Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Vascular plant response to slashburning and clearcutting in central British Columbia : a 20 year study… Chandler, Julia Rae 2014

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

Item Metadata

Download

Media
24-ubc_2014_september_chandler_julia.pdf [ 4.79MB ]
Metadata
JSON: 24-1.0167539.json
JSON-LD: 24-1.0167539-ld.json
RDF/XML (Pretty): 24-1.0167539-rdf.xml
RDF/JSON: 24-1.0167539-rdf.json
Turtle: 24-1.0167539-turtle.txt
N-Triples: 24-1.0167539-rdf-ntriples.txt
Original Record: 24-1.0167539-source.json
Full Text
24-1.0167539-fulltext.txt
Citation
24-1.0167539.ris

Full Text

      VASCULAR PLANT RESPONSE TO SLASHBURNING AND CLEARCUTTING IN CENTRAL BRITISH COLUMBIA: A 20 YEAR STUDY OF PLANT FUNCTIONAL TYPE RESILIENCE    by  JULIA RAE CHANDLER  B.A., Simon Fraser University, 2004 M.Sc., Simon Fraser University, 2007      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)         July 2014  © Julia Rae Chandler, 2014 ii  Abstract How resilience is understood and measured has become increasingly challenging for ecologists, particularly as terrestrial ecosystems are undergoing radical change as climate changes.  This body of work proposes a specific approach to studying resilience and applied it to Interior Cedar-Hemlock (ICH), Sub-Boreal Spruce (SBS) and Engelmann Spruce-Subalpine Fir (ESSF) forests extending across central British Columbia, Canada.  Repeated measurements (% cover and height) of vascular plants were collected between 1981 and 2008 (1, 2, 3, 5, 10 and 20 years after clearcutting and slashburning) in permanent research installations.  Individual species sensitive to the forestry treatment (recorded exclusively pre-burn) included Rhododendron albiflorum, Menziesia ferruginea and Prosartes hookeri in the ICH; Rubus pedatus in the SBS; and Orthilia secunda, Listera cordata and Moneses uniflora in the ESSF.  Post-burn shifts in species dominance consisted of substantial loss of Abies lasiocarpa, Oplopanax horridus and Listera cordata, and increases in Alnus spp., Salix spp., Epilobium spp. and Calamagrostis spp., indicating possible transition from conifer forest to mixed forest or open meadow ecosystems at several study sites. To overcome the difficulty of evaluating ecosystem resilience from measurements of 183 individual species recorded in experimental plots, I created plant functional types (PFTs) based on 15 common plant traits.  PFTs were determined by grouping together plants that behave in similar ways or produce similar outcomes despite having different physical characteristics or evolutionary paths.  PFT models of abundance and richness along gradients of soil nitrogen and fire severity over time indicated linear and non-linear response trends, and lasting and temporary effects.  Structural equation modeling (SEM) was used to measure the relative importance of factors driving the responses observed.  The SEM indicated that mean annual precipitation (MAP) negatively influenced fire severity; mean annual temperature (MAT) positively influenced fire severity and soil nutrients; and MAP and MAT directly and/or indirectly influenced most PFTs.  My research suggests that clearcutting and slashburning do not alone alter the diversity or function of mesic ESSF, SBS and ICH forests; however, past and future anthropogenic disturbances combined with non-historical climate and interrelated edaphic factors may place long-term stability of these ecosystems at risk.    iii  Preface In 2007 and 2008, I, accompanied by Evelyn Hamilton, Suzanne Simard, and Sybille Haeussler as well as several field assistants, revisited seven permanent research installations to locate existing plots, collect 20-year post-burn vegetation data measurements, and collect soil samples.  The primary tasks of this dissertation, works to call my own, has been creating a plant trait dataset and developing plant functional types of central B. C. (Chapter 2); measuring resistance and resilience of ICH, ESSF and SBS forest 20 years after clearcutting and prescribed burning in central B. C. (Chapter 3); investigating response of plant functional types of central B. C. to gradients of fire severity and soil nitrogen over time (Chapter 4); and developing a structural equation model to quantify the relative importance of factors associated with resilience of conifer forest in central B. C. (Chapter 5).      iv  Table of Contents Abstract .................................................................................................................................................. ii Preface................................................................................................................................................... iii Table of Contents ....................................................................................................................................iv List of Tables .......................................................................................................................................... vii List of Figures ........................................................................................................................................ viii List of Abbreviations................................................................................................................................ ix Glossary .................................................................................................................................................. x Acknowledgements ................................................................................................................................. xi Dedication.............................................................................................................................................. xii Chapter 1. General introduction .............................................................................................................. 1 Approaches for measuring the state of managed forests in central British Columbia ..................... 3 Reducing forest ecosystem complexity using plant functional types (Chapter 2) ........................ 3 Measuring stability with vascular plant species and plant functional types (Chapter 3) .............. 4 Measuring resilience of vascular plants along ecological gradients (Chapter 4) .......................... 6 Measuring resilience with structural equation modeling (Chapter 5) ......................................... 7 Thesis objectives ........................................................................................................................... 8 Study area ..................................................................................................................................... 9 Chapter 2. Complex dataset reduction: 183 vascular plants characterized by 9 functional types ............ 15 Methods ..................................................................................................................................... 16 Field procedures ..................................................................................................................... 16 Data ........................................................................................................................................ 17 Statistical analysis ................................................................................................................... 17 Results ........................................................................................................................................ 18 9 plant functional types ........................................................................................................... 18 Discussion ................................................................................................................................... 19 Plant functional types and their ecology .................................................................................. 19 Conclusion .................................................................................................................................. 21   v  Chapter 3. Measuring resistance and resilience in ESSF, ICH and SBS forests of central British Columbia 23 Methods ..................................................................................................................................... 25 Measuring resistance .............................................................................................................. 25 Measuring resilience ............................................................................................................... 25 Results ........................................................................................................................................ 26 Resistance ............................................................................................................................... 26 Resilience ................................................................................................................................ 28 Discussion ................................................................................................................................... 30 Resistance ............................................................................................................................... 30 Resilience ................................................................................................................................ 31 Conclusion .................................................................................................................................. 32 Chapter 4. Plant functional type response over time to gradients of fire severity and soil nitrogen ....... 43 Methods ..................................................................................................................................... 45 Field procedures - measuring fire severity ............................................................................... 45 Field procedures - measuring soil nutrients ............................................................................. 45 Laboratory analysis - measuring soil %N and %C ...................................................................... 46 Statistical analysis ................................................................................................................... 46 Results ........................................................................................................................................ 47 PFT response to fire severity gradient ..................................................................................... 47 PFT response to soil nitrogen gradient..................................................................................... 48 Curvilinear models .................................................................................................................. 48 Discussion ................................................................................................................................... 49 Fire severity ............................................................................................................................ 50 Soil nitrogen gradient .............................................................................................................. 51 Time ........................................................................................................................................ 52 Conclusion .................................................................................................................................. 52   vi  Chapter 5. Modeling the relative importance of environmental factors controlling ecosystem resilience in ESSF, ICH and SBS forests of central British Columbia .................................................................. 57 Methods ..................................................................................................................................... 57 Structural equation modeling features .................................................................................... 57 The structural equation modeling method .............................................................................. 59 The measurement and structural model .................................................................................. 59 Hypotheses ............................................................................................................................. 60 Assessing model fit .................................................................................................................. 61 Data set................................................................................................................................... 62 Results ........................................................................................................................................ 62 Discussion ................................................................................................................................... 64 Influences of precipitation and temperature on fire severity and soil nutrients ....................... 64 Influences of fire severity and soil nutrients on plant functional types ..................................... 65 Influences of precipitation and temperature on plant functional types .................................... 66 Conclusion .................................................................................................................................. 68 Chapter 6. General discussion and conclusions ...................................................................................... 73 Limitations of this study .............................................................................................................. 76 Suggestions for future research ................................................................................................... 78 Bibliography .......................................................................................................................................... 80 Appendix A. Species list of 183 vascular plants from 16 study sites in central British Columbia .............. 94 Appendix B. Traits used to develop plant functional types ..................................................................... 97 Appendix C. Dendrogram of cluster analysis results for 183 vascular plants ......................................... 102 Appendix D. Plant functional type descriptions.................................................................................... 103 Appendix E. Plant functional type response to gradients of fire severity and soil N .............................. 104 Appendix F. Structural equation modeling (SEM) R code ..................................................................... 108 Appendix G. Covariance matrix for the variables included in the structural equation model ................ 110 Appendix H. Structural equation modeling (SEM) results ..................................................................... 111   vii  List of Tables Table 1.1. Description of the 16 study sites located in the temperate and montane coniferous forests of central interior British Columbia. .................................................................................................... 12 Table 3.1. Species recorded exclusively pre-burn (underlined) or post-burn for each PFT, noted by site. See Appendix 1 for species codes. ................................................................................................... 34 Table 3.2. Pre-burn and post-burn representatives identified as 3 most dominant species, measured by greatest total percent cover, within each PFT noted by site (species with <1% cover excluded). See Appendix 1 for species codes. ......................................................................................................... 35 Table 3.3. Mean percent cover (3.3a) and richness (3.3b) of PFT before and 17-22 years after slashburn for all five sites (n = 5). Significant differences (p<0.05) are indicated by bold p-values. ................... 36 Table 3.4. Mean percent cover (3.4a) and richness (3.4b) of PFT before and 17-22 years after slashburn for each of the three BEC zones (n = number of plots): ESSF (n = 44), ICH (n = 126), and SBS (n = 6). Significant differences (p<0.05) are indicated by bold p-values........................................................ 37 Table 3.5. Mean percent cover (3.5a) and richness (3.5b) of PFT before and 17-22 years after slashburn for five sites (n = number of plots): Goat River (n = 6, 19 yrs), Herron (n = 10, 17 yrs), Mackenzie (n = 6, 19 yrs), Otter Creek (n = 34, 20 yrs) and Walker Creek (n = 120, 21 yrs). Significant differences (p<0.05) are indicated by bold p-values. .......................................................................................... 38 Table 4.1. Wilcoxon tests with the Bonferroni correction found significant differences (p<0.05) in abundance and richness for low (I), moderate (II) and high (III) classes of % LFH consumption. Symbols are as follows: very low (─ ─); low (─); high (+); very high (+ +); not different from + or ─ (±); no difference between classes (.); PFT not recorded (na). See Appendix E for mean and standard deviation for abundance and richness of each PFT for each gradient class. ..................................... 53 Table 4.2. Wilcoxon tests with the Bonferroni correction found significant differences (p<0.05) in abundance and richness for low (I), low-moderate (II), moderate-high (III) and high (IV) classes of soil N. Symbols are as follows: very low (─ ─); low (─); high (+); very high (+ +); not different from + or ─ (±); different only from + (. +); different only from ─ (. ─); no difference between classes (.); PFT not recorded (na). See Appendix E for mean and standard deviation for abundance and richness of each PFT for each gradient class. ............................................................................................................. 