Canopy Macrolichen Distribution In A Very Wet Oldgrowth Forest Landscape Of The Upper Fraser River Watershed David N. Radies Thesis Submitted In Partial Fulfillment Of The Requirements For The Degree Of Master Of Science in Natural Resources And Environmental Studies (Biology) The University Of Northern British Columbia June 2008 © David N. Radies 1*1 Library and Archives Canada Bibliotheque et Archives Canada Published Heritage Branch Direction du Patrimoine de I'edition 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue Wellington Ottawa ON K1A0N4 Canada Your file Votre reference ISBN: 978-0-494-48766-2 Our file Notre reference ISBN: 978-0-494-48766-2 NOTICE: The author has granted a nonexclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or noncommercial purposes, in microform, paper, electronic and/or any other formats. AVIS: L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par Plntemet, prefer, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conformement a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these. While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada ABSTRACT Windward slopes of the inland mountain ranges in British Columbia support a unique temperate rainforest ecosystem. Continued fragmentation and loss of old-growth forests in this globally rare ecosystem, has led to calls for the identification of conservation priorities between remaining stands. This thesis addresses this concern by surveying the relative abundances of 37 canopy macrolichens over a 70-km2 area of remaining oldgrowth (>140 years) forest in the upper Fraser River watershed, British Columbia, Canada. To ensure adequate representation of landscape-scale old-growth forest characteristics, we divided study plots equally among leading tree species and between broadly defined sites of "wet" and "dry" relative soil moisture. Other variables included: minimum mean annual temperature, mean annual precipitation, solar loading, and canopy openness. This thesis integrates two statistical techniques: Nonmetric Multidimensional Scaling ordination for analysis of lichen assemblages and logistic regression to evaluate the habitat conditions of a subset of 8 lichen species previously identified as "old-growth associated". Ordination suggested that community assemblages were greatly influenced by both the presence and abundance of bipartite cyanolichens. These communities correlated well with increasing levels of relative soil moisture, temperature, precipitation, and canopy openness, with little to no significant effect of tree leading species. Logistic regression models identified relative soil moisture and temperature in all parsimonious models. Leading tree species, in combination with moisture and temperature, were important factors explaining the presence or absence of 5 of 8 modeled lichen species. ii The results of this thesis emphasize the importance of maintaining representative areas of old-growth forests that are potentially less prone to natural disturbances such as fire. Of concern to the maintenance of lichen populations in old-growth inland temperate rainforests is the continued forest harvesting of low-elevation water-receiving sites. It is recommended that the conservation of these wet topographic positioned areas be identified spatially to meet remaining provincially set old-growth threshold targets for the purpose of maintaining biological diversity and ecological integrity. in TABLE OF CONTENTS Abstact ii Table of Contents iv-v List of Tables vi-vii List of Figures vii-x Acknowledgments Xi Preface xii Introduction 1 Chapter One: Not all old-growth is equal: Ecological attributes and lichen 3 biodiversity in an inland temperate rainforest landscape Introduction 3 Methods 7 Results 13 Discussion and conclusion 15 Chapter Two: Parallels between old-growth forest retention targets and the 23 children's game Kerplunk! Introduction 23 Methods 26 Results 31 Discussion and conclusion 32 Works Cited 39 Tables 49 IV Figures 62 v LIST OF TABLES Table 1-1. Macrolichen presence/absence by number of sites containing each species and abundance distribution (in brackets) in the upper Fraser River watershed. Abundances denote frequency of occurrence of each species based on a three point scale: very low equals 1-3 thalli, low equals 3-10 thalli, and common equals more than 10 thalli per site. One exception, is L. pulmonaria*, which was measured by number of handful sizes (10 x 20 cm2) as opposed to thalli measurements. Table 1-2. Multiple linear regression estimates for the log-transformed environmental variables temperature, precipitation (mean minimum temperature, March to October), solar loading, canopy openness, relative soil moisture, and basal area of cedar (Cw), hemlock (Hw), spruce (Sx), and fir (Sf) calculated against Axis 1 (R2 = 0.548, f-ratio = 5.529, P < 0.001) and Axis 2 (R2 = 0.656, f-ratio = 8.701, P < 0.001) ordination scores (n = 51) Table 1-3. Mean and standard deviation of stand variables. N = 27 and 26 respectively for dry and wet broad relative soil moisture (BRSM) stands in the upper Fraser River watershed. Table 1-4. Predicted best model sets (> 0.1 AICcw) for presence of the old-forest associated macrolichens and receiver operating characteristic (ROC) results for best models in the upper Fraser River watershed. Abbreviations for model variables are as follows: broad relative soil moisture (BRSM); leading tree species (LEAD); canopy openness (OPEN), mean annual precipitation (PRECIP) and; mean minimum temperature, March - October (TEMP). Abbreviations in brackets —cedar (Cw), hemlock (Hw) and wet BRSM (wet) — represent categorical variables with greatest influence. Table 2-1. List of macrolichen species and abbreviations surveyed for in the upper Fraser River watershed. Table 2-2. Independent variable abbreviations and categorical coding (in brackets) used to predict the probability of occurrence of the 8 'old-growth dependent' lichen species (see Radies et al. 2008) including Cavernularia hultenii, Lobaria retigera, and Nephroma occultum. Table 2-3. 13 models tested for each species in the upper Fraser River watershed. Table 2-4. Site sampling breakdown in the upper Fraser River watershed. Table 2-5. Ecosystem representation as a % of total area of ICHvk2 biogeoclimatic zone in the upper Fraser River watershed of old-growth forests with wet versus those with dry broad soil moisture regime status and the proportion of these old-forest stands that currently falls in the non-timber harvesting landbase (NHLB), and in two types of protected areas: old-growth management areas (OGMA), and provincial parks. ROGC = Remaining Conifer species (consisting mainly of Abies lasiocarpa). Total forested area of ICHvk2 is 130,571 ha. Base year = 2001. VII LIST OF FIGURES Figure 1-1. Landscape distribution of old-growth cedar, hemlock, and spruce-leading forests (separated by both wet and dry broad relative soil moisture conditions) in the very wet cool Interior Cedar Hemlock (ICHvk2) and adjacent very-wet cool Sub-Boreal Spruce (SBSvk) biogeoclimatic zones of the upper Fraser River watershed. Forests north of the Fraser River are part of the Rocky Mountain formation, while forests south of the Fraser River are part of the Columbia Mountain formation. Insert 1 indicates study location in British Columbia, Canada. Insert 2 and 3 identifies old-growth forest type on north and south-facing aspects of Driscoll ridge respectively. Reference points indicate plot-sampling locations. Figure 1-2. Overlay of species abundance in the upper Fraser River watershed in 2005 on stand ordinations for the tripartate macrolichen Lobaria pulmonaria, and the bipartate cyanolichens: L. retigera, Sticta fuliginosa, Nephroma isidiosum, Nephroma occultum, and Pseudocyphellaria anomala Figure 1-3. Coefficients and 95% confidence intervals of independent variables for each of the eight "old-growth associated" lichen species in the upper Fraser River watershed generated using logistic regression. Abbreviations of variables as follows: dry relative soil moisture (dry); wet relative soil moisture (wet); canopy openness (Densio); mean minimum temperature (Temp); Mean annual precipitation (Precip); solar loading (Solar); cedar-leading (cw); hemlock-leading (hw); and spruce-leading (sx). Vlll Figure 2-1. Distribution of Interior Cedar Hemlock wet and very wet biogeoclimatic variants within the Inland Temperate Rainforest (ITR) boundary (Inset). Abbreviations as follows: w = wet; v = very wet; k = cool; c=cold. From Craighead and Cross 2004. Figure 2-2. Clear-cut logging activity in past 50 years in the northern portion of the ITR. ICHwk and vk forests in colour. From Craighead and Cross 2004. Figure 2-3. Distribution of remaining class 8 (140-250 years) and class 9 (250 years or older) old-growth forests in the ICHwk and ICHvk. From Craighead and Cross 2004. Figure 2-4. Clear-cut logging in the past 50 years in and surrounding the ICHvk2 and Sub Boreal Spruce very wet (SBSvk) biogeoclimatic zones. Portion of SBSvk that appears unharvested is Timber Forest License 30, in which logging activity was unavailable at time of writing. From Craighead and Cross 2004. Figure 2-5. Location of study sites. Note that the ICHvk2 has been more recently been divided into the ICHwk4 and ICHvk2 (see Fig. 2-6). Figure 2-6. Remaining coniferous leading forests greater than 140 years and distribution by dominant tree species. "Other species" is mostly leading in Abies Lasiocarpa. Figure 2-7. Sample plots used for data collection: A. Stand Structure Plot (r=17.84m); B. Lichen Plot - area inside the 40m x 100m (0.4ha) rectangle. Locations of densiometer measurements are identified by point locations (square point indicating plot centre). Figure 2-8. Ordination distance matrix, identifying macrolichen species presence and absence in wet and dry Broad Relative Soil Moisture (BSMR) sites by tree leading species. Codes are as follows: Cw = cedar-leading; Hw = Hemlock leading; Sx = Spurce leading. IX Figure 2-9. Predicted distribution of wet and dry BSMR old-growth forests leading in Thujaplicata or Tsuga heterophylla in the ICHvk2. Based on data from 2006. Figure 2-10. Predicted probability of occurrence (very low, low, medium, high, very high) for Cavernularia hultennii in the ICHvk2 biogeoclimatic subzone in the upper Fraser River watershed. See Fig. 2-1 for geographic placement of area. Figure 2-11. Predicted probability of occurrence (very low, low, medium, high, very high) for Lobaria retigera in the ICHvk2 biogeoclimatic subzone in the upper Fraser River watershed. See Fig. 2-1 for geographic placement of area. Figure 2-12. Predicted probability of Nephroma occultum occurrence (very low, low, medium, high, very high) for in the ICHvk2 biogeoclimatic subzone in the upper Fraser River watershed. See Fig. 2-1 for geographic placement of area. x ACKNOWLEDGEMENTS First, I would like to thank all my friends and family for their encouragement, patience and support. I am particularly grateful to: M.E. Gauthier, C. Helenius, J. Kelly, D. Khurana and C. LeBoutier for assistance with fieldwork and; D. Coxson, S. Stevenson, P. Sanborn and C. DeLong as supporting members of my committee. C. Johnson and K. Konwiki for all their technical support in developing the predictive models and maps. T. Goward for assistance with lichen identifications; Baden Cross (Applied Conservation GIS mapping) and the Valhalla Wilderness society for providing the Inland Temperate Rainforest Maps (Figures 2-1 to 2-4). The Canadian Forest Service supplied climate data. The Sustainable Forest Management Network, the B.C. Forest Science Program, Mountain Equipment Coop, and the University of Northern B.C provided funding. XI PREFACE Chapter 1, "Not all old growth is equal: Ecological attributes and lichen biodiversity in an inland temperate rainforest landscape''' is currently in submission as a co-authored manuscript of the same title to the journal "Ecological Applications". The coauthors of this submission (listed in order of contribution) were David N. Radies, Darwyn S. Coxson, Chris J. Johnson, and Ksenia Konwicki. Darwyn Coxson and David Radies jointly wrote the "Ecological Applications" manuscript, with additional contributions from Chris J. Johnson, and Ksenia Konwicki. Logistic regression analysis was conducted by David Radies and Chris Johnson, GIS queries and mapping was conducted by Ksenia Konwicki and David Radies, and ordination analyses and related statistical testing was carried out by Darwyn Coxson. The analysis of the data sets presented in Chapter 2 "Parallels between old-growth forest retention targets and the childrens game Kerpluck!" was respectively conducted by Ksenia Konwicki and David Radies (GIS queries and mapping) and Darwyn Coxson (distance matrix). Writing responsibilities for Chapter 2 reside solely with David Radies. David Radies provided the primary data set on which the analysis conducted for Chapter 1 and 2 above were based, using field data collected under his M.Sc. thesis project. The data sets presented in this thesis were prepared under contract for the Forest Science Program of British Columbia by the University of Northern British Columbia (Darwyn