size and relief were used as watershed attributes: the Melton ratio (ruggedness number) and the relief ratio (Section 5.2). The linear regression procedure of SAS (Statistical Analysis System Version 8.2) was used to identify the best 10 models based on the adjusted coefficient of multiple determination (a statistic to test the ability of a model to describe the variability in the data) (adjusted R2). The "best" model selection process is similar to step-wise regression. The dependent variable is the number of events and the independent variable(s) are watershed attributes. The number of independent variables was set at: one when the sample size was less than 13, two variables when the sample size was between 13 and 25, three variables when the sample size was between 26 and 32, and four variables when the sample size was 33 or greater. The limit on the number of variables was set due to the relatively small sample sizes and the potential situation where multi-variables could 88 Table 6.1. Watershed attributes. Hydrogeo-morphic category Watershed attribute Abbrev. Description Units Area Area Topographically defined area of the watershed. km2 Watershed length Length The straight-line length from the fan apex to the most distant point on the watershed boundary. km Shape Shape Watershed area divided by watershed length squared. kmVknf Peak flow generation Length of channels Channels The total length of stream channels identified on TRIM maps. km Drainage density DrainDen The total length of stream channels divided by watershed area. km/km2 Hypsometric integral Hypso The hypsometric curve is a pbt of the relative watershed area above a relative height. The hypsometric integral is the area under the curve. %/% Relief Relief The difference between the highest and bwest points in a watershed. km Environmentally sensitive areas for soil stability ESA ESAs are forest cover map attributes that are identified by forest classifiers or terrain specialists. ESAs are map polygons that contain the initiation sites for natural mass wasting. ESAs are expressed as a percent of the total watershed area. % Sediment production Environmentally sensitive areas for soil stability and other factors ESASx ESASxs are forest cover map attributes that are identified by forest classifiers or terrain specialists. These map polygons contain initiation sites for natural mass wasting and other factors that are sensitive to forestry activities (e.g., wildlife habitat, reforestation issues such as high moisture levels, visual or landscape retention objectives). ESASxs are expressed as a percent of the total watershed area and for the purpose of this study include the extent of ESAs. % Commercial forest cover Comm The percent of a watershed with commercial forest cover, defined as areas of mature and immature forest, and areas that are not satisfactorily restocked as a result of tagging or natural disturbances (e.g., wildfire). % Extent of terrain greater than 30, 35, or 40° G30 G35 G40 The percent of the watershed that is greater than 30, 35, or 40°. % Extent of terrain between 30 and 40° B3040 The percent of the watershed that is between 30 and 40°. % 89 Table 6.1. Continued. Hydrogeo-morphic category Watershed attribute Abbrev. Description Units Sediment production The proportion of steep terrain in a corridor adjacent to streams. G3025 G3050 G30100 G3525 G3550 G35100 G4025 G4050 G40100 The proportion of area within a 25, 50 or 100m wide corridor abng all streams on TRIM maps that is greater than 30, 35 or 40°. kirrVkm2 The proportion of terrain between 30 & 40° in a corridor adjacent to streams. B304025 B304050 B3040100 The proportion of area within a 25, 50 or 100m wide corridor along all streams on TRIM maps that is between 30 and 40°. krrrVkm2 General morpho-metry Maximum elevation MaxEl The maximum elevation in a watershed. km Minimum elevation MinEl The elevation of the fan apex. km Mean elevation MeanElev The watershed is divided in 25 by 25 m grid cells and the weighted mean is calculated for the watershed. km Ratios Melton Ruggedness Index Melton Watershed relief divided by the square root of watershed area. km/km Relief ratio ReHefRatio Watershed relief divided by watershed length. km/km describe the sample as opposed to the population. However, the limit was a rule of thumb which could be modified if adding another variable significantly changed the adjusted-R2 value. An adjustment to R2 was undertaken to account for differences in sample sizes and the number of model parameters; adjusted R2 = 1- (1-R2) (n - 1) / (n - p), where n is sample size and p is number of model parameters. Regression analysis was conducted to identify equations for predicting the total number (certain plus probable) of hydrogeomorphic events for each category within the classification. In cases where the sample size of combinations was fewer than four, samples were combined with a higher or lower power class or disturbance extent to provide sufficient statistical power. The regression screening was done and adjusted R2 values were 90 determined. The next step was to determine the regression equations where the sample size was greater than three and the adjusted R2 values were greater than 0.4 (i.e., at least 40% of the variability in the dependent variable (number of events) is explained by the independent variable (watershed attribute)). Using the linear regression procedure (PROC Reg) of SAS, equations were determined and P-values were calculated. The P-value is a test of the significance of the regression coefficient (slope of the regression line). It is expressed as the probability of the coefficient being zero when the null hypothesis is true. The null hypothesis states that there is no relationship between the dependent variable (total number of events) and the independent variables(s) (watershed attributes). Thus, a large P-value indicates that the coefficient is not different from zero, therefore the null hypothesis cannot be rejected, while a small P-value indicates that the coefficient is different from zero and the null hypothesis should be rejected. Only equations with P-values less than 0.1 were considered to provide sufficient evidence to reject the null hypothesis and accept the equation. 6.3 RESULTS Results are presented under the following headings: skeleton plots and summary of events, regression equations, details on ESASx polygons, and details on site-level debris flow watersheds. 6.3.1 Skeleton plots and summary of events Skeleton plots were constructed for the 59 study fans using the dendroecological data. The establishment of cohorts or scars plus abrupt growth changes were used to identify "certain" events. "Probable" events were identified in situations where there were no cohorts or scars, but there was evidence of abrupt growth changes and wood anatomical features caused by physiological stress. An example skeleton plot with a summary of events for that fan is presented in Appendix E. A summary of events for the 59 study fans is presented in Appendix F. 91 Table 6.2. Regression equations to predict the total number of events during a 50-year period. Process Power/Disturbance level Equation for events Degrees of freedom Adj. R 2 P-value Floods Stand + site levels 6.1 0.4 + 1.05(ESASx) 7 0.86 0.0005 Low-power 6.2 -5.6 + 2.8(Length) 5 0.67 0.0291 Debris floods Stand level 6.3 -17.9 + 8.3(MinEl) + 37.0(Hypso) 15 0.46 0.0072 Site level 6.4 -3.0 + 4.0(DrainDen) 12 0.44 0.0082 Low-power 6.5 13.2 - 0.97(Length) 3 0.73 0.0927 Debris flows Stand level 6.6 30.0 - 0.8(Comm) 6 0.73 0.0089 Site level 6.7 9.7 - 0.2(G40) 4 0.59 0.0799 6.3.2 Regression equations No significant equations were found to predict the number events in a 50-year period when the data were unclassified. Table 6.2 presents regression equations to predict the number of "total" events during a 50-year period with the data classified based on field observations (Appendix G). Stand and site-level floods were combined for the analysis because there were only two site-level flood watersheds. 6.3.3 Details on ESASx polygons The mean slope of ESASx polygons in stand- and site-level flood watersheds is 28°, with a range of 19° to 37°. The areal extent of ESASx polygons in these watersheds ranges from 1% to 21%, with a mean of 7.5%. In comparison, the areal extent of terrain greater than 30° in these watersheds ranges from 0% to 62%, with mean of 22%. The ESASx polygons apply only to the forested portion of watersheds, which ranges from 30% to 89%, with a mean of 57%. The location of forests, and hence the ESASx polygons is primarily in valley bottoms and lower slopes of high-power flood watersheds (Figure 6.5). 92 Figure 6.5. ESAs are forested areas with steep, failing slopes. Characteristically ESAs are in the valley bottom, adjacent to streams (Winfield). Table 6.3. Details on the site-level debris flow watersheds. Sediment sources: Mb -morainal blanket (greater than lm thick); Mv - morainal veneer (less than lm thick); Cv - colluvial veneer. Limiting factor relates to transport or weathering with regards to the occurrence of debris flows (Bovis and Jakob 1999). Watershed Total events Predicted events Bedrock type Sediment source Area (km2) G40 (%) Limiting factor Carriganl 10 8 Sedimentary Mb 1.39 10 Transport Irapline 9 9 Volcanic Mv and Cv 0.68 2 Transport 3Copper2 5 7 Plutonic Cv (limited) 0.91 12 Weathering Kits3 4 3 Sedimentary Talus 0.21 32 Weathering Legate 2 4 Plutonic Cv (limited) 1.32 29 Weathering 93 6.3.4 Details on site-level debris flow watersheds Details on the five site-level debris flow watersheds are presented in Table 6.3. This information is used in the Discussion (Section 6.4) to explain the inverse relationship in Equation 6.7 between the number of events occurring in a 50-year period and G40 (the areal extent of terrain greater than 40°). 6.4 DISCUSSION The application of dendroecological techniques has become a common practice in geomorphologic research. The commonly used techniques of dating scars and cohorts were used to identify what are termed here as "certain" events. Additional techniques for identifying abrupt growth and wood anatomical features were applied to determine the dates of "probable" events. The reason for the additional techniques was that cohorts or trees with scars could be removed by subsequent events, thus leading to an underestimation of hydrogeomorphic events on fans. Certain and probable events were added to determine the total number of events on each fan. Dendroecological data was collected on 59 of the 65 study fans. The data indicates that hydrogeomorphic events have occurred in all 59 watersheds during the past 50 years, with all but four watersheds having at least one characterizing event. It is clear from these data that hydrogeomorphic events are not extreme or rare events within the forestry context (i.e., rotation lengths of 80 to 120 years and drainage structure designs for 50 to 100 year events). Similar situations have been identified in other areas (Innes 1985; Jakob and Jordan 2001). 6.4.1 The equations and their variables The equations for predicting the number of events in a 50-year period are presented in Table 6.2. The watershed variables in the equations should not be viewed in isolation because the watersheds have already been classified based on a series of variables as described in Chapter 5. The variables discussed in this section are thus a refinement to the description of specific hydrogeomorphic processes. 94 The extent of ESASx is the variable in Equation 6.1 to predict the number of events from stand-and site-level flood watersheds. This variable is directly related to the number of events. As previously noted, ESASx identifies the extent of unstable terrain within the portion of a watershed with forest cover. Forest cover in high-power flood watersheds ranges from 30 to 89%, with a mean of 57%. While this results in only partial identification of unstable sites in a watershed, it is possible that the location of ESASx polygons is a key factor. These polygons are typically found on lower slopes and in valley bottoms (Figure 6.5). From a sediment transfer perspective, these areas are generally coupled with the stream channel system. Thus ESASx may reflect the extent of directly connected, actively failing terrain. The mean slope angle of ESASx polygons in high-power flood watersheds is 28°. Since the "steep" slopes used in the regression analysis start at 30° (i.e., percent of watershed greater than 30°) it is possible that ESASx is a surrogate for slope classes below 30°. However, given the location of these polygons, they may define erosion-prone sites on lower gradient slopes that provide sediments to streams (e.g., a more narrowly defined class than just slopes greater than 25°). What is striking is the limited extent of ESASx units in the watersheds: a mean of 7.5% and a maximum of 21 %. In comparison, the mean extent of G30 in flood watersheds is 22% and the maximum is 62%. In conclusion, ESASx captures a combination of sediment supply and slope steepness, and is thus an appropriate variable for the prediction of hydrogeomorphic flood events. Watershed length is the variable in Equation 6.2 for the prediction of the total number of low-power flood events. Watershed length is directly related to the number of flood events. The power of events from flood watersheds is directly related to watershed size (Section 5.4.2). Watershed length is a measure of watershed size, and it is reasonable to conclude that as watershed size increases to the 15 km2 limit for low-power floods, the number of events in a 50-year period that create low-power disturbances also increases. Watershed length is also the variable in Equation 6.5 for the prediction of number of low-power debris flood events. Watershed length is inversely related to the number of debris flood events. 95 The inverse relationship between watershed length and the number of events could by spurious due to the small sample size (four) and the limited variation in the number of events (9 to 11), but the relationship could well have a physical basis. The potential for sediment storage from hillslope failures and in-channel mobilization decreases as a watershed becomes shorter. Thus, watersheds with shorter lengths should experience more debris flood events on their fans than longer watersheds, because there is a higher potential for runoff events to deliver sediments to the fan. While future sampling should be undertaken to validate the equation, the physical explanation of the inverse relationship indicates that watershed length is a reasonable variable to predict events from low-power debris flood watersheds. Minimum watershed elevation and the hypsometric integral are the two variables in Equation 6.3 to predict the number of stand-level debris floods in a 50-year period. Both variables are directly related to the number of events, and both are associated with the generation of peak flows. Snowmelt is a major element in the generation of peakflows in the study area (Chapter 3). The higher the minimum watershed elevation, the greater the potential extent of snow in a watershed, and hence the higher probability of snow-generated peak flows. The higher the hypsometric integral, the larger percentage of the watershed at higher elevations, and hence the greater the potential for snow accumulation and subsequent snow-generated peak flows. Within the context of stand-level debris floods as outlined in Chapter 5 (e.g., attributes related to sediment delivery: Melton ratio, relief and extent of commercial forest cover) it is logical that the number of events should be related to factors associated with the generation of peak flows. Drainage density is the variable in Equation 6.4 to predict the number of site-level debris floods. This watershed attribute is directly related to the routing of water, and hence the generation of peak flows (Carlston 1963; Patton 1988). Two aspects of drainage density are related to potential sediment input and delivery to a fan. The first is hillslope dissection by stream channels and the resulting steep slopes with a potential for erosion. The second is the proximity of eroding sites to stream channels. In both cases an increase in drainage density increases the potential for 96 sediment production and delivery to a fan. Drainage density is thus a logical variable to predict the number of site-level debris floods. The areal extent of commercial forest cover is the variable in Equation 6.6 to predict the number of stand-level debris flows. The extent of forest cover is inversely related to the number of hydrogeomorphic events. This is logical given the well established influence of forests on hillslope stabilization and moderation of runoff (Sidle et al. 1985; Hetherington 1987). The areal extent of terrain greater than 40° (G40) is the variable in Equation 6.7 to predict the number of site-level debris flows. The variable has an inverse relationship to the number of hydrogeomorphic events. A similar inverse relationship exists for the Melton ratio and all of the steep slope classes (G30, G35, G40, G40100) identified in the 10 best regression equations with one variable. These relationships are contrary to what should be expected; steeper watersheds should produce more events (Ellen and Mark 1993). An examination of the five site-level debris flow watersheds provides insights into the observed relationships (Table 6.3). The two watersheds with the highest number of events have relatively less steep slopes and sediment sources are deep, failing morainal deposits (Carriganl with 10 events) or small failures in shallow morainal or colluvial deposits (Trapline with 9 events). Carriganl is underlain by sedimentary rock and Trapline is underlain by volcanic rock. Both of these rock types have relatively high weathering rates. These watersheds are characteristic "transport-limited" watersheds as described by Bovis and Jakob (1999). Since sediment input to the channel is not a limiting factor, the frequency of debris flows is controlled primarily by precipitation or snowmelt events (hence, transport-limited). In contrast, 3Copper2 has a slightly smaller watershed than Carriganl but has half the number of events (five). It is underlain by plutonic bedrock and has no deep surficial deposits. This is a characteristic "weathering-limited" watershed as described by Bovis and Jakob (1999) and illustrates the situation where a slow recruitment of sediments to the channel limits the frequency of debris flows. Kits3 had four events, less than half the number of events of Carriganl, although it too is underlain by sedimentary rock. Kits3 has very coarse 97 colluvial (talus) deposits and no morainal deposits were identified. This is the smallest watershed in the study (0.21 km2) and potentially requires extremely large and rare precipitation or snowmelt events to generate debris flows. While this may appear to be a "transport-limited" watershed, the size of sediment is generally too large for the stream to transport, so Kits3 was classified as a "weathering-limited" watershed. Legate was one of the steepest, largest debris flow watersheds yet it had the lowest number of events (two). It is underlain by plutonic rock and has very limited surficial deposits. This is a characteristic "weathering-limited" watershed with associated low frequency of events (Bovis and Jakob 1999). In addition, snow avalanches influence a major portion of the stream channel in the Legate watershed, thus potentially reducing the number of debris flow events by spreading the limited amount of sediment along the channel rather than allowing sediment to accumulate and subsequently fail. Three conclusions can be drawn from the site-level debris flow watershed review. The first is that the sample size (five) is too small to determine a relationship between the number of events and watershed attributes given the observed variation in the watersheds (e.g., geology, surficial materials and size). This conclusion points to the need for additional sampling in future studies. The second conclusion is that watershed steepness is masked by bedrock geology (i.e., the high number of events in the less steep watersheds is due to bedrock with a high weathering rate). This conclusion does not reflect contemporary bedrock weathering as only one watershed (Kits3) had bedrock as the principal source of sediment for debris flows. However the situation could reflect thicker, finer glacial moraine deposits in watersheds with sedimentary and volcanic bedrock. These deposits would be higher sediment producers than the thinner, coarser morainal deposits in the plutonic watersheds. The third conclusion, complementary to the second, is that the inverse relationship between watershed steepness and the number of events is possibly the result of historic erosion of surficial deposits in the steep site-level debris flow watersheds in the study area; steep watersheds are sediment- or weathering-limited. Thus recognizing the small 98 sample size, the inverse relationship between the number of site-level debris flow events and G40 is reasonable for the study area. 6.5 CONCLUSIONS Dendroecological techniques were used to identify the number of certain and probable hydrogeomorphic events on study fans during the past 50 years. Statistically significant relationships were found between the total number of hydrogeomorphic events occurring during the 50-year period and watershed attributes. In addition to a limited number of basic watershed morphometric measurements, two forest attributes (Comm and ESASx) were used as independent variables in the predictive equations. The use of these forest attributes to describe hydrogeomorphic processes is novel and marks a contribution to the science. A significant finding is that hydrogeomorphic processes in the study area occur relatively frequently and can not be considered as extreme events. Future research is required to increase the sampling database and to validate the equations. 99 C H A P T E R 7. F O R E S T R Y A C T I V I T I E S O N F A N S 7.1 INTRODUCTION This chapter explores the relations between forestry practices and hydrogeomorphic processes. In particular the exploration was intended to: • Identify watershed or fan features that can be used to indicate hazards for forestry activities. • Identify forestry practices that increase the disturbance caused by hydrogeomorphic processes (i.e., inappropriate prescriptions). • Identify forestry practices that do not increase the disturbance caused by hydrogeomorphic processes (i.e., appropriate prescriptions). Forestry activities occurred on 55 of the 65 study fans. Activities on the fans ranged from a single road crossing at the apex to complete clearcutting of the forests. Forestry activities were undertaken across a considerable time period (pre-1950 through to 2000). Given the range of activities, the long time period, and the gradually evolving awareness of other resources in forestry prescriptions and practices, this sample provides considerable insights into the effects of forestry activities on fans. 7.2 METHODS Staff from forest licencees and government agencies (Fisheries and Oceans Canada, BC Ministry of Forests, and BC Ministry of Water, Land and Air Protection) provided a list of fans in the study area with forestry activities. Study fans were selected based on the following criteria: • Reasonable access (vehicle preferred but a helicopter was used for several fans), • Availability of historic information regarding the forestry activities, • Forestry is the only land use activity on a fan (e.g., fans with a complex history of logging, gas pipelines, highways and other activities obscure the objective of focusing on forest land use practices), 100 • Sample population should be a reasonable cross-section of the hydrogeomorphic processes, power and disturbance extent combinations, • No or very limited forestry activities in the contributing watershed. Pre-logging aerial photographs were used to identify five fan and watershed features: multiple (distributary) channels on fans (Figure 7.1); evidence of high sediment load in the stream channels on the fan and in the watershed (Figure 7.2); abrupt disappearance of a stream channel on a fan (Figure 7.3); abrupt angles in stream channels on fans (Figure 7.4); and, sediment sources (landslides) near the watershed mouth (lower 25% of the watershed) (Figure 7.5). Fans were classified in the field using features from the hydrogeomorphic processes: power (as expressed in damage to the forest cover) and disturbance extent (a width greater or less than 20 metres of stand cleared) (Chapter 4). The events that had the most powerful effect on the forest stand are referred to as characterizing events (e.g., high-power stand level floods). Dendroecology was the principal method of determining when characterizing events occurred (Chapter 6). The other method of determining event dates used the records of forest licencees and government agencies. Forestry activities were described according to Table 7.1. Management issues on fans were identified based on discussions with forest licencees, government agencies, logging contractors, consultants and field observations (dendroecologic sampling, and description of forestry practices and damage). Damage to five types of features were identified: drainage structures, roads, fish habitat, plantations and forest sites. Impacts were classified as major (PI) if damage was significant, for example drainage structures required replacement, roads washed out, streams impacted to the point where there is no or very poor fish habitat, and plantations buried in sediments or eroded. Impacts were classified as limited (P2) if there was localized erosion on roads and limited impacts to drainage structures, plantations, forest sites and fish habitat. "Nil" was used to describe fans on which there were no significant management issues, even though hydrogeomorphic events may have occurred. 