54  viii  List of Figures Figure 1.1. Stability agents, matter and information (Larsen 1995), can be described in terms of stability concepts (such as resistance and resilience) that are associated with a specific ecological context (Grimm and Wissel 1997). This study was interested in vascular pant species at individual species or plant functional type (PFT) levels of description providing measures of species richness (i.e. information) and abundance (i.e. matter). The reference state refers to the ecosystem as noted; and the single disturbance examined was clearcutting with slashburning. The study includes site, ecosystem and study area spatial scales; and the temporal scale refers to the year post-burn. ....... 13 Figure 1.2. Map of study locations in British Columbia; site descriptions are in Table 1.1. ...................... 14 Figure 2.1. Divisions, characteristics and species of PFTs produced by cluster analysis; GSM = ground surface material. ............................................................................................................................. 22 Figure 3.1. Mean annual temperature (MAT) and mean annual precipitation (MAP) from 1980 to 2002 at five sites: Goat River (GR), Herron (HR), Mackenzie (MA), Otter Creek (OC), and Walker Creek (WC). Data derived from ClimateBC (Wang et al. 2006)............................................................................. 39 Figure 4.1. Abundance (4.1a) and richness (4.1b) of PFTs for three classes of forest floor (LFH) consumption (10-20%; 20-30%; and >30%) modeled over time. Models with R2adj < 0.25 in all classes not reported; solid lines indicate R2adj ≥ 0.25. .................................................................................. 55 Figure 4.2. Abundance (4.2a) and richness (4.2b) of PFTs for four classes of soil nitrogen (0-0.1%; 0.1-0.2%; 0.2-0.3%; and 0.3-0.4%) modeled over time. Models with R2adj < 0.25 in all classes not reported; solid lines indicate R2adj ≥ 0.25. ........................................................................................ 56 Figure 5.1. The initial structural equation meta-model (SEMM) represents major categories of influence on relative dominance of plant functional types (PFTs) and resulting ecosystem types (represented by dashed rectangles) - conifer forest, deciduous forest, and heathland; it illustrates indirect and direct effects of climate on soil N and soil C, fire severity and PFTs. Specifically, the SEMM delineates (i) influence of precipitation and temperature on conifer forest, deciduous forest and heathland ecosystem types, (ii) influence of precipitation and temperature on fire severity and soil %N and %C , (iii) influence of fire severity on soil %N and %C , and (iv) influence of fire severity and soil %N and %C  on conifer forest, deciduous forest and heathland ecosystem types. ........................................ 70 Figure 5.2. Hypothesis for the structural equation model presented as a path diagram. A “1” indicates parameters that were fixed to “1”. PFT latent variable indictors were represented by the PFT number followed by a C (% cover) or an R (richness). Gamma and beta regression coefficients were represented by single-headed arrows from one latent variable to another; regression coefficients of indicators predicted by latent variables (lam) were represented by single-headed arrows; the covariance of errors for each latent dependent variable (psi) were represented by double-headed arrows; and the covariances among errors associated with indicators predicted from latent variables (the) were represented by double-headed arrows. ......................................................................... 71 Figure 5.3. The results of the SEM presented as the structural model includes the standardized path coefficients indicating the direction and strength of the relationships between latent variables, and the R2 value indicating the amount of variation in each endogenous latent variable that has been accounted for. Beneath the standardized path coefficients (indicating direct effects) the total effects and mediating variables are noted in italics. Model χ2 was 60.06 with 57 degrees of freedom and p = 0.365. The root mean square error of approximation (RMSEA) = 0.006 and the Tucker-Lewis Index (NNFI) = 0.999. ................................................................................................................................ 72 ix  List of Abbreviations BEC Biogeoclimatic Ecosystem Classification ESSF* Engelmann Spruce-Subalpine Fir biogeoclimatic zone EFA  exploratory factor analysis ICH* Interior Cedar Hemlock biogeoclimatic zone LFH* organic soil horizons L, F, H MAP  mean annual precipitation MAT mean annual temperature PFT* plant functional type SBS* Sub-boreal Spruce biogeoclimatic zone SMC squared multiple correlation SEM* structural equation modeling SEMM structural equation meta-model WD* woody debris  *Abbreviations with an asterisk are defined in the Glossary.  x  Glossary Engelmann Spruce-Subalpine Fir biogeoclimatic zone (ESSF): The common sub-alpine zone in the southern interior with severe climate characterized by a long cold winter and a short cool summer; Engelmann spruce and sub-alpine fir dominate wetter areas, with lodgepole pine as a pioneer after disturbance and mountain hemlock in higher snowfall areas.  Interior Cedar Hemlock biogeoclimatic zone (ICH): A zone characterized by cool and wet winters and warm and dry summers; the most productive interior forest zone with a great diversity of tree species, including western hemlock, western red cedar, and interior spruce in moist areas. Douglas-fir and lodgepole pine occur on drier sites and as seral communities. Devil's club and skunk cabbage dominate in wetter sites.  organic soil horizons (LFH): this includes the accumulation of fresh litter (L horizon); the accumulation of partly decomposed organic matter that may be occupied by filamentous fungi (F horizon); and the accumulation of decomposed organic matter with indiscernible organic structures (H horizon).   plant functional type (PFT): A group of species that has similar effects on ecosystem processes; each group describes response to, and has effects on, multiple environmental factors and ecosystem processes (Chapin et al. 1996a,b).  resilience: returning to a reference state (or dynamic) after a temporary disturbance, i.e. some change after disturbance but no shift to alternative structures and/or processes, or stability domains.  resistance: staying essentially unchanged despite the presence of disturbances, i.e. no change after sudden disturbance.  stability: an idea or collection of concepts about an ecosystem's response to stress by integrating a combination of functions that represent matter and information; stability statements should include clearly defined, measurable properties that are related to their given ecological situation.  structural equation modeling (SEM): a collection of statistical techniques that allow a set of relationships between one or more independent variables, either continuous or discreet, and one or more dependent variables, either continuous or discreet, to be examined. Both independent variables and dependent variables can be either factors or measured variables. Structural equation modeling is also referred to as causal modeling, causal analysis, simultaneous equation modeling, analysis of covariance structures, path analysis, or confirmatory factor analysis (Tabachnick and Fidell 2007).  Sub-boreal Spruce biogeoclimatic zone (SBS): Occurs on gently rolling plateaus in the Central and Boreal Interior with a wide range of temperatures; interior and White spruce and sub-alpine fir are the climatic climax, interspersed with abundant wetlands and large expanses of seral lodgepole pine resulting from past wildfire.  woody debris (WD): dead organic material of plant origin found on the forest floor such as twigs, fallen branches, bark. xi  Acknowledgements This body of work is the result of the efforts of many dedicated individuals who measured and remeasured vegetation and other attributes at long-term research installations in central B. C. throughout the past several decades.  Further acknowledgements extend to the British Columbia Ministry of Forests, Lands and Natural Resource Operations for its continued support; this extensive collection of records over the years has generated numerous technical reports, journal articles and graduate studies. My study was initiated by Evelyn Hamilton and pursued by my academic supervisor Dr. Suzanne Simard, without either of whom would I have undertaken such an endeavor.  I would also like to thank Dr. Sybille Haeussler for contributing her expertise throughout this entire process including lengthy discussions, assistance with plant identification of samples retrieved from the field, and for her exhaustive assistance with the species traits dataset. The British Columbia Ministry of Forests, Lands and Natural Resource Operations provided equipment and site logistics with special assistance from Dr. Marty Kranabetter, Karen McKeown, Bruce Rogers, Reg Newman and Stewart Philpott.  Review of thesis chapters was carried out by Dr. Suzanne Simard, Dr. Sybille Haeussler, Evelyn Hamilton, Dr. Gary Bradfield, and Dr. Michael Feller.  Funding was provided by the Forest Investment Account (FIA) Forest Science Program Award and Graduate Student Pilot Project Award; the University of British Columbia Graduate Research Assistantship and PhD Tuition Award; and the Mathematics of Information Technology and Complex Systems (MITACS) Accelerate BC Award and Travel Award.   xii  Dedication I extend special thanks and dedicate this work to all those very steadfast individuals who assisted with fieldwork including Evelyn Hamilton, Dr. Suzanne Simard, Dr. Sybille Haeussler, Megan Anderson, Dr. Joseph Hope, Brian Wallace, Ian Eichinger, and Lynn and Tony Walford.  To you all, your undiminished enthusiasm during many hours finding our way through the labyrinth of forest service roads, endurance amidst clouds of blackfly and mosquitos, and unfaltering hands to the ground searching for plot pins have led me to this end.  These were some of my most memorable days, thank you for sharing them with me.1  Chapter 1. General introduction While researchers have made long strides in defining, redefining and classifying terms of resilience, understanding the many connections between disturbance and ecosystem stability involves several remaining challenges.  Examples of such challenges include: (i) determining the controlling variables and processes acting upon ecosystem thresholds (Swift 2008); and (ii) identifying threshold limits (Zak et al. 1997) prior to an ecosystem’s shift to an alternate state.  Considering threshold theory, including reversibility (Grace and Pugesek 1997; Briske et al. 2006), questions of physiological (within species) and evolutionary (between species) adaptation become central aspects of stability assessment (Larsen 1995; Whitham et al. 2006).  Once threshold limits and ecosystem functions are better understood, we may seek approaches that take into account the influence of processes at one spatial and/or temporal scale that can influence process at another scale (Holling 1986). Currently, Northern forest ecosystems are relatively stable (Stark et al. 2006; Aubin et al. 2007; Hamilton and Haeussler 2008), but changing environmental conditions along with forest management practices may reduce the ability of these ecosystems to absorb stress and recover to the same stability domain in the future.  This may provoke a threshold trajectory, where negative feedback functions become positive feedback functions, moving the ecosystem as a whole towards a 'switch' (Petraitis and Latham 1999; Briske et al. 2006) to another state.  There is evidence to suggest that forest recovery towards pre-fire conditions is unlikely under present-day conditions in some areas and there is a threat of ecosystem switches occurring in forest ecosystems via loss of abiotic and biotic processes.  For example, a history of low-severity surface fire exclusion during the past century may shift ponderosa pine (Pinus ponderosa) forest of the southwest towards a non-forested grass or shrub community (Savage and Mast 2005); and increased winter precipitation and resulting floods since fire have reduced conifer density and establishment of Quebec’s boreal floodplain (Bouchon and Arsenault 2004).  2  Although fire regimes have been found to shift over periods of centuries or millennia (Hallett et al. 2003), current rapid change in climate has been accepted largely as an effect of anthropogenic activities (IPCC 2007; Spittlehouse 2008) and evidence suggests that increasing temperatures will continue to accelerate wildfire frequency (Gillett et al. 2004).  A general trend of increasing area burned by wildfires in Canada has already been noted (Van Wagner 1987; Stocks et al. 2003; Gillett et al. 2004).  Although more difficult to quantify, fire severity is also expected to increase with climate change in North America (IPCC 2007).  This will be caused at least partially by changes in species composition (Collins et al. 2007).  Fire frequency is not expected to be homogeneous across the regions of circumboreal forest (Flannigan et al. 1998) and recently it has been found that insect outbreaks can have a greater impact than fire in some ecoregions (Parish et al. 1999; Burton et al. 2005; Kurz et al. 2008).  It is now of concern, regardless of fire, that increasing global temperatures alone may conceivably extirpate species with pre-adapted alleles from cooler climates (Aitken et al. 2008) and attention is being paid to mechanisms underlying species persistence, including dispersal (Clark et al. 2003; Higgins et al. 2003) and genetic diversity (Petit et al. 2003, 2004; Hamrick 2004).   Between 1912 and 1994, 8,470,340 ha of British Columbia’s forest were harvested; and between 1913 and 1993, 1,744,789 ha of these harvested lands underwent prescribed burn (Voller and Harrison 1998).  Prescribed burning (known as slashburning) was widely used after clearcutting as a means of fire hazard reduction, cost reduction of site preparation for planting (Vyse and Muraro 1973) and control of vegetation assumed to compete with conifer crop species (McMinn 1982).  The resulting second growth coniferous forests are still in early to mid-seral development stages today (B.C. Ministry of Forests, Mines and Lands 2010).  As a result, there is a dearth of experiments in regenerating forests that are old enough to provide a meaningful window of assessment for forest recovery and research into the effects of slashburning on forest ecosystem structure and function in British Columbia has been sparse and short-term. 3  There are many factors that determine the effect of fire on ecosystems, including, but not limited to: pre-existing conditions (previous fire regime, tree mortality due to insects or disease, forest management, and species present); climate and weather (leading up to and following fire); properties of the fire (intensity, severity, extent, patchiness, source of ignition); the landscape (factors that affect fire behaviour such as slope, aspect, and topographical discontinuities that may also afford opportunities for refugia); and physiological attributes of the organisms present that have adapted to survive fires such as thick bark, resinous cones in the canopy, or seeds with hard coats (Bradstock and Kenny 2003). There are also many ways to measure fire effects on forest ecosystem stability including: plant biomass (Mack et al. 2008; Shenoy et al. 2011), plant community diversity and composition (Romme 1982; Johnstone and Chapin III 2006; Hamilton and Haeussler 2008), soil nutrients (Kranabetter and Macadam 2007), and tree productivity measures such as site index (Nigh 1996).  Because many of these factors and effects are directly or indirectly related to one another, merely the exploratory process of constructing simplified theoretical frameworks (e.g., plant functional types) may reveal relevant indicator properties of stability.  Examples of such indicators might include individual species, functional groups of species, or physiological mechanisms.  These indicators of ecosystem response may be useful for introducing approaches to forest management that may be more readily and sustainably applied in an environment of directional climatic change and increased disturbance (Dale et al. 2001; Haughian et al. 2012) that is occurring in the forests of central British Columbia today. Approaches for measuring the state of managed forests in central British Columbia Reducing forest ecosystem complexity using plant functional types (Chapter 2) One approach to simplifying this complex problem is to reduce the large number of individual species that are found in these forests to relatively few groups determined by a combination of the many ways in which individuals sustain life, i.e. their basic functions.  Physical traits differentiate the ability of plants to access resources for survival and growth such as solar radiation, soil nutrients and 4  water, as well as functional differences related to length of season for photosynthetic activity and nutrient requirement for high leaf turnover (Chapin III et al. 1996b).  Seed traits address the abilities of plants to colonize from outside the site, or establish via long-term viability of seeds banked in the soil (Stark et al. 2006) or protected in cones in the canopy.  Plants also capitalize on chance by producing more seeds than can be consumed by birds and small mammals, increasing the odds for successful colonization after disturbance.  Plant root traits are also important because they relate to the potential for vegetative regrowth, adult survival of fire, and the ability to develop mutualistic relationships with fungi that aid in absorption of minerals.  Environmental traits can identify plants that may have advantage over others by being able to re-establish on nutrient-rich, nutrient-poor or undeveloped soils, or survive adverse winters and direct sun in the absence of overstory vegetation.  Chapter 2 of this dissertation offers an example of plant functional type (PFT) construction using the traits mentioned above while Chapters 3, 4 and 5 provide examples of how these PFTs can be useful in the investigation of ecosystem stability. Measuring stability with vascular plant species and plant functional types (Chapter 3) There have been several approaches to describe stability: (i) the system state represented by either a large ball (high resistance) or a small ball (low resistance) situated within the hollow of a 3-dimensional “stability landscape” of varying degrees of resilience (Larsen 1995); (ii) a combination of resilience (width of the hollow) and stability (height of the hollow) (Gunderson 2000; Hamilton and Haeussler 2008); or (iii) a combination of the probability of change versus system recovery, response, and resilience (Chapin III et al. 2002).  