natural disturbance dynamics in coastal british columbiab. dorner and c. wong, natural disturbance...

Natural Disturbance Dynamics in Coastal British

Columbia

Prepared by

Brigitte Dorner, Ph.D., Carmen Wong, MRM. for the Coast Information Team

B. Dorner and C. Wong, Natural Disturbance Dynamics in Coastal BC May 15, 2003

i

Acknowledgements

This review is based in part on a review of natural disturbance dynamics the authors prepared for the North Coast LRMP. While we retained the original structure, the content of individual sections was updated to incorporate additional information that has since become available and to reflect the larger, more diverse area covered in this report.

We would like to thank again the many people who contributed to the report on North Coast disturbance dynamics:

• Hannah Horn, Technical Coordinator, North Coast LRMP • Allen Banner, MOF Research Ecologist • John Corsiglia • Davide Cuzner, MSRM LIM Officer • Brian Fuhr, MSRM • Kevin Haukaas, MOF Protection Branch • Rachel Holt, Veridian Ecological Consulting • Ken Lertzman, SFU • Sarma Liepins, MSRM Strategic Planning Biologist • Jim MacDonald, UNBC • Andy MacKinnon, Ministry of Sustainable Resource Management • Charles Menzies, UBC • John Parminter, MOF Research Ecologist • Audrey Pearson, Consultant • Brad Pollack, Acer Consulting • Jim Pojar, MOF, Research Ecologist • Shawn Reed, MSRM GIS • Don Reid, MSRM, Wildlife Inventory Habitat Specialist • Holger Sandmann, Consultant • Doug Steventon, MOF, Wildlife Habitat Biologist • Nancy Turner, University of Victoria • Len Vanderstar, WLAP, Ecosystem Specialist • Amy Waterhouse, MOF GIS • Ken White, MOF, Regional Entomologist • Alex Wood, MOF, Forest Pathologist

In addition, we thank Andy MacKinnon, John Parminter, and Jim Pojar for helping us to track down additional material. Thanks also especially to Steve Taylor for making available his unpublished analysis of the provincial fire database, and to Audrey Pearson and Joseph Fall for reviewing earlier drafts of this manuscript.


Page ii

Executive Summary

Natural disturbances, along with the activities of First Nations peoples, have historically shaped the composition and structure of forests on the coastal landscape. Natural disturbances can impact timber production by damaging or killing trees, but they are also agents of diversity and renewal that ensure a steady supply of important habitat elements such as snags, fallen trees, and browsing opportunities for many species. A good understanding of natural disturbance dynamics is thus an important first step towards maintaining a sustained timber supply, as well as sustaining ecosystem function and habitat characteristics to which native species have adapted. This report provides background information on natural disturbance regimes and identifies forest management issues arising from our understanding of natural forest dynamics in Haida Gwaii, the North and Central Coast, and northern Vancouver Island.

The landscape in the area is varied, consisting of lowlands, intricate coastlines, extensive floodplains, and coastal mountain ranges with deeply incised valleys and steep-sided fjords. The moist, cool climate, along with physiographic and topographic characteristics, are influential drivers of forest pattern and disturbance dynamics. In the low-lying coastal sections and exposed slopes near the outer coast, patches of productive forest often occur within a matrix of non-forested bog lands and stunted bog woodland. In the watersheds of the insular and mainland mountain ranges, the forest mosaic typically consists of large, continuous tracts of all-aged, structurally diverse old-growth coastal western hemlock (CWH) forest, separated by cliffs, gullies, wetlands, and shrub-covered avalanche and landslide tracks.

Over the majority of the coastal landscape, stand-replacing disturbances are rare and stands are very old. Death and reestablishment of trees primarily occurs in small canopy gaps of ten trees or fewer. These gaps are typically created by a combination of wind and pathogens, and make up around 10 to 30 percent of the area in old-growth stands. Geomorphic disturbances are the most important naturally occurring high-severity, stand-replacing events in the area. Wind and fire may also occasionally create larger canopy openings, but return intervals for these larger events are on the scale of millennia throughout the majority of the area. Flooding is a key element of forest dynamics in floodplains and estuaries. Other agents that may occasionally injure or kill trees include snow, ice, frost, and drought, as well as insects and mammals, most notably porcupines, beavers, and deer. The different types of disturbances do not occur homogeneously across the forest landscape. Rather, most of the agents are confined to, or occur predominantly in specific stand types and on specific site types or landscape positions. Below, we give a brief description of the most important disturbance agents, followed by management considerations arising from the current scientific knowledge of natural disturbance dynamics in the area.

Blowdown, largely attributable to the frequent cyclonic storms originating over the Pacific, is one of the key disturbance agents in the coastal landscape. The severity of wind effects is determined by topographic exposure to prevailing winds, as well as tree species and growth form, stand structure and site characteristics, and the amount of precipitation accompanying


Page iii

the storm. The vast majority of openings created by wind are small canopy gaps. Patches of partial or complete canopy blowdown up to several hundred hectares in size have been reported along the coast of British Columbia and southeast Alaska. However, stand replacing blowdown is likely very rare in the CIT area, except possibly for the most wind-exposed parts of the landscape, such as the west coast of Haida Gwaii. The primary mode of mortality during windstorms is stem snap. Fallen trees serve an important ecological role as large woody debris, which is a key feature of both terrestrial and hydro-riparian habitat. Uprooting of trees during windstorms, although comparatively rare, provides habitat for various species and helps to maintain soil productivity.

Geomorphic disturbances, such as landslides and snow avalanches are the primary natural agents of stand-replacing disturbance in coastal forests. Most geomorphic disturbances occur on steep slopes and are favoured by wet conditions. Tsunamis associated with earthquakes or landslides may play an important role in low-lying coastal areas. Geomorphic disturbances dislodge trees, soil, and rocks and can deliver sediments and woody debris considerable distances to valley floors, often modifying stream channels and floodplains in the process. Jams of large woody debris alter flow patterns, create a complex channel morphology, and protect downstream habitat and back channels from flood scouring. Geomorphic disturbances also create early-seral communities that provide forage opportunities in close proximity to cover for many species. Slides generate more soil disturbance than snow avalanches and can contribute towards maintaining long-term forest productivity in those areas where mineral soil is deposited. Any disturbance that removes tree cover and changes local drainage patterns, including clearcut harvesting and road building, can substantially increase the risk of landslides and avalanchesincidence of landslides is approximately 10 to 20 times higher in logged areas than in natural terrain. Severe rainstorms and earthquakes have, in the past, triggered major landslide episodes in natural terrain, consisting of a large number of slide events scattered throughout a wide area. Since the most recent of these events occurred in the first half of the 20th century, their potential impact on harvested terrain is still unknown.

Floods, triggered by rainstorms, rain on snow, or rapid snowmelt, can cause varying degrees of tree mortality. However, flooding rarely kills a substantial number of mature trees unless debris flows or torrents are triggered. Flooding and associated mass movements are confined to specific landscape positions, and typically influence only a small proportion of area in a watershed. Floodplains, fans and estuaries subject to flooding are highly productive biodiversity hotspots. Depending on disturbance frequency and severity, they support forests of various composition and ages, including deciduous forests, stands of massive spruce and hemlock, as well as shrub, berry and herb communities.

With the exception of the drier, easternmost parts of the Coast Mountains, lightning caused fires are rare on the coast, due to the cool, wet climate. Research on historical (pre-European-settlement) fire regimes in coastal BC is limited to Vancouver Island and the southern mainland. The closest published study of fire in coastal western hemlock forests (CWHvm1) stems from Clayoquot Valley, Vancouver Island. There, the average time since the most recent fire ranged between 750 and 4500 years ago. Dry, south-facing slopes are more likely to burn than other parts of the landscape, as are stands dominated by western redcedar occurring on dry sites. Intentional low-severity burning by First Nations peoples to enhance berry production has been documented for the area. Aboriginal burning of cedar to


Page iv

hollow out canoes is also a probable cause of origin for several fire-regenerated cedar stands observed on the mid- and north coast. Fire affects post-disturbance succession by reducing thick organic layers, improving nutrient availability, providing a good seed bed for western redcedar and removing advance regeneration of mistletoe infected western hemlock.

Decay fungi, including heart, butt, and root rots, are the most important biotic disturbance agents in coastal BC. The parasitic hemlock dwarf mistletoe also plays an important role. Although pathogens rarely cause direct mortality in older trees, they can predispose infected trees to blowdown and other modes of disturbance. Thus, pathogens enhance structural complexity and accelerate succession to all-aged stands with old-growth characteristics. Decay fungi spread primarily through air-born spores and enter trees through wounds, which may be caused by falling trees during a windstorm, mammal damage, or activities related to forest management. Both fungi and mistletoe infections seem to develop relatively slowly at higher latitudes, and infection may show serious effects only after trees reach several hundred years of age.

Insects play a less important role in coastal forests than in many other forest types in BC, though they have had considerable impacts in some instances. Mature western hemlock can suffer growth reduction and some mortality after consecutive years of defoliation by several types of budworm and loopers. Stands of young Sitka spruce can be deformed by weevils, and feeding by aphids has caused high mortality in some areas. The recent outbreak of mountain pine beetle in interior British Columbia has caused considerable mortality in lodgepole pine stands in the easternmost part of the Central Coast.

Mammals may also affect forest dynamics where conditions are favourable. Porcupines, which are limited to the mainland coast, may kill or injure trees and thus contribute to gap dynamics, especially in second-growth stands. Beavers use saplings and pole-sized trees for food and dam construction and can substantially affect hydrology and vegetation dynamics in hydroriparian zones. Heavy browsing by deer can sometimes denude the understory and impede tree regeneration. The impact of deer is especially dramatic on Haida Gwaii, where deer were introduced in the early 1900s and have since spread throughout the islands, favoured by the mild climate and virtually unchecked by natural predators. Beavers are also an introduced species on Haida Gwaii and have been associated with substantial alterations in wetland dynamics and flooding of important shoreline habitat.

Timber harvesting in the area is having a substantial impact not only on stand and landscape structure, but also on natural disturbance regimes. In many cases, harvesting increases the frequency and/or severity of natural disturbances. Thus, harvesting promotes occurrence of landslides and erosion, and can increase the severity of flooding. By opening up the canopy, harvesting also makes stands more susceptible to blowdown, and the wounding of trees during harvesting operations encourages fungal infections. Furthermore, several disturbance agents, including root rots, porcupines, and some defoliating insects, are more prevalent in harvested landscapes since they primarily affect younger, second-growth stands. On the other hand, harvesting is likely to reduce the incidence of disturbance agents that are slow to develop or affect primarily older stands, such as heart and butt rots and mistletoe.

Climate change will likely also increase the frequency or severity of many natural disturbance types. Future climate change in coastal BC is predicted to bring increases in average temperatures by 1.5 to 4°C, as well as increases in winter precipitation and severity


Page v

of winter storms. Predictions also include glacial retreat, increased flooding especially in the spring, and flooding of coastal areas. More severe winter storms will likely increase the incidence and severity of blowdown, and higher precipitation during winter will likely result in more, and more severe, landslides, erosion, and flooding. Milder winter temperatures favour most biotic agents, increasing both their impact and their range. By affecting fuel moisture, climate change will likely also influence fire regimes. However, the direction of change is uncertain in this case, as there is a fair amount of uncertainty in projections regarding summer precipitation levels.

The character of natural disturbance dynamics in coastal forests raises some important issues and concerns for forest management. We have subdivided these issues into three general areas: spatial and temporal scales of natural processes, use of natural disturbance patterns to guide forest management, and species conservation and habitat considerations.

(1) What are appropriate spatial and temporal scales for planning and management?

• Landscape dynamics on the coast are strongly influenced by the underlying mosaic of physiography and topography. The paradigm of a shifting mosaic of seral stagesi that underlies the management approach proposed in the Biodiversity Guidebook does not fit this type of landscape very well, since much of the early-seral vegetation is confined to a few areas subject to repeated disturbance, whereas most of the operable, productive forests of interest to harvesting rarely experience severe, large-scale disturbance. A planning approach that fails to take this into consideration will make it difficult to incorporate knowledge about natural dynamics in a meaningful way. A stratification into watershed and landscape sections dominated by different disturbance types and regime would be a logical subdivision of the landscape in terms of maintaining ecological processes and their outcomes. Such a stratification could also help identify planning areas where special management is required to avoid undesirable interactions with dominant disturbance agents.

• Climate is also a significant driver of forest dynamics on the coast. Predicted changes in future climate will likely increase the incidence and/or severity of most disturbance agents and affect how forests respond to these agents and other processes. Consideration of intermediate and long-term climatic trends is therefore essential for anticipating and planning future forest development.

• The coastal forest is, in many respects, a system of long time scales. Tree species are long-lived, intervals between disturbances are measured on the scale of centuries, and processes of decay and decomposition progress equally slowly. Thus, undesired effects of forest management, such as potential negative impacts on terrestrial and aquatic habitat because of reduced inputs of large woody debris, may take a long time to manifest and even longer to reverse. This suggests a need for careful long-term

i This paradigm applies to a landscape where forest dynamics are similar across the landscape and occurrence of natural

disturbances is not restricted to particular topographic or other site characteristics. In such a landscape, the spatial distribution of young and old forest throughout the landscape is similarly independent of topography or other site characteristics, and the forest mosaic therefore ‘shifts’ at random across the landscape over time as disturbances create new openings and older openings gradually merge back into the matrix of old forest.


Page vi

planning and continued monitoring if negative ecosystem impacts are to be minimized.

• While coastal forests are primarily driven by gap dynamics, larger, more severe natural disturbance events do occur episodically in some parts of the landscape. Proactive planning and forecasting over long time horizons will be required to avoid negative impacts of such rare events on sustained timber supply and biodiversity. This could include considering the potential effect of earthquakes, tsunamis, and landslides on infrastructure and taking into account losses to episodic severe natural disturbances in timber supply analyses as well as land use planning and conservation strategies.

(2) How can natural disturbance regimes serve as a guide for forest management?

• The forests on the BC coast are predominantly considered NDT (Natural Disturbance Type) 1 in the Biodiversity Guidebook, with intervals between stand replacing disturbances averaging 250 to 350 years. Although some small sections of the coastal landscape may historically have experienced such relatively frequent stand-replacing disturbances, there is no evidence to support the disturbance intervals postulated in the Guidebook on a broader scale. The scientific evidence also indicates that in forests considered productive and operable, disturbance return intervals for large, stand-replacing events were generally much longer than commercial rotation periods.

• Because coastal forests are primarily driven by gap dynamics, natural disturbance regimes provide no template for clearcutting systems, or traditional approaches to even-aged management in general. If the objective of forest management is to emulate natural disturbance patterns, a silvicultural system that emulates a gap-phase replacement regime, such as variable retention with high retention levels, is the closest match. In cases where high retention levels are rejected because of economical considerations, the potential risk of negative ecological impacts from silvicultural practices that do not maintain natural structure and dynamics may be reduced through extended rotations, and by constraining the proportion of the landbase over which harvesting occurs. However, if a reduction in harvesting landbase is to be offset by more intensive management zones, it will be important to carefully consider the allocation of such areas, since in many cases the most productive areas from a timber growth perspective also tend to be of primary importance for biodiversity.

• Most of the interactions between harvesting and natural disturbance on the coast are synergistic, i.e., harvesting is likely to increase occurrence of natural disturbances. Since climate change will most likely also increase the incidence or severity of most natural disturbances, overall disturbance rates are likely to go up in the future even without harvesting, and harvesting will almost certainly lead to disturbance rates substantially higher than the landscape has historically experienced. This emphasizes the need to monitor landscape conditions to recognize and avert potential negative impacts on ecosystem function and biodiversity.


Page vii

• Although natural disturbances are a key aspect of forest dynamics, forests are shaped and maintained by a variety of factors that are partially or completely outside the scope of this report. Even if harvesting is designed to emulate natural disturbances very closely, consideration of other factors is essential for maintaining ecosystem integrity and productivity. These factors include watershed hydrology, nutrient cycling, soil development, availability of seed sources, as well as genetic health of tree populations.

(3) One of the key principles of ecosystem-based management is that species are adapted to, and therefore most likely to persist under the environment and processes they have experienced historically. What are the key habitat characteristics historically maintained by natural dynamics, and how is forestry affecting these characteristics?

• The majority of the coastal forest was historically in all-aged stands dominated by gap-phase replacement. Reducing the supply of this type of habitat may negatively impact old-growth-dependent species.

• Historically, stand replacing disturbances were largely constrained to susceptible locations within the landscape, whereas late seral forest was persistent throughout the remainder of the forested area. Current areas of persistent, late seral forest can act as habitat refugia for species dependent on them. Continued existence of such refugia can be ensured by setting aside old growth areas and by managing parts of the landscape to develop and retain old growth characteristics.

• Stands regenerating after clearcut harvesting tend to go through a relatively long period where the canopy is very dense and little light reaches the understory. Understory plants, and species dependent on them, may drop out of the stand during this period. In areas where clearcut harvesting is extensive, requirements for persistence and recolonization of such species may therefore need to be considered if these species are to be maintained in the landscape.

• The historical landscape mosaic provided routes for movement and dispersal of species dependent on particular habitat characteristics. Where harvesting or other forms of development interrupt these routes, the ability of species to move and disperse throughout the landscape may be impacted.

• Several species in the area rely on the juxtaposition of early-seral herb and shrub communities to forest cover for protection and denning. Where forest management creates larger openings and reduces the amount of small canopy gaps characteristic of old-growth forests, the result may be a lack of suitable cover for some species.

• Standing dead and downed wood are important habitat elements in both terrestrial and aquatic ecosystems. Natural disturbances generally ensure an ample, sustained supply of snags, root wads, and downed wood. Reduced recruitment of snags and large woody debris in managed stands may have long-term effects on nutrient cycling, tree regeneration, and habitat quality.

• Productive valley bottom areas and estuaries are focal areas for forest planning and management since they often have high economic timber value as well as very high


Page viii

habitat value. Narrow riparian buffers are susceptible to blowdown and may not provide effective protection of hydro-riparian ecosystems. In order to sustain input of large woody debris into stream channels at historical levels and allow for the natural meandering of floodplain channels throughout the floodplain area, most, if not the entire valley bottom might need to be protected in some cases.

• Introduced species have had substantial impacts on the island ecosystem of Haida Gwaii, as well as on the North American continent in general. Forestry, through silvicultural trials and harvesting operations, has in the past contributed to the introduction and spread of forest pests and weeds. Since many of the species that turn into ‘problem species’ are strong competitors in disturbed areas, it is important for forest managers to be aware of the potential risks posed by introduced species, as well as the means by which these species are spread.


Page ix

Table of Contents

Acknowledgements __________________________________________________________ i

Executive Summary ________________________________________________________ ii

1 Introduction __________________________________________________________ 1 1.1 Background and Rationale __________________________________________________ 1

1.2 Study Area________________________________________________________________ 3 1.2.1 The Physical Landscape ________________________________________________________ 3 1.2.2 Physiography_________________________________________________________________ 4 1.2.3 The Climate _________________________________________________________________ 5 1.2.4 The Forests __________________________________________________________________ 6 1.2.5 First Nations _________________________________________________________________ 7 1.2.6 European Settlers______________________________________________________________ 8

1.3 Climate Trends and Climate Change Predictions________________________________ 8

2 Natural Disturbance Agents______________________________________________ 9 2.1 Wind ___________________________________________________________________ 10

2.1.1 Disturbance dynamics _________________________________________________________ 10 2.1.2 Disturbance effects and post-disturbance succession _________________________________ 12 2.1.3 Factors influencing susceptibility, severity of effects, and post-disturbance succession ______ 13 2.1.4 Direct and indirect effects on habitat______________________________________________ 14 2.1.5 Interactions with other disturbance agents, forest management, and climate change _________ 14

2.2 Snow, Ice, and Hail________________________________________________________ 15

2.3 Geomorphic Disturbances __________________________________________________ 15 2.3.1 Earthquakes _________________________________________________________________ 16 2.3.2 Tsunamis ___________________________________________________________________ 16 2.3.3 Volcanic eruptions____________________________________________________________ 17 2.3.4 Snow avalanches _____________________________________________________________ 17 2.3.5 Mass wasting________________________________________________________________ 18 2.3.6 Interactions with other disturbance agents, forest management, and climate change _________ 21

2.4 Flooding_________________________________________________________________ 22 2.4.1 Disturbance dynamics _________________________________________________________ 22 2.4.2 Interactions with other disturbance agents, forest management, and climate change _________ 25

2.5 Fire_____________________________________________________________________ 26 2.5.1 Disturbance dynamics _________________________________________________________ 26 2.5.2 Traditional burning by First Nations ______________________________________________ 27 2.5.3 Disturbance effects and post-disturbance succession _________________________________ 28 2.5.4 Factors influencing susceptibility and severity of effects ______________________________ 29 2.5.5 Interactions with other disturbance agents, forest management, and climate change _________ 29

2.6 Short-term Climatic Disturbances ___________________________________________ 30

2.7 Pathogens _______________________________________________________________ 31 2.7.1 Fungi ______________________________________________________________________ 31 2.7.2 Hemlock dwarf mistletoe ______________________________________________________ 33 2.7.3 The role of pathogens in the forest ecosystem_______________________________________ 34 2.7.4 Interactions with other disturbance agents, forest management, and climate change _________ 34


Page x

2.8 Insects __________________________________________________________________ 35 2.8.1 Defoliators__________________________________________________________________ 35 2.8.2 Bark Beetles ________________________________________________________________ 35 2.8.3 Interactions with other agents of change ___________________________________________ 36

2.9 Browsing by Mammals ____________________________________________________ 37 2.9.1 Porcupines __________________________________________________________________ 37 2.9.2 Beavers ____________________________________________________________________ 38 2.9.3 Deer_______________________________________________________________________ 38 2.9.4 Interactions with other agents of change ___________________________________________ 40

3 Natural Disturbance Regimes Throughout the Coastal Landscape: An Overview__ 41

4 Effect of Natural Disturbances on Habitat Requirements of Forest-dwelling Species45

5 Management Issues ___________________________________________________ 46

Appendix 1: Summary of Disturbance History Studies in Table Form _______________ 63

List of Tables

Table 1: Fire incidence by ecoregion for the period 1950 - 1992 _____________________ 27

Table 2: Common fungal species in coastal British Columbia. ______________________ 33

Table A. Characteristics of stand-replacing disturbance agents______________________ 42

Table B: Disturbance agents that vary substantially with site characteristics____________ 43

Table C: Disturbance agents that vary little with site characteristics __________________ 43

Summary Table 1: Gap characteristics for coastal British Columbia and southeast Alaska 63

Summary Table 2: Landslide characteristics for coastal British Columbia. ____________ 64

Summary Table 3: Blowdown characteristics for southeast Alaska and coastal British Columbia. _______________________________________________________________ 66

Summary Table 4: Snow avalanche characteristics for the northern Cordillera. _________ 67

Summary Table 5: Wildfire characteristics for coastal British Columbia ______________ 68


Page 1

1 Introduction

1.1 Background and Rationale

Natural disturbances, along with the activities of First Nations peoples, have historically maintained the composition and structure of forest stands and landscapes on the coast, and natural disturbances will almost certainly continue to play a significant role in future landscape dynamics. Thus, a good understanding of the factors driving the frequency, severity, and spatial pattern of natural disturbances, and of the effects natural disturbances have on the ecosystem is an important first step towards achieving forest management goals such as sustained timber supply and maintenance of habitat characteristics to which native species have adapted.

This report provides background information on natural disturbance regimes and forest management issues arising from our understanding of natural forest dynamics in coastal British Columbia. For each disturbance agent, we document the currently available knowledge on historical occurrence, effects on forest and landscape structure, and habitat characteristics considered important for maintaining biodiversity, as well as interactions with other disturbance agents, forest management and climate change. We further present a conceptual framework for understanding natural disturbance dynamics, vegetation structure, and implications for biodiversity in a landscape context, as well as a list of strategic planning issues arising out of current scientific understanding of natural disturbance dynamics in coastal ecosystems. The final section of the report outlines the most important gaps in knowledge about natural disturbances on the BC coast.

In preparing this report, we relied on published scientific literature, augmented by unpublished material and information provided by specialists. Where available, studies from the area covered by the CIT (see Figure 1) were given the highest priority. However, we also drew heavily from the published literature from other coastal regions, particularly southeast Alaska, the southern mainland, the part of Vancouver Island not covered by the CIT, and the US Pacific Northwest. We discuss potential limitations and caveats in cases where we believe that the applicability of results and conclusions to the CIT area might be limited.


Page 2

Figure 1: Area covered by the CIT (Source: Coast Information Team, http://www.citbc.org).


Page 3

1.2 Study Area

1.2.1 The Physical Landscape (Holland 1965, Clague 1989, Church 1994) British Columbia’s coastline lies at the junction of three tectonic plates, the Pacific, Juan de Fuca/Explorer, and America plates. As a result, the area is both volcanically and seismically active. The primary center of activity along the BC coast lies offshore along the Queen Charlotte Fault, but moves onshore over subducting margins both in Alaska and south of the US border.