101 Figure 7.1. Multiple channels are present on this fan (Powerline). Photo 30BCC594 #15 at a scale of 1: -12 000 and taken in 1972. Source: MSRM. 102 Figure 7.2. High sediment loads are being transported by this stream as evidenced by: a change in channel morphology downstream of the fan; the fan is pushing the main stream across the valley; and mid-channel bars and braided reaches in the stream channel on the fan (Winfield). The arrow indicates flow in the main channel. Photo 30BCC93067 #159 at a scale of 1: -19 000 and taken in 1993. Source: MSRM. 103 Figure 7.3. The stream channel disappears from view on the aerial photographs (arrow) indicating a broadcasting of flows under the forest and/or multiple channels that do not have the power to clear a swath through the forest stand (Gosnell 1). Photo 30BCC96050 #81 at a scale of 1: ~4 050 and taken in 1996. Source: MSRM. 104 Figure 7.4. A large debris jam in the main channel (white arrow) led to the formation of a second channel (black arrow) which initiates at an abrupt angle (black arrow). Flow occured in the second channel for a period of years but is now totally in the main channel (Sibola). Photo 30BCC96169 #179 at a scale of 1: ~5 800 and taken in 1996. Source MSRM. 105 Figure 7.5. Major sediment sources near the watershed mouth can provide direct delivery of sediments and debris to fans due to limited opportunity for channel and bank storage (Whitebot). Photo SRSB985953 #421 at a scale of 1: ~9 200 and taken in 1998. Photo with permission. 106 Table 7.1. Description of forestry activities. Forestry activity Measurement or description Road profile using a Suunto and 50 metre tape Roads, ditches and borrow Dimensions - width, depth, cut bank height pits Level of maintenance or deactivation and quality of ditchbfocks Evidence of erosion (surface and mass wasting) Dimensions Drainage structures and Condition (e.g., crushed culverts) channel excavations Engineering works (dykes and rip-rap) Logged areas including Dimensions of riparian reserves landings Zones of erosion or deposition (including skid trails, fire guards and yarding disturbance) The causes of management issues were established primarily through field observations and aerial photographic analysis, although consideration was given to the observations and opinions of staff from forest licencees, government agencies and consultants. Five classes of causes were established: inadequate drainage structures (including channel excavation), roads (including ditches), riparian logging, removal of large woody debris from streams, and soil mass wasting associated with forestry activities in watersheds. Attempts were made to determine the financial cost of problems, with the focus primarily on the costs associated with restoring access across fans. Very few situations were found where costs were available since road problems are dealt with in many ways. Contractors, individual licencees, or partnerships between licencees or licencees and government agencies have corrected problems on impacted study fans. In almost all cases, specific records for individual fans are not kept. Where records are kept they generally include only contract costs and exclude staff time. As a result, financial costs could not be evaluated. 107 7.3 RESULTS The results are presented in a spreadsheet in Appendix H. Section 7.3.1 presents an overview of the results from the 55 study fans with forestry activities. Section 7.3.2 presents specific aspects of forestry activities and impacts. Fan reference numbers are presented in Appendix H. 7.3.1 Overview 7.3.1.1 Pre-logging aerial photographic features A summary of features identified on pre-logging aerial photographs and the relation to management issues is presented in Table 7.2. Road crossings were present below the distributary point in the zone of multiple channels on nine fans. Management-related issues arose on five fans (56% of the situations) (fan #1, 20, 22, 23, and 33). The issues involved roads crossing the multiple channels and inadequate drainage structures when the proportion of flow changed in the various channels. On four fans, the portion of streamflow in the multiple channels had not changed since road construction (#14, 29, 36, and 42). Road crossings were present above the distributary point on six fans, avoiding issues with multiple channels (#4, 5, 6, 7, 16, and 21); however, drainage structures were inadequate in four cases (#4, 5, 6, and 7). In one case, a road may have crossed in the zone of multiple channels but no evidence of the structures remained at the site (#2 with 1950s logging). Table 7.2. Pre-logging aerial photographic features and relation to management issues. Feature Fans with feature Related to management issue Multiple channels on fan 16 5 (31%) High sediment load in channel 10 10 (100%) Disappearing channel 7 5 (71%) Abrupt stream angle on fan 3 3 (100%) Sediment source near mouth of watershed 17 15 (88%) 108 Aerial photographic interpretation indicated high sediment loads were noted in the stream channels on 10 fans (Figure 7.2). Management issues arose on all 10 fans. Drainage structures were overwhelmed on six fans (#3, 4, 5, 6, 15, and 20). Avulsions occurred on six fans. Avulsions occurred above forestry activities on three fans, leading to impacts on roads (#5, 12, and 33). Avulsions occurred in areas with riparian logging on two fans (#8 and 9) and just downstream of an inadequate drainage structure on one fan (#3). Abruptly disappearing channels were present on seven fans. Two of these fans had no management issues (#21 and 40). Management issues arose on five fans related to disappearing channels. The issue on two of these fans was related to an inadequate or no riparian zones (#28 and 29). The issue on the other three fans involved intercepted broadcast flow by a road (#39) or skid trail (#26 and #27) and subsequent erosion. Abrupt angles in stream channels were found on three fans. All were associated with management issues. An avulsion at an abrupt stream angle led to the failure of a drainage structure on one fan (#13). The other two were multiple channel situations with logjams blocking flows down one channel. In one case the jam was breached (#20) and in the other a jam was formed (#33). In each case there was a significant change in the amount of water carried by each channel. Drainage structures and roads were impacted in both cases. Sediment sources near the mouths of watersheds were found in 17 cases. Management issues were present on 15 of these fans (#1, 4, 5, 6, 9, 12, 13, 23, 30, 39, 46, 47, 51, 52, and 55). Management issues were associated with drainage structures, riparian logging, and roads. No issues were present on the other two fans. One fan had a road crossing below the deposition zone of flood events (#31). The other fan had not experienced a characterizing event since the 1999 logging and road construction (#41). 109 7.3.1.2 Characterizing events Characterizing events influencing a fan are those events that have the most powerful effect on the fan and its forest stands. These events are used to classify a fan (e.g., high-power stand level flood fan). Dating of characterizing events is done primarily using dendroecology although specific dates are available from written, oral or photographic records in some situations. Dendroecology sampling was undertaken on 49 of the 55 study fans with forestry activity. Sampling was not possible due to extensive clearcutting on four fans and sampling was not undertaken on two other fans. The occurrence of characterizing events on these six fans was determined through aerial photographs and the presence of recent sediment splays. A summary of characterizing events is presented in Table 7.3. Eighty percent of the fans experienced characterizing events and fans in all impact classes have experienced characterizing events since the start of forestry activities. Characterizing events occurred on 24 (89%) of the PI (major) impact class fans, 11 (78%) of the P2 (limited) impact class fans, and nine (64%) of the Nil impact fans ("Nil" impacts as a result of the forestry activities). Events that had an influence on forest cover but were less powerful than characterizing events occurred on almost 18% of the fans; three (11 %) of the PI impact class fans, three (21 %) of the P2 impact class fans, and four (28%) of the Nil impact class fans. One Table 7.3. Occurrence of characterizing events since forestry activities. Impact class PI P2 Nil Totals Number of fans 27 14 14 55 Number of fans with characterizing event(s) since forestry activity 24 11 9 44 Number of fans with non-characterizing event(s) only since forestry activity 3 3 4 10 Number of fans with no events since forestry activity (done in 2001) 0 0 1 1 110 Nil impact category fan experienced very recent forestry activities and had not experienced an event. 7.3.1.3 Overview of forestry activity relations and impacts A total of 41 (74%) of the 55 fans with forestry activities had impacts to drainage structures, roads, forest sites and fish habitat. No impacts were apparent on 14 study fans (25%). The PI (major) impact class accounted for 27 (49%) of the fans with forestry activities, and 14 (25%) of the fans had P2 (limited) impacts. The number of fans in each impact class is summarized according to the forest stand-based hazard classification scheme in Table 7.4. All seven high-power stand-level flood fans with forestry activities sustained major impacts. Four high-power stand-level debris flood fans had no impacts associated with forestry activities while eight had major impacts and four had limited impacts. One high-power stand-level debris flow fan had major impacts and five had limited impacts. Two high-power site level flood fans were sampled, and one had no impacts associated with forestry activities while the other had major impacts. Four high-power site-level debris flood Table 7.4. Impacts by category of the forest stand-based hazard classification. Power High-power Low-power Disturbance level Stand level Site level N/A Totals Process Flood Debris flood Debris flow Flood Debris flood Debris flow Flood Debris flood Number of fans 7 16 6 2 10 3 7 4 55 (100%) Impact class PI 7 8 1 1 5 0 3 2 27 (49%) P2 0 4 5 0 1 1 1 2 14 (25%) Nil 0 4 0 1 4 2 3 0 14 (25%) 111 fans had no impacts associated with forestry activities, five had major impacts and one had limited impacts. Two high-power site-level debris flow fans had no impacts associated with forestry activities and one had limited impacts. Three low-power site-level flood fans had no impacts associated with forestry activities, three had major impacts and one had limited impacts. Two low-power site-level debris flood fans had major impacts associated with forestry activities and two had limited impacts. Multiple forestry activities were associated with impacts on 17 of the 41 impacted fans. Multiple impacts to improvements and resource values were identified on 27 of the 41 fans. Associated forestry activities and impacts for the three impact classes are presented in Table 7.5. Table 7.5. Associated forestry activities and impacts by impact classes. Percentages are presented by impact class for causes and impacts. Multiple causes and impacts result in percentage values summing to greater than 100%. Class of impact Total PI (major) P2 (limited) PI +P2 Nil Total 27 14 41 14 55 49.1% 25.4% 74.5% 25.4% 99.9% Associated forestry activity Drain, struct /chan. exc. 23 (85%) 7 (50%) 30 (73%) 30 (54%) Roads / ditches 18 (67%) 6 (42%) 24 (58%) 24 (44%) Riparian tagging 16 (59%) 7 (50%) 23 (56%) 23 (42%) LWD removal 1 (4%) 1 (7%) 2 (5%) 2 (4%) Mass wasting related to 1 (4%) 0 1 (2%) 1 (2%) forestry activities Impacts Roads 22 (81%) 10 (71%) 32 (78%) 32 (57%) Plantations/forest sites 22 (81%) 6 (43%) 28 (68%) 28 (51%) Drainage structures 19 (70%) 4 (28%) 23 (56%) 23 (42%) Fish habitat 14 (52%) 1 (7%) 15 (36%) 15 (27%) 112 Inadequate drainage structures and channel excavations were associated with 73% of impacted fans and were found on 54% of all study fans. Roads and ditchlines were associated with 58% of the impacted fans (44% of all study fans). Riparian logging was associated with 56% of the impacted fans (42% of all study fans). Removal of large woody debris from stream channels and mass wasting related to forestry activities had minor associations on impacted fans. Comparing the relation between forestry activities and impact class, it is more common for forestry activities to be associated with major rather than minor impacts. Damage to roads occurred on 78% of impacted fans (57% of all study fans). Plantation and forest site damage occurred on 68% of the impacted fans (51% of all study fans). Damage to drainage structures occurred on 56% of impacted fans (42% of all study fans). Damage to fish habitat occurred on 36% of the impacted fans (27% of all study fans). Comparing the incidence of impacts in all categories to impact class, it is more common for impacts to be major rather than minor. Associated forestry activities by impact classes are summarized according to the forest stand-based hazard classification scheme in Tables 7.6 and 7.7. Impacted resources are summarized by impact classes according to the forest stand-based hazard classification scheme in Tables 7.8 and 7.9. Specific aspects of impacts are presented in a following section. 113 Table 7.6. Forestry activities associated with PI impacts by category of the forest stand-based hazard classification. Power High-power Low-power Disturbance level Stand level Site level N/A Totals Process Flood Debris flood Debris flow Flood Debris flood Debris flow Flood Debris flow Number of fans 7 8 1 1 5 0 3 2 27 Associated forestry activity Drainage structure/channel excavation 5 8 0 1 4 0 3 2 23 Riparian logging 2 5 1 1 3 0 2 2 16 Roads/ditches 4 5 1 1 3 0 3 1 18 LWD removal from streams 0 0 0 0 0 0 1 0 1 Mass wasting from forestry activity 1 0 0 0 0 0 0 0 1 Table 7.7. Forestry activities associated with P2 impacts by category of the forest stand-based hazard classification. Power High-power Low-power Disturbance level Stand level Site level N/A Totals Process Flood Debris flood Debris flow Flood Debris flood Debris flow Flood Debris flood Number of fans 0 4 5 0 1 1 1 2 14 Associated forestry activity Drainage structure/channel excavation 0 3 0 0 1 2 0 1 7 Riparian fogging 0 2 4 0 0 0 0 1 7 Roads/ditches 0 3 1 0 0 1 1 0 6 LWD removal from streams 0 0 0 0 0 0 0 1 1 Mass wasting from forestry activity 0 0 0 0 0 0 0 0 0 114 Table 7.8. PI impacts by category of the forest stand-based hazard classification. Power High-power Low-power Disturbance level Stand level Site level N/A Totals Process Flood Debris flood Debris flow Flood Debris flood Debris flow Flood Debris flood Number of fens 7 8 1 1 5 0 3 2 27 Impact resource Roads 4 6 1 1 5 0 3 2 22 Plantations/forest site 6 5 1 1 4 0 3 2 22 Drainage structures 6 6 0 1 3 0 1 2 19 Fish Habitat 3 4 0 1 2 0 2 2 14 Table 7.9. P2 impacts by category of the forest stand-based hazard classification. Power High-power Low-power Disturbance level Stand level Site level N/A Totals Process Flood Debris flood Debris flow Flood Debris flood Debris flow Flood Debris flood Number of fens 0 4 5 0 1 1 1 2 14 Impact resource Roads 0 4 2 0 1 1 1 1 10 Plantations/forest site 0 0 4 0 0 0 0 2 6 Drainage structures 0 1 0 0 1 1 0 1 4 Fish Habitat 0 0 0 0 0 0 0 1 1 7.3.2 Specific details A summary of specific details on forestry prescriptions grouped according to their role in exacerbating hydrogeomorphic events is presented in Table 7.10. Specific details are presented in the following sections. 115 Table 7.10. A summary of specific details on forestry prescriptions grouped according to their role in exacerbating hydrogeomorphic events. Prescription Summary of results Appropriate prescriptions Road crossing at apex Observed on ten fans. Four were original locations, six were the result of relocations. This location avoids most problems from hydrogeomorphic events. Seasonal deactivation Observed on four fans with Nil impacts. Adequate permanent drainage Present on 28 fans (51% of fans with forestry activities). Ten are replacements structures following hydrogeomorphic events. Engineered structures Observed on five fans. All were functioning but one will confine flows during hydrogeomorphic events and potentially lead to failure of a drainage structure. Special maintenance attention for Frequent inspections and works on three fans reduced the level of impacts. drainage structures Road crosses stream in uniform Observed on 16 fans. Only one had a problem related to the structure - 3m gradient reach scour at the outfall. Roads with design feaures that Observed on 39 fans. Problems were avoided on 10 fans with adequate climb to stream crossings drainage structures, rolling grades, crossing low on a fan, regular maintenance of drainage structures, effective deactivation and no ditchlines. Roads drop to stream crossings Observed on seven fans. Four are Nil impacts, two have localized erosion, and one re-diverted an event back to channel. Roads cross below the zone of Observed on 12 fans. Led to no damage in 10 cases, limited damage in one erosion and deposition case and major damage in another. Overlanded roads Observed on two Nil impact fans. No ditchlines Observed on seven fans. Five were Nil impact fans. Two were on major impact fans, with the road diverted the event off the road in one case (outsloped road) and channelling water down the road in another (slightly entrenched road became deeply entrenched). Rolling grade in road Observed on five fans. Three were Nil impact fans and two had limited road impacts. Retention of large woody debris Observed on eight logged fans. Prescribed hydrogeomorphic role on five fans. Partial cutting Observed on four fans. Maintained the hydrogeomorphic role of the forest. 116 Table 7.10. Continued. Prescription Summary of results Inappropriate prescriptions Roads climb to stream crossings Roads at slope break in stream/ fan (from steep to gentle) Inadequate ditchblocks and cross drains as roads traverse fans Multi-span drainage structures Inadequate drainage structures Channel excavation - no riprap Inadequate drainage structures in multiple channel situations Non-engineered structures Roads on fans are not deactivated Logging of the hydrogeomorphic riparian zone Skid trails on fans intercept and concentrate flows Observed on 39 fans. Aggravation of natural events occurred in 26 cases (66%) and roads were impacted in 29 cases (74%). Nine roads were constructed at a slope break and all had drainage structure failures due to sediment aggradation. Observed on 15 fans with problems: eight were linked to failure of main stream drainage structure, five to broadcast flow interception, one to subsurface interception, one eroded during a high flow. Observed on four fans. All have been replaced due to blockage of bedload. Observed on 30 fans (54% of the fans with forestry activities and 73% of impacted fans). Too small for or damaged by hydrogeomorphic events, and changed channel hydraulics (wrong cross-sectional shape). Observed on five fans. Major problems on two fans and localized problems on three fans. Multiple channels observed on 16 fans with nine having road crossings in the zone of multiple channels. Five had impacts related to inadequate drainage structures when the proportion of flow changed. Observed on four fans. All have, or will shortly fail to achieve their objectives. Observed on nine fans, resulting in major problems on seven fans and localized problems on two fans. Observed on 24 of the 41 impacted fans. Channel widening or damage to plantations occurred on 23 fans of the 24 fans. Observed on seven fans with major disturbances on three and limited disturbances on four fans. 7.3.2.1 Roads and drainage structures Drainage structures and channel excavations were the forestry activity most frequently associated with both the major and limited impact classes (30 of 55 fans or 54%). Road-related channel excavations were associated with PI impacts on two fans and P2 impacts on three fans. Three situations involved excavations of the channel to accommodate drainage structures or the road. In all cases, rip-rap was not placed to support the abrupt channel bed drop. In two situations borrow pits in or very close to the channel were associated with avulsions. 117 Multi-span drainage structures across a single channel were originally installed on four fans. All of these structures have been replaced or removed. Removal in one case was associated with the relocation of the road to cross at the fan apex. Inadequate ditchblocks and cross drains were associated with impacts on 15 fans. Eight cases were linked to failure of the main stream drainage structure, five were associated with ditchline interception of broadcast flows, one was due to subsurface water interception (based on field evidence), and one was due to ditchblock erosion during a high flow event (the ditch is now a high-stage distributary). Adequate seasonal drainage structures had been installed on four Nil impact fans (a seasonal drainage structure is installed for a short period of time and removed prior to high runoff conditions). Adequate permanent drainage structures were present on 28 fans. However, 10 of these structures were replacements due to past damage by hydrogeomorphic processes damage or recognition that the original structures were too small. On four of the 10 fans, structures have been replaced more than once. One fan has had three successive replacement bridges at a cost of more than $600,000. Roads were relocated from mid-fan to the apex on four of the 10 fans. Engineered structures were constructed on five fans. One was a functional dyke located immediately upstream of a mid-fan bridge crossing. The second engineered structure was an armoured bank that was constructed in 1993 for $260,000. The third was a dyke/armoured bank to protect a road that descended from the adjacent hillslope onto the margin of a lower fan. The fourth was an armoured bank installed following an avulsion (associated with riparian logging and an inadequate drainage structure). Engineered dykes were installed on one fan; however this constricted the channel, modifying sediment transport and impacting fish habitat. In this case the existing bridge has adequate capacity to pass water but not sediments associated with debris floods (seven events in the past 50 years and four since the road was constructed). This structure has a width of just under 10 metres and a clearance of just over 1.5 metres. Channel scour and 118 elevated deposits upstream of the dyke area indicate that sediments and water associated with debris floods in this stream were estimated to have a cross sectional area 25 metres wide by 2 metres deep. In-channel excavations and non-engineered dykes were present on four fans. An excavator removed bedload and woody debris from one channel in an attempt to reduce the amount of bedload delivered to a drainage structure (#37). This work was not successful at eliminating the problem and a larger structure is being designed. Waste from this operation was placed on the streambanks, which had been logged. This activity resulted in impacts to two hectares of plantations and forest sites. Dykes of local fan materials were constructed on two fans. One dyke eroded, leading to a major channel avulsion that damaged 200 metres of road (a climbing road) (#23). This dyke was replaced with an engineered dyke and a longer bridge was installed. One dyke of local materials is presently eroding but has not yet breached (#9). A dyke of streambed materials was constructed on one fan to block off a distributary channel (to address a road-related issue) (#39). Engineered structures were needed, but not present on two streams to stabilize channel excavations (the roads were built across the toes of steep fans) (#17, 42). No cases were found where riprap was used to stabilize excavations in stream channels. Riprap was generally "dumped" around bridge abutments. Commonly the material was not placed ("keyed-in") or set below the depth of scour. In most cases riprap was not significantly larger than the D 9 0 of the adjacent streambed. Special maintenance attention was given to three fans (i.e., inspections prior to, during and after major runoff events with work prescribed as necessary compared to annual inspections). Two of the fans had P2 level impacts (#8, 9) and one had PI level impacts (#15). When stream channels have a change in slope from steep to gentle, deposition of sediments is a natural process. Nine drainage structures were located across streams in such reaches and all 119 failed due to aggradation of sediments (#4, 6, 7, 17, 23, 30, 37, 49, 51). Sixteen fans had drainage structures in reaches with uniform channel gradients. Although other factors are involved, only one of these fans has had a problem with a drainage structure (i.e., a culvert with a 2.5 metre deep scour hole at the outlet that is undermining the road) (#34). Ten fans are Nil impact, three are P2 impact fans and three are PI impact fans. Roads and ditches were the second most frequent cause of major (PI) impacts and the third most frequent cause of limited (P2) impacts. Data were collected on a series of aspects of roads and ditches. Table 7.11 presents results with regards to road profiles on fans. Roads climb to the stream in at least one direction on 39 of the 55 fans with forestry activity. In the 24 cases where roads were associated with PI and P2 impacts, all had a climbing grade to the stream. In 32 cases where roads were impacted by events, 29 had climbing grades to the stream (the three other cases involved localized washouts at drainage structures). Table 7.12 summarizes details of the 24 situations where climbing roads were associated with impacts. The leading causes were inadequate ditch blocks and cross drains, and inadequate drainage structures on the main streams. Breaching of stream banks by road construction was noted as a cause of avulsions on three fans with climbing roads. It is possible that the number of fans with breached banks is higher but the evidence was not clear. Table 7.13 summarizes details of situations where climbing roads were not associated with impacts and sustained limited or no damage. In 10 of the 15 situations where climbing roads were not associated with impacts the drainage structures on the main streams Table 7.11. Summary of roads that climb to streams in at least one direction. Number of roads that Road was associated with _ , . . , Impact class • Road was impacted climb to stream impacts PI 22 18 20 P 2 11 6 9 Nil 6 0 0 24 Total 39 29 120 Table 7.12. Details of cases where climbing roads are associated PI and P2 impacts. Details Number of cases Percent of total (26) Inadequate ditch blocks/inadequate cross drains 20 83% Inadequate drainage structure on main stream 18 73% Road became stream channel 17 71% Ditchline interception or channelling flow 14 58% Road not deactivated promptly or adequately 8 33% Streambanks breached 3 12% Borrow pit cbse to or in stream stream 2 8% Table 7.13. Details of cases where climbing roads did not cause impacts. Details Number of cases Percent of total (13) Adequate drainage structure or effective back-up structure 10 67% Rolling grade in road 3 20% Road crosses on bwer fan 3 20% Regular maintenance of drainage structures 3 20% Effective deactivation 2 13% No ditchline and outsfoped road 2 13% were either adequate or there was an effective back-up structure. Other factors that contributed to climbing roads not causing problems were rolling grades, crossing low on a fan (beyond the most active zone of deposition and erosion), regular maintenance of drainage structures, effective deactivation, and using an outsloped road with no interior ditchline. Failure to deactivate roads was associated with impacts on nine fans (seven PI impacts and two P2 impacts). Prompt road deactivation was undertaken on three Nil impact fans. Roads became avulsion channels on 13 fans. Of these, six were corrected shortly after the hydrogeomorphic events, but seven continued to act as stream channels 10 to 15 years after the events. 