Coxson, Principal Investigator), which retains copyright in these data sets for the purposes of meeting contract deliverables with the British Columbia Forest Science program. INTRODUCTION In British Columbia, Canada, water creates two distinct old-growth mountain forest ecosystems — described as coastal and inland temperate rainforests. These temperate mountain ecosystems trap massive volumes of snow at higher elevations in long winter periods, slowly releasing a migration of ground water beneath coniferous forests from tree line to valley bottom far into the summer months. Importantly, this movement of melt-water, in combination with rainfall, is not equally distributed either between these two mountainous ecoregions or within them. Soil water is orchestrated by a vast array of climatic, topographic and edaphic conditions, which create a broad range of soil moisture conditions that affect the natural disturbance regime and forest conditions at a variety of scales. Continued loss of habitat in primary temperate rainforest in British Columbia is of concern. The provincial government of BC continues to allow clear-cut forest harvesting in both its coastal and inland temperate old-growth rainforests due to the fact that approximately half of old-growth mountain forests will remain in their natural state. However, these remaining old-growth forest stands are primarily in forested regions considered "inoperable". Inoperable old-growth temperate rainforest tend to reside in mountain regions that are considered not economically viable from a forest industry perspective mostly due to geographic positioning (i.e. steep mountain topography) and/or timber quality or in areas restricted from harvest (i.e. provincial parks). Given that it is unlikely that inoperable old-growth will have the same ecological attributes as operable old-growth landscapes, it is reasonable to determine whether retaining inoperable areas of Radies 1 old-growth forest will maintain ecological function and integrity of these globally significant ecosystems. This thesis — as one part of a required body of scientific work needed to address ecological differences between operable and inoperable old-growth forests — inquires as to whether canopy macrolichen diversity will be impacted if timber harvesting in inland temperate old-growth rainforests does not recognize differences among site or stand conditions within the biogeoclimatic subzone scale. To answer this question, the following four objectives were identified: 1. to examine the influence of dominant ecological characteristics (i.e. soil moisture, temperature) and forest composition (i.e. tree leading species, canopy openness) on arboreal foliose macrolichens as communities and as individuals — in particular "old-growth associated" lichens — in the Inland Temperate Rainforests of the upper Fraser River watershed; 2. to compare and contrast the ecological requirements of lichen communities versus individual lichen species; 3. to examine the effectiveness of predicting the presence of "old growth-associated" arboreal lichens using coarse filter landscape attributes applicable to GIS; and 4. to compare and contrast stand and site attributes within and outside the timber harvesting landbase to determine if old-growth forests considered inoperable are capable of maintaining arboreal macrolichen diversity at the landscape scale I have divided this thesis into two main chapters, the first is a multi-authored paper submitted to a scientific journal intended for an international scientific audience and the second is intended for provincial forest managers, biologists, politicians and the general public. It is my hope that this work evokes a deep respect for the complex ecology of old-growth temperate rainforests, of which we know so little. Radies 2 CHAPTER 1: NOT ALL OLD-GROWTH IS EQUAL: ECOLOGICAL ATTRIBUTES AND LICHEN BIODIVERSITY IN AN INLAND TEMPERATE RAINFOREST LANDSCAPE INTRODUCTION Wet-temperate rainforest ecosystems are widely recognized as an important repository of biodiversity, particularly for organisms that live within the forest canopy (Kitching et al. 1993, McCune et al. 2000, Castellon and Sieving 2007). In British Columbia (B.C.), important steps have been taken for conserving large regions of coastal temperate rainforests (Coast Information Team 2004). However, a second major wettemperate rainforest ecosystem is found on the windward slopes of interior mountain ranges. This inland temperate rainforest (ITR) has many unique characteristics, including globally significant assemblages of canopy lichens (Goward 1994, Arsenault and Goward 2000, Goward and Spribille 2005), serving as habitat for endangered populations of mountain caribou (Stevenson et al. 2001), and supporting headwater spawning runs for many of the Fraser River salmon populations (Kew 1992). A major difference between coastal and inland temperate rainforest ecosystems in B.C. is that ITR ecosystems receive approximately half the annual precipitation of the former. Therefore, the development of "rainforest" attributes in the ITR is more dependent on patterns of snowmelt that influence ground moisture conditions. Eng (2000) noted that stands located on cool north facing slopes showed a near 10-fold reduction in stand destroying fire frequency compared to stands associated with warm south facing aspects in inland mountainous forests of central B.C. Beaty and Taylor Radies 3 (2001) reiterated this influence of aspect and further identified a reduced fire frequency in lower slope, water-receiving topographic positions. DeLong (1998) reported that annual fire return intervals in wet montane forests of the upper Fraser River watershed ranged from 244 to over 1600 years while Sanborn et al. (2006) found a median time since fire in wet inland temperate rainforests between 800-1200 years These results suggest that disturbance processes in the wettest portions of ITR are more similar to coastal temperate rainforests, where single-tree gap dynamics dominate due to tree age (Lertzman et al. 2002), while in drier parts of the ITR, stand replacing fires will occur, yet infrequently. In the upper Fraser River watershed, this resistance to fire in wet toe-slope valley bottom positions has favored the development of forest stands that contain western red cedar {Thuja plicata) trees of exceptional age (i.e. 1000 years old) and stature (i.e. 3 meters in diameter) (Benson and Coxson 2002). Goward and Arsenault (2000) identify some of these forest stands as "antique": sites where the last major stand replacement disturbance event, such as fire, happened well before the current generation of trees established. Preliminary studies (Goward 2003) in the upper Fraser River watershed have suggested that forest stands in water-receiving toe-slope positions contain highly diverse communities of arboreal lichens. Human impacts on ITR watersheds of the upper Fraser River valley have occurred mostly in valley-bottom locations. Forest harvesting (and accidental fires) accompanied railroad development along the upper Fraser River valley in the early 1900s, followed by highway development on north facing slopes in the mid 1960s. Thus, while DeLong (2007) identified a natural range of variability between 76-84% in the cover of oldgrowth forests in wet mountain trench ecosystems of the upper Fraser River watershed, Radies 4 current cover estimates of approximately 64-68% (Anonymous 2005) suggest that these old forest types are under-represented relative to natural disturbance regimes. These trends parallel those in coastal wet temperate rainforests, where historical harvesting has similarly targeted old-growth forests in valley bottom locations (Moola et al. 2004) Current landscape level management policies in the upper Fraser River watershed specify an old-growth management threshold of no less than 53% old-forest cover greater than 140 years of age (Anonymous 2004). This target, which overlooks stand or site variability within the biogeoclimatic subzone unit, does not necessarily ensure that oldforest stands of high biological value will be retained in future landscapes. Indeed, the opposite may be true, in that the placement of transportation corridors through toe-slope stands in the upper Fraser River watershed has resulted in disproportionate clear-cut harvesting of old-forest stands in surrounding valley bottom positions. In the U.S. Pacific Northwest and elsewhere canopy macrolichens have been used as indicators of stand age (Hyvarinen et al. 1992, Campbell and Fredeen 2004) and environmental conditions on forested landscapes (McCune et.al. 2000, 2002, Liden and Hilmo 2005). Furthermore, macrolichens have proven to be useful indicators of total lichen diversity (Bergamini et al. 2005) and other taxa (Negi and Gadgil 2002) when applied appropriately (sensu Niemi and McDonald 2004). Therefore, we suggest that a landscape-level assessment of canopy macrolichens in ITR, could provide major advances in our understanding of the role old-growth site and stand structural attributes contribute to the biological integrity of these ecosystems. We addressed this premise by evaluating the composition and abundance of canopy macrolichens in relation to structural and site attributes in 53 old-forest stands, Radies 5 located within a 70-km2 area of the upper Fraser River watershed. Landscape-level sampling was stratified to ensure equal representation of wet and dry broad relative soil moisture (BRSM) site conditions. Reasons for stratifying our study design by soil moisture are twofold. First, soil moisture affects the ecology of forested landscapes including the development and structure of forest stands (Lertzman 2002, Spies et al. 2006), plant species numbers (Zinko et al 2005), and underlying ecological processes (Pastor and Post 1986, Turner 1989). Second, relative soil moisture is a forest management tool, applied from site-specific forest practices (DeLong et al. 2003) to coarse-filter landscape-level planning implementation using G.I.S. technology (Anonymous 1999, Iverson et al. 2000). We further stratified each of the two BRSM categories into an equal number of "cedar-" (Thuja plicata), "hemlock-" (Tsuga heterophylla), and "spruce-" (Picea glauca x P. engelmannii) leading stands, identifying the potential influence of substrate on lichen diversity (Goward and Arsenault 2003). Our analysis uses ordination approaches to examine community level responses of canopy lichens, and logistic regression to examine autecological responses of a subset of individual species identified by Goward (1994) as "old-growth associated". Our research is intended to provide guidance on the interpretation of canopy macrolichens as indicators of both microclimatic conditions and forest stand continuity in old-growth inland temperate rainforests. We also assess the importance of site and standlevel old-growth representation for managing sensitive macrolichen species, which may be a gap in current ecosystem-based management strategies (Anonymous 2004, Coast Information Team 2004, Price et al. 2008). The ecology and conservation status of lichens, as well as the range and magnitude of threats facing them are not well understood Radies 6 (Fazey et. al 2005). This study represents the first landscape-level analysis of arboreal lichen habitat attributes in old-forest stands of the inland temperate rainforest of western North America. There is a pressing need for this type of study, given the conversion of representative valley bottom temperate rainforests to plantation management and the attendant loss of old-growth associated species. METHODS Study Area The study area is located in east-central B.C., Canada, in the upper Fraser River watershed (Fig. 1-1, inset 1). This region is part of the inland rainforest formation or "interior wet-belt" (Stevenson et al. 2001) of the Rocky and Columbia mountain formations that consists of: high elevation wet and very wet Engelmann Spruce Subalpine Fir (ESSF) forests (in blue) (between 49° - 57° latitude); mid to low mountain elevation, wet and very wet Interior Cedar-Hemlock (ICH) forests (in green) (between 51° and 54° latitude); and extreme valley bottom locations of the very wet-cool Sub Boreal Spruce (SBSvk) forests (in yellow) (between 53° and 55° latitude). We focused on forests in the Slim variant of the very wet-cool ICH biogeoclimatic subzone (ICHvk2) (DeLong 2003), and to a limited extent, adjacent valley bottom forests within the SBSvk (Figure 1). ICHvk2 forests are dominated by western redcedar {Thujaplicatd), hereafter referred to as cedar and western hemlock (Tsuga heterophylla), hereafter referred to as hemlock, with some Douglas-fir {Pseudotsuga menziesii), hybrid white spruce {Picea engelmanni x P. glauca), hereafter referred to as spruce, and sub-alpine fir {Abies lasiocarpa), hereafter referred to as fir. SBSvk forests are dominated by spruce and fir (Fig 1-1). Radies 7 The distribution of remaining areas of old forest in the ICHvk2 and SBSvk varies greatly across the upper Fraser River watershed (Fig. 1-1). Many of the tributary valleys have been heavily logged and have little remaining old-forest cover. Furthermore, based on location