My approach to the idea of stability includes three distinct and interrelated components (Figure 1.1): 1. ecological context that delimits the domain of validity of stability statements.  This includes the variable of interest (vascular plants), the level of description (plant species and PFTs), the reference state (as described in the Study Area section of this chapter), the disturbance 5  under consideration (slashburning after clearcutting), and the spatial and temporal scales (the study includes site, ecosystem and study area spatial scales; and the temporal scale refers to the year in relation to slashburning up to 20 years post-burn) (Grimm and Wissel 1997); 2. stability agents consisting of matter (biogeochemical components of ecosystems) and information (the genetic composition of ecosystems) (Larsen 1995); these stability agents integrate combinations of functions; and 3. a collection of concepts about an ecosystem's response to disturbance and their interpretations bound by the ecological context and stability agents.  Such concepts include ecosystem persistence, resistance and resilience (Grimm and Wissel 1997). The ontology I present in this study provides for direct measures of stability agents (matter measured by abundance, and information measured by species richness) that govern by key ecosystem functions, and facilitates the application of these data to test hypotheses of stability.  This study was interested in vascular pant species at individual species or plant functional type (PFT) levels of description; the reference state refers to the ecosystem as noted; and the disturbance includes clearcutting and slashburning.  The spatial scale includes either site, ecosystem or study area levels and the temporal scale refers to the year in relation to slashburning.  Hence, inferences about ecosystem stability and function were able to be statistically validated and directly related to stability agents and concepts that were rooted to a specific ecological context. Functional diversity of plant communities is important to our understanding of ecosystem response because it can reduce the response of a large number of individual species to relatively few key ecosystem processes and it helps us predict the rate and degree of recovery following disturbance (MacGillivray et al. 1995).  While ecosystem models have sought to examine and predict the effect of harvest and/or fire on long-term tree biomass production (Kimmins et al. 2010) and stand development 6  (Coates 2002; Coates et al. 2003), none have examined the long-term resilience of vascular plant communities to slashburning after clearcutting.   In Chapter 3 of this dissertation, I identified pre-burn or post-burn indicators (species recorded exclusively pre-burn or post-burn) and PFT representatives (the three most dominant species, measured by greatest total percent cover for each PFT).  For each PFT, I measured abundance and richness before and 17-21 years after slashburn in ESSF, ICH and SBS forest of central British Columbia. Measuring resilience of vascular plants along ecological gradients (Chapter 4) Prescribed burning after clearcut harvesting may not always fall within the range of natural variability in the forest ecosystems of central British Columbia (Kopra 2003).  Where it does not, the ability of the system to recover to its pre-disturbance state may be compromised.  Following severe fire that removes a large portion of the forest floor, for example, ecosystem services such as forest productivity or plant diversity may be reduced (Perry et al. 2011).  The effects of slashburning on plant communities are not always obvious or linear, and responses can vary by species along multiple gradients of resource availability and disturbance severity (Kranabetter et al. 2006).  In addition, relative differences between species responses can change over time.  However, mechanisms of succession described by classical ecological theory (Clements 1916; Gleason 1926) can assist with prediction and lend to groupings based on species traits as mechanisms for coping with fire (Rowe 1983). The response of forest systems altered by logging and burning has been investigated in many ways.  Ecological researchers have gained knowledge in the areas of successional patterns (Halpern 1989; Franklin et al. 2002; Johnstone and Chapin 2006; Stark et al. 2006) and evolutionary strategies of plants under stress (Grime 1977; Newland and Deluca 2000; Schimel and Bennett 2004).  Attention to ecosystem recovery following fire has also focused on soil properties and processes such as soil nitrogen cycling (Kranabetter et al. 2007), soil carbon (Kranabetter and Macadam 2007) and the diversity of soil fungi (Durall et al. 2006; Treseder et al. 2004).  Theories of ecological stability (Holling 1973), complexity 7  (Puettmann et al. 2008), and self-organization in ecosystems (Perry 1995; Levin 2005) underscore the multitude of factors, effects, and scales forming our understanding of ecological recovery. One approach to studying these complex systems in a relatively simple way is to observe the relationship between plant community response and ecosystem or disturbance properties along gradients.  This provides a method of accounting for and conceptualizing nonlinear variation in factors driving and servicing ecosystem processes.  Explanatory gradients can include gradients of resources such as nutrients, water, and light (Chen et al. 2004; Kranabetter et al. 2010); environmental conditions such as temperature and pH; or they can involve factors such as topography, slope or latitude that indirectly determine resource availability or environmental conditions (Allen 1991; Franklin 1995). Ecosystem stability can also be linked to disturbance gradients; for example, vegetation succession has been shown to depend on fire regimes (Hamilton and Haeussler 2008; Johnstone and Chapin 2006) and soil properties (Tilman 1987; Fenton 2006).  To facilitate comprehensive study of variation in ecosystem stability with disturbance, PFTs have been examined along different types of gradients including soil resources (Kleyer 1999), fire disturbance (Kessell, 1979, Boer and Smith 2003) and climate (Diaz 1998).  Chapter 4 of this dissertation uses the PFTs developed in Chapter 2 to measure resilience along gradients of fire severity and soil nutrient availability. Measuring resilience with structural equation modeling (Chapter 5) Until recently, analysis of natural systems has generally employed methods such as bivariate or multivariate regression techniques (an overview is provided by Tabachnick and Fidell 2007), or methods that consider the relationship between one group of dependent variables and one group of independent variables, such as canonical correspondence analysis (ter Braak 1986).  Structural equation modeling (SEM) provides an alternative technique for carrying out ecological inquiries (Iriondo et al. 2003; Pugesesk et al. 2003; Grace 2006) provoking a non-traditional approach to multivariate methods 8  (Guarino 2004; Alavifar et al. 2012).  In simplest terms, SEM is the combination of exploratory factor analysis (EFA) and multiple regression analysis (Tabachnick and Fidell 2007). While the domain of social science has been using SEM as a standard for decades, only recently has there been an upwelling of practical applications of SEM in ecology.  It has become a useful tool in plant ecology, including investigations of plant functional strategies during old-field succession in southern France (Vile et al. 2006), species richness and soil properties in Pinus ponderosa forests of southwestern United States (Laughlin et al. 2007), and effects of oak on species richness through the contribution of litter in Pinus palustris woodlands of southeastern United States (Hiers et al. 2007).  SEM has been used to consider the application of prescribed fire in California with studies on postfire plant diversity in shrublands (Grace and Keeley 2006) and factors controlling invasion of alien annuals after wildfire in Pinus ponderosa forest (Keeley and McGinnis 2007).  Recently, SEM has been used to determine the relative importance of factors driving carbon storage in boreal forest of northern Sweden (Jonsson and Wardle 2010).  Chapter 5 of this dissertation postulates a SEM scenario that uses the PFTs developed in Chapter 2 to determine the relative importance of ecological factors on resilience of ESSF, ICH and SBS forest of central British Columbia. Thesis objectives The goal of my research was to investigate the stability of Sub-Boreal Spruce (SBS), Engelmann Spruce-Subalpine Fir (ESSF) and Interior Cedar-Hemlock (ICH) forest ecosystems of central B. C. (Figure 1.1).  My objectives were to (i) delineate PFTs for these forest ecosystems of central British Columbia (Chapter 2), (ii) determine which indicator species and PFTs succeeded or failed in response to slashburning after clearcutting by using measurements of abundance and richness (Chapter 3), (iii) explore the relationship between resilience and gradients of fire severity and soil resource availability (Chapter 4), and (iv) develop a structural equation model to assess relative importance of climate, fire severity and soil nutrients to the resilience of forest ecosystems in central B. C. (Chapter 5). 9  Study area The study area included 16 long-term research installations (Table 1.1) located in central British Columbia (Figure 1.2), and maintained by the British Columbia Ministry of Forests and Range.  These 16 study sites represent three biogeoclimatic zones:  The ICH zone occupies low to mid elevations in the mountainous terrain of interior British Columbia generally characterized by cool, wet winters and warm relatively dry summers.  Temperatures in the ICH are milder than the SBS and ESSF, and it is more humid and has longer fire return intervals than the SBS.  The ICH sites in this study are relatively wet and cold northern sites, with one described as the Nass Moist Cold (ICHmc1) variant (Banner et al. 1993) and the other two as transitional to the ESSF Cariboo Wet Cool (ESSFwk1) variant or the SBS Very Wet Cool (SBSvk) subzone (DeLong 2003).  The SBS zone occupies plateau landscapes of central British Columbia with a wider range of temperature, lower precipitation and shorter fire return intervals than the ICH and ESSF.  Its continental climate has severe, snowy winters and relatively warm, moist, short summers.  The SBS sites in this study are well distributed across this zone, with subzones and variants described as Babine Moist Cold (SBSmc2) and Mossvale Moist Cool (SBSmk1) (DeLong et al. 1993), as well as Willow Wet Cool (SBSwk1) and Very Wet Cool (SBSvk) (DeLong 2003).  The ESSF is a subalpine forest zone that lies above the other two zones and has a colder climate; it has a cool short growing season, longer winters with heavy snowfall and long fire-free intervals.  The ESSF study sites are described as the Moist Cold (ESSFmc) subzone (Banner et al. 1993) and Northern Monashee Wet Cold (ESSFwc2) variant (Lloyd et al. 1990). Of the 16 study sites, five are located on the west side of the Interior Plateau and in the Skeena Mountains of west-central British Columbia near the towns of Smithers, Hazelton and Burns Lake (Herron (HR), Kinskuch (KI), Walcott (WA), Helene (HE) and McKendrick (MK)).  The soils are developed 10  on glacial till and have sandy loam to loamy texture and moderate coarse fragment content (20-40%).  Soil orders include Luvisols, Brunisols, and Podzols (Soil Classification Working Group 1998).  Prior to harvest, the four SBS and ESSF sites supported mature or old-growth stands of lodgepole pine (Pinus contorta var. latifolia), interior spruce (Picea glauca x engelmanii), and subalpine fir (Abies lasiocarpa) with forest floors ranging from 3 to 22 cm thick.  The ICH site (KI) also had western hemlock (Tsuga heterophylla) and western redcedar (Thuja plicata) with forest floors averaging 5 cm thick.  Further details on these five west central British Columbia study sites and sampling methods are provided in Kranabetter and Macadam (2007). The other eleven sites are located on the east side of the Interior Plateau and in the Omineca and Columbia Mountains of east-central British Columbia near Prince George, McBride and Blue River (Brinks Mill (BM), Chuchinka Creek (CH), Francis Lake (FL), Genevieve Lake (GL), Goat River (GR), Haggen Creek (HC), Indianpoint Creek (IP), Mackenzie (MA), Otter Creek (OC), Walker Creek (WC) and West Twin (WT)).  Parent material at the SBS sites (Brinks Mill (BM), Francis Lake (FL), Genevieve Lake (GL) and Mackenzie (MA)) includes morainal till with glaciofluvial, fluvial, and lacustrine sediments along rivers.  Soil orders include Luvisols, Brunisols, Podzols, Gleysols, as well as Organics in depressions (Canadian System of Soil Classification 1998).  Forest floors were typically 5-10 cm thick prior to burn.  Soils in the ESSF zone (Otter Creek (OC) and West Twin (WT)) were Podzols on morainal till, with forest floor 4-7 cm thick prior to burn.  Soils at the ICH  sites were Podzols, with clay loam to silty clay loam texture and forest floor 4-6 cm thick (Goat River) or silty clay loam texture with forest floor up to 10 cm thick (Walker Creek).  Prior to harvest, these 11 sites supported mature or old-growth stands of white spruce (Picea glauca), Engelmann spruce (Picea engelmanii) and their hybrids; lodgepole pine; subalpine fir; and some western hemlock at Goat River.  Understory vegetation ranged from ericaceous shrubs including black huckleberry (Vaccinium membranaceum) and false azalea (Menziesia ferruginea) and feathermosses (Pleurozium schreberi, Hylocomium splendens) on well-drained sites to devil’s club 11  (Oplopanax horridus), oak fern and lady fern (Gymnocarpium dryopteris, Athyrium filix-femina), leafy mosses (Plagiomnium drummondii), and liverworts (Marchantia polymorpha) on moister and more nutrient-rich sites.  Further details on eight of the eleven east-central British Columbia study sites (Chuchinka, Goat River, Herron, Mackenzie, McKendrick, Otter Creek, Walker Creek, and West Twin) and sampling methods are provided in Hamilton and Peterson (2003, 2006) and Hamilton (2006a, 2006b, 2006c, 2007); and ten-year vegetation response is reported by Hamilton and Haeussler (2008). All 16 study sites were clearcut and slashburned in the 1980’s (Table 1.1).  A year after slashburning, the sites were planted (except Otter Creek which was two years after slashburning); interior spruce was planted at Chuchinka Creek, Goat River, Haggen Creek, Indianpoint Creek, Mackenzie, Walker Creek and West Twin; lodgepole pine was planted at Helene, Herron, Kinskuch, McKendrick and Walcott; both interior spruce and lodgepole pine were planted at Brinks Mill, Francis Lake and Genevieve Lake; and interior spruce, lodgepole pine and subalpine fir were plated at Otter Creek. Fire effects and vegetation monitoring between 1985 and 2000 was conducted by Evelyn Hamilton at study sites located on the east side of the Interior Plateau and in the Omineca and Columbia Mountains of east-central British Columbia near Prince George, McBride and Blue River (Brinks Mill, Chuchinka Creek, Francis Lake, Genevieve Lake, Goat River, Haggen Creek, Indianpoint Creek, Mackenzie, Otter Creek, Walker Creek and West Twin).  Brad Hawkes from the Canadian Forest Service assisted with fire effects data at Walker Creek, and Michael Feller collected the fire effects data at Otter Creek.  At study sites located on the west side of the Interior Plateau and in the Skeena Mountains of west-central British Columbia near Smithers, Hazelton and Burns Lake (Helene, Herron, Kinskuch, McKendrick and Walcott), fire effects and soils monitoring was conducted by Rick Trowbridge and Anne Macadam, and 20-year soils monitoring was carried out by Marty Kranabetter and Marcel Lavigne.  Vegetation monitoring at these sites was carried out by Allen Banner and Karen McKeown; and Karen McKeown, Ben Heemskerk and Will MacKenzie assisted with access to datasets. 12  Table 1.1. Description of the 16 study sites located in the temperate and montane coniferous forests of central interior British Columbia.     Sitea BEC Site Seriesb Latitude LongitudeMAT(°C)cYear LoggedYear BurnedDOBd (cm)LFHe (cm) n Years sampledgBM SBSvk/01 53°26'15"N 121°33'39"W 910 3.6 704 1985 1986 1.6 5.9 Sx, Pl 3 25 1,2,3,5,10,21CH SBSwk/06-07 54°31'13"N 122°31'09"W 820 2.8 851 1986 1987 1.3 8.9 Sx 29 25 -1,1,2,4,5,10FL SBSmk/01 53°44'56"N 122°20'52"W 850 3.4 686 1985 1986 4.1 13.0 Sx, Pl 6 25 1,2,3,5,10,21GL SBSmk/01-07 53°15'35"N 122°22'29"W 920 3.4 651 1985 1986 1.8 9.0 Sx, Pl 4 25 1,2,3,5,10,21GR ICH(ESSF)wk1/01 53°21'46"N 120°41'13"W 1040 3.1 1057 1987 1988 1.0 6.2 Sx 6 25 -1,1,2,3,5,10,20HC SBSvk/01 53°30'27"N 121°30'27"W 1020 3.2 798 1983 1986 2.8 8.3 Sx 8 25 1,2,3,5HL SBSmc2/01 54°16'40"N 125°03'45"W 1050 2.7 608 1981 1982 2.4 8.7 Pl 15 25 1,2,3,4,5,7,10,18,21HR ESSFmc/01 54°20'25"N 125°10'00"W 1335 1.6 644 1982 1983 1.3 9.8 Pl 10 25 -1,1,2,3,4,8,17IC SBSvk/02 53°26'37"N 121°33'03"W 950 3.6 724 1985 1986 0.7 6.5 Sx 3 25 1,2,3,5,10KI ICHmc1/01 55°33'22"N 128°59'40"W 270 5.3 1031 1981 1982 2.9 5.3 Pl 30 25 1,2,3,4,5,7,10MA SBSwk/08 55°05'42"N 122°59'07"W 770 2.7 755 1987 1988 3.1 12.7 Sx 6 25 -1,1,2,3,5,10,20MK ESSFmc/01 54°51'50"N 126°43'45"W 1075 1.9 653 1984 1985 2.3 7.9 Pl 15 25 -1,1,2,3,4,5,7OC ESSFwc2/01-07 51°45'33"N 119°12'46"W 1560 1.3 1391 1987 1989 0.9 5.1 Sx, Pl, Bl 34 9 -1,1,2,3,5,11,18WA SBSmc2/01 54°31'20"N 126°55'00"W 830 3.0 588 1981 1982 3.7 10.4 Pl 15 25 1,2,3,4,5,7,10,16,21WC ICH(SBS)vk/01 53°54'30"N 120°35'45"W 1050 3.0 1064 1985 1986 1.9 11.5 Sx 120 0.79 -1,1,2,3,5,10,21,22WT ESSFwk1/01 53°24'59"N 120°34'18"W 1400 2.3 955 1987 1989 3.5 6.0 Sx 30 25 -1,1,2,3,5,11g Measurements collected after clearcutting but before prescribed burn is indicated by -1; and 1-22 indicates growing seasons after prescribed burn.f Tree species planted: Sx = interior spruce, Bl = subalpine fir, Pl = lodgepole pine.d DOB, mean depth (cm) of burn of forest floor (LFH) layers.Plot Size (m2)Note: Site locations are shown in Figure 2.Elevation (m)MAP(mm)cc Average mean annual temperature (°C) and mean annual preciptitation (mm) for 1980 - 2002 (ClimateBC, Wang et al. 2006, 2012).e LFH, depth (cm) of forest floor layers including woody debris <1 cm in diameter. Species Plantedfa Site abbreviations: BM = Brinks Mill, CH = Chuchinka Creek, FL = Francis Lake, GL = Genevieve Lake, GR = Goat River, HC = Haggen Creek, HL=Helene,            HR = Herron, IC = Indianpoint Creek, KI = Kinskuch, MA = Mackenzie, MK = McKendrick, OC = Otter Creek, WA = Walcott, WC = Walker Creek, WT = West Twin.b Biogeoclimatic ecosystem classification site series (Meidinger et al. 1988; Lloyd et al. 1990; Banner et al. 1993; DeLong et al. 1993; DeLong 2003).13     Figure 1.1. Stability agents, matter and information (Larsen 1995), can be described in terms of stability concepts (such as resistance and resilience) that are associated with a specific ecological context (Grimm and Wissel 1997). This study was interested in vascular pant species at individual species or plant functional type (PFT) levels of description providing measures of species richness (i.e. information) and abundance (i.e. matter). The reference state refers to the ecosystem as noted; and the single disturbance examined was clearcutting with slashburning. The study includes site, ecosystem and study area spatial scales; and the temporal scale refers to the year post-burn.   