121 Roads cross at the apex of four Nil impact fans. Of the four, three have roads that are not on the fan and one climbs the fan at a distance from the entrenched channel. Of the seven cases where roads were relocated, six now have crossings at the apex and the seventh is higher on the fan in a reach with uniform gradient and an entrenched channel. Roads cross below the general zone of erosion and deposition on 12 fans. Six are Nil impact fans. Four are P2 impact fans with no impacts to the roads. Two are PI impact fans with one having limited impacts to the road and one having major impacts. Overall, roads that crossed low on fans posed no problems in 10 of 12 cases. Roads dropped to the stream crossing and then rose again on seven fans. Four are Nil impact fans. One is a P2 impact fan with localized road erosion due to an inadequate drainage structure. One is a PI impact fan with limited road erosion. One is a PI fan on which the descending road diverted sediment back toward the stream channel, limiting impacts below the road. Roads were overlanded on two Nil impact fans (built above the fan surface with imported material). In both cases adequate cross drains were present. Roads conformed to old channels with rolling grades on five fans of which three were Nil impact fans and two were PI impact fans. The extent of impacts to roads was limited due to the rolling grades on the two PI impact fans. No ditchlines were present on seven fans of which five were Nil impact fans and two were PI impact fans. One PI impact fan had a climbing road but the debris flow was directed off the road by a gentle outslope. On the other PI fan, water was diverted approximately 500 metres down the road (a non-deactivated trail from 1940s logging). 7.3.2.2 Logging Riparian zones were completely clear-cut logged on 24 of the 40 impacted fans. Damage to plantations from broadcast sediments or channel widening occurred on 23 of the 24 fans. The 122 one non-impact situation had an entrenched channel. Logging-related damage to streambanks such as gouging or scalping was not observed on the study fans. No situations were observed where the stream channels on fans were used as skid trails or roads (i.e., logging was not observed to directly impact stream channels). Natural large woody debris was removed from streams during the logging of two fans. One stream had extensive channel cleaning and the fan was classified as a PI due to impacts to the drainage structure, road, plantation and fish habitat. The other stream had localized channel cleaning and the fan was classified as a P2 due to impacts to the plantation. Large woody debris in the streams on three fans was burned during broadcast burning. All three were classified as PI impacted fans due to impacts to plantations, fish habitat, roads and drainage structures. Landings were close to stream channels on two fans. One was a Nil impact fan that was logged in 1999 and had not sustained a characterizing event. The second fan was logged without a riparian reserve in the mid-1950s. After logging, a 160-metre wide swath adjacent to the stream was cleared of stumps and woody debris by high-power stand-level floods. A portion of this cleared area was used as a dry-land sort in the mid-1990s. Landslides associated with forestry activities were present in the watersheds of three fans. Landslides in one case were road-related and located near the mouth of the watershed. The large amount of sediment that discharged to the stream channel contributed to the PI designation level of the fan. In two other cases, the amount of sediment from landslides in logged areas did not appear to be significant given the evidence of high levels of natural sediment sources. Skid roads and haul roads within logged areas became stream channels on seven fans. The resulting disturbance was major on three fans and limited on four fans. Large woody debris was left scattered on-site on eight fans. This practice was prescribed on five fans to trap broadcast sediment and provide obstructions to concentration of flow. Stumps 4 m 123 high were deliberately left in the special management zone beyond the riparian zone on two fans logged in 1999 and 2000 to provide future recruitment of large woody debris to the fan surface. Partial cutting was undertaken on four fans. Two of the fans were selectively logged in the 1940s and 1950s for cedar poles. One fan was logged in 1986 in a zone of relatively low hydrogeomorphic activity. Logging on the fourth fan was conducted in 2000 as an experiment to determine the feasibility of partial cutting on an active low-power flood fan. This fan had not sustained a characterizing event. Regardless of the silvicultural system used, ground disturbance was low on all of the logged fans. 7.3.3 Impacts Road access problems that were short-term in nature occurred on 21 fans, with access re-established within a week after repair of drainage structures and roads. Long-term access problems (greater than a week) occurred on 12 fans that required installation of new structures, major road reconstruction or road relocation. Drainage structures were replaced on 23 fans due to damage from hydrogeomorphic events. Roads were relocated from mid-fan to apex locations on seven fans. Engineered structures were installed on four fans to protect roads and major drainage structures. Natural hydrogeomorphic events were aggravated by forestry activities on fans with resulting impacts to commercial forest stands and plantations. Commercial forest stands were affected by sediment deposits on seven fans, with impacted areas ranging from 1 to 5 hectares. Plantations were affected on 25 fans by sediment deposition or erosion, with impacted areas ranging from 1 to 15 hectares. Fish habitat was directly affected on 15 fans. Twelve streams sustained major damage and three streams sustained moderate damage. Major damage involved loss of habitat complexity and stability (e.g. significant channel length with unstable substrate, wide and shallow flow, and de-124 watering during low flow periods due to sediment accumulation). Moderate damage involved more localized and less dramatic loss of habitat complexity and stability. 7.4 DISCUSSION Forestry activities exacerbated the effects of natural hydrogeomorphic events on 40 of the 55 study fans. Drainage structures that were too small to pass hydrogeomorphic events were the cause of problems on 30 fans. Roads channelled sediments and water on 24 fans. Riparian logging led to channel widening and increased spread of sediments on 23 fans. Damage to roads occurred on 32 fans, plantations on 28 fans, drainage structures on 23 fans, and fish habitat on 15 fans. It is clear that current forest management practices on fans are not sustainable. It is also clear that the re-construction and relocation of roads and drainage structures is a costly undertaking. Results from this study provide guidance for improved forestry planning and operations. 7.4.1 Pre-logging aerial photographic features Multiple channels indicating distributary stream systems were present on 16 study fans. Roads crossed seven fans in the single channel zone and nine fans in the multiple channel zone. Major impacts to roads and drainage structures resulted on five fans with multiple crossings when upstream diversions changed the proportion of flow in the channels and drainage structures where inadequate to accommodate the additional water, sediments and woody debris. Given appropriate circumstances (e.g., logjam formation) the same situation could develop on three other fans with multiple crossings. One road that crosses in the zone of multiple channels is most likely "safe" from impacts because it is located low on the fan, where there are multiple small channels and the gradient is very low. Multiple channels indicate a major hazard if roads cross below the distributary point. It is apparent that drainage structures and roads in this zone must be designed to accommodate the total flow with allowance made for bedload and woody debris. 125 Evidence of high sediment loads in the stream channels on fans and in the contributing watershed includes mid-channel bars, braided reaches, fans building into water bodies or across valleys, and changes in channel form downstream of the confluence of the fan stream with a larger stream. Such evidence was present in 10 cases. In all cases, major impacts were sustained as a result of forestry activities in combination with natural hydrogeomorphic events. Drainage structures were under-designed or had multiple spans in six cases. In-filling of the channel in reaches upstream of forestry activities led to avulsions that impacted roads in three cases. Riparian logging on two fans removed the hydrogeomorphic role of vegetation, leading to channel widening and a more extensive broadcasting of sediments. It is clear that a major hazard for forestry activities exists when there is evidence of high sediment loads in stream channels. When stream channels on fans abruptly disappear from view on aerial photographs, it indicates that the channel has lost confinement and has developed multiple smaller channels or that flows are being broadcast. This situation was found on seven fans. Roads crossing the fan below the point of "disappearance" sustained major damage by broadcast flows on two fans (#33 and 39). A n inadequate riparian zone on two disappearing channels led to sedimention in plantations (#27 and 28). The hazards of multiple channels were recognized at the forestry planning stage on three fans and no impacts were sustained. In one case the road was constructed on the slope adjacent to the fan and the stream was crossed at the apex (#21). One fan had a narrow seasonal road with a rolling grade that conformed to old channels and was deactivated immediately after logging (#26). Another had a permanent road with no ditchline and three drainage structures, each of which was large enough for all the flow and sediment (#40). In one case, an older road crossed near the toe of the fan, well below the point of disappearance in a zone with gentle gradient and dispersed flow (#29). "Disappearing" stream channels represent a high hazard for roads that requires careful field inspection for evidence of multiple channels and broadcast flow. Abrupt stream angles on fans are generally not due to bedrock control, rather they are typically caused by logjams. The hazard for forestry activities is that the jams may break, allowing water 126 to flow down the original channel. Abrupt stream angles were found on three fans, all of which sustained impacts. An avulsion through a narrow neck led to the failure of a drainage structure on one fan (#13). In two cases the abrupt angle was at the point of departure for multiple channels. In one case logjams formed on two channels, diverting a major portion of flow down a third, very old channel that was the location of a landing and a road (#33). In another case a log jam breached, sending all of the water down one channel, removing a bridge (#20). Abrupt stream angles on fans represent a high hazard for forestry activities. Major sediment sources that are visible on aerial photographs and are located near the mouth of a watershed (lower 25% of the watershed) can provide direct delivery of sediments and debris to a fan. Due to their location, there can be limited opportunity for channel and bank sediment storage or gradual routing of sediments to the fan. Sediment sources were found near the mouths of 17 watersheds. Management issues were present on 15 of the fans, of which 12 were major impacts. High sediment loads led to damage of drainage structures and roads. No issues were present on one fan where the road crossed below the deposition zone of flood events, or on a fan that had recent forestry activity and no characterizing events. Major sediment sources near the mouths of watersheds represent a significant hazard for forestry activities. 7.4.2 Characterizing events Characterizing events were found to have occurred within the past 50 years on 80% of the fans with forestry activity, with a range of one to 18 events (based on dendroecology sampling). Characterizing events occurred on 83% of the impacted fans while only 50% of the non-impacted fans experienced characterizing events. This difference could be partially due to the more recent forestry activity on many non-impacted fans. The high number of fans with characterizing events suggest that the hydrogeomorphic processes are not rare occurrences, rather quite frequent given that the oldest forestry activity occurred only 50 years ago. A similar observation on frequency was made in a study of hydrogeomorphic events in mountain streams in southern British 127 Columbia (Jakob and Jordan 2001). It follows that information about characterizing hydrogeomorphic events is required for planning forestry activities. 7.4.3 Forestry activities 7.4.3.1 Drainage structures Inadequate drainage structures were associated with 73% of the impacted fans and 54% of the 55 fans with forestry activities. Inadequate means that the structures were too small to pass water, bedload and debris, or their geometry changed channel hydraulics (e.g., confining a wide stream into a high, narrow structure) causing a localized increase in scour potential and subsequent downstream deposition (Figure 7.6). The B C Forest Practices Code requires an engineered design for structures on streams with a flow of greater than 6 m3s 1 (Anon. 1995c). Commonly, the design flood for a stream is determined using one or several methods: rainfall frequency-Figure 7.6. The maximum width of this structure is 8.4 meters while the channel is 15 meters wide. The resulting change in channel geometry can lead to scour and subsequent downstream deposition. 128 duration analysis, regional analysis of streamflow data, or extrapolation from gauged streams. Estimates of bedload and woody debris volumes are added to the flow calculation in some cases to determine the specifications for a drainage structure. However, debris flows and debris floods are not generally considered. For streams less than 6 m-V, site measurements of gradient and cross-sectional area (using indications of high flow such as sediment deposits) are used by foresters or technicians to determine drainage structure size. Site observations regarding debris flows and debris floods are not generally made. Given that hydrogeomorphic events are frequent occurrences in the study streams, it is not surprising that drainage structures designed for flood water alone often fail. Jakob and Jordan (2001) describe a similar situation in southern British Columbia. They determined that hydrogeomorphic events could be one to two orders of magnitude greater than calculated floods. Their recommendation was to include a geomorphic approach when designing drainage structures. This study provides the information to apply a geomorphic approach: determination of dominant process and its power, extent of disturbance and number of events in the past 50 years. A series of issues were identified with regard to the location of roads and drainage structures on fans. Roads that cross at the apex had no issues associated with drainage structures. At this location, the channel is generally confined by hillslopes or bedrock if the channel is entrenched. The confinement limits lateral movement and undermining of drainage structures. Of note, the majority of relocated roads now have crossings at the apex. Roads that cross in mid-fan locations had few problems if the stream reach had a uniform gradient. When the gradient decreased from steep to gentle in the area of crossing, there was always a problem related to bedload aggradation. Attempts to confine streams with dykes and maintain bedload movement through a structure were generally unsuccessful. In some cases the dykes sustained considerable damage because they were not engineered and were constructed of local (fan) materials. Even where dykes were engineered they were generally ineffective at maintaining bedload movement. Roads that crossed 129 low on fans, or below the zone of erosion and aggradation generally had very few problems, although roads did not always have adequate drainage structures or armouring to accommodate changes in the proportion of flow down multiple channels. When drainage structures fail it was common for water to be diverted out of the channel and down the path of least resistance. Every case involving a drainage structure failure with a climbing road resulted in the stream flowing down the ditchline or road (20 fans) (Figure 7.7). Several design and construction features limited the damage to climbing roads when main stream drainage structures failed. In one situation, an outsloped road with no ditch allowed water to cross with very limited damage to the road. Rolling grades that conformed to old stream channels diverted water from two climbing roads. Effective ditchblocks and back-up drainage structures limited damage in one case. Figure 7.7. Excavation into the stream channel on this fan was not stabilized with rip-rap. The result was retrogression of the cut face 50 metres upstream with the sediments plugging the culvert and leading to major impacts to the road (200m eroded to a depth of 1.4m) (22Shedin). 130 When climbing roads do not have ditches it is critical to have rolling dips or gentle outslopes so water is not channelled. These features were not present in two cases and the roads channelled flows for almost a kilometre causing erosion of the road and significantly altering streamflow patterns. Three features of ten climbing roads led to no impacts: crossing low on a fan; effective deactivation prior to the first seasonal runoff events; and providing extra maintenance attention (frequent inspections during high runoff situations and immediate action if required, as compared to one annual inspection, generally after high runoff). Roads that drop in elevation to stream crossings in both directions had no or limited erosional impacts, even if the drop was just over a short distance. In cases where the drainage structure failed, erosion was localized. In one case an avulsion occurred above a descending road, water and sediments were channelled back toward the original stream channel. While the road sustained a minor degree of erosion, there were no impacts to a plantation below the road. Channel excavations were made in three cases to allow for drainage structure placement. In all cases, rip-rap was not placed to support the abrupt channel bed drop, and headward erosion of the cut face produced sediments that filled drainage structures. In one case the headward erosion was arrested when buried logs 50 metres up the channel acted as knickpoints, however the eroded sediments overwhelmed the drainage structure and the diverted water eroded 200 metres of a road to a depth of 1.5 metres. In the other two cases, sediments from the headward erosion plugged the drainage structures with every large flow (several times a year), and occasionally lead to complete road closures. Engineered structures are clearly required. All channel excavations require the placement of effective material to act as knickpoints. 7.4.3.2 Riparian logging Riparian zones were logged on 24 of the 40 impacted fans. As no direct streambank damage was noted and there was no evidence that stream channels were used as skid trails, the impacts were explored from the perspective of tree removal. Damage to plantations from broadcast sediments 131 or channel widening occurred on 23 of the 24 impacted fans with riparian logging (Figure 7.8). The one non-impact situation had an entrenched channel. Riparian logging was the second highest forestry activity associated with impacted fans, equal with roads and ditches, but less than drainage structures. It was generally clear that roads, ditches and drainage structures caused natural hydrogeomorphic events to have greater than expected impacts to fan surfaces. With riparian logging this causal relationship was not always clear. Detailed site measurements of logged and unlogged riparian zones were not undertaken in this study. Rather, general observations were made on the nature of disturbances related to hydrogeomorphic processes in natural and logged riparian zones, and qualitative and process-based comparisons were made between the two. Channel widening and/or broadcasting of sediments was observed on 23 of the 24 study fans with riparian logging. The exception occurred on a fan with an entrenched stream channel. Figure 7.8. Pre- and post-harvesting aerial photographs illustrating the effects of logging the hydrogeomorphic riparian zone. There has been a significant increase in the area impacted by sediment deposition (Big Wdnl). Left photo 3765 5241-3 taken in 1968, right photo SRS5856 #276 taken in 1997, both at a scale of 1: ~11 000. Photos with permission. 