The landscape of coastal British Columbia is mountainous, dominated by the coastal part of the Canadian Cordillera, a series of rugged mountain ranges extending along the continental margin. Formed as a result of tectonic uplift beginning approximately 10 million years ago, these coastal mountain ranges rise more than 2500 m above sea level, with an additional submarine relief of 750 m. Whereas the inner coast is now either stable or subsiding, uplift continues at the outer coast at rates of up to 2 mm/a for some areas.

The western section of the mainland coast is mostly underlain by intrusive igneous rocks. Bedrock formations on Haida Gwaii consist of a mix of folded sedimentary and volcanic rocks (mainly in the southern part), as well as flat lying or gently dipping lavas (on most of Graham Island) and sedimentary rocks (on the northeastern tip of Graham Island). In the northern part of Vancouver Island, folded and faulted sedimentary and volcanic rocks dominate, though intrusive igneous rocks are also present.

Glaciation, beginning in the northern Canadian Cordillera more than 9 million years ago, has profoundly shaped the character of the coastal landscape. Glacial scouring has sculpted valleys with steep, glacially eroded slopes and created a highly dissected coastline, with fjords extending up to 150 km inland. As glaciers retreated, they left a legacy of unstable sediments susceptible to erosion and mass wasting. Near the coast, delayed recovery from depression of the land surface under the weight of ice resulted in flooding of coastal valleys and accumulation of highly erodible marine silts.

Due to the steep topography in the area, geomorphic processes, covered in detail in Section 2.3, play an important role in shaping both vegetation and physical landscape. The modern landscape is vertically zoned with respect to its geomorphology, largely owing to its tectonic and glacial history. The highest summits are glaciated and primarily subject to glacial and peri-glacial processes of erosion and deposition, with frost weathering affecting exposed areas. The steep slopes and gullies along the sides of the trough-shaped valleys are susceptible to snow avalanches and episodic mass wasting. Lower down, on the gentler slopes near the valley bottom, more continuous fluvial processes dominate. Whereas in valleys with narrow valley bottoms sediments are delivered directly into streams, fluvial processes in larger glacial troughs often result in sediment build-up in form of colluvial footslopes, as well as debris or alluvial fans.


Page 4

1.2.2 Physiography (Holland 1965, Demarchi 1996) The area covered in this review consists of Haida Gwaii (Queen Charlotte Islands) a group of approximately 150 islands located 50 to 130 km off the north coast of British Columbia, as well as the central coast and north coast sections of the British Columbia mainland coast, and the northern part of Vancouver Island. The area is part of the Western System of the Canadian Cordillera, which is divided into three major subdivisions: the Outer Mountain Area, comprising (within the CIT area) the insular mountain ranges on Haida Gwaii and Vancouver Island; the Coast Mountain Area, comprising (within the CIT area, from north to south) the Boundary, Kitimat, and Pacific Ranges; and the Coastal Trough, a lowlying, partially submerged area separating the Outer Mountain and Coast Mountain Areas.

The Queen Charlotte Mountains and the Vancouver Island Ranges lack the extreme relief and extensive glaciation characteristic of the St. Elias Mountains, which form the northern extension of the Outer Mountain Area onto the mainland. The southern part of the Queen Charlotte Mountains, comprising the Queen Charlotte Ranges rises just above 1000 m above sea level and is characterized by very rugged terrain and serrated peaks. The northern part of the Queen Charlotte Mountains consists of the Skidegate Plateau, a dissected, eastward-sloping plateau landscape. The Vancouver Island Ranges are similar in relief to the Coast Mountains, with a very dissected topography especially in the western part (the Windward Island Mountains), and peaks well exceeding 2,000 m.

The terrestrial part of the Coastal Trough comprises, within the CIT area, the Queen Charlotte Lowlands, in the northeastern part to Graham Island, the Hecate Lowland, a strip of islands and low-lying country extending westwards of the Coast Mountain Area, as well as the Nahwitti Lowland on the northern Tip of Vancouver Island and the northern part of the

Figure 2: Coastal ecoregions


Page 5

Georgia Depression (Outer Fiordland) straddling the eastern part of Vancouver Island and the mainland coast. The terrain within the Coastal Trough is generally glacially abraded and often poorly drained. Wetlands are common, and forest is often stunted and occurs in discontinuous patches. However, whereas the Queen Charlotte and Nahwitti Lowlands have a comparatively gentle topography, dominated by plains and low hills, the Hecate Lowland and Outer Fiordland are characterized by rugged terrain with an intricate coastline and deeply incised fjords.

The Coast Mountain Ranges are characterized by a steep relief and often very rugged terrain, with serrated peaks, deeply incised fjords, and glacially eroded valleys. Mountains in the area rise well above 2000 m, sometimes straight from sea level. The Boundary, Kitimat and Pacific Ranges are separated from each other by the Nass River and the Burke Channel-Bella-Coola River valley respectively, and from the St. Elias Mountains to the northwest by the Tatsenshini and Chilkat Rivers. Relief in the Kitimat Ranges is somewhat more subdued than in the surrounding mountain ranges, allowing considerable amounts of moisture to penetrate into the interior. Therefore, the central portion of the CIT area, consisting of Kitimat Ranges and the Hecate Lowland, is termed the Coastal Gap Ecoregion in British Columbia's Ecoregion classification.

1.2.3 The Climate (Church 1994, Pojar et al. 1999, Tamblyn and Horn 2001, Environment Canada 2002)

Moderated by the Pacific Ocean, the climate in the area is generally maritime, with moist, cool summers and mild, wet winters. Annual average temperatures near sea level range around 8°C, with July means between 12 and 16°C and January means between -3 and 4°C. Although prolonged spells of warm, dry weather sometimes occur during the summer months, cloud cover is common throughout the year, and fog is frequent, especially near the ocean. Cyclonic storm systems forming in the North Pacific and the Gulf of Alaska bring large amounts of precipitation, especially during the winter months. Windward slopes receive the brunt of this precipitation. Values above 3 m/year are common, and extremes may reach more than 6 m/year on upper mountain slopes. While much of the winter precipitation falls as snow at higher elevations and in inland areas, low-lying coastal areas typically experience little snow accumulation.

The Coast Mountains form a barrier to westerly atmospheric ciculation, resulting in strong east-west climatic gradients. On Haida Gwaii annual precipitation ranges from an estimated

Figure 3: Hecate Lowland

Figure 4: Kitimat Ranges


Page 6

5+m/year on some windward slopes of the Queen Chalotte Ranges to approximately 1 m/year in the eastern lowlands. On the mainland and Vancouver Island, similar precipitation gradients prevail between windward and leeward sections of the coastal mountain ranges, moderated on the mainland coast by the deeply incised inlets, which allow clouds to penetrate deep into the mountain ranges. Between outer and inner coast mountains, the east-west precipitation gradient is also accompanied by a more general climatic gradient, as hyper-maritime climate characteristic of the coastal lowlands gradually shifts towards a more continentally influenced climate with warmer, drier summers and colder winters in the eastern parts.

1.2.4 The Forests (Banner et al. 1989, Church 1994, Pojar et al. 1999, Prince Rupert Protected Areas Team 2001a and b, Pojar 2003)

Coniferous temperate rainforests dominate the vegetation in the area. In the watersheds of the coastal mountain ranges, the forest mosaic typically consists of large, continuous tracts of all-aged, structurally diverse old-growth coastal western hemlock (CWH) forest, separated by cliffs, gullies, wetlands, and shrub-covered avalanche and landslide tracks. In the low-lying coastal sections, patches of productive forest occur within a matrix of non-forested bog lands and stunted bog woodland.

In the northern part of the mainland coast and on Haida Gwaii, western Hemlock (Tsuga heterophylla, Hw), western redcedar (Thuja plicata, Cw), and Sitka Spruce (Picea Sitchensis, Ss) dominate closed canopy forests at lower elevations, replaced by mountain hemlock (Tsuga mertensiana, Hm) and yellow-cedar (Chamaecyparis nootkatensis, Cy) at higher elevations. In the southern part of the mainland coast, Amabilis Fir (Abies amabilis, Ba) often forms a substantial component of stands, especially at middle and higher elevations, whereas Sitka Spruce tends to be mostly confined to wetter sites, especially floodplain forests. Subalpine fir (Abies lasiocarpa) occurs in the subalpine zone of the inner coastal mountain ranges, and is also sometimes found at lower elevations, on sites subject to drainage of cold air from glacier-headed valleys. Douglas-fir (Pseudotsuga menziesii, Fd) can be found in the southern part of the mainland coast, and on Vancouver Island, especially on warmer aspects and in the rainshadow of the mountain ranges. Lodgepole pine (Pinus contorta var. latifolia, Pl) and Engelmann spruce (Picea engelmanni) also occur in interior forests in the easternmost portion of the mainland coast.

Various types of wetlands are common throughout the area. Riparian forests and swamps are dominated by Sitka spruce, western hemlock, and red alder (Alnus ruba) joined by yellow-cedar at higher elevations, and black cottonwood and willows (Salix spp.) in the eastern part of the area. Hypermaritime peatlands and scrubby bog forests and woodlands commonly occur on poorly drained terrain, especially in the lowlands of the Figure 5: Stunted bog forest


Page 7

Coastal Trough (Hecate and Queen Charlotte Lowlands) and on the eastern Skidegate Plateau. Forested and non-forested boglands are also often found on windward slopes underlain by nutrient-poor, resistant bedrock. Bog forests are typically composed of western redcedar, western and/or mountain hemlock, shore pine (Pinus contorta var. contorta), and often yellow-cedar.

1.2.5 First Nations (Hebda and Mathewes 1984, Ames and Maschner 1999, Boyd 1999, Dalton 2003)

Evidence of human habitation on the Pacific Coast dates back 10,000 years, and it is likely that the first migrants came to the area much earlier.ii First Nations life was generally focused around the shorelines, where the abundant marine life provided an easily accessible food source, and waterways could be used as routes of transport through an otherwise difficult to penetrate landscape. The people lived as hunter-gatherers, congregating in large towns or villages with sometimes over 1,000 inhabitants during winter, and often scattering into smaller groups in summer to take advantage of the resources in the area. Forests were used to hunt game, harvest medicinal and food crops, and to collect firewood, as well as raw materials for buildings and various cultural objects. Western redcedar in particular has

played a dominant role in First Nations culture in coastal BC, providing a light, easily shaped, and rot-resistant material for long houses, sea-going canoes, totem poles and other ceremonial objects, as well as cooking and storage containers. In addition, the bark of the cedar tree was used to fabricate clothing, rope, and matting.

Estimates of historical (pre-contact) population levels vary widely. However, archaeologists generally agree that they were high for hunter-gatherer societies, ranging from somewhere around 200,000 to well over a million people (Ames and Maschner 1999). Thus, pre-contact population densities may well have exceeded modern-day population densities in parts of the coast. Following European contact, populations were sharply decimated by a variety of diseases introduced by the Europeans. The most deadly of these was smallpox, which swept the coast in a series of epidemics, reducing populations by as much as 90%, or more. The first documented epidemic occurred in the late 18th century, roughly coincident with the arrival of the first Europeans in the area. However, it is possible that smallpox introduced by the Spanish to Mexico in the early

ii Although much of coastal BC and Alaska was glaciated during the last glaciation ending approximately 12,000 years ago, an

increasing body of evidence indicates that at least some of the outer coastline remained ice-free, allowing for human migration into and through the area (see Dalton 2003).

Figure 6: Haida Totem pole at K'uuna (Skedans)


Page 8

1500’s may have reached the coast much earlier, resulting in population decline well-preceding the arrival of Europeans to the coast.

Given the high pre-contact population densities in the area, First Nations people likely had substantial impact on local ecosystems, especially around village sites. Probably the most significant impact of First Nations on forest dynamics was through fire, either started accidentally, or set intentionally to enhance berry production.

1.2.6 European Settlers (Banner et al. 1989, Pojar et al. 1999, Tamblyn and Horn 2001)

Although activities such as trapping, mining, and clearing of land for building and agriculture and introduction of new species have had substantial impacts on terrestrial and aquatic ecosystems on the coast, the largest and most wide-spread impacts of European settlers have likely been through logging and the associated road network. Logging on the coast begun early, with the arrival of the first settlers during the 19th century. The first larger, commercial sawmills started up around the turn of the 19th century, but most operations stayed relatively small and consisted of A-frame logging in areas easily accessible from shore. Towards the middle of the 19th century, logging became more intensive and industrialized and moved further inland. Since the 1940s, logging has been carried out almost exclusively through clearcutting, although the use of variable retention systems is increasing, especially on Haida Gwaii. Harvesting patterns are constrained by topography and distribution of merchantable timber. On Haida Gwaii and Vancouver Island, logging often extends all the way up the mountain sides. In the Coast Mountain Ranges on the mainland coast, harvesting is mostly constrained to the more productive, more easily accessible valley bottom areas.

The scope of this review is limited to natural disturbance agents and does not address harvesting or other forms of forest disturbance associated with European settlement per se. However, we do address interactions between natural and anthropogenic disturbances, as well as expected effects of anthropogenic disturbances and climate change on natural disturbance dynamics.

1.3 Climate Trends and Climate Change Predictions

Climate is an important driver for various aspects of natural disturbances – temperature and precipitation patterns influence insect populations, distribution of diseases (Williams et al. 2000), mass wasting (Septer and Schwab 1995), blowdown (Everham and Brokaw 1996), fire frequency and severity (Gavin 2000), as well as soil development, tree growth, and ecosystem composition (DeGroot et. al. 2002). It is now well accepted that greenhouse gas emissions are having significant impacts on global climate and will likely continue to do so in the future (Albritton and Filho 2001). If knowledge of natural disturbance dynamics is to be used to inform and guide future forest management, it is essential to understand how climate change may affect these dynamics. In the following, we give a brief overview of recent climatic trends and scientific predictions for the next century. Specific implications regarding expected trends in frequency and severity for different disturbance agents are discussed in the sections pertaining to these agents.


Page 9

In the past century, average temperature in coastal British Columbia has increased by 0.4°C (Taylor and Taylor 1997). Growing Degree Days, an indicator of heat energy available for plant growth, increased by 13% in the Coastal Mountains over the same time period (British Columbia Ministry of Land Water and Air Protection 2002). Predictions of the Canada Country Study (CCS, Taylor and Taylor 1997) for coastal BC to the year 2050 include:

• A temperature increase of 1 to 4.5°C in winter and 1.5 to 4°C in summer.

• Up to 40% increase in winter precipitation. Predictions for summer precipitation are less consistent and range from a 30% decrease to a 30% increase.

• It is not clear, whether the combination of milder temperatures and increased precipitation in winter will increase or decrease snow accumulation. Potentially there might be more snow at high elevations, but less snow at low elevations.

• Increased intensity of the Aleutian Low, resulting in more severe winter storms; the Canadian Climate Centre (CCC) model predicts 4% decrease in overall frequency of cyclones in the northern hemisphere, but increases in the frequency of intense cyclones.

• A 30 cm rise in sea levels on the northern BC coast. This is less than in other areas because of continuing crustal rebound and tectonic uplift.

• Increased spring flooding.

• Increased sedimentation and coastal flooding in low gradient intertidal areas.

• Increased summer drought.

• Glacial retreat and accompanying flow reductions in glacier-fed rivers.

2 Natural Disturbance Agents

The following section provides an overview of abiotic and biotic disturbance agents of relevance in coastal forests. Gap dynamics, driven by wind and pathogens, are the primary mode of disturbance throughout coastal forests in BC. Geomorphic disturbances can play an important role in susceptible parts of the landscape. They are the main naturally occurring high-severity, stand-replacing events in the area, although wind and fire may also occasionally create larger canopy openings. Flooding is a driving agent of forest dynamics in floodplains and estuaries. Other agents that may occasionally injure or kill trees include snow, ice, and hail, frost and drought, insects, and mammals, most notably porcupines, beavers, and deer. In this section we review the available scientific information for each of these agents. As far as possible, we describe size, spatial distribution, and frequency, as well as disturbance effects and post-disturbance succession. We also briefly outline the role of each agent in maintaining ecosystem function and biodiversity, and discuss interactions with


Page 10

other disturbance agents, forest management, and climate change. A synthesis of disturbance dynamics for different sections of the coastal landscape can be found in Section 3. Appendix 1 summarizes in table format estimates for disturbance return intervalsiii and opening sizes for the most important stand-replacing disturbance agents, as provided in the scientific literature.

2.1 Wind

2.1.1 Disturbance dynamics Wind is a major disturbance agent on the BC Coast. High winds on coastal BC typically occur in conjunction with cyclonic storm systems forming over the Northern Pacific, the majority of which occur during the fall and winter months (Salmon 1997). Wind direction during these storms is roughly southerly, veering from easterly to westerly as the frontal system passes through. However, topographic features such as inlets and mountain valleys may channel or deflect winds, creating substantial local variation in primary wind direction. Pacific frontal systems tend to weaken as they move further inland, creating a regional gradient of wind exposure that mirrors the general climatic east-west gradient characteristic of the area. In the mainland valleys and inlets, katabatic (outflow) winds can be locally more important as agents of wind disturbance than pacific weather systems (Harris 1989, Mitchell 1998, Pojar et al. 1999).

The effects of strong winds on trees range from loss of foliage and branches to small gap formation to partial or complete overstory blowdown (Harris 1989, Foster and Boose 1992, Everham and Brokaw 1996, Parminter 1998, Pearson 2000, Ulanova 2000). The heavy precipitation that tends to accompany storm systems saturates soils, reduces root adhesion, and increases the weight of tree crowns and thus makes trees more susceptible to damage (Everham and Brokaw 1996).

Disturbance frequency and temporal dynamics Events causing mild to moderate damage to trees and creating canopy gaps occur frequently enough to be considered endemic. Total area in gaps in coastal productive forests is often around 10 to 20%, though values as low as 4% and as high as 34% have been observed (Lertzman et al. 1996, Veblen and Alaback 1996, Nowacki and Kramer 1998, Ott and Juday 2002, see also Summary Table 1 in Appendix 1).iv Because of the uncertainty surrounding the time it takes for gaps to fill, it is difficult to estimate return intervals for gap replacement events. Estimates of turnover time (average interval between successive gap creation events) range from 230 to 920 years in southeast Alaska (Ott and Juday 2002), and from 280 to 1000 years on the South Coast of British Columbia (Lertzman and Krebs 1991).

iii The term '(disturbance) return interval' is used in this report to mean the average time between consecutive disturbance

events at a given point in the landscape. Thus, the return interval is the inverse of disturbance frequency, which is the average, or expected number of disturbance events per year occurring at a given point.

iv Note that in studies of gap dynamics a clear distinction between wind and other causes of tree fall is not always possible. Thus, not all of the gaps observed in the cited studies may be attributable to wind.


Page 11

Large-scale blowdown resulting in larger canopy openings is primarily an episodic phenomenonrare, severe events are typically responsible for most such catastrophic blowdown (Harris 1989, Sinton 1996, Mitchell 1998, Nowacki and Kramer 1998). On the islands and possibly also the mainland coast of southeast Alaska, catastrophic blowdown is frequent enough in wind-exposed areas to control forest development and structure over a considerable part of the land base (Harris 1989, Nowacki and Kramer 1998, Kramer in prep., Kramer et al. 2001). For those locations most prone to severe blowdown, intervals between partial or complete canopy replacement are short enough to prevent development of all-aged stand structure, which likely translates into estimated return intervals of 300 years or less (Nowacki and Kramer 1998, Kramer et al. 2001). While it is possible that catastrophic blowdown occurs with similar frequencies locally in some very wind-prone areas along the BC coast, such catastrophic events appear to be rare throughout most, if not all of the CIT area (Schoonmaker and Poerll 1997, Mitchell 1998, Pearson 2000, Pearson 2003, see also Summary Table 3 in Appendix 1). This conclusion is also supported by several studies of stand structure on the South Coast (Lertzman 1992, Arsenault 1995, Daniels et al. 1995), Vancouver Island (Pearson 2000), and the Central Coast (Schoonmaker and Poerll 1997), none of which found evidence of the single or multi-cohort age structure expected after relatively frequent catastrophic wind disturbance.

Spatial pattern While maximum opening sizes observed for southeast Alaska, British Columbia, and other parts of the west coast of North America can range to hundreds of hectares (see Summary Table 3 in Appendix 1), the vast majority of openings crated by wind are small (e.g., Schoonmaker and Poerll 1997, Harris 1989, Sinton 1996, Veblen and Alaback 1996, Mitchell 1998, Nowacki and Kramer 1998, Parminter 1998, Pearson 2003). The size distribution of openings is typically ‘reverse-J’ shaped, and most of the openings consist of small canopy gaps (< 10 trees). The size distribution for canopy gaps is typically also skewed to the left, i.e., small gaps occur more frequently than bigger gaps (Lertzman et al. 1996, Veblen and Alaback 1996, Nowacki and Kramer 1998 see also Summary Table 1 in Appendix 1).

Whereas gap-sized openings occur throughout the landscape, larger openings are more typically found in forests susceptible to blowdown (see Section 2.1.3 below for a description of factors that make stands susceptible). Where gap dynamics are the primary mode of disturbance, the spatial ‘grain’ of the forest mosaic is very fine, determined by canopy gap size. At coarser scales of observations the forest appears as a homogeneous blanket of multi-aged trees. Where large openings are part of the wind regime, the resulting landscape mosaic can be patchy at multiple scales, made up of single cohort stands, stands with multiple even-aged cohorts, and all-aged stands maintained primarily through gap dynamics (Harris 1989, Everham and Brokaw 1996, Nowacki and Kramer 1998, Ulanova 2000, Kramer et al. 2001).

Like fire, catastrophic blowdown does not always lead to complete canopy replacement and the internal remnant structure in blowdown patches is quite variable. For example, in the North Coast forest district the proportion of trees blown down within individual openings ranged from 30%, which was the cut-off for defining a patch as ‘blown down’, to 100% (Mitchell 1998). There was no clear link between the magnitude of disturbance events, as measured in terms of total area blown down, and the proportion of canopy trees blown down


Page 12

in the openings created during those events (Mitchell 1998). The same seems to hold for other areas: large windstorms create a variety of opening sizes with a variety of internal remnant structure, and there is no clear link between the magnitude of a particular event and the severity of blowdown within the openings created by that event (e.g.,Harris 1989, Everham and Brokaw 1996, Nowacki and Kramer 1998, Ulanova 2000, Kramer et al. 2001)

2.1.2 Disturbance effects and post-disturbance succession Tree fall during wind events may be through stem snap or uprooting, and trees may be blown down directly or brought down by other falling trees (Everham and Brokaw 1996). In the coastal forests of British Columbia and southeast Alaska, the primary mode of mortality is stem snap, whereas uprooting is comparatively rare (e.g., Lertzman et al. 1996, Veblen and Alaback 1996, Pearson 2000, see also Summary Table 1 in Appendix 1). Blowdown primarily affects canopy trees and sub-canopy trees, although mechanical damage from falling trees and branches and litter accumulation tend to cause some understory mortality (Everham and Brokaw 1996, Veblen and Alaback 1996).

Because the understory is mostly left intact, regeneration after wind disturbance tends to follow different pathways than regeneration after mass wasting or fire. Gaps created by windfall are primarily filled through lateral growth of adjacent canopy trees or through the release of seedlings and saplings present in the understory (Deal et al. 1991, Ott 1997). Presence of large, decaying logs further favours establishment of advance regeneration (e.g., Harmon and Franklin 1989, Ott 1997). Whereas western hemlock, western redcedar, and amabilis fir can maintain themselves in a gap replacement regime, the somewhat less shade-tolerant Sitka spruce may need larger openings to regenerate successfully (Ott 1997).

Depending on the size of the opening, the proportion of canopy trees blown down, and the return time of wind events, post-disturbance stands may develop even-aged, multi-cohort, or all-aged characteristics (Kramer et al. 2001). However, even in single-cohort stands established after severe wind disturbance, structural and species diversity is significantly higher than in even-aged stands established after clearcutting (Nowacki and Kramer 1998, Price et al. 1998).

Root mounds created by uprooting play an important role in the dynamics of forests prone to windthrow. The pit and mound micro-topography results in a diverse range of microhabitats on the forest floor (Deal et al. 1991, Ott 1997). Uprooting also leads to mixing of soil strata. Sitka spruce in particular may benefit from exposed mineral soil and the absence of the otherwise abundant hemlock advance regeneration (Deal et al. 1991, Ott 1997).