Grimm and Wissel 1997 STABILITY CONCEPTS  What stability properties are addressed?  1. Resistance 2. Resilience Larsen 1995 STABILITY AGENTS  Matter Biogeochemical cycle  Information  Genetic Diversity ECOLOGICAL CONTEXT  To which ecological situation does the stability statement refer?  1. Variable of interest 2. Level of description 3. Reference state 4. Disturbance 5. Spatial scale 6. Temporal scale Grimm and Wissel 1997 14     Figure 1.2. Map of study locations in British Columbia; site descriptions are in Table 1.1.   15  Chapter 2. Complex dataset reduction: 183 vascular plants characterized by 9 functional types Plant functional traits describing species vital attributes that reflect adaptations specific to fire have been outlined by Cattelino et al. (1979), Noble and Slatyer (1980) and Bradstock and Kenny (2003).  These attributes can be related to (i) persistence strategies that include both propagule production and vegetative reproduction, including colonization from outside the site, long-term viability of seedbank stored in the soil, seeds protected in cones in the canopy, vegetative regrowth, and/or adult survival; (ii) establishment characteristics such as the resource tolerance, intolerance, or requirements of an established community; or (iii) life history stages, including propagule longevity or germination, maturation, loss, and local extinction.  Rowe (1983) described life history classes that indicate adaptation to fire intensity, with classes defined by species favoured by (i) high intensity fire (endurers, evaders, and invaders), (ii) low intensity fire (resisters), and (iii) fire exclusion (avoiders). PFTs have been used successfully as indicators of disturbance and ecosystem processes in a range of ecosystem types including serpentine grassland (Hooper and Vitousek 1998), Brazilian woodland (Batalha and Martins 2004), and temperate deciduous forests (Aubin et al. 2007).  Chapin III et al. (1996a) identified 10 emergent groups (generally representing life form) involving 37 vascular and non-vascular plants in tundra ecosystems on the south slope of the Brooks Range in Alaska.  According to the paleorecord, some of these PFTs responded predictably to past climatic changes, whereas field experiments also show that these PFTs respond predictably to changes in soil resources, but less predictably to temperature.  Aubin et al. (2007) identified 13 emergent groups (generally representing a combination of life form and mode of dispersal) that included 214 species in sugar maple (Acer saccharum) dominated stands in Quebec.  They found that nine of the 13 PFTs varied with the type of human disturbance (single tree selection in a forested matrix, or originating from pasture abandonment or undergoing maple syrup production in an agricultural matrix).  Over time there was an increase in wind-dispersed exotics and an increase in seedling density, but a decrease in spring geophytes and 16  certain shade tolerant herbs.  I am aware of no such studies for forest ecosystems of central British Columbia. For the purpose of this study, plant functional types (PFTs) are defined as “groups of species that have similar effects on ecosystem processes” (Chapin III et al. 1996b) and are the result of organizing individual species into groups that share the same or overlapping traits expressing establishment, succession or change within the ecosystem.  In creating the PFTs, we can overcome the difficulty presented to us by complexity in nature by, for example, studying two (and more) plants that have different evolutionary paths or physical characteristics, but that can engineer in the same way or evoke like outcomes in an ecological context. PFTs may be useful in (i) measuring resistance and resilience of forest ecosystems after clearcutting and slashburning; and (ii) measuring vegetation response to physical environmental gradients subject to sudden disturbances and/or chronic stress such as climate change (Chapin III et al. 1996a).  I predicted that the classification of PFTs may serve as a useful primary predictor of ecosystem recovery from disturbance because PFTs represent ecosystem processes (Hooper et al. 2002) in contrast to taxonomic or phylogenetic classifications of plants.   Methods Field procedures The sites were clearcut between 1981 and 1987; harvesting was done in the winter to help protect the forest floor and understory vegetation during the logging operation.  Between one and three years after logging, the sites were slashburned in the fall (except Otter Creek, which had blocks burned in spring and fall).  To measure fire severity, depth-of-burn (DOB) pins were inserted into the soil, and forest floor depths were measured prior to and immediately after the burn (McCrae et al. 1979; Trowbridge et al. 1987).  The mean depth of burn ranged from 0.7 to 4.1 cm, indicating that the prescribed burns ranged from low to moderate severity (Feller 1996) across the 16 study sites (Table 17  1.1).  After slashburning, interior spruce (Picea glauca x engelmanii) and/or lodgepole pine (Pinus contorta var. latifolia) (and subalpine fir (Abies lasiocarpa) at Otter Creek) were planted.  Plot size and number of plots varied by site, and vegetation was re-measured (percent cover and height of each vascular plant species) in permanent plots 1, 2, 3, 5, 10, and 20 years after slashburning for most sites.  Vegetation data collected after harvest but before burning exists for eight of the 16 study sites.  See Table 1.1 for details regarding clearcut and burn years, fire severity, species planted and the years of vegetation measurements for each site. Data A species-trait dataset was created that included all 183 vascular plants measured on the 16 study sites (Appendix A provides a species list).  This dataset was used to establish the traits of individual species for the cluster analysis input in order to delineate PFT membership.  Fifteen traits were selected to differentiate strategies, requirements, and responses of individual species following logging and slashburning, thereby representing recovery potential of ecosystem function and processes.  These measurable attributes included plant functional traits related to physical properties (plant height, plant duration, leaf duration, N-fixation); seed properties (seed quantity, seed dispersal, seed longevity, seed size); and root properties (depth of rooting, vigour of sprouting, rate of lateral spread, and dominant mycorrhizal guild).  I also included traits associated with environment: ground surface material, climate, and light index.  See Appendix B for details on trait categories and descriptions, traits of each individual species in the dataset, and data sources. Statistical analysis Cluster analysis was performed with the vegan package (Oksanen et al. 2013) using the average linkage method based on Euclidean distance.  This method avoids distortion caused by outliers by finding spherical groups where the distance between groups is the average of all distances for all pairs of individuals (McCune and Grace 2002). Processing was performed in R (The R Foundation for Statistical 18  Computing 2013).  All  183 vascular plant species and all traits in the species-traits dataset were used for cluster analysis input and  each species was allowed to occur only once in a single cluster.  Initially, the dendrogram was pruned (McCune and Grace 2002) at a distance of 2.0, producing 10 unique clusters, including one with only a single species, northern green rein orchid (Platanthera aquilonis).  To avoid single-species groups, the number of clusters was reduced to nine and northern green rein orchid was reassigned to another group. Results 9 plant functional types The cluster analysis grouped all species into nine PFTs, summarized in Figure 2.1.  See Appendix C for the dendrogram.  The first division separated the conifers (PFT 1; 9 members) from all other plants in the dataset.  The response of the conifers is particularly important because conifers were the keystone species in these forests; they are tall and influence their surroundings chemically and physically.  The flowering plants were then divided by the type of humus form they are associated with (primarily mor or primarily moder/mull humus form).  Those not associated with mor were generally deciduous, including deciduous trees and actinorhizal shrubs (PFT 4; 8 members), which may also have a relatively large effect on early ecosystem recovery (Shenoy et al. 2011) and may serve well as response indicators.  Compared with conifers, deciduous tree species can sprout from disturbed root stocks, have greater seed quantity and recycle nutrients more rapidly via litterfall (LePage et al. 2000; Simard and Vyse 2006).  Deciduous tree leaf litter decomposition rate is initially greater than that of needle litter (Prescott et al. 2000), and they have higher net carbon uptake in the growing season (Bradley and Fyles 1996).  The understory flowering plants not associated with mor humus form were generally deciduous and represented a gradient of shade tolerance that included shade-tolerant surface water specialists that are found in the alpine tundra and adapted to late snow-melt sites (PFT 5; 7 members), semi-shade 19  tolerant gap specialists (PFT 2; 79 members) associated with mull/moder humus forms, and shade-intolerant gap specialists (PFT 3; 42 members) associated with exposed mineral soil. The remaining four PFTs consisted of 38 species commonly associated with mor humus forms that have adaptations enabling them to persist with limited solar radiation and soil nutrients (PFTs 6, 7, 8 and 9).  Members of PFT 6 are relatively transient and do not tolerate well the earliest stage of succession, particularly exposure to full sunlight.  This includes all members of Lycopodiaceae, and nearly all dwarf woody plants in the dataset.  Most of these deep shade specialists are of the Orchidaceae and Ericaceae families and are able to acquire a portion of their nutrient demands through mycorrhizal fungi (Gebauer and Meyer, 2003; Tendersoo et al. 2007; Massicotte et al. 2008), thereby supplementing the sun’s energy deficient in the closed canopy environment.  Unsuited to canopy gaps, the ability to disperse between patches of closed forest and back into the disturbed area once the canopy closes is an essential mechanism facilitated by wide dispersal of minute seeds.  PFT 6 therefore provides a good indicator of stability because these species are best adapted to the conifer forest environment of deep shade and mor humus form, and reflect return to the original state over time. Discussion Plant functional types and their ecology The unique combination of attributes of individual species presents a difficulty in generalizing an abundant number of species into relatively few PFTs, and not all traits were discretely classified into distinct groups; the range of overlap of traits among PFTs can appear when the dendrogram (Appendix C) is perceived as a 3-dimensional object, similar to how one would envision a mobile with all nodes able to rotate in space.  These traits guide interactions among the plants, soil chemicals, and soil organisms (fungi and bacteria), thereby assisting in plant resource acquisition and playing a role in community structure and organization.  Because PFT classification was based on functional traits important to disturbance response, and produced groupings of species that share a common set of traits distinct from 20  those of other PFTs, we can gain insight to ecosystem response to slashburning after clearcutting in these observed forests. The PFT classification divided several genera among different PFT indicating a range of functionality within genera (Appendix B).  For example, northern green rein orchid (PFT 5) was placed in a different PFT than one-leaved rein orchid (Platanthera obtusata) and slender rein orchid (Platanthera stricta) (PFT 6), even though it also shares the same potential for mixotrophy and wide seed dispersal typical of the species in PFT 6.  The aerenchymatous roots of Platanthera spp. (Currah and Zelmer 1992; Taylor et al. 2002) are an adaptation to water-logged soils: one-leaved rein orchid is associated with mor humus forms; slender rein orchid is associated with mor humus forms but can also be found in soils with surface water; and northern green rein orchid is associated with soils with surface water; this adaptive radiation (Darwin 1859) within Platanthera serves to divide the genus niche space.  Thus, they are important indicators of the recovery rate of wet sites because they are able to colonize from outside the site and occur where other species cannot, particularly in early stages of succession when trees have not yet had the time to generate the biomass for substantial water uptake. In contrast to PFT 6, the eight species in PFT 7 are found growing under deciduous trees as well as conifers, and they tolerate the least shade of the mor associated plants.  PFT 8 requires slightly less light than PFT 7 and as a group has the greatest average seed size.  Both PFT 7 and 8 have moderate sprouting vigour; however PFT 7 has deeper rooting while PFT 8 has greater lateral root spread.  Members of PFT 7 and 8 are predominantly arbuscular mycorrhizal except for western mountain-ash (Sorbus scopulina) and Sitka mountain ash (Sorbus sitchensis) of PFT 7, which are ectomycorrhizal; and velvet-leaved blueberry (Vaccinium myrtilloides) of PFT 8, which is ericoid mycorrhizal. Members of PFT 9 are ericoid mycorrhizal and prefer the acidic and nutrient deficient conditions characterized by mor humus forms.  Ericoid mychorrhizal fungi are able to assist in the decomposition process and aid in mobilization of nutrients otherwise unavailable in organic form (Read and Perez-21  Moreno, 2003).  This explains the ability of species in PFT 9 to acquire the resources required for growing tall and sustaining high leaf turnover, despite shaded conditions and nutrient deficient soils, and this ability greatly reduces stress from competition.  An additional adaptation of species in PFT 9 that confers stability after disturbance is the high capacity for resprouting from rhizomes following fire.  Velvet-leaved blueberry was separated from the subalpine ericaceous shrubs (PFT 9) because it tends to occur at lower elevations, an example of adaptation to climatic conditions due to landscape position, rather than edaphic conditions dividing a genus into different PFT. Conclusion Cluster analysis proved to be a useful method for identifying PFT groups, thereby simplifying a dataset that included 183 species, representing 183 combinations of interspecific and species-environment interactions (i.e. there were no two plants with exactly the same traits), into only nine groups.  These nine PFTs distinguish species with seed cones (conifers); species generally associated with mor humus form (shade specialists, ericaceous shrubs, montane boreal and cool temperate species, and ground surface material generalists); and species that are generally not associated with mor humus form (deciduous trees and tall shrubs, surface water specialists and gap specialists that are either semi-shade tolerant or shade intolerant).  The results of this classification are expected to provide a simplified approach to measuring and possibly predicting ecological response to disturbance.  22     Figure 2.1. Divisions, characteristics and species of PFTs produced by cluster analysis; GSM = ground surface material.  GymnospermsAngiosperms and Pteridophyta1Abies amabilisAbies lasiocarpaPicea engelmannii x glaucaPicea glaucaPinus contortaPseudotsuga menziesii var. glaucaThuja plicataTsuga heterophyllaTsuga mertensianaModer/Mull Humus FormSemi-shade tolerant3Exposed Mineral SoilShade-intolerant24Deciduous Trees5Surface Water SpecialistsShade-tolerant79 Species representing 28 families; between 9 and 12 spp. of Asteraceae, Poaceae and Rosaceae familiesAlnus crispaAlnus tenuifoliaBetula papyriferaPopulus balsamiferaPopulus tremuloidesSalix barclayiSalix bebbianaSalix scoulerianaCaltha leptosepalaChrysosplenium tetrandrumEquisetum scirpoidesLeptarrhena pyrolifoliaPetasites frigidusPetasites sagittatusPlatanthera aquilonis42 Species representing 17 families; between 6 and 11 spp. of Asteraceae, Cyperaceae and Poaceae familiesGap SpecialistsVascular PlantsofCentral British ColumbiaSmall-staturedArbuscular MycorrhizalDeciduousShade-tolerant820 Species;Lycopodiaceae, Orchidaceae and dwarf woody members of Caprifoliaceae, Cornaceae and Ericaceae6EvergreenDeep Shade SpecialistsModer/Mull Humus Forms as Secondary GSM7Clintonia unifloraFestuca occidentalisMelampyrum lineareOryzopsis asperifoliaPaxistima myrsinitesSorbus scopulinaSorbus sitchensisViola orbiculataGeocaulon lividumMaianthemum canadenseVaccinium myrtilloidesEricaceous ShrubsShade-adaptedSubalpine9Menziesia ferrugineaRhododendron albiflorumVaccinium alaskaenseVaccinium caespitosumVaccinium membranaceumVaccinium ovalifoliumVaccinium parvifoliumNot Generally Associated with Moder/Mull Humus FormsNot Generally Associated with Mor Humus FormDeciduousMor Humus Form as  Primary GSMEvergreen and DeciduousMixotrophic orArbuscular Mycorrhizal23  Chapter 3. Measuring resistance and resilience in ESSF, ICH and SBS forests of central British Columbia In 1973, Holling defined resilience as 'a measure of the persistence of systems and of their ability to absorb change and disturbance and still maintain the same relationships between populations or state variables'.  