132 Some cases of channel widening involved entrenchment with subsequent bank erosion, since stumps were found in the channel. It is possible that given the same entrenchment situations on forested fans, trees would have been in the channel. If that were the case, the large woody debris would have offered some resistance to flow, and hence reduced erosion, and provided structures for sediment storage. This scenario would possibly produce less channel entrenchment and bank erosion although the total area with sediment deposition could be greater. Fans with both logged and unlogged riparian areas appeared to have a greater amount of channel widening in the logged reaches. Logged riparian zones generally did not have large woody debris on the ground to act as sediment storage elements. Stumps offered only a limited amount of storage capacity until they were buried. Trees that could retain transported woody debris and form support for sediment levees were gone. With limited woody debris, there are few obstructions that tend to disperse broadcast flows and greater opportunity for concentration of flow exists, leading to erosion of new channels. Forests in unlogged riparian zones stored sediment storage and obstructed flows, causing the stream to return to, or stay in, the original channel. The observations in unlogged riparian zones support the conclusion that logging in riparian zones has led to an increase in the extent of broadcasting of sediments. To complete the discussion on riparian reserves it is necessary to address the question: what is an appropriate riparian reserve on a fan from a hydrogeomorphic perspective? To a degree the answer is simple: maintain the forest where it is performing a hydrogeomorphic role and where it is being directly influenced by events. Evidence includes sediment storage by logs and trees, scars on stems, growth effects as a result of events, and the presence of cohorts established on sediment deposits connected to the channel. A time frame for placing importance on observations is present in Table 7.14. The time frames are based on the nature and longevity of the evidence and on silvicultural obligations, rotation lengths, and criteria for drainage structures. 133 Table 7.14. A forestry time frame for the significance of hydrogeomorphic disturbances on fans. Time period Category 1 - 50 years Significant 50 - 100 years Important 100 - 200 years Moderate 200+ years Low Several other points require attention regarding hydrogeomorphic riparian zones. It is necessary to maintain the recruitment of L W D in reaches where this material is an integral component of channel architecture. In situations where the riparian vegetation is not being influenced by hydrogeomorphic processes it may be a key element in maintaining stream bank integrity. Such situations include when the stream channel is elevated above the surrounding fan surface (a relatively common situation) or when the stream banks are relatively low and tree roots are reinforcing the soil mass. A central issue with hydrogeomorphic riparian reserves is the need to locate a reserve on a specific fan based on site information. It is inappropriate to use a fixed width such as 30 metres. This can be wider than necessary where a stream is confined by debris flow levees or entrenchment, or much too narrow in situations where broadcasting sediments and distributary channels are present for 100 metres from the main channel. Since the intention of a reserve is to function over a long period of time, a management zone may be required to ensure that the reserve is not subject to abnormal blowdown. Several study fans had reserves blow down and it could be argued that the resulting bedload aggradation presented a worse situation (i.e., avulsion) than if the forest was logged to the bank. A final comment is that riparian reserves should be considered on fans regardless of whether fish are present or not. The objective is to maintain the 134 forest on a landform that has zones of hydrogeomorphic activity where the influence of the forest is providing a degree of stability. Large woody debris (LWD) is recognized as an important element in stream channels (Thomson 1991). This material was removed or reduced in abundance by forestry activities on four fans, one with "stream cleaning" of natural LWD, two with prescribed burning and one with burning and cleaning. Two of the prescribed burns were broadcast with no indication that attempts had been made to protect the streams. One case involved an escaped prescribed burn, although no riparian reserve was left on the logged side. Major impacts were sustained on three fans and the fourth was limited only because the cleaning was limited. Characteristically, impacted channels were entrenched in some reaches and other reaches were wide, shallow, avulsion-prone channels behind debris jams. Dates of logging on these fans ranged from 1979 to 1992 and reflect to some degree the "practices of the day". Few logged areas are now burned and only isolated cases of stream cleaning occur. However, the dramatic impact of LWD removal is perhaps an indication of what can be expected in several decades in reaches with no riparian reserves (Bilby and Ward 1991). The recruitment of LWD in these channels will be limited to second-growth material that in most cases is significantly smaller in size than the original old-growth stems (Bilby and Ward 1991; Gomi et al. 2001). 7.4.3.3 Soils and mass wasting Generally, logged fans had low levels of significant ground disturbance (e.g., rutting) and there was not an obvious interaction with hydrogeomorphic processes (e.g., channelling of surface flows or interception of subsurface flows). One reason for this finding is that most fans had coarse-textured, well-drained soils with a low susceptibility for rutting. Given this texture, it is fortunate that rutting was not common as the low soil cohesion can lead to rapid, deep erosion (e.g., as observed on some skid trails). 135 Most logged study fans had landings. Upon completion of logging, landings are generally left to reforest, with no attempt made to re-contour levelled areas or replace large woody debris on the cleared areas. Thus, if a hydrogeomorphic event entered a landing there would be no obstructions to impede the flow. However, no cases were found to support this concern. Only two study fans had landings in close proximity to the stream channel. One fan was recently logged and has not experienced a characterizing event. The second fan had a landing established after a high-power stand level flood created a 160 metre wide zone. That fan was logged without a riparian reserve in the mid-1950s. Forestry activities on unstable hillslopes increase the frequency of mass wasting (Sidle et al. 1985). However, very few of the study fans had logging in their watershed, and only three watersheds had landslides associated with logging or road construction. In two cases the landslides were minor sources of sediment, but in one case the sediment production was significant and most likely contributed to a major channel avulsion. 7.4.3.4 Silvicultural systems Most study fans were clearcut logged, however four fans had partial or selective cutting. Two of these fans had evidence of active sediment broadcasting. One was logged approximately 50 years ago and showed no evidence that the logging exacerbated hydrogeomorphic processes on the fan. Abundant large woody debris was available for sediment storage and skid trails were not apparent. However, the haul road that ran up the fan had not been deactivated and intercepted and channelled broadcast flows and was eroded up to 2 metres deep for several hundred metres. The second partial cut fan was logged in 2000 and has not had a characterizing event. However, it is likely that the logging will have limited impacts as abundant woody debris was left in place, logging was conducted with a 1-metre snowpack that limited skidding disturbance, no mechanical site preparation was undertaken, and advanced regeneration was largely left undamaged. 136 7.4.3.5 Utilization standards The general trend in forestry in British Columbia over the past decade has been to limit the amount of woody debris left on-site. The change in utilization standards occurred at the same time as forest ecologists, hydrologists, aquatic biologists, and wildlife biologists were recognizing the significance of coarse woody debris for biodiversity. At the present time, utilization standards take precedence over requirements for coarse woody debris (Anon. 1995b). The implications of utilization standards on fans, particularly where woody debris is performing a hydrogeomorphic role, could also be significant. Fortunately there appears to be some flexibility as long as a prescription identifies the need to leave woody debris. LWD was left on site, by prescription, on five recently logged fans. The material was left in areas that were subject to broadcasting of sediments. In addition to downed woody debris, stumps up to four metres were left on two fans to provide future recruitment of large woody debris. 7.4.3.6 Recognition of impacts Recognition of forestry-related impacts on fans are mixed. In many cases the fact that an activity is located on a fan is not recognized (i.e., issues are not placed in a landform context). However, impacts to roads and drainage structures are recognized as costly in terms of repairs, relocations and road closures. Foresters and engineers showed a high level of interest in this project as they wanted to reduce the number of incidents that have proven costly. Recognition of impacts to forest sites and plantations is generally low. This could in part be due to the generally limited extent of impacts on most fans (less than 2 hectares on 22 of 32 fans). Impacts to fish habitat are generally recognized as inappropriate. However once habitat is impacted on a fan there are limited options for restoration and sites are generally left to rehabilitate naturally (a process that could take decades). Two study fans with impacts to fish habitat were investigated by Fisheries and Oceans staff, however court action did not result. In the future it is likely that more fans with impacts to fish habitat will be investigated for prosecution under either the Federal Fisheries Act 137 or the new BC Forest Practices Code (T. Pendray, pers.com., Senior Habitat Biologist, Fisheries and Oceans Canada, 2002). 7.5 CONCLUSIONS This study has demonstrated that on the majority of study fans, forestry activities are exacerbating the impacts of natural hydrogeomorphic processes. Forestry activities on fans require closer attention in order to be more cost effective and less damaging to both forest and fisheries resources. Both appropriate and inappropriate prescriptions have been described. For forestry planning purposes, specific watershed and fan features were identified that represent hazards for forestry activities. In addition, the previous chapters have identified that for the study area, the power, disturbance extent and frequency of hydrogeomorphic processes can be forecast from simple watershed measurements. Application of this information requires fieldwork to develop tailored prescriptions for each fan. Results presented in this chapter represent the first comprehensive examination of forestry activities on fans in BC and clearly identify appropriate sustainable forest management practices. 138 C H A P T E R 8. CONCLUSIONS This study was initiated because conventional forest practices on fans were considered to exacerbate the disturbance resulting from natural hydrogeomorphic processes (floods, debris floods and debris flows). Conventional practices include clearcut logging, roads with ditches, drainage structures designed to pass 50- or 100-year floods, and riparian reserves only on fish streams or streams that flow directly into fish streams. This study found clear evidence that conventional forestry practices exacerbate the disturbance caused by hydrogeomorphic processes (Chapter 7), and points to the need for hazard recognition in the planning phase of forest management. The conventional planning process identifies unstable terrain (initiation zones but not run-out zones) and neither a hazard classification nor a summary of practices is available for forest management on fans. The approach taken in this study was that forestry staff should be empowered to identify hazards by focusing on features that can be operationally measured and recognized by foresters. For this reason, a hazard classification was proposed using forest stand characteristics, site features, and basic watershed attributes (e.g., area, relief). The classification was tested on 65 fans and their watersheds in west central British Columbia and determined to be statistically valid. Consequently, I conclude that natural forest stands on fans can be used as indicators of the power, disturbance extent and frequency of natural, contemporary hydrogeomorphic processes. The forest stand-based hazard classification has three basic elements: power, disturbance extent and hydrogeomorphic process. Power refers to the ability of a hydrogeomorphic event to clear a forest. High power events clear forests and result in clearings (disturbance extent) that are wider than 20 m (stand level) or less than 20 m wide (site level). Low power events spread sediment under a forest stand and do not result in the physical removal of trees. There are three hydrogeomorphic processes: floods, debris floods and debris flows. Thus the classification has 9 potential categories. However during field classification of the 65 study fans, no low-power debris flows were described, thus the forest stand-based classification has eight categories. 139 Initially an attempt was made to determine if just power and disturbance extent (three categories) could be used to identify hazards; however, no statistically significant watershed attributes could be identified. This is reasonable considering that the three different hydrogeomorphic processes are generated from significantly different watersheds (Chapter 4). As a result, the fans were grouped into processes and differentiating watershed attributes were identified using analysis of variance (Table 5.9). The Melton ratio was used to differentiate flood watersheds from debris flood and debris flow watersheds, as has previously been done by Jackson et al. (1987) working in the Canadian Rocky Mountains of south western Alberta and south eastern British Columbia.. The Melton ratio and watershed length were used to differentiate debris flood and debris flow watersheds. This is a key scientific contribution of the study as no literature citations were found for differentiating the watersheds of these two processes. Statistically significant watershed attributes were identified for the power and disturbance combinations for the hydrogeomorphic processes (Table 5.10). This is a fundamental finding that demonstrates that forest stands on fans can be used as indicators of the power and disturbance extent of natural, contemporary hydrogeomorphic processes (contemporary being within the last 200 years). Dendroecology techniques were used to determine the number of events that had occurred on 59 of the study fans. In addition to the commonly used techniques of dating scars, cohorts (groups of trees established on the sediment from an event) and reaction wood (e.g., Jakob 1996), an additional technique of dating the commencement of growth responses (primarily suppression but also release) was also used to identify the occurrence of hydrogeomorphic events (Strunk 1997). A significant finding was that hydrogeomorphic events had occurred in all 58 watersheds during the past 50 years, with all but three watersheds having at least one characterizing event (the event that led to the classification of fan, for example high power stand-level debris flow). This finding supports the observation of other researchers that hydrogeomorphic events are not necessarily extreme or rare events (Innes 1985; Jakob and Jordan 2001). This is an important finding, because there is an assumption amongst some forestry staff that fans are essentially "fossil" 140 features, having formed in paraglacial times and no longer subject to modification. This study, together with an increasing number of others, has shown that fans in mountain areas can be very active, even if the bulk of their volume developed in the early part of the Holocene. Linear regression analysis was used to develop models for predicting the number of events in the past 50 years. Only one category had a sufficient number of observations to allow for two independent variables (watershed attributes). The analysis identified seven statistically significant models (the two high-power flood categories were combined because of the limited number of observations) (Table 6.2). All but one of the models have one independent variable due to the small sample sizes. Two of the models use forest cover attributes (Comm is the extent of commercial forest cover and includes mature and immature forests, and areas that are not satisfactorily restocked; ESASx refers to areas that are the initiation zones of natural mass wasting). While forest cover in a watershed has been referred to in the literature as a variable that influences the occurrence of hydrogeomorphic events, no examples were found where forest attributes were variables in predictive frequency models. The use of these forest attributes to describe the frequency of hydrogeomorphic events is thus novel and marks a contribution to the science. Forestry activities had been conducted on 55 of the study fans, with practices ranging from a single road traversing a portion of a fan to complete clearcutting including the riparian zone (no or very limited forestry activities were present in the watersheds of the study fans). Conventional forest practices on 75% of the study fans exacerbated the impacts of hydrogeomorphic processes, resulting in channel avulsions and broadcasting of sediment. The leading cause of impacts was drainage structures that were too small to transfer the sediment and debris loads carried by hydrogeomorphic events (Table 8.1). The blocked structures led to the diversion of streams (avulsions). Roads on fans, primarily roads that climb to stream crossings, were the second most common cause of impacts as a result of dispersion of flows across a fan surface. Approximately half of the study fans had their riparian zones clearcut, and in almost all cases this led to impacts 141 T a b l e 8.1. Forest practices and impacted features associated with hydrogeomorphic events. Percentages are presented by the total in each category. Total in each category Impacts identified Total fans with forestry activities and total impacted fans by impact class 55 100% 41 74.5% Associated forestry practice Drainage structure / Channel excavation 55 (100%) 30 (54%) Roads / Ditches 53 (100%) 24 (45%) Riparian fogging 24 (100%) 23 (96%) LWD removal or burned 4 (100%) 4 (100%) Mass wasting related to forestry activities 3 (100%) 1 (33%) Impacted features Roads 53 (100%) 32 (60%) Plantations/forest sites 37 (100%) 28 (76%) Drainage structures 55 (100%) 23 (42%) Fish habitat 26 (100%) 15 (58%) due to the lack of forest influence: channel confinement, reduced storage capacity for sediment, and reduced capacity to resist the erosion of new channels. The exacerbation of natural hydrogeomorphic impacts on fans resulted in damage to roads (60%), plantations (76%), fish habitat (58%) and drainage structures (42%). I conclude that conventional practices, although complying with the BC Forest Practices Code, are inappropriate, as they do not account for hydrogeomorphic activity on fans. Although there is a requirement to identify the initiation zones of mass wasting, this is not the case for runout zones, which would include debris flood and debris flow fans. While the requirement for drainage structures is that they pass a 50- or 100-year flood event, not enough attention is being paid to sediment and debris associated with these events, and it is apparent that very little attention is being paid to debris floods or debris flows. 142 An observation that I made during the description of natural forests on fans was that the riparian forest was storing sediment, maintaining the channel location, and reinforcing the fan surface and thereby increasing resistance to the formation of avulsion channels. This zone has a series of characteristic features that includes scarred trees, buried trees, log steps, log retaining walls, and tree holes. It is very active and, given that these features can be easily recognized and given the important role of the riparian forest and associated woody debris, I have called this zone the hydrogeomorphic riparian zone. The width of the zone ranges from a few metres to 100s of metres, and does not generally conform to the BC Forest Practices Code definition of riparian zone, which is limited to a 30 m wide strip along the banks of fish-bearing streams. The importance of this zone became apparent in the review of fans with forestry activities. Where the zone was clearcut there was an apparent increase in the area covered with sediment splays and in the occurrence of avulsions. This zone also provides information on the power, disturbance extent, and frequency of hydrogeomorphic events, information that is directly required for the design of drainage structures. It is clear that attention must be paid to this zone if natural stability levels on fan surfaces are to be maintained. This zone has not previously been recognized in the literature. A key aspect of the forest practices review was the identification of appropriate and inappropriate prescriptions (Tables 8.2 and 8.3). This is the first hydrogeomorphic synopsis of practices on fans and represents a significant contribution to forest management. Extension of the results to forestry professionals was an integral component of this study. Presentations where made at six conferences, field training session were held within the study area and at four locations around the province, two technical articles were published, and numerous consulting projects were completed. There was enthusiasm to hear the results of this study and apply them operationally. An apparent impetus for the interest is the desire by most foresters to manage forests in a sustainable manner. It is clear that conventional practices fall short in two important criteria: (1) maintenance of forest ecosystem health and vitality and (2) conservation 143 Table 8.2. A summary of hydrogeomorphically appropriate forestry prescriptions on fans. Prescription Summary comment Road crossing at apex Seasonal deactivation Adequate permanent drainage structures Engineered structures Special maintenance attention for drainage structures Road crosses stream in uniform gradient reach Roads with special design features that climb to stream crossings Roads drop to stream crossings Roads cross below the zone of erosion and deposition Overlanded roads No ditchlines Rolling grade in road Retention of large woody debris This location avoids most problems from hydrogeomorphic events. Of structures not designed to withstand hydrogeomorphic events. Designed to withstand hydrogeomorphic events. Designed to maintain roads or drainage structures, but should maintain channel hydraulics and fish habitat. For drainage structures that are high risk because of hydrogeomorphic events or their location on a fan (e.g., zone of deposition due to a channel gradient break). To avoid the problem of sediment aggradation in the drainage structure. Roads that climb to a stream crossing are inappropriate unless there are special design features: drainage structures designed to accommodate hydrogeomorphic events, rolling grades, crossing low on a fan, regular maintenance of drainage structures, effective deactivation, and no ditchlines. The optimum profile for a road crossing a stream. A location that avoids aggradation or degradation in the drainage structure. Avoids problems with ditchlines but requires adequate cross drainage. Avoids interception of broadcast and intercepted subsurface flows, but road must have armoured rolling dips or cross drains. Uses old stream channels to allow water to cross the road rather than intercepting and channelling surface flows. To maintain the hydrogeomorphic role following tree removal. Partial cutting Maintains a degree of the hydrogeomorphic role of the forest. and maintenance of soil and water resources (Montreal Process Working Group 1998). While this has been recognized in academic circles (e.g., Schreier 2000), it is still to be translated into practice in many parts of the world, including British Columbia. This is one of the first studies to explore forestry activities on fans and to develop an operational system for the recognition of hydrogeomorphic processes. As such, there remain many 144 Table 8.3. A summary of forestry prescriptions that exacerbate hydrogeomorphic events on fans. Prescription Summary comment Roads climb to stream crossings Roads at slope break in stream/ fan (from steep to gentle) Inadequate ditchblocks and cross drains as roads traverse fans Multi-span drainage structures Inadequate drainage structures Channel excavation - no riprap Inadequate drainage structures in multiple channel situations Non-engineered structures Roads on fans are not deactivated Logging of the hydrogeomorphic riparian zone Skid trails on fans intercept and concentrate flows Road becomes an avulsion channel if drainage structure is blocked. Drainage structures failure due to sediment aggradation. Roads and ditches become stream channels with the lack of adequate cross drainage. Interfere with bedload movement leading to sediment aggradation. Trap woody debris. Drainage structures that are too small for, or damaged by, hydrogeomorphic events, and altered channel hydraulics (wrong cross-sectional shape) leading to channel scour. Unsupported channel excavations erode upstream, leading to sediment accumulation in drainage structures. A common problem where roads cross in zones of multiple channels. At the time of construction, drainage structures are installed based on the flow volumes. Volumes can change significantly if the proportion of flow changes. Dikes and berms built of local streambed materials that can be readily eroded. Roads on fans that are not used and maintained regularly should be deactivated because roads can exacerbate natural hydrogeomorphic events. Removes the forest influence: sediment storage, maintenance of channel location and reinforcement of the soil mass to avulsions. Particularly an issue with bladed skid trails that climb toward the stream. unanswered scientific questions. Future research should be directed at providing a quantitative understanding of the hydrogeomorphic riparian zone, including the effects of partial cutting. Research to date on partial cutting has focused on stand establishment (light levels and ground disturbance) primarily on stable sites (e.g., Coates and Burton 1999; LePage et al. 2000), although there is a long history of using partial cutting to maintain root systems on unstable sites (i.e., Protection Forests) (Troup 1928). These questions are similar to a degree, but the environments are considerably different to hydrogeomorphically active areas of fans. Issues such 145 as the periodic deposition of sediments resulting in unstable conditions for the establishment of regeneration, and the need to maintain large woody debris as an integral structural stand element may require more study. Additional sampling within the study area would test the models and provide greater confidence in the predictions. Sampling in areas beyond the study area may lead to the development of additional models that account for regional differences. The degree of impacts to forest sites and fish habitat indicate the need for restoration research. A key result from this project that is directly applicable to restoration research is the frequency of natural hydrogeomorphic events influencing fans. As mentioned above, it is a common assumption that natural fans are stable landforms that are only destabilized by forestry activities. Another area requiring research is forest engineering on fans. It is important to define appropriate protocols and standards for the design of roads and drainage structures on active fan areas. Avoiding fans in mountainous areas is not always an option. Future research will strengthen the ability to predict hydrogeomorphic hazards on fans and will improve the level of forest management prescriptions. However, this study has demonstrated quantitatively that natural forest stands on fans can be used as indicators of the power, disturbance extent and frequency of natural, contemporary hydrogeomorphic processes. This study has developed a series of predictive tools and knowledge, the application of which will be a considerable step toward sustainable forest management on fans. 146 L I T E R A T U R E C I T E D Alestalo, J. 1971. Dendrochronological interpretation of geomorphic processes. Soc. Geographica Fenniae. 