Figure 7: Stem snap


Page 13

In the cool, wet climate of the outer coast soil mixing is of special concern, since it can maintain soil productivity by reversing the built-up of a thick layer of dense organic soil, which eventually leads to waterlogging and bog development (paludification) (Ugolini and Mann 1979, Banner and Rouse 1983, Bormann and Kramer 1998). However, soil disturbance is only one of several factors influencing build-up of organic soil and bog formation. Climate and topo-edaphic (site) factors also play a large, ultimately perhaps more significant role (e.g., Ugolini and Mann 1979, Banner and Rouse 1983, Alaback 1991, Emili et al. 1998). Bormann and Kramer (1998) estimate that soil mixing would have to occur at intervals of 200 years or less to maintain soil productivity. Given the low incidence of uprooting as a mode of mortality it thus seems unlikely that windthrow has historically played a key role in maintaining soil productivity over large portions of the landscape.

2.1.3 Factors influencing susceptibility, severity of effects, and post-disturbance succession

One major factor affecting blowdown frequency and severity is exposure to prevailing winds, which is largely controlled by local and regional topography. Valley bottoms in valleys open to the south, and east to west facing slopes near ridge tops are most exposed, receive the greatest amounts of precipitation (Jakob 2000), and also tend to experience the highest rates of catastrophic blowdown (Harris 1989, Nowacki and Kramer 1998, Kramer et al. 2001). Local patterns such as channelling of winds, turbulence on leeward slopes near ridge tops, and increases in wind velocity in constricted areas also have considerable influence on blowdown patterns (Foster 1988, Harris 1989, Nowacki and Kramer 1998, Kramer et al. 2001).

Other site conditions that affect susceptibility to storm damage include growing conditions and soil properties. Rooting depth is restricted in shallow soils over bedrock or in riparian areas, which makes the trees on these sites more susceptible to blowdown (e.g., Harris 1989, Pearson 2000). Stands established after blowdown may be more susceptible to further blowdown because trees established on upturned soil and roots tend to be stilt-rooted and hence weaker (Harris 1989).

Biotic factors that affect susceptibility to wind disturbance and its severity include tree species, diameter and height, strength and elasticity of the bole, rooting strength, and crown shape, all of which influence the amount of force wind exerts on a tree and the ability of the tree to withstand this force (e.g., Everham and Brokaw 1996). Conifers are generally less windfirm than hardwoods (Everham and Brokaw 1996). Among the conifer species of coastal British Columbia, cedar is the most windfirm (Harris 1989, Keenan 1993). Compared to other species, mortality in cedar is more often by uprooting, rather than stem snap (Pearson 2000, see also Summary Table 1 in Appendix 1). Tall trees with top-heavy canopies provide most leverage to wind (Stathers et al. 1994), and trees that extend above the canopy experience more wind drag and turbulence (Raupach 1989). However, trees exposed to wind can adapt their growth pattern to increase windfirmness against winds from common directions (Harris 1989). Thus, a fringe of trees along exposed sea and lakeshores may withstand a southeasterly gale while trees further inland are blown down (Foster 1988, Harris 1989).


Page 14

Stand characteristics and landscape context also affect susceptibility to wind. Bog woodlands and low productivity cedar stands are little affected by wind because of their openness and small, tapered tree canopies (Harris 1989, Keenan 1993, Kramer 1997, Pearson 2003). Productive stands in valley bottom areas are among the most affected by wind disturbance, likely because of their large, top-heavy canopies and because the high water table in riparian forests restricts rooting depth (Harris 1989). Openings and unevenness in the forest canopy create turbulence and allow for deeper wind penetration. Thus, dense, even-aged stands tend to be relatively windfirm, but become susceptible to blowdown once their canopy has opened up (Gardiner et al. 1997), particularly in the first five years following the opening of the canopy (Mitchell et al. 2001).

2.1.4 Direct and indirect effects on habitat Wind disturbance fulfills an important role in creating and maintaining habitat for a variety of species. Foliage and lichens blown down during storms provide an important winter food source for deer (Harris 1989). Increased production of shrubs and herbs in canopy openings also increases browsing opportunities for a range of mammals, although fallen stems and branches may impede movement in some cases (Harris 1989, Alaback 1991). The combination of frequent landslides and blowdown results in input of both sediments and large woody debris into streams, two attributes that, supplied in moderate amounts, are critical for maintaining stream productivity and salmon habitat (e.g., Bryant and Everest 1998, see also Section 2.4). Woody debris also provides food and shelter for many terrestrial species (see Caza 1993 and Lofroth 1998 for reviews). Root wads and hollows under large fallen logs are used as denning sites by bears and other mammals (Bormann and Kramer 1998).

2.1.5 Interactions with other disturbance agents, forest management, and climate change

The synergistic interaction between wind and diseases and pathogens is a key factor driving gap dynamics in coastal forests (see also Section 2.7). Pathogens such as heart and butt rots gradually reduce the structural integrity of infected trees and make them more susceptible to wind damage (Hennon 1995). Pathogens are also responsible for the predominance of stem snap as the primary mode of mortality. Since rots develop relatively slowly, the predominance of stem snap over uprooting increases with stand age (Nowacki and Kramer 1998). Also, since cedar is less susceptible to pathogens than other species, predominance of stem snap over uprooting is less pronounced for cedar (Nowacki and Kramer 1998, Pearson, 2000). Wounds created by falling trees during a windstorm provide entry points for fungal infection, and wind damage thus in turn facilitates the spread of heart rots (Hennon 1995, Hennon and DeMars 1997).

Synergetic effects between rainstorms, windthrow, and debris slides and avalanches are also well documented (Harris 1989, Schwab 1998a, Jakob 2000). Windthrow and the swaying of trees in the wind create soil disturbance that may trigger soil movements. Conversely soil movements reduce anchoring strength and thus may trigger windthrow. Wind-exposed cutblock edges on or near slide-prone terrain are particularly at risk from the combined effect of windthrow and soil mass movements (Swanston 1967).


Page 15

Forest management can both increase and decrease the severity of wind disturbance. Since young stands are less affected by decay and are generally less susceptible to blowdown than older, taller stands with uneven canopies, clearcut harvesting with short rotation periods to some degree decrease the likelihood of blowdown. However, harvesting also exposes trees to wind stress, without giving them time to develop wind resistance. Thus, openings in the canopy typically lead to increased blowdown at the edges of the remaining stands (e.g., Moore 1977, Harris 1989, Mitchell 1998, Mitchell et al. 2001). Mitchell (1998) predicts that increased cut rates and the implementation of the relatively small cutblocks mandated under the Forest Practices Code will lead to and increase in exposed edges, resulting in increased losses to blowdown. Lack of wind resistance of retained trees and patches may also be a serious concern in variable retention systems (Scott and Beasley 2002). Finally, blowdown after harvesting is especially a problem in narrow riparian buffers, since buffers that experience a large degree of blowdown may not be able to fulfil their ecological function of moderating microclimate and ensuring a sustained supply of large woody debris (see review in Price and McLennan 2001). Moreover, blowdown in buffers may result in excessive input of woody debris into streams in a short time period, which in turn may jam creeks and increase the risk of debris torrents (Grizzel and Wolff 1998, see also Section 2.4).

Although there are no detailed studies on the effect of future climate change on ecosystem dynamics in coastal forests, some predictions can be made based on our understanding of past and current dynamics and the predictions of climate change models. Some of these models predict increases in the frequency and intensity of storms (Everham and Brokaw 1996, Taylor and Taylor 1997). Predicted increases in precipitation during the fall and winter storm season may also increase the severity of wind damage and possibly lead to relatively more uprooting, soil mixing and increased site productivity (Taylor and Taylor 1997). Higher average temperatures and milder winters may result in increased growth rates for butt and heart rots and dwarf mistletoe, potentially causing more and earlier stem snap. In summary, global climate change will likely lead to increases in the frequency and severity of blowdown.

2.2 Snow, Ice, and Hail

Similar to wind, heavy snowfall, ice accumulation, or severe hail can result in uprooting, wounding or breakage of stems and branches, and loss of foliage (British Columbia Ministry of Forests 1997). Together with wind, snow, ice, and hail thus contribute towards driving gap dynamics in coastal forests. As in the case of blowdown, infection with fungi or mistletoe makes trees more susceptible to damage, and wounding may in turn promote fungal infection.

2.3 Geomorphic Disturbances

Because of the steep terrain, wet climate, often unstable soils, and seismic activity in the area, geomorphic disturbances play a significant role in coastal watersheds (Church 1994). Geomorphic disturbances derive their name from the fact that they not only kill or damage


Page 16

vegetation, but also shape the physical landscape by modifying slopes, disturbing soils, and delivering sediment and woody debris considerable distances to valley floors, thereby altering stream channels and floodplains (Naiman et al. 2000). Below we describe the dynamics and effects of a variety of geomorphic disturbances occurring on the BC coast.

2.3.1 Earthquakes The coast of BC is adjacent to a series of fault systems stretching from the US Pacific Northwest along the west side of Haida Gwaii and through southeastern Alaska. Three of the ten largest earthquakes in the world in this century originated in the northern part of this fault system (Hansen and Combellick 1998). Between 1870 and 1988 coastal BC has experienced eleven significant earthquakes, ranging in magnitude between 6.0 and 8.1 on the Richter scale (Brun and Etkin 1997). Combellick and Motyka (1995) estimate that earthquakes measuring 7.9 on the Richter scale can reoccur in Alaska every 120 years, whereas earthquakes of magnitude 6 would be expected every five years. Hyndman et. al. (2001) arrive at similar frequency estimates for subduction quakes along the Queen Charlotte fault and Cascadia subduction zone (see Figure 8).v

Earthquakes can trigger tsunamis, avalanches, and landslides, which affect forest structure and productivity, and precipitate channel change and large inputs of sediment and woody debris into streams (Clague 1996, Allen et al. 1999). Another potential source of forest disturbance by earthquakes is liquefaction, which primarily affects unconsolidated sediments in estuaries, deltas, and floodplains (Clague 1996).

2.3.2 Tsunamis

Tsunamis are rapidly-moving wave fronts most commonly triggered by earthquakes or landslides (Clague 1996). If large enough, they can destroy buildings, uproot trees, and bury vegetation under a layer of marine silt carried inland by the water (Clague 1996, Clague et al. 2000). Tsunamis have also caused large numbers of fatalities in the past. Low-lying coastal sections are most likely to be affected, as are deep inlets surrounded by steep slopes, which channel wave energy and where waves may oscillate for days (Clague 1996). Computer models indicate that within the CIT area, Graham Island, the West Coast of Vancouver Island, and the central mainland coast are especially at risk (Clague 1996).

v The Cascadia subduction zone, lying offshore of the southern tip of Vancouver Island and extending from there to the south,

has been inactive during recent time. It primarily seems to produce rare, very severe quakes (e.g., Hyndman et. al. 2001).

Figure 8: Frequency distribution for subduction eathquakes along the Queen Charlotte fault and Cascadia subduction zone. M is the magnitude of the quake on the Richter scale (Source: Hyndman et. al. 2001)


Page 17

Clague et. al. (2000) estimate that large tsunamis triggered by major (subduction) earthquakes strike the British Columbia coast on average once every several hundred years, though local events, triggered by landslides or crustal earthquakes, may be more frequent. Between 1737 and 1996, 52 Tsunamis were recorded along the Pacific Coast of North America (NOAA/NWS West Coast/Alaska Tsunami Warning Center 2002). Of these, 12 had a runup (maximum wave elevation at shore) of over 10 m, 5 had a runup of over 50 m, and 3 had a runup of over 100 m. The largest tsunami in terms of regional significance recorded during the 1737 - 1996 period was caused by a large subduction earthquake at the Aleutian-Alaska megathrust zone in 1964. It produced a wave height of 67 m in Valdez Inlet and 4.3 m on the California coast, and caused damage and fatalities all along the west coast of North America (Sokolowski 2000). A tsunami of probably even greater magnitude was likely triggered by a subduction earthquake in the Cascadia subduction zone in 1700. It was felt as far away as Japan, where waves up to 5 m high caused significant damage (Clague et al. 2000).

While tsunamis triggered by large subduction earthquakes typically cause the largest impacts in terms of regional extend, the local impact of smaller events can also be considerable. The tsunami with the largest recorded runup was a local event triggered by a landslide in 1958 in Lituya Bay, Alaska (Geertsema 2000). The waves stripped trees to an elevation of 525 m on the opposite shore. Locally significant tsunamis triggered by (non-seismic) landslides also occurred in or near the CIT area, in Kitimat Arm (Clague 1996) and in Troitsa Lake (Schwab 1999). The Troitsa lake tsunami resulted from an underwater landslide off a steep slope of a fan delta and carried a mat of uprooted trees 150 m inland (Schwab 1999). Fan deltas like the one near Troitsa lake, where deposited sediment is perched on steep slopes underwater, are found along the side walls of coastal fjords and steep-sided fjord-like lakes throughout coastal BC. These fan deltas are very productive but highly unstable, with past landslide evidence often concealed under water.

2.3.3 Volcanic eruptions A series of volcanoes extending along the Pacific Coast from Northern California to Alaska bear evidence of past volcanic activity in the area (Church 1994). The closest currently active volcanoes, Mt. St. Helens and the Aleutian and Wrangell volcanic centres, are several hundred kilometres away to the south and north of the British Columbia border respectively. However, eruptions may still cause ash fallout on the BC coast (Combellick and Motyka 1995). An eruption 1,500 years ago in the Wrangell-St. Elias Mountain deposited ash over 300,000 km2 in eastern Alaska and southern Yukon, and similarly large ash fallouts occurred after the eruptions of Mt. St. Helens and Mazama south of the border (Brun and Etkin 1997). Depending on the chemical composition and thickness of the ash layer, ash fallout can inhibit tree growth by blocking photosynthesis and affect soil productivity, often leading to bog or den formation (Veblen and Alaback 1996).

2.3.4 Snow avalanches Snow avalanches are high-severity disturbances that occur in mountainous terrain with sufficient snowfall, such as the Coast Mountains. Avalanches are triggered by weaknesses in the snowpack, which may be caused by rain on snow events, by windloading, or by rapid warming of the surface layer in a deep snowpack as might occur on southerly aspects (Septer


Page 18

and Schwab 1995). Since studies of avalanche dynamics for the Coastal Cordillera are scarce, the summary below is based primarily on research from the Interior Cordillera.

Most avalanches occur within well-defined paths at relatively short intervals, sometimes repeatedly within the same year. The size of avalanche tracks in the Invermere Forest District ranged from 2 ha to 885 ha (Ferguson and Pope 2001), in Clayoquot Sound from 2.4 to 34 ha (Pearson 2000, see also Summary Table 4 in Appendix 1). The locations at which avalanche paths develop within a watershed are determined by local snow climate, topography, as well as properties of the underlying material: avalanches tend to initiate on moderate (25-35°) slopesvi, on southerly or easterly aspects, on substrate that erodes relatively easily and thus allows formation of gullies conducive to subsequent avalanching (Butler and Walsh 1990). The frequency and severity of avalanches within a given path depends on the regional climate (likelihood of heavy snowfall, wind, temperature, etc.), and the terrain (e.g., slope of the path, and starting zone and runout elevation). In Kananaskis Valley, Alberta, return interval within a path was less than 20 years for the upper, steeper part of the path, ranging to a maximum of 130 years for the gentler slopes near the runout zone (Johnson 1987). Since trees become increasingly susceptible to breakage with increasing stem height and diameter, frequent avalanching prevents development of mature forest and maintains a characteristic mosaic of upland and wetland herb communities, deciduous/shrub, coniferous/shrub and subalpine meadow communities along the core area of the path (Walsh et al. 1994, Ferguson and Pope 2001). These herb and shrub communities provide key foraging habitat in close vicinity to shelter. The runout zone in particular is used by grizzly and black bears in the spring and early summer (Ferguson and Pope 2001).

2.3.5 Mass wasting Mass wasting includes a wide range of geomorphic processes which transport substantial quantities of rock or soil downhill. Mass wasting processes occur at different rates and intensities, depending upon the surface material, slope shape and inclination, vegetation and soil saturation (Chatwin et al. 1994). Factors that make an area conducive to mass wasting include steep and potentially unstable slopes, shallow soils, and high annual rainfall. On Haida Gwaii approximately a third of the area is estimated to be susceptible to mass wasting, and similar figures likely apply for the mainland coast (Schwab 1998b).

Creeps, slumps, and earthflows Creeps are slow downslope movements of surficial material at rates of centimeters per year to millimeters per year (Chatwin et al. 1994). Slumps and earthflows involve the movement of a block of material either slowly or rapidly (meters per second). Bedrock slumps typically occur below the surface of the earth as a relatively slow displacement of bedrock. They are common on steep slopes underlain by fractured bedrock. Creeps often precede slumps and earthflows. Excess rainfall can cause weaknesses particularly in clay material, resulting in creeps, slumps or earthflows on slopes as gentle as 2-3 degrees.

vi Steep slopes generally do not accumulate enough snow to trigger avalanches.


Page 19

Rock slides, falls, and avalanches Rock slides, falls and avalanches result from bedrock failures, typically originating on steep rock faces. Whereas rock falls tend to be small, rock avalanches can be very large and travel a long distance, leaving fields of unstable rocky debris often far beyond the foot of the failed slope (Clague 1996). Rock slides are broken masses of rock moving downhill by sliding on underlying discontinuities in the parent material. Under the right conditions, they can become rock avalanches and deposit material over a large area (Clague 1996). These debris fields are sometimes covered in vigorous forests (Clague 1996, Schwab 1998b).

Debris slides, flows, torrents, and avalanches Debris slides are relatively rapid, shallow landslides, typically originating on steep (35º - 60º) slopes (Combellick and Long 1983, Weir, 2000, Chatwin et al. 1994). They are visible throughout the landscape as relatively narrow, linear tracks, often covered with early-seral vegetation. If there is water present, debris slides can become debris flows moving at rates of metres per minute to metres per second. Debris torrentsvii occur when a mixture of water, earth and vegetation debris rapidly moves through a steep, well-defined stream channel. Damming of this channel by woody debris may occur. During major storms these dams may fail moving large volumes of water, wood, soil and rock extremely rapidly. When channels drop below eight degrees, such as when meeting tributaries, significant amounts of sediment, organic material and coarse woody debris are deposited (Chatwin et al. 1994). Sediment deposition from debris flows and torrents is thought to move as a gradual wave through a drainage network (Naiman et al. 2000).

Debris slides and flows typically occur when soils are saturated with water after heavy or prolonged rainfall and/or rapid snowmelt (Schwab 1998b). For Haida Gwaii, Hogan and Schwab (1991) estimate that precipitation exceeding 22 mm within a 24-hour period is needed to trigger slope failures if soils were previously wet, more (up to 29 mm / 24-hour period) if soils were previously dry. Rainstorms heavy enough to meet this condition occur on Haida Gwaii with a frequency of 2 to 10 years. Although many slides on Haida Gwaii are triggered by such small, frequently occurring weather systems (Hogan and Schwab 1991), major slide events in natural terrain appear to be episodic (Septer and Schwab 1995, Schwab 1998b, Schwab 1998a) and often associated with unusually severe rainstorms, which are approximately as frequent as severe windstorms that have caused major blowdown events.

Estimates of landslideviii frequency in the CIT area range from 0.06 slides per year per km2 to 1.1 slide per year per km2 (see Summary Table 2 in Appendix 1). Note, however, that these estimates of slide frequency are derived from study areas that include both logged and

vii The terminology used to distinguish between different types of debris movement is not always consistent. In particular, the

term debris torrent is also often used synonymously with the term debris flow. viii Studies often do not differentiate between different types of mass wasting. The term 'landslide' is commonly used to

encompass different types of rock and/or soil movement that leave visible scars in the landscape.


Page 20

unlogged terrain. Since logging and road building can substantially increase landslide frequency (see below), these numbers likely do not reflect historical (pre-settlement) landslide frequency in the area.

Similar to size distributions for other natural disturbances such as wind and fire, the size distribution of landslides is reverse J-shaped, i.e., most slides are small, both in terms of area (< 1ha, Jakob 2000) and in terms of debris volume transported (< 1000 m3, Rood 1990). However, as with other disturbance types, the relatively large rare slides account for most of the total area affected (Smith et al. 1986).

Slope inclination seems to be the primary factor governing slide frequencymost slides originate on steep slopes (e.g., Smith et al. 1986, Gimbarzevsky 1988, Rood 1990, Millard 1999, Rollerson et al. 2001, Rollerson et al. 2002, see also Summary Table 2 in Appendix 1). Local and regional precipitation and runoff patterns can also have a strong influence on slide occurrence. Slides predominantly occur on concave or complex slopes and in gullies that channel surface drainage (Gimbarzevsky 1988, Rood 1990, Millard 1999, Rollerson et al. 2001, Rollerson et al. 2002). On a larger scale, slide frequency is governed by precipitation patterns and generally decreases from the outer to the inner coast (Pojar et al. 1999, Jakob 2000, Rollerson et al. 2001, Trainor 2001). Some studies have also found higher slide frequencies on windward slopes (typically east to south facing aspects) most exposed to incoming pacific weather systems (Gimbarzevsky 1988, Jakob 2000). This last relationship seems to be stronger in some areas than in others, however.

Landslides generate more soil disturbance than snow avalanches, with the top two-thirds of the slide typically scoured to bedrock and the soil deposited in the lower third (Smith et al. 1984, Smith et al. 1986). Mineral soil and bedrock exposure decreases as slides age – the lower portions of slides are vegetated more quickly than the upper portions (Smith et al. 1986, Veblen and Alaback 1996). As discussed in Section 2.1.2, the exposure and/or deposition of mineral soil can counteract the built-up of organic soil that has been associated with reduced forest productivity (Ugolini and Mann 1979, Banner et al. 1983, DeGroot et al. 2002). By depositing mineral soil, silt, or debris, mass wasting and other geomorphic disturbances can therefore contribute towards maintaining long-term forest productivity in the Figure 9: Alder saplings on a fresh debris fan


Page 21

affected areas.

Similar to succession after snow avalanches, succession after landslides is dominated by early-seral communities of deciduous shrubs, red alder, Sitka spruce, western hemlock and redcedar (Smith et al. 1984, Smith et al. 1986). On Haida Gwaii, red alder was generally the most abundant tree species, especially on the lower part of slide tracks, at low elevations and on fine-textured soils (Smith et al. 1986). Conifer regeneration in the lower part of the slides was suppressed by the abundant, but short-lived red alder for the first 70 - 90 years after the slide, but conifers dominated in the middle and upper parts of slide tracks, especially at high elevations and on coarse-textured, acidic soils. Viewed in a wider landscape context, snow avalanche and landslide tracks fragment the lateral connectivity of upland forest and create early-seral corridors between high elevations and valley floors. However, the terrain or type of slide material may inhibit use by wildlife.

Debris from landslides, especially debris flows, can be key for triggering forest succession in riparian areas (Naiman et al. 2000). Jams of large woody debris in particular, can alter local flow patterns and provide refugia against flood scour which allow vegetation to develop downstream (Naiman et al. 2000). Woody debris is important for salmon habitat and riparian function. However, at extreme levels debris can also greatly modify riparian channels and destroy habitat (e.g., Bryant and Everest 1998, see also discussion in Section 2.4).


Geomorphic disturbances often closely interact with each other, as well as with other forms of disturbance. Most of these interactions are synergistic. For example, snow avalanches and mass wasting can increase susceptibility to blowdown by opening up the forest canopy (Mitchell 1995) and increase susceptibility to pathogens by wounding trees. If snow avalanches have been particularly erosive, they can increase the likelihood of debris flows. Damming of streams or lakes by the runout of mass wasting events can lead to flooding, which in turn can undercut slopes and reduce slope stability (Chatwin et al. 1994). In general, geomorphic disturbances act in concert with flooding to shape landforms along hillslopes and valley floors–the relative timing and size of both disturbance types affects the degree to which channels and floodplains are modified (Naiman et al. 2000).

There is ample evidence that earthquakes can cause various types of terrestrial and subaqueous landslides. For example, an earthquake on Vancouver Island in 1946 measuring 7.2 on the Mercalli scaleix triggered over 300 slides within an area of 20,000 km2 (Mathews 1979), and there are many other documented examples of seismically-triggered landslides from Vancouver Island (Mathews 1979), Puget Sound (Jacoby et al. 1992; Schuster et al. 1992), and other parts of the world (Keefer 1984). In general, the bigger the quake, the bigger and more widespread the slide events associated with it (Keefer 1984). However, the limited accuracy of methods available for dating historical slides make it difficult to assess the contribution of seismic activity to overall slide frequency in coastal British Columbia

ix The Mercalli is based on the expected damage resulting from the quake. Average return time for M 6-7 quakes in

southwestern BC is 15 years (Clague 1996).


Page 22

(Clague 1996). It seems that rainstorms are most likely to trigger smaller, shallow slope failures such as debris slides and flows, whereas earthquakes may be more important as a trigger for rock falls and rock avalanches, and large, deep-seated slides (though likely not slow-moving failures such as bedrock slumps and earthflows; Clague 1996, Schwab 1998b).