of most logging clearcuts, it is evident that harvesting patterns have targeted low elevation wet broad relative soil moisture (BRSM) sites, most notably spruce and secondly cedar In general, wet BRSM are primarily found on north-facing slopes in mid to lower valley positions, though they occupy a more topographically restricted band in toe-slope positions on south facing slopes (i.e. compare inset 2 and inset 3, Fig. 1-1). Wet hemlock-leading sites are more spatially confined, often occurring in lower valley topographically flat positions with standing surface water. Spruce-leading forests, both wet and dry, are found both in upslope topographically cold locations and extreme low elevation sites, in part, subject to cold air pooling. Mean annual precipitation of the ICHvk2 is 839.8 mm (374.3 mm in summer and 465.5mm in winter) with a mean summer temperature of 14.7 °C and a mean winter temperature of-12.1°C. Recorded mean annual snowfall is 306.8mm persisting on the ground for up to 8 months of the year (Reynolds 1997). Slow-melting snow packs in higher mountain elevations tends to keep soil moisture levels high in the ICH during the summer months (Ketcheson et al. 1991), particularly on north-facing aspects. Study Design Field data were collected in summer 2004 and 2005. We randomly selected 53 GIS polygons to sample from a total of 120 candidate polygons (Fig. 1-1). Selection criteria for eligible polygons (hereafter called stands) included: a) location in the ICHvk2 and in the adjacent SBSvk (within 5 km of the ICHvk2) biogeoclimatic subzones; b) Radies 8 forest greater than 140 years in age; c) cedar, hemlock, or spruce leading; d) 500 m or less from road access (for logistical purposes) and; e) at least 50 m from cutblock edges, riparian areas, and deciduous forest types. We used the B.C. Ministry of the Environment Predictive Ecosystem Mapping (PEM) database (an ecosystem mapping database conducted at a 1:50,000 scale; see Anonymous (1999)) and the B.C. Ministry of Forests Vegetation Resources Inventory (VPJ) database (a forest inventory mapping; see Anonymous (1998)) to identify candidate polygons that met our selection criteria. Sampling within old growth forests was stratified to ensure representation from each of cedar-, hemlock-, and spruce-leading stands (using VRI), and from stands representing both wet or dry BRSM conditions (using PEM). At each stand, we laid out 2 plots that shared a common centre. The lichen assessment plot was a rectangular survey area 40m x 100m, with the long axis parallel to the slope contours to avoid marked topographic changes. Each plot was assessed for 37 possible arboreal foliose lichens (checklist adapted from Goward et al. 1994) using survey methods of McCune et al. (2000). Each macrolichen species observed was given an abundance rating between 0 and 4 (with the exception of Lobariapulmonaria): 0 = absent, 1 = rare [1-3 individuals/plot], 2 = uncommon [4-10 individuals/plot], 3 = common [>10 individuals], 4 = very abundant [covering more than half of available substrates]. For L. pulmonaria, similar categorical measurements were made, but because of its ubiquity and high abundances in this ecosystem, we used a measure of "hand-size" (approximately 10 x 20 cm2)/lichen plot in place of "individuals/lichen plot". Melanelia and Parmelia lichens were surveyed at the genus level. Radies 9 The stand structure plot was a circular plot with a radius of 17.8 m. The purpose of the stand structure plot was to provide more detailed information on the structural components of the forest stand structure and on Relative Soil Moisture (RSM) conditions. Diameter at breast height (DBH) (1.3 m) was measured for all stems greater than 16.5 cm DBH, categorized by live and dead stems and identified by tree species. RSM of each stand was classified on a seven point scale using the moisture regime key in DeLong (2003), which incorporates measurements of slope gradient, meso slope position, aspect and soil texture. Soil samples for texture analysis were obtained from within a soil pit dug to approximately one meter at plot centre. At each of 13 locations, one in the centre and two on each of the six transects, four replicate measurements of canopy openness were taken using a spherical densiometer. These measurements at each location were taken at 90° intervals, and then averaged. We subsequently pooled all 13 averages to obtain the overall "openness" of the stand. Predicted mean annual precipitation and temperature for each stand was obtained from the Canadian Forest Service (CFS) regional climate database (Hutchinson 1995). We used mean monthly minimum temperature for the months March-October in our analysis; this reflected the seasonal time period during which most lichen growth occurs (Coxson and Stevenson 2007). Potential solar insolation was calculated using SAGAGIS Version 2.0 (Scilands GmbH, Gottingen, Germany) solar radiation model. Data Analysis Nonmetric multidimensional scaling (NMS) ordination was used to examine trends in lichen community composition across stands (PC-ORD V. 4.0, McCune and Radies 10 Mefford 1999). We then used a general linear regression model to evaluate the following variables against ordination scores for Axes 1 and 2: temperature (TEMP), precipitation (PRECIP), solar insolation (SOLAR), canopy openness (OPEN), relative soil moisture (RSM), and basal area of cedar (BACw), hemlock (BAHw), spruce (BASx), and subalpine fir (BASf). Variables were tested for multicollinearity (Stata Corporation 2002, College Station, Texas). For all variables, we used a tolerance score of < 0.2 to indicate significant multicollinearity (Menard 2002). We used logistic regression to identify important environmental factors that influenced the distribution of lichens observed in our sample plots. We fit logistic regression models to presence-absence data for Cavernularia hultenii, Lobaria retigera, Nephroma isidiosum, Nephroma occultum, Platismatia norvegica, Peltigera collina, Sticta fuliginosa and Sticta oroborealis, species previously identified as old-growth associated by Goward (1994). Independent variables assessed for each lichen species included: categorized Broad Relative Soil Moisture (BRSM) categorized as wet or dry (see below), canopy openness (OPEN), average minimum temperature (TEMP), average annual precipitation (PRECIP), solar insolation (SOLAR), and leading tree species (LEAD, categorized as Cw, Hw, or Sx; see below) Plots were classified as wet when the BRSM was above 4 on the 7 point relative soil moisture scale, and dry if the BRSM was less than 4 (corresponding to mesic or submesic sites in DeLong's (2003) moisture regime key). When stands were classified as a 4 (mesic), vegetation and soil characteristics were used to separate wet versus dry BRSM categories. When identifying the leading species (LEAD), a stand was determined leading in cedar (Cw), hemlock (Hw), or spruce (Sx), based on the tree Radies 11 species that had the highest Basal Area in the stand (to be consistent with VRI classification methodology). We tested thirteen combinations of independent variables that served as plausible explanatory hypotheses for the distribution of each lichen species: l.BRSM X OPEN X TEMP X PRECIP; 2. BRSM X OPEN X TEMP X SOLAR; 3. BRSM X PRECIP; 4. BRSM X TEMP; 5. LEAD X BRSM; 6. LEAD X OPEN X TEMP; 7. LEAD X OPEN X PRECIP; 8. LEAD X OPEN X SOLAR; 9.LEAD X PRECIP; 10. LEAD X TEMP; 11. OPEN X PRECIP X SOLAR X TEMP; 12. OPEN X PRECIP; 13.OPEN X TEMP. We used Akaike's Information Criterion with a correction for small sample size (AICC) (Johnson and Omland 2004) to identify the most parsimonious logistic regression model. All AICC values were subtracted from the lowest AICC value in each model set to derive the AIC difference (AICC dif). We then calculated the AICC weights (AICcw) and interpreted this value as the approximate probability that the model with the largest value was the most parsimonious of the set (Johnson and Omland 2004). We calculated the area under the Receiver Operating Characteristic (ROC) curve for the top-ranked models (Munoz and Felicisimo 2004). ROC scores allowed us to evaluate the ability of the most parsimonious model to predict the distribution of lichens on the landscape. We used Multi-Model Inference (MMI) to determine the relative importance of the predictor variables for each species (Johnson and Omland 2004). MMI uses the AICcw to average the coefficients from all variables within the set of models for each lichen species and thus accounts for variation attributed to model selection uncertainty. We used 95% confidence intervals, corrected for model selection uncertainty, to assess the strength of effect of each predictor covariate on the dependent variable. Radies 12 RESULTS Of the 37 species and 2 genera {Melanelia spp. and Parmelia spp.) of arboreal lichens surveyed within the 53 study plots, 18 of the 19 cyanolichens were either more frequent and/or occurred with greater abundance in stands that had wet BRSM conditions (Table 1-1). The only exception to this pattern was the cyanolichen, Sticta wrightii which occurred rarely (less than 4 thalli present) in 3 sites, in which 2 of the 3 sites were classified as dry BRSM. The chlorolichens occurred with relatively even frequency and abundance between stands with wet and dry soil moisture conditions, the noticeable exception being Cavernularia hultenii, which had a much higher frequency of occurrence in stands with wet BRSM conditions. Stand ordinations showed clustering of bipartite cyanolichens in the upper left quadrant of the plot (Fig. 1-2). This included regionally rare species such as Lobaria retigera and Nephroma occultum and the more commonly abundant bipartite cyanolichens such as Nephroma isidiosum, Pseudocyphellaria anomala, and Sticta fuliginosa. The tripartite cyanolichen Lobaria pulmonaria, was found widespread throughout the ordination, although it displayed some tendency of increasing abundance in the upper left quadrant of the plot (Fig. 1-2). Most of the chlorolichens were widely distributed, showing no strong placement preference along the two ordination axes. Temperature and precipitation were significantly correlated with both axis 1 and axis 2 ordination scores (Table 1-2). When fit to a linear regression, data for Axis 2 demonstrated the best fit (R2 = 0.656), with the variables relative soil moisture (P = 0.002), temperature (P = 0.005), canopy openness (P = 0.010), precipitation (P = 0.020), and basal area of spruce (P = 0.028) accounting for a significant proportion of the Radies 13 variation. Only temperature (P = 0.001) and precipitation (P = 0.021) were correlated with Axis 1 ordination scores (R2 = 0.548). Mean annual temperature, relative soil moisture, basal area, and canopy openness co-varied (Table 1-3). When fitting and assessing the suite of logistic regression models we noted few similarities among the 8 old-growth associated lichen species (Table 1-4). The exceptions were N. occultum and C. hultenii, for which the best predictive models consisted of leading tree species and wet or dry BRSM status (although these two lichens selected for different leading tree species). The variables associated with the most parsimonious models for each lichen species were: temperature (6 species), leading tree species (5 species), relative soil moisture condition (4 species), openness (4 species), and precipitation (1 species) (Table 4). ROC scores ranged from 0.7175 {P. norvegica) to 0.9183 (N. isidiosum), indicating a good to excellent fit for each of the lichen species (Table 1-4). Averaged coefficients suggested that all variables, other than precipitation and insolation, had some influence (positive or negative) on the distribution of one or more old-growth associated lichen species (Fig. 1-3). Dry BRSM status had a 0 to negative effect on the presence of the eight lichen species, and wet BRSM status showed positive effects (both with relatively small confidence intervals). Canopy openness (OPEN) showed 0 or slightly positive effects. The temperature variable had the largest effect on the distribution of N. isidiosum or 5*. oroborealis, though confidence intervals were quite large. In combination with temperature and/or BRSM, L. retigera, P. norvegica, and S. oroborealis showed greatest affinities to stands leading in hemlock and to a lesser degree cedar. C. hultenii, a chlorolichen, was related almost exclusively to hemlock and Radies 14 negatively to both cedar and spruce leading stands whereas N.occultum, a cyanolichen, showed preference to cedar dominated forests and a negative association with both hemlock and spruce dominated forests. Spruce had either no effect or a negative influence on the presence of the 8 lichen species we tested. As with temperature, confidence intervals were often quite large for leading tree species (Fig. 1-3). DISCUSSION The first major question of our study was whether distinct assemblages of oldgrowth dependant foliose macrolichens were distributed equally across an old-growth forest landscape of Inland Temperate Rainforest (ITR). A homogenous distribution of lichens would suggest that arboreal lichens do not respond to stand