105: 1-140. Allen, J.R. 1965. A review of the origin and characteristics of recent alluvial sediments. Sedimentology. 5: 89-191. Alley, N.F. and Young, G K . 1978. Environmental significance of geomorphic processes in the northern Skeena Mountains and southern Stikine Plateau. BC Min. Env., Res. Analysis Br. Bulletin 3. 83pp. Anon. 1992. Forest Inventory Manual. BC Min. Forests. Anon. 1995a. Riparian management area guidebook. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 68pp. Anon. 1995b. Forest Road Engineering Guidebook. Forest Practices Code. BC Min. Forests and BC Environment. 153pp. Anon. 1995c. Biodiversity Guidebook. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 99pp. Anon. 1998. Channel Assessment Procedure. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 37pp. Anon. 1999a. Watershed Assessment Procedures. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 45pp. Anon. 1999b. Mapping and Assessing Terrain Stability Guidebook. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 36pp. Anon. 2001. Gully Assessment Procedure. Fourth Ed. Forest Practices Code Guidebook. BC Min. Forests and BC Environment. 60pp. Banner, A., MacKenzie, W., Haeussler, S., Thomson, S., Pojar, J., and Trowbridge, R. 1993. A field guide to site identification and interpretation for the Prince Rupert Forest Region. Land Management Handbook 26. BC Min. Forests. Beaty, C.B. 1970. Age and estimated rate of accumulation of an alluvial fan. White Mountains, California. Amer. Journ. of Sci. 268: 50-77. Benda, L.E and Cundy, T. W. 1990. . Predicting deposition of debris flows in mountain chan-nels. Can. Geotech. Jour. 27:409-417. 147 Beschta, R.L., Bilby, R.E., Brown, G.W., Holtby, L.B., and Hofstra, T.D. 1987. Stream tempera-ture and aquatic habitat: fisheries and forestry interactions. In: Salo, E.O. and Cundy, T.W. (eds.) Streamside Management: Forestry and Fishery Interactions. U. of Wash. Institute of Forest Resources, pp. 191-232. Bilby, R.E. and Ward, J.W. 1991. Characteristics and function of large woody debris in streams draining old-growth, clear-cut, and second-growth forests in southwestern Washington. Can. J. Fish. Aquat. Sci. 48: 2499-2508. Bisson, P.A., Bilby, R.E., Bryant, M.D., Dolloff, C.A., Grette, G.B., House, R.A., Murphy, M.L., Koski, K.V., and Sedell, J.R. 1987. Large woody debris in forested streams in the Pacific northwest: past, present, and future. In: Salo, E.O. and Cundy, T.W. (eds.) Streamside Management: Forestry and Fishery Interactions. U. of Wash. Institue of Forest Resources, pp. 143-190. Bjornn, T C . and Reiser, D.W. 1991. Habitat requirements of salmonids in streams. In: Meehan, W.R. (ed.). Influences of forest and rangeland management on salmonid fishes and their habitats. Amer. Fish. Soc. Special Pub. 19. pp. 83-138. Bovis, M.J. and Dagg, B.R. 1988. A model for debris accumulation and mobilization in steep mountain channels. Hydrological Sciences Journal 33: 589-604. Bovis, M.J. and Jakob, M. 1999. The role of debris supply conditions in predicting debris flow activity. Earth Surf. Process. Landforms 24: 1039-1054. Bull, W.B. 1964a. Alluvial fans and near-surface subsidence in western Fresno county, Califor-nia. US Geol. Survey Prof. Paper 437-A. pp. 1-71. Bull, W.B. 1964b. Geomorphology of segmented alluvial fans in western Fresno County, Cali-fornia. US Geol. Survey Prof. Paper 552-F. pp. 89-129. Bull, W.B. 1977. The alluvial-fan environment. Progress in Physical Geography. 1:222-270. Burns, R.M. and Honkala, B.H. 1990. Silvics of North America. USDA, FS Agric. Hdbk. No. 654. Vol. 1. 675pp. Vol.2. 877pp. Caine, N. 1980. Rainfall intensity-duration control of shallow landslides and debris flows. Geograf. Ann. 62A: 23-27. Carlston, C.W. 1963. Drainage density and streamflow. US Geol. Survey Prof. Paper 422-C. pp.1-8. Church, M. (In prep.). Mountains and montane channels. In: Allison, R.J. and Burt, TP. (eds.). The Sediment Cascade. J. Wiley. Chapter 2. 148 Church, M. and Miles, M.J. 1987. Meteorological antecedents to debris flow in southwestern British Columbia; some case studies. In: J.E. Costa and Wieczorek, G.F. (eds). Debris flows/avalanches: Process, Recognition, and Mitigation. Reviews in Eng. Geol. Vol. VII. Geol. Soc. Am. pp. 63-79. Clague, J.J. 1984. Quaternary geology and geomorphology, Smithers-Terrace-Prince Rupert area, British Columbia. Geol. Survey of Can. Memoir 413. 71pp. Clague, J.J., Evans, S.G., and Blown, I.G 1985. A debris flow triggered by the breaching of a moraine-dammed lake, Klattasine Creek, British Columbia. Can. J. Earth Sci. 22: 1492-1502. Coates, K.D. and Burton P.J. 1999. Growth of planted tree seedlings in response to ambient light levels in northwestern interior cedar - hemlock forests of British Columbia. Can. J. For. Res. 29: 1374-1382. Cook, E.R. and Kairiukstis, L.A. 1990. Methods of dendrochronology. Applications in the environmental sciences. International Institute for Applied Systems Analysis. Kluwer Academic Pub., Dordrecht. 394pp. Costa, J.E. 1984. Physical geomorphology of debris flows. In: Costa, J.E. and Fleisher, J.P. (eds.). Developments and applications of geomorphology. Springer-Verlag. New York, pp. 268-317. Costa, J.E. 1988. Rheologic, geomorphic, and sedimentologic differentiation of water floods, hyperconcentrated flows, and debris flows. In: Baker, V.R, Kochel, R.C., and Patton, P.C. (eds.). Flood Geomorphology. John Wiley. New York. pp. 113-122. Creuze des Chatelliers, M. , Poinsart, D., and Bravard, J.-P. 1994. Geomorphology of alluvial groundwater ecosystems. In: Gilbert, J., Danielopol, D.L., and Stanford, J.A. (eds.). Groundwater Ecology. Academic Press Inc. pp. 157-185. Dagg, B.R. 1987. Debris supply to torrent-prone channels on the east side of Howe Sound, British Columbia. M.Sc. Thesis. Dept. ofGeog. U.B.C. 243pp. Denny, C.S. 1965. Alluvial fans in the Death Valley region, California and Nevada. U.S. Geol. Survey Prof. Paper 466. Denny, C.S. 1967. Fans and pediments. Am. J. Sci. 265:81-105 Dury, G H . 1973. Magnitude-frequency analysis and channel morphology. In: Morisawa, M. (ed). Fluvial Geomorphology. SUNY, Binghamton: Pubs. In Geomorphology. pp. 91-121. Eisbacher, G.H. and Clague, J.J. 1984. Destructive mass movements in high mountains: hazard and management. Geol. Survey of Can. Paper 84-16. 230pp. 149 Ellen, S.D and Mark, R.K. 1993. Mapping debris flow hazard in Honolulu using a DEM. Hy-draulic Engineering '93, Proceedings 2: 1774-1779. Evans, S.G. 1984. The landslide response of tectonic assemblages in the sourthern Canadian Cordillera. Proceedings, IV Inter. Symp. on Landslides, Toronto. 1: 495-502. Fowells, H.A. 1965. Silvics of forest trees of the United States. USDA, FS. Agric. Hdbk. No. 271. 762pp. Freeze, R.A. and Cherry, J.A. 1979. Groundwater. Prentic-Hall Inc. Englewood Cliffs, NJ. 604pp. Gomi, T , Sidle, R.C. Bryant, M.D., and Woodsmith, R.D. 2001. The characteristics of woody debris and sediment distribution in headwater streams, southeastern Alaska. Can. J. For Res. 31: 1386-1399. Gomi, T , Sidle, R .C , and Richardson, J.S. 2002. Understanding processes and downstream linkages of headwater systems. Bioscience. 52(10): 905-916. Haupt, H.F. 1959. Road and slope characteristics affecting sediment movement from logging roads. Jour. For. 57: 329-332. Heede, B.H. 1972. Flow and channel characteristics of two high mountain streams. USDA FS. Research Paper RM-96. 12p. Hetherington, E.D. 1987. The importance of forests in the hydrological regime. In: Healey, M.C. and Wallace, R.R. (eds.). Canadian aquatic resources. Can. Bull. Fish. Aquat. Sci. 215. pp. 179-211. Hogan, D.L., Bird, S.A., and Hassan, M A . 1998. Spatial and termporal evolution of small coastal gravel-bed streams: influence of forest management on channel morphology and fish habitats. In: Kingeman, P C , Beschta, R.L., Komar, P.D., and Bradley, J.B. (eds.). 1998. Gravel-Bed Rivers in the Environment. Water Resources Publications, LLC, High-lands Ranch, Colorado, U.S.A. pp. 365-392. Holland, S.S. 1964. Landforms of British Columbia, A Physiographic Outline. Bulletin No. 48. BC Dept. of Mines and Petroleum Resources. 138pp. Hooke, R.LeB. 1968. Steady-state relationships on arid-region alluvial fans. J. Geology 75: 438-460. Hooke, R.LeB. 1967. Processes on arid-region alluvial fans. J. Geol. 75: 438-460. 150 Hooke, R.LeB., and Rohrer, W.L. 1977. Relative erodability of source-area rock types, as deter-mined from second-order variations in alluvial-fan size. Geol. Soc. Amer. Bull. 88: 1177-1182. Howes, D.E and Kenk, E. 1997. Terrain Classification System for British Columbia. MOE Manual 10. BC Min. Envir., Lands and Parks. 90pp. Howes, D.E and Swanston, D.N. 1994. A technique for stability hazard assessment. In: Chatwin, S.C., Howes, D.E., Schwab, J.W., and Swanston, D.N. A guide for management of landslide-prone terrain in the Pacific Northwest. BC Min. For. Land Manage. Handb. No. 18(2nded.). Victoria, B.C. 220pp. Hungr, O., Evans, S.G., Bovis, M.J., and Hutchinson, J.N. 2001. A review of the classification of landslides of the flow type. Environmental and Engineering Geoscience. 7(3): 221-238. Hunt, C.B. and Mabey, D.R. 1966. Stratigraphy and structure. Death Valley, California. US Geol. Survey Prof. Paper 494-A. 162pp. Hupp, C R . 1983. Geo-botanical evidence of late Quaternary mass wasting in block field areas of Virginia. Earth Surf. Process. Landforms. 8:439-450. Innes, J.L. 1982. Debris flow activity in the Scottish highlands. U. of Cambridge. Unpublished PhD thesis. 638pp. Innes, J.L. 1983. Debris flows. Progress in Physical Geogeography 7:469-501. Innes, J.L. 1985. Magnitude-frequency relations of debris flows in northwest Europe. Geograf. Ann. 67A: 23-32. Irasawa, M., Ishikawa, Y., Fukumoto, A., and Mizuyama, T. 1991. Control of debris flows by forested zones. In: Proceedings of the Japan-U.S. Workshop on Snow Avalanche, Land-slide, Debris Flow Prediction and Control. Sept. 30 - Oct. 2, 1991. Tsukuba, Japan. Sci-ence and Tech. Agency of Japanese Gov. pp. 543-550. Iverson, R.M. 1997. The physics of debris flows. Reviews of Geophysics. 35:245-296. Jackson, L.E., Kostaschuk, R.A., and MacDonald, G M . 1987. Identification of debris flow hazard on alluvial fans in the Canadian Rocky Mountains. In: Costa, J.E. and Wieczorek, G.F. (eds.). Debris flows/avalanches: Process, Recognition, and Mitigation. Reviews in Eng. Geol. Vol. VII. Geol. Soc. Am. pp. 115-124. Jakob, M. 1996. Morphometric and geotechnical controls of debris flow frequency and magni-tude in southwestern British Columbia. PhD Thesis. Dept. Geog. UBC. 232pp. 151 Jakob, M. and Jordan, P. 2001. Design flood estimates in mountain streams - the need for a geomorphic approach. Can. J. Civ. Eng. 28: 425-239. Jordan, P. 1994. Debris flows in southern Coast Mountains. Ph.D. Thesis. Dept. Geog. UBC. 260pp. Johnson, A. 1970. Physical Processes in Geology. Freeman, Cooper and Co. San Francisco. 577pp. Keller, E.A. and Swanson, F.J. 1979. Effects of large organic material on channel form and fluvial processes. Earth Surf. Process. Landforms 4: 361-380. Kellerhals, R. and Church, M. 1990. Hazard management on fans, with examples from British Columbia. In: Rachochi, A.H. and Church, M . (eds.). Alluvial fans: a field approach. John Wiley and Sons. New York. pp. 335-354. Kilpatric, F.A. and Barnes, H.H. 1964. Channel morphology of piedmont streams as related to frequency of floods. US Geol. Survey Prof. Paper 422E. pp. 1-10. Kirkby, M.J. and Chorley, R.J. 1967. Throughflow, overland flow and erosion. Int. Assoc. Sci. Hydrol.Bull. 12:5-21. Kittredge, J. 1948. Forest Influences. New York. McGraw-Hill. 394pp. Lecce, S.A. 1990. The alluvial fan problem. In: Rachocki, A.H. and Church, M. (eds.) Alluvial Fans: A Field Approach. John Wiley and Sons Ltd. New York. pp. 3-24. Leopold, L.B. and Maddock, R. 1953. The hydraulic geometry of stream channels and some physiographic implications. US Geol. Survey Prof. Paper 252. LePage, P., Canham, C D . , Coates, K.D., and Bartemucci, P. 2000. Seed abundance versus substrate limitation of seedling recruitment in northern temperate forests of British Colum-bia. Can. J. For. Res. 30:415-427 Luckman, B.H. 1992. Debris flows and snow avalanche landforms in the Lairig Ghru, Criangorm Mountains, Scotland. Geogr. Ann. 74A: 109-121. Marston, R.N. 1982. The geomorphic significance of log steps in forest streams. Annals of the Assoc. of Amer. Geographers. 72: 99-108. Melton, M A . 1957. An analysis of the relation among elements of climate, surface properties and geomorphology. Office of Nav. Res. Dept. Geol. Columbia Univ., NY. Tech. Rep. 11. 102pp. 152 Melton, M.A. 1965. The geomorphic and paleoclimatic significance of alluvial deposits in southern Arizona. J. Geol. 73: 1-38. Miles, M.J. and Kellerhals, R. 1981. Some engineering aspects of debris torrents. Proceedings of the Can. Soc. of Civil Eng., 5th Can. Hydrotechnical Conf. Fredricton, New Brunswick, pp. 395-420. Montreal Process Working Group. 1999. Criteria and indicators for the conservation and sus-tainable management of temperate and boreal forests. (2nd ed.). 27 Jan. 2003. http://www.mpci.org/rep-pub/1999/ci_e.html Mukerji, A.B. 1990. Terminal fans of inland streams in Suglej-Yamauna plain, India. Zeit. F. Geomorph. 20: 190-204. Murphey, J.B., Wallace, D.E., and Lane, L.J. 1977. Geomorphic parameters predict hydrograph characteristics in the southwest. Water Resour. Bull. 13: 25-38. Murphy, W.L. and Meehan, W.R. 1991. Stream Ecosystems. In: Meehan, W.R. (ed.). Influences of forest and rangeland management on salmonid fishes and their habitats. Amer. Fish. Soc. Special Pub. 19. pp. 17-46. Naiman, R.J., Bilby, R.E., and Bisson, P.A. 2000. Riparian ecology and management in the Pacific coastal rain forest. Bioscience 50(11): 996-1011. Nilsen, T.H. 1982. Alluvial fan deposits. In: Scholle, P.A. and Spearing, D. (eds.) Sandstone depositional environments. Amer. Assoc. of Petroleum Geologists. Memoir 31 pp. 49-86. Nussbaum, A.F. 1996. Site index curves and tables for British Columbia: coastal species. 2nd edition. BC Min. For. Res. Br. Land Manage. Handb. Field Guide Insert 3. Ogrosky, H.O. and Mockus, V. 1964. Hydrology of agricultural lands. In: Chow, V.T. (ed.). Handbook of Applied Hydrology. McGraw-Hill Book Co. New York. Chapter 21: 1-97. O'Hara, K.L. 1998. Silviculture for structural diversity: a new look at multiaged systems. Jour. For. 96(7): 4-10. Oliver, C D . 1981. Forest development in North America following major disturbances. Forest Ecology and Management. 3: 153-168. Oliver, C D . and Larson, B.C. 1996. Forest Stand Dynamics. John Wiley and Sons, Inc. New York. 520pp. 153 Osterkamp, W.R. and Hupp, C R . 1987. Dating and interpretation of debris flows by geologic and botanical methods at Whitney Creek Gorge, Mount Shasta, California. In: Costa, J.E. and Wieczorek, G.F. (eds.). Debris flows/avalanches: Process, Recognition, and Mitigation. Reviews in Eng. Geol. Vol. VII. Geol. Soc. Am. pp. 157-163. Patton, P C 1988. Drainage basin morphometry and floods. In: Baker, V.R., Kochel, R .C , and Patton, P C (eds.). Flood Geomorphology. New York. John Wiley, pp. 51-64. Patton, P C and Baker, V.R. 1976. Morphometry and floods in small drainage basins subject to diverse hydrogeomorphic controls. Water Resour. Res. 12: 941-52. Pickett, S.T.A. and White, PS. (eds.). 1985. The Ecology of Natural Disturbances and Patch Dynamics. Academic. New York. 472pp. Pierson, T C and Scott, K.M. 1985. Downstream dilution of a lahar: transition from debris flow to hyperconcentrated streamflow. Water Resources Res. 21:1511-1524. Pierson, T . C and Costa, J.E. 1987. A rheologic classification of subaerial sediment-water flows. In: Costa, J.E. and Wieczorek, GF. (eds.). Debris flows/avalanches: process, recognition, and mitigation. Geol. Soc. Amer. Reviews in Engineering Geology. Vol. VII. pp.1-12. Ritter, J.B., Miller, J.R., Enzel, Y, Howes, S.D., Nadon, G , Grubb, M.D., Hoover, K.A., Olsen, T , Reneau, S.L., Sack, D., Summa, C.L., Taylor, I., Touysinhthiphonexay, K.C.N., Yodis, E.G., Schneider, N.P, Ritter, D.F., and Wells, S.G 1993. Quaternary evolution of Cedar Creek alluvial fan, Montana. Geomorphology. 8:287-304. Runka, GG. 1972. Soil resources of the Smithers-Hazelton area. BC Dept. Agr., Soil Survey Division. Kelowna. BC. 234p. Ryder, J.M. 1971a. The stratigraphy and morphology of paraglacial alluvial fans in south-central British Columbia. Can. J. Earth Sci. 8: 279-98. Ryder, J.M. 1971b. Some aspects of the morphometry of paraglacial alluvial fans in south-central British Columbia. Can. J. Earth Sci. 8:1252-1264. Sato, T. 1991. Flood disaster with drifted logs and sand in Aso Volcano. In: Proceedings of the Japan-U.S. Workshop on Snow Avalanche, Landslide, Debris Flow Prediction and Control. Sept. 30 - Oct. 2, 1991. Tsukuba, Japan. Science and Tech. Agency of Japanese Gov. pp.497-506. Schreier, H. 2000. Research, planning, and implementation of watershed management. In: Price, M.F. and Butt, N. (eds.). Forests in sustainable mountain development. A state of knowledge report for 2000. IUFRO Research Series 5. CABI Publishing, Wallingford. pp. 380-389. 154 Schumm, S.A., Mosley, M.P., and Weaver, W.E. 1987 Experimental Fluvial Geomorphology. John Wiley. New York. 413pp. Schwab, J.W. 1983. Mass wasting: October-November 1978 storm, Rennell Sound, Queen Charlotte Islands. BC Min. For. Res. Note No. 91. 23pp. Schweingruber, F.H. 1996. Tree Rings and Environment Dendroecology. Paul Haupt Publishers. Berne. 609pp. Schweingruber, F.H., Eckstein, D., Serre-Bachet, F., and Braker, O.U. 1990. Identification, presentation and interpretation of event years and pointer years in dendrochronology. Dendrochronologia. 8: 9-37. Septer, D. and Schwab. J.W. 1995. Rainstorm and flood damage: northwest British Columbia 1891-1991. BC Min. For. Land Man. Handbook 31. 196pp. Sharp, R.P. 1942. Mudflow levees. J. Geomorphology. 5:222-227. Shreve, R.L. 1966. Statistical law of stream numbers. J. Geology. 74:17-37. Shreve, R.L. 1967. Infinite topographically random channel networks. J. Geology. 75:178-186. Sidle, R.C. 1991. A conceptual model of changes in root cohesion in response to vegetation management. J. Environ. Qual. 20(1): 43-52. Sidle, R .C , Pearce, A.J., and O'Loughlin, C.L. 1985. Hillslope stability and land use. Water Res. Monograph 11. Amer. Geophysical Union. Washington, D.C. 140pp. Sigafoos, R.S. 1964. Botanical evidence of floods and flood-plain deposition. US Geol. Survey Prof. Paper 485-A. 35pp. Sokal, R.R. and Rohlf, F.J. 1995. Biometry: the principles and practice of statistics in biological research. 3d ed. W.H. Freeman and Co. NY. 887pp. Smith, D.G. 1976. Effect of vegetation on lateral migration of a glacier meltwater river. Geol. Soc. Amer. Bull. 87: 857-860. Smith, G.A. 1986. Coarse-grained nonmarine volcaniclastic sediment: terminology and depositional process. Geol. Soc. of Amer. Bulletin. 97:1-10. Sterling, S.M. 1997. The influence of bedrock type on the magnitude, frequency and spatial distribution of debris torrents on northern Vancouver Island, British Columbia. M.Sc. Thesis. Dept. of Geography. UBC. 117pp. 155 Strahler, A.N. 1952. Hypsometric (area-altitude) analysis of erosional topography. Geol. Soc. Amer. Bull. 63: 1117-42. Strahler, A.N. 1958. Dimensional analysis applied to fluvially eroded landforms. Geol. Soc. Amer. Bull. 69: 279-99. Strunk, H. 1997. Dating of geomorphological processes using dendro-geomorphological meth-ods. Catena 31: 137-151. Suwa, H. 1988. Focusing mechanism of large boulders to a debris flow front. Transactions, Japanese Geomorphological Union. 9(3): 151-178. Swanson, F.J., Johnson, S.L., Gregory, S.V., and Acker, S.A. 1998. Flood disturbance in a forested mountain landscape. Bioscience 48(9): 681-689. Swanston, D.N. and Swanson, F.J. 1976. Timber harvesting, mass erosion and steepland forest geomorphology in the Pacific Northwest. In: Coates, D.R. (ed.). Geomorphology and Engineering. Stroudsburg, Pa. Downden, Hutchinson and Ross. pp. 199-211. Takahashi, T. 1981. Debris flows. Annu. Rev. Fluid Mech. 13: 57-77. Thrower, J.S., Nussbaum, A.F., and Di Lucca, C M . 1994. Site index curves and tables for British Columbia: interior species. 2nd edition. BC Min. For. Res. Br. Land Manage. Handb. Field Guide Insert 6. Thurber Consultants. 1983. Debris Torrent and Flooding hazards, Highway 99, Howe Sound. Report to the BC Min. of Trans, and Highways. 25pp. Thomson, B. 1991. Annotated bibliography of large organic debris (LOD) with regards to stream channels and fish habitat. BC Min. Envir. Tech. Rept. 32. Victoria. 93pp. Troup, R.S. 1928. Silvicultural systems. Oxford University Press. 199pp. VanDine, D.F. 1985. Debris flows and debris torrents in the southern Canadian Cordillera. Can. Geotechnical J. 22: 44-68. Watanabe, M . and Ikeya, H. 1981. Investigation and analysis of volcanic mud flows on Mount Sakurajima, Japan. Erosion sediment transport measurement. Int. Assoc. Hydrol. Sci. Publ. 133: 245-256. Weaver, W.E. and Schumm, S.A. 1974. Fan-head trenching an example of a geomorphic thresh-old. Geol. Soc. Amer. Abstracts with Programs. 1974. p.481. Weaver, W.E. 1984. Geomorphic thresholds and the evolution of alluvial fans. Geol. Soc. Amer. Abstracts with Programs. 1984. p.688. 156 Wells, S.G. and Harvey, A.M. 1987. Sedimentologic and geomorphic variations in storm-gener-ated alluvial fans, Howgill Fells, northwest England. Geol. Soc. Amer. Bulletin. 98: 182-198. Whittow, J. 1984. The Penguin Dictionary of Physical Geography. Penguin Books Ltd. London, England. 591pp. Wilford, D.J. and Schwab, J.W. 1982. Soil mass movements in the Rennell Sound Area, Queen Charlotte Islands, B.C. Proceedings of the Can. Hydrol. Symp. '82, Hydrological Proc-esses of Forested Areas. Fredricton, NB. Nat. Res. Coun. of Can. pp. 521-541. Wilford, D.J. 1984. The sediment-storage function of large organic debris at the base of unstable slopes. In: Meehan, W.R., Merrell, T.R., and Hanley, T.A. (eds.). Fish and Wildlife Rela-tionships in Old-Growth Forests: Proceedings of a symposium held in Juneau, Alaska. April 1982. Amer. Inst. Fish. Res. Biol. pp. 115-119. Wilford, D.J. 1998. Tributary Alluvial Fans. Extension Note #30. BC Min. For. Smithers, B.C. Can. 7pp. Wilford, D.J. and Sakals, M.E. 2002a. Forest management on fans: recognition of hydrogeomorphic hazards and general prescriptions. Manuscript in preparation for the BC Min. For. Research Working Paper series. Wilford, D.J., Sakals, M.E., and Innes, J.L. 2002b. Forest Management and Restoration on Fans. Watershed Restoration Tech. Bulletin Streamline. 