This definition of resilience was followed in 1986 with further distinction of 'persistence' by asserting that an ecosystem’s capacity for stability is attributed to its ability to resist or recover from disturbance (Holling 1986).  Since then, the scope of literature in the domain of ecosystem resilience has flourished with more than 160 definitions available by 1997 (Grimm and Wissel 1997).  Because resilience thresholds are not always immediately apparent, care must be taken in measuring ecosystem sensitivity and/or defining stability.  This includes considering response measurements in the context of: (i) time scales relevant to the processes under study (Chapin III et al. 2002); (ii) sudden 'triggers' (Briske et al. 2006); and (iii) long term conditions such as directional change in the environment (Starfield and Chapin III 1996). For the purposes of the present study, the three stability concepts range from a narrow equilibrium perspective to a broad non-equilibrium perspective: (i) persistence - constancy through time of an ecological system, i.e. no change occurs despite chronic environmental change or stress; (ii) resistance - staying essentially unchanged despite the presence of disturbances, i.e. no change after sudden disturbance; and (iii) resilience - returning to a reference state (or dynamic) after a temporary disturbance, i.e. some change after disturbance but no shift to alternative structures and processes, or stability domains (Holling 1973; Johnstone et al. 2010).  Therefore, persistence is a comparatively simple measure of stability whereas resistance and resilience become increasingly complex and more difficult to understand and measure.  “Resistance” was measured at the species scale and adds the dimension of succession and/or factors related to the disturbance itself (i.e. climate; fire severity or frequency).  Unlike “resistance”, “resilience” does not require exactly the same species to return after disturbance, 24  and the ecosystem is deemed “resilient” provided that the functional integrity (and the associated structures) of the ecosystem is retained. Three biogeoclimatic zones of central interior British Columbia were represented by the study sites measured in this chapter:  the moist temperate Interior Cedar Hemlock (ICH) zone; the cool mountain Sub-boreal Spruce (SBS) zone; and the cold, snowy montane Engelmann Spruce-Subalpine fir (ESSF) zone (Meidinger and Pojar 1991).  Climate drives fire events (Hallett et al. 2003) and accordingly, each zone has a unique relationship with fire including the length of time between fires (DeLong 1998) and specific patterns of succession that determine forest community composition and structure (Parminter 1983).  Plants that have evolved in forest ecosystems with a prior history of wildfire (i.e. shorter fire return interval) are expected to have a greater capacity to recover from fire, producing ecosystems with less change in species composition after slashburning (Hamilton and Haeussler 2008). The objective of this chapter was to use PFTs to measure ecosystem recovery across a range of study sites previously logged and slashburned.  Because of the interrelationships among climate, fire, and species adaptations to fire, I hypothesized that resistance to slashburning (measured by change in species composition) would decline with declining expected frequency of fire using a combination of temperature (MAT: >2°C=warm; <2°C=cold) and moisture availability (MAP: >1000mm =wet; <1000mm=dry) such that resistance decreases in the order: SBS (Mackenzie = warm and dry) > ICH (Goat River/Walker Creek = warm and wet) > ESSF (Herron = cold and dry; Otter Creek = cold and wet).  I expected that the role of local climate in the response of PFTs to slashburning would produce unequal distribution of PFTs among the subzones, indicating which PFTs are better adapted to slashburning and why.  By testing the relationship between resistance and resilience of plant communities to an ecological stress across a climatic gradient, this study explores a potential approach for evaluating the responses and properties of forests at risk in the midst of a directionally changing global climate (Chapin et al. 2003). 25  Methods Measuring resistance Of the eight study sites with post-harvest/pre-burn vegetation data, vascular plant percent cover was re-measured at five sites beyond 10 years post-burn: Goat River (re-measured at 20 years, n = 6 plots; ICH), Herron (17 years, n = 10; ESSF), Mackenzie (20 years, n = 6; SBS), Otter Creek (20 years, n = 34; ESSF) and Walker Creek (21 years, n = 120; ICH).  MAT and MAP for these five sites (Figure 3.1) were derived using ClimateBC (Hamann and Wang 2005; Wang et al. 2006).  On these five sites, 168 species were measured and recorded.  These measurements were used to identify two types of indicator species: (i) bioindicator species (species recorded exclusively post-burn or pre-burn), which are species thought to be sensitive to, and therefore may serve as an early warning indicator of, environmental changes such as global warming or modified fire regimes; and (ii) pre- and post-burn representative species (species with the greatest total percent cover) which are dominant species that provide much of the biomass or number of individuals in an area, and in the case of our study were used to assess plant community resistance and identify shifts in species dominance within PFT.  Change in percent cover of each species is a good measure of resistance because it identifies those species that thrive, decline, appear or disappear with disturbance. Measuring resilience To determine plant community resilience, I tested for significant differences in vegetation abundance and species richness prior to and 17-21 years post-burn using repeated measures.  Paired-sample t-tests were carried out for each of three levels: for all sites combined, for each site individually, and for each biogeoclimatic zone.  For all sites, a site-level average (n = 5) was calculated for total vegetation abundance by PFT and for total (pooled across plots) species richness by PFT.  For individual sites (n-values as in paragraph above) and for each biogeoclimatic zone (ESSF (n = 44), ICH (n = 126), and SBS (n = 6)) a plot-level average was calculated for total vegetation abundance and total species richness 26  by PFT.  Vegetation abundance and species richness by PFT are considered good indicators of resilience because they indicate the ability of plant groups identified as having important ecosystem functions to recover. Results Resistance At the subset of five sites, 64 species were recorded pre-burn and 105 species were recorded in the years post-burn.  Of these, 60 species remained after burning, 45 new species arrived after burning, and four species did not return to any of the five sites after burning.  The four species to not return were recorded only once pre-burn with <1% cover and were grouped in PFT 2 (northern bedstraw (Galium boreale) and skunk currant (Ribes glandulosum)); and PFT 3 (purple-leaved willowherb (Epilobium ciliatum) and an unidentified buttercup (Ranunculus sp.)).  Species not recovering at a site nor gained at a site of the same ecosystem included three members of Ericaceae, two members of Rosaceae, and one member each of Pyrolaceae, Orchidaceae and Liliaceae.  These were, respectively: white-flowered rhododendron (Rhododendron albiflorum) and false azalea (Menziesia ferruginea) (both PFT 9) at Goat River; single delight (Moneses uniflora, PFT 6) at Herron; red raspberry (Rubus idaeus; PFT 3) and five-leaved bramble (Rubus pedatus; PFT 6) at Mackenzie; one-sided wintergreen (Orthilia secunda) and heart-leaved twayblade (Listera cordata) (both PFT 6) at Otter Creek; and Hooker's fairybells (Prosartes hookeri; PFT  2) at Walker Creek (Table 3.1). Species indicating potentially low resistance to slashburning declined in percent cover (reported as % reduction) relative to pre-burn/postharvest levels as follows: heart-leaved twayblade (PFT 6), -98% and no longer occurring at Otter Creek; enchanter's-nightshade (Circaea alpina; PFT 2), -78% and only occurring at Walker Creek; subalpine fir (Abies lasiocarpa) (PFT 1), -75%; devil’s club (Oplopanax horridus (PFT 2), -68%; white-flowered rhododendron (PFT 9), -43% and no longer occurring at Goat River; and spiny wood fern (Dryopteris expansa; PFT 2), -34% (Tables 3.1 and 3.2).  Subalpine fir, the 27  most abundant conifer at every site pre-burn, did not return as the first representative species at any site; devil’s club did not return as a representative species in the SBS and transitional ICH sites; and white-flowered rhododendron did not return as a representative species in the ESSF and transitional ICH sites (Table 3.2). At the ecosystem level, our results showed that, not including species lost in one ecosystem but gained in another or with <1% cover, only one species disappeared from each biogeoclimatic zone with slashburning: heart-leaved twayblade (ESSF), white-flowered rhododendron (ICH), and five-leaved bramble (SBS).  Considering PFT indicators of low ecosystem resistance, we found: (i) loss of subalpine fir as a representative species of PFT 1 in all zones; (ii) loss of devil’s club as a representative species of PFT 2 on SBS and ICH(SBS) sites; and (iii) loss of white-flowered rhododendron as a representative species of PFT 9 on ESSF and ICH(ESSF) sites. Species with the greatest increase in percent cover following treatment included grasses (Calamagrostis spp.; PFT 2), fireweed (Epilobium angustifolium; PFT 2) and willows (Salix spp.; PFT 4) (Tables 3.1 and 3.2).  Post-harvest and pre-burn, Calamagrostis spp. occurred only once at Otter Creek, Epilobium angustifolium occurred on eight plots at Walker Creek averaging 2.5% cover per plot, and Salix spp. occurred only once at Walker Creek.  However, 17-22 years following slashburning, Calamagrostis spp. increased at all sites except Mackenzie and occurred on 21 plots averaging 4% cover per plot; and percent cover at all sites increased for Epilobium angustifolium (162 plots averaging 18% cover per plot) and Salix spp. (17 plots averaging 18% cover per plot).  Of all plants in the dataset, Alnus spp. (all designated to PFT 4) had the greatest percent cover post-harvest at Herron, Mackenzie, and Walker Creek, both pre- and post-burn.  Post-harvest and pre-burn, Alnus spp. was found on 17 plots averaging 15% cover per plot; 17-22 years post-burn Alnus spp. increased in abundance, occupying 40 plots averaging 40% cover per plot. 28  Vegetation in all zones appeared moderately resistant to slashburning following clearcutting and generally did not follow the hypothesized climatic severity gradient (SBS > ICH > ESSF) at any level of observation.  Each ecosystem lost a single bioindicator, subalpine fir, as a representative species, and lost a total of either 4 or 5 species.  The number of species gained showed greater differences in resistance among ecosystems: SBS (4 species lost, 30 species gained; total change of 34 species) > ESSF (4 species lost, 34 species gained; total change of 38 species) > ICH (5 species lost, 46 species gained; total change of 51 species).  Results at the site level were counter to our hypothesis, where resistance decreased (i.e. total change of species in the plant community increased) as follows:  Otter Creek (ESSF - cold and wet; change of 19 compared to 30 initial species, or 63%) > Herron (ESSF - cold and dry; 25 compared to 15 initial species, or 167%) > Goat River (ICH(ESSF); 27 compared to 15 initial species, or 180%) > Walker Creek (ICH(SBS); 32 out of 40 initial species, or 80%) > Mackenzie (SBS; 34 out of 31 initial species, or 110%). Resistance measured by ecosystem type found differences in the total number of species change post-burn (SBS > ESSF > ICH), while at the site level the inverse of the hypothesis was possible (ESSF > ICH > SBS).  However, where there is greater functional redundancy, and therefore increased ability to recover, an increase in resilience would be expected.  At the ecosystem level, by counting the number of PFTs associated with conifer forest that had significant increases in abundance and/or richness, results suggest differences in resilience by ecosystem type (ICH > ESSF > SBS) would be opposite resistance. Resilience Of the eight PFTs with group members found at the five sites, percent cover significantly (p<0.05) increased on average over all sites 17-22 years after slashburning for conifers (PFT 1), gap specialists (PFT 2 and 3), and ground surface material generalists (PFT 7); and richness significantly increased for PFTs 1, 2 and 3 (p<0.05) (Table 3.3).  Of all the PFTs, understory deciduous shade-intolerant gap specialists associated with exposed mineral soil (PFT 3) had the greatest increase in percent cover (35x) 29  and richness (14x) while the semi-shade tolerant gap specialists (PFT 2) had the greatest mean percent cover and richness pre- and post-burn over all sites (Table 3.3). When sites were grouped by biogeoclimatic zone, significant changes (p<0.05) in mean percent cover  (Table 3.4a) and richness (Table 3.4b) pre- and post-burn of PFTs were as follows: PFT 1 and 2 increased in percent cover and richness in all zones; PFT 3 increased in percent cover in the ESSF and ICH zones, and increased in richness at all zones; PFT 4 increased in percent cover and richness in the ICH and SBS zones; PFT 6 increased in percent cover in the ESSF and ICH zones, and increased in richness in the ICH zone; and PFT 9 increased in percent cover and richness in the ICH zone.  None of the PFTs significantly decreased in cover or richness in any zone. At the individual site level, significant (p<0.05) differences in mean percent cover and richness pre- and post-burn were found for PFTs 1, 2, 3, 6 and 9 (Table 3.5).  PFT 1 increased in percent cover at all sites except Otter Creek, and increased in richness at all sites except Goat River; PFT 2 increased in percent cover and richness at all sites; PFT 3 increased in percent cover at all sites except Mackenzie, and increased in richness for all sites; PFT 6 increased in percent cover at Goat River, Herron, and Walker Creek, and increased in richness at Walker Creek; PFT 9 increased in percent cover and richness at Walker Creek, but decreased in percent cover at Goat River. At 20 years post-burn, vegetation recovery at Goat River (MAT = 3.1°C; MAP = 1057mm) cover of shade specialists was greater than that of conifers (Figure 3.2a); and after 17 years this trend was also seen at Herron (MAT = 1.6°C; MAP = 644mm) (Figure 3.3a).  After two decades post-burn at Walker Creek (MAT = 3.0°C; MAP = 1063mm), the semi-shade tolerant gap specialists clearly dominated at 122% cover (Figure 3.2b).  Otter Creek (the coldest, wettest site at the highest elevation; MAT = 1.3°C; MAP = 1390mm) appears to be an open meadow ecosystem 18 years post-burn rather than a conifer forest (Figure 3.3b).  This site remained dominated by understory semi-shade tolerant gap specialists (108% cover) with scattered subalpine ericaceous shrubs (59% cover) and sporadic (7% cover) conifers or the 30  remaining (shade-intolerant) gap specialists.  At 20 years post-burn, Mackenzie (MAT = 2.7°C; MAP = 755mm) showed regeneration of conifers (34% cover) along with deciduous trees and tall shrubs (43% cover), but semi-shade tolerant gap specialists dominated at 80% cover (Figure 3.4). Discussion Resistance Species that decreased in percent cover indicate low resistance to slashburning and are generally shade-tolerant or shade-requiring and may also have had difficulty thriving without a mature overstory for protection from solar radiation.  Enchanter's-nightshade and devil’s club are the two most shade tolerant species of PFT 2 (Appendix B), and both may be considered indicators of low resistance.  Enchanter's-nightshade occurred only at Walker Creek and regained only 20% of its pre-burn cover post-burn; devil’s club also occurred at Walker Creek and Mackenzie, regaining only 33 and 13% of its original cover, respectively.  These results emphasize both the low resistance and narrow fundamental niche of these two species.  Devils’ club is the most widely used medicinal plant in British Columbia in terms of both the number of cultural groups that use it and the number of illnesses it is used to treat (Johnson 2006); and its loss due to slashburning in ICH forest will have profound cultural as well as ecological implications. Another especially vulnerable species that may serve as a valuable indicator of ecosystem stability loss in nitrogen-poor soils is heart-leaved twayblade; there were 15 counts of this orchid species prior to clearcut and slashburn at Otter Creek and Walker Creek, but only a single count at Walker Creek 21 years after treatment.  Heart-leaved twayblade is a putative mixotrophic species, and thus may depend on mycorrhizal connections to neighbouring trees as a source of carbon (Selosse and Roy 2009).  Like other orchids, it relies on wide seed dispersal to move between patches of suitable habitat to overcome temporal discontinuity of habitats.  On the other hand, Goodyera spp. (PFT 6) and Platanthera spp. (PFT 5 and PFT 6), both also putative mixotrophs (Downie 1943; Taylor et al. 2002), were recorded at sites 31  where they were not found prior to treatment.  On the subset of 5 sites, northern green rein orchid (Platanthera aquilonis) was the single species representing PFT 5, and was recorded once (at Walker Creek) 21 years after slash burning.  This rare occurrence provides an example of unpredictable plant response to slashburning after clearcutting. Resilience While PFT 1 had the greatest increases of percent cover, these results may be biased by the planting of interior spruce.  