Tree cover generally reduces the likelihood of geomorphic disturbances. Trees intercept snow and rain in their canopies and moderate temperature fluctuations of the snow surface, thereby reducing the likelihood of weak layers in the snowpack (McClung 2001). They also anchor soil and decrease the point at which soils become saturated. Therefore, any disturbance that removes tree cover, anthropogenic or natural, can increase the risk of mass wasting, as well as snow avalanches. Jakob (2000) estimates that it takes approximately 20 years after tree establishment for soils to stabilize.

Logging and road building in particular have substantially increased the overall frequency of slope failures throughout coastal BC. Although clearcuts appear to be responsible for the majority of harvesting-related mass wasting in coastal areas, roads affect drainage patterns and thus also increase the risk of debris slides (Jakob 2000). Slides originating from clearcuts can occur on flatter slopes, tend to be smaller in volume, and travel less distance than those originating in natural terrain (Swanston et al. 1996, Jakob 2000). In Clayoquot Sound the frequency of landslides in logged terrain was approximately nine times higher than the natural rate (Jakob 2000). Similar or even larger (up to 30 times) increases in landslide frequency in response to clearcut harvesting and road building have been recorded for other areas of Vancouver Island, Haida Gwaii, and parts of the North Coast Forest District (Schwab 1998b, Jordan 2001). According to Schwab (1998b), logged terrain on the coast has yet to experience rainstorms severe enough to trigger the serious natural debris slides that have occurred in the past (see also Septer and Schwab 1995). The same can be said for big earthquakes (Clague 1996). The potential effects of such severe events on harvested areas are unknown at this point.

Climate change is also likely to result in increased incidence and/or severity of several types of geomorphic disturbance. Landslide frequency will likely increase as winter precipitation rises, rainstorms become more frequent (Evans and Clague 1997), and retreating glaciers expose unstable slopes of glacial till. Similarly, increased precipitation in winter likely means more snow accumulation and more snow avalanches at higher elevations, though warmer winter temperatures may result in less snow accumulation and hence fewer snow avalanches at lower elevations. Tsunamis triggered by slide events may also become more frequent, and the impact of tsunamis may be felt further inland as sea levels rise.

2.4 Flooding

2.4.1 Disturbance dynamics Flooding is an important disturbance agent in flat valley bottoms, estuaries, and low-lying coastal areas. Non-marine flooding occurs when runoff exceeds the capacity of drainage channels, as a result of intense and/or prolonged rainfall, ice jams, rapid snowmelt, or a


Page 23

combination of these factors. Flooding events between 1891 and 1991 for many rivers and creeks in the area are detailed in Septer (1995). Four types of events occur:

1. Rainstorm triggered floods. Rainstorms severe enough to cause flooding may occur throughout the year, although the most intense storms occur during the fall (Harris 1989). Melting of shallow snowpacks due to continuous fall rains or rapid melting of snow in late spring may also cause substantial flooding. Floods triggered by rainstorms have caused extensive debris flow and torrent activity throughout the BC coast.

2. Icejam floods occur from November to April during the freeze-up or break-up of ice. These have occurred on the Skeena River.

3. Glacial outbursts occur when there is a sudden release of water stored behind a glacial ice dam, particularly in the summer (Combellick and Long 1983). These are relatively rare flooding events (at least 27 in northwestern BC since 1852) but are usually of high-severity because the amount of water discharged is often much more than the expected runoff after rainstorms for most watersheds (Septer and Schwab 1995).

4. Marine flooding. Tidal flooding of estuaries occurs regularly at high tides. More extreme marine flooding of low-lying coastal sections can occur due to a combination of winter storms and high tides, or tsunamis (see Section 2.3.2). Tidal flooding of estuaries mixes marine water and sediment with freshwater and river sediment to create a productive mosaic of unique forest wetlands, shrub thickets, sedge and grassland ecosystems, salt, brackish, and freshwater marshes, and mudflats (Price and McLennan 2001).

Because flooding is so strongly associated with climatic events, flooding can occur region-wide –e.g., a rainstorm in 1978 caused synchronous events observed from Haida Gwaii across to Terrace (Septer and Schwab 1995). Flooding is also influenced by topography and soil characteristics. The steep slopes with thin soils in the Coast Mountains can transport water rapidly to streams, increasing the likelihood of flooding.

Flooding, through erosion and deposition of sediment and coarse organic material Figure 10: River estuary on Haida Gwaii


Page 24

downstream, support the development of floodplains in wide valleys and fans where steep streams meet the valley floor. Extensive floodplain-fan complexes occur in many watersheds in the Outer Coast Mountains (Price and McLennan 2001). Floods can cause varying degrees of tree mortality, but most events are stand-maintaining unless debris flows or torrents are triggered.

Floodplains and fans subject to flooding are highly productive and support forests of various composition and ages, depending on disturbance frequency and severity (see review in Price and McLennan 2001). Portions of fans that are frequently flooded or experience debris torrents support red and slide alder, deciduous shrub and herb communities. Very large, widely spaced Sitka spruce and western hemlock are common in less active, well-drained areas of floodplains and fans. Portions that are not well-drained may support shrub and sedge wetlands. On the Central Coast, forests comprised on average 15% of the floodplain area, and most of these stands were old (Pearson 2003).

While the flooding regime affects the composition and dynamics of the surrounding forest ecosystems, the forests, both riparian and upland, in turn fulfil an important role in regulating floodplain ecosystem dynamics (Friele 1997, Price and McLennan 2001). Root systems of trees stabilize slopes and regulate the amount of debris reaching the floodplain from mass wasting events, while tree crowns intercept precipitation and return some of the moisture to the air by transpiration. Thus, presence of upland forests limits surface erosion, reduces soils saturation and overall runoff, and regulates peak runoff during rainstorms and snowmelt, which in turn reduce the severity of flood events and limit debris and sediment input into streams. The forests on the floodplain help to stabilize stream banks and unconsolidated sediments. Standing or downed trees in the floodplain can also temporarily trap sediments

Figure 11: Debris jam in a small stream


Page 25

and thus reduce the amount of sediment entering rivers during flooding events.

Downed wood in particular plays a keystone role in riparian and floodplain ecosystems. Trees in floodplains often grow on logs and root mounds, indicating that these structures may be needed for successful regeneration of these forests (Harmon and Franklin 1989). Log jams affect channel morphology and flow regime by interrupting the pool-riffle morphology of smaller streams, trapping sediments, and reducing the kinetic energy of streams, thus slowing the transport of water and bedload materials (Hogan et al. 1994, Abbe and Montgomery 1996). Newly formed log jams can drastically alter the hydrological regime of a stream reach, often precipitating channel changes and largely destroying existing in-stream fish habitat (Hogan et al. 1994, Tripp 1994). However, as log jams age they create instream backeddies and protect newly formed backchannels from scour, thereby providing and maintaining new sources of highly productive habitat for salmon and other aquatic species (Tripp 1994, Abbe and Montgomery 1996). On a smaller scale, single large pieces of woody debris fulfil a role similar to that of logjams, especially in smaller streams (Naiman et al. 2000, Price and McLennan 2001). A relatively continuous supply of large woody debris is thus critical for maintaining diversity and natural variability of riparian and floodplain ecosystems. Either a lack of large woody debris input or input of excessive amounts over short periods of time can substantially alter ecosystem dynamics and endanger local fish populations (Tripp 1994, Bryant and Everest 1998).


Frequency and severity of flooding is often closely interlinked with other disturbances affecting hydroriaprian systems. Debris from avalanches and mass wasting can dam streams and cause flooding (see section 2.3 on geomorphic disturbances above for a description of how flooding interacts with geomorphic disturbances to modify channels and floodplains). Beaver dams can also alter hydrological regimes and lead to flooding (see Section 2.9.2). Climate change in BC is expected to increase spring flooding and increase the frequency of extreme storm events (Beckmann et al. 1997). It is uncertain whether the increase in precipitation predicted for coastal BC can offset the effects of glacial retreat on flow to glacial fed streams. The expected rise in sea level and increase in precipitation is expected to increase coastal flooding and erosion of estuaries (Beckmann et al. 1997).

Harvesting can have substantial impacts on watershed hydrology, both through the trees it removes and through the slash it leaves behind. Thus, removal of trees reduces canopy interception, which can increase the incidence of flooding relative to unlogged landscapes. Clearcuts and roads without proper drainage increase the risk of slope failures, and thus also the risk of flooding (see Section 2.3.6). Last, but not least, logging can substantially affect input of sediments and large woody debris into streams. In the short run, large volumes of slash and debris often enter streams as a direct or indirect result of logging operations. This increases the risk of debris torrents and can substantially alter channel morphology, with potentially serious impacts on aquatic habitat (Hogan et al. 1994, Tripp 1994)x. In the long x Even though the Forest Practices Code contains provisions for stream protection designed to alleviate these problems, it

seems that the effectiveness of these measures has been limited in practice (see Sierra Legal Defense Fund 1997b).


Page 26

run, harvesting of productive valley bottom forests will almost certainly lead to a scarcity of large woody debris in streams, since it removes future sources of recruitment (e.g., Bryant and Everest 1998).

Ensuring that recruitment of large woody debris will remain high enough to sustain current levels might require substantial limitations on harvesting in valley bottom areas. Pollock (1997) suggests that an annual flux rate of 15% should be maintained to ensure continuation of current debris loadings in the Kawesas watershed (a part of the Greater Kitlope Ecosystem). For the Kawesas, this implies exempting almost all the old growth forest on the valley floor from harvesting, a conclusion in line with the observation that channel changes have essentially consumed the entire floodplain forest over the last several hundred years. Pearson (2003) argues along similar lines: since the natural meander of the river often encompasses the entire valley floor area, protection of the entire valley bottom may be indicated if natural floodplain dynamics are to be maintained.

2.5 Fire

2.5.1 Disturbance dynamics Because of the wet climate, the role of fire as a disturbance agent on the BC coast is limited (Banner et al. 1993, British Columbia Ministry of Forests 1995a, Pojar et al. 1999). Except for the drier, more continental inner Coast Mountain areas, evidence of fire is rare throughout the current coastal landscape (British Columbia of Forests 1999, Pearson 2003)modern (post-1950) fire return intervals in the ecoregions of the CIT area range from 2,000 to well over 10,000 years (Table 1). Like the distribution of opening sizes for most other disturbance agents, the frequency distribution for fire size is reverse-J-shaped: the vast majority of fires are small, although fires can grow to several thousand hectares in size in suitable weather conditions (Pearson 1968, see also Table 1 and Summary Table 5 in Appendix 1).

There is no research specific to the CIT area on pre-European-settlement fire regimes. Charcoal evidence from two lake cores on the west side of southern Vancouver Island suggest fire return intervals of approximately 3000 years since 7,000 BP, although lakes cores from the east side of Vancouver Island indicate a much higher incidence of fire (Brown and Hebda 1998). In coastal western hemlock forests (CWHvm1) in Clayoquot Valley the median time since last fire ranged from 750 to 4500 years ago (Gavin et al. 2003, see also Summary Table 5 in Appendix 1). Fires rarely exceeded 250 m in diameter in this study, but sampling may have been too limited to lower elevations to detect fires that spread to higher elevations. Evidence from other sites in coastal BC and Washington support the conclusion that fires were historically infrequent in coastal western hemlock forests (Huff 1995, Green et al. 1999), although many stands on the South Coast and on the east side of Vancouver Island contain a significant component of Douglas-fir and thus have likely originated after fire (Arsenault 1995, Daniels et al. 1995, Green et al. 1999). Fire regimes in coastal western hemlock forests are thought to be of mixed severity: a combination of infrequent, high-severity crown and low-severity surface fires causing partial or complete stand-replacement (Parminter 1990).


Page 27

Fires historically also occurred rarely in the subalpine Mountain Hemlock zone. Hallet et. al. (2003) found return intervals of 300 to 7000 years, with time since the most recent fire ranging between 50 and 9200 years for subalpine sites in the Coastal Mountains (see also Summary Table 5 in Appendix 1).

In the drier, eastern portion of the Central Coast plan area, fires were likely historically more frequent than in coastal zones at similar elevations (Wong et al. in prep.). No fire history studies have been conducted in this area, and while research on historical fire regimes has occurred in the same biogeoclimatic zones further in the interior of the province, it is difficult to extrapolate results because of the proximity of the transition to coastal climate in the study area. Analyses in similar dry ecosystems near the coastal transition in the Fraser TSA have found return intervals for stand-replacing disturbances somewhere between 200 and 450 years for the CWHds1 and CWHxm1, and somewhere between 200 and 420 years for the ESSFmw (Dorner et. al. 2003). The fire regimes in these drier ecosystems were likely mixed-severity – a combination of low-severity fires (fires which kill many small saplings but few large trees), moderate-severity fires (fires which cause patchy mortality) and high-severity fires (fires which kill many large trees; Wong et al. in prep.). Resulting stand structure may range from uneven-aged to even-aged.

Number of fires

Area burned (ha)

Number of fires / million

ha / yr Ha burned /

million ha / yr Average fire

size (ha)

Ecoregion Total area (ha)

lightn

ing

huma

n

lightn

ing

huma

n

lightn

ing

huma

n

lightn

ing

huma

n

lightn

ing

huma

n

Maximum fire size

(ha)

Estimated return

interval (yrs)

Queen Charlotte Ranges 722,954 5 123 1 249 0.2 4.0 0 8 0 2 99 124,348 Queen Charlotte Lowland 327,621 0 56 0 202 0.0 4.0 0 14 N/A 4 69 69,706 Northern Coastal Mountains 327,133 0 2 0 0 0.0 0.1 0 0 0 0 0 35,166,818 Northern Coastal Mountains 2,431,685 63 58 7,830 552 0.6 0.6 75 5 124 10 6,086 12,475 Coastal Gap 5,209,042 302 704 5,816 6,542 1.3 3.1 26 29 19 9 3,569 18,125 Pacific Ranges 6,182,537 1,854 2,918 70,368 56,909 7.0 11.0 265 214 38 20 8,583 2,089 Western Vancouver Island 2,043,739 404 1,015 3,262 13,929 4.6 11.5 37 158 8 14 2,226 5,112

Table 1: Fire incidence by ecoregion for the period 1950 - 1992 (source: Steve Taylor, unpublished data).

2.5.2 Traditional burning by First Nations Intentional low-severity burning by aboriginals has been documented for the Haisla, Nisga’a, Tsimshian and the Haida peoples on Haida Gwaii before concentrated European settlement (John Corsiglia pers. com., Turner 1991, McDonald, in prep.). The Haisla enhanced berry production of various Vaccinium species (e.g., Alaskan blueberry, red huckleberry) and the Haida also burned to stimulate salal (Gautheria shallon) (Turner 1991, Johnson Gottesfeld 1994). The interior Gitksan and Wet’suwet’en peoples in the Skeena and Bulkley Valleys near Hazelton burned lowbush blueberry in valley bottoms and black huckleberry patches “half way up the mountain” every four years in the fall (Johnson Gottesfeld 1994). Some place names in Tsimshian territory such as “burnt mountain top” suggest similar burning practices (McDonald in prep.). The Nisga’a are also thought to have burned every four years


Page 28

for berry production as recorded in the Oolichan Petition (John Corsiglia pers. com). Further evidence of aboriginal burning is suggested by 100-200 year old western redcedar stands, some as large as 30-40 ha, observed along the central and north coast. These stands contain culturally modified trees and burned out cedar stumps and are believed to have originated after fires escaped from attempts to burn out cedars for canoes or from clearing for village sites (A. Banner pers. com, B. Fuhr pers. comm.).xi A study is under way to document First Nation resource management practices in the North Coast (contact: Charles Menzies, UBC).

As First Nations have lived on the BC coast for thousands of years, intentional (and possibly also unintentional) setting of fires by aboriginal peoples likely played a key role in driving fire regimes in certain areas. Like in other regions in British Columbia, aboriginal burning on the coast probably decreased as populations declined with the small pox epidemic in 1862 and ceased with Forest Service policies in the 1930s (Turner 1991, Johnson Gottesfeld 1994, Parminter 1995).

2.5.3 Disturbance effects and post-disturbance succession Fire affects post-disturbance succession by reducing or eliminating advanced western hemlock regeneration, making it easier for spruce, amabilis fir, and western redcedar to regenerate (Veblen and Alaback 1996). Very successful regeneration of western redcedar has been observed in stands established after fire on the North Coast (A. Banner pers. comm.). These stands are thought to have regenerated after aboriginal burning and while uncommon, they illustrate the role that fire can play in regeneration in coastal forests. Lodgepole pine, as well as the relatively shade-intolerant Douglas-fir and aspen are common regenerating species after fires in drier inland areas (Green and Klinka 1994, Pojar et al. 1999). Lodgepole pine and Douglas-fir are often found there together on dry, recently burned sites (Pojar et al. 1999). Fire can be especially important for lodgepole pine regeneration, since it opens the serotinous cones of this species (Muir and Lotan 1985).

In subalpine forests, post-fire communities are often Vaccinium meadows. Tree regeneration in these systems appears to be largely unpredictable and only partially correlated with time-since-disturbance, as it is also substantially affected by browsing, snow creepxii and the duration of the snowpack (Agee and Smith 1984).

xi As these stands are smaller than the resolution of forest cover maps, more detailed surveys are required to document the

distribution of such stands. xii Slow movement of snow downslope.

Figure 12: Burned-out cedar near the Haida village of Hlk'yah (Windy Bay) on Lyell Island


Page 29

Like most other disturbance agents, fire provides important habitat elements for many species, including snags, large coarse woody debris, and Vaccinium communities, which provide nutritious early seral forage for wildlife species such as bears and deer (Bunnell 1995, Gray and MacKenzie 2001).

2.5.4 Factors influencing susceptibility and severity of effects Topography in the area exerts influence on fuel moisture and fire spread at both regional and local scales. On regional scales, the drier climate in the rainshadow of the mountain ranges is more conducive to fires, a pattern apparent on the mainland coast as well as Vancouver Island and Haida Gwaii (Pearson 1968, Brown and Hebda 1998, Pojar et al. 1999). On local scales, drier southerly aspects are more susceptible than northerly aspects and valley bottom areasin Clayoquot Valley, fires were more than 25 times more likely on south-facing slopes than in less susceptible landscape positions (Gavin et al. 2003a).

Forest structure also influences the susceptibility to fire (e.g., Agee 1993). Thus, Pearson (1968) notes that the low-productivity bog woodlands of the Queen Charlotte Lowland are usually fire resistant, but may become very conducive to burning in unusually dry weather because of their high cedar content, large number of snags, and peaty soils. In Clayoquot Valley, fire intervals were shorter in cedar-salal forests than in other forest types (Gavin 2000). Gavin (2000) suggests that fires might be involved in a feedback mechanism maintaining salal and stunted cedar forests, whereby nitrogen volatization results in poorer productivity, which in turn favours a forest structure containing drier fuels. In contrast, A. Banner (pers. comm.) indicates that on the North Coast, fire in cedar hemlock stands of low productivity may actually improve productivity by reducing the organic layer and improving nutrient availability. However, evidence in both areas points at a potential link between high cedar content and fire incidence (cf. Section 2.5.2). This connection is likely explained by the high flammability of cedar, as compared to other tree species (Pearson 1968), as well as the fact that fire seems to favour western redcedar regeneration (or continued persistence of western redcedar, cf. Section 2.5.3 above).

Finally, fire occurrence is also strongly tied to long-term climatic trends. Area return intervals of fires occurring around Clayoquot Lake increased from 50 years to 350 years between 900 and 1100 AD, coincident with the onset of the Little Ice Age (Gavin 2000, Table 2). Hallet et. al. (2003) report similar fluctuations in response to climatic changes for subalpine forests in the Fraser Valley.


By removing advance regeneration of mistletoe infected western hemlock, fire can reduce the incidence of mistletoe. By stressing trees, fires can also increase susceptibility to other disturbance agents such insects (Bradley and Tueller 2001) and fungal infection (Geiszler et al. 1980). Other disturbance agents such as blowdown can in turn increase fuel loadings and thus may increase susceptibility to fire (Turner et al. 1999). In the subalpine, fallen snags from fires can prevent snowcreep and inhibit browsing (Agee and Smith 1984).


Page 30

Because of cessation of traditional burning by First Nations, different land management practices, and fire suppression, present-day fire patterns likely differ from those that occurred before European settlement. Consequently, statistics derived from recent fire history do not necessarily reflect historical (pre-settlement) dynamics. For example, human ignitions associated with post-settlement lifestyles were responsible for two thirds of the fires in the ecosections listed in Table 1, and many of the biggest fires that occurred in Haida Gwaii and the North Coast since 1950 were associated with logging (Pearson 1968, Dorner and Wong 2002). Pearson (1968) points out that harvesting increases the probability of fire, since logging slash is highly flammable when fresh. Slash, especially cedar slash, can remain a fire hazard until the young stand reaches crown closure, at which point fuels dry out less easily (Pearson 1968). It is difficult to estimate the combined effect of the different post-settlement influences on fire frequency. However, it seems likely that modern fires tend to burn in different locations and possibly also in different seasons from pre-settlement fires.

We expect that climate change will influence fire regimes by affecting fuel moisture during the summer. However, the direction of change is unknown as there is a fair amount of uncertainty in projections regarding summer precipitation levels for coastal BC.

2.6 Short-term Climatic Disturbances

Fluctuating climatic factors such as frosts and periods of severe cold or drought can injure tree canopies and roots systems. New growth in spring is especially sensitive to frost damage, but frost injury may also result after a sudden drop in temperature before tissues have hardened in the fall (British Columbia Ministry of Forests 1997). Frost lesions on the stem provide access for decay fungi (Allen and Walsh 1996). Frost heaving can also kill seedlings, particularly in low-lying depressions, and soil freezing may cause root injury (British Columbia Ministry of Forests 1997).

During periods of drought, dieback of the leader, or top-kill, can occur, particularly in microsites such as gravel bars, which poorly retain water. Drought mortality may also occur in winter, when trees lose water through transpiration during unseasonable warm, drying winds and are unable to replace that water because the ground is frozen (Allen and Walsh 1996). “Red belts” of injury across mountain slopes often indicate the elevational zone of the drying winds. Drought caused injury is rare on the outer coast, but some sites in the inner Coast Mountains are subject to moisture stress in summer (J. Pojar, pers. comm., Pojar et al. 1999).

Climatic events occur across large regions but resulting effects on growth or tree mortality can be patchy, depending on tree species and microsite conditions. If adverse climatic conditions persist, trees may gradually deteriorate, often resulting in death (Williams et al. 2000). The decline of yellow cedar in southeast Alaska is thought to be due to climatic warming beginning in 1880 (Williams et al. 2000). Yellow cedar decline is also occurring on the North Coast of British Columbia, but not to the same extent as in southeast Alaska (A. Woods, pers. comm.). The decline in southeast Alaska is believed to result from a combination of reduced snowpack and extreme freeze-thaw events, resulting in root injury


Page 31

from soil freezing – these events are not as likely at sea level on the BC coast. Climatic warming is also hypothesized to affect western hemlock. If global warming continues as predicted under a doubling of CO2, one model predicts that CWH forests may undergo catastrophic collapse, as trees become highly susceptible to frost events because of an inadequate chilling period to induce winter cold hardiness (Burton and Cumming 1995).

2.7 Pathogens

Pathogens are a key component of forest dynamics in coastal forests (Hennon 1995). In southeast Alaska, heart rots are ecologically most importantHennon (1995) estimates that for most forests in coastal Alaska annual loss to fungal decay is at or near equilibrium with volume of wood produced annually. The parasitic hemlock dwarf mistletoe also plays an important role, and root rots can be a significant disease in younger stands (Hennon 1995, Trummer et al. 2000). Further south, the relative importance of root rots and mistletoe increases (Haak and Byler 1993, Hennon and DeMars 1997, Garbutt 1998).

2.7.1 Fungi The coastal climate is very conducive to the development and spread of fungal infections. The literature generally distinguishes between three types of rots, based on the effect the fungal infection has on the tree. Heart rots affect the bole, root rots infect the root system, and butt rots affect the butt and lower bole of the tree (e.g., British Columbia Ministry of Forests 1995b). There are a large number of root rot and decay fungal species endemic to coastal forests. Some of the most common ones are listed in Table 2. While some of these fungi are constrained to either the root system or the butt or bole of the tree, others can affect both. Thus, some butt rots are caused by root rot fungi spreading into the base of the tree, and Heterobasidium annosum, one of the most important species on the coast, may independently infect either the root system, or the butt or bole.