characteristics other than the criteria of reaching ages of 140 years or older. Our ordination plots show a clear assemblage of lichen species within a representative subset of old-growth forests greater than 140 years of age. These assemblages occurred mostly at low elevations (as lower elevations in the temperature models were identified by warmer temperatures) with topography that favored the accumulation of soil moisture. Of the old-forest associated species identified by Campbell and Fredeen (2004) the following more regionally rare species, C. hultenii, L. hallii, N. isidiosum, andN. occultum, were either limited to or more frequent and abundant in stands with wet BRSM conditions. Other old-growth associated lichens designated by Campbell and Fredeen (2004), such as H.vittata, L. scrobiculata, N. helveticum, N. parile, P. anomala and S. fuliginosa, were more widely distributed in our old forest stands, but were still far more abundant on wet BRSM sites. A third set of old-growth associated lichens identified by Campbell and Fredeen (2004), Radies 15 chlorolichen species such as Hypogymnia rugosa and P. norvegica, were widely distributed in both wet and dry stands greater than 140 years of age. From these results, we can identify an appropriate use of macrolichens as indicators at this regional scale. First, the presence and high abundance of L. pulmonaria across most of our research sites (wet or dry) proved not to be a sensitive indicator to site-specific conditions and lichen diversity (as suggested by Campbell and Fredeen (2004)). This result concurs with Kalwij et al. (2005) that suggested that L.pulmonaria at the site level is not sensitive to landscape disturbances, but could be a useful indicator of lichen diversity and disturbance frequencies when comparing among regional landscapes. We postulate that within regional landscapes, the presence and abundance of bipartite cyanolichens would serve as more appropriate indicators of potential biodiversity "hotspots" (sensu Bergamini et al. 2005), canopy microclimate conditions (Stevenson and Coxson 2008) and site disturbances (Goward 1994). We must caution, however, that the proliferation of a guild of bipartite cyanolichen species is a reflection of individual species' ability to disperse, establish and persist at a particular site. Therefore, while we may identify that L.retigera and N.occultum are both found in greater numbers and frequencies in sites of wet BRSM, we must also be aware that co-varying ecological attributes found in wet BRSM (i.e. soil moisture, temperature, and site disturbance) could influences the presence of these regionally rare macrolichens differently. Our data suggests that sites of wet BRSM status are biodiversity hotspots due to a combination of optimal microclimatic conditions (due to lichen establishment limitations) and potential stand continuity (due to lichen dispersal limitations). Radies 16 Glavich et al. (2005) pointed to temperature and moisture as major predictors of cyanolichen diversity over a vast region of coastal wet temperate rainforests. Our correlation of environmental attributes with ordination scores suggests that gradients of temperature and moisture continue to play an important role in shaping cyanolichen assemblages at finer landscape scales. However, we must stress here, that our climate variables are derived from climate models (as with Glavich et al. (2005)), not from stand level measurements. Temperature has long been inferred as an important environmental variable in structuring ITR cyanolichen communities. Canopy cyanolichens rapidly diminish in abundance as stand composition shifts from cedar-hemlock to spruce-fir with increasing elevation (Goward 1994). Studies on cyanolichen physiology suggest that processes of carbon assimilation and nitrogen fixation are highly rate-limited at low temperatures (Sundberg et al. 1996). Interactions with precipitation, which can be viewed as a proxy for the duration of thallus hydration, are also fairly straightforward. Lichen growth models are highly sensitive to the duration of physiological activity (Sundberg et al. 1996). Coxson and Stevenson (2007) showed that most growth in L. pulmonaria populations from the ITR coincides with precipitation events in the spring and summer (the exception being some snowmelt events in the early spring). Although thalli are often hydrated for long time periods in the late fall and winter, they are commonly frozen and experience very low light availability, and hence cannot realize much growth potential. The opposing trends in temperature and precipitation with increasing elevation in ITR mountain valleys would appear to have major constraints on the establishment of canopy cyanolichens. However, site moisture and relative humidity are not related solely Radies 17 to precipitation. The amount of slope above a position on the landscape is a major factor in determining site moisture (Ketcheson et al. 1991). Relative soil moisture status was significantly correlated with axis 2 of the NMS ordination and was a significant variable in a majority of best model sets predicted by logistic regression for individual species. Where wet BRSM status coincides with warmer valley bottom conditions, lichen communities escape the constraints that would otherwise be placed on their development by regional gradients of temperature and precipitation. We hypothesize that the greater relative humidity found within the lower canopy of these stands extends the duration of periods of metabolic activity experienced by canopy lichens after precipitation events. This, in combination with the form in which the precipitation falls (i.e., rain versus snow), would identify more optimal habitat — particularly for cyanolichens, which need direct contact with water to resume physiological activity (Budel and Lange 1991). This is evident where one finds adjacent dry BRSM stands with much lower canopy lichen diversity, notwithstanding very similar exposure to temperature and absolute precipitation. Old-forest stands that develop in wet BRSM areas tend to share many common attributes. Basal areas of cedar, fir, and spruce are greater than in dry BRSM areas, presumably reflecting the influence of subsurface water on tree growth, and indirectly, the greater exclusion of fire as a major natural disturbance agent (Eng 2000, Beaty and Taylor 2001). Importantly, wet stands tended to have a more open canopy structure, reflecting the greater role of gap dynamics within old-forest stands in water-receiving positions (Benson and Coxson 2002, Lertzman et al. 2002, Radies and Coxson 2004). For canopy cyanolichens, this combination of abundant light in a humid lower canopy Radies 18 environment creates ideal conditions for growth and establishment (Coxson and Stevenson 2007). Spies and Franklin (1991) postulated that receipt of groundwater flow and attendant transfer of nutrients was a major determinant of the overall growth, development and structure of coastal wet temperate rainforests. Spies et al. (2006) further pointed out the importance of recognizing these site-specific factors when developing plans for the conservation of old-growth forests. Conversely, higher fire return intervals on upslope positions, particularity those of south-facing aspects (Eng 2000), have most likely limited the accumulation of rare lichen species — due to both dispersal limitations (Sillett et al 2000, Hilmo and Sastad 2001) and unfavorable site and structural characteristics for lichen establishment (as discussed). Our second major question was whether old-growth lichens, when examined individually, would select similar environmental variables. The same variables played an important role in predicting the presence or absence of individual species. All 8 of the so-called "old-growth associated" lichen species had either temperature or BRSM as major predictive variables in their best model sets (> 0.1 Akaike's weight), with 6 of these species having both predictor variables present. However, of the 8 species, only 2 species shared similar parsimonious models and 5 demonstrated varying affinities to leading old-growth stand type — indicating habitat limitations for some lichen species across our study area and the importance of stand representation. Presence of the endangered cyanolichen species N. occultum was predicted best by wet cedar-leading stands. Evidence suggests that undisturbed wet cedar-leading stands have persisted for time periods well in excess of the age of the oldest trees, fulfilling definitions of antique forest stands (Goward and Arsenault 2000). This Radies 19 suggests that N. occultum is most likely dispersal limited, proliferating only in stands that reach exceptional ages. Thalli of N. occultum, however, may also be influenced by greater nutrient availability in water receiving (wet BRSM) stands. These sites receive groundwater flow from upslope catchment areas, potentially representing a landscape level transfer of soluble nutrients. Further, on sites where cedar dominates, enrichment of exchangeable soil calcium can occur through deposition of CaCCb in litterfall (Graff et al. 1999), a potentially important factor in subsequent enrichment of throughflow precipitation as it passes over canopy foliage. Most of the "old-growth associated" lichens that selected for leading species also demonstrated wide confidence intervals around the respective coefficients. These large intervals can be explained by the broad autoecological requirements of individual species themselves and/or the broad scale at which forest stands were measured (in this case, leading stand type). Importantly, our measurement of "leading-tree species" does not identify the gradient of mixed conifer forest of cedar, hemlock, spruce and fir most often found in these forest types. However, the association of lichens within this gradient of mixed stand types could help explain the similar positive influence of both cedar and hemlock leading forests on the presence of L. retigera, P.norvegica, and S. oroborealis. This finding also suggests a broader ecological tolerance to stand conditions by certain "old-growth associated" lichens, but not by all. From the perspective of setting priorities for the conservation of canopy lichen communities, the retention of representative old-forest stands in areas of water-receiving valley bottom positions should be a high priority for land-use planners. This strategy will also ensure old-growth ITR landscapes are more resilient by mimicking natural Radies 20 disturbance regimes (sensu Drever et al. 2006) and protecting regions of temperate mountain coniferous forests less prone to fire disturbance —such as north-facing aspects (Eng 2000), moderate slope terrain, and valley bottom locations (Beaty and Taylor 2000). We also suggest that forest harvesting practices in these regions needs to be more complex including variable retention single tree and small patch cuts that reflect the natural range of variability in disturbance events that characterizes these wet old-growth forest systems (Lertzman et al 2002, Radies and Coxson 2004). Across dry BRSM sites, particularly those areas more prone to both insect outbreak and fire, clearcut harvesting may be a more suitable prescription. We emphasize that harvesting strategies must recognize spatial and temporal scale and pattern discrepancies between wet and dry BRSM forest types. Ecosystem-based forest management would not "borrow" disturbance patterns from one moisture type and prescribe it to another (i.e. frequent clear-cuts in wet BRSM sites). Current spatial representation of old-forests in protected areas in the ICHvk2 is approximately 6%, — well below the ecosystem-based threshold target of 53% set for the upper Fraser River watershed (Anonymous 2004). Given that many of the rare lichens were only found in one or two wet sites (and not the same ones), it is highly unrealistic to expect that current protected areas will maintain canopy lichen diversity within regional landscapes. Furthermore, continued clear-cut forest harvesting in representative wet, low elevation old-growth forest stands, will likely ensure a greater loss of macrolichen biodiversity prior to reaching regionally set ecosystem-based old-growth threshold targets (Anonymous 2004). Indeed, past harvesting of old-forest stands in wet BRSM sites may Radies 21 already have incurred an "extinction debt" (Berglund and Jonsson 2005) in the upper Fraser ITR. CONCLUSION The application of indicator species requires an understanding of both the natural history of the organism and the appropriate use of scale and measurement. Our data suggest that at site-specific scales in mountainous old-growth ITR, bipartite cyanolichens could prove useful in the determination of areas of high biological value due to both optimal site conditions (relating to lichen establishment) and potential longevity of the stand itself (relating to lichen dispersal). Given the reduction in the total area of these site types due to anthropogenic activities, we recommend that old-forest stands in the upper Fraser River watershed that have wet BRSM status be given immediate consideration for protected area status. These sites represent significant biodiversity hotspots for canopy lichens that are essential for the maintenance of biodiversity within larger regional landscapes. The distribution of these organisms highlights the importance of appropriately identifying soil moisture site discrepancies in old-growth forests for the purposes of executing effective ecosystem-based management thresholds and strategies. Therefore, when applying old-growth thresholds to temperate old-growth rainforests, the question, "how much is really enough?" (Price et al. 2008), must also be coupled with, "of what kind, in what location, and in what context?". Otherwise, the objective of setting threshold targets for the purposes of maintaining biodiversity will most likely miss its mark. Radies 22 CHAPTER 2: PARALLELS BETWEEN OLD-GROWTH FOREST RETENTION TARGETS AND THE CHILDREN'S GAME KERPLUNK! INTRODUCTION Remember the children's game "KerPlunk!"