6(3): 1-8. Wilford, D.J., Sakals, M.E., and Innes, J.L. 2002c. Forestry on Fans: Identifying hydrogeomorphic hazards. In: Jordan, P. and Orban, J. (eds.) Terrain stability and forest management in the interior of British Columbia. BC Min. For. Res. Program, Victoria, B.C. Tech. Rep. 3. pp. 189-201. Williams, G.P. 1978. Bank-full discharge of rivers. Water Res. Res. 14:1141-1154. Wolman, M.G. and Gerson, R. 1978. Relative scales of time and effectiveness of climate in watershed geomorphology. Earth Surf. Process. Landforms. 3:189-208. Wolman, M.G. and Miller, J.P. 1960. Magnitude and frequency of forces in geomorphic proc-esses. J. Geology 68: 54-74. Yoshida, K., Kikuchi, S., Nakamura, F., and Noda, M. 1997. Dendrochronological analysis of debris flow disturbance on Rishiri Island. Geomorphology. 20: 135-145. 157 A P P E N D I X A . O P E R A T I O N A L N E T W O R K The following individuals provided assistance and technical input throughout the project. • Dino Diana, R.P.F., Skeena Sawmills Ltd., Terrace • David Coates, Ph.D., R.P.F., Research Silviculturist, MOF, Smithers • Howard DeBeck, P.Eng., Forest Engineer, MOF, Smithers • Alan Harrison, Forest Technician, MOF, Hazelton • Dan Hogan, M.Sc, P.Geo., Fluvial Geomorphologist, MOF, Vancouver • Loren Kelly, P.Eng., Forest Engineer, MOF, Smithers • Norm Larson, Engineering Technician, MOF, Hazelton • Chris Lind, Forest Technician, MOF, Terrace • Jeff Lough, Fisheries Biologist, WLAP, Smithers • Denis Maynard, M.Sc, P.Geo., Terrain Specialist, Vancouver • Herb Neubrand, Forest Technician, MOF, Houston • Tom Pendray, Fisheries Biologist, Fisheries & Oceans Canada, Smithers • Dave Rebagliati, Engineering Technician, MOF, Houston • Dave Ripmeister, R.P.F., West Fraser Mills Ltd, Smithers • Jim Schwab, R.P.F., P.Geo., Forest Geomorphologist, MOF, Smithers • Tim Smith, Woods Manager, Canfor, Houston • Bruce Thomson, P.Geo. Terrain Specialist, WLAP, Surrey • Steve Webb, R.P.F, Silvicon Services Inc., Smithers • Irene Weiland, M.Sc, P.Geo..Terrain Specialist, Smithers • Ted Wilson, Engineering Technician, MOF, Terrace • Rob Ziegler, R.P.F., Skeena Cellulose Inc, Terrace 158 A P P E N D I X B. S U M M A R Y OF D I S T U R B A N C E A G E N T S This project is focusing on hydrogeomorphic factors as agents of disturbance. However, forest stands can be established and maintained by a range of disturbance factors. Most disturbance factors have key diagnostic features. Windthrow is a common factor in old growth forests. Characteristically individual trees are toppled leaving a hollow in the ground, and the root wad is tipped up. As the root wad and tree bowl decay, the evidence remaining is the hollow. The hollow may resemble a scour hole (a pool formed by running water). A characteristic in the soil profile is of disturbed soil horizonation and a lack of sorting of the mineral material in situations where sorting is expected or present below the disturbed layers. This mixing of the soil by vegetation is referred to as "floralpedoturbation". Over a long period of time, individual windthrown trees can produce a multi-aged stand structure, similar to what may be found in cases of hydrogeomorphic disturbance. However the soil profile will not show evidence of aggradation. Trees may also be broken by wind. For a long period of time the evidence should be present - stumps or broken tree bowls and downed large woody debris. Large scale or catastrophic windthrow events can also produce single cohort stands. However, the soil profile will not show evidence of aggradation. In addition, large-scale events would also extend beyond the fan to the surrounding landscape positions. Thus, the whole landscape should have a similar vegetation pattern following such a disturbance. There is a range of wood boring and bark beetle insects that are endemic to natural forests. Several features are key to identifying insect disturbances to vegetation communities. First, the disturbance is generally not limited to the fan. Second, it is usual for the disturbance to be limited to one species or diameter class, unless several insects are present. Third, for a considerable period of time, evidence of the insects is apparent either under the bark or a short distance into the wood (e.g., galleries and frass). A key feature is that hydrogeomorphic evidence will not be present (scars, deposition of sediments around the tree bowl). There may be an apparent 159 confounding of insect disturbance since a hydrogeomorphic event may stress trees and predispose them to insect attacks (e.g., Ips sp.). Forest fires can have a range of intensities and result in single and multicohort stands. However, fire is generally not limited to specific landforms, and if a fire has occurred the evidence should be present in the forest stands beyond the fan. It is interesting to note that characteristically stands on fans escape fire due to the higher moisture levels in the ecosystem (e.g., vegetation, soils, and microclimate). Evidence of fire is present for a considerable period of time; charred trees, stumps and logs and charcoal in the soil profile. Snow avalanches can initiate on the slopes above a fan. In this case, the hillslope stands can be used as an indicator of the disturbance. Snow avalanches can also initiate well up a watershed and flow down the stream channel. The track and deposit of debris may resemble a debris flow, however, several factors should be considered. First, the initiation zone of snow avalanches can be similar to mass wasting initiation zones, however, snow avalanches initiate within a snow pack rather than within the soil. Site inspections of soil should be undertaken for clarification. Second, a snow avalanche is primarily snow, with some entrained debris (rocks and organic material). Thus the deposit should lack evidence of significant fine mineral material and the transport zone should exhibit less evidence of scour than with a debris flow. The key features of hydrogeomorphic events are presented in Chapter 2. In summary, trees will be scarred, buried by sediments, or totally removed from a portion of the fan. Sediment deposits are a key diagnostic feature of the different hydrogeomorphic processes. Human (anthropomorphic) activities can range from individual tree removal (e.g., cedar poling) through to stand clearing and drainage/diversions. Site features include windrow piles, disturbances not connected to the stream channel or adjacent hillslopes, stumps, and evidence of old roads (e.g., berms, ruts, or strips of cohorts). 160 A P P E N D I X C . A B B R E V I A T I O N S F O R T R E E S P E C I E S A N D E C O S Y S T E M S Ecosystems From: Banner, A., MacKenzie, W., Haeussler, S., Thomson, S., Pojar, J. ,and Trowbridge, R. 1993. Afield guide to site identification and interpretation for the Prince Rupert Forest Region. Land Management Handbook 26. BC Min. Forests. SBSmc2 - Babine Variant of the Moist Cold subzone of the Sub-Boreal Spruce Biogeoclimatic Zone 01c - Sxw - Huckleberry association, submesic phase of the zonal site series, 02 - PI - Huckleberry - cladonia association, dry forested site series. 05 - Sxw - Twinberry - Coltsfoot association, fresh to wet forested site series 06 - Sxw - Oak fern association, fresh to wet forested site series. ESSFmc - the Moist Cold subzone of the Engelmann Spruce-Subalpine fir Biogeoclimatic Zone 02 - P1B1 - Juniper - Cladonia association, drier forested site series 05 - Bl - Huckleberry - Thimbleberry association, wetter forested site series 06 - Bl - Oak fern - Heron's-bill association, wetter forested site series 161 Tree species From: Anon. 1992. Forest Inventory Manual. BC Min. Forests Banner, A., MacKenzie, W., Haeussler, S., Thomson, S., Pojar, J. ,and Trowbridge, R. 1993. A field guide to site identification and interpretation for the Prince Rupert Forest Region. Land Management Handbook 26. BC Min. Forests. • Ba or Bl - Abies lasiocarpa - subalpine fir • Hm - T suga mertensiana- mountain hemlock • Pa - Pinus albicaulis - whitebark pine • PI - Pinus contorta var. latifolia - lodgepole pine • S - Picea sp. - Spruce - undifferentiated • Sxw - Picea glauca x engelmannii - hybrid white spruce 162 A P P E N D I X D . A T T R I B U T E S O F T H E S T U D Y W A T E R S H E D S U S E D F O R T H E A N A L Y S I S O F V A R I A N C E 163 APPENDIX D. Process Name Area Relief Length Channels Melton (kmA2) (km) (km) (km) Alice 38.55 1.78 10.88 114.46 0.29 Dasque 77.32 2.07 16.03 191.23 0.24 Hunter 63.25 2.07 9.57 241.10 0.26 Flood Sinclair 57.71 1.11 13.05 112.80 0.15 Tszkwa 99.27 1.58 18.46 178.15 0.16 Winfield 39.58 1.11 10.66 84.53 0.18 Middle 72.60 1.89 13.44 208.64 0.22 16Wllmse 2.07 1.18 2.92 3.71 0.82 16Wllmsw 1.57 1.21 2.33 4.46 0.96 Big_Wdn2 1.61 1.23 2.73 5.55 0.97 Cranberry 14.88 1.65 6.17 58.30 0.43 Compass 18.22 1.47 7.87 48.35 0.34 Gosnell8 7.55 1.09 4.35 20.09 0.40 Ktwancool 14.83 1.33 5.85 41.93 0.34 25UKit 3.74 1.42 3.06 12.10 0.73 Debris 39UKit 2.67 1.41 3.51 8.04 0.86 Flood 3Copper1 0.74 1.03 2.24 1.94 1.20 8McDnll3 4.69 1.56 3.81 9.17 0.72 8McDnll1 2.70 1.52 3.74 4.86 0.93 ManyBear 4.64 1.47 3.84 14.03 0.68 Sibola 9.48 0.88 6.22 23.07 0.29 Tetlock 4.68 1.29 3.71 18.93 0.60 Powerline 5.21 1.17 4.07 16.64 0.51 Whitebot 12.77 1.60 6.79 44.98 0.45 Skilokus 16.43 1.65 7.81 53.10 0.41 Big_Wdn1 4.08 1.43 4.68 8.35 0.71 Fernando 0.97 1.06 2.11 2.51 1.08 Kitsl 0.49 0.55 1.13 3.04 0.78 Debris Kits2 0.25 0.61 0.98 1.16 1.21 Flow Kits4 0.42 0.56 1.22 3.25 0.86 Rico 2.31 1.42 2.62 8.58 0.94 Wan 1.29 0.74 1.89 2.99 0.66 Z_Cascad 2.29 1.44 2.43 11.95 0.95 Power Disturbance) Extent a> 3 o Q. i . C d> > _ i TJ C ca 164 APPENDIX D. Continued. Power Disturbance Extent Process Name Area (kmA2) Relief (km) Length (km) Channels (km) Melton Flood Shelford 37.32 0.46 11.81 145.13 0.08 Ailport 19.46 0.50 7.69 35.09 0.11 8McDnll2 1.80 1.52 3.66 4.24 1.13 Herb 6.30 0.78 5.13 26.59 0.31 Luno 31.38 1.71 10.73 77.16 0.31 nwfM "55 OMcDnll 3.91 0.88 4.20 8.48 0.44 nwfM > CD —I 18SKit 20.64 1.35 4.26 21.51 0.30 0. 1 Debris Flood 19SKit 12.42 1.37 6.98 49.77 0.39 .c x V (75 Gosnell7 Mill 5.43 1.41 0.94 1.04 3.57 2.16 14.82 3.68 0.40 0.88 Miller 6.79 1.25 4.66 12.08 0.48 22 Shedin 4.82 1.02 5.41 12.41 0.46 Gosnell 1 3.10 1.08 3.58 5.65 0.61 Gosnell4 4.23 1.10 4.55 13.25 0.54 Gosnell6 8.48 1.02 4.38 23.01 0.35 a> 3Copper2 0.91 1.05 2.12 1.21 1.10 High-Power > a> _ i a> Debris Flow Carriganl Kits3 Legate 1.39 0.21 1.32 0.87 0.55 1.24 2.31 0.78 2.44 2.98 1.59 5.30 0.74 1.19 1.08 CO Trapln 0.68 0.86 2.03 1.66 1.04 Canyon 9.18 0.73 5.31 15.58 0.24 3D 9.17 0.92 5.91 13.52 0.24 CP095-1 5.32 0.46 3.48 17.14 0.20 Flood CP095-2 2.55 0.44 3.57 8.88 0.27 M3 9.66 1.15 5.80 20.48 0.37 o ft McKndrkl 5.82 0.60 4.53 9.37 0.25 i > z Newcmbl 1.35 0.57 2.27 4.32 0.49 o Carrigan2 1.28 0.86 1.98 3.17 0.76 _J Debris Flood SprCmp 1.77 0.79 3.30 3.46 0.60 Tableland Carrigan3 3.82 1.57 0.69 0.48 2.90 1.68 12.37 3.95 0.35 0.38 Poplar 4.24 0.53 4.14 8.66 0.26 165 A P P E N D I X D. Continued. R e l i e f R a t i o D r a i n D e n P r o c e s s N a m e H y p s o S h a p e (km/km A 2 ) Alice 0.41 0.33 0.16 2.97 Dasque 0.45 0.30 0.13 2.47 Hunter 0.46 0.69 0.22 3.81 Flood Sinclair 0.52 0.34 0.08 1.95 Tszkwa 0.37 0.29 0.09 1.79 Winfield 0.53 0.35 0.10 2.14 Middle 0.48 0.40 0.14 2.87 16Wllmse 0.59 0.24 0.40 1.79 16Wllmsw 0.52 0.29 0.52 2.85 Big_Wdn2 0.45 0.22 0.45 3.45 Cranberry 0.60 0.39 0.27 3.92 Compass 0.42 0.29 0.19 2.65 Gosnell8 0.49 0.40 0.25 2.66 Ktwancool 0.54 0.43 0.23 2.83 25UKit 0.60 0.40 0.46 3.24 Debris 39UKit 0.57 0.22 0.40 3.01 Flood 3Copper1 0.64 0.15 0.46 2.63 8McDnll3 0.46 0.32 0.41 1.95 8McDnll1 0.51 0.19 0.41 1.80 ManyBear 0.51 0.32 0.38 3.02 Sibola 0.61 0.25 0.14 2.43 Tetlock 0.61 0.34 0.35 4.04 Powerline 0.60 0.31 0.29 3.20 Whitebot 0.53 0.28 0.24 3.52 Skilokus 0.56 0.27 0.21 3.23 Big_Wdn1 0.63 0.19 0.30 2.05 Fernando 0.56 0.22 0.50 2.60 Kitsl 0.54 0.39 0.49 6.17 Debris Kits2 0.49 0.26 0.62 4.64 Flow Kits4 0.45 0.28 0.46 7.74 Rico 0.49 0.34 0.54 3.71 Wan 0.50 0.36 0.39 2.33 Z_Cascad 0.55 0.39 0.60 5.21 P o w e r Ex ten t o a. O) a> > _ i TJ C (0 166 A P P E N D I X D. Continued. P o w e r D i s t u r b a n c e Exten t P r o c e s s N a m e H y p s o S h a p e R e l i e f R a t i o D r a i n D e n (km/km A 2 ) Flood Shelford 0.56 0.27 0.04 3.89 Ailport 0.72 0.33 0.06 1.80 8McDnll2 0.41 0.14 0.42 2.35 Herb 0.56 0.24 0.15 4.22 Luno 0.58 0.27 0.16 2.46 owei OMcDnll 0.58 0.22 0.21 2.17 owei > 18SKit 0.62 1.14 0.32 1.04 Q_ I _ i Debris Flood 19SKit 0.57 0.26 0.20 4.01 ' CO Gosnell7 Mill 0.54 0.54 0.43 0.30 0.26 0.48 2.73 2.61 Miller 0.43 0.31 0.27 1.78 22 Shedin 0.50 0.16 0.19 2.57 Gosnell 1 0.48 0.24 0.30 1.82 Gosnell4 0.53 0.20 0.24 3.13 Gosnell6 0.60 0.44 0.23 2.71 3Copper2 0.59 0.20 0.49 1.34 High-Power > _l Debris Flow Carriganl Kits3 Legate 0.53 0.49 0.52 0.26 0.35 0.22 0.38 0.70 0.51 2.14 7.52 4.02 CO Trapln 0.58 0.16 0.42 2.46 Canyon 0.47 0.33 0.14 1.70 3D 0.61 0.26 0.12 1.47 CP095-1 0.60 0.44 0.13 3.22 0) Flood CP095-2 0.71 0.20 0.12 3.48 M3 0.38 0.29 0.20 2.12 o n McKndrkl 0.61 0.28 0.13 1.61 > z Newcmbl 0.52 0.26 0.25 3.20 > O Carrigan2 0.50 0.33 0.43 2.47 _l Debris SprCmp 0.51 0.16 0.24 2.00 Flood Tableland Carrigan3 0.55 0.65 0.45 0.56 0.24 0.28 3.24 2.52 Poplar 0.54 0.25 0.13 2.04 167 A P P E N D I X D. Continued. P o w e r D isturbance) Ex ten t P r o c e s s N a m e C o m m (%) E S A (%) E S A S x (%) Alice 40 0 5 Dasque 16 1 7 Hunter 39 6 12 Flood Sinclair 62 7 10 Tszkwa 65 4 5 Winfield 58 16 21 Middle 34 0 4 16Wllmse 28 0 0 16Wllmsw 43 25 25 Big_Wdn2 81 4 20 Cranberry 29 10 12 Compass 68 14 34 Gosnell8 10 10 12 Ktwancool 44 18 27 25UKit 28 0 0 Debris 39UKit 50 0 10 Flood 3Copper1 50 35 35 8McDnll3 27 6 22 8McDnll1 23 8 19 ManyBear 9 1 1 Sibola 23 3 6 Tetlock 51 5 13 Powerlin'e 56 16 23 Whitebot 49 1 9 Skilokus 37 3 7 Big_Wdn1 31 1 11 Fernando 20 0 0 Kitsl 34 30 34 Debris Kits2 30 22 29 Flow Kits4 32 38 38 Rico 40 14 23 Wan 26 3 25 Z_Cascad 1 0 0 0) o Q. > a> _ i •o c (0 CO 168 A P P E N D I X D. Continued. P o w e r D i s t u r b a n c e Exten t P r o c e s s N a m e C o m m (%) E S A (%) E S A S x (%) Flood Shelford Ail port 86 89 1 1 1 2 8McDnll2 29 2 24 Herb 20 4 4 Luno 30 5 5 OWGI "o> OMcDnll 33 14 30 OWGI > to 18SKit 5 5 5 High-P Site Li Debris Flood 19SKit Gosnell7 Mill 55 10 71 9 1 53 14 6 53 Miller 40 3 16 22 Shedin 56 0 31 Gosnell 1 43 4 9 Gosnell4 32 0 6 Gosnell6 13 3 10 3Copper2 74 69 69 High-Power ite Lev Debris Flow Carriganl Kits3 Legate 95 16 56 2 22 29 5 22 29 CO Trapln 87 0 0 Canyon 60 6 7 3D 91 0 0 CP095-1 91 0 3 a> Flood CP095-2 66 5 5 M3 63 3 13 o n McKndrkl 50 10 13 • > z Newcmbl 80 28 59 O Carrigan2 74 15 17 —1 Debris Flood SprCmp Tableland 83 53 7 0 19 26 Carrigan3 99 0 21 Poplar 89 0 4 169 A P P E N D I X D. Continued. P o w e r D is tu rbance ! Ex ten t P r o c e s s N a m e G 3 0 G 3 5 G 4 0 B 3 0 4 0 (%) (%) (%) (%) Alice 45 32 20 25 Dasque 63 46 31 32 Hunter 62 45 28 33 Flood Sinclair 5 2 1 4 Tszkwa 6 3 2 4 Winfield 7 3 1 6 Middle 52 32 16 35 16Wllmse 42 18 4 38 16Wllmsw 42 24 12 31 Big_Wdn2 24 7 1 23 Cranberry 37 24 11 26 Compass 29 17 8 21 Gosnell8 38 18 7 31 Ktwancool 61 42 19 42 25UKit 50 29 12 37 Debris 39UKit 23 10 3 21 Flood 3Copper1 26 17 13 14 8McDnll3 30 17 8 22 8McDnll1 33 20 10 23 ManyBear 67 51 34 33 Sibola 6 2 1 5 Tetlock 53 34 18 35 Powerline 52 30 10 42 Whitebot 55 32 13 42 Skilokus 46 28 13 33 Big_Wdn1 21 7 1 20 Fernando 58 37 19 40 Kitsl 79 63 38 41 Debris Kits2 69 50 29 41 Flow Kits4 68 54 31 37 Rico 75 59 38 37 Wan 57 41 25 32 Z_Cascad 76 55 31 45 a> o Q_ i JZ cn 0) > a> _i TJ C (0 co 170 A P P E N D I X D. Continued. P o w e r D i s t u r b a n c e P r o c e s s N a m e G 3 0 G 3 5 G 4 0 B3040 (%) (%) (%) (%) Ex ten t Flood Shelford 0 0 0 0 Ailport 1 0 0 1 8McDnll2 27 15 6 22 Herb 2 0 0 1 Luno 36 22 10 26 CO "55 OMcDnll 12 5 2 10 o > CO 18SKit 12 7 4 8 0. _l 19SKit 48 31 16 32 A CO Gosnell7 36 16 5 31 I CO Flood Mill 63 30 4 60 Miller 6 2 1 6 22 Shedin 1 0 0 1 Gosnell 1 24 7 1 23 Gosnell4 28 13 3 25 Gosnell6 35 16 4 32 "55 3Copper2 42 25 12 29 ± 1 > CD Debris Carriganl 45 24 10 35 _ l Kits3 76 59 32 44 = £ Flow Legate 69 52 29 40 CO Trapln 21 7 2 20 Canyon 3 1 0 2 3D 1 0 0 1 CP095-1 4 1 0 4 Flood CP095-2 10 6 3 8 a> M3 0 0 0 0 o McKndrkl 2 1 0 2 1 > z Newcmbl 5 2 0 4 > o Carrigan2 58 38 18 40 —I Debris SprCmp 4 1 0 4 Tableland 9 3 1 8 Flood Carrigan3 15 5 1 14 Poplar 1 0 0 1 171 A P P E N D I X E. A N E X A M P L E S K E L E T O N PLOT A N D S U M M A R Y OF EVENTS CONTENTS E . l . FIELD NOTES E.2. LEGEND FOR SKELETON PLOT E.3. SKELETON PLOT E.4. SUMMARY OF EVENTS 172 E . l . FIELD NOTES - 39KUKIT Type of sample: C = core, D = disk, W = wedge Species: B - Abies amabilis, H - T suga heterophyllaS - Picea sp., Ac - Populus trichocarpa Aspect of sample: D = downstream, TJ = upstream, XS = across slope facing stream, XN = across slope away from stream Sampling Dates: April 19, 2001 (#1-8), May 15, 2001 (#9-33), June 13, 2001 (#34) # Type Species Sample ht. (cm) Aspect Tree Dia./Ht (cm/m) Notes 1 C B 130 D 50.5/26 On left bank of contemporary channel. GPS #1. 2 D H 0 - /1.0 On recent sediments on left bank, near toe of fan. GPS #1. 3 D S 0 - /1.0 On recent sediments on left bank, near toe of fan. GPS #1. 4 D Ac 0 - /1.0 On recent sediments on left bank, near toe of fan. GPS#1. 5 C S 130 U 58.5/16 Broken top, appears buried, on left bank. GPS #1. 6 W B 80 XS Scar 1.0 x 0.10m on upstream side of tree. On right bank. GPS #2. 7 C S 10 xs 6.5/4 Just upstream of Tree 6, on right bank, in cohort of 12 trees. GPS #2. 8 D S 0 - / l Just on downstream side of proposed cribbing, 9 m downstream of centre line. GPS #3 9 C s 130 u On upstream side of proposed cribbing on levee, has adventitious roots, beside creek. GPS #17. 10 C s 130 xs 2 m further from right bank, otherwise as Tree 9 without adventitious roots. GPS #17. 11 C s 130 u Repeat of Tree 10. GPS #17. 12 C s 20 u /6 Cohort of 3 trees, behind lobe just above proposed cribbing. GPS #17. 13 D s 0 - 7.5m below the two big spruce on lobe, 1 m from channel right bank. GPS #17. 14 C B 80 xs 10 m N of big spruces, up on fan surface, dug around bole - buried. GPS #18. 173 # Type Species Sample ht. (cm) Aspect Tree Dia./Ht (cm/m) Notes 15 D H 0 - Event marker, on top of log which fell before event and forms log step, just upstream of Tree 14. GPS #18. 16 C S 130 X N Beside Trees 14 and 15, dug around bole - buried. GPS# 18. 17 W H 110 U 2 m further from stream than Tree 16, scar 1 m above ground, 0.3 m high and 0.15 m wide. GPS #18. 18 C B 130 D Has scar on upstream side, 6 m upstream of Tree 17, 1.5 m to creek, near edge of deposit area. GPS #18. 19 C S 130 D 36.5/ Does not appear buried, 15 m from right bank. GPS #19. 20 C S 130 U 21/ Does not appear buried, 20 m from right bank. GPS #19. 21 C s 130 U 10/ Does not appear buried, 20 m from right bank. GPS #19. 22 D H 0 ~ /3 Potentially knocked over, looking for onset of compression wood, on right side of fan, in younger channel bottom cohort. GPS #20. 23 C s 30 X N 16/ In younger channel bottom cohort, just upstream of Tree 24, rotation could be from germinating on log. GPS #21. 24 C H 130 D Veteran in younger cohort (Tree 23) has scar 1.5 m high x 0.5 m wide, one of few older trees in vicinity. GPS #21. 25 D S 0 ~ Broken top, on channel side of right bank deposit lobe/lateral channel bar, 2 m from right bank. GPS #22. 26 C B 130 D At tail end of levee, appears buried, just right of more active deposit and 5 m from right bank, may have been scarred. GPS #23. 27 C S 130 U Upstream from Tree 26, also buried. GPS #23. 28 D S 0 - /4 In right bank cohort (3 x 25 m) with an older cohort further from the channel, 5 m from right bank. GPS #24. 29 C B 60 xs 14/10 In older cohort described above, 2 m wide x 15 m long. GPS #24. 30 C S 60 X N /14 Same as above, scar of questionable origin. GPS #24. 174 # Type Species Sample ht. (cm) Aspect Tree Dia./Ht (cm/m) Notes 31 C S 40 D 24.4/ 2.5 m from left bank, buried, must have survived event as it is among a younger cohort (3.5 x 25 m). GPS #25. 32 D S 0 - /3 Just 10 m below Tree 31 in the younger cohort. GPS #25. 33 D S 0 - / l 30 m below road centre line, 5 m from right bank on sediment lobe. GPS #26. 34 C S 10 u/xs 6.5/4 Repeat of Tree 7. Cohort of 12 trees. GPS #2 175 E.2. LEGEND FOR SKELETON PLOT Abbreviations S - scar A - abrupt growth change (+ or -) F - frost ring R - radial crack C - compression wood T - traumatic resin canal (resin duct) E - early wood L - late wood () - other side of pith Notation * - an event 0 - date of establishment, a dashed circle indicates an estimate due to stem burial, - over the circle indicates an estimate # - date of pith - actual if skeleton plot line solid, estimated if line dashed • - date of end of core or wedge 4—• - core or wedge continues, establishment date not estimated Visual Growth Analysis Growth reductions: 1 i - - slight - 40 - 55% —I r|— - moderate - 56 - 70% -A —I = i— - strong - >71% 176 Growth increases: • - slight - 50 - 100% • 3 - moderate - 101 - 200% i j - strong-> 201% In cases where growth changes are gradual two approaches are used in the Skeleton Plots >• - "release" / "suppression" indicates a gradual change starting at that point. This is used in cases where the growth is in long-term change due most likely to a change in stand conditions. •'-...[^ - a dashed diagonal line from the start of change to the point where growth has changed enough to achieve the required ring width. This is used in cases where growth change is gradual, but most likely due to a hydrogeomorphic event - the tree generally returns to normal growth after a period of time. 177 E.3. SKELETON PLOT 178 I -3 5' Q - - -«y— a*: 9 0* 0 o ® 4 « d ] " T 1-4. 