Ingress of natural regeneration takes much longer and would likely have resulted in greater dominance of some of the other PFTs (Lieffers et al. 1996).  As conifers increased the most rapidly and became more important in driving the pattern for all vegetation, this provided a competitive advantage to the shade tolerant and semi-tolerant understory plants (PFT 2 and 7) by shading out the sun-requiring understory pioneers, and possibly hosting the understory mixotrophs (PFT 6) and shrub layer (PFT 9) via mycorrhizal networks (Simard et al. 2012).  The greatest increase in conifer diversity occurred 6-11 years post burn.  This puts into question long-term dominance of the plant community by the planted spruce and suggests that (i) natural ingress from the surrounding forests occurred over a very short period of time, and (ii) forest floor substrate was receptive to conifers for only a limited time due to rapid invasion by other early successional species.  The significant increase in percent cover and richness of PFTs 2 and 3 may have been attributed to the absence of overstory vegetation complemented by their having the greatest seed longevity of all PFTs (see Appendix B). There was an unequal response of PFTs to slashburning among the biogeoclimatic zones, as hypothesized.  The ICH and SBS sites were returning more quickly to a forest state than the ESSF sites (albeit a mixed forest in the SBS), with the greatest increase in conifers and evergreen deep shade specialists typically found beneath conifers (PFT 6) occurring at Goat River (ICH).  Walker Creek (ICH) remained a fairly open meadow dominated by PFT 2; however there was an increasing presence of conifers in the second decade post-burn; a mixture of PFTs 3, 4, 6 and 9; and rare occurrences of PFTs 5 32  and 7.  Slow regeneration of conifers, PFT 6 and PFT 9 occurred at Herron (ESSF). However, these three groups showed a trend towards dominance 10 years post-burn.  Otter Creek (ESSF) may have transitioned to an alternate stability domain, an open meadow or potentially a heathland ecosystem dominated by PFTs 2 and 9.  Therefore, resilience of conifer forest in the ESSF appears limited by the combination of low MAT and high MAP, while the inverse trends in climate may be responsible for the current but possibly temporary dominance of deciduous trees at Mackenzie. There is a natural pattern of succession that is expected with establishment of gap specialists and deciduous trees and tall shrubs following disturbance (Coates 2002; Clark et al. 2003; Simard et al. 2004).  Over time, these pioneering species should decline, beginning with the gap specialists, and particularly those not well adapted to shade.  Clear patterns of the rapid rise and fall of the shade intolerant species (PFT 3) occurred within 5 or 10 years post-burn in the ICH and SBS as expected.  It could be several decades longer before the conifers begin replacing the deciduous trees and tall shrubs.  Without additional disturbance it could take 250-300 years for the conifers to assume dominance in the SBS and ICH, and generally ESSF ecosystems will take longer, approximately 400 years or more (Parminter 1983).  Therefore, care must be taken to not mistake shifts in PFT dominance that occur with natural succession for indicators of high/low resilience. Similar to our results, Aubin et al. (2007) found that the understory of sugar maple forests generally maintained its functionality despite over 200 years of human disturbance that has degraded tree-stem quality of northern hardwood forest (Coulombe et al. 2004).  Their results indicate that, despite disturbance, the ecological integrity of sugar maple forests measured by biodiversity and ecosystem services remains and provides the potential of the tree stratum of these forests to recover naturally. Conclusion This chapter studied the response of conifer forest to logging and slashburning by measuring both individual species as well as PFTs.  There were no substantial differences in changes in species 33  composition, before and after slashburning, between ecosystem types as expected.  However, measuring the response of individual species identified several species that were sensitive to the forestry treatment addressed.  I also found that the response of PFTs to slashburning was unequally distributed among the subzones, identifying shifts in PFT dominance that suggest the plant communities are on trajectories towards conifer or mixed forest, heathland and/or open meadow ecosystems.  34  Table 3.1. Species recorded exclusively pre-burn (underlined) or post-burn for each PFT, noted by site. See Appendix 1 for species codes.      PFT Pre-burn Post-burn1 TSUGHET THUJPLI2 ARNI_SP ATHYFIL CALACAN EPILANG GALITRF PTERAQU RUBUPARSAMBRAC STREAMP3 ANAPMAR CARE_SP CINNLAT CIRS_SP FRAGVIR HIERALB HIERUMBLUZUPAR POA_SP RUBUIDA TARAOFF4 POPUBAL SALIX_SP6 DIPHCOM9 RHODALB MENZFERHerronPFT Pre-burn Post-burn1 PICEENE PINUCON2 RIBEGLA ACHIMIL CALACAN EPILANG LUPI_SP OSMOBER RIBELAC SAMBRACSTRELAN SYMPFOL VALESIT3 AGROEXA CARE_SP HIERALB LUZUPAR RIBELAX RUBUIDA TARAOFF4 SALI_SP6 MONEUNI DIPHCOM PYROASA7 FESTOCCMackenziePFT Pre-burn Post-burn2 GALIBOR ACHIMIL AMELALN ARALNUD ARNI_SP BOTRLUN BOTRVIR CASTMINEPILANG MAIARAC OSMOBER PETAPAL PROSHOO RIBEHUD VIBUEDU3 RANU_SP RUBUIDA ANAPMAR EQUIARV HIERALB HIERAUR POA_SP RHINMIN SENEPAUTARAOFF4 BETUPAP POPUBAL SALI_SP6 RUBUPED CHIMUMB DIPHCOM GOODOBL MONEUNI7 SORBSCOOtter CreekPFT Pre-burn Post-burn1 PICEGLA PINUCON2 EPILANG GALITRF LUPI_SP MITEPEN SAMBRAC SENETRI TRISCER3 ANAPMAR CARE_SP EQUIARV HIERALB RUBUIDA4 SALI_SP6 ORTHSEC LISTCOR LYCOANN LYCOCLAPFT Pre-burn Post-burn1 PICEENE THUJPLI TSUGHET TSUGMER2 PROSHOO ACTARUB ARNI_SP BOTRVIR CALACAN ELYMGLA MITENUD PETAPALRUBUPUB TRISCER URTIDIO VAHLATR VALESIT3 EPILCIL ANAPMAR GEUMMAC HIERALB HIERUMB4 BETUPAP POPUBAL POPUTRE5 PLATAQU6 DIPHCOM HUPEOCC LYCOCLA PLATOBT PLATSTR PYROMINWalker CreekGoat River35  Table 3.2. Pre-burn and post-burn representatives identified as 3 most dominant species, measured by greatest total percent cover, within each PFT noted by site (species with <1% cover excluded). See Appendix 1 for species codes.      Goat RiverPFT1 ABIELAS TSUGHET PICEENE2 GYMNDRY DRYOEXP STRELAN EPILANG GYMNDRY CALACAN3 RIBELAX CARE_SP RIBELAX HIERALB4 SALI_SP POPUBAL6 RUBUPED CORNCAN LYCOANN CORNCAN RUBUPED LYCOANN7 SORBSCO SORBSCO9 VACCMEM MENZFER RHODALB VACCMEM VACCOVAHerronPFT1 ABIELAS PINUCON ABIELAS PICEENE2 ARNI_SP RUBUPAR EPILANG SYMP_SP RUBUPAR3 RUBUIDA HIERALB RIBELAX4 ALNU_SP ALNU_SP SALI_SP6 CORNCAN ORTHSEC LINNBOR CORNCAN LINNBOR ORTHSEC7 SORBSCO CLINUNI SORBSCO CLINUNI9 VACCMEM VACCMEMMackenziePFT1 ABIELAS PICEENE ABIELAS2 GYMNDRY OPLOHOR ACERGLA GYMNDRY RUBUPAR EPILANG3 TARAOFF HIERALB ANAPMAR4 ALNU_SP ALNU_SP POPUBAL BETUPAP6 RUBUPED CORNCAN CORNCAN ORTHSEC GOODOBL7 CLINUNI CLINUNI SORBSCOOtter CreekPFT1 ABIELAS PICEENE PICEENE PICEGLA ABIELAS2 GYMNDRY VALESIT TIARTRI2 ARNI_SP VALESIT GYMNDRY3 LUZUPAR POA_SP ANAPMAR HIERALB LUZUPAR4 SALI_SP6 RUBUPED LISTCOR PLATSTR RUBUPED LYCOCLA7 CLINUNI SORBSIT CLINUNI SORBSIT9 MENZFER RHODALB VACCMEM VACCMEM MENZFER VACCOVAWalker CreekPFT1 ABIELAS PICEENE THUJPLI TSUGHET2 OPLOHOR GYMNDRY RUBUPAR GYMNDRY EPILANG ATHYFIL3 RUBUIDA RIBELAX EQUIARV EQUIARV CINNLAT RIBELAX4 ALNU_SP SALI_SP ALNU_SP SALI_SP BETUPAP5 PLATAQU6 RUBUPED CORNCAN LYCOANN CORNCAN LYCOANN RUBUPED7 CLINUNI CLINUNI9 VACCOVA MENZFER VACCMEM VACCOVA MENZFER VACCMEMPre-burn Post-burnPre-burn Post-burnPre-burn Post-burnPre-burn Post-burnPre-burn Post-burn36  Table 3.3. Mean percent cover (3.3a) and richness (3.3b) of PFT before and 17-22 years after slashburn for all five sites (n = 5). Significant differences (p<0.05) are indicated by bold p-values.  Table 3.3a.    PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 Pre-burn 3.4 33.9 0.2 1.4 0.0 5.6 0.9 0.0 14.2 Post-burn 26.6 71.1 8.1 12.7 0.0 18.8 2.0 0.0 17.3 p-value 0.015 0.028 <0.001 0.185 0.374 0.058 0.030 NA 0.293  Table 3.3b.    PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 Pre-burn 0.6 5.4 0.2 0.2 0.0 1.6 0.4 0.0 1.3 Post-burn 1.3 8.9 3.5 0.7 0.0 2.0 0.5 0.0 1.2 p-value 0.036 0.004 0.024 0.105 0.374 0.215 0.207 NA 0.693   37  Table 3.4. Mean percent cover (3.4a) and richness (3.4b) of PFT before and 17-22 years after slashburn for each of the three BEC zones (n = number of plots): ESSF (n = 44), ICH (n = 126), and SBS (n = 6). Significant differences (p<0.05) are indicated by bold p-values.  Table 3.4a.   PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 ESSF Pre-burn 3.2 59.1 0.1 0.0 0.0 1.6 2.2 0.0 43.8 ESSF Post-burn 10.4 89.0 8.1 0.9 0.0 7.3 3.3 0.0 49.0 p-value <0.001 0.016 <0.001 0.293 NA 0.022 0.027 NA 0.218 ICH Pre-burn 0.6 41.0 0.8 1.8 0.0 2.6 0.0 0.0 2.4 ICH Post-burn 27.4 117.1 8.4 14.4 0.0 11.1 0.2 0.0 11.7 p-value <0.001 <0.001 <0.001 <0.001 1.000 <0.001 0.062 NA <0.001 SBS Pre-burn 1.0 44.2 0.1 4.7 0.0 0.9 0.7 0.0 0.0 SBS Post-burn 34.3 79.6 5.9 43.0 0.0 9.7 2.7 0.0 0.0 p-value 0.031 0.031 0.031 0.031 NA 0.156 0.590 NA NA  Table 3.4b.   PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 ESSF Pre-burn 0.4 6.2 0.2 0.0 0.0 1.0 0.5 0.0 2.9 ESSF Post-burn 1.3 8.8 2.5 0.1 0.0 1.0 0.6 0.0 2.8 p-value <0.001 <0.001 <0.001 0.484 NA 0.630 0.182 NA 0.379 ICH Pre-burn 0.1 5.5 0.3 0.1 0.0 0.8 0.0 0.0 0.3 ICH Post-burn 0.8 8.0 1.2 0.4 0.0 1.0 0.0 0.0 0.6 p-value <0.001 <0.001 <0.001 <0.001 1.000 0.032 0.233 NA <0.001 SBS Pre-burn 0.5 11.5 0.5 0.5 0.0 1.3 0.7 0.0 0.0 SBS Post-burn 1.5 16.8 5.2 1.8 0.0 2.8 1.0 0.0 0.0 p-value 0.048 0.059 0.034 0.053 NA 0.149 0.346 NA NA    38  Table 3.5. Mean percent cover (3.5a) and richness (3.5b) of PFT before and 17-22 years after slashburn for five sites (n = number of plots): Goat River (n = 6, 19 yrs), Herron (n = 10, 17 yrs), Mackenzie (n = 6, 19 yrs), Otter Creek (n = 34, 20 yrs) and Walker Creek (n = 120, 21 yrs). Significant differences (p<0.05) are indicated by bold p-values.  Table 3.5a.   PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 GR Pre-burn 9.2 2.3 0.2 0.0 0.0 21.2 0.8 0.0 5.5 GR Post-burn 43.7 22.4 8.5 2.6 0.0 45.5 1.7 0.0 0.6 p-value 0.004 0.003 0.007 0.218 NA 0.024 0.259 NA 0.023 HR Pre-burn 3.5 5.3 0.0 0.2 0.0 3.0 0.3 0.0 9.7 HR Post-burn 21.3 23.4 10.8 2.8 0.0 28.7 1.6 0.0 14.9 p-value <0.001 0.001 <0.001 0.336 NA <0.001 0.130 NA 0.198 MA Pre-burn 1.0 44.2 0.1 4.7 0.0 0.9 0.7 0.0 0.0 MA Post-burn 34.3 79.6 5.9 43.0 0.0 9.7 2.7 0.0 0.0 p-value <0.001 0.029 0.052 0.028 NA 0.266 0.300 NA NA OC Pre-burn 3.1 74.9 0.1 0.0 0.0 1.2 2.8 0.0 53.8 OC Post-burn 7.2 108.4 7.3 0.3 0.0 1.0 3.8 0.0 59.0 p-value 0.144 0.010 <0.001 0.325 NA 0.638 0.107 NA 0.267 WC Pre-burn 0.2 43.0 0.8 1.9 0.0 1.6 0.0 0.0 2.2 WC Post-burn 26.5 121.9 8.4 15.0 0.0 9.4 0.1 0.0 12.2 p-value <0.001 <0.001 <0.001 <0.001 0.319 <0.001 0.159 NA <0.001  Table 3.5b.   PFT 1 PFT 2 PFT 3 PFT 4 PFT 5 PFT 6 PFT 7 PFT 8 PFT 9 GR Pre-burn 1.5 1.0 0.2 0.0 0.0 2.8 0.3 0.0 1.7 GR Post-burn 1.3 5.5 6.0 0.8 0.0 3.0 0.3 0.0 1.2 p-value 0.611 <0.001 <0.001 0.093 NA 0.741 NA NA 0.203 HR Pre-burn 0.7 1.0 0.0 0.2 0.0 2.2 0.4 0.0 1.0 HR Post-burn 1.5 4.0 3.4 0.3 0.0 2.7 0.4 0.0 1.0 p-value <0.001 0.007 <0.001 0.678 NA 0.096 1.000 NA NA MA Pre-burn 0.5 11.5 0.5 0.5 0.0 1.3 0.7 0.0 0.0 MA Post-burn 1.5 16.8 5.2 1.8 0.0 2.8 1.0 0.0 0.0 p-value 0.012 0.014 0.003 0.010 NA 0.076 0.175 NA NA OC Pre-burn 0.3 7.7 0.2 0.0 0.0 0.7 0.5 0.0 3.5 OC Post-burn 1.2 10.2 2.2 0.0 0.0 0.4 0.6 0.0 3.4 p-value <0.001 <0.001 <0.001 0.325 NA 0.133 0.103 NA 0.379 WC Pre-burn 0.0 5.8 0.3 0.1 0.0 0.7 0.0 0.0 0.2 WC Post-burn 0.7 8.2 1.0 0.4 0.0 0.9 0.0 0.0 0.6 p-value <0.001 <0.001 <0.001 <0.001 0.319 0.039 0.181 NA <0.001  39        Figure 3.1. Mean annual temperature (MAT) and mean annual precipitation (MAP) from 1980 to 2002 at five sites: Goat River (GR), Herron (HR), Mackenzie (MA), Otter Creek (OC), and Walker Creek (WC). Data derived from ClimateBC (Wang et al. 2006).   40  Figure 3.2a. Goat River     Figure 3.2b. Walker Creek     Figure 3.2. Plant functional type response to slashburning after clearcutting over time measured at Goat River (3.2a) and Walker Creek (3.2b). These sites are Interior Cedar-Hemlock (ICH) forest ecosystems located in central British Columbia.   41  Figure 3.3a. Herron     Figure 3.3b. Otter Creek     Figure 3.3. Plant functional type response to slashburning after clearcutting over time measured at Herron (3.3a) and Otter Creek (3.3b). These sites are Engelmann Spruce-Subalpine Fir (ESSF) forest ecosystems located in central British Columbia.  42      Figure 3.4. Plant functional type response to slashburning after clearcutting over time measured at Mackenzie River. This site is a Sub-Boreal Spruce (SBS) forest ecosystem located in central British Columbia.  43  Chapter 4. Plant functional type response over time to gradients of fire severity and soil nitrogen  Many plant species in sub-boreal and temperate forests depend on some aspect of fire to complete their life cycle.  Those plants where seed release is determined by fire are termed pyriscent and can be influenced by fire in many ways including properties of the fire itself (such as its intensity, speed, and timing), and/or fire effects on edaphic factors (such as woody debris and char or the presence of smoke).  Seed germination requirements of pyriscent plants can vary; nitrogenous compounds and/or charred wood can promote seed germination of some fire annuals (Thanos and Rundel 1995).  Brown and Johnstone (2012) found that a shortened fire return interval can reduce on-site seed availability and therefore limit recruitment of black spruce (Picea mariana [(Mill.) BSP.]).  Relationships between fire severity and seed success after logging can be nonlinear; seed-origin plants are more likely to establish on sites that are severely burned but less likely to establish on lightly burned sites compared to unburned sites (Stark et al. 2006).  Moreover, species-poor communities on mesic soils can be readily invaded by exotic species following burning, while in the absence of fire, the relationship between species richness and seed invasion may be less evident (MacDougall 2005).  The above findings indicate that establishment and composition of plant communities following fire depends not only on individual plant traits, but also on (i) fire severity, (ii) site productivity, and (iii) the diversity, composition and fire-sensitivity of the existing plant community. Although severity and type of disturbance favour species differently, forest soils are not necessarily degraded by fire.  This is true even in nutrient poor ecosystems where fire is relatively infrequent (Haeussler and Kneeshaw 2003; Kranabetter et al. 2007).  Between 10 and 20 years after harvest and moderate slashburn (mean depth of burn between 1.3 and 3.7 cm), lodgepole pine (Pinus contorta var. latifolia [Douglas]) plantations in the sub-boreal forests of north central British Columbia had produced sufficient biomass that the rate of carbon accumulation exceeded ongoing carbon losses, despite total carbon levels remaining lower than pre-harvest levels (Kranabetter and Macadam 2007). 44  Because of the large legacy of nitrogen stored in soils of these forests, total nitrogen may change little over time since fire (Driscoll et al. 1999).  In contrast to total nitrogen, available nitrogen has been found to increase immediately following fire (i.e., assart effect; Kimmins 2004), followed by declines due to immobilization by the developing plant community and microbial biomass (Forge and Simard 2000; Wan et al. 2001).  However, available soil nitrogen has been shown to increase and C:N ratio to decrease along gradients of increasing tree productivity in fire-origin boreal forests of British Columbia (Kranabetter et al. 2007).  Plant diversity has also been found to control microbial nitrogen mineralization (Zak 2003), and tree species composition can influence net nitrification both spatially and temporally (Aubert et al. 2005).  These findings highlight the complexity of the relationship between plants and soils, and how they may be affected by fire over time.  However, there still remains a shortage of long term studies that monitor plant communities and soils in areas of relatively infrequent fire (Bradstock and Kenny 2003). The objective of this study was to use a PFT approach to assess plant community recovery from slashburning following clearcutting along gradients of soil nitrogen, fire severity and time.  PFTs may be a good approach to analyzing plant responses to gradients for several reasons.  First, there is merit to being able to condense a large dataset representing a great number of species (in this case 163 species) into a much smaller number of groups that represent core functions and strategies within a complex ecosystem.  Because of the reduction in sparsity (total number of zeros in a matrix) that occurs with the grouping of species into PFT, the analysis of groups rather than individual species may be simpler and produce more reliable and intelligible results.  This becomes apparent when examining results indicating how each PFT responds to each ecological gradient, the interrelationships of the PFTs, and the individuals represented by each.  For this chapter, I used the PFTs produced by cluster analysis and reported in Chapter 2.  See Appendix D for a summary of PFTs and their characteristics. 45  My first hypothesis is that all PFTs representing early seral species (including gap specialists, deciduous trees and shrubs, and species adapted to open habitats and surface water) will increase in cover and richness with increasing fire severity.  My second hypothesis is that representatives of mature conifer forests, generally species that are shade tolerant and thrive beneath a closed canopy, will decline with increasing fire severity (this includes mixotrophs, ericaceous shrubs, and the remaining species with mor humus form as the primary ground surface material) while the opposite will occur for pyriscent conifers.  The third hypothesis is that all PFTs will increase in cover and richness with increasing soil nitrogen with the exception of the conifers and species most adapted to the nutrient poor and acidic soil conditions typical of developed conifer forest (e.g., mixotrophs and ericaceous shrubs).  