Although they are sometimes caused by the same fungal species, the symptoms of root rot are quite different from those of heart or butt rot. Heart and butt rots rarely cause direct mortality. Their primary effect is a gradual undermining of structural integrity, which makes the tree more susceptible to breakage during windstorms or under heavy snow loads (Hennon 1995). Typical symptoms of root rots include reduced growth, as well as yellowing and thinning of foliage, sometimes accompanied by a distress cone crop, and, in severe cases, death of the affected tree (Canadian Forest Service 2002). Because they weaken the root system, root rots also make infected trees more susceptible to windthrow. On the coast, root

Figure 13: Fruiting bodies of Ganoderma applanatum


Page 32

rot is primarily a disease of younger stands, whereas heart and butt rots are the primary forms of fungal disease occurring in older stands (Pojar et al. 1999, Canadian Forest Service 2002).

Fungal diseases spread by two primary mechanisms: through airborne spores and through contact with infected roots, although insects can also be carriers (British Columbia Ministry of Forests 1997). Heart rot fungi most commonly spread through airborne spores, whereas root and butt rots may be spread either through spores or through root contact, or both. Infection from airborne spores occurs primarily through wounds exposing heartwood or dead sapwood, although some fungal species can infect trees through small twigs and branch stubs (Hennon 1995, Hennon and DeMars 1997). Since Annosus rot can be transmitted both through airborne spores and through root contact, Annosus infection may spread first through airborne spores to cutfaces on freshly cut stumps and then via root contact to nearby trees. The way in which a particular fungal species spreads to some degree determines the spatial distribution of disease throughout the landscape. Thus, occurrence of root diseases spread through root contact is often patchy, whereas distribution of heart rots, spread by airborne spores, tends to be more uniform (van der Kamp 1991).

The relatively thin bark of coastal tree species facilitates injury and subsequent fungal infection (Hennon 1995). In general, wounds low on the bole are more likely to be infected than higher wounds (British Columbia Ministry of Forests 1997), and the larger the wound, the faster the spread of the infection and associated development of decay (Hennon and DeMars 1997). Wound size may also be correlated with likelihood of infection. Some studies report that wounds less than 1 ft2 are rarely infected, whereas larger wounds usually are (see review in Hennon and DeMars 1997). However, Hennon and DeMars (1997), who observed an infection rate of close to 100% irrespective of wound size, argue that fungal infection is likely present in smaller wounds as well, but harder to detect because of the slower spread of infection from smaller wounds.

Development of fungal decay is generally a slow process, and the prevalence of heart and butt rot in coastal forests is at least partially attributable to the longevity of the dominant tree species (British Columbia Ministry of Forests 1997). Hennon and DeMars (1997) indicate that speed at which fungal decay develops varies with latitudespread is fastest in the US Pacific Northwest and slowest in Alaska, with intermediate rates on the BC coast. In southeast Alaska, trees show little evidence of decay at 100 years of age. At 200 yrs, 65% of western redcedar, 50% of western hemlock, and 20% of Sitka spruce contained some decay (Kimmey 1956).

As decayed trees fall, they often injure nearby trees, thereby creating entry points for new infections and contributing to the persistence of decay fungi in the stand. However, because of the long time it takes for decay to develop after wounding, the bole breakage – wounding cycle is not likely to result in expanding canopy gaps, but rather in a pattern of scattered gaps across landscape (Hennon 1995).

All coastal conifer species are susceptible to fungal infections, although some fungal species show distinct preferences for certain tree species (Hennon 1995, see also Table 2). Cedars are somewhat more decay-resistant than other species, due to protective compounds in the heartwood that inhibit fungal growth. However, this protection is often eventually overcome


Page 33

by a succession of fungi, and heart and butt rot is very common in older western redcedars (van der Kamp 1975).

Scientific name Common name Host(s)

Armillaria spp. Armillaria root rot, honey mushroom Conifers and broadleafs Echinodontium tinctorium Indian paint fungus, brown stringy trunk rot Conifers (rarely Fd or Sx) Fomitopsis pinicola Red belt fungus, brown crumbly rot Conifers and broadleafs Ganoderma applanatum Artist's conk, white mottled rot Conifers and broadleafs Heterobasidion annosum Annosus root and butt rot Conifers and broadleafs Leptographium wageneri Black stain root disease Fd, Pl (rare on Hw) Phaeolus schweinitzii Schweinitzii butt rot, velvet top fungus,

dyer's polypore All conifers

Phellinus pini Pini (red ring) rot or white pocket rot All conifers Phellinus weirii (cedar form) Cedar butt (yellow ring) rot Cw Pholiota spp. Yellow cap fungus B, Hw, P, Sx Postia sericeomollis Brown cubical butt and pocket rot Cw Rhizina undulata Rhizina root rot Conifer seedlings on burned

sites Stereum sanguinolentum Bleeding fungus, red heart rot Conifers except Cw

Table 2: Common fungal species that attack live trees in coastal British Columbia (adapted from British Columbia Ministry of Forests 1997).

2.7.2 Hemlock dwarf mistletoe Hemlock dwarf mistletoe is primarily a parasite of hemlock, although it may occasionally affect Sitka spruce, lodgepole pine, and amabilis fir (British Columbia Ministry of Forests 1995b, Trummer et al. 2000). Infected trees develop swellings and abnormal growth on boles and branches commonly known as ‘brooms’. Severe infection slows down tree growth and weakens the bole of the tree around infection points (Wallis et al. 1980, Thomson et al. 1984, British Columbia of Forests 2001). Hemlock dwarf mistletoe affects stands of all age classes (Garbutt 1998), but is more prevalent in older and all-aged stands (Nevill et al. 1996, Trummer et al. 2000). On Kuiu and Chicagoff Islands in southeast Alaska all hemlocks older than 300 years were infected (Trummer et al. 2000). The parasite propagates by ejecting seeds up to 15 m away. It can spread horizontally from crown to crown and vertically from infected canopy trees to surrounding sub-canopy trees (British Columbia Ministry of Forests 1995b). As a result, the distribution of mistletoe-infected trees tends to be patchy (Smith 1977).

Figure 14: Mistletoe on western hemlock


Page 34

Like fungal decay, mistletoe spreads and develops more slowly at higher latitudes. In southeast Alaska, researchers do not expect infection levels to be damaging for rotation lengths of 90 –120 years (Trummer et al. 2000, Shaw III, 1982). In contrast, on southern Vancouver Island loss of volume is estimated to be 25% in severely infected trees, and 15% in moderately infected trees by age 80 (Thomson et al. 1984). Reasons for slower development at higher latitudes are likely climatic; e.g., freezing can rupture fruit capsules and reduce seed dispersal by up to 95% (Trummer et al. 2000).

2.7.3 The role of pathogens in the forest ecosystem Pathogens and parasites impair the vigour or the structural integrity of affected trees, but they also play an important role as agents of diversity and structural complexity (van der Kamp 1991). Van der Kamp (1991) suggests that pathogens may induce genetic diversity in host populations. By reducing the competitive ability of western hemlock, hemlock dwarf mistletoemay contribute to the persistence of other tree species in hemlock-dominated forests (Trummer et al. 2000, van der Kamp, 1991). Dwarf mistletoe has also been shown to enhance wildlife habitat, including nesting for marbled murrelets (Smith 1982, Bennetts et al. 1996). Decay fungi break down cellulose and lignin and thus play an important part in nutrient cycling (British Columbia Ministry of Forests 1997). Decayed wood provides food and shelter for many organisms, and rots create access to the inner tree bole, which is critical for cavity-nesting birds, especially if rot develops before a tree dies (see Lofroth 1998 and MacKinnon 1998 fore reviews). Finally, decay fungi and mistletoe are often the principal cause of gap formation and thus promote structural complexity and accelerate the succession to gap-phase dynamics after large-scale high-severity disturbances (Hennon 1995).


Injuries by falling trees in a windstorm provide an entry point for fungal infection. Injuries that predispose trees to fungal infection may also be caused by porcupines, bears, beavers, and mistletoe, as well as forest management activities such as harvesting, road building, and thinning (British Columbia Ministry of Forests 1995b, Hennon 1995, Hennon and DeMars 1997, Zeglen 1997). Whereas clearcutting reduces infection sources and thus reduces incidence of mistletoe, partial harvesting may promote infection by maintaining infected overstory trees (British Columbia Ministry of Forests 1995b, Trummer et al. 2000). Both deliberate wounding and management to maintain a certain level of mistletoe-infection have been proposed as a strategy to manage for wildlife habitat and structural diversity in second-growth stands (Hennon and DeMars 1997, Trummer et al. 2000).

Because of the link between pathogen spread and climatic conditions, climate change may lead to changes in pathogen dynamics. The milder temperatures predicted by global climate models will likely accelerate the development of infections and make trees succumb faster to pathogens (Taylor and Taylor 1997). This could accelerate development of structural diversity in second-growth stands. However, increased rates of decay could also mean that the lifespan of woody debris, snags and other structural elements is reduced.


Page 35

2.8 Insects

In coastal forests insects are not typically major disturbance agents but can be influential within some stands (British Columbia of Forests 1999). Most insects occur at endemic levels in these forests, but populations can build to outbreaks when there are favourable climatic conditions and abundant host tree species.

2.8.1 Defoliators Defoliators such as the western blackheaded budworm (Acleris gloverana), green-striped forest looper (Melanolophia imitata), and western hemlock looper (Lambdina fiscellaria lugubrosa), feed primarily on the foliage of western hemlock and may kill trees if defoliation occurs several years in a row. Large outbreaks have been infrequent in the Central Coast (Pojar et al. 1999) and Haida Gwaii (Turnquist et al. 2000). Mortality during outbreaks can be high in severely defoliated areas– e.g., green-striped forest looper may cause up to 25% mortality in stands (K. White pers. comm.). Growth may also be reduced by up to 50% for approximately four years after budworm outbreaks (British Columbia Ministry of Forests 1995b). Historically these defoliators have preferred old and mature stands of western hemlock, primarily in valley bottoms (Koot 1993, British Columbia Ministry of Forests 1995b, Hoggett 2000). However, the recent 1996-2001 outbreak of western blackheaded budworm on Haida Gwaii has caused severe defoliation of young western hemlock stands and nearby mature Sitka spruce (Turnquist et al. 2000, British Columbia Ministry of Forests 2003). The western blackheaded budworm is coincident with outbreaks of hemlock sawfly (Neodiprion tsugae).

Plantations of Sitka spruce in the Central Coast are often attacked by the spruce leader weevil (Pissodes strobi), the woolly adelgid (Pineus spp.), and the green spruce aphid (Elatobium abietinum) (Pojar et al. 1999). Weevils feed on the terminal leader of young trees (0.5 – 12 m high) and do not directly cause mortality but decrease growth and deform stems (Alfaro 1989). Attacks in Sitka spruce plantations begin five years after planting, increase rapidly until stands are 30 years old and then decrease to affect approximately 5% of trees, which is considered an endemic level for coastal stands (British Columbia Ministry of Forests 1996). High-severity attacks of this weevil can cause a shift in species from planted Sitka spruce to western hemlock or amabilis fir. Aphids feed on the needles of all ages of trees. Infestations lasting one to three years have been reported along most of the coast of Haida Gwaii since 1960 (Koot 1991, Garbutt and Vallentgoad 1992). Tree mortality in stands along the east coast of the islands has reached up to 67%.

Severe defoliation can significantly alter the nitrogen cycle – in sites of high precipitation or gravely soils, the nitrogen from needles is lost easily from the forest floor into streams; in other sites, the nitrogen in insect frass and biomass may be retained in the forest floor (Lovett 2002).

2.8.2 Bark Beetles Mountain pine beetle (Dendroctonus ponderosae) has historically been a key natural disturbance in interior lodgepole pine stands (Taylor and Erickson 2003). Feeding by


Page 36

mountain pine beetle and the accompanying blue stain fungus weakens and often kills affected trees (Shore et al. 2000). The beetle prefers mature trees of larger diameter, and major outbreaks tend to be episodic, limited in part by the abundance (or lack) of susceptible stands throughout the landscape. Over the last century, periods between outbreak peaks in BC ranged between 13 and 60 years, while individual epidemics extended over 5-8 years (Wood and Unger 1996). Mountain pine beetle is currently attacking lodgepole pine in the eastern portion of the Central Coast Plan area composed of interior biogeoclimatic zones. In 2002, an area of 166,959 ha in the Central Coast Forest District was affected – most of it in Tweedsmuir Park. 62% of the attacked area experienced greater than 30% pine mortality in the first year of this attack (British Columbia Ministry of Forests 2003). The infestation is currently moving westward down the Dean River to the Pacific Ocean.

Other bark beetles also occur in the Central Coast Plan area. All high elevation stands of subalpine fir contain chronic, but mild, scattered infestations of balsam bark beetle, totaling 4,292 ha in 2002. Very little Douglas-fir and spruce bark beetle was detected in 2002 (British Columbia Ministry of Forests 2003). The effects of these bark beetles on stand structure are similar to those from endemic levels of mountain pine beetle. Hazard ratings by biogeoclimatic subzone for each of the insects discussed above, in addition to some other minor ones can be found in Green and Klinka (1994).

Snags resulting from beetle kill trees offer habitat opportunities for various species (Lofroth 1998). At endemic levels, beetles can open up small patches in a stand, create snags, and increase understory vegetation (Stuart et al. 1989). Outbreaks causing large areas of pine mortality can accelerate succession to other species (Drever and Hughes 2001).

2.8.3 Interactions with other agents of change Insect attacks often weaken trees and predispose them to other disturbance agents (Shepherd 1994). It is hypothesized that weevils can introduce decay fungi, which further weaken trees. The western blackheaded budworm is coincident with outbreaks of hemlock sawfly (Neodiprion tsugae). Mountain pine beetles can increase fuel loads and thus make affected areas more susceptible to stand-replacing fires (Lotan et al. 1984). However, where beetle-related mortality is low to moderate, the hazard of high intensity crown fires may actually be reduced as there is less continuity between canopy fuels (as observed in some Yellowstone stands, Turner et al. 1999).

Temperature and precipitation patterns greatly influence the population dynamics of insects (Williams et al. 2000). For example, heavy rains during the flight period can reduce egg laying and survival of western hemlock looper (Koot 1994) and outbreaks of spruce aphids typically follow mild winters and are inhibited by prolonged cool and overcast periods (Koot 1991). Thus, milder winters may increase the probability of multiple generations of spruce aphids in a year, and increases or decreases in summer precipitation may reduce or increase western hemlock looper populations. Increases in temperatures will likely also benefit the mountain pine beetle, extending its range further north and leading to more frequent and/or more severe outbreaks (British Columbia Ministry of Land Water and Air Protection 2002). Because the climatic triggers of outbreaks are unknown for the green striped looper, it is unclear how climate change may impact this looper (K. White, pers. comm.).


Page 37

Forest management can also influence insect populations. For example, harvesting of large, contiguous blocks of forest may create landscape conditions very favourable to mountain pine beetle outbreaks as soon as the stands reach the age at which trees become susceptible. Replanting with lodgepole pine, rather than other species or a species mix, is also likely to make the post-harvesting landscape more susceptible to mountain pine beetle attack. Similarly, planting stands of pure Sitka spruce increases the susceptibility of landscapes to attacks first by spruce weevils and later by spruce aphids (Garbutt 1998). Since most defoliators of western hemlock prefer mature and old stands, harvesting may decrease the susceptibility of landscapes to outbreaks in the short term (Garbutt 1998). However, the recent widespread defoliation of juvenile stands on Haida Gwaii calls this hypothesis into question.

2.9 Browsing by Mammals

Porcupines (Erethizon dorsatum), beavers (Castor canadensis), and black-tailed deer (Odocoileus hemionus) have substantially affected forest dynamics in coastal BC and Alaska. Cyclic increases of vole (Microtus spp.) populations have also occasionally resulted in clipping, scarring and sometimes mortality of seedlings (Garbutt and Vallentgoad 1992).

2.9.1 Porcupines Porcupines are present throughout the mainland coast of British Columbia, although they are less common on the coast than in other areas of BC (Harestad 1982). They do not occur on Vancouver Island or Haida Gwaii.

Porcupines feed on a diet of grasses, herbs and forbs in summer, but turn to the vascular tissue and foliage of pole-sized or mature trees in winter (Sullivan and Cheng 1989). Vigorous dominants or co-dominants are preferred, likely because of the larger radial increment in vascular tissue in vigorous stems (British Columbia Ministry of Environment Lands and Parks 1998, Zimmerling and Croft 2001). In coastal forests, the tree species most frequently attacked is western hemlock (Sullivan et al. 1986, Garbutt and Vallentgoad 1992, Zimmerling and Croft 2001).

Porcupine feeding can substantially impact stand development in affected stands. Porcupines can partially or completely girdle trees, which may result in stunted growth, crown dieback, or death of the affected tree (Sullivan et al. 1986). The incisor marks also provide an entry point for heart rot (Hennon and DeMars 1997). Thus, porcupine feeding promotes development of canopy gaps and structural complexity and may also counteract the dominance of western hemlock (Alaback 1991).

While the importance of porcupines as a disturbance agent in coastal old-growth is not well documented, porcupine feeding seems to be more prevalent in young, even-aged forest or pole-size second-growth, where it can result in significant losses for commercial timber production (Sullivan et al. 1986, Alaback 1991, D. Steventon pers. comm.). Most of porcupine winter feeding occurs in close proximity to dens, creating a patchy pattern of impact hotspots (Sullivan and Cheng 1989). In the Shames Valley near Terrace, porcupines


Page 38

forage for trees in an area of approximately one to two hectares around denning sites, attacking on average one new tree per day (Zimmerling and Croft 2001). In the Kalum TSA, some degree of porcupine feeding was apparent on 490 out of a total of 720 hectares (Garbutt and Vallentgoad 1992). Feeding was concentrated on lower west- or south-facing slopes and in valley bottom areas. Tree mortality was < 1% in younger stands, though the majority of trees showed evidence of feeding. A study in 30-year old second-growth stands in Khutzeymateen inlet found that about half of the trees showed signs of porcupine feeding, and approximately 5% of previously unaffected trees were being attacked every year (Sullivan et al. 1986). The authors estimate that if losses continue at this rate there will be few commercially viable trees available for harvest at the completion of the rotation. Preliminary results from permanent sample plots in the Scotia River drainage also show some substantial porcupine impacts (British Columbia Ministry of Forests 1999). However, such high impacts may be localized, and it is not clear to what degree these observations are representative of the situation in coastal BC in general. Ongoing studies may help to shed some more light on the question of regional importance of porcupine feeding.

2.9.2 Beavers Beaver activity is generally restricted to the vicinity around streams and wetlands. Where beavers occur, their influence on the aquatic and surrounding terrestrial ecosystem can be substantial. While there are, to our knowledge, no studies of beaver activity specific to the the North or Central Coast, studies from other areas have identified beavers both as a significant forest disturbance and as a keystone species maintaining aquatic ecosystems (e.g., Naiman et al. 1986, Donkor and Fryxell 1999, Gibson 2001). Beavers feed on saplings and the bark of small trees and fall trees and dam streams, often flooding the surrounding forest. Their activity modifies stream channels, increases sediment retention, alters the hydrologic regime of the area, modifies species composition and stand structure in the riparian zone, and affects nutrient cycling in and around the stream (Naiman et al. 1986).

Beavers occur throughout the CIT area, but while they are endemic on the mainland coast, they are considered an introduced species on Haida Gwaii, where they were brought by the BC Game Commission in the first half of the 20th century to enhance the fur trade. Whereas beaver activity is an important factor for maintaining historical ecosystem conditions in wetlands along the mainland coast, the ability of beavers to drastically alter wetland dynamics is a cause for concern on Haida Gwaii, where beaver dams have changed drainage patterns, flooded shorebird nesting sites, and affected salmon spawning grounds (Englestoft 2002).

2.9.3 Deer The mild coastal climate and limited snow cover at lower elevations favours deer and other ungulates (RGIS 2002). Deer browsing has especially large impacts on the islands of the outer coastal environment, where winters are mild and wolves and other large predators are often absent (Klein 1965, Alaback 1991). Perhaps the most dramatic example of the effect of deer on island ecosystems in the CIT area is Haida Gwaii, where deer were introduced around the turn of the 19th century to enhance game hunting (Englestoft 2002). Like some of the more remote islands off the mainland coast, Haida Gwaii is free of wolves and cougars,


Page 39

although black bears may occasionally take fawns. Thus deer were able to spread essentially unchecked, only limited by food supply and human hunting pressure.

Intensive browsing by deer may virtually denude the forest floor, with dense understory vegetation primarily confined to areas out of reach of deer (Klein 1965). Thus, a study by Stockton (summarized in RGIS 2002) comparing vegetation on a series of islands found that understory composition and structure were greatly simplified on islands with deer when compared to islands without deer. Some of the understory shrubs and herbs most heavily impacted by deer browsing include vaccinium species and other ericaceous shrubs, as well as devil's club, skunk cabbage, and various ferns (Pojar 2003). Pojar (2003) notes that elimination of ericaceous shrubs in particular may have far-reaching impacts on ecosystem function, since the mycorrhizae associated with these shrubs fulfil a key role in nutrient cycling.

Deer also browse tree seedlings, with potentially serious consequences for forest regeneration and stand composition. While browsing of tree seedlings is not always lethal, repeated browsing tends to reduce or even arrest height growth, at least until the trees 'escape' out of reach of the deer (Vila et al. 2002). Among tree seedlings, deer generally favour cedar (both types, but especially redcedar) over hemlock and spruce (Pojar and Banner 1984, Pojar 2003). This is likely especially true on Haida Gwaii, where cedar has evolved until very recently without browsing pressure, and thus has lower levels of the defensive chemical compounds and is more palatable to deer than trees on the mainland coast (Vourc'h and Martin 2001). Levels of chemical defense compounds will likely increase with selective browsing, thus possibly reducing the pressure on cedar in the future. Nonetheless, the preference of deer for cedar will likely lead to shifts in species composition towards spruce and especially hemlock, which is least affected by deer browsing (Vila et al. 2002, Pojar 2003).

By reducing the understory, deer browsing also impacts insect and songbird populations. Thus, Allombert observed a substantial reduction in invertebrate abundance, and Martin found that islands with a long history of deer population had an impoverished bird fauna, likely attributable to reduced breeding success because of increased nest predation (both reviewed in RGIS 2002).

Although deer have impacted all terrestrial ecosystems on Haida Gwaii to some degree, some areas are more heavily affected than others. Browsing impacts are most severe in shoreline plant communities and floodplain forests, though substantial impacts are also noticeable in

Figure 15: Heavily browsed young spruce on southern Moresby Island


Page 40

upland forests, and to a lesser degree in alpine and subalpine areas (Pojar and Banner 1984, Pojar 2003). Browsing pressure in high-elevation areas has increased over the last 25 years, possibly as a result of increased deer populations and/or milder winter temperatures.

2.9.4 Interactions with other agents of change Like most other relationships between natural disturbance agents on the coast, interactions between mammal browsing and other agents of forest disturbance are often synergistic. Deer populations generally benefit from early-seral habitat, especially if openings are small, in close vicinity to cover, as is typically the case in coastal landscapes subject only to natural disturbance dynamics. Thus, natural disturbances on the coast tend to favour deer. In turn, deer browsing, if severe, may destabilize steep slopes by removal of the understory and thus increase the likelihood of slope failures (Pojar 2003). Porcupine feeding scars provide entry points for fungi, thus increasing the likelihood of fungal diseases, whereas beavers alter wetland dynamics and thus may affect the likelihood and severity of flooding.

Forest management and global climate change could substantially influence the role of browsing animals. The association between porcupine feeding and second-growth was already noted above. The early-seral habitat created by harvesting is favourable to deer at least in summer, though in winter they tend to seek out forests with old growth structure for their increased thermal cover, reduced snow depth, and abundant winter browse (McNay 1996, see also Sharpe in RGIS 2002). By removing tree cover needed for shelter, harvesting can also drive deer to concentrate in narrow riparian buffer strips, making them more susceptible to predation (S. Liepins, pers. comm.). If global warming will result in milder winters with reduced snow cover, this may lead to increases in deer and other ungulate populations presently limited by access to food during winter (Taylor and Taylor 1997). Conversely, in inland areas and at higher elevations, where winter temperatures tend to be more severe, higher snowpacks due to increased winter precipitation may decrease ungulate populations.

Another significant factor influencing deer populations is increased human presence and improved access for hunters and poachers in actively managed forests. On Haida Gwaii, browsing impacts were significantly reduced in areas with easy human access, although the number of deer actually killed by hunters was likely not large enough to control population levels (Martin and Baltzinger 2002). However, increased human access and hunting pressure has also been shown to extirpate large predators (e.g., Person et al. 1996). In areas on the mainland coast where predators are currently controlling deer populations, reduction in predator populations due to expansion of the road network and increased human access to backcountry areas may ultimately outweigh the increased pressure on deer populations by human hunters.


Page 41

3 Natural Disturbance Dynamics Throughout the Coastal Landscape: An Overview

From the preceding review of natural disturbance agents it should be apparent that natural disturbances do not occur homogeneously across the coastal landscape. Rather, several of the agents are confined to, or occur predominantly in specific stand types and on specific terrain types. Using information on natural disturbance regimes effectively therefore requires a clear understanding of the particular disturbance dynamics each part of the landscape is subject to. In this section, we present a summary of disturbance dynamics across the coastal forest landscape.