3* U H V -3 4 : J J J J J ;f J . Pf o CD CJ) CD O) CM CD O) co CD CJ) CO O) LO CO O) CD CD CJ) CD O) co co CJ) OMcDnll site/DFIood P X X 16Wllmse stand/DFIood P 16Wllmsw stand/DFIood P X 18SKit site/DFIood X P X P 19SKit site/DFIood X P X X 22Shedin site/DFIood X 25UKit stand/DFIood X X 39UKit stand/DFIood X 3Copper1 stand/DFIood X P 3Copper2 site/DFlow P 3D low/Flood X P 8McDnll1 stand/DFIood X P 8McDnll2 site/DFIood X 8McDnll3 stand/DFIood X P Ail port site/Flood X Alice stand/Flood P Big_Wdn1 stand/DFlow P P p Big_Wdn2 stand/DFIood Canyon low/Flood P P P CP095-1 low/Flood CP095-2 low/Flood P P Carriganl site/DFlow X X Carrigan2 site/DFIood X P p Carrigan3 (log) low/DFIood Compass stand/DFIood Cranberry stand/DFIood P P X Dasque (log) stand/Flood Fernando stand/DFlow P X P X p Gosnell 1 site/DFIood X X GosnelW site/DFIood P Gosnell6 site/DFIood Gosnell7 site/DFIood X Gosnell8 stand/DFIood X P Herb site/DFIood X P X P X P Hunter stand/Flood P P P X Kitsl stand/DFlow P Kits2 stand/DFlow Kits3 site/DFlow Kits4 stand/DFlow Kitwancool stand/DFIood P X X Legate site/DFlow 185 A P P E N D I X F. Continued. o CM CO LO CO 00 CD o CM CO LO CO 00 LO LO LO LO LO LO LO LO LO LO co CO CO CO CO co co co CO F a n C a t e g o r y CD o> CD O) CD CD CD CD OJ o> O) o> o> CD CD CD OJ CD CD Luno site/DFIood McKndrkl low/Flood M3 site/Flood X P P P P ManyBear stand/DFIood X P P Middle stand/Flood X P Mill site/DFIood P X Miller site/DFIood X X Newcmbl site/Flood Poplar site/DFIood P P P P X Pwrline stand/DFIood P P Rico stand/DFlow N N N N Shelf ord site/Flood Sibola stand/DFIood X X P P P X Sinclair stand/Flood X P P Skilokus (log) stand/DFIood SprCmp low/DFIood P P Tablelnd low/DFIood P P P X Tetlock stand/DFIood X P X Trapln site/DFlow X X X X Tszkwa stand/Flood N Whitebot stand/DFIood P X X Wan stand/DFlow P P P P P Winfield stand/Flood P P P X P X P p Z Cas-no data stand/DFlow # certain events 2 3 3 3 2 4 2 6 4 0 2 1 1 2 2 7 3 3 6 # probable events 2 2 1 3 0 8 3 3 6 3 5 8 2 6 2 6 5 2 7 # non-characterizing events 0 0 0 0 0 0 0 1 0 1 0 1 0 0 0 1 1 0 0 Total # events 4 5 4 6 2 12 5 10 10 4 7 10 3 8 4 14 9 5 13 Total # samples 186 APPENDIX F. Continued. Fan Category co o o> CD CNJ O) CO O CO m CD c o CD c n co CO CD CD o 0 0 CD oo CD CM 0 0 CD CO oo CD 0 0 c n LO oo CD CD 0 0 o> co CD OMcDnll site/DFIood P X 16Wllmse stand/DFIood X P X 16Wllmsw stand/D Flood X X X X 18SKit site/DFIood P P P X X p 19SKit site/DFIood X p P X 22Shedin site/DFIood 25UKit stand/DFIood X X 39UKit stand/DFIood X 3Copper1 stand/DFIood p X 3Copper2 site/DFlow X 3D low/Flood X p P P 8McDnll1 stand/DFIood P p P X 8McDnll2 site/DFIood X X 8McDnll3 stand/DFIood X P X P Ailport site/Flood X X X Alice stand/Flood P X P Big_Wdn1 stand/DFlow X X P Big_Wdn2 stand/DFIood Canyon low/Flood P X P P X CP095-1 low/Flood CP095-2 low/Flood P P p Carriganl site/DFlow X X X X X Carrigan2 site/DFIood P p P P Carrigan3 (log) low/DFIood Compass stand/DFIood N N Cranberry stand/DFIood X X Dasque (log) stand/Flood Fernando stand/DFlow X p X P P X GosnelH site/DFIood X X X X Gosnell4 site/DFIood P X p Gosnell6 site/DFIood P X p X Gosnell7 site/DFIood P X X Gosnell8 stand/DFIood P X X Herb site/DFIood P X p X Hunter stand/Flood p X P P Kitsl stand/DFlow X P X Kits2 stand/DFlow P Kits3 site/DFlow Kits4 stand/DFlow X X Kitwancool stand/DFIood p X Legate site/DFlow X X 187 APPENDIX F. Continued. O ) o CM co LO CO CO O ) o CM oo • t f LO co CD 1^ OO CD 00 00 oo co oo 00 Fan Category O ) CTJ a) CJ) O ) CD CJ) CJ) CJ) O ) o o> O ) CJ) CJ) O ) O ) 0) O ) Luno site/DFIood X P X McKndrkl low/Flood M3 site/Flood X P ManyBear stand/DFIood P P P X Middle stand/Flood P Mill site/DFIood X X Miller site/DFIood P P Newcmbl site/Flood Poplar site/DFIood P X Pwrline stand/DFIood P P Rico stand/DFlow N Shelford site/Flood X X Sibola stand/DFIood P X P P X P P Sinclair stand/Flood X P X P P Skilokus (log) stand/DFIood SprCmp low/DFIood P X P P Tablelnd low/DFIood X P P P P Tetlock stand/DFIood X P X X Trapln site/DFlow P X X Tszkwa stand/Flood N N Whitebot stand/DFIood X X Wan stand/DFlow X X P Winfield stand/Flood P X X P P P X X X Z_Cas-no data stand/DFlow # certain events 4 2 11 1 1 4 6 5 8 6 3 3 12 2 2 6 6 4 5 # probable events 3 6 4 5 3 2 3 1 10 5 4 4 10 4 3 3 3 3 3 # non-characterizing events 1 0 0 0 1 1 0 0 0 1 0 0 0 1 0 0 0 0 0 Total # events 8 8 15 6 5 7 9 6 18 12 7 7 22 7 5 9 9 7 8 Total # samples 188 A P P E N D I X F. Continued. F a n C a t e g o r y c o o o C D C D o o C D o C D C D C D C D C M C D C D C O C D C D C D C D m C D C D c o C D C D C D C D o o C D C D C D C D C D 2000 # Certain # Probable Total # # Samples OMcDnll site/DFIood X X p X 6 3 9 17 16Wllmse stand/DFIood X X 4 2 6 6 16Wllmsw stand/DFIood X X X 8 1 9 13 18SKit site/DFIood X X X P X 8 7 15 21 19SKit site/DFIood X p P X 7 5 12 17 22Shedin site/DFIood 1 0 1 10 25UKit stand/DFIood X X 6 0 6 30 39UKit stand/DFIood X 3 0 3 34 3Copper1 stand/DFIood X X 4 1 5 15 3Copper2 site/DFlow P X p 2 3 5 9 3D low/Flood P X P P 3 7 10 14 8McDnll1 stand/DFIood X 3 4 7 22 8McDnll2 site/DFIood X X P 5 1 6 12 8McDnll3 stand/DFIood X X 5 3 8 14 Ailport site/Flood X 5 0 5 9 Alice stand/Flood X p p X 3 5 8 18 Big_Wdn1 stand/DFlow X X X 5 4 9 12 Big_Wdn2 stand/DFIood Canyon low/Flood X X 4 6 10 21 CP095-1 low/Flood 0 0 0 4 CP095-2 low/Flood p X P 1 7 8 7 Carriganl site/DFlow P P X 8 2 10 32 Carrigan2 site/DFIood P P P P 1 10 11 21 Carrigan3 (log) low/DFIood Compass stand/DFIood N 0 0 0 15 Cranberry stand/DFIood X X 5 2 7 18 Dasque (log) stand/Flood Fernando stand/DFlow P p P X 6 9 15 17 Gosnelh site/DFIood X X X 9 0 9 20 Gosnell4 site/DFIood X X X 4 3 7 16 Gosnell6 site/DFIood X p 3 3 6 13 Gosnell7 site/DFIood X 4 1 5 17 Gosnell8 stand/DFIood X X P 5 3 8 17 Herb site/DFIood P p X X X X X p 10 8 18 25 Hunter stand/Flood P X X P p 4 9 13 14 Kitsl stand/DFlow X 3 2 5 43 Kits2 stand/DFlow p p 0 3 3 2 Kits3 site/DFlow p X X X 3 1 4 5 Kits4 stand/DFlow X 3 0 3 4 Kitwancool stand/DFIood X X 5 2 7 16 Legate site/DFlow 2 0 2 39 A P P E N D I X F. Continued. F a n C a t e g o r y CO GO CD CD co CD o CD CD •i— CD CD CM CD CD co CD CD • t f CD CD LO CD CD CD CD CD 1^ CD CD oo CD CD CD CD CD 2000 # Certain # Probable Total # # Samples Luno site/DFIood X X X X X 7 1 8 32 McKndrkl low/Flood M3 site/Flood P X P P 3 8 11 15 ManyBear stand/DFIood X P X 4 6 10 21 Middle stand/Flood X 2 2 4 9 Mill site/DFIood 3 1 4 9 Miller site/DFIood X X 4 2 6 28 Newcmbl site/Flood X 1 0 1 2 Poplar site/DFIood P X 3 6 9 21 Pwrline stand/DFIood X X 2 4 6 7 Rico stand/DFlow N N N 0 0 0 42 Shelford site/Flood 2 0 2 13 Sibola stand/DFIood X X X 8 8 16 26 Sinclair stand/Flood X X X 6 5 11 20 Skilokus (log) stand/DFIood SprCmp low/DFIood X P P P 2 8 10 12 Tablelnd low/DFIood X X 4 7 11 16 Tetlock stand/DFIood X X P 7 3 10 27 Trapln site/DFlow X X 8 1 9 16 Tszkwa stand/Flood N 0 0 0 16 Whitebot stand/DFIood X 5 1 6 23 Wan stand/DFlow P X X X 5 7 12 33 Winfield stand/Flood X X X X X X 13 10 23 29 Z_Cas-no data stand/DFlow # certain events 11 12 9 8 15 11 7 12 5 4 6 0 1 247 # probable events 8 5 4 1 7 2 3 3 2 4 4 0 1 197 # non-characterizing events 1 0 0 0 1 0 1 0 1 0 0 1 15 Total # events 20 17 13 9 23 13 11 15 8 8 10 1 2 459 Total # samples 1056 190 A P P E N D I X G. D A T A SETS F O R R E G R E S S I O N A N A L Y S I S 191 A P P E N D I X G . P r o c e s s N a m e T o t a l A r e a M a x E l M inEI Re l ie f (km) (km) (km) E v e n t s (kmA2) Alice 8 38.55 2.03 0.25 1.78 Hunter 13 63.25 2.37 0.30 2.07 Flood Sinclair 11 57.71 1.86 0.75 1.11 Tszkwa 0 99.27 2.33 0.75 1.58 Winfield 23 39.58 1.87 0.76 1.11 Middle 4 72.60 1.95 0.06 1.89 16Wllmse 6 2.07 1.83 0.64 1.18 16Wllmsw 9 1.57 1.80 0.60 1.21 Cranberry 7 14.88 2.01 0.37 1.65 Compass 0 18.22 2.01 0.54 1.47 Gosnell8 8 7.55 2.14 1.05 1.09 Ktwancool 7 14.83 1.81 0.48 1.33 25UKit 6 3.74 1.72 0.31 1.42 Debris 39UKit 3 2.67 1.86 0.45 1.41 Flood 3Copper1 5 0.74 1.20 0.17 1.03 8McDnll3 8 4.69 2.54 0.98 1.56 8McDnll1 7 2.70 2.51 0.98 1.52 ManyBear 10 4.64 2.25 0.79 1.47 Sibola 16 9.48 1.91 1.03 0.88 Tetlock 10 4.68 1.58 0.29 1.29 Powerline 6 5.21 1.36 0.19 1.17 W_Bottm 6 12.77 1.73 0.14 1.60 Big_Wdn1 9 4.08 1.71 0.28 1.43 Fernando 15 0.97 1.63 0.57 1.06 Debris Kitsl 5 0.49 1.73 1.18 0.55 Kits2 3 0.25 1.77 1.17 0.61 Flow Kits4 3 0.42 1.75 1.19 0.56 Rico 0 2.31 1.57 0.14 1.42 Wan 12 1.29 1.85 1.11 0.74 P o w e r D i s t u r b -a n c e 0i S o Q_ • JZ U) X a> > a> _ i TJ C (0 +-* CO 192 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events Area (kmA2) MaxEl (km) MinEI (km) Relief (km) Flood Shelf ord 2 37.32 1.51 1.04 0.46 Ailport 5 19.46 1.40 0.90 0.50 8McDnll2 6 1.80 2.54 1.02 1.52 Herb 18 6.30 1.92 1.14 0.78 Luno 8 31.38 2.18 0.46 1.71 OMcDnll 9 3.91 1.90 1.02 0.88 18SKit 15 20.64 1.66 0.31 1.35 Power Debris Flood 19SKit 12 12.42 1.68 0.30 1.37 Power "35 > _j Gosnell7 Mill 5 4 5.43 1.41 2.03 1.54 1.09 0.50 0.94 1.04 igh-l Miller 22Shedin 6 1 6.79 4.82 2.26 1.81 1.01 0.79 1.25 1.02 X GosnelH 9 3.10 1.96 0.88 1.08 Gosnell4 7 4.23 2.02 0.92 1.10 Gosnell6 6 8.48 2.03 1.01 1.02 3Copper2 5 0.91 1.20 0.15 1.05 Debris Flow Carriganl Kits3 10 4 1.39 0.21 1.55 1.79 0.68 1.24 0.87 0.55 Legate 2 1.32 1.46 0.23 1.24 Trapln 9 0.68 1.44 0.58 0.86 Canyon 10 9.18 1.83 1.10 0.73 3D 10 9.17 1.21 0.48 0.92 I— Flood CP095-1 0 5.32 1.48 1.02 0.46 CP095-2 8 2.55 1.50 1.06 0.44 o n M3 11 9.66 2.04 0.89 1.15 > z Newcmbl 1 1.35 1.57 1.00 0.57 > o —I Debris Carrigan2 SprCmp 11 10 1.28 1.77 1.55 1.74 0.69 0.95 0.86 0.79 Flood Tableland 11 3.82 1.66 0.96 0.69 Poplar 9 4.24 1.61 1.08 0.53 193 APPENDIX G. Continued. Process Name Total Mean E lev Length Channels Melton Events (km) (km) (km) Alice 8 0.99 10.88 114.46 0.29 Hunter 13 1.25 9.57 241.10 0.26 Flood Sinclair 11 1.33 13.05 112.80 0.15 Tszkwa 0 1.34 18.46 178.15 0.16 Winfield 23 1.34 10.66 84.53 0.18 Middle 4 0.97 13.44 208.64 0.22 16Wllmse 6 1.34 2.92 3.71 0.82 16Wllmsw 9 1.22 2.33 4.46 0.96 Cranberry 7 1.35 6.17 58.30 0.43 Compass 0 1.16 7.87 48.35 0.34 Gosnell8 8 1.58 4.35 20.09 0.40 Ktwancool 7 1.19 5.85 41.93 0.34 25UKit 6 1.16 3.06 12.10 0.73 Debris 39UKit 3 1.25 3.51 8.04 0.86 Flood 3Copper1 5 0.83 2.24 1.94 1.20 8McDnll3 8 1.70 3.81 9.17 0.72 8McDnll1 7 1.76 3.74 4.86 0.93 ManyBear 10 1.53 3.84 14.03 0.68 Sibola 16 1.57 6.22 23.07 0.29 Tetlock 10 1.08 3.71 18.93 0.60 Powerline 6 0.90 4.07 16.64 0.51 W_Bottm 6 0.98 6.79 44.98 0.45 Big_Wdn1 9 1.18 4.68 8.35 0.71 Fernando 15 1.17 2.11 2.51 1.08 Debris Flow Kitsl 5 1.48 1.13 3.04 0.78 Kits2 3 1.47 0.98 1.16 1.21 Kits4 3 1.45 1.22 3.25 0.86 Rico 0 0.84 2.62 8.58 0.94 Wan 12 1.48 1.89 2.99 0.66 Power Disturb-ance 0) o a. Ui x > a> - i TJ C (0 CO 194 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events Mean E lev (km) Length (km) Channels (km) Melton Flood Shelford 2 1.30 11.81 145.13 0.08 Alport 5 1.26 7.69 35.09 0.11 8McDnll2 6 1.65 3.66 4.24 1.13 Herb 18 1.58 5.13 26.59 0.31 Luno 8 1.47 10.73 77.16 0.31 OMcDnll 9 1.53 4.20 8.48 0.44 18SKit 15 1.14 4.26 21.51 0.30 3 ower "CD > CD i Debris Flood 19SKit Gosnell7 Mill 12 5 4 1.08 1.60 1.06 6.98 3.57 2.16 49.77 14.82 3.68 0.39 0.40 0.88 igh-l CD to Miller 22Shedin 6 1 1.55 1.29 4.66 5.41 12.08 12.41 0.48 0.46 X GosnelM 9 1.40 3.58 5.65 0.61 Gosnell4 7 1.50 4.55 13.25 0.54 Gosnell6 6 1.62 4.38 23.01 0.35 3Copper2 5 0.77 2.12 1.21 1.10 Debris Flow Carrigani Kits3 Legate 10 4 2 1.14 1.51 0.87 2.31 0.78 2.44 2.98 1.59 5.30 0.74 1.19 1.08 Trapln 9 1.08 2.03 1.66 1.04 Canyon 10 1.45 5.31 15.58 0.24 3D 10 0.92 5.91 13.52 0.24 Flood CP095-1 0 1.30 3.48 17.14 0.20 CD CP095-2 8 1.37 3.57 8.88 0.27 O n M3 11 1.33 5.80 20.48 0.37 L L > z Newcmbl 1 1.29 2.27 4.32 0.49 > O _l Debris Carrigan2 SprCmp 11 10 1.12 1.35 1.98 3.30 3.17 3.46 0.76 0.60 Flood Tableland 11 1.34 2.90 12.37 0.35 Poplar 9 1.37 4.14 8.66 0.26 195 APPENDIX G. Continued. Power Disturb-ance Total Shape ReliefRatio DrainDen Process Name Events Hypso (km/kmA2) Alice 8 0.41 0.33 0.16 2.97 Hunter 13 0.46 0.69 0.22 3.81 Flood Sinclair 11 0.52 0.34 0.08 1.95 Tszkwa 0 0.37 0.29 0.09 1.79 Winfield 23 0.53 0.35 0.10 2.14 Middle 4 0.48 0.40 0.14 2.87 16Wllmse 6 0.59 0.24 0.40 1.79 16Wllmsw 9 0.52 0.29 0.52 2.85 Cranberry 7 0.60 0.39 0.27 3.92 Compass 0 0.42 0.29 0.19 2.65 Gosnell8 8 0.49 0.40 0.25 2.66 Ktwancool 7 0.54 0.43 0.23 2.83 25UKit 6 0.60 0.40 0.46 3.24 Debris 39UKit 3 0.57 0.22 0.40 3.01 Flood 3Copper1 5 0.64 0.15 0.46 2.63 8McDnll3 8 0.46 0.32 0.41 1.95 8McDnll1 7 0.51 0.19 0.41 1.80 ManyBear 10 0.51 0.32 0.38 3.02 Sibola 16 0.61 0.25 0.14 2.43 Tetlock 10 0.61 0.34 0.35 4.04 Powerline 6 0.60 0.31 0.29 3.20 W_Bottm 6 0.53 0.28 0.24 3.52 Big_Wdn1 9 0.63 0.19 0.30 2.05 Fernando 15 0.56 0.22 0.50 2.60 Debris Flow Kitsl 5 0.54 0.39 0.49 6.17 Kits2 3 0.49 0.26 0.62 4.64 Kits4 3 0.45 0.28 0.46 7.74 Rico 0 0.49 0.34 0.54 3.71 Wan 12 0.50 0.36 0.39 2.33 CD o Q. . C U> CD > CD _ l •o c CQ CO 1 9 6 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events Hypso Shape Relief Ratio DrainDen (km/kmA2) Flood Shelford 2 0.56 0.27 0.04 3.89 Ailport 5 0.72 0.33 0.06 1.80 8McDnll2 6 0.41 0.14 0.42 2.35 Herb 18 0.56 0.24 0.15 4.22 Luno 8 0.58 0.27 0.16 2.46 OMcDnll 9 0.58 0.22 0.21 2.17 18SKit 15 0.62 1.14 0.32 1.04 3 ower "53 > o i Debris Flood 19SKit Gosnell7 Mill 12 5 4 0.57 0.54 0.54 0.26 0.43 0.30 0.20 0.26 0.48 4.01 2.73 2.61 igh-l o to Miller 22Shedin 6 1 0.43 0.50 0.31 0.16 0.27 0.19 1.78 2.57 X GosnelM 9 0.48 0.24 0.30 1.82 GosnelW 7 0.53 0.20 0.24 3.13 Gosnell6 6 0.60 0.44 0.23 2.71 3Copper2 5 0.59 0.20 0.49 1.34 Debris Flow Carriganl Kits3 Legate 10 4 2 0.53 0.49 0.52 0.26 0.35 0.22 0.38 0.70 0.51 2.14 7.52 4.02 Trapln 9 0.58 0.16 0.42 2.46 Canyon 10 0.47 0.33 0.14 1.70 3D 10 0.61 0.26 0.12 1.47 t_ Flood CP095-1 0 0.60 0.44 0.13 3.22 z Newcmbl 1 0.52 0.26 0.25 3.20 o _ J Debris Carrigan2 SprCmp 11 10 0.50 0.51 0.33 0.16 0.43 0.24 2.47 2.00 Flood Tableland 11 0.55 0.45 0.24 3.24 Poplar 9 0.54 0.25 0.13 2.04 197 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events Comm (%) ESA (%) ESASx (%) G30 (%) Alice 8 39 0 5 45 Hunter 13 32 6 12 62 Flood Sinclair 11 62 7 10 5 Tszkwa 0 63 4 5 6 Winfield 23 57 16 21 7 Middle 4 30 0 4 52 16Wllmse 6 28 0 0 42 16Wllmsw 9 43 25 25 42 Cranberry 7 29 10 12 37 Compass 0 68 14 34 29 Gosnell8 8 10 10 12 38 Ktwancool 7 44 18 27 61 25UKit 6 50 0 0 50 Debris 39UKit 3 50 0 10 23 Flood 3Copper1 5 50 35 35 26 8McDnll3 8 27 6 22 30 8McDnll1 7 23 8 19 33 ManyBear 10 9 1 1 67 Sibola 16 23 3 6 6 Tetlock 10 46 5 13 53 Powerline 6 56 16 23 52 W_Bottm 6 49 1 9 55 Big_Wdn1 9 32 1 11 21 Fernando 15 20 0 0 58 Debris Flow Kitsl Kits2 5 3 34 30 30 22 34 29 79 69 Kits4 3 32 38 38 68 Rico 0 40 14 23 75 Wan 12 26 3 25 57 CD 5 o Q. • g> x Oi > Oi _ l TJ C CO 198 APPENDIX G . Continued. Power Disturb-ance Process Name Total Events Comm (%) ESA (%) ESASx (%) G 3 0 (%) Flood Shelford Ailport 2 5 81 89 1 1 1 2 0 1 8McDnll2 6 29 2 24 27 Herb 18 20 4 4 2 Luno 8 30 5 5 36 OMcDnll 9 33 14 30 12 18SKit 15 38 5 5 12 Power "55 > i Debris Flood 19SKit Gosnell7 Mill 12 5 4 55 10 71 9 1 53 14 6 53 48 36 63 • _: © Miller 6 40 3 16 6 C O 22Shedin 1 70 0 31 1 X Gosnell 1 9 42 4 9 24 Gosnell4 7 31 0 6 28 Gosnell6 6 13 3 10 . 35 3Copper2 5 74 69 69 42 Debris Flow Carrigani Kits3 Legate 10 4 2 93 17 57 2 22 29 5 22 29 45 76 69 Trapln 9 87 0 0 21 Canyon 10 60 6 7 3 3D 10 91 0 0 1 I— a> g Flood CP095-1 0 80 0 3 4 CP095-2 8 65 5 5 10 o M3 11 61 3 13 0 1 > z Newcmbl 1 79 28 59 5 > o _l Debris Carrigan2 SprCmp 11 10 77 83 15 7 17 19 58 4 Flood Tableland 11 53 0 26 9 Poplar 9 89 0 4 1 199 APPENDIX G. Continued. Power Disturb- Process Name Total G35 G40 B3040 G3025 ance Events (%) (%) (%) Alice 8 32 20 25 0.45 Hunter 13 45 28 33 0.65 Flood Sinclair 11 2 1 4 0.04 Tszkwa 0 3 2 4 0.03 Winfield 23 3 1 6 0.11 Middle 4 32 16 35 0.46 16Wllmse 6 18 4 • 38 0.23 16Wllmsw 9 24 12 31 0.63 Cranberry 7 24 11 26 0.36 Compass 0 17 8 21 0.37 Gosnell8 8 18 7 31 0.31 Ktwancool 7 42 19 42 0.47 a> 75 > 25UKit 6 29 12 37 0.57 g o a> _ i Debris 39UKit 3 10 3 21 0.32 0 . • TJ Flood 3Copper1 5 17 13 14 0.33 High Stan 8McDnll3 8 17 8 22 0.39 High Stan 8McDnll1 7 20 10 23 0.45 Stan ManyBear 10 51 34 33 0.60 Sibola 16 2 1 5 0.15 Tetlock 10 34 18 35 0.57 Powerline 6 30 10 42 0.45 W_Bottm 6 32 13 42 0.51 Big_Wdn1 9 7 1 20 0.31 Fernando 15 37 19 40 0.75 Debris Kitsl 5 63 38 41 0.87 Kits2 3 50 29 41 0.76 Flow Kits4 3 54 31 37 0.70 Rico 0 59 38 37 0.85 Wan 12 41 25 32 0.30 200 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events G35 (%) G40 (%) B3040 (%) G3025 Flood Shelford Ailport 2 5 0 0 0 0 0 1 0.00 0.02 8McDnll2 6 15 6 22 0.43 Herb 18 0 0 1 0.03 Luno 8 22 10 26 0.29 OMcDnll 9 5 2 10 0.15 18SKit 15 7 4 8 0.53 Power "CD > CD i Debris Flood 19SKit Gosnell7 Mill 12 5 4 31 16 30 16 5 4 32 31 60 0.50 0.37 0.35 igh-l CD CO Miller 22Shedin 6 1 2 0 1 0 6 1 0.05 0.02 X Gosnell 1 9 7 1 23 0.20 Gosnell4 7 13 3 25 0.21 Gosnell6 6 16 4 32 0.28 3Copper2 5 25 12 29 0.58 Debris Flow Carrigani Kits3 Legate 10 4 2 24 59 52 10 32 29 35 44 40 0.64 0.85 0.72 Trapln 9 7 2 20 0.47 Canyon 10 1 0 2 0.07 3D 10 0 0 1 0.01 >_ CD Flood CP095-1 0 1 0 4 0.03 CP095-2 8 6 3 8 0.18 O f\ < M3 11 0 0 0 0.09 Urn 1 > z Newcmbl 1 2 0 4 0.04 > o _l Debris Carrigan2 SprCmp 11 10 38 1 18 0 40 4 0.70 0.14 Flood Tableland 11 3 1 8 0.13 Poplar 9 0 0 1 0.03 201 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events G3050 G30100 G3525 G3550 Alice 8 0.47 0.49 0.32 0.33 Hunter 13 0.66 0.68 0.49 0.50 Flood Sinclair 11 0.05 0.06 0.02 0.02 Tszkwa 0 0.03 0.04 0.01 0.01 Winfield 23 0.10 0.10 0.04 0.04 Middle 4 0.47 0.51 0.28 0.30 16Wllmse 6 0.27 0.34 0.00 0.00 16Wllmsw 9 0.59 0.57 0.42 0.38 Cranberry 7 0.38 0.39 0.26 0.27 Compass 0 0.36 0.37 0.23 0.22 Gosnell8 8 0.31 0.31 0.13 0.13 Ktwancool 7 0.52 0.58 0.33 0.36 25UKit 6 0.52 0.51 0.39 0.33 Debris 39UKit 3 0.30 0.30 0.16 0.14 Flood 3Copper1 5 0.33 0.35 0.18 0.20 8McDnll3 8 0.39 0.41 0.25 0.26 8McDnll1 7 0.39 0.42 0.28 0.25 ManyBear 10 0.62 0.67 0.46 0.48 Sibola 16 0.15 0.12 0.08 0.08 Tetlock 10 0.57 0.57 0.40 0.38 Powerline 6 0.49 0.51 0.24 0.27 W_Bottm 6 0.53 0.57 0.31 0.32 Big_Wdn1 9 0.32 0.30 0.14 0.14 Fernando 15 0.76 0.80 0.50 0.53 Debris Flow Kitsl 5 0.89 0.88 0.68 0.74 Kits2 3 0.78 0.76 0.62 0.64 Kits4 3 0.73 0.74 0.58 0.60 Rico 0 0.89 0.90 0.70 0.75 Wan 12 0.37 0.43 0.23 0.25 o Q. i JZ Ui i f > a> _ i TJ C (0 CO 202 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events G3050 G30100 G3525 G3550 Flood Shelf ord 2 0.01 0.01 0.00 0.00 Ailport 5 0.02 0.02 0.01 0.01 8McDnll2 6 0.40 0.38 0.30 0.25 Herb 18 0.03 0.02 0.01 0.01 Luno 8 0.30 0.32 0.19 0.20 OMcDnll 9 0.21 0.23 0.06 0.09 18SKit 15 0.52 0.51 0.33 0.32 3 ower Debris Flood 19SKit 12 0.50 0.51 0.36 0.35 3 ower "55 > 0) i Gosnell7 Mill 5 4 0.38 0.45 0.39 0.54 0.17 0.13 0.17 0.18 igh-l V (7) Miller 22Shedin 6 1 0.06 0.01 0.06 0.02 0.01 0.00 0.01 0.00 X GosnelM 9 0.18 0.20 0.08 0.07 Gosnell4 7 0.21 0.23 0.09 0.09 Gosnell6 6 0.28 0.30 0.16 0.13 3Copper2 5 0.72 0.75 0.40 0.53 Debris Flow Carriganl Kits3 10 4 0.62 0.82 0.64 0.82 0.38 0.67 0.38 0.66 Legate 2 0.72 0.74 0.52 0.54 Trapln 9 0.47 0.40 0.23 0.21 Canyon 10 0.09 0.07 0.02 0.03 3D 10 0.01 0.01 0.00 0.00 i— Flood CP095-1 0 0.05 0.05 0.01 0.02 a> S CP095-2 8 0.16 0.14 0.12 0.11 o M3 11 0.07 0.07 0.02 0.01 LL. 1 > z Newcmbl 1 0.04 0.04 0.01 0.02 O _ l Debris Carrigan2 SprCmp 11 10 0.80 0.10 0.82 0.09 0.54 0.04 0.66 0.02 Flood Tableland 11 0.15 0.13 0.05 0.06 Poplar 9 0.02 0.02 0.00 0.00 203 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events G35100 G4025 G4050 G40100 Alice 8 0.35 0.21 0.22 0.23 Hunter 13 0.51 0.32 0.33 0.33 Flood Sinclair 11 0.02 0.01 0.01 0.01 Tszkwa 0 0.02 0.00 0.00 0.01 Winfield 23 0.04 0.02 0.01 0.01 Middle 4 0.32 0.14 0.15 0.17 16Wllmse 6 0.02 0.00 0.00 0.00 16Wllmsw 9 0.34 0.22 0.17 0.17 Cranberry 7 0.27 0.13 0.14 0.13 Compass 0 0.23 0.12 0.12 0.12 Gosnell8 8 0.14 0.03 0.03 0.04 Ktwancool 7 0.41 . 0.16 0.17 0.19 25UKit 6 0.30 0.22 0.18 0.15 Debris 39UKit 3 0.13 0.06 0.05 0.05 Flood 3Copper1 5 0.23 0.12 0.15 0.17 8McDnll3 8 0.26 0.14 0.14 0.13 8McDnll1 7 0.29 0.18 0.15 0.16 ManyBear 10 0.52 0.27 0.28 0.33 Sibola 16 0.06 0.03 0.03 0.02 Tetlock 10 0.38 0.23 0.22 0.22 Powerline 6 0.29 0.08 0.10 0.10 W_Bottm 6 0.34 0.11 0.12 0.12 Big_Wdn1 9 0.11 0.06 0.05 0.04 Fernando 15 0.60 0.20 0.22 0.32 Debris Flow Kitsl 5 0.73 0.43 0.47 0.46 Kits2 3 0.58 0.31 0.35 0.32 Kits4 3 0.60 0.31 0.33 0.34 Rico 0 0.77 0.49 0.51 0.52 Wan 12 0.28 0.13 0.16 0.19 CD —I TJ C CO CO 204 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events G35100 G4025 G4050 G40100 Flood Shelford 2 0.00 0.00 0.00 0.00 Ailport 5 0.01 0.00 0.00 0.00 8McDnll2 6 0.22 0.17 0.14 0.09 Herb 18 0.01 0.00 0.00 0.00 Luno 8 0.21 0.10 0.11 0.11 OMcDnll 9 0.10 0.01 0.03 0.03 18SKit 15 0.31 0.16 0.16 0.16 Power Debris Flood 19SKit 12 0.35 0.21 0.21 0.20 Power "55 > i Gosnell7 Mill 5 4 0.18 0.22 0.06 0.01 0.06 0.01 0.06 0.02 igh-l CD •#-» (75 Miller 22Shedin 6 1 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 X GosnelM 9 0.06 0.02 0.01 0.01 Gosnell4 7 0.10 0.03 0.03 0.02 Gosnell6 6 0.13 0.04 0.04 0.03 3Copper2 5 0.59 0.13 0.21 0.31 Debris Flow Carriganl Kits3 10 4 0.39 0.64 0.20 0.26 0.21 0.31 0.19 0.34 Legate 2 0.56 0.27 0.30 0.33 Trapln 9 0.15 0.11 0.07 0.05 Canyon 10 0.03 0.00 0.01 0.01 3D 10 0.00 0.00 0.00 0.00 i _ Flood CP095-1 0 0.02 0.00 0.00 0.00 CP095-2 8 0.08 0.06 0.05 0.04 o M3 11 0.02 0.01 0.00 0.00 L L > z Newcmbl 1 0.02 0.01 0.01 0.01 > o Debris Carrigan2 SprCmp 11 10 0.63 0.03 0.32 0.00 0.44 0.00 0.36 0.00 Flood Tableland 11 0.05 0.01 0.02 0.01 Poplar 9 0.00 0.00 0.00 0.00 205 APPENDIX G. Continued. Power Disturb- Process Name Total B304025 B304050 B3040100 ance Events Alice 8 0.26 0.27 0.28 Hunter 13 0.36 0.37 0.38 Flood Sinclair 11 0.04 0.04 0.05 Tszkwa 0 0.02 0.03 0.04 Winfield 23 0.10 0.09 0.09 Middle 4 0.33 0.35 0.36 16Wllmse 6 0.23 0.27 0.34 16Wllmsw 9 0.46 0.46 0.43 Cranberry 7 0.25 0.27 0.29 Compass 0 0.27 0.26 0.27 Gosnell8 8 0.29 0.28 0.28 Ktwancool 7 0.34 0.38 0.43 55 "53 25UKit 6 0.37 0.36 0.38 o CD i Debris 39UKit 3 0.27 0.25 0.26 Q. TJ Flood 3Copper1 5 0.22 0.20 0.19 .C CO I Stan 8McDnll3 8 0.27 0.26 0.30 Stan 8McDnll1 7 0.28 0.26 0.29 Stan ManyBear 10 0.37 0.38 0.38 Sibola 16 0.13 0.13 0.11 Tetlock 10 0.37 0.37 0.38 Powerline 6 0.39 0.42 0.44 W_Bottm 6 0.42 0.44 0.47 Big_Wdn1 9 0.26 0.28 0.27 Fernando 15 0.61 0.59 0.53 Debris Kitsl 5 0.48 0.48 0.48 Kits2 3 0.49 0.49 0.48 Flow Kits4 3 0.45 0.42 0.43 Rico 0 0.40 0.43 0.43 Wan 12 0.19 0.24 0.26 206 APPENDIX G. Continued. Power Disturb-ance Process Name Total Events B304025 B304050 B3040100 Flood Shelford 2 0.00 0.01 0.00 Ailport 5 0.02 0.02 0.02 8McDnll2 6 0.27 0.27 0.30 Herb 18 0.03 0.03 0.02 Luno 8 0.21 0.21 0.23 OMcDnll 9 0.14 0.19 0.21 18SKit 15 0.39 0.38 0.38 Power "35 > o i Debris Flood 19SKit Gosnell7 Mill 12 5 4 0.31 0.32 0.35 0.32 0.33 0.45 0.34 0.35 0.54 igh-l CD CO Miller 22Shedin 6 1 0.05 0.02 0.06 0.01 0.06 0.02 X GosnelM 9 0.19 0.17 0.19 GosnelW 7 0.18 0.19 0.21 Gosnell6 6 0.25 0.25 0.28 3Copper2 5 0.49 0.58 0.51 Debris Flow Carriganl Kits3 Legate 10 4 2 0.46 0.62 0.49 0.44 0.56 0.47 0.47 0.52 0.46 Trapln 9 0.38 0.42 0.36 Canyon 10 0.07 0.08 0.07 3D 10 0.01 0.01 0.01 i_ Flood CP095-1 0 0.03 0.05 0.05 CO CP095-2 8 0.12 0.11 0.10 o ft M3 11 0.08 0.06 0.07 UL > z Newcmbl 1 0.03 0.04 0.04 > o _l Debris Carrigan2 SprCmp 11 10 0.42 0.14 0.41 0.10 0.50 0.09 Flood Tableland 11 0.13 0.14 0.12 Poplar 9 0.03 0.02 0.02 207 A P P E N D I X H . D A T A S U M M A R Y O F F O R E S T R Y A C T I V I T I E S O N F A N S 208 A P P E N D I X H . 1. Dendroecology sampling did not identify events but site features provide evidence of events. 2. A second road on a fan. 3. Groundwater ponds along toe of fan were dewatered following channel entrenchment due to road and drainage structure erosion. F a n n a m e Alice Dasque Hunter Sinclair Tsezakwa Winfield Middle F a n n u m b e r 1 2 3 4 5 6 7 F a n t y p e H i g h - p o w e r s t a n d l e v e l f l o o d F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes yes yes yes yes yes Evid. of high sediment load (in w/s or fan) yes yes yes yes Abrupt disappearing stream on fan Abrupt stream angle on fan Sediment source near w/s mouth yes yes yes yes yes M a n a g e m e n t i s s u e P1-major problem/ P2-limited problem/ Nil P1 P1 P1 P1 P1 P1 P1 Associated forestry activity rip.log rip.log bridge bridge road bridge bridge road bridge Pit road slides road Impacts plantat bridge bridge bridge bridge bridge bridge fish fish plantat. road road plantat plantat road plantat fish plantat forest road Date of management activity (r - road, r-1956 r-1956 r-1987 50-80's r-60's 50-80's r-1956 I = logging) 1-1956 1-1956 1-1990 l-60's 1-1956 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 8 13 11 0 23 4 No. char, events since road 3+5 2+3 4+5 0+0 8+10 3+1 No. char, events since logging 3+5 2+2 4+5 0+0 3+1 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes1 yes yes yes yes Non-characterizing event occurred yes No natural event Channel avulsion Channel re-occupation (multi situation) 209 A P P E N D I X H . Continued. F a n n a m e Alice Dasque Hunter Sinclair Tsezakwa Winfield Middle F a n n u m b e r 1 2 3 4 5 6 7 F a n t y p e H i g h - p o w e r s t a n d l e v e l f l o o d R o a d s - n e g a t i v e - o r i g i n a l Road climbing to creek yes yes yes yes Borrow pit in creek yes Ditchline interception/channelling yes yes yes Channel excavated - no rip rap Poor ditchblocks/inadequate X drains yes yes Banks damaged Road at slope break (steep to gentle) yes yes yes Multi channels - inadeq. drain, struct. Drain, struct. Inadequate/none yes yes yes yes yes Multi span struct - bedload/debris issue yes yes Rd not deactivated promptly or inadequ yes Inappropriate engineering works Fan truncated by road Road became creek (avulsion) yes yes yes yes L o g g i n g - n e g a t i v e No/inadequate rip reserve yes yes yes Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream yes Logging related landslides yes yes Stream down skid road/on-block road yes 210 A P P E N D I X H . Continued. F a n n a m e Alice Dasque Hunter Sinclair Tsezakwa Winfield Middle F a n n u m b e r 1 2 3 4 5 6 7 F a n t y p e H i g h - p o w e r s t a n d l e v e l f l o o d R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex new new new new new new Road crosses in area of uniform gradient yes Road crosses on lower fan/stable chan. Road drops across fan to creek then up Road overlanded No ditchline and road outsloped yes Rolling grade conforms to old channels yes Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal Adequate drain struct (perm) or backup yes new new new new Engineered struct to control stream yes Regular maint of structures Deactivation - guick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes yes High stumps Woody debris left on site Partial cut Limited ground disturbance R o a d i m p a c t s Short-term access yes Long-term access yes yes yes yes yes yes Replace structures yes(x8) yes(x3) yes yes yes yes(x4) Relocate road yes yes yes yes yes yes Engineered structures reguired yes yes yes Cost $600k+ 211 A P P E N D I X H . Continued. F a n n a m e Alice Dasque Hunter Sinclair Tsezakwa Winfield Middle F a n n u m b e r 1 2 3 4 5 6 7 F a n t y p e H i g h - p o w e r s t a n d l e v e l f l o o d T i m b e r i m p a c t s Area impacted 5ha 3ha 4ha P l a n t a t i o n o n f a n yes yes yes yes yes Area impacted 15ha 12ha 15ha 2ha F i s h h a b i t a t p r e s e n t yes yes yes yes yes yes yes Impact - minor, moderate, major major major major 212 A P P E N D I X H . Continued. F a n n a m e 16Wllmse 16Wllmsw Big Wdn2 Cranberry Compass Skilokus Ktwancool 25UKit F a n n u m b e r 8 9 10 11 12 13 14 15 F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d ( c o n ' t ) F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes Evid. of high sediment load (in w/s or fan) yes yes yes yes Abrupt disappearing stream on fan Abrupt stream angle on fan yes Sediment source near w/s mouth yes yes yes M a n a g e m e n t i s s u e P1 -major problem/ P2-limited problem/ Nil P2 P2 Nil P1 P1 P1 P2 P1 Associated forestry activity rip.log rip.log rip.log road rip.log bridge rip.log ditch Pit bridge culvert bridge bridge road Impacts ditch road fish road bridges road bridge road forest road bridge fish plantat plantat Date of management activity (r - road, r-1988 r-1988 r-1976 r-1973 r-1995 r-1971 r-1985 r-1987 I = logging) 1-1990 1-1990 1-1991 1-1974 l-40's 1-1971 1-1986 1-1989 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 6 9 7 0 7 6 No. char, events since road 2+0 2+0 4+0 0+0 2+0 2+0 No. char, events since logging . 