Finally, with hypothesis 4, I expected that time since burn moderates PFT response to both nitrogen and fire severity. Methods Field procedures - measuring fire severity I use the term fire severity as it relates to fire effects as opposed to fire intensity which refers to measures of energy output (Keeley 2009).  Forest floor (L, F and H layers) and woody fuel loads were measured using the line intersect method along fuel assessment triangles (Trowbridge et al. 1987) prior to and immediately after burning.  Along each 30 m segment of each triangle, five depth-of-burn pins were evenly distributed and inserted to carry out depth measurements.  Depth-of-burn pins were also established in the vegetation sampling quadrats with metal rods marking the corners of square quadrats and the centres of circular quadrats (Hamilton and Haeussler 2008).  These data were provided by the BC Ministry of Forests, Research Branch. Field procedures - measuring soil nutrients For direct measurement of soil nutrients (%N and %C), I collected composite mineral soil samples between September 3rd and 9th, 2008.  Each sample comprised of three subsamples for each 25 m2 plot 46  at the 5 study sites with burn triangles (Brinks Mill, Francis Lake, Genevieve Lake, Goat River and Mackenzie).  At Otter Creek, sample plots (3 × 3 m) were located at 15 m intervals along two 150 m transects and three 75 m transects within burned areas (see Hamilton and Peterson 2003 for further details on sample design at Otter Creek).  Two composite samples were collected along the three 75 m transects and four composite samples were collected along the 150 m transects.  At Walker Creek, three samples were collected at each of the 5 blocks.  In total, I collected 56 composite mineral soil samples that were analyzed at the Forestry and Technical Services Section, Research Branch Laboratory in Victoria B.C.  Soil nutrient data for the remaining sites used in this analysis for this chapter (Helene, Herron and Walcott) was provided by the BC Ministry of Forests, Lands and Natural Resource Operations, Skeena Region Research Section. Laboratory analysis - measuring soil %N and %C The air-dried soil samples were ground to -200 mesh using a Rocklabs “Ring Grinder”.  The analysis was performed on this sample material using a Fisons (Carlo-Erba) NA-1500 combustion elemental analyzer, providing % total N and C by weight.  A separate sub-sample for each sample was oven-dried to determine moisture content and this value was used to correct the air-dry results from the instrument to an oven-dry basis. Statistical analysis Of the 16 sites in the dataset, 10 had vegetation surveys completed more than 10 years after slashburning, as well as fire severity and soil %N measurements.  These 10 sites are: Brinks Mill, Francis Lake, Genevieve Lake, Goat River, Helene, Herron, Mackenzie, Otter Creek, Walcott, and Walker Creek.   To address differences in the response of PFT % cover and richness to gradients of fire severity and soil nitrogen (Hypotheses 1, 2 and 3), heuristic classes for each of the two explanatory variables were created.  Fire severity classes (% LFH consumed) were: low (10-20%), n=30; moderate (20-30%), n=168; 47  and high (>30%), n=21.  Soil nitrogen classes (% total nitrogen) were: low (0.0-0.1%), n=36; moderate-low (0.11-0.2%), n=20; moderate-high (0.21-0.3%), n=126; and high (0.31-0.4%), n=37. The species matrix was comprised of 168 species x 219 plots (from 10 sites) recorded at 5 time intervals (1, 3, 5, 10 and 20 years post-burn) for a total of 1095 data points.  Out of these 1095 data points, 92% of entries were zeros.  However, for the PFT matrices (PFT 1 through 9 abundance or richness, each with 1095 observations), 59% of entries were zeros, thereby reducing dataset sparsity by 33%.  Accordingly, the dataset was not normally distributed.  Therefore, pairwise Wilcoxon rank sum tests (Wilcoxon 1945), non-parametric statistical hypothesis tests, were used to test for significant differences (p≤0.05) in abundance and richness among classes of fire severity and soil nitrogen at 1, 3, 5, 10, and 20 years.  The Bonferroni method was used to correct the p-values for multiple comparisons. Relationships between soil N and years since slashburning were examined to determine whether there were lasting effects of disturbance on soil N, and to account for species influences on soil N over time.  To determine if PFT responses to either of the two gradients varied with time, relationships were assessed between time (years post-burn) and PFT abundance and richness for each corresponding gradient class of fire severity and soil nitrogen.  Three models (linear, quadratic, and cubic) were tested for each PFT for each gradient class and from these, PFTs with relationships of R2adj ≥ 0.25 in at least one or more gradient classes were reported (Hamilton and Haeussler 2008; Kranabetter et al. 2010).  Statistical analysis for this chapter was carried out with R, version 3.0.1 (The R Foundation for Statistical Computing 2013). Results PFT response to fire severity gradient   High fire severity favoured abundance and richness of conifers (PFT 1), deciduous trees (PFT 4), surface water specialists (PFT 5), and montane boreal and cool temperate understory specialists (PFT 8).  However, PFT responses were not consistent through time, response patterns were not linear, and the 48  response patterns of abundance and richness were not always the same (Table 4.1; Appendix E).  By year 20, most PFTs had greatest abundance and richness in the highest severity burn class (>30% LFH consumption).  Exceptions to this included semi-shade tolerant gap specialists (PFT 2) with highest abundance and richness at moderate fire severity (20-30% LFH consumption); and shade intolerant gap specialists (PFT 3) with higher richness, and ericaceous shrubs (PFT 9) with higher abundance and richness at the lowest fire severity class (10-20% LFH consumption). PFT response to soil nitrogen gradient Generally, by year 20, response of most PFTs to nitrogen was the inverse of the response to fire severity (Table 4.2; Appendix E).  As with most PFTs, conifers had higher abundance and richness at low (0-0.1%) or moderate-low (0.1-0.2%) levels of soil N compared to moderate-high (0.2-0.3%) and high (0.3-0.4%) levels of soil N.  As with fire severity, exceptions included semi-shade tolerant gap specialists with higher abundance and richness at moderate-high and high levels of soil N; shade intolerant gap specialists with higher abundance and richness at the moderate -low levels of soil N; and ericaceous shrubs with highest abundance and richness at the highest levels of soil N. Curvilinear models  Models of percent cover and richness over time for each PFT and for both fire severity and soil N were significant (p ≤ 0.05), except for PFT 7 at the lowest LFH consumption class (p = 0.10).  Models with R2adj < 0.25 in all gradient classes of either fire severity or soil N were not reported.  Fire severity was correlated with abundance of PFTs 1, 4, 6, 7 and 9, and with richness of PFTs 1, 3 and 6.  Specifically, abundance of PFTs 1, 4, 6, 7 and 9 had highest correlation with >30% LFH consumption (R2adj = 0.79, R2adj = 0.27, R2adj = 0.41, R2adj = 0.26, and R2adj = 0.29 respectively) (Figure 4.1).  Richness of PFTs 1 and 3 were most strongly correlated with 10-20% LFH consumption (R2adj = 0.33 and R2adj = 0.27 respectively); and richness of PFT 6 had highest correlation with >30% LFH consumption (R2adj = 0.25) (Figure 4.1b).  Soil nitrogen was correlated with abundance of PFTs 1, 4, 6 and 9, and with richness of PFTs 1 and 6.  49  Abundance of PFT 1 had highest correlation at 0-0.1% Nsoil (R2adj = 0.69); abundance of PFTs 4 and 9 had highest correlation at 0.3-0.4% Nsoil (R2adj = 0.28, R2adj = 0.59 respectively); and abundance of PFT 6 had highest correlation at 0.1-0.2% Nsoil (R2adj = 0.43) (Figure 4.2a).  Richness of PFTs 1 and 6 had highest correlation at 01.-0.2% Nsoil (R2adj = 0.55, R2adj = 0.29 respectively) (Figure 4.2b). Discussion I found that the abundance and richness of PFTs were related to fire severity and soil N, consistent with previous studies (Kleyer 1999; Chen et al. 2004).  My results showed that these relationships can be complex, relate to specific functional types and indicate relationships that change through time.  Because these findings were derived from a variety of study sites that not only vary in fire severity and soil N level, but also in other factors important to these ecological systems such as elevation and climate, it is possible that my results reflect differences among study sites in addition to the two gradients.  For example, these sites have different  climate (Figure 3.1) and results from Chapter 3 illustrate differences in PFT response vis-à-vis individual sites over time where the initial measurement (year 0) was conducted prior to slashburning and the following measurements were conducted post-burn (Figures 3.2, 3.3 and 3.4). Generally, PFTs exhibited different patterns of response to fire severity and soil N.  Richness of PFT 4 and 6 indicate the same pattern with respect to soil N for all years, with greater richness at lower levels of soil N.  The strategy of both groups involves early establishment of disturbed areas low in soil N.  However, their methods differ in both establishment strategy and mechanisms for survival in soils with limited soil N.  As mentioned in Chapters 2 and 3, PFT 4 produces larger seeds in great numbers, whereas PFT 6 produces minute seeds with high potential for dispersal.  PFT 4 represents actinorhizal species while PFT 6 represents mixotrophic species. 50  Fire severity Like Stark et al. (2006), I found a positive relationship between fire severity and abundance of seed origin plants.  My findings also agree with Hamilton and Haeussler (2008), whose results depended on the stability agent of interest (matter or information; see Figure 1.1).  In the current study, the response of several PFTs, including the conifers, differed depending on whether abundance or richness was considered.  Overall, my results support my first hypothesis that abundance and richness of ruderal species increase with increasing fire severity, although it appears that the optimal fire severity (i.e. as indicated by high abundance and richness) differs among PFTs.  All measurements across time since burn showed that abundance of the deciduous trees and shrubs (PFT 4) was generally highest with high fire severity.  The response of the two groups of gap specialists differed: initially PFT 2 abundance was lower at >30% LFH consumption while cover of PFT 3 was higher indicating a possible threshold (with inverse effects) for these two groups. Of the mature forest species, only the ericaceous shrubs (PFT 9) generally had lowest abundance and richness at high fire severity, contrary to my second hypothesis predicting reduced mature forest species with increasing fire severity.  During the first twenty years of regeneration, the abundance and richness of the conifer group (PFT 1) was highest on the highest severity burns.  The mixotrophs (PFT 6) and ground surface material generalists (PFT 7) had greatest cover and richness first where burns were most severe, then later where they were least severe.  These findings may be explained by the dominance of gap specialists, which rapidly occupy available space; these originated from the seed- and budbank on low to moderate burns (10-30% LFH consumption for PFT 2) or moderate to high severity burns (20->30% LFH consumption for PFT 3), rather than re-establishing from off-site seed sources, as is often the case for PFT 6.  Montane boreal plants (PFT 8) were most abundant on sites with highest fire severity (similar to the shade-intolerant early seral species, PFT 3).  Although the species of PFT 8 (false toad-flax (Geocaulon lividum), wild lily-of-the-valley (Maianthemum canadense) and velvet-leaved 51  blueberry (Vaccinium myrtilloides)) are usually associated with mor humus form, they rapidly regenerate vegetatively from underground rhizomes following fire while also producing berries whose seeds are widely dispersed by birds and mammals, providing two mechanisms for recovery after fire. Soil nitrogen gradient My results indicate specialized temporal relationships between PFT composition and soil N, consistent with the results of Aubert et al. (2005) and supporting my third hypothesis.  My results show that plant abundance and species richness increase with soil N, at least for certain PFTs, up to certain levels of soil N, and within a certain time periods.  Two groups of gap specialists had opposing preferences in soil N; semi-shade tolerant plants (PFT 2) had higher abundance and richness at higher levels of soil N while abundance and richness of shade intolerant gap specialists (PFT 3) were generally highest at moderate levels of soil N. In the second decade post-burn, resilience measures of deciduous trees and shrubs (PFT 4) corresponded exactly with higher abundance and richness at lower levels of soil N throughout all years.  This included year 20, when resilience was low only at the highest level of soil N.  This makes sense because PFT 4 includes all deciduous trees and shrubs recorded on the study sites, including Sitka alder (Alnus crispa) and mountain alder (Alnus tenuifolia) which were representative species for this PFT pre- and post-burn on most sites.  Alnus spp. are unique because they are actinorhizal - able to fix N from the atmosphere - a strategy that allows them to readily outcompete other early seral species where soils are nutrient limited (Binkley 1982).  Ground surface material generalists (PFT 7) tended to have greater abundance and richness at the highest and lowest levels of soil N through the entire time sequence measured. Abundance and richness of the conifers and mixotrophs (PFTs 1 and 6) were greater at lower soil N levels.  Because of the dominance of conifers, it may be possible that the nitrogen status was due to species influence on soil N rather than an accurate predictor for vegetation response alone.  As 52  expected, the mixotrophs closely reflected the trends exhibited by the conifers.  Of all the PFTs, the mixotrophs require the least light, preferring the shade of the closed canopy.  Most of these species are evergreen, thereby reducing their nutritional requirements and acquire nutrition from surrounding plants.  Ericaceous shrubs (PFT 9) had consistently higher richness at low levels of soil N compared to moderate low and/or moderate-high soil N, but surprisingly highest richness was associated with the highest level of soil N. Time Time played a critical role in moderating the PFT abundance and richness response to both fire severity and soil N status.  Contrary to the initial findings, results indicated that into the second decade post-burn, (i) abundance of the shade intolerant gap specialists (PFT 3) was generally greater at higher fire severity; (ii) richness of conifers (PFT 1) was generally greater at higher fire severity; (iii) and abundance and richness of surface water specialists (PFT 5) was influenced by fire severity and soil N only after first records of these species in years 10 and 20. Conclusion Both 20-year and shorter term trends and thresholds in PFT abundance and richness in the context of gradients of fire severity and soil N were identified.  PFTs representing early seral species generally increased in cover and richness with increasing fire severity; however, the semi-shade tolerant species significantly declined at the highest level of fire severity and indicated a long-term trend in optimum abundance at moderate fire severity.  Representatives of mature conifer forest declined at the level of moderate fire severity exhibiting optimum at either low or high fire severity.  These non-linear trends closely resembled those of the conifers, with the exception of ericaceous shrub abundance which declined with higher fire severity after the third year.  For most PFTs, optimum soil N was either low or moderate; exceptions included the gap specialists which reached optimum at either moderate or high levels of soil N, and the ericaceous shrubs which reached optimum at high levels of soil N.53  Table 4.1. Wilcoxon tests with the Bonferroni correction found significant differences (p<0.05) in abundance and richness for low (I), moderate (II) and high (III) classes of % LFH consumption. Symbols are as follows: very low (─ ─); low (─); high (+); very high (+ +); not different from + or ─ (±); no difference between classes (.); PFT not recorded (na). See Appendix E for mean and standard deviation for abundance and richness of each PFT for each gradient class.      % LFH PFT1 PFT2 PFT3 PFT4 PFT5 PFT6 PFT7 PFT8 PFT9 % LFH PFT1 PFT2 PFT3 PFT4 PFT5 PFT6 PFT7 PFT8 PFT9Year 1 Year 1I ─ ─ ─ ─ na + + na + I ─ ± ─ ─ na + + na + +II ─ + + ─ na ─ ─ na ─ II ─ + + ─ na ─ ─ na ─III + ─ + + na + + + na + III + ─ + + na + + na +Year 3 Year 3I + + ─ ─ na + + ± + I + ± ─ ─ na + + ± +II ─ + + + ─ na ─ ─ ─ ─ II ─ + + ─ na ─ ─ ─ ─III + ─ + + na + + + ─ III + + ─ ─ + na + + + ─Year 5 Year 5I + + ─ ─ na + + na + I + ± ─ ─ na + + na +II ─ + + + ─ na ─ ─ na ─ II ─ + + ─ na ─ ─ na ─III + + ─ + + na + + na ─ III + + ─ ─ + na + + + na ─Year 10 Year 10I ─ ─ . ─ ─ ─ + ± + + I ─ ─ . ─ ─ + + ± +II ─ + . ─ + ─ ─ ─ + II ─ + . ─ ─ ─ ─ ─ ─III + ─ . + + + + + + ─ III + ─ . + + + + + + +Year 20 Year 20I ─ + ─ ─ ± ± + ± + I ± ± + ─ ± ± + ± + +II ─ + + + + ─ ─ ─ ─ ─ II ─ + ─ ± ─ ─ ─ ─ ─III + ─ + + + + + + + ─ III + ─ ─ + + + + + + +RichnessAbundance54  Table 4.2. Wilcoxon tests with the Bonferroni correction found significant differences (p<0.05) in abundance and richness for low (I), low-moderate (II), moderate-high (III) and high (IV) classes of soil N. Symbols are as follows: very low (─ ─); low (─); high (+); very high (+ +); not different from + or ─ (±); different only from + (. +); different only from ─ (. ─); no difference between classes (.); PFT not recorded (na). See Appendix E for mean and standard deviation for abundance and richness of each PFT for each gradient class.    