Landform and substrate play an important role in shaping ecosystem dynamics and regulating disturbance regimes in mountainous landscapes (Swanston et al. 1988, Swanson et al. 1993, Montgommery 1999), especially in the mountainous terrain of the BC coast. The subdivision into biogeoclimatic subzones and variants, and the coarse stratification into natural disturbance types used in the Biodiversity Guidebook, are of limited use for describing disturbance dynamics in this type of system, since they do not capture landform or substrate differences at a fine enough scale. As an alternative, we propose a stratification into disturbance units based primarily on similarity in disturbance regimes and vegetation dynamics. These units can be described based on their relative position in the landscape with respect to watershed topography, as well as topo-edaphic characteristics and, in some instances, predominant vegetation type.

Below, we present three tables that summarize disturbance characteristics for the coastal landscape. The first table (Table A) gives a summary of patch size, frequency and spatial distribution for stand-replacing disturbances on the coast. More detailed tables summarizing the information presented in the available disturbance history studies can be found in Appendix 1. The second table (Table B) summarizes disturbance dynamics in the CIT area stratified by disturbance unit and covers wind, geomorphic disturbances and flooding. The most important disturbance agents for each unit are highlighted in red and italics. The third table (Table C) summarizes characteristics of other disturbance agents which are either of relatively minor importance on the coast or, in the case of pathogens, relatively unaffected by landscape position and dominant vegetation type.


Page 42

Table A: Compiled estimates of patch size, frequency and spatial distribution for stand-replacing disturbance agents. The tables in Appendix 1 summarize the existing information used to derive the estimates presented below.

Disturbance Type

Mean patch

size (ha)

Frequency or return intervala Spatial distribution Comments

Wind 5-15

Return interval Exposed sites: ~ 50 – 300 years Sheltered sites: > 300 years - never

More susceptible sites: • in valley bottoms directly exposed

to southerly storms • on SW to SE slopes, especially

near the ridge • recent edges • closer to coast • Ba, Hw, or Ss leading upland

stands

• Frequency estimates are uncertain since the literature does not provide formal analyses for blowdown frequency.

• Blowdown statistics include events where up to 90% of trees are left standing and should therefore be considered only partially stand-replacing.

• The occurrence of stand-replacing blowdown is dependent on exposure to strong winds. The actual proportion and distribution of susceptible (i.e., ‘exposed’) sites in the CIT area has not been formally determined. Most likely areas where exposed sites may occur include the west coast of Haida Gwaii, northern Vancouver Island, as well as potentially some outer coast areas on the mainland coast.

Landslide 1-10

Unlogged: ~ 0.005-0.05 /yr/km2

Logged: 10 – 30 times higher than natural

Unlogged: on steep, complex or concave slopes, more frequent closer to coast Logged: On moderate to steep slopes, more frequent on slopes closer to coast E, SE, S slopes

• Frequency estimates depend on assumptions made about how long landslides are evident in air photos.

• Harvesting primarily increases the frequency of debris flows and slides, whereas frequency of other types of mass-wasting is less affected. Frequency estimates for logged areas include slides originating from roads and most were based on pre- Forest Practices code data (but see Sierra Legal Defense Fund 1997a).

Snow Avalanche 10-50

< 20 years in non- treed zone of path < 130 years in treed zone

More frequent on open slopes or gullies > 30-50o and on those likely to experience snowpack warming or snow loading from winds.

More frequent in the start zone than the run out zone.

• No data available for the CIT area or nearby coastal areas. • Return interval is the average time between events within an

established path.

Fire < 1 - 100 ~ 300 years – never More likely on south facing slopes or well drained Cw/Hw slopes

Estimates are very imprecise and based on very little historical data specific to the CIT area.

_________________________ a) Disturbance frequency is the average number of disturbance events one would expect per year at a given point in the landscape (or, in the case of landslide frequency, within a given

fixed area). Return Interval is the inverse of disturbance frequency and refers to the average interval between disturbance events at a given point in the landscape. See also footnote iii on page 10.


Page 43

Table B: Disturbance agents that vary substantially with site characteristics. The first line in each cell describes the frequency of the disturbance agent in the disturbance unit, the second line describes size distribution of opening, and the third describe disturbance severity. Disturbance agents in red and italics are the most prevalent on the disturbance unit, disturbances characteristics in bold are the most likely for each agent (see following page for explanation of symbols).

Disturbance Unit

Wind

Geomorphic

Flooding

Floodplains and fans endemic rJ, gap

sl-fT, soil, part.

freq. to infreq. rJ

soil, part.- compl

endemic, freq. to infreq. rJ, gap

soil, part.

steep, exposed

endemic, infreq.? to rare rJ, gap

sl-fT, soil, part.-compl.?

freq. rJ

soil, part.- compl.

N/A

steep, sheltered

endemic gap

sl-fT, soil

freq. rJ

soil, part.- compl. N/A

gentle, exposed

endemic, infreq.? to rare rJ, gap


rare rJ

soil, part.- compl

N/A

Productive upland forest, (Hw/Ss leading or Hw/Ba leading)

gentle, sheltered

endemic gap

sl-fT, soil

rare rJ

soil, part., compl

N/A

steep endemic

gap sl-fT, soil

freq. rJ

soil, part., compl

N/A

Upland forest, (Cw/Hw leading)

gentle rare gap

sl-fT, soil

rare rJ

soil, part.- compl

N/A

Bog forests, bog woodland, blanket bogs

rare gap

sl-fT, soil

rare (?) rJ

soil, part.- compl

endemic gap

soil, part.

Avalanche tracks N/A freq.

rJ soil, part.- compl

N/A

Subalpine forest and alpine

endemic gap

sl-fT c), soil

freq.? to rare rJ

soil, part.- compl N/A

Gullies endemic, rare?

rJ, gap sl-fT, soil, part.-compl.?

freq. rJ

soil, part.- compl N/A

Estuaries endemic rJ, gap


rare rJ

soil, part.- compl

endemic gap

soil, part.

Salt spray zone endemic rJ, gap


rare rJ

soil, part., compl

endemic gap

soil, part.

Table C: Disturbance agents that vary little with site characteristics or are generally of minor significance on the North Coast (see following page for explanation of symbols).

Characteristics Fire Short-term climatic Pathogens Insects Browsing

Frequency rare rare end.-indirect endemic, end.-indirect endemic Size distribution of openings rJ gap gap gap rJ

Severity part.- compl. sl sl-fT sl, fT, part. sl–fT


Page 44

Key for Tables B and C: Site type: steep: steep slope (>30%) gentle: slope < 30% exposed: Areas regularly subject to strong winds capable of causing severe damage to trees, including windward slopes in topographically exposed locations (to southeasterly winter storms or katabatic winds); valleys and ridges parallel to the prevailing storm direction; constrictions and saddles; mountain flanks where winds may ‘wrap around’ from exposed slopes. sheltered: Areas not subject to prevailing winds or areas where winds are rarely strong enough to cause blowdown. Disturbance attributes: (In cases where multiple attributes apply, predominant ones are marked in boldface. Question marks indicate uncertainty about the prevalence or effects of a disturbance agent.) Disturbance rates/intervals: endemic: an agent that occurs continually or in frequent outbreaks. Impact is persistent and results in mortality but not catastrophic, affects individual trees or small groups of trees. end. - indirect: an agent that is endemic, but mainly weakens trees, making them susceptible to other forms of disturbance. freq.: severe disturbance occurs frequently (interval < 100yrs, often much shorter). These forests rarely reach maturity/rotation age. infreq.: disturbance occurs frequently enough to prevent succession to all-aged forest. Stand structure in these forests is single-cohort or multi-cohort. (intervals between 100 and 300 yrs). rare: agent rarely affects stands. Not significant on the landscape level, but may occur in specific stands. Size distribution of openings: (For ‘freq.’, ‘infreq.’, and ‘rare’ types we give size range and shape of the distribution of opening sizes where possible) gap: small canopy gaps, typically less than 10 trees. rJ (‘reverse J’-shaped): is used to represent a size distribution where most openings are small and frequency of openings rapidly decreases with increasing size. Disturbance severity (proportion of overstory mortality, mode of death, and severity of understory disturbance): sl .sublethal effects such as reduced radial and height growth fT: falling canopy trees cause mechanical damage to adjacent canopy and sub-canopy trees (wounding or breakage; implies some damage to understory). soil: soil disturbance/exposure of mineral soil (implies damage to understory). part.: partial canopy replacement. compl.: complete canopy replacement; survival of canopy trees constrained to single veterans or small patches.


Page 45

4 Effect of Natural Disturbances on Habitat Requirements of Forest-dwelling Species

Natural disturbances, by affecting the development of vegetation patterns across the landscape, influence the population dynamics of many forest-dwelling species (Oliver et al. 1998). Animal and plant species are thought to be adapted to certain natural disturbance regimes and the forest structure and landscape patterns they create. For example, the proportion of species which breed in late-successional forests or use downed wood to breed in, tend to increase with longer disturbance return intervals in British Columbia (Bunnell 1995). In Section 2, which describes disturbance dynamics and effects by disturbance agent, we have already detailed some of the more important relationships to habitat. Below, we briefly summarize how natural disturbances, through their influence on stand and landscape structure, maintain habitat characteristics important for species and communities in the CIT area.

• Many forest-dwelling species depend on the structural attributes of old-growth forests – e.g., downed trees for shelter and denning, blown-down lichen and foliage for winter browse, easy to hollow-out snags for feeding and cavity nesting, or mistletoe platforms for nesting of murrelets. Wind and pathogens, the two key drivers of gap-phase replacement, are primarily responsible for ensuring a continuous supply of these habitat elements, although most other natural disturbance agents also create snags and downed wood and thus contribute to the structural complexity of coastal forests.

• Many species have seasonal requirements for mature and old-growth forest. E.g., marbled murrelets use old-growth stands for nesting and mountain goats and black-tailed deer seek out old stands for shelter and food in winter (Tamblyn and Horn 2001, Blood et al. 2002). Due to the absence of frequent stand-replacing events, the coastal landscape has historically provided an ample supply of suitable late-successional habitats.

• Certain species, such as grizzly bears or moose, require early seral habitat and wetlands for forage and browse (MacHutchon et al. 1993, Ferguson and Pope 2001, Pollard 2001). On the coast, natural disturbances provide early-seral and wetland habitats primarily as a persistent feature in specific landscape positions – e.g., avalanche tracks, floodplains, mass wasting, riparian areas.

• Certain species require that different habitat types used for foraging, bedding and cover from predators, or different habitat types used during different seasons, occur within close enough proximity to allow the animals access to each of them, when required. For example, grizzly bears feed in estuaries or avalanche chutes but tend to bed in mature to old-growth forest (MacHutchon et al. 1993) and mountain goats in the winter feed and take shelter in mature forests within 400 m of steep escape terrain (Tamblyn and Horn 2001). Because the coastal landscape


Page 46

is naturally partitioned into landform units with distinct dynamics, juxtaposition of different habitat types has historically been a persistent landscape feature.

• Forests and forest disturbances have a substantial influence on watershed hydrology and thus help to maintain the habitat of species dwelling in the rivers, wetlands, and riparian areas. Moderate, sustained inputs of sediments and large woody debris supplied by geomorphic disturbances, flooding, and wind in concert with pathogens are particularly important for shaping and maintaining the complex channel system characteristic of coastal floodplains, which serves as habitat for many species, including salmon. Thus, the services of natural disturbances may be a critical factor for the persistence of salmon, a ‘cornerstone species’ that is at the core of a complex food web supporting both aquatic and terrestrial ecosystems (Naiman et al. 2002).

5 Management Issues

Natural disturbances, and the way they interact with coastal forest ecosystems, can teach us several key lessons regarding the future management of these systems. The most important implications arising from the characteristics of disturbance dynamics described in the previous sections are summarized below, organized into three general categories: observations regarding the spatial and temporal scales of forest dynamics; observations pertaining to the question how, and to what degree natural disturbances patterns can serve as a guideline for harvesting; and implications of disturbance dynamics for species habitat.

(1) What are appropriate spatial and temporal scales for planning and management?

• Landscape dynamics on the coast are strongly influenced by the underlying mosaic of physiography and topography. The paradigm of a shifting mosaic of seral stagesxiii that underlies the management approach proposed in the Biodiversity Guidebook does not fit this type of landscape very well, since much of the early-seral vegetation is confined to a few areas subject to repeated disturbance, whereas most of the operable, productive forests of interest to harvesting rarely experience severe, large-scale disturbance. A planning approach that fails to take this into consideration will make it difficult to incorporate knowledge about natural dynamics in a meaningful way. A stratification into watershed and landscape sections dominated by different disturbance dynamics would be a logical subdivision of the landscape in terms of maintaining ecological processes and their outcomes. Such a stratification could also help identify planning areas where special management is required to avoid undesirable interactions with dominant disturbance agents.

xiii This paradigm applies to a landscape where forest dynamics are similar across the landscape and occurrence of natural

disturbances is not restricted to particular topographic or other site characteristics. In such a landscape, the spatial distribution of young and old forest throughout the landscape is similarly independent of topography or other site characteristics, and the forest mosaic therefore ‘shifts’ at random across the landscape over time as disturbances create new openings and older openings gradually merge back into the matrix of old forest.


Page 47

• Climate is also a significant driver of forest dynamics on the coast. Predicted changes in future climate will likely increase the incidence and/or severity of most disturbance agents and affect how forests respond to these agents and other processes. Consideration of intermediate and long-term climatic trends is therefore essential for anticipating and planning future forest development.

• The coastal forest is, in many respects, a system of long time scales. Tree species are long-lived, intervals between disturbances are measured on the scale of centuries, and processes of decay and decomposition progress equally slowly. Thus, undesired effects of forest management, such as potential negative impacts on terrestrial and aquatic habitat because of reduced inputs of large woody debris, may take a long time to manifest and even longer to reverse. This suggests a need for careful long-term planning and continued monitoring if negative ecosystem impacts are to be minimized.

• While coastal forests are primarily driven by gap dynamics, larger, more severe natural disturbance events do occur episodically in some parts of the landscape. Proactive planning and forecasting over long time horizons will be required to avoid negative impacts of such rare events on sustained timber supply and biodiversity. This could include considering the potential effect of earthquakes, tsunamis, and landslides on infrastructure and taking into account losses to episodic severe natural disturbances in timber supply analyses as well as land use planning and conservation strategies.

(2) How can natural disturbance regimes serve as a guide for forest management?

• The forests on the BC coast are predominantly considered NDT (Natural Disturbance Type) 1 in the Biodiversity Guidebook, with intervals between stand replacing disturbances averaging 250 to 350 years. Although some small sections of the coastal landscape may historically have experienced such relatively frequent stand-replacing disturbances, there is no evidence to support the disturbance intervals postulated in the Guidebook on a broader scale. The scientific evidence also indicates that in forests considered productive and operable, disturbance return intervals for large, stand-replacing events were generally much longer than commercial rotation periods.

• Because coastal forests are primarily driven by gap dynamics, natural disturbance regimes provide no template for clearcutting systems, or traditional approaches to even-aged management in general. If the objective of forest management is to emulate natural disturbance patterns, a silvicultural system that emulates a gap-phase replacement regime, such as variable retention with high retention levels, is the closest match. In cases where high retention levels are rejected because of economical considerations, the potential risk of negative ecological impacts from silvicultural practices that do not maintain natural structure and dynamics may be reduced through extended rotations, and by constraining the proportion of the landbase over which harvesting occurs. However, if a reduction in harvesting landbase is to be offset by more intensive management zones, it will be important to carefully consider the allocation of such areas, since in many cases the most


Page 48

productive areas from a timber growth perspective also tend to be of primary importance for biodiversity.

• Most of the interactions between harvesting and natural disturbance on the coast are synergistic, i.e., harvesting is likely to increase occurrence of natural disturbances. Since climate change will most likely also increase the incidence or severity of most natural disturbances, overall disturbance rates are likely to go up in the future even without harvesting, and harvesting will almost certainly lead to disturbance rates substantially higher than the landscape has historically experienced. This emphasizes the need to monitor landscape conditions to recognize and avert potential negative impacts on ecosystem function and biodiversity.

• Although natural disturbances are a key aspect of forest dynamics, forests are shaped and maintained by a variety of factors that are partially or completely outside the scope of this report. Even if harvesting is designed to emulate natural disturbances very closely, consideration of other factors is essential for maintaining ecosystem integrity and productivity. These factors include watershed hydrology, nutrient cycling, soil development, availability of seed sources, as well as genetic health of tree populations.

(3) One of the key principles of ecosystem-based management is that species are adapted to, and therefore most likely to persist under the environment and processes they have experienced historically. What are the key habitat characteristics historically maintained by natural dynamics, and how is forestry affecting these characteristics?

• The majority of the coastal forest was historically in all-aged stands dominated by gap-phase replacement. Reducing the supply of this type of habitat may negatively impact old-growth-dependent species.

• Historically, stand-replacing disturbances were largely constrained to susceptible locations within the landscape, whereas late seral forest was persistent throughout the remainder of the forested area. Current areas of persistent, late seral forest can act as habitat refugia for species dependent on them. Continued existence of such refugia can be ensured by setting aside old growth areas and by managing parts of the landscape to develop and retain old growth characteristics.

• Stands regenerating after clearcut harvesting tend to go through a relatively long period where the canopy is very dense and little light reaches the understory. Understory plants and species dependent on these plants may drop out of the stand during this period. In areas where clearcut harvesting is extensive, requirements for persistence and recolonization of such species may therefore need to be considered if these species are to be maintained in the landscape.

• The historical landscape mosaic provided routes for movement and dispersal of species dependent on particular habitat characteristics. Where harvesting or other forms of development interrupt these routes, the ability of species to move and disperse throughout the landscape may be impacted.


Page 49

• Several species in the area rely on the juxtaposition of early-seral herb and shrub communities to forest cover for protection and denning. Where forest management creates larger openings and reduces the amount of small canopy gaps characteristic of old-growth forests, the result may be a lack of suitable cover for some species.

• Standing dead and downed wood are important habitat elements in both terrestrial and aquatic ecosystems. Natural disturbances generally ensure an ample, sustained supply of snags, root wads, and downed wood. Reduced recruitment of snags and large woody debris in managed stands may have long-term effects on nutrient cycling, tree regeneration, and habitat quality.

• Productive valley bottom areas and estuaries are focal areas for forest planning and management since they often have high economic timber value as well as very high habitat value. Narrow riparian buffers are susceptible to blowdown and may not provide effective protection of hydro-riparian ecosystems. In order to sustain input of large woody debris into stream channels at historical levels and allow for the natural meandering of floodplain channels throughout the floodplain area, most, if not the entire valley bottom might need to be protected in some cases.

• Introduced species have had substantial impacts on the island ecosystem of Haida Gwaii, as well as on the North American continent in general. Forestry, through silvicultural trials and harvesting operations, has in the past contributed to the introduction and spread of forest pests and weeds. Since many of the species that turn into ‘problem species’ are strong competitors in disturbed areas, it is important for forest managers to be aware of the potential risks posed by introduced species, as well as the means by which these species are spread.


Page 50

References Abbe, T. B. and Montgomery, D. R. 1996. Large woody debris jams, channel hydraulics, and habitat

formation in large rivers. Regulated Rivers: Research and Management 12:201 - 221.

Agee, J. K. 1993. Fire Ecology of the Pacific Northwest Forests. Island Press, Washington, D.C.

Agee, J. K., and L. Smith. 1984. Subalpine tree re-establishment after fire in the Olympic Mountains, Washington. Ecology 65:810-819.

Alaback, P. B. 1991. Comparative ecology of temperate rainforests of the Americas along analogous climatic gradients. Revista Chilena de Historia Natural 64:399-412.

Albritton, D. L., and L. G. Filho. 2001. Technical summary: Climate change 2001: the scientific basis. Contribution of the Working Group 1 to the Third Assessment by the Intergovenmental Panel on Climate Change.

Alfaro, R. 1989. Probability of damage to Sitka spruce by the Sitka spruce weevil, Pissodes strobi. J. Entomol. Soc. Brit. Columbia 86:48-54.

Allen, R. B., P. J. Bellingham, and S. K. Wise. 1999. Immediate damage by an earthquake to a temperate montane forest. Ecology 80:708-714.

Allen, T. R., and S. J. Walsh. 1996. Spatial and compositional pattern of alpine treeline, Glacier National Park, Montana. Photogrammetric Engineering & Remote Sensing 62:1261-1268.

Ames, K. M., and H. D. G. Maschner. 1999. Peoples of the Northwest Coast: Their Archeology and Prehistory. Thames and Hudson, London, UK.

Arsenault, A. 1995. Pattern and Process in Old-growth Temperate Rainforests of Southern British Columbia. Ph.D. thesis. University of British Columbia, Vancouver, BC.

Banner, A., W. MacKenzie, S. Haeussler, S. Thomson, J. Pojar, and R. Trowbridge. 1993. A field guide to site identification and interpretation for the Prince Rupert Forest Region. Land Management Handbook No. 26, British Columbia of Forests.

Banner, A., J. Pojar, and G. E. Rouse. 1983. Postglacial paleoecology and successional relationships of a bog woodland near Prince Rupert, British Columbia. Canadian Journal of Forest Research 13:938-947.

Banner, A., J. Pojar, J. Schwab, and R. Trowbridge. 1989. Vegetation and soils of the Queen Charlotte Islands: Recent impacts of development. Pages 261-279 in G. G. E. Scudder and N. Gessler, editors. The Outer Shores. Based on the proceedings of the Queen Chalotte Islands First International Symposium, Vancouver, BC.

Banner, A., and G. E. Rouse. 1983. Postglacial paleoecology and successional relationships of a bog woodland near Prince Rupert, British Columbia. Canadian Journal of Forest Research 13:938-947.

Beckmann, L., M. Dunn, and K. Moore. 1997. Chapter 8: Effects of climate change on coastal systems in British Columbia and Yukon. Pages 8.1 - 8.26 in Canada Country Study: Climate impacts and adaptation. Environment Canada.

Bennetts, R. E., G. C. White, F. G. Hawksworth, and S. E. Severs. 1996. The influence of dwarf mistletoe on bird communities in Colorado ponderosa pine forests. Ecological Applications 6:899-909.


Page 51

Blood, D. A., K. MacIntosh, I Robertson, E. Robertson, and P. Eredics. 2002. Coastal Black-tailed Deer Winter Range Inventory and Mapping in the Mid-Coast Forest District - First Year Report. Ministry of Environment Lands and Parks Wildlife BranchCariboo Region (5).

Bormann, B. T., and M. G. Kramer. 1998. Can ecosystem-process studies contribute to new management strategies in the coastal Pacific Northwest and Alaska? Northwest Science 72:77-83.

Boyd, R., editor. 1999. Indians, fire, and the land in the Pacific Northwest. Oregon State University Press, Corvallis, OR.

Bradley, T., and P. Tueller. 2001. Effects of fire on bark beetle presence on Jeffrey pine in Lake Tahoe Basin. Forest Ecology and Management 142:205-214.

British Columbia Ministry of Environment Lands and Parks. 1998. Inventory Methods for Mountain Beaver, Bushy-tailed Woodrat & Porcupine. Standards for Components of British Columbia's Biodiversity No. 27, Victoria, BC.

British Columbia Ministry of Forests. 1995a. Biodiversity Guidebook. Forest Practices Code Guidebook, B.C. M.o.F. Communications Branch, Victoria, B.C.

British Columbia Ministry of Forests. 1995b. Defoliator Management Guidebook. Forest Practices Code Guidebook, B.C. M.o.F. Communications Branch, Victoria, B.C.

British Columbia Ministry of Forests. 1996. Management of Terminal Weevils in British Columbia. Forest Practices Code Guidebook, B.C. M.o.F. Communications Branch, Victoria, B.C.

British Columbia Ministry of Forests. 1997. Tree Wounding and Decay Guidebook, Forest Practices Code Guidebook, B.C. M.o.F. Communications Branch, Victoria, B.C.

British Columbia Ministry of Forests. 1999. North Coast Timber Supply Area Analysis Report. Online at http://www.for.gov.bc.ca/tsb/tsr2/tsa/tsa26/report/21ts99ar-final.pdf

British Columbia Ministry of Forests. 2001. Simulating Western Hemlock Dwarf Mistletoe Impact in Coastal British Columbia. http://www.for.gov.bc.ca/hfp/pubsforhealth.htm.

British Columbia Ministry of Forests. 2003. Forest Health Conditions in the Vancouver Forest Region - 2002. Annual Report.

British Columbia Ministry of Land Water and Air Protection. 2002. Climate Change and Terrestrial Ecosystems. Online at http://wlapwww.gov.bc.ca/air/climate/indicat/index.html.