1+0 2+0 4+0 0+0 2+0 1+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes1 yes yes1 yes yes Non-characterizing event occurred yes No natural event Channel avulsion Channel re-occupation (multi situation) 213 APPENDIX H. Continued. Fan name CD in E 5 CO 16Wllmsw Big Wdn2 Cranberry Compass Skilokus Ktwancool 25UKit Fan number 8 9 10 11 12 13 14 15 Fan type High-power stand level debris flood (con't) Roads - neqative - oriqinal Road climbing to creek yes yes yes yes yes yes yes Borrow pit in creek yes Ditchline interception/channelling yes yes Channel excavated - no rip rap Poor ditchblocks/inadequate X drains yes yes Banks damaged Road at slope break (steep to gentle) Multi channels - inadeq. drain, struct. Drain, struct. Inadeguate/none yes yes yes Multi span struct - bedload/debris issue yes Rd not deactivated promptly or inadequ yes Inappropriate engineering works yes yes Fan truncated by road Road became creek (avulsion) yes Logging - negative No/inadequate rip reserve yes yes yes yes Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides yes Stream down skid road/on-block road 214 A P P E N D I X H . Continued. F a n n a m e CD E 5 C D 16Wllmsw Big Wdn2 Cranberry Compass Skilokus Ktwancool 25UKit F a n n u m b e r 8 9 10 11 12 13 14 15 F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d ( c o n ' t ) R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient yes yes yes yes2 Road crosses on lower fan/stable chan. yes yes yes2 Road drops across fan to creek then up yes yes2 Road overlanded No ditchline and road outsloped Rolling grade conforms to old channels Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal yes Adequate drain struct (perm) or backup yes yes yes2 yes Engineered struct to control stream Regular maint of structures yes yes yes Deactivation - quick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes yes yes yes High stumps Woody debris left on site yes yes Partial cut yes yes Limited ground disturbance yes yes yes R o a d i m p a c t s Short-term access yes yes yes yes Long-term access yes Replace structures yes yes yes yes Relocate road Engineered structures required yes yes Cost 215 A P P E N D I X H . Continued. F a n n a m e 16Wllmse 16Wllmsw Big Wdn2 Cranberry Compass Skilokus Ktwancool 25UKit F a n n u m b e r 8 9 10 11 12 13 14 15 F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d ( c o n ' t ) T i m b e r i m p a c t s Area impacted 1ha P l a n t a t i o n o n f a n yes yes yes yes yes yes yes Area impacted 2ha 2ha F i s h h a b i t a t p r e s e n t yes yes yes yes Impact - minor, moderate, major major major 216 A P P E N D I X H . Continued. F a n n a m e 39UKit 3Copper1 c Q o c o 8McDnll3 Sibola Tetlock Powerline Whitebot F a n n u m b e r 16 17 18 19 20 21 22 23 F a n t y p e ( c o n ' t ) H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes yes yes yes yes Evid. of high sediment load (in w/s or fan) yes Abrupt disappearing stream on fan yes Abrupt stream angle on fan yes Sediment source near w/s mouth yes M a n a g e m e n t i s s u e P1 -major problem/ P2-limfted problem/ Nil Nil P2 Nil P1 P1 Nil P1 P1 Associated forestry activity road bridge bridge culvert road ch.exc road road bridge culvert rip.log rip.log Impacts bridge bridge bridge road bridge road road fish road plantat fish plantat Date of management activity (r - road, r-2001 r-1960 r-pre50 r-pre50 r-pre50 r-1987 r-1973 r-50's I = logging) 1-1994 1-1985 1-1957 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 3 5 7 8 16 10 6 6 No. char, events since road 0+0 3+2 3+4 5+3 8+8 2+1 2+1 5+1 No. char, events since logging 0+0 2+0 5+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes yes yes yes yes Non-characterizing event occurred yes No natural event yes Channel avulsion Channel re-occupation (multi situation) yes 217 A P P E N D I X H . Continued. F a n n a m e 39UKit 3Copper1 c D o CO 8McDnll3 Sibola Tetlock Powerline Whitebot F a n n u m b e r 16 17 18 19 20 21 22 23 F a n t y p e ( c o n ' t ) H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d R o a d s - n e q a t i v e - o r i q i n a l Road climbing to creek yes yes yes yes yes Borrow pit in creek Ditchline interception/channelling Channel excavated - no rip rap yes Poor ditchblocks/inadeguate X drains yes Banks damaged Road at slope break (steep to gentle) yes yes Multi channels - inadeq. drain, struct. yes yes Drain, struct. Inadequate/none yes yes yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ yes Inappropriate engineering works yes Fan truncated by road yes Road became creek (avulsion) yes yes L o g g i n g - n e g a t i v e No/inadeguate rip reserve yes yes Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides Stream down skid road/on-block road 218 A P P E N D I X H . Continued. F a n n a m e 39UKit 3Copper1 8McDnll1 8McDnll3 Sibola Tetlock Powerline Whitebot F a n n u m b e r 16 17 18 19 20 21 22 23 F a n t y p e ( c o n ' t ) H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d R o a d s - p o s i t i v e Road off fan yes yes Road crosses at or close to apex yes yes yes Road crosses in area of uniform gradient yes yes Road crosses on lower fan/stable chan. Road drops across fan to creek then up yes yes Road overlanded No ditchline and road outsloped yes Rolling grade conforms to old channels yes Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal Adequate drain struct (perm) or backup yes yes yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes High stumps Woody debris left on site Partial cut yes Limited ground disturbance yes R o a d i m p a c t s Short-term access yes yes Long-term access yes yes yes Replace structures yes yes yes Relocate road yes Engineered structures required yes yes Cost $2k/yr 219 A P P E N D I X H . Continued. F a n n a m e 39UKit 3Copper1 c Q o CO 8McDnll3 Sibola Tetlock Powerline Whitebot F a n n u m b e r 16 17 18 19 20 21 22 23 F a n t y p e ( c o n ' t ) H i g h - p o w e r s t a n d l e v e l d e b r i s f l o o d T i m b e r i m p a c t s Area impacted P l a n t a t i o n o n f a n yes yes yes Area impacted 10ha 1ha F i s h h a b i t a t p r e s e n t no3 yes Impact - minor, moderate, major mod mod 220 A P P E N D I X H . Continued. F a n n a m e Big Wdn1 Fernando Kitsl Kits2 Kits4 Wan Shelford Ailport F a n n u m b e r 24 25 26 27 28 29 30 31 H i g h - p o w e r s i t e l e v e l f l o o d F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o w F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes Evid. of high sediment load (in w/s or fan) Abrupt disappearing stream on fan yes yes yes yes Abrupt stream angle on fan Sediment source near w/s mouth yes yes M a n a g e m e n t i s s u e P1 -major problem/ P2-limited problem/ Nil P1 P2 P2 P2 P2 P2 P1 Nil Associated forestry activity rip. log rip-log skid rd. rip.log rip.log rip.log culvert road skid rd. road rip.log Impacts plantat plantat skid rd. plantat plantat plantat culvert road skid rd. road plantat fish Date of management activity (r - road, r-1976 r-1987 r-2000 r-1993 r-1992 r-1992 r-1996 I = logging) 1-1991 1-1987 I-93/00 1-1993 1-1992 1-1993 1-1972 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 9 15 5 3 3 12 2 5 No. char, events since road 4+1 2+3 0+0 0+0 0+0 3+0 2+0 0+0 No. char, events since logging 3+0 2+3 0+0 0+0 0+0 3+0 2+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes1 yes1 yes yes Non-characterizing event occurred yes1 yes1 No natural event Channel avulsion yes yes Channel re-occupation (multi situation) 221 A P P E N D I X H . Continued. F a n n a m e Big Wdn1 Fernando Kitsl Kits2 Kits4 Wan Shelford Ailport F a n n u m b e r 24 25 26 27 28 29 30 31 H i g h - | s o w e r e v e l o d F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o w s i t e f l o R o a d s - n e q a t i v e - o r i q i n a l Road climbing to creek yes yes yes yes yes yes Borrow pit in creek Ditchline interception/channelling yes Channel excavated - no rip rap Poor ditchblocks/inadequate X drains yes Banks damaged Road at slope break (steep to gentle) yes Multi channels - inadeq. drain, struct. Drain, struct. Inadequate/none yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ yes Inappropriate engineering works Fan truncated by road Road became creek (avulsion) yes yes L o g g i n g - n e g a t i v e No/inadequate rip reserve yes yes yes yes yes yes Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides Stream down skid road/on-block road yes yes 222 A P P E N D I X H . Continued. F a n n a m e Big Wdn1 Fernando Kitst Kits2 Kits4 Wan Shelford Ailport F a n n u m b e r 24 25 26 27 28 29 30 31 H i g h - p o w e r s i t e l e v e l f l o o d F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o w R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient Road crosses on lower fan/stable chan. yes yes yes yes yes yes Road drops across fan to creek then up yes2 Road overlanded No ditchline and road outsloped yes yes Rolling grade conforms to old channels yes Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal yes Adeguate drain struct (perm) or backup yes yes yes yes yes yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal yes L o g g i n g - p o s i t i v e Riparian reserve yes High stumps yes Woody debris left on site yes Partial cut Limited ground disturbance yes R o a d i m p a c t s Short-term access yes Long-term access yes Replace structures yes Relocate road Engineered structures reguired Cost 223 A P P E N D I X H . Continued. F a n n a m e Big Wdn1 Fernando Kitsl Kits2 Kits4 Wan Shelford Ailport F a n n u m b e r 24 25 26 27 28 29 30 31 H i g h - p o w e r e v e l o d F a n t y p e H i g h - p o w e r s t a n d l e v e l d e b r i s f l o w s i t e f l o T i m b e r i m p a c t s Area impacted P l a n t a t i o n o n f a n yes yes yes yes yes yes yes Area impacted 4ha 1ha 1ha 2 ha 1ha 5ha F i s h h a b i t a t p r e s e n t yes yes Impact - minor, moderate, major major 2 2 4 A P P E N D I X H . Continued. F a n n a m e 8McDnll2 Herb OMcDnll 18SKit 19SKit Miller F a n n u m b e r 32 33 34 35 36 37 F a n t y p e H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d ( c o n ' t ) F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes yes Evid. of high sediment load (in w/s or fan) yes Abrupt disappearing stream on fan Abrupt stream angle on fan yes Sediment source near w/s mouth M a n a g e m e n t i s s u e P1-major problem/ P2-limited problem/ Nil P2 P1 P1 Nil Nil P1 Associated forestry activity culvert rip.log culvert culvert road rip.log ch.exc rip.log Impacts culvert road road culvert road plantat plantat road fish fish plantat forest Date of management activity (r - road, r-pre50 r-1993 r-pre50 r-1990 r-1990 pre50 I = logging) 1-1993 1-1983 1-1991 1-1991 1-1985 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 6 18 9 15 12 6 No. char, events since road 5+1 3+1 6+3 3+1 1+2 4+2 No. char, events since logging 3+1 4+1 3+1 1+2 2+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes yes yes yes Non-characterizing event occurred No natural event Channel avulsion yes Channel re-occupation (multi situation) yes 225 A P P E N D I X H . Continued. F a n n a m e 8McDnll2 Herb OMcDnll 18SKit 19SKit Miller F a n n u m b e r 32 33 34 35 36 37 F a n t y p e H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d ( c o n ' t ) R o a d s - n e q a t i v e - o r i q i n a l Road climbing to creek yes yes yes Borrow pit in creek Ditchline interception/channelling yes Channel excavated - no rip rap Poor ditchblocks/inadequate X drains yes Banks damaged Road at slope break (steep to gentle) yes Multi channels - inadeq. drain, struct. yes Drain, struct. Inadequate/none yes yes yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ yes yes Inappropriate engineering works yes Fan truncated by road Road became creek (avulsion) yes L o g g i n g - n e g a t i v e No/inadequate rip reserve yes yes yes Damage to banks LOD removed from stream LOD in stream burned yes Channel used as skid trails/road Landing intercept water/close to stream yes Logging related landslides Stream down skid road/on-block road yes 226 A P P E N D I X H . Continued. F a n n a m e 8McDnll2 Herb OMcDnll 18SKit 19SKit Miller F a n n u m b e r 32 33 34 35 36 37 F a n t y p e H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d ( c o n ' t ) R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient yes Road crosses on lower fan/stable chan. Road drops across fan to creek then up yes Road overlanded No ditchline and road outsloped Rolling grade conforms to old channels Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal Adequate drain struct (perm) or backup yes yes Enqineered struct to control stream Regular maint of structures Deactivation - quick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes yes High stumps Woody debris left on site Partial cut Limited ground disturbance R o a d i m p a c t s Short-term access yes yes yes Long-term access Replace structures yes yes yes Relocate road Engineered structures required Cost 227 A P P E N D I X H . Continued. F a n n a m e 8McDnll2 Herb OMcDnll 18SKit 19SKit Miller F a n n u m b e r 32 33 34 35 36 37 F a n t y p e H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d ( c o n ' t ) T i m b e r i m p a c t s Area impacted 1ha P l a n t a t i o n o n f a n yes yes yes Area impacted 2ha 1ha 2ha F i s h h a b i t a t p r e s e n t yes yes yes yes yes Impact - minor, moderate, major major major 228 A P P E N D I X H . Continued. F a n n a m e 22 Shedin Gosnelh GosnelW Gosnell6 F a n n u m b e r 38 39 40 41 ( c o n ' t ) H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d F a n t y p e F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan Evid. of high sediment load (in w/s or fan) Abrupt disappearing stream on fan yes yes Abrupt stream angle on fan Sediment source near w/s mouth yes M a n a g e m e n t i s s u e P1-major problem/ P2-limited problem/ Nil P1 P1 Nil Nil Associated forestry activity ch.exc road road bridge Impacts culvert road road bridge forest Date of management activity (r - road, r-1999 r-1996 r-1998 r-1998 I = logging) 1-1996 1-1999 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 1 9 7 6 No. char, events since road 0+0 0+0 1+0 0+1 No. char, events since logging 0+0 1+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes1 yes yes Non-characterizing event occurred yes1 No natural event Channel avulsion Channel re-occupation (multi situation) yes 229 A P P E N D I X H . Continued. F a n n a m e 22 Shedin GosnelH Gosnell4 Gosnell6 F a n n u m b e r 38 39 40 41 ( c o n ' t ) H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d F a n t y p e R o a d s - n e q a t i v e - o r i q i n a l Road climbing to creek yes yes yes Borrow pit in creek Ditchline interception/channelling yes yes Channel excavated - no rip rap yes Poor ditchblocks/inadequate X drains yes yes Banks damaged yes Road at slope break (steep to gentle) Multi channels - inadeq. drain, struct. Drain, struct. Inadequate/none yes yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ Inappropriate engineering works yes Fan truncated by road Road became creek (avulsion) yes yes L o g g i n g - n e g a t i v e No/inadequate rip reserve Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides Stream down skid road/on-block road 230 A P P E N D I X H . Continued. F a n n a m e 22 Shedin GosnelM GosnelM Gosnell6 F a n n u m b e r 38 39 40 41 ( c o n ' t ) H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d F a n t y p e R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex yes Road crosses in area of uniform gradient yes yes Road crosses on lower fan/stable chan. yes Road drops across fan to creek then up yes Road overlanded yes No ditchline and road outsloped yes Rolling grade conforms to old channels Rip-rap used to stabilize ck. excavation Adeguate drain struct - seasonal Adeguate drain struct (perm) or backup yes yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes yes High stumps Woody debris left on site yes yes Partial cut Limited ground disturbance yes yes R o a d i m p a c t s Short-term access yes yes Long-term access Replace structures yes yes Relocate road Engineered structures required Cost $10k 231 A P P E N D I X H . Continued. F a n n a m e 22 Shedin Gosnelh GosnelW Gosnell6 F a n n u m b e r 38 39 40 41 ( c o n ' t ) H i g h - p o w e r s i t e l e v e l d e b r i s f l o o d F a n t y p e T i m b e r i m p a c t s Area impacted 1ha P l a n t a t i o n o n f a n yes yes Area impacted F i s h h a b i t a t p r e s e n t yes Impact - minor, moderate, major 232 A P P E N D I X H . Continued. F a n n a m e 3Copper2 Carriganl Trapline F a n n u m b e r 42 43 44 H i g h - p o w e r s i t e l e v e l d e b r i s f l o w F a n t y p e F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes Evid. of high sediment load (in w/s or fan) Abrupt disappearing stream on fan Abrupt stream angle on fan Sediment source near w/s mouth yes M a n a g e m e n t i s s u e P1 -major problem/ P2-limited problem/ Nil P2 Nil Nil Associated forestry activity road ch.exc culvert Impacts road bridge Date of management activity (r - road, r-1960 r-1999 r-1998 I = logging) I-2000 1-1999 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 5 10 9 No. char, events since road 2+3 0+0 0+0 No. char, events since logging 0+0 0+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes Non-characterizing event occurred yes yes1 yes1 No natural event Channel avulsion Channel re-occupation (multi situation) 233 A P P E N D I X H . Continued. F a n n a m e 3Copper2 Carrigani Trapline F a n n u m b e r 42 43 44 H i g h - p o w e r s i t e l e v e l d e b r i s f l o w F a n t y p e R o a d s - n e q a t i v e - o r i q i n a l Road climbing to creek yes yes yes Borrow pit in creek Ditchline interception/channelling Channel excavated - no rip rap yes Poor ditchblocks/inadequate X drains Banks damaged Road at slope break (steep to gentle) Multi channels - inadeq. drain, struct. Drain, struct. Inadequate/none yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ Inappropriate engineering works Fan truncated by road yes Road became creek (avulsion) L o g g i n g - n e g a t i v e No/inadeguate rip reserve yes Damage to banks LOD removed from stream LOD in stream burned Channel used as skid trails/road Landing intercept water/close to stream yes Logging related landslides Stream down skid road/on-block road 234 A P P E N D I X H . Continued. F a n n a m e 3Copper2 Carriganl Trapline F a n n u m b e r 42 43 44 H i g h - p o w e r s i t e l e v e l d e b r i s f l o w F a n t y p e R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient yes yes Road crosses on lower fan/stable chan. Road drops across fan to creek then up Road overlanded No ditchline and road outsloped yes Rolling grade conforms to old channels yes yes Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal yes yes Adequate drain struct (perm) or backup yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal yes yes L o g g i n g - p o s i t i v e Riparian reserve yes yes High stumps yes Woody debris left on site yes yes Partial cut Limited ground disturbance yes yes R o a d i m p a c t s Short-term access yes Long-term access Replace structures Relocate road Engineered structures required yes Cost $1k/yr 235 A P P E N D I X H . Continued. F a n n a m e 3Copper2 Carriganl Trapline F a n n u m b e r 42 43 44 H i g h - p o w e r s i t e l e v e l d e b r i s f l o w F a n t y p e T i m b e r i m p a c t s Area impacted P l a n t a t i o n o n f a n yes yes Area impacted F i s h h a b i t a t p r e s e n t Impact - minor, moderate, major 236 A P P E N D I X H . Continued. F a n n a m e Q CO CP095-1 CP095-2 CO Mckndrki Newcmbi Canyon F a n n u m b e r 45 46 47 48 49 50 51 F a n t y p e L o w - p o w e r f l o o d F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan Evid. of high sediment load (in w/s or fan) Abrupt disappearing stream on fan Abrupt stream angle on fan Sediment source near w/s mouth yes yes yes M a n a g e m e n t i s s u e P1 -major problem/ P2-limited problem/ Nil Nil P2 P1 Nil P1 Nil P1 Associated forestry activity ditch culvert Iwd rip.log road road ditch culvert bridge rip.log Impacts road road culvert road plantat road fish plantat forest fish plantat Date of management activity (r - road, r-1999 r-1984 r-1982 r-pre50 r-1985 r-2000 r-pre50 I = logging) I-2000 1-1984 1-1982 1-1986 1-1987 1-1987 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 10 8 11 1 10 No. char, events since road 0+1 0+1 3+8 0+0 4+6 No. char, events since logging 0+0 0+1 1+3 2+0 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes yes1 yes Non-characterizing event occurred yes1 yes1 No natural event Channel avulsion Channel re-occupation (multi situation) 237 APPENDIX H. Continued. Fan name Q CO CP095-1 CP095-2 CO Mckndrki Newcmbi Canyon Fan number 45 46 47 48 49 50 51 Fan type Low-power flood Roads - neqative - oriqinal Road climbing to creek yes yes yes yes yes Borrow pit in creek Ditchline interception/channelling yes yes yes Channel excavated - no rip rap Poor ditchblocks/inadeguate X drains yes yes yes Banks damaged yes yes Road at slope break (steep to gentle) yes yes Multi channels - inadeq. drain, struct. Drain, struct. Inadeguate/none yes yes yes Multi span struct - bedload/debris issue Rd not deactivated promptly or inadequ yes yes Inappropriate engineering works Fan truncated by road Road became creek (avulsion) yes yes yes Logging - negative No/inadeguate rip reserve yes yes Damage to banks LOD removed from stream yes LOD in stream burned yes Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides Stream down skid road/on-block road yes 238 APPENDIX H. Continued. F a n n a m e Q CO CP095-1 CP095-2 CO Mckndrki Newcmbi Canyon F a n n u m b e r 45 46 47 48 49 50 51 F a n t y p e L o w - p o w e r f l o o d R o a d s - p o s i t i v e Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient yes yes yes Road crosses on lower fan/stable chan. yes yes Road drops across fan to creek then up Road overlanded yes No ditchline and road outsloped yes Rolling grade conforms to old channels Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal Adequate drain struct (perm) or backup yes yes yes yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal L o g g i n g - p o s i t i v e Riparian reserve yes yes yes yes High stumps Woody debris left on site yes Partial cut yes Limited ground disturbance yes yes yes yes yes yes R o a d i m p a c t s Short-term access yes yes yes yes Long-term access yes Replace structures yes Relocate road Engineered structures required Cost 239 A P P E N D I X H . Continued. F a n n a m e Q CO CP095-1 CP095-2 CO Mckndrki Newcmbi Canyon F a n n u m b e r 45 46 47 48 49 50 51 F a n t y p e L o w - p o w e r f l o o d T i m b e r i m p a c t s Area impacted 4ha P l a n t a t i o n o n f a n yes yes yes yes yes Area impacted 1ha 2ha 1ha F i s h h a b i t a t p r e s e n t yes yes yes Impact - minor, moderate, major major major 240 A P P E N D I X H . Continued. F a n n a m e SprCmp Tableland Carrigan 3 Poplar F a n n u m b e r 52 53 54 55 F a n t y p e L o w - p o w e r d e b r i s f l o o d F e a t u r e s f r o m p r e - l o g g i n g a i r p h o t o s Multiple channels on fan yes Evid. of high sediment load (in w/s or fan) Abrupt disappearing stream on fan Abrupt stream angle on fan Sediment source near w/s mouth yes yes M a n a g e m e n t i s s u e P1-major problem/ P2-limited problem/ Nil P2 P1 P2 P1 Associated forestry activity culvert rip.log Iwd culvert culvert rip.log road rip.log Impacts road culvert plantat culvert culvert road road fish plantat plantat forest fish fish forest Date of management activity (r - road, r-pre50 r-1979 r-1978 r-1981 I = logging) 1-1979 1-1992 1-1981 N o . c h a r a c t e r i z i n g e v e n t s - p a s t 5 0 y r s 10 11 9 No. char, events since road 2+8 2+4 2+1 No. char, events since logging 2+4 2+1 N a t u r a l h y d r o g e o m o r p h i c e v e n t Characterizing event occurred yes yes yes Non-characterizing event occurred yes1 No natural event Channel avulsion Channel re-occupation (multi situation) 241 A P P E N D I X H . Continued. Fan name SprCmp Tableland Carrigan 3 Poplar Fan number 52 53 54 55 Fan type Low-power debris flood Roads - neqative - oriqinal Road climbing to creek yes yes yes Borrow pit in creek Ditchline interception/channelling yes Channel excavated - no rip rap Poor ditchblocks/inadequate X drains yes Banks damaged Road at slope break (steep to gentle) yes yes Multi channels - inadeq. drain, struct. Drain, struct. Inadequate/none yes yes yes Multi span struct - bedload/debris issue yes Rd not deactivated promptly or inadequ yes Inappropriate engineering works Fan truncated by road Road became creek (avulsion) yes Logging - negative No/inadequate rip reserve yes yes yes Damage to banks LOD removed from stream yes LOD in stream burned yes Channel used as skid trails/road Landing intercept water/close to stream Logging related landslides Stream down skid road/on-block road yes 242 A P P E N D I X H . Continued. Fan name SprCmp Tableland Carrigan 3 Poplar Fan number 52 53 54 55 Fan type Low-power debris flood Roads - positive Road off fan Road crosses at or close to apex Road crosses in area of uniform gradient yes Road crosses on lower fan/stable chan. Road drops across fan to creek then up Road overlanded No ditchline and road outsloped Rolling grade conforms to old channels Rip-rap used to stabilize ck. excavation Adequate drain struct - seasonal Adequate drain struct (perm) or backup yes Engineered struct to control stream Regular maint of structures Deactivation - quick, seasonal Logging - positive Riparian reserve High stumps Woody debris left on site Partial cut Limited ground disturbance yes yes yes Road impacts Short-term access yes yes yes Long-term access Replace structures yes yes yes Relocate road Engineered structures required Cost 243 A P P E N D I X H . Continued. F a n n a m e SprCmp Tableland Carrigan 3 Poplar F a n n u m b e r 52 53 54 55 F a n t y p e L o w - p o w e r d e b r i s f l o o d T i m b e r i m p a c t s Area impacted 1ha 2ha P l a n t a t i o n o n f a n yes yes yes Area impacted 1ha 1ha 2ha F i s h h a b i t a t p r e s e n t yes yes yes Impact - minor, moderate, major mod major major 244