Abundance RichnessSoil N PFT1 PFT2 PFT3 PFT4 PFT5 PFT6 PFT7 PFT8 PFT9 Soil N PFT1 PFT2 PFT3 PFT4 PFT5 PFT6 PFT7 PFT8 PFT9Year 1 Year 1I + ─ ± + na + + + na + I + ─ ± + na + + na +II ─ ─ ± + na + + na + II ─ ± ± + na + + na +III ─ + + ─ na ─ ─ na ─ III ─ + + ─ na ─ ─ na ─IV ─ + ─ ─ na ─ + na + + IV ─ + + ─ ─ na ─ + na + +Year 3 Year 3I + + ─ ─ ─ + na + + . ─ I + + ─ ─ + na + + . +II + ─ ± + na . ─ + . ─ II + ± ± + na + + . +III ± ± + ─ na . + ─ . + III ─ + + ─ na ─ ─ . ─IV ─ + ─ ─ na ─ + . + + IV ─ + + ± ─ na ─ + . + +Year 5 Year 5I + + ─ ─ + na + ± na ─ I + ─ ─ + na + ± na +II + ─ ± + na . ─ ─ na ─ II + ± ± + na + ─ na +III ± + + ─ na . + ─ ─ na + III ─ + + ─ na ─ ─ ─ na ─IV ─ + ─ ─ na ─ + na + + IV ─ + + + ─ na ─ + na + +Year 10 Year 10I + ─ . + ± ± + + + I + ─ ± + ± + + + +II + ─ . + + ─ + + + II + ─ ± + + + + + +III ─ + + . ─ ─ + ─ ─ ─ III ─ + ─ ─ ─ ─ ─ ─ ─IV ─ + . ─ ± ─ ─ + + + + IV + + + + + ─ ± ─ + + + +Year 20 Year 20I + ─ ─ + + + + + + ─ I ± ─ ─ + + + + + + + +II . ─ ─ + + ± + + + ± ± II + . + + + + ± + + + ± ─III . + + . + + ─ + ─ ─ + III ─ ─ + ─ + ─ + ─ ─ ─IV ─ + . ─ ─ ± ─ + ± + + IV ─ + + + ─ ± ─ + ± + +55  4.1a. Abundance PFT 1    PFT 4    PFT 6   PFT 7    PFT 9    4.1b. Richness PFT 1    PFT 3    PFT 6     Figure 4.1. Abundance (4.1a) and richness (4.1b) of PFTs for three classes of forest floor (LFH) consumption (10-20%; 20-30%; and >30%) modeled over time. Models with R2adj < 0.25 in all classes not reported; solid lines indicate R2adj ≥ 0.25. 56  4.2a. Abundance PFT 1    PFT 4    PFT 6      PFT 9    4.2b. Richness PFT 1    PFT 6    Figure 4.2. Abundance (4.2a) and richness (4.2b) of PFTs for four classes of soil nitrogen (0-0.1%; 0.1-0.2%; 0.2-0.3%; and 0.3-0.4%) modeled over time. Models with R2adj < 0.25 in all classes not reported; solid lines indicate R2adj ≥ 0.25.  57  Chapter 5. Modeling the relative importance of environmental factors controlling ecosystem resilience in ESSF, ICH and SBS forests of central British Columbia What we learned from the previous chapters is that there are many forms of matter (e.g., biogeochemistry) and expressions of information (e.g., species), and a myriad of intricate relationships that may occur between the two.  The use of methods to simplify and reduce complexity (i.e. developing PFTs) can assist in improving the understanding of how these processes coalesce to affect stability.  In the previous chapters, I have demonstrated that functional groups within forest systems respond differently to slashburning after clearcutting; for most groups, the response is dependent on soil nutrient availability and/or fire severity, and PFT response to fire severity and soil nutrients (soil %N and %C) is moderated by time.  This highlights the multiple factors involved in secondary succession within these forest communities. Structural equation modeling (SEM) with PFTs provides an opportunity to examine a complex forest system in a simplified form that provides insight to the relative importance of factors behaving simultaneously in that system.  Additionally, each PFT factor in the model includes both abundance and richness and therefore incorporates both information and matter into the analysis.  The purpose of this chapter is to test a more holistic approach to the ecological question of which factors (including precipitation, temperature, soil nutrients and fire severity) are important to the resilience of conifer forest after clearcutting and slashburning and which factors influence the transition to an alternate stability domain. Methods Structural equation modeling features The SEM technique is rooted in path analysis (Wright 1918) and can measure relationships between observed and latent variables (the measurement model) as well as relationships among latent variables (the structural model) (Grace 2010).  In practice, statistical procedures usually assume that independent variables are measured without error; however, SEM determines measurement error, and in doing so, 58  addresses bias in the model parameters and the overestimation of error in the dependent variables (Bollen 1989, Chapter 5; Pugesek et al. 2006, Chapter 7).  A Chi-square test statistic is used to test the difference between the sample covariance and the estimated population covariance matrices, and further model parameterization allows assessment of competing hypotheses (Tabachnick and Fidell 2007). SEM provides the capability to construct latent variables, that is, variables that are "conceptual" and therefore not directly measured but instead representative of a single variable or collection of measureable variables.  The latent variables provoke recognition of possible domains of abstractions or collections that can be meaningful in a multivariate quantitative space; the strength of their relationship to one another is measureable, and the relationship between them and the observed variables from which they are derived is measureable (Tabachnick and Fidell 2007).  Further, because the indicators are employed by a latent variable simultaneously, the need for running multiple separate analyses can be avoided thereby increasing efficiency for the researcher. SEM also provides output as a path diagram that assists interpretation and presentation of the model.  The path diagram includes basic recognizable shapes: variables that could not be measured but were represented by measurable variables (i.e. latent variables) are represented by ellipses, the measured variables (i.e. indicators) are represented by squares or rectangles, and relationships between the objects consist of paths represented by arrows that indicate the direction of a relationship.  Single-headed arrows represent regression coefficients (pointing to dependent variables and away from independent variables) while double-headed arrows represent variances or covariances.  Curved double-headed arrows represent variances of exogenous variables or error variances.  Where no path was present from one variable to another, it is assumed in the analytical procedure that no relationship exists. 59  The structural equation modeling method All statistical analysis for this chapter was performed in R (The R Foundation for Statistical Computing 2013).  SEM solutions can be based on the covariance matrix or the raw data, and many parameter estimation techniques are available, including generalized least squares (GLS), asymptotically distribution free (ADF; Browne 1984) and robust estimators (Satorra and Bentler 1994, 2010), but with most software, maximum likelihood (ML) is the default (Mueller and Hancock 2008). While the ADF method is recommended for continuous non-normal data, Schermelleh et al. (2003) advise that ML should be preferred because simulation studies show that it performs better than ADF with or without correction for non-normality (Hu et al. 1992; Olsson et al. 2000; Boomsma and Hoogland 2001). I carried out the analysis with the sem package (Fox et al. 2013; Fox 2006) using the covariance matrix and the ML estimation technique.  SEM output in R provided overall evaluation of the model with five types of information: goodness-of-fit measures, analysis of the residuals, measures of the variation that has been accounted for (R2) for endogenous latent variables, the standard errors and correlations of the parameter estimates, and model modification indices.  The analysis also included additional output by using separate functions in the sem package: the standardizedCoefficients function, which produced a solution where both the latent and the observed variables are standardized; the modIndices function, which provided modification indices; and the effects function, which generated the direct, indirect (i.e. mediation) and total effects (i.e., the sum of all direct and indirect effects) between variables (Fox 1980).  The R code is available in Appendix F; the sample covariance matrix in Appendix G. The measurement and structural model To begin my analysis, I created a structural equation meta-model (SEMM) (Grace et al. 2010) to identify the variables under study and conceptualize their relationships to one another (Figure 5.1).  Next, I replaced the ideas presented in the SEMM by measured variables and used this to develop the 60  measurement and structural model as input for the analysis.  The SEM hypothesis consisted of a path diagram that illustrated both a measurement model (relationships between observed and latent variables) and a structural model (relationships among latent variables).   Initially the SEM included all the PFTs.  However, several PFTs were removed in sequence, beginning with PFT 8, which consistently produced a coefficient estimate of “not applicable” (this PFT was represented by a single specimen of wild lily-of-the-valley (Maianthemum canadense) observed at the same plot at Francis Lake over five time intervals); followed by two PFTs with low R-square (PFT 3 R2 = 0.079; PFT 5 R2  = 0.015); and finally PFT 7 because its inclusion did not result in a model with acceptable model-data fit and it had relatively low abundance and richness compared to the other remaining PFTs.  Therefore, the SEMM includes only the five remaining PFTs.  The notation used in constructing the measurement model follows that of McDonald and Hartmann (1992) and Fox (2006).  To assist with identification, start values for selected parameters were fixed to “1” (Grace et al. 2010).  The 2+ emitted paths rule (Bollen and Davis 2009) was employed in the initial model.  R also provides an error message when a model run with sem is underidentified. Hypotheses It was expected that an acceptably fitted model would indicate the relative strength and direction of the relationships between environmental predictor variables and PFTs represented by vegetation indicators (richness and abundance).  Five PFTs that produced a significant model were: conifers (PFT 1), semi-shade tolerant gap specialists (PFT 2), deciduous trees and shrubs (PFT4), shade specialists (PFT6) and ericaceous shrubs (PFT9).  The indicators of the latent environmental predictor variables were: mean annual precipitation (MAP); mean annual temperature (MAT); soil nutrients - %N and %C; and fire severity - % LFH consumption and % woody debris (WD) consumption (see methods for determining soil nutrients and fire severity in Chapter 4).  My hypotheses are presented as a path diagram (Figure 5.2): (i) 61  precipitation and temperature influence fire severity and soil nutrients (%N and %C), (ii) fire severity and soil nutrients influence PFTs, and (iii) precipitation and temperature influence PFTs. Assessing model fit I present model results after the exclusion of paths that were either not significant (p>0.05) or did not improve model fit, and after the inclusion of modifications suggested by the sem routine that were logical in the model.  There are a number of fit indices available in R and several types of these were included because it is recommended that several indices be evaluated to determine an acceptable model (Hu and Bentler 1999; Schermelleh-Engel et al. 2003).  In the results, I included the χ2 statistic, the p-value of the χ2 statistic, and two measures of fit: one absolute (the root-mean-square error of approximation - RMSEA; Steiger and Lind 1980); and one relative (the non-normed fit index, also known as the Tucker-Lewis index - NNFI; Tucker and Lewis 1973; Bollen 1986). The p-value of the χ2 statistic indicates the significance of the difference between the sample covariance matrix and the estimated population covariance matrix (Tabachnick and Fidell 2007).  However, it is possible for the probability levels to be inaccurate either due to low sample size or when assumptions underlying the χ2 statistic are violated (Bentler 1995).  Therefore, to use χ2 as a descriptive goodness-of-fit index, Jöreskog and Sörbom (1993) suggest comparing the magnitude of χ2 with the number of degrees of freedom (the expected value of the sample distribution): E(χ2) = df.  The smallest possible ratio of χ2/df is desired; a ratio between 2 (“good” data-model fit) and 3 (“acceptable" data-model fit) is a reasonable target.  The RMSEA analyzes the discrepancies between the hypothesized model and the population covariance matrix.  For an acceptable model, the RMSEA should be <0.1, and RMSEA<0.01 indicates a very good model (MacCallum, Browne and Sugawara 1996).  The NNFI measures discrepancies between the χ2 of the hypothesized and the null models, and has a range that falls between 0 and 1; the cutoff for an acceptable model fit was set at >0.9 (Bentler and Bonett 1980).  62  It is also important to acknowledge that these tests only test model fitness and do not necessarily test the hypothesis. Data set The vegetation measurements were derived from 10 sites with measurements beyond 10 years post-burn: Brinks Mill, Francis Lake, Genevieve Lake, Goat River, Helene, Herron, Mackenzie, Otter Creek, Walcott and Walker Creek (Table 1.1).  The % cover (abundance) and richness for each species in the dataset (each of which belongs to a PFT) was measured at each of 219 plots in the 10 sites repeatedly, generally at 1, 2, 3, 5, 10 and 20 years post-burn.  Exceptions to the timing of vegetation surveys (see Table 1.1) are as follows: Brinks Mill, Francis Lake, Genevieve, Helene, Walcott and Walker Creek were surveyed at years 21 and/or 22 instead of 20; Herron was surveyed at years 4, 8, and 17 instead of years 5, 10 and 20; and Otter Creek was surveyed at years 11 and 18 instead of years 10 and 20.  In the initial dataset, MAP and MAT corresponded to the year and location of each vegetation survey, fire severity measurements were sampled directly after the prescribed burn (i.e. year 0) and soil N and C were measured in year 20 or the last vegetation survey at each site as noted above.  Abundance and richness of each PFT was the total % cover and richness of species for each plot year surveyed (n = 1314). Results All fit indices suggest an excellent model-data fit.  The χ2 statistic = 60.06 with 57 degrees of freedom and a χ2 to df ratio = 1.05.  The statistical chi-square test to assess how well the hypothesized model fit the data indicated acceptable fit (p = 0.365); the RMSEA index = 0.006 (with a 90% confidence interval of 0, 0.018); and the Tucker-Lewis (NNFI) = 0.999.  Normalized residuals for the model tested had a mean (-0.006) and median (0.019) both near zero, and minimum (-2.00) and maximum (1.61) within the acceptable range of ±2.58, indicating high correspondence between the hypothesized and the actual covariance matrices (Hair et al. 1998; Khamis and Hanoon 2010). 63  Non-significant (p>0.05) paths were removed from the model and parameters were added, as suggested by modification indices. The three paths added to the model were from conifers (PFT 1) to: (i) semi-shade tolerant gap specialists (PFT 2); (ii) deciduous trees and shrubs (PFT 4); and (iii) shade specialists (PFT 6).  The final model included seven mediation paths (indirect effect from one variable to another via a third variable) that were calculated following the method of Judd and Kenny (1981) using the standardized coefficients.  Full model results (Appendix H) include the structural model and measurement model paths and covariances, and the unstandardized parameter estimates (along with the standard error, z-value and p-values for each). The path diagram of the structural model (including all significant paths, standardized path coefficients, and R2 for the endogenous latent variables) is presented in Figure 5.3.  Standardized path coefficients showed direct positive effects of precipitation on temperature (0.44) and temperature on fire severity (0.70).  The direct effect of precipitation on fire severity (-0.63) was mediated by temperature (for a total effect of -0.32).  Temperature had a direct positive effect on soils (1.0) which was mediated through fire severity (for a total effect of 0.59).  Soils had a direct negative effect on shade specialists (PFT 6) (-0.42), fire severity had a significant direct positive effect on conifers (PFT 1) (0.68), and fire severity had significant direct negative effects on soils (-0.60), semi-shade tolerant gap specialists (PFT 2) (-0.73) and ericaceous shrubs (PFT 9) (-0.59).  Conifers (PFT 1) mediated the effect of fire severity on semi-shade tolerant gap specialists (PFT2) (for a total effect of -0.62).  Standardized path coefficients also showed that precipitation had positive direct effects on conifers (PFT 1) (0.11; when mediated by temperature had a total effect of -0.21) and shade specialists (PFT 6) (0.11).  Temperature had direct effects on four PFTs:  a strong negative effect on conifers (PFT 1) (-0.73); weak negative effects on deciduous trees and shrubs (PFT 4) (-0.21) and ericaceous shrubs (PFT 9) (-0.25); and a strong positive effect on semi-shade tolerant gap specialists (PFT 2) (0.84).  Fire severity mediated the positive effect of temperature on semi-shade tolerant gap specialists (PFT 2) (for a total 64  effect of 0.32), and the negative effects of temperature on conifers (PFT 1) (for a total effect of -0.25) and ericaceous shrubs (PFT 9) (for a total effect of -0.67).  Conifers (PFT 1) had direct positive effects on semi-shade tolerant gap specialists (PFT 2) (0.14), deciduous trees and shrubs (PFT 4) (0.29) and shade specialists (PFT 6) (0.28).  Conifers also mediated the effect of fire severity on semi-shade tolerant gap specialists (PFT 2) (for a total effect of -0.62), the effect of precipitation on shade specialists (PFT 6) (for a total effect of 0.15), and the effect of precipitation on deciduous trees and shrubs (PFT 4) (for a total effect of -0.42). Results for R2, or squared multiple correlation (SMC), of each endogenous latent variable indicates the percent variance explained in that variable.  The R2 results are as follows in descending order: soils (R2 = 0.94), semi-shade tolerant gap specialists (PFT 2; R2 = 0.61), ericaceous shrubs (PFT 9; R2 = 0.54), fire severity (R2 = 0.50), conifers (PFT 1; R2 = 0.47), deciduous trees and shrubs (PFT 4; R2 = 0.34), shade specialists (PFT 6; R2 = 0.25), and temperature (R2 = 0.19).  Precipitation was the only exogenous variable (R2 = 1.0).  While the parameter estimates for the paths between all factors in the final model were significant (p<0.0001; Appendix H), R2 values <0.25 for temperature, shade specialists (PFT 6) and deciduous trees and tall shrubs (PFT 4) suggest that the equations for these variables may be omitting relevant explanatory variables (Mitchell 1992; Iriondo et al. 2003). Discussion Influences of precipitation and temperature on fire severity and soil nutrients While previous empirical studies have determined that climate influences fire severity and soils, which in