Brown, K. J., and R. J. Hebda. 1998. Long-term fire incidence in coastal forests of British Columbia. Northwest Science 72:64-66.

Brun, S. E., and D. Etkin. 1997. Part 2: Natural hazards in Canada. in Coping with natural Hazards in Canada: Scientific, Government and Insurance Industry Perspectives. Environmental Adaptation Research Group, Environment Canada, and Institute for Environmental Studies, University of Toronto, Toronto, ON.

Bryant, M. D., and F. H. Everest. 1998. Management and conditions of watersheds in southeast Alaska: the persistence of anadromous salmon. Northwest Science 72:249-267.

Bunnell, F. L. 1995. Forest-dwelling vertebrate faunas and natural fire regimes in British Columbia: patterns and implications for conservation. Conservation Biology 9:636-644.

Burton, P. J., and S. G. Cumming. 1995. Potential effects of climatic change on some western canadian forests, based on phenological enhancement to a patch model of forest succession. Water, Air and Soil Pollution 82:401-414.


Page 52

Butler, D. R., and S. J. Walsh. 1990. Lithologic, structural, and topographic influences on snow-avalanche path location, eastern Glacier National Park, Montana. Annals of the Association of American Geographers 80:362-378.

Canadian Forest Service. 2002. Common Tree Diseases of British Columbia. Online at http://www.pfc.forestry.ca/diseases/CTD/.

Caza, C. L. 1993. Woody Debris in the Forests of British Columbia: A Review of the Literature and Current Research. Land Management Report No. 78, BC Ministry of Forests, Victoria, BC.

Chatwin, S. G., D. E. Howes, J. W. Schwab, and D. N. Swanston. 1994. A Guide for Management of Landslide-prone Terrain in the Pacific Northwest. Land Management Handbook No. 18, BC Ministry of Forests Research Program.

Church, M. 1994. The landscape of the Pacific Northwest. in D. L. Hogan , P. J. Tschapinski, and S. Chatwin, editors. Carnation Creek and Queen Charlotte Islands Fish/Forestry Workshop: Applying 20 Years of Coast Research to Management Solutions, Queen Charlotte City, BC.

Clague, J. J. 1989. Quaternary geology of the Canadian Cordillera. in R. J. Fulton, editor. Quarternary Geology of Canada and Greenland. Geological Survey of Canada.

Clague, J. J., editor. 1996. Paleoseismology and Seismic Hazards, Southwestern British Columbia. Geological Survey of Canada, Bulletin No. 494.

Clague, J. J., P. T. Bobrowsky, and I. Hutchinson. 2000. A review of geological records of large tsunamis at Vancouver Island, British Columbia, and implications for hazard. Quaternary Science Reviews 19:849-863.

Combellick, R. A., and W. E. Long. 1983. Geologic Hazards in Southeastern Alaska: an Overview. Alaska Report of Investigations 83-17, State of Alaska, Department of Natural Resources, Division of Geological and Geophysical Surveys.

Combellick, R. A., and R. J. Motyka. 1995. Geologic Hazards in the Yakataga Planning Area, Southeastern Alaska: an Overview. Report by the State of Alaska, Department of Natural Resources, Division of Geological and Geophysical Surveys.

Dalton, R. 2003. The coast road. Nature 422:10-12.

Daniels, L. D., P. Marshall, R. E. Carter, and K. Klinka. 1995. Age structure of Thuja plicata in the tree layer of old-growth stands near Vancouver, British Columbia. Northwest Science 69:175-183.

Deal, R. L., C. D. Oliver, and B. T. Borman. 1991. Reconstruction of mixed hemlock-spruce stands in coastal Alaska. Canadian Journal of Forest Research 21:643-654.

DeGroot, A., A. Banner, P. LePage, and J. Shaw. 2002. The Blanket Bog - Upland Forest Complex of North Coastal British Columbia: Succession, Disturbance and Implications for Management. Extension Note No. 52, Forest Sciences, Prince Rupert Forest Region, Smithers, BC.

Demarchi, D. A. 1996. An Introduction to the Ecoregions of British Columbia. Wildlife Branch, Ministry of Environment, Lands and Parks, Victoria, BC, Online at http://srmwww.gov.bc.ca/rib/wis/eco/bcecode.html.

Donkor, N. T., and J. M. Fryxell. 1999. Impact of beaver foraging on structure of lowland boreal forests of Algonquin Provincial Park, Ontario. Forest Ecology and Management 118:83-92.

Dorner, B., H. Sandmann, and C. Wong. 2003. Analysis of Pre-1860 Disturbance Intervals in the Fraser Timber Supply Area. Report prepared for the British Columbia Ministry of Sustainable Resource Management, Victoria, BC.


Page 53

Dorner, B., and C. M. Wong. 2002. Natural Disturbance Dynamics on the North Coast. Background Report prepared for the North Coast LRMP, Prince Rupert, BC.

Drever, R., and J. Hughes. 2001. Salvaging Solutions. David Suzuki Foundation, Forest Watch of British Columbia Society, and Canadian Parks and Wilderness Society, Vancouver, B.C.

Emili, L. A., C. L. Konig, J. Turunen, J. S. Price, B. G. Warner, R. Aravena, and A. Banner. 1998. Hydrogeochemistry and peat development in a coastal cedar-hemlock bog-forest complex. Northwest Science 72:60-63.

Englestoft, C. 2002. Section 1: Introduced Species Account. Pages 1-33 in Restoration Priorities Associated with Introduced Species Impacts on Haida Gwaii/Queen Charlotte Islands: Perspectives, and Strategies. Report prepared for the Council of the Haida Nation (CHN) Forest Guardians and the Ministry of Water, Land and Air Protection, Queen Charlotte City, BC.

Environment Canada. 2002. Canadian Climate Normals 1971 - 2000. Online at http://www.msc-smc.ec.gc.ca/climate/climate_normals/index_e.cfm.

Evans, S. G., and J. J. Clague. 1997. Chapter 7: The impact of climate change on catastrophic geomorphic processes in the mountains of British Columbia, Yukon and Alberta. Pages 7.1-7.16 in Canada Country Study: Climate impacts and adaptation. Environment Canada.

Everham, E. E., and N. V. L. Brokaw. 1996. Forest damage and recovery from catastrophic wind. The Botanical Review 62:114-185.

Ferguson, R. S., and B. Pope. 2001. Inventory and classification of avalanche paths in Tree Farm Licence No. 14, Invermere Forest District, BC. Report for Tembec Industries Inc. FRBC project no - 00-RIP-FRBC-405.

Foster, D. R. 1988. Disturbance history, community organization and vegetation dynamics of the old-growth Pisgah Forest, south-western New Hampshire, USA. Journal of Ecology 76:105-134.

Foster, D. R., and E. R. Boose. 1992. Patterns of forest damage resulting from catastrophic wind in central New England, USA. Journal of Ecology 76:135 - 151.

Friele, P. 1997. Chapter 2: Terrain Analysis. in Kawesas Watershed Assessment. Ecotrust Canada. Online at http://www.inforain.org/kawesas/.

Garbutt, R. 1998. Common pests in successional forests on Vancouver Island. Northwest Science 72:106-109.

Garbutt, R., and J. Vallentgoad. 1992. Forest Insect and Disease Conditions: Prince Rupert Forest Region 1992. FIDS report 93-5, Canadian Forest Service, Victoria, BC.

Gardiner, B. A., G. R. Stacey, R. E. Belcher, and C. J. Wood. 1997. Field and wind tunnel assessments of the implications of respacing and thinning for tree stability. Forestry 70.

Gavin, D. G. 2000. Holocene fire history of a coastal temperate rain forest, Vancouver Island, British Columbia, Canada. Ph.D. Dissertation. University of Washington, Seattle.

Gavin, D. G., L. B. Brubaker, and K. P. Lertzman. 2003a. An 1800-year record of the spatial and temporal distribution of fire from the west coast of Vancouver Island, Canada. Canadian Journal of Forest Research 33:573-586.

Gavin, D. G., L. B. Brubaker, and K. P. Lertzman. 2003b. Holocene Fire history of a coastal temperate rain forest based on soil charcoal radiocarbon dates. Ecology 84:186-201.

Geertsema, M. 2000. Tsunamis in the Pacific Northwest. Online at http://www.emporia.edu/earthsci/student/geert1


Page 54

Geiszler, D. R., R. I. Gara, C. H. Driver, V. F. Gallucci, and R. E. Martin. 1980. Fire, fungi, beetle influences on a lodgepole pine ecosystem of south-central Oregon. Oecologia 46:239-243.

Gibson, C. P. 2001. The general ecology of beavers (Castor spp.), as related to their influence on stream ecosystems and riparian habitats, and the subsequent effects on fish - a review. Reviews in fish biology and fisheries 10:439-461.

Gimbarzevsky, P. 1988. Mass Wasting on the Queen Charlotte Islands: A Regional Inventory. Land Management Report No. 29. BC Ministry of Forests, Victoria, BC.

Gray, R. W., and K. L. MacKenzie. 2001. Mule deer browse species response to thinning and burning in interior Douglas-fir forests in BC. in Fire in temperate, boreal and montane ecosystems, Tall Timbers 22nd Fire Ecology Conference, Kananaskis, AB.

Green, R., and K. Klinka. 1994. A Field Guide to the Site Identification and Interpretation in the Vancouver Forest Region. Land Management Handbook No. 28, Victoria, B.C.

Green, R. N., B. A. Blackwell, K. Klinka, and J. Dobry. 1999. Partial reconstruction of fire history in the Capilano Watershed. Chapter 12 in Greater Vancouver Regional District watershed ecological inventory program: Annex: Analysis reports watershed management plan No. 5.

Haak, R. A., and J. W. Byler. 1993. Insects and pathogens: regulators of forest ecosystems. Journal of Forestry 91:32-37.

Hallet, D. J., D. Lepofsky, R. W. Mathewes, and K. P. Lertzman. 2003. 11 000 years of fire history and climate in the mountain hemlock rain forests of southwestern British Columbia based on sedimentary charcoal. Canadian Journal of Forest Research 33:292-312.

Hansen, R. A., and R. A. Combellick. 1998. Planning scenario earthquakes for southeast Alaska. Miscellaneous Publication 34, State of Alaska, Department of Natural Resources, Division of Geological and Geophysical Surveys.

Harestad, A. 1982. Natural History of Vertebrate Pests in Forests of British Columbia. WWHR-4, BC Ministry of Forests, Research Branch, Victoria, BC.

Harmon, M. E., and J. F. Franklin. 1989. Tree seedlings on logs in Picea-Tsuga forests of Oregon and Washington. Ecology 70:48-59.

Harris, A. S. 1989. Wind in the Forests of Southeast Alaska and Guides for Reducing Damage. General Technical Report PNW-GTR-244, USDA Forest Service.

Hebda, R. J., and R. W. Mathewes. 1984. Holocene history of cedar and native Indian cultures of the North American Pacific coast. Science 225:711-713.

Hennon, P. E. 1995. Are heart rot fungi major factors in disturbance in gap-dynamic forests? Northwest Science 69:284-292.

Hennon, P. E., and D. J. DeMars. 1997. Development of woody decay in wounded western hemlock and Sitka spruce in southeast Alaska. Canadian Journal of Forest Research 27:1971-1978.

Hocker, K. M. 1990. Influence of canopy gaps on forest dynamics in southeast Alaskan hemlock-spruce forests. Senior Thesis. Harvard University, Cambridge, MA.

Hogan, D. L., S. A. Bird, and S. Rice. 1994. Stream channel morphology and recovery processes. Pages 77-96 in D. L. Hogan , P. J. Tschapinski, and S. Chatwin, editors. Carnation Creek and Queen Charlotte Islands Fish/Forestry Workshop: Applying 20 Years of Coast Research to Management Solutions, Queen Charlotte City, BC.


Page 55

Hogan, D. L., and J. W. Schwab. 1991. Meteorological Conditions Associated with Hillslope Failures on the Queen Charlotte Islands. Land Management Report No. 73, BC Ministry of Forests, Victoria, BC.

Hoggett, A. 2000. Western Hemlock Looper and Forest Disturbance in the ICH wk3 of the Robson Valley. Stage 2: The Effects of Western Hemlock Looper -Report and Silviculture Recommendations. Unpublished report Faculty of Foresty, University of British Columbia.

Holland, S. S. 1965. Landforms of British Columbia, a Physiographic Outline. Bulletin 48 Department of Mines and Petroleum Resources.

Huff, M. H. 1995. Forest age structure and development following wildfires in the western Olympic Mountains, Washington. Ecological Applications 5:471-483.

Hyndman, R.D., G.C. Rogers, H. Dragert, K. Wang, D. Oleskevich, J. Henton, J.J. Clague, J. Adams and P.T. Boborowsky. 2001. Giant Earthquakes Beneath Canada's West Coast. Natural Resources Canada, Geological Survey of Canada - Sidney Subdivision. Online at http://www.pgc.nrcan.gc.ca/geodyn/mega1.htm.

Jacoby, G. C., Williams, P. L. and Buckley, B. M. 1992. Tree ring correlation between prehistoric landslides and abrupt tectonic events in Seattle, Washington. Science 258:1621 - 1623.

Jakob, M. 2000. The impacts of logging on landslide activity in Clayoquot Sound, British Columbia. Catena 38:279-300.

Johnson, E. A. 1987. The relativel importance of snow avalanche disturbance and thinning on canopy plant populations. Ecology 68:43-53.

Johnson Gottesfeld, L. M. 1994. Aboriginal burning for vegetation management in northwest British Columbia. Human Ecology 22:171-188.

Jordan, P. 2001. Regional incidence of landslides. Pages 237-247 in Watershed assessment in the southern Interior of British Columbia: Workshop Proceedings, Penticton, BC.

Keefer, D. K. 1984. Landslides caused by earthquakes. Geological Society of America Bulletin 95:406-421.

Keenan, R. J. 1993. Structure and Function of Wester Redcedar and Western Hhemlock Forests on Nnorthern Vancouver Island. Ph.D. Thesis, University of British Columbia, Vancouver, BC.

Kimmey, J. W. 1956. Cull factors for Sitka spruce, western hemlock, and western redcedar in southeast Alaska. Forest Research Center. Station Paper No. 6, US Forest Service, Alaska.

Klein, D. R. 1965. Ecology of Deer Range in Alaska. Ecological Monographs 35:259-284.

Koot, H. P. 1991. Spruce Aphid. Forest Insect and Disease Survey, Forest Pest Leaflet 16, Forestry Canada.

Koot, H. P. 1994. Western Hemlock Looper. Forest Pest Leaflet 21, Pacific Forestry Centre, Victoria.

Koot, P. 1993. Green-striped Forest Looper. Forest Pest Leaflet 22, Pacific Forestry Centre, Victoria.

Kramer, M. G. 1997. Abiotic controls on windthrow and forest dynamics in a coastal temperate rainforest, Kuiu Island, southeast Alaska. M.Sc. Montana State University, Bozeman, MT.

Kramer, M. G., A. J. Hansen, and M. L. Taper. 2001. Abiotic controls on long-term windthrow disturbance and temperate rain forest dynamics in southeast Alaska. Ecology 82:2749 - 2768.

Lertzman, K., D. Gavin, D. Hallet, L. Brubaker, D. Lepofsky, and R. Mathewes. 2002. Long-term fire regime estimated from soil charcoal in coastal temperate rainforests. Conservation Ecology 6:5 (Online at http://www.consecol.org/vol6/iss2/art5).


Page 56

Lertzman, K., and C. J. Krebs. 1991. Gap-phase of a subalpine old-growth forest. Canadian Journal of Forest Research 21:173-1741.

Lertzman, K., T. Spies, and F. Swanson. 1996. Disturbances, forest dynamics and ecosystem management in temperate rain forests. in Schoonmaker et. al., editors. Environments and people of the coastal temperate rainforest. Island Press, Washington, DC.

Lertzman, K. P., G. D. Sutherland, A. Inselberg, and S. C. Saunders. 1996. Canopy gaps and the landscape mosaic in a coastal temperate rain forest. Ecology 77:1254-1270.

Lertzman, K. P. 1992. Patterns of gap-phase replacement in a subalpine, old-growth forest. Ecology 73:657-669.

Lofroth, E. 1998. The dead wood cycle. Pages 185-214 in J. Voller and S. Harrison, editors. Conservation Biology Principles for Forested Landscapes. UBC Press, Vancouver, BC.

Lotan, J. E., J. K. Brown, and L. F. Neuenschwander. 1984. Role of fire in lodgepole pine forests. Pages 133-152 in D. M. Baumgartner, R. G. Krebill, J. T. Arnott, and G. F. Weetman, editors. Lodgepole Pine: The Species and Its Management. Washington State University, Spokane, WA.

Lovett. 2002. Insect defoliation and nitrogen cycling in forests. BioScience 52.

MacHutchon, A. G., S. Himmer, and C. A. Bryden. 1993. Khutzeymateen Valley Grizzly Bear Study. Wildlife Habitat Research Report No. 31, BC Ministry of Environment, Lands and Parks and Ministry of Forests, Victoria.

MacKinnon, A. 1998. Biodiversity in old-growth forests. Pages 146-184 in J. Voller and S. Harrison, editors. Conservation biology principles for forested landscapes. UBC Press, Vancouver, BC.

Martin, J., and C. Baltzinger. 2002. Interaction among deer browsing, hunting, and tree regeneration. Canadian Journal of Forest Research 32:1254-1264.

Mathews, W. H. 1979. Landslides of central Vancouver Island and the 1946 earthquake. Bulletin of the Seismological Society of America 69:445-450.

McClung, D. M. 2001. Snow Avalanche Activity and Decision Support in Forested and Harvested Terrain. Interim FRBC Report KB97241-0RE1.

McNay, S. 1996. How Black-tailed Deer React to Logging in their Winter Habitat. Coastal Black-tailed Deer Study, Brochure No. 40.

Millard, T. 1999. Debris Flow Initiation in Coastal British Columbia Gullies. Forest Resarch Technical Report TR-2 BC Forest Service, Vancouver Forest Region, Nanaimo, BC.

Mitchell, J. M. 1998. Report Determining the Annual Extent of Natural and Harvest Related Blowdown in the North Coast Forest District. Strathinnes Forestry Consultants Ltd., Victoria, BC.

Mitchell, S. J. 1995. The windthrow triangle: a relative windthrow hazard assessment procedure for forest managers. Forestry Chronicle 71:446-450.

Mitchell, S. J., T. Hailemariam, and Y. Kulis. 2001. Empirical modeling of cutblock edge windthrow risk on Vancouver Island, Canada, using stand level information. Forest Ecology and Management 154:117-130.

Montgommery, D. R. 1999. Process domains of the river continuum. Journal of the Water Resources Association 35:1-14.

Moore, M. K. 1977. Factors Contributing to Blowdown in Streamside Leave Strips on Vancouver Island. Land Management Report 3, BC Ministry of Forests, Victoria, BC.


Page 57

Muir, P. S., and J. E. Lotan. 1985. Disturbance history and serotiny of Pinus contorta in western Montana. Ecology 66:1658-1668.

Naiman, R. J., R. E. Bilby, and P. A. Bisson. 2000. Riparian ecology and management in the Pacific Coastal rain forest. BioScience 50:996-1011.

Naiman, R. J., R. E. Bilby, D. E. Schindler, and J. M. Helfield. 2002. Pacific salmon, nutrients, and the dynamics of freshwater and riparian ecosystems. Ecosystems 5:399-417.

Naiman, R. J., J. M. Melillo, and J. E. Hobbie. 1986. Ecosystem alteration of boreal forest streams by beaver (Castor canadensis). Ecology 67:1254-1269.

Nevill, R., N. Humphreys, and A. Van Sickle. 1996. Five-Year Overview of Forest Health Surverys in Young Managed Stands in British Columbia 1991-1995. FRDA report 262, Canadian Forest Service and the BC Ministry of Forests, Victoria ,BC.

NOAA/NWS West Coast/Alaska Tsunami Warning Center. 2002. Past Tsunamis along the U.S. West Coast, British Columbia, and Alaska. Online at http://wcatc.gov.

Nowacki, G., and M. G. Kramer. 1998. The Effects of Wind Disturbance on Temperate Rain Forest Structure and Dynamics of southeast Alaska. Gen. Tech. Rep. PNW-GTR-421, USDA Forest Service, Portland, OR.

Oliver, C. D., A. Osawa, and A. Camp. 1998. Forest dynamics and resulting animal an plant population chages at the stand and landscape levels. Journal of Sustainable Forestry 6:281-312.

Ott, R. A. 1997. Natural Disturbance at the Site and Landscape Levels in Temperate Rainforests of Southeast Alaska. Ph. D. University of Alaska, Fairbanks, Fairbanks, AK.

Ott, R. A., and G. P. Juday. 2002. Canopy gap characteristics and their implications for management in the temperate rainforests of southeast Alaska. Forest Ecology and Management 159:271-291.

Parminter, J. 1990. Fire history and effects on vegetation in three biogeoclimatic zones of British Columbia. Pages 263-277 in Fire and the Environment: Ecological and Cultural Perspectives. Southeastern forest experiment station, Knoxville, Tenessee.

Parminter, J. 1995. Human influence on landscape pattern in the Pacific Region: impacts of burning by First Nations and early European settlers. Paper presented at the Landscape Ecology Symposium, 76th Annual Meeting of the Pacific Division, American Association for the Advancement of Science. Vancouver, B.C., June 20, 1995.

Parminter, J. 1998. Natural disturbance ecology. Pages 3-41 in J. Voller and S. Harrison, editors. Conservation Biology Principles for Forested Landscapes. UBC Press, Vancouver, BC.

Pearson, A. F. 2000. Natural Disturbance Patterns in a Coastal Temperate Rain Forest Watershed, Clayoquot Sound, British Columbia. Ph.D. Dissertation. University of Washington, Seattle, WA.

Pearson, A. F. 2003. Natural and Logging Disturbances in the Temperate Rain Forests of the Central Coast, British Columbia. Draft Report.

Pearson, W. J. 1968. A Review and Analysis of the Fire History of the Queen Charlotte Islands. Unpublished report submitted to the Association of British Columbia Professional Foresters, Vancouver BC.

Person, D. K., M. Kirchhoff, V. V. Ballenberghe, G. C. Iverson, and E. Grossman. 1996. The Alexander Archipelago wolf: a conservation assessment. Gen. Tech. Rep. PNW-GTR-384, U.S Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, Oregon.

Pojar, J. 2003. Changes in vegetation of Haida Gwaii in historical time. Draft Manuscript.


Page 58

Pojar, J., and A. Banner. 1984. Old-growth forests and introduced black-tailed deer on the Queen Charlotte Islands, British Columbia. Pages 247-257 in W. R. Meehan, T. R. J. Merell, and T. A. Hanley, editors. Fish and wildlife relationships in old-growth forests: Proceedings of a symposium held in Juneau, Alaska, 12-15 April 1982. American Institute of Fishery Research Biologists, Juneau, Alaska.

Pojar, J., C. Rowan, A. MacKinnon, D. Coates, and P. LePage. 1999. Silvicultural Options in the Central Coast. British Columbia Ministry of Forests, Victoria, BC.

Pollard, B. T. 2001. Moose Winter Range Mapping for the Prince Rupert Forest District. Report prepared for the North Coast LRMP.

Pollock, M. 1997. Chapter 4: Aquatic habitat and salmonids. in Kawesas Watershed Assessment. Ecotrust Canada, online at http://www.inforain.org/kawesas/.

Price, K., and D. McLennan. 2001. Hydroriparian Ecosystems of the North Coast. Background report prepared for the North Coast LRMP.

Price, K., J. Pojar, A. Roburn, L. Brewer, and N. Poirier. 1998. Windthrow or clearcut - what's the difference. Northwest Science 72:30-33.

Prince Rupert Protected Areas Team. 2001a. Hecate Lowlands Ecosection. Prince Rupert, BC.

Prince Rupert Protected Areas Team. 2001b. Kitimat Ranges Ecosection. Prince Rupert, BC.

Raupach, M. R. 1989. Stand overstorey processes. Philosophical Transactions of the Royal Society of London, Series B 324:175 - 190.

RGIS (Research Group on Introduced Species). 2002. Lessons Form The Islands: Introduced Species and What They Tell Us About How Ecosystems Work. Summary of the RGIS Conference held in Queen Charlotte City and Skidegate in October 2002. Research Group on Introduced Species, Queen Charlotte City, BC.

Rollerson, T., T. Millard, C. Jones, K. Trainor, and B. Thomson. 2001. Predicting Post-Logging Landslide Activity Using Terrain Attributes: Coast Mountains, British Columbia. Forest Resarch Technical Report TR-11, BC Forest Service, Vancouver Forest Region, Nanaimo, BC.

Rollerson, T., T. Millard, and B. Thomson. 2002. Using Terrain Attributes to Predict Post-Logging Landslide Likelihood on Southwestern Vancouver Island. Forest Resarch Technical Report TR-15 BC Forest Service, Vancouver Forest Region, Nanaimo, BC.