? For those who have forgotten, or were never exposed, the game was very simple. Players first created a crisscross horizontal web of tiny plastic colored sticks strategically placed through holes about the centre region of a vertically standing plastic tube. Marbles were then added to the top of the tube, in which the sticks in the center of the tube immediately halted their downward momentum. At this point, the game begins. Players remove the sticks, and "with a bit of skill and a little luck, you can keep a lot of marbles from going KerPlunk!" With this in mind, let's draw some comparisons between this game and maintaining biodiversity and ecological integrity using old-growth retention targets in B.C.'s Inland Temperate Rainforests. The purpose of setting old-growth retention targets is to maintain ecological integrity and biodiversity associated with these older forests (Anonymous 2004, Price et al. 2008). In some areas, determining these threshold targets are based on the determination of the Natural Range of Variability (NRV) — a set amount of old-growth found within an ecological unit. The assumption in NRV theory is that biological diversity will be impacted less if forest management follows similar temporal and spatial patterns of natural disturbances (sensu Drever et al. 2007). However, identifying appropriate scale (both temporally and spatially) is critical for the effectiveness of such a management system (Grumbine 1994, Drever et al. 2007). Radies 23 In British Columbia, when old-growth threshold targets are applied, they are most often set at the biogeoclimatic subzone scale for forests greater than 140 years (Wells et al. 2003, Anonymous 2004, Coast Information Team 2004). Assuming these to be appropriate spatial and temporal units of scale, the question: "how much [old-growth] is really enough?" (Price et al. 2008), to maintain biodiversity is the next logical question in forest management; in other words, how far can we push old-growth threshold targets without ecological consequences? Although this may sound like a seemingly easy question, it is not. However, the province is accepting a "low risk to biodiversity" if an old-growth threshold target of 70% the minimum NRV is implemented in forests greater than 140 years at the biogeoclimatic subzone unit in wet mountain forests (Wells et al. 2003, Anonymous 2004). An example of this old-growth management strategy can be found in the Prince George Timber Supply Area (PGTSA) (Anonymous 2004). In the PGTSA, most ecological input for setting old-growth threshold targets followed two scientific works: 1. DeLong's (2007) estimation of the Natural Range of Variability (NRV) in the Prince George Forest region and; 2. DeLong et al. (2004) work comparing wildlife tree habitat value between mature (class 8, 140-250 years of age) and old (class 9, 250 years or older) forests. The NRV study results identified a NRV for forest >140 years within the inland rainforest (called Wet Trench-Valley) at 76-84% (DeLong 2007). The second study by Delong et al. (2004) found that wildlife tree habitat measurements did not seem to discriminate between age class 8 and 9 stands (thus allowing the incorporation of class 8 stands to fulfill old-growth threshold requirements). Given these results — in combination with the accepted setting of old-growth thresholds at 70% the minimum NRV — a 53% non-spatial old-growth threshold target for Wet Radies 24 Trench-Valley forests over 140 years of age was accepted into regional policy (Anonymous 2004, Delong 2007). In other words, the only requirement for meeting the 53% old-growth retention target is that forests are 140 years of age. Thus, old-growth threshold targets can be non-spatially met primarily in the Non Timber Harvesting Landbase (NHLB) that constitutes approximately half of the remaining old-growth forest landbase in the PGTSA (Anonymous 2005). But will this old-growth threshold approach be capable of maintaining biodiversity? One group of organisms that has been of considerable concern in ITR ecosystems is canopy lichens (Goward and Arsenault 2000, Goward and Spribille 2005). Arboreal lichens have been shown at the site level to be influenced by both age (Campbell and Fredeen 2004, Radies and Coxson 2004) and stand characteristics (Goward and Arsenault 2003, Coxson and Stevenson 2007, Stevenson and Coxson 2008). Therefore, arboreal macrolichens seem to be an appropriate group of organisms with witch to test the effectiveness of an aspatial old-growth theshold policy in ITR forests. To do this, I compared the presence or absence of 37 macrolichen species in forests greater than 140 years of age by two broad-scale ecological variables — leading tree species and broad soil moisture regime status. My reasoning for choosing these two main variables was: 1. they both greatly influence the NRV (i.e. fire return, insect outbreak), stand structure and tree composition in mountain forests (Spies and Franklin 1991, Eng 2000, Beaty and Taylor 2001) and; 2. they are both practical in landscape-level queries, such as building species predictability maps, using provincial GIS tools (most notably, the Vegetation Resource Inventory (VRI) (Anonymous 1998) for leading tree species and the Predictive Ecosystem Mapping (PEM) (Anonymous 1999) database for soil moisture regime). The Radies 25 questions I therefore explored using these two central forest variables in combination with presence/absence macrolichen data were as follows: • Do canopy macrolichens significantly vary in their presence between old-growth forests greater than 140 years of age in wet soil moisture conditions versus dry relative soil moisture conditions • What percentage of the ICHvk2 old-gorwth forest is in wet and dry relative soil moisture conditions and of this, what proportion is inside or outside the timber harvesting landbase? • Can we manage old-growth forests for biodiversity and ecological integrity using threshold targets set at the biogeoclimatic unit scale? • Can we make any inferences from our lichen results that would suggest that areas outside or inside the THLB are more susceptible to natural disturbances? • Can we use VRI and PEM to build meaningful coarse filter predictability maps for canopy lichens of concern? It is my hope that the overall findings in this report will provide useful information for forest managers in BC on the benefits and pitfalls of old-growth threshold targets set at the biogeoclimatic unit scale. METHODS Study area The study area is located in one of two variants of the very wet Interior Cedar Hemlock Biogeoclimatic SubZone — The ICHvk2 of the upper Fraser River watershed (Fig. 2-1). ICHvk systems are the wettest low elevation cedar-leading forests in the interior of the province and are found in only two locations, one in the south near Radies 26 Revelstoke (the ICHvkl) and the other near Prince George (the ICHvk2) (Fig. 2-1). These vk systems are located in the northern portion of the Inland Temperate Rainforest (ITR) bioregion (Inset Fig. 2-1) and are part of the ICH "wet-belt" region (ICH wet and very wet subzones) (Fig. 2-1). The ICH wet-belt and its surrounding watersheds have had extensive clearcut logging in the past 50 years (Fig. 2-2). The majority of remaining class 8 (140-250 years) and class 9 (250 years or older) ICH wet-belt forests are restricted and most extensive in the northern ITR region, (Fig. 2-3), in which provincial park protection of ICHvkl and vk2 old-growth forests (Fig. 2-2) is spatially minimal (Fig. 23). Figure 2-4 identifies the extent of clear-cut forest harvesting in and around the ICHvk2 and lower elevation very wet Sub Boreal Spruce (SBSvk) biogeoclimatic subzone of the upper Fraser River watershed (note that the area that appears to have no cutblocks is TFL 30 for which the information was not available at time of writing). Figure 2-5 outlines study area and plot locations in the SBSvk and ICHvk2 (divided into the vk2 and wk4 during time of study (see Inset Fig. 2-6)) and Figure 2-6 identifies the varying distribution of cedar {Thuja plicata), hemlock (Tsuga heterophylla), spruce (Picea glauca x engelmannii), and "other" coniferous leading forests greater than 140 years of age in the ICHvk2 and ICHwk4 (using the VRI). Minimum and maximum climate data for the 53 study sites (from Chapter 1) were as follows: Absolute precipitation, 736mm-917mm; Average minimum temperature, 3.14°C to -1.65°C; Average maximum temperature, 7.96°C to 9.92°C; Mean annual average temperature: 2.44°C to 4.03°C; Average 8 month physiological temp range: 0.553°C to 2.39°C. Radies 27 Study design Research plots were established in 53 randomly stratified mature and old-growth forest polygons in the SBSvk and ICHvk2 (now divided into the ICHvk2 and ICHwk4). My primary sampling objective was to collect from an equal number cedar, hemlock, or spruce-leading old-growth forests (greater than 140 years of age) in both wet and dry Broad Soil Moisture Regime (BSMR) sites. At each stand, we laid out 2 plots that shared a common centre (Fig. 2-7). The lichen assessment plot was a rectangular survey area 40m x 100m, with the long axis parallel with the slope contours to avoid marked topographic changes. Each plot was assessed for 37 possible arboreal foliose lichens (checklist adapted from Go ward et al. 1994, see Table 2-1 for list of species and abbreviations) using survey methods of McCune et al. (2000). Melanelia and Parmelia lichens were surveyed at the genus level. The stand structure plot was a circular plot with a radius of 17.8 m for the purpose of measuring stand structure and relative soil moisture conditions. We used densiometer measurements at 13 locations in each lichen assessment plot to obtain the overall "openness" of the stand. Predicted mean annual precipitation and temperature for each stand was obtained from the Canadian Forest Service (CFS) regional climate database (Hutchinson 1995). We used mean monthly minimum temperature for the months March-October in our analysis; this reflected the seasonal time period during which most lichen growth occurs (Coxson and Stevenson 2007). Potential solar insolation was calculated using SAGAGIS Version 2.0 (Scilands GmbH, Gottingen, Germany) solar radiation model. To analyze data for the purposes of fitting field data to PEM and VRI data, we categorized both site and stand variables. Sites were classified as wet BSMR when the Radies 28 soil moisture regime was above 4 on the 7-point soil moisture scale, and dry BSMR if the soil moisture regime was less than 4 (corresponding to mesic or submesic sites in DeLong's (2003) soil moisture regime key). When stands were classified as a 4 (mesic), vegetation and soil characteristics were used to separate wet versus dry BSMR categories (i.e. mesic wet versus mesic dry site conditions). Leading tree species (LEAD) was categorized as stands leading in cedar (Cw), hemlock (Hw), or spruce (Sx), based on the tree species that had the highest Basal Area in the stand (to be consistent with VRJ classification methodology). For analysis of "rareness" to "commonness" of species present in sites surveyed, a presence/absence dendrogram analysis for species was applied for all sample locations (by BRSM site category) following methods in McCune and Mefford (1999) (PC-ORD V. 4.0). We used beta coefficients derived from logistic regression results from Chapter 1 to map, using VRI and PEM, important environmental factors that influenced the distribution of 3 of 8 old-growth associated lichens — Cavernularia hultenii, Lobaria retigera, and Nephroma occultum. Independent variables used in Chapter 1 to assess for each lichen species included: BRSM (categorical), canopy openness (OPEN), average minimum temperature (TEMP), average annual precipitation (PRECIP), solar insolation (SOLAR), and LEAD (categorical) (see Table 2-2). We tested thirteen combinations of independent variables that served as plausible explanatory hypotheses for the distribution of each of the 3 lichen species (Table 3-3). We used Akaike's Information Criterion with a correction for small sample size (AICC) (Johnson and Omland 2004) to identify the most parsimonious logistic regression model. We calculated the area under the Receiver Operating Characteristic (ROC) curve for the top-ranked models (Munoz and Felicisimo Radies 29 2004). ROC scores allowed us to evaluate the ability of the most parsimonious model to predict the distribution of lichens on the landscape. We used Multi-Model Inference (MMI) to determine the relative importance of the predictor variables for each species (Johnson and Omland 2004). We used 95% confidence intervals, corrected for model selection uncertainty, to assess the strength of effect of each predictor covariate on the dependent variable. To map the probability of occurrence for L .retigera, N. occultum, and Cavernularia hultennii across the study area we combined beta coefficient values (from Chapter 1) to VRI, PEM, and climate data. Probabilities for each cell were calculated according to eqn 1 and represented the sum of each beta coefficient multiplied by the respective GIS value. Eq. 1. y = PiX] + P2X2... -PNXN- Relative to eqn 1, the p is the average MMI coefficient (i.e. for BSMR, LEAD, SOLAR), and X is derived from the GIS databases from PEM (for BSMR), VRI (for LEAD and OPEN), CFS climate data (for TEMP and PRECIP), and solar insolance (for SOLAR). BSMR site conditions were determined from queries in PEM. We assigned polygons with a dominant (41-100%) site series of mesic, subhygric, or hygric to a wet BSMR status, and polygons with a submesic, subxeric or xeric dominant site series (41100%) to a dry BSMR status. Polygons with a primary mesic site series between 41 to 90% combined with a secondary (10-49%) submesic site series, were classified as having a dry SMR status. This analysis was combined with forest inventory data to estimate the remaining area of old-forest in the ICHvk2 by age class (140-250 years old or > 250 years old), leading tree species (spruce, hemlock and cedar), and BSMR status (wet or dry) inside the Timber Harvesting Landbase (THLB) and outside the THLB (NHLB) (including parks and Old Growth Management Areas). Radies 30 RESULTS Sites were relatively evenly distributed between BSMR and LEAD categories (Table 2-4). Based on the presence/absence dendrogram for all species measured, we found trends that identify similarity of both site and macrolichen species (Fig. 2-8). The species axis categorized 3 major subgroups, in which (from left to right) macrolichens are "rare" (found in less than 50% of the sites), "less common" (found in more than 50% of the sites), or "common" (found present in almost all sites). On the site axis, we see a division into 2 major subgroups, in which (from top to bottom) dry sites dominate (18 of 26) in the first subgroup and wet BSMR sites dominate in the second subgroup (19 of 25 times). Overall, wet BSMR subgroups had higher levels of macrolichen diversity, particularly for the bipartite cyanolichen species. PEM and VRI derived data compared well with field-derived variables, with the exception of canopy openness, for which accuracy from the VRI database was inconsistent (<25% accurate), forcing us to construct maps of lichen distribution without this variable. Leading species field data was found to be 77% accurate when compared with the VRI database and BSMR categorization was 74% accurate within the PEM database (following outlined definitions of wet and dry BSMR polygons). The resulting maps illustrate the distribution of wet (mesic wet to subhygric) and dry (mesic dry to xeric) BSMR old-growth sites (Fig. 2-9) and our selected lichen species (Fig. 2-10, 2-11, and 2-12). Spatial predictions placed L. retigera predominantly in areas with warmer temperatures and in hemlock- or cedar- leading stands (Fig. 2-11). C. hultennii (Fig. 210) and N. occultum (Fig. 2-12), both predicted by the model BSMR X LEAD, illustrate the usefulness of leading tree species (VRI data) combined with BSMR (PEM data) to Radies 31 create predictive landscape maps. However, although our ROC scores indicated a good predictability when applying these models to PEM and VRI, it is beyond the scope of this paper to fully analyze the accuracy of the PEM and VRI databases. From the forest inventory data (as of 2006), 64.3% (67,769 ha) of the total forested area of the ICHvk2 biogeoclimatic zone is composed of cedar-, hemlock-, and spruce-leading stands > 140 years old (Table 2-5). Of this area, representation is near equal between wet and dry sites (29% and 35% respectively). Total protection is also relatively similar between wet and dry BSMR sites at approximately 6% each. However, areas identified as not in the Timber Harvesting Landbase (NHLB) differed between wet and dry BSMR conditions, in which 11.3% (39.7% of 29) wet BSMR sites and 19.3% (54.7% of 35.3) dry BSMR old-growth sites are found in the NHLB. DISCUSSION Dendrogram results identified that canopy macrolichens are not equally present in mid to low elevation coniferous mountain forests greater than 140 years of age in the upper Fraser River watershed. Notably, wet soil moisture sites (mesic wet and wetter), contained a greater number of rare and uncommon canopy lichen occurrences than those dry soil moisture sites (mesic dry or drier). This result is somewhat intuitive, as moisture has been shown to play a critical role in the distribution of lichen species near streams (Liden and Hilmo 2005) and in temperate regions of increasing precipitation (Goward and Spribille 2005). However, understanding the influence ground moisture plays on the presence of canopy cyanolichens is not well understood. In Chapter 1, Radies et al. hypothesized that increasing soil moisture would increase air humidity, allowing for the establishment and proliferation of a lichen species Radies 32 restricted by the requirement of longer periods of thallus hydration. However, Radies et al. (Chapter 1) also recognized a combination of other ecological characteristics in sites of wet soil moisture that co-vary and likely influence the presence or absence of rare canopy lichens, namely: 1. a difference in canopy structure that favors the establishment of cyanolichen species, such as a more open canopy structure that allows for an increase in direct thallus wetting episodes, lighting periods, and substrate availability; 2. higher temperatures, due to lower topographic positioning, and; 3. an increase in site continuity due to the absence of natural disturbances (i.e. stand destroying fires) on topographically wet sites, augmenting the probability of rare lichen species occurrences. Notable here, is that the first two points assume forests after reaching 140 years of age will contain an equal opportunity to house rare species if structural and microclimatic conditions remain constant. Our species predictability maps are based on this assumption (following current NRV forest policy), delineating the probability of species occurrence based on microclimate and stand-level characteristics. But why then, are rare species absent in regions delineated suitable in our habitat models? For example, if we consider the predictability map of the threatened species Nephroma occultum, we see that it has a higher probability of selecting for wet cedar-leading forests in the ICHvk2. Is this selection for wet cedar leading forests (that are believed exceptional in age) a result of a species limited by establishment or dispersal? Although there is evidence that favorable lichen establishment environments in old-growth forests constitutes a critical phase in the proliferation of arboreal lichens (Hilmo and Sastad, 2001), successful transplant experiments in younger forest stands suggests that for some lichens, dispersal is a limitation (Rosso et al. 2000; Sillet et al. Radies 33 2000). This is a critical point; as dispersal limitation highlights the critical role temporal scale plays in the landscape distribution of rare epiphytic species. In this case, similar site and structural characteristics will not be as important to species as forests that are extremely old (as the probability of species presence, limited by dispersal, increases with time). This emphasizes the importance of temporal scale in NRV management that is vital to biological diversity. Therefore, to improve NRV forest management in British Columbia, we must readdress how we intend to identify, describe and manage old-growth forests after they reach 140 years of age. Managing for natural disturbances at the biogeoclimatic subzone scale: doing our best given what we know It is well known that fire is distributed unequally across mountain landscapes (Spies et al. 2006). This distribution of fire frequency and intensity is most often correlated with wet or dry forest stand conditions at multiple scales (Eng 2000, Beaty and Taylor 2001). This is why we recognize, at a broad scale, that ITR forests burn more frequently than coastal rainforests — as ITR receives approximately half the annual precipitation as its coastal counterpart. At finer scales, spatial patterns of fire emerge as a result of aspect, slope and soil characteristics in mountainous forests (Beaty and Taylor 2001). For example, while studying drier mountain forests near our study area in the upper Fraser River watershed Eng (2000) noted an almost 10-fold increase in fire intensity on south facing slopes. This pattern of dry and wet site conditions clearly emerges when we review Insets 2 and 3 from Figure 1-1 showing south-facing slopes of Driscoll Ridge dominated by dry BSMR sites (mostly mesic dry and submesic sites) and north-facing slopes of the ridge dominated by wet BSMR sites (mostly mesic wet and Radies 34 subhygric conditions) in the ICHvk2 and SBSvk. This query demonstrates the usefulness of PEM when developing and implementing NRV in old-growth forest landscapes, which can identify forested regions that are potentially more prone to fire disturbance. If we compare forest stands inside and outside the timber harvesting landbase (THLB) (Table 2-5), we find a higher amount of wet BSMR sites potentially available for future harvesting throughout all forest stand types. The potential result of this timber supply model would be a forested landscape where remaining old-growth stands were more susceptible to fire disturbance (as there would be more water shedding old-growth forest on the landscape). This demonstrates how aspatial objectives for setting oldgrowth threshold targets can run counter to NRV forest policy objectives set by the BC Ministry of the Environment for the purpose of maintaining species diversity (Anonymous 2004). In other words, if NRV forest policies do not recognize the implications of aspatial old forest targets, which do not require representation of forests in wet BSMR conditions, how is it possible to maintain ecological integrity in these oldgrowth forest systems (sensu Drever et al. 2006)? For example, given current NRV policy in the PGTSA it is possible to harvest 10% of the remaining forest which could potentially remove all remaining class 9 cedar-leading stands (outside protected area) regardless of their spatial or temporal distribution within a biogeoclimatic subzone (Table 2-5). Furthermore, the reduction of the NRV old-growth management thresholds from a predicted "natural" baseline of approximately 80% to 53%, due to timber supply concerns, highlights why it is critical to maintain stand and site conditions that are optimal to maintain lichen species richness in the upper Fraser River watershed. Radies 35 One approach that could address these stand and site level concerns is to include both spatial and temporal forest stand representation in old-growth management policy at the biogeoclimatic subzone scale. This means that forests would need to be proportionally represented on the landscape both spatially (i.e. based on tree species composition and site condition) and temporally (i.e. higher or lower old-growth threshold targets based on NRV). The result of this approach would be an old-growth forest landscape equally represented in proportions more reflective of a natural landscape (to the best of our understanding). Notable here, is that the increasingly popular application (provincially) of identifying ecological risk by representing site series by equal area (for example see Ecosystem Representation in TFL 30) does not align well with most NRV or Ecosystem-based Management (EBM) literature (sensu Grumbine 1994, Drever et al. 2006). The criticism here is that the site series within an biogeoclimatic subzone will not be represented by the natural spatial and temporal distribution found within the ecological unit. SUMMARY Temperate inland mountain old-growth forests of BC are complex at regional scales; ranging from younger dry hemlock dominated forest on topographically dry ridges to older wet cedar dominated forests in water receiving areas. Recognizing these spatial and temporal landscape patterns are a first step to improving both NRV and/or EBM goals and objectives for maintaining biodiversity and ecosystem integrity in BC. The results of this study concur with NRV theory by identifying an increase in the presence of rare and uncommon canopy macrolichens in water receiving regions of the landscape that are more than likely less disturbed by natural disturbances such as fire. We therefore Radies 36 recommend that NRV policy address appropriate spatial and temporal representation of old-growth forest stands within biogeoclimatic subzones. To compare these old-growth threshold targets to the game Kerplunk!, we recognize that the removal of each tree, or hectare of forest, is not always equal for the purpose of maintaining biological diversity; likewise, removing a given number of sticks from a Kerplunk! tube for the purpose of preventing marbles from dropping does not always have equal results. Recommendations for implementing EBMfor the purpose of maintaining biodiversity and ecological integrity in old-growth inland temperate rainforests. 1. Recognize and identify in policy, given the best of current knowledge, the temporal and spatial distribution of natural disturbances in forests at the biogeoclimatic subzone scale. Importantly, recognize the role class 9 forests (identified as greater than 250 years) play in the distribution of species known or believed limited by dispersal. 2. Update provincial databases to identify previous stand composition and structure in previously harvested old-growth sites. For example, wet cedar-leading forests, multi-aged, with a secondary layer of hybrid spruce. 3. Recognize that although human activity is part of NRV and/or EBM forest management, we must first construct (to the best of our ability) how stand Radies 37 processes would have occurred across the landscape without human intervention. This becomes an essential baseline layer in which to measure human impacts. 4. Measure other taxonomic groups other than lichens to determine if they are also following similar ecological patterns. 5. Focus future research to address knowledge gaps in above recommendations. CONCLUSION The purpose of this thesis was to ask if macrolichen species differed across a regional old-growth landscape, and if so, what ecological characteristics emerge to influence this distribution. From our results we identify that for some groups, such as the chlorolichens, the distribution is relatively equal, while for others, such as the cyanolichens, the distribution in old-growth forested landscapes is distinctly unequal, concentrating in lower elevation water-receiving slope positions. Reasons or hypotheses for this concentration of uncommon and rare species occurrences in these regions are numerous (some of which are identified in this thesis), yet still not entirely known. However, it is unlikely that a "one shoe fits all" hypothesis will unravel this cryptic ecological message, in which we find species of the same guild selecting for differing stand variables within water receiving slope positions. As with most scientific inquiries, more questions than answers have emerged. From the perspective of managing (i.e. doing forestry) in these old-growth forests, consideration of this study must be put into context. First, we must remember that this Radies 38 landscape study measured 37 species out of a total that is well into the thousands, and that many taxonomists other than lichenologists have yet to set foot in this forested system. For this reason alone, the size of refuge needed for the conservation of biodiversity in these inland old-growth temperate rainforests is nothing short of a well informed guess. Therefore, it is best that we err on the side of caution and recommend that old-growth forest in water receiving positions be conserved (and appropriately represented by stand composition and structure) until further scientific research indicates otherwise. LITERATURE CITED Anonymous. 1998. Vegetation Resources Inventory: Implementation strategy to integrate management, provincial and national inventories. Resources Inventory Branch, B.C. Ministry of Forests, Victoria, B.C. Available at: . Anonymous. 1999. Standards for Predictive Ecosystem Mapping, version 1.0. The resources inventory committee, Province of B.C, Victoria, B.C. Available at: . Anonymous. 2004. Legal establishment of Natural Disturbance Unit's and subsequent biodiversity objectives within the Prince George Timber Supply Area. Government of B.C. Integrated Land Management Bureau. Available at: Radies 39 Anonymous. 2005. Old forest retention results to March 31, 2005. Prince George Public Advisory Group. Unpublished report. B.C. Ministry of Forests. Prince George, B.C. Arsenault, A. and T. Goward. 2000. Ecological characteristics of inland rainforests. Ecoforestry 15: 20-23. Beaty, R.M. and A.H. Taylor. 2001. Spatial and temporal variation of fire regimes in a mixed conifer forest landscape, Southern Cascades, California, USA. Journal of Biogeography 28:955-966. Benson, S., and D.S. Coxson. 2002. Lichen colonization and gap structure in wettemperate rainforests of northern interior British Columbia. The Bryologist 105:673-692. Bergamini, A., Scheidegger, C, Stoffer, S., Carvalho, P., Davey, S., Dietrich, M., Dubs, F., Farkas, E., Groner, U., Karkkainen, K. Rkkainen, K., Keller, C, Lokos, L., Lommi, S.,Maaguas, C, Mitchell, R., Pinho, P., Rico, V.J., Aragon, G., Truscott, A.M., Wolesly, P. and A. Watt. 2005. Performance of macrolichens and lichen genera as indicators of lichen species richness and composition. Conservation Biology 19:1051-1062 Berglund, H., and B.G. Jonsson. 2005. Verifying an extinction debt among lichens and fungi in northern Swedish boreal forests. Conservation Biology 19:338-348. Budel, B., and O.L. Lange. 1991 Water status of green and blue green phycobiots in lichen thalli after hydration by water vapour uptake? Do they become turgid? BotanicaActa 104:361-366. Radies 40 Campbell, J., and D.S. Coxson. 2001. Canopy microclimate and arboreal lichen loading in subalpine spruce-fir forest. Canadian Journal of Botany 79:537-555. Campbell J, and A.L. Fredeen. 2004. Lobaria pulmonaria abundance as an indicator of macrolichen diversity in Interior Cedar hemlock forests of East-Central British Columbia. Canadian Journal of Botany 82:970-982. Castellon, T.D., and K.E. Sieving. 2007. Patch network criteria for dispersal-limited endemic birds of South American temperate rain forest. Ecological Applications 17:2152-2163. Coast Information Team. 2004. EBM Planning Handbook. Available at . Coxson, D.S., and S.K. Stevenson. 2007. Growth rate responses of Lobaria pulmonaria to canopy structure in even-aged and old-growth cedar-hemlock forests of centralinterior British Columbia, Canada. Forest Ecology and Management 242:5-16. Craighead, L. and B. Cross. 2004. A conservation area design for the inland temperate rainforest of Canada. Available at: DeLong, S.C. 1998. Natural disturbance rate and patch size distribution of forests in northern British Columbia: Implications for forest management. Northwest Science 72:35-48. DeLong, S.C. 2003. A field guide for site identification and interpretation for the southeast portion of the Prince George Forest Region. Research branch, B.C. Ministry of Forests, Victoria, B.C. Land Management Handb. No. 51 Radies41 DeLong, S.C., Harrison, M., and P.Burton. 2004. Assessing the relative quality of oldgrowth forest: An example from the Robson Valley, British Columbia. BC Journal of Ecosystem Management 4:1-18 DeLong, S.C. 2007. Implementation of natural disturbance-based management in northern British Columbia. Forestry Chronicle 83:326-227. Drever, C.R., Peterson, G., Messier, C, Bergeron, Y., and M. Flannigan. 2006. Can forest management based on natural disturbances maintain ecological resilience? Canadian Journal of Forest Research 36:2285-2299. Ecosystem Representation in Tree Farm License #30. Final Report March 31, 2006 Forest Ecosystem Solutions Ltd. #227-998 Harbourside drive North Vancouver BC V7P 3R5. Eng, M. 2000. Fire analysis for the Robson Valley Forest District. Unpublished Report. Research Branch. Ministry of Forests. Victoria, B.C. . Fazey, I, Fisher, J., and D.B. Lindenmayer. 2005. What do conservation biologists publish? Biological Conservation 124:63-73. Glavich, D.A., L.H. Geiser, and A.G. Mikulin. 2005. Rare epiphytic coastal lichen habitats, modeling, and management in the Pacific Northwest. The Bryologist 108:377-390. Goward, T. 1994. Notes on old-growth dependent epiphytic macrolichens in inland British Columbia, Canada. Acta Botanica Fennica 150:31-38. Goward, T. 2003. Lichens of the Robson Valley provincial parks: Grizzly-Sugarbowl, West Twin Creek, Slim Creek. Unpublished report. B.C. Parks. Prince George. Radies 42 Goward, T., and A. Arsenault. 2000. Inland old-growth rain forests: safe havens for rare lichens? In Darling, L.M. (editor). Proceedings, Biology and Management of Species and Habitats At Risk, Kamloops, B.C., 15-19 February 1999. B.C. Min. Environ., Lands and Parks, Victoria, B.C. and University College of the Cariboo, Kamloops, B.C:759-766. Goward, T. and A. Arsenault. 2003. Notes on the Populus "dripzone effect" in well ventilated stands in humid inland east-central British Columbia. The Canadian Field-Naturalist 117:61-65. Goward, T., McCune, B., and D.V. Meidinger. 1994. The Lichens of British Columbia Illustrated Keys Part 1: Foliose and Squamulose Species. B.C. Ministry of Forests, Research Branch. Special Report Series. Goward, T., and T. Spribille. 2005. Lichenological evidence for the recognition of inland rainforests in western North America. Journal of Biogeography 32:12091219. Graff, J.E., R.K. Hermann, and J.B. Zaerr. 1999. Ionic balance and organic acids in western redcedar, western hemlock, and Douglas-fir seedlings grown in low- and high-N soils. Canadian Journal of Forest Research 29:669-678. Grumbine, R.E. 1994. What is Ecosystem Management? Conservation Biology 8:27-38. Hilmo, O., and S.M. Sastad. 2001. Colonization of old-forest lichens in a young and an old boreal Picea abies forest: an experimental approach. Conservation Biology 102:251-259 Hutchinson, M. F. 1995. Interpolating mean rainfall using thin plate smoothing splines. International Journal of Geographical Information Systems 9:385-403. Radies 43 Hyvarinen, M., P. Halonen, and M. Kauppi. 1992. Influence of stand age on the epiphytic lichen vegetation in the middle-boreal forests of Finland. The Lichenologist 24:165-180. Iverson, L.R., Martin, E.D., Scott, C.T. and A. Prasad. 2000. A GIS-derived integrated moisture index to predict forest composition and productivity of Ohio forests (U.S.A.). Forest Ecology and Management 129:93-100 Johnson, J.B. and Omland, K.S. 2004. Model selection in ecology and evolution. Trends in Ecology and Evolution 19:101-108. Kalwij, J.M., Wagner, H.H., and C. Scheidegger. 2005. Efferts of stand-level disturbances on the spatial distribution of a lichen. Ecological Applications 15:2015-2024. Ketcheson, M.V., T.F. Braumandl, D. Meidinger, G. Utzig, D.A. Demarchi, and B.M. Wikeem. 1991. Interior cedar-hemlock zone. In Ecosystems of British Columbia, Meidinger, D., and Pojar, J., eds. Ch 11. BC Ministry of Forests, Special Report Series No. 6. Kew, M. 1992. Salmon availability, technology, and cultural adaptation in the Fraser River Watershed. In: A Complex Culture of the British Columbia Plateau. Traditional Stl'atl'imx Resource Use. Edited by B. Hayden. UBC Press, Vancouver, B.C. Kitching, R.L., Bergelson, J.M., Lowman, M.D., Mclntyre, S., and G. Carruthers. 1993. The biodiversity of arthropods from Australian rainforest canopies: General introduction, methods, sites and ordinal results. Australian Journal of Ecology 18.181-191. Radies 44 Lertzman, K., Gavin, D., Hallett, D., Brubaker, L., Lepofsky, D. and R. Mathewes. 2002. Long-term fire regime estimated from soil charcoal in coastal temperate rainforests. Conservation Ecology 6 . Liden, M. and O. Hilmo. 2005. Population characteristics of the suboceanic lichen Platismatia norvegica in core and fringe habitats: Relations to macroclimate, substrate, and proximity to streams. The Bryologist 108:506-517. McCune, B., and M.J. Mefford. 1999. PC-ORD. Multivariate analysis of ecological data. Version 4.0. MJM Software, Glenden Beach, Oregon, U.S.A.. McCune, B., R. Rosentreter, J.M. Ponzetti, and D.C. Shaw. 2000. Epiphyte habitats in an old conifer forest in western Washington, U.S.A. The Bryologist 103:417-427. McCune, B., J. Hutchinson, and S. Berryman. 2002. Concentration of rare epiphytic lichens along large streams in a mountainous watershed in Oregon, U.S.A. The Bryologist 105:439-450. Menard, S. 2002. Applied Logistic Regression Analysis. Second Edition. Sage Publications Inc., California. Moola, F.M., D. Martin, B. Wareham, J. Calof, C. Burda, and P. Grames. 2004. The coastal temperate rainforests of Canada: The need for ecosystem-based management. Biodiversity 5:9-15. Munoz, J. and A. Felicisimo. 2004. Comparison of statistical methods commonly used in predictive modelling. Journal of Vegetation Science 15:285-292. Negi, R.N. and M. Gadgil. 2002. Cross-taxon surrogacy of biodiversity in the Indian Garhwal Himalaya. Biological Conservation 105:143-155. Radies 45 Niemi, G.J. and M.E. McDonald. 2004. Application of ecological indicators. Annual Review of Ecology, Evolution, and Systematics 35:89-111. Pastor, J., M. and W.M. Post. 1986. Influence of climate, soil moisture, and succession on forest carbon and nitrogen cycles. Biogeochemistry 2:3-27. Price, K., Holt, R.F., and L. Kremsater. 