Rollerson, T., B. Thompson, and T. Millard. 1997. Identification of Coastal British Columbia Terrain Susceptible to Debris Flows. in First International Conference on Debris Flow Hazards Mitigation: Mechanics, Prediction and Assessment. American Society of Civil Engineers, San Francisco, California.

Rollerson, T. P. 1992. Relationships Between Landscape Attributes and Landslide Frequencies After Logging: Skidegate Plateau, Queen Charlotte Islands. Land Management Report 76, British Columbia Ministry of Forests, Victoria, BC.

Rood, K. M. 1990. Site Characteristics and Landsliding in Forested and Clearcut Terrain, Queen Charlotte Islands, B.C. Land Management Report 64, BC Ministry of Forests, Victoria, BC.

Salmon, D. K. 1997. Oceanography of the eastern North Pacific. Pages 7-22 in P. K., Schoonmaker, B. von Hagen, and E. C. Wolf, editors. The Rain Forests of Home: Profile of a North American Bioregion. Island Press, Washington, DC.


Page 59

Schoonmaker, P. K., and J. Poerll. 1997. Chapter 3: Vegetation: distribution, characteristics, and dynamics. in Kawesas Watershed Assessment. Ecotrust Canada, online at http:/www.inforain.org/kawesas/.

Schuster, R. L., Logan, R. L. and Pringle, P. T. 1992. Prehistoric rock avalanches in the Olympic Mountains, Washington. Science 258:1620 - 1621.

Schwab, J. W. 1983. Mass wasting: October-November 1978 storm, Renell Sound, Queen Charlotte Islands, British Columbia. Research Note 91, BC Ministry of Forests, Victoria, BC.

Schwab, J. W. 1998a. Debris-avalanche flows on British Columbia's North Coast. Pages 259-260 in D. L. Hogan , P. J. Tschapinski, and S. Chatwin, editors. Carnation Creek and Queen Charlotte Islands Fish/Forestry Workshop: Applying 20 Years of Coast Research to Management Solutions, Queen Charlotte City, BC.

Schwab, J. W. 1998b. Landslides on the Queen Charlotte Islands: Processes, rates and climatic events. Pages 41-48 in D. L. Hogan , P. J. Tschapinski, and S. Chatwin, editors. Carnation Creek and Queen Charlotte Islands Fish/Forestry Workshop: Applying 20 Years of Coast Research to Management Solutions. British Columbia Ministry of Forests Land Management Handbook 41, Queen Charlotte City, BC.

Schwab, J. W. 1999. Tsunamis on Troitsa Lake, British Columbia. Extension Note 35, Forest Sciences, Prince Rupert Region.

Scott, R., and B. Beasley. 2002. Variable Retention Windthrow Monitoring Report: A Survey of Eleven Variable Retention Cutblocks in the Long Beach Model Forest, B.C. Long Beach Model Forest Society.

Septer, D., and J. Schwab. 1995. Rainstorm and flood damage: Northwest British Columbia 1891-1991. Land Management Handbook 31, BC Ministry of Forests, Victoria, BC.

Shaw III, C. G. 1982. Development of dwarf mistletoe in western hemlock regeneration in southeast Alaska. Canadian Journal of Forest Research 12:482-488.

Shepherd, R. F. 1994. Management strategies for forest insect defoliators in British Columbia. Forest Ecology and Management 68:303-324.

Shore, T. L., L. Safranyik, and J. P. Lemieux. 2000. Susceptibility of lodgepole pine stands to the mountain pine beetle: testing of a rating system. Canadian Journal of Forest Research 30:44-49.

Sierra Legal Defense Fund. 1997a. Going Downhill Fast: Landslides and the Forest Practices Code. Online at http://www.sierralegal.org/.

Sierra Legal Defense Fund. 1997b. Stream Protection Under the Code: The Destruction Continues. Online at http://www.sierralegal.org/.

Sinton, D. S. 1996. Spatial and temporal patterns of windthrow in the Bull Run Watershed, Oregon. Ph.D. Oregon State University, Portland, OR.

Smith, G. W. 1982. Habitat use by porcupines in a ponderosa pine/Douglas-fir forest in northeastern Oregon. Northwest Science 56:236-240.

Smith, R. B. 1977. Overstory spread and intensification of hemlock dwarf mistletoe. Canadian Journal of Forest Research 7:632-640.

Smith, R. B., P. R. Commandeur, and M. W. Ryan. 1984. Vegetative succession, soil development and forest productivity on landslides, Queen Charlotte Islands, British Columbia, Canada. Pages 109-116 in Proceedings of the IUFRO Symposium on effects of forest land use on erosion and slope stability, Honolulu, Hawaii.


Page 60

Smith, R. B., P. R. Commandeur, and M. W. Ryan. 1986. Soils, Vegetation, and Forest Growth on Landslides and Surrounding Logged and Old-growth Areas on the Queen Charlotte Islands. Land Management Report 41, BC Minstry of Forests, Victoria, BC.

Sokolowski, T. J. 2000. The great Alaskan eatrthquake and tsunamis of 1964. West Coast & Alaska Tsunami Warning Center, Palmer, Alaska.

Stathers, R. J., T. P. Rollerson, and S. J. Mitchell. 1994. Windthrow Handbook for British Columbia. BC Ministry of Forests, Victoria, BC.

Stuart, J. D., J. K. Agee, and R. I. Gara. 1989. Lodgepole pine regeneration in an old, self-perpetuating forest in south central Oregon. Canadian Journal of Forest Research 19:1096-1104.

Sullivan, T. P., and C. Cheng. 1989. Impact of Porcupine Feeding Damage on Intensive Forest Management in the Prince Rupert Forest Region: Problem Analysis. FRDA Report 83, FRDA.

Sullivan, T. P., W. T. Jackson, J. Pojar, and A. Banner. 1986. Impact of feeding damage by the porcupine on western hemlock - Sitka spruce forests of north-coastal British Columbia. Canadian Journal of Forest Research 16:642-647.

Swanson, F. J., J. A. Jones, D. O. Wallin, and J. H. Cissel. 1993. Natural Variability - Implications for Ecosystem Management. General Technical Report PNW-GTR-318, USDA Forest Service, Pacific Northwest Research Station, Portland, OR.

Swanston, D. F., T. K. Kratz, N. Caine, and R. G. Woodmansee. 1988. Landform effects on ecosystem patterns and processes. BioScience 38:92-98.

Swanston, D. N. 1967. Debris Avalanching in Thin Soils derived From Bedrock. Research Note PNW-64, USDA Forest Service, Pacific Northwest Forest and Range Experiment Station, Portland, OR.

Swanston, D. N., C. G. Shaw III, W. P. Smith, G. A. Cellier, and F. H. Everest. 1996. Scientific Information and the Tongass Land Management Plan: Key Findings From the Scientific Literature, Species Assessments, Resource Analyses, Workshops, and Risk Assessment Panels. General Technical Report PNW-GTR-386, USDA Forest Service.

Tamblyn, G. C., and H. Horn. 2001. Current Conditions Report: North Coast Land and Resource Management Plan. Report prepared for the Prince Rupert Interagency Management Committee, Prince Rupert, BC.

Taylor, E. M., and W. G. Taylor, editors. 1997. Responding to Global Climate Change in British Columbia and Yukon. Volume I of the Canada Country Study: Climate Impacts and Adaptation. BC Ministry of Environment, Lands and Parks, Victoria, BC.

Taylor, S., and B. Erickson. 2003. Historical Mountain Pine Beetle Activity. Online at http://www.pfc.forestry.ca/entomology/mpb/historical/index_e.html.

Thomson, A. J., R. B. Smith, and R. I. Alfaro. 1984. Growth patterns in immature and mature western hemlock stands infected with dwarf mistletoe. Canadian Journal of Forest Research 14:518-522.

Trainor, K. 2001. Geomorphological/Hydrological Assessment of the Central Coast Plan Area. Technical report prepared for the Joint Solutions Project reporting to the Central Coast Land and Resource Management Plan.

Tripp, D. 1994. Evolution of fish habitat structure and diversity at log jams in logged and unlogged streams subject to mass wasting. Pages 97-108 in D. L. Hogan , P. J. Tschapinski, and S. Chatwin, editors. Carnation Creek and Queen Charlotte Islands Fish/Forestry Workshop: Applying 20 Years of Coast Research to Management Solutions, Queen Charlotte City, BC.


Page 61

Trummer, L. M., P. E. Hennon, E. M. Hansen, and P. S. Muir. 2000. Modeling the incidence and severity of hemlock dwarf mistletoe in 110-year-old wind-disturbed forests in Southeast Alaska. Canadian Journal of Forest Research 28:1501-1508.

Turner, M. G., W. H. Romme, and R. H. Gardner. 1999. Prefire heterogeneity, fire severity, and early postfire plant reestablishment in subalpine forests of Yellowstone National Park, Wyoming. International Journal of Wildland Fire 9:59-77.

Turner, N. J. 1991. Burning mountainsides for better crops: aboriginal burning in British Columbia. Archaeology in Montana 32:57-93.

Turnquist, R., R. Garbutt, and V. Nealis. 2000. Report on forest pest conditions, Queen Charlotte Islands, British Columbia. Prepared for the South Moresby Forest Replacement Account.

Ugolini, F. C., and D. H. Mann. 1979. Biopedological origin of peatlands in Southeast Alaska. Nature 281.

Ulanova, N. 2000. The effects of windthrow on forests at different spatial scales: a review. Forest Ecology and Management 135:155-167.

van der Kamp, B. J. 1975. The distribution of microorganisms associated with decay in western redcedar. Canadian Journal of Forest Research 5:61-67.

van der Kamp, B. J. 1991. Pathogens as agents of diversity in forested landscapes. Forestry Chronicle 67:353-354.

Veblen, T. T., and P. B. Alaback. 1996. A comparative review of forest dynamics and disturbance in the temperate rainforests of North and South America. Pages 173-213 in R. G. Lawford, P. B. Alaback, and E. Fuentes, editors. High-latitude Rainforests and Associated Ecosystems of the West Coast of the Americas. Springer-Verlag, New York.

Vila, B., G. Vourc'h, D. Gillon, J. Martin, and F. Guibal. 2002. Is escaping deer browse just a matter of time in Picea sitchensis? A chemical and dendroecological approach. Draft manuscript.

Vourc'h, G., and J. Martin. 2001. Defensive adaptations of Thuja plicata to ungulate browsing: a comparative study between mainland and island populations. Oecologia 126:84-93.

Wallis, G. W., D. J. Morrison, and D. W. Ross. 1980. Tree hazards in recreation sites in British Columbia: management guidelines. Joint Report 13, Ministry of Lands, Parks, and Housing; Canadian Forest Service, Victoria, BC.

Walsh, S. J., D. R. Butler, T. R. Allen, and G. P. Malanson. 1994. Influence of snow patterns and snow avalanches on the alpine treeline ecotone. Journal of Vegetation Science 5:657-672.

Walsh, S. J., D. R. Butler, D. G. Brown, and L. Bian. 1990. Cartographic modeling of snow avalanche path location within Glacier National Park, Montana. Photogrammetric Engineering & Remote Sensing 56:615-621.

Weir, P. 2000. Handbook for Forest Management on Snow Avalanche Prone Terrain. Draft report to the Ministry of Forests, Research Branch, Victoria. B.C.

Williams, D. W., R. P. Long, P. M. Wargo, and A. M. Liebhold. 2000. Effects of climate change on forest insect and disease outbreaks. Ecological studies: analysis and synthesis 139:455-494.

Wong, C., H. Sandmann, and B. Dorner. 2002. Estimating Historical Variability of Natural Disturbances in B.C. Part 1: Literature Review of Current Knowledge and Assessment of Knowledge Gaps. Report prepared for the BC Ministry of Forests, Victoria, BC.


Page 62

Wood, C. S., and I. Unger. 1996. Mountain Pine Beetle: A history of Outbreaks in Pine Forests in British Columbia, 1910 to 1995. Natural Resources Canada, Forest Service, Pacific Forest Research Centre, Victoria, BC.

Zeglen, S. 1997. Tree wounding and Partial-cut Harvesting. BC Ministry of Forests, Forestry Division Services Branch, Victoria, BC.

Zimmerling, T. N., and C. D. Croft. 2001. Resource selection by porcupines: winter den site location and forage tree choices. WJAF 16:53-57.


Page 63

Appendix 1

Summary Table 1: Gap characteristics for coastal British Columbia and southeast Alaska.

Gapmaker Type Forest Type

(leading tree species) Region % area in gaps

% area in

expanded gapsa

Gap area (m2) (mean and

range)

Number of gapmakers (mean and

range) %Snapped % Root-

thrown

% Dead standing or leaning

Source

T. heterophylla SE Alaska 8.7 27.4 63 (6 – 264) 2 (1 – 7) 75.9 16.5 7.6 (Ott and Juday 2002) (n = 3) T. heterophylla – P. sitchensis SE Alaska 3.8 1,162ft2(269-2625) 2.4 (1 – 5) 87 13 (Hocker 1990) T. heterophylla – P. sitchensis SE Alaska 7.4 18.2 (Ott 1997) T. heterophylla – P. sitchensis – C. nootkatensis SE Alaska 33.7 (Ott 1997)

T. heterophylla – C. nootkatensis SE Alaska 23.1 (Ott 1997) T. heterophylla (CWHvm1) Clayoquot Sound 14.9 33.3 (Lertzman et al. 1996) (n = 34) T. heterophylla (CWHvm2) Clayoquot Sound 16.0 43.5 (Lertzman et al. 1996) (n = 7) T. mertensiana (MHmm1) Clayoquot Sound 5.8 15.0

5.0 3.7 3.7

(1 – 19) 42.6 24.6 33.0 (Lertzman et al. 1996) (n = 3)

T. heterophylla – A. amabilis Clayoquot Sound 37 38 3.4 - 17 (1 - 17) 60.3 8.3 9.6 (Pearson 2000) (n = 3) T. plicata - T. heterophylla Clayoquot Sound 25 38 3.6 - 5.5 (1 - 8) 22 47 10 (Pearson 2000) (n = 5)

T. mertensiana (MH) Southern Coast Mountains 18 71 77 (5 – 535) 5.7 (1 – 16) 31 13 55 (Lertzman and Krebs 1991)

(n = 4) ________________________________________ a Includes area in the primary gap.


Page 64

Summary Table 2: Landslide characteristics for coastal British Columbia.

Landslide Studies Haida Gwaii and North

and Central Coast

Skidegate Plateau, Haida Gwaiia

(Rollerson 1992)

Haida Gwaii (Gimbarzevsky 1988) (logged and unlogged

terrain)

Rennell Sound, Haida Gwaii (Schwab 1983) in (Schwab

1998b) (logged and unlogged terrain)

Haida Gwaii (Rood 1990)

(logged and unlogged terrain) Central Coastd (Pearson 2003)

Central Coast (Trainor 2001)

(logged and unlogged terrain)

Total area studied 3380 ha All islands 150 km2 27 basins 350 km2 1,500,000 ha (573,068 ha forested) 140,000 ha sampled

Number of disturbance patches 1006b? 8240 264 1337 926 38,766

Total area disturbed (%) Moderate (1-3/km2): 24%

Severe (4-7/km2): 7% Extreme (>8/km2): 1%

4.3% of clearcut terrain (126) 1.9% of road (25)

0.1% of forest (112) 1.4% of forested area

Proportion associated with roads or cutblocks 43 times more in clearcuts

17 times more from roads

Minimum patch size mapped (or data source) 500 m2 1:50,000 air photos Forest cover polygons

Reference period 15 yrs 40? yrs 1 storm in 1978 20? years 140 years Patch size range 0.01 x 10-4 - 255.3 Mean patch size 10.2 (+ 18.9 stdev) Median patch size 4.7

Density 17/km2 2.6/km2 average, but up to 18/km2 in some areas

1.76/km2 15 times more in roads and

clearcuts

3.82/km2 (logged and unlogged)

34 times more in logged terrain

Outer Coast Mountains: 3.22c

Inner Coast Mountains: 1.50c

Hecate Lowlands: 1.41c Frequency 1.1/yr/km2 0.065 /yr/km2 N/a 0.19/yr/km2(logged and logged)

More likely sites

Steep slopes, headwater basins and stream escarpments,

dissected terrains, gullies

Steep slopes, upper slope area, areas with

high precipitation Steep slopes, gullies, concave

or complex slopes debris slides more likely

in Hecate Lowland, debris flows more likely

in Coast Mountains

________________________________________ a Only slides associated with roads and clearcuts. b Number of polygons with landslides > 500 m2. c No units given in report. d Includes avalanche tracks.


Page 65

Landslide Studies Vancouver Island and

southern mainland coast

Clayoquot Valley (Pearson 2000)

(unlogged terrain)

Clayoquot Sound (Jakob 2000)

(logged and unlogged terrain)

West Coast of Vancouver Islanda

(Rollerson et al. 1997) in (Rollerson et al. 2001)

Southwestern Vancouver Islanda

(Rollerson et al. 2002) Coast Mountainsa

(Rollerson et al. 2001)

Total area studied 7608 ha 2546 km2 3491 ha 8490 ha Number of disturbance patches 49 1004 470b 127b Total area disturbed 183 ha 2205 ha

Total area disturbed (%) 2.4% Logged: 3% Unlogged: 0.37%

Proportion associated with roads or cutblocks

Minimum patch size mapped 500m2 500 m2 500 m2 500 m2

Reference period 1900?-1988 Logged: 20 yrs Unlogged: 40-60 yrs. 15 yrs 15 yrs 15 yrs

Patch size range 0.2-29 ha

Mean patch size 3.7 + 5.2 ha Logged: 1.59 + 1.62 ha Unlogged: 1.67 + 1.86 ha

Median patch size Logged: 1 ha Unlogged: 1 ha

Density

(0.64/km2)

0.39/km2 (forested)

0.12- 0.85/km2

Logged: 1.94/km2 Unlogged: 0.22/km2

8/km2 6 /km2 0.8-1.2/km2

Frequency 0.0073/yr/km2 Logged: 0.0097 yr/ km2 Unlogged: 0.0037-0.0055/yr/ km2 0.5yr/km2 0.05 – 0.08 /yr/km2

More likely sites Steep slopes, most above 500 m elevation

Steep, concave or straight slopes, high precipitation, E, SE, S slopes Steep slopes, gullies,

dissected terrain Steep slopes, gullies,concave

and complex slopes, areas with high precipitation

________________________________________ a Only slides associated with roads and clearcuts. b Number of polygons with landslides > 500 m2.


Page 66

Summary Table 3: Blowdown characteristics for southeast Alaska and coastal British Columbia.

Wind Studies Prince of Wales and associated islands

SE Alaska (Harris 1989)

Kuiu Island SE Alaska

(Kramer 1997, Kramer et al. 2001)

SE Chichagof Island

SE Alaska (Nowacki and Kramer 1998)

NE Chichagof Island

SE Alaska (Nowacki and Kramer 1998)

Clayoquot Sound Vancouver Island (Pearson 2000)

1906 Vancouver Island Windstorm

(Keenan 1993, Pearson 2003)

Central Coast (Pearson 2003)

West Coast of Haida Gwaii

(Pearson 2003)

North Coast Forest District

(based on data from Mitchell 1998)

Total area studied (ha) 870,000 197,000 110,000 104,000 48,000 78,766 (54,398 forested)

1,500,000 (573,068 forested) 232,766 119,130

(Proportion of) area in productive forest 59.5% 65% 54.9% 64.1% 222,372 119,130

Number of blowdown patches 1010 1886 1118 444 0 208 162 62

Total area disturbed (ha) 7414.8 26,588 4737.6 6196.8 0 2,569 1,750 5,586 1531

Total area disturbed (%) 0.9% 13.5% 4.3% 6% 0% 3.2% (4.7% forested) 0.1% (0.3% forested) 2.4% 1.3%

Productive forest disturbed (%) 1.6% 20.8% 7.9% 9.3% 0% 2.5% 1.3%

Proportion associated with/adjacent to clearcuts 13.9% N/a 8%

Minimum patch size mapped (ha) (or data source) 0.8 0.8 0.4 0.8 4.5 0.2 Forest cover

polygons Forest cover polygons

Reference period 15+ years 150 years ~ 140 years N/a 140 years 37 years

Includes partial blowdown? yes

(up to 90% of trees left standing)

yes yes yes yes yes yes

(up to 70% of trees left standing)

Patch size range (ha) 0.8 – 70 0.8 – 400 0.4 – 100 0.8 – 308 N/a 0.2 - 508 0.001 - 157.8 0.4 - 68 0.8 - 118

Mean patch size (ha) 7.2 15.6 4 14 N/a 14 9.7 (+ 19.6 stdev) 7.5 (+ 8.1 stdev) 10.8

Median patch size (ha) 5.6 2 7.6 N/a 4 3.4 4.8 6.2

Interval between major (stand-replacing or partially stand-replacing) events

50 – 300 years in

exposed locations, gap-phase dynamics

elsewhere N/a N/a 4400 years average

for study area > 3000 yrs average

for NC area (see Dorner and Wong

2002)

Predominant vegetation type hemlock / spruce hemlock hemlock N/a hemlock / amabilis fir

Predominant topographic position windward slopes, flats and sideslopes parallel to predominant storm

direction

ridge noses and flanks, exposeda stands N/a

Predominant soil characteristics shallow and/or highly productive nonhydric nonhydric N/a

________________________ a Determined using EXPOSE simulation model; elevation and slope were also significant factors.


Page 67

Summary Table 4: Snow avalanche characteristics for western Canada.

Avalanche Studies Clayoquot Valley (Pearson 2000)

Invermere TFL 14 (Ferguson and Pope 2001)

Glacier Nat. Park (Walsh et al. 1990)

Kananskis (Johnson 1987)

Total area studied 7000 ha 151,866 ha Number of disturbed patches 14 336 121 2 Total area disturbed 176 ha 14,731 ha Total area disturbed (%) 2.3% 9.7% Minimum patch size mapped 0.5 ha Reference period 1900? -1988 Patch size range 2.6-34 ha 2-885 ha Mean patch size 12.5 + 8.9 ha 44 ha

Interval between events < 10 years in a path < 20 years in non treed zone of path, < 130 years in treed zone

More likely areas Elevations > 500 m Slopes > 30-50o and/or on those likely to experience snowpack warming such as on south aspects (Weir 2000)

Steep (30-45o), open slopes or gullies Smooth or unvegetated terrain

More frequent in top than bottom of path


Page 68

Summary Table 5: Wildfire characteristics for coastal British Columbia (see also Table 1).

Fires Fraser Valley (MH)

soil charcoal (Lertzman et al. 2002,

Hallet et al. 2003)

Fraser Valley (MH) lake sedimenta (2 sites)

(Hallet et al. 2003)

Clayoquot Valley soil charcoal

(Lertzman et al. 2002, Gavin et al. 2003a, Gavin et al. 2003b)

Clayoquot Valley lake sedimenta

(Gavin et al. 2003b) Central Coast

(Pearson 2003) North Coast

(Dorner and Wong 2002)

Total area studied 730 ha 110 ha 1,500,000 ha (573,068 ha

forested) 1.76 million ha

Number of disturbed patches 100 45 Total area disturbed 8.5 ha Proportion associated with human ignitions 0% 83%

Minimum patch size mapped (or data source) Forest cover

polygons MOF Protection

Branch Database

Reference period ~ 9,000 yrs BP to present 11,000 yrs BP to present

tree dates: AD 1550-1886 radiocarbon dates: 280-10,330

BP AD 200-present 140 years 1950-1997

Patch size range 0.001 – 1,375.3 ha 0-4 ha

Mean patch size Rarely extended beyond 250 m in diameter 74.9 ha

(+ 183.2 ha stdev) 0.19 ha

(+ 0.59 ha stddev) Median patch size 18.7 ha

Interval between events rangeb: 300 – 7,000 yrs medianb: 1,200 yrs

161 and 204 yrs av. resp. for entire period; 174 and 271 yrs av. resp. over the last

1,500 yrs

rangeb: 450 – 8,760 yrs medianb: 2,380 yrs

fire interval for lake catchment:

AD 200-900: 50 yrs AD 1100 - present: 350 yrs

greater 1000 yrs on

average for the study area

More likely areas 25 times more likely on south facing slopes, more likely on well drained Cw/Hw slopes

___________________________________________ a Lake sediments capture evidence of fires throughout the entire catchment area of the basin. Thus, fire statistics refer to occurrence of fires somewhere in the watershed, and

intervals do not necessarily indicate expected return times at any fixed location. Estimated catchment area for the Clayoquot Sound study is extends 500 m from the lake shore (Gavin et al. 2003a).

b Because of the low temporal resolution of the charcoal record only intervals ≥ 300 years could be detected.

natural disturbance dynamics in coastal british columbiab. dorner and c. wong, natural disturbance...

Documents