2008. How much is really enough?: Informing old growth targets with threshold science. Conservation Biology (in submission). Radies, D. N., and D.S. Coxson. 2004. Macrolichen colonization on 120-140 year old Tsuga heterophylla in wet temperate rainforests of central-interior British Columbia: a comparison of lichen response to even aged versus old-growth stand structures. The Lichenologist 36:235-248. Radies, D.N., Coxson, D.S., Johnson, C.J., and K. Konwiki. 2008. Not all old-growth is equal: Ecological attributes and lichen biodiversity in an inland temperate rainforest landscape. Ecological Applications. In Submission. Reynolds, R.R. 1997. Climatic data summary for the biogeoclimatic zones of British Columbia. British Columbia Ministry of Forests, Research Branch. Unpublished Report. Victoria, B.C. Rosso, A.L., McCune, B., and T.R. Rambo. 2000. Ecology and conservation of a rare, old-growth-associated canopy lichen in a silvicultural landscape. The Bryologist: 103:117-127. Sanborn, P., M. Geertsema, A.J.T. Jull, and B. Hawkes. 2006. Soil and sedimentary charcoal evidence for Holocene forest fire in an inland temperate rainforest, eastcentral British Columbia, Canada. Holocene 16:415-427. Radies 46 Sillett, S.C., B. McCune, J.E. Peck, T.R. Rambo, and A. Ruchty. 2000. Dispersal limitations ofepiphytic lichens result in species dependent on old-growth forests. Ecological Applications 10:789-799. Spies, T.A., and J.F. Franklin. 1991. The structure of natural young, mature, and oldgrowth Douglas-Fir Forests in Oregon and Washington. Published by: U.S. Department of Agriculture, Forest Service Pacific Northwest Research Station Portland, Oregon General Technical Report PNW-GTR-285. Spies, T.A, M.A. Hemstrom, A. Youngblood, and S. Hummel. 2006. Conserving oldgrowth forest diversity in disturbance-prone landscapes. Conservation Biology 20:351-362. Stevenson, S.K., H.M. Armleder, M.J. Ml, D.G. King, B.N. McLellan, and D.S. Coxson. 2001. Mountain caribou in managed forests: recommendations for managers: Second Edition. B.C. Min. Environ., Lands, and Parks. Wildlife Rep. No. R-26. Victoria, B.C. Stevenson, S.K., and D.S. Coxson. 2008. Edge effects on Lobaria retigera in B.C.'s inland temperate rainforest. Forest Ecology and Management (in-submission). Sundberg, B., K. Palmqvist, P.A. Esseen, and K.E. Renhorn. 1996. Growth and vitality ofepiphytic lichens. II. Modelling of carbon gain using field and laboratory data . Oecologia 109:10-18. Turner, G.T. 1989. Landscape Ecology: The Effect of Pattern on Process. Annual Review of Ecology and Systematics 20:171-197. Walker, B.H. 1992. Biodiversity and ecological redundancy. Conservation Biology 6:18-23. Radies 47 Wells, R.W., Bunnell, F.L., D. Haag, and G. Sutherland. 2003. Evaluating ecological representation within differing planning objectives for the central coast of British Columbia. Canadian Journal of Forest Research. 33: 2141-2150. Zinko, U., Seibert, J., Dynesius, M., and C. Nilsson, C. 2005. Plant species numbers predicted by a topography-based groundwater flow index. Ecosystems 8:430441. Radies 48 TABLES Table 1-1. Species Dry Stand Plots Wet Stand Plots (27 sites total) (26 sites total) Number of site occurrences and occurrences by abundance scale: very low (VL), low (L), common (C) CHLOROLICHENS # of dry sites (VL, L,C) Cavernularia hultenii 7 (5,2,0) 14 (7,3,4) Cetraria cetroides 0 1 (1,0,0) Hypogymnia austerodes 2 (2,0,0) 1 (1,0,0) H. bitteri 18(3,7,8) 17 (2,6,9) H Jmshaugii 0 1 (1,0,0) H. metaphysodes 19(4,4,11) 14 (7,3,4) H .occidentalis 26(0,1,25) 26 (0,0,26) H .oroborealis 3 (3,0,0) 2 (2,0,0) H. physodes 27 (0,0,27) 26 (0,0,26) H. rugosa 19(4,11,4) 20(4,10,6) H. tubulosa 27 (0,4,23) 26 (0,2,24) H. vittata 23 (4,9,10) 25 (0,4,21) Melanelia spp. 15(11,3,1) 15(7,7,1) Parmelia spp. 27 (0,0,27) 26 (0,0,26) Parmeliopsis ambigua 26(3,11,12) 26(1,9,16) Radies 49 # of wet sites (VL, L, C) P. hyperopta 27 (2,6,19) 26(1,4,21) Platismatia glauca 27 (0,0,27) 26 (0,0,26) P. norvegica 18(5,8,5) 17(5,7,5) Tuckermannopsis chlorophylla 27(1,9,17) 26(2,1,23) T. orbata 7(3,3,1) 2 (2,0,0) Vulpicida pinastri 12 (6,4,2) 14(7,4,3) CYANOLICHENS # of dry sites (VL,L,C) # of wet sites (VL,L,C) Leptogium burnetiae 0 1 (1,0,0) L. saturninum 2(1,0,1) 4 (2,0,2) Lobaria hallii 0 3 (3,0,0) Lobaria pulmonaria * 27(1,1,25) 26 (0,0,26) Lobaria retigera 2 (2,0,0) 12 (6,2,4) Lobaria scrobiculata 19(9,7,3) 23(1,10,11) Nephroma bellum 11(4,5,2) 18(7,7,4) Nephroma helveticum 23 (4,6,13) 26 (0,3,23) Nephroma isidiosum 11(4,6,1) 23 (6,5,12) Nephroma occultum 3 (3,0,0) 11(10,1,0) Nephroma parile 24(4,7,13) 26 (0,3,23) Nephroma resupinatum 1 (0,1,0) 1 (0,0,1) Peltigera collina 1 (0,1,0) 5 (3,1,0) Polychidium dendriscum 0 7 (2,3,2) Pseudocypellaria anomala 19 (8,8,3) 26(3,6,17) Sticta fuliginosa 17(10,4,3) 23 (2,4,17) Radies 50 Sticta limbata 0 2 (2,0,0) Sticta oroborealis 3 (1,2,3) 12 (5,4,3) Sticta wrightii 2 (2,0,0) 1 (1,0,0) Radies 51 Table 1-2. Variable Coefficient SE t P CONSTANT 40.199 18.270 2.200 0.033 Temperature -6.139 1.726 -3.557 < 0.001 Precipitation -13.848 5.751 -2.408 0.021 Solar Loading 1.240 0.751 1.652 0.106 Basal Area (Hw) 0.180 0.127 1.420 0.163 Basal Area (Cw) 0.110 0.101 1.089 0.283 Basal Area (Sx) -0.146 0.137 -1.068 0.292 Basal Area (Sf) 0.209 0.199 1.049 0.301 Relative Soil Moisture -0.806 0.943 -0.855 0.398 Canopy Openness -0.301 0.479 -0.628 0.534 -61.304 22.412 -2.733 0.009 Relative Soil Moisture 3.760 1.158 3.247 0.002 Temperature 6.139 2.118 3.002 0.005 Canopy Openness 1.58 0.588 2.689 0.01 Precipitation 17.019 7.060 2.411 0.02 Basal Area (Sx) -0.383 0.168 -2.280 0.028 Basal Area (Hw) 0.263 0.155 1.694 0.098 Solar Loading 1.429 0.992 1.551 0.129 Basal Area (Sf) 0.322 0.245 1.317 0.195 AXIS1 AXIS 2 CONSTANT Radies 52 Basal Area (Cw) 0.115 0.124 Radies 53 0.928 0.359 Table 1-3. Stand Stand Variables Type Mean Minimum Annual Solar Loading % Canopy Temperature Mar.-Oct. Precipitation (kWh/m2)1 Openness2 (°C) ' Temperature (mm) TO1Mean SD Mean SD Mean SD Mean SD Mean SD DRY 3~1 04 LI 04 831.3 4L2 1208.7 238.0 8^6 ~2X WET 3.5 0.4 1.8 0.3 811.5 32.9 1218.0 108.2 14.1 6.3 Stand Stand Variables Type Stand Basal Basal Area Basal Area- Basal Area- Basal Area- Area Fir Cedar Hemlock Spruce (m 2 /ha) 2 (m2/ha)2- (m 2 /ha) 2 ' (m 2 /ha) 2 (m 2 /ha) z Mean SD Mean SD Mean SD Mean sT3 Mean SD~ DRY 76.6 27/7 23 2^8 348 4L0 16l 1A9 8^4^ 9^ WET 80.2 48.8 4.0 4.9 43.1 64.9 13.7 21.3 11.0 11.5 Radies 54 Table 1-4. Species Best Model Sets (>0.1 AICcw) AICcw ROC Cavernularia hultenii LEAD (Hw) X BRSM (wet) 0.972 0.856 Lobaria retigera LEAD (Hw) X OPEN X TEMP 0.607 0.914 Nephroma isidiosum BRSM (wet) X OPEN X TEMP X PRECIP 0.419 0.918 Nephroma occultum LEAD (Cw) X BRSM (wet) 0.881 0.815 Platismatia norvegica LEAD (Hw) X TEMP 0.525 0.718 Peltigera collina BRSM (wet) X TEMP 0.333 0.737 Sticta fuliginosa OPEN X TEMP 0.458 0.850 Sticta oroborealis LEAD (Hw) X OPEN X TEMP 0.956 0.956 Radies 55 Table 2-1. Species Abbreviation CHLOROLICHENS Cavernularia hultenii CAVEHULT Cetraria cetroides CETRCETR Hypogymnia austerodes HYPOAUST H. bitteri HYPOBITT H .imshaugii HYPOIMSH H. metaphysodes HYPOMETA H .occidentalis HYPOOCCI H .oroborealis HYPOOROB H. physodes HYPOPHYS H. rugosa HYPORUGO H. tubulosa HYPOTUBU H. vittata HYPOVITT Melanelia spp. MELASPP. Parmelia spp. PARMSPP. Parmeliopsis ambigua PARMAMBI P. hyperopta PARMHYPER Platismatia glauca PLATGLAU P. norvegica PLATNORV Radies 56 Tuckermannopsis chlorophylla TUCKCHLO T. orbata TUCKORBA Vulpicida pinastri VULPPINA CYANOLICHENS Leptogium burnetiae LEPTBURN L. saturninum LEPTSATU Lobaria hallii LOBAHALL Lobaria pulmonaria LOBAPULM Lobaria retigera LOBARETI Lobaria scrobiculata LOBASCRO Nephroma bellum NEPHBELL Nephroma helveticum NEPHHELV Nephroma isidiosum NEPHISID Nephroma occultum NEPHOCCU Nephroma parile NEPHPARI Nephroma resupinatum NEPHRESU Peltigera collina PELTCOLL Polychidium dendriscum POLYDEND Pseudocypellaria anomala PSEUANOM Sticta fuliginosa STICFULI Sticta limbata STICLIMB Sticta oroborealis STICOROB Sticta wrightii STICWRIG Radies 57 Table 2-2. Variable Description BSMR Broad Soil Moisture Regime measured as Dry (0) or Wet (1) OPEN Average densiometer measurements in lichen plot TEMP Average minimum temperature between March to October PRECIP Average (absolute) annual precipitation SOLAR Measure of solar insolence, based on aspect, topography, and elev LEAD Leading tree species. Cw (0), Hw (1), Sx (2) Radies 58 Table 2-3. 1. BSMR (2) X OPEN X TEMP X PRECIP E. BSMR (2) X OPEN X TEMP X SOLAR 3. BSMR (2) X PRECIP 4. BSMR (2) X TEMP 5. LEAD (3) X BSMR (2) 6. LEAD (3) X OPEN X TEMP 7. LEAD (3) X OPEN X PRECIP 8. LEAD (3) X OPEN X SOLAR 9. LEAD (3) X PRECIP 10. LEAD (3) X TEMP 11. OPEN X PRECIP X SOLAR X TEMP 12. OPEN X PRECIP 13. OPEN X TEMP Radies 59 Table 2-4. DRY SMR WET SMR Cedar 11 11 Hemlock 10 6 Spruce 6 11 27 sites 26 sites Radies 60 Table 2-5. DrySMR Leading Wet SMR Tree Stand % Total % Protection % Total % Protection Species Age (% NHLB) (Park, OGMA) (%NHLB) (Park, OGMA) Cedar 140-250 8.5(51) 0.81 (0.41,0.4) 5.9(27) 0.65(0.53,0.12) 250+ 5.8 (64) 1.19(0.31,0.88) 7.1(38) 1.44(0.27,1.17) Hemlock 140+ 4.25 (59) 1.1(0.4,0.7) 1.6(44) 0.9(0.7,0.2) Spruce 140+ 13.2 (50) 2.0(1.5,0.5) 12.7(45) 2.62(2.1,0.5) Other 140 + 3.5 (63) 0.65(0.5,0.15) 1.7(47) 0.32(0.3,0.02) 35.3 (54.7) 5.8(3.1,2.7) 29 (39.7) 5.9 (3.9, 2.0) Total Radies 61 Figure 1-1. •-Zf-i."*. •! ^ • ^ ^ xv ' **" a>n-ccdll Riclse ?%•? J%'' '*% .. , %, Insert .V Dnscoll Ridge - Souili Slope LEGEND • B F a r l y Scroll <40ye.irsl Cudui. Di\ • H Ccxhii, Wet , Hemlock. Drv M I tan toe k. Wei Spruce. Dry • • Spruce, Wet o Field Inspections •f^f^ 8 "-*"^. Radies 61 Figure 1-2. 1 1 1 1 A AA A AA Lobaria pulmonaria A^A A A A * "* * ' * . A * » » *' * * 1 ** * A - A A A A A A A A ±A A A Sticta fulginosa A AA A A * jA A A A * • * A A A • A ' . • A A • A A .A ' A A ' A A * . A A A A A * * A A * Nephroma occultum A - A * A : * A A * A ' AA A . -£^\ AA A A * • 4 f A A A * * A Pseudocyphellaria A anomala - A A * f " A ' ' A ^ A ^ * • * A A A A A A - * . A 1 A A A A A A A A A -1.0 -0.5 ' A Nephroma isidiosum A • ». * * * ^ A A A ' ' ' A * * * A Lobaria retigera .*. / ± A A A A A A > V- - AA ^ A4 ^ 0.0 A A 0.5 1.0 -0.5 Axisl Radies 62 0.0 A A 0.5 1.0 1.5 Figure 1-3. 4 2 0 -2 -4 4 2 O -2 -4 4 2 0 -2 aria retlgera Lobaria • • , , • Nephroma • _ + • J- H h- Nephroma occultum i * 1 Platismatia Peltigera $ • 4 2 O -2 -4 1 1 I $ 1- H h- H H colllna ^ * 1 H Sticta fuliginosa Sticta oroborealis { 1 1- J * J 1 H 4 2 0 -2 -4 1- h H 4 2 0 -2 -4 H norvegica H 4 2 0 -2 I ? I isidiosum 3 H 4 2 O -2 -4 I 1 1 1 1 1 Cavern ularia hultenii dry w e t Denslo Temp Preclp Solar cw Environmental variables Radies 63 H X hw sx Figure 2-1. ./"f.-v " V. /•• >'•-' Interior Cedar Hemlock > -*; V -y , , J '. >Woaeoclimatic zone "wk" and "vk" \ ' M f c . , , " •••• J ., •;,- i -J :ttf >* > i* f • ^>:v:-r.„\f. ^.j *.-#. ..<*/ • • • w i ' : n . ••y • * • * / •• ^ ^ * J - - § - •, • *• •«" * . " " • ' • > - • . •• 4 :,^#/Jr^ 1 Radies 64 Figure 2-2. * < * - . , - - --, Interior Cedar Hemlock latic zone "wk" and "vk" sub variants logged areas (approx last SO years) ' ••'• * i y .•J #fef!*«J iipB . . . . . • • - * * ' t / * * • " ' ' •••• - * ' - , : * ! / " " • i f . « • v*»- ; ... *>£")•• I ,-? •*,-S«£5>4 \:4~: Radies 65 Figure 2-3. Interior Cedar Hemlock biogeoclimatic zone "wk" and "vk" sub variants | Age Class 9 Age Class 8 '">?••' •'-- -•jesss. Protected Areas r favors,'. .••*•.:".'•. •1* * ; . : - r - r; o- • 'J, •-•'-' U' . e," .>: ^ ,f: r i ^ i W ? •*# • -**$- . j r . -»ft. • * m ' i -. Radies 66 Figure 2-4. 0~ 4" wSS^ \ j » j , SflS vk and IOTvk2 IOT vk2 ' : Bio^coclimatk si itfc suKkt is k****-! Mm** < # !»$$# w«W- if mi mm , ?3Lrfl%lri» . t^m% m •:sm ~4\*< ' • * & .iiiii 3KJW Radies 67 Figure 2-5. Radies 68 Figure 2-6. Other Species >140 years Piceii glaucn \ engclmuniiii, >140 years Thuja plicatu, >140 years Tsugu heterophyila, >140 years Radies 69 Figure 2-7. Om 20m 40m Om 10m 30m 50m Radies 70 70m 90m 100m Figure 2-8. Matrix Coding I Presence • Absence Information Remaining {%) 0 25 50 75 100 -Q •cE tf -CE ^ £ •tf i_ jOQ.=jo.jQ.oicLO-liJlij •••• ••••••••••• • zzz • ••• • • •••••••• •• • • • • Pi • • IB • flflflflflflfl • • • • • • • •B •B DDflfl • r u n ••«Z BBBBBBB B I BB • • • • • BB BBBBBBB BB • • • mi1 BBBBBBB BBBBBBB BB • • • • • • BBBBBBB BB • B*II' IB BBBBBBB BB U, BBBBBBB B ^ BBBBBBB B ••• • • •••_!• 1 BBBBBBB BB • • • • • • fl BBBBBBB BB ••• • B BBBBBBB BB • • • • • • •1 BBBBBBB BB • • 1 •DDDI ••DDE • • • • •• • 1 •• • ••• •• •• • • •• ••••••• • • • • • • • • • • BBC BBBBBBB BBB BBBBBBB B BBBBBBB B BBBBBBB • •nar •••• • •_• • • ....... ....... ....... .....BB . • • • ••••••• • ami i :>ij ••••• •• •• • ••• ••• • •• • •••• • • ••• • • ....... I . B • B B B • • •••a... • ....... • BBBBBBB • BBBBBBB B BBBBBBB • BBBBBBB E BBBBBBB fl ••••••• • •••••BB fl Figure 2-9. 0 2.5 5 10 15 20 25 Tsuga heterophylla 'Dry' Thuja plicata 'Dry' Thuja plicata 'Wet' Tsuga heterophylla 'Wet' Radies 72 Figure 2-10. 0 2.5 5 ™ ™ •" 10 — 15 20 | 25 1^—hn | Very Low r~iLo» ^ ^ | Medium H Radies 73 Very High Figure 2-11. 0 2.5 5 10 15 20 25 I | Very Low r i LOW i m Medium HHigh ^ | Very High Radies 74 Figure 2-12. 0 2.5 5 10 15 20 25 | | Very Low [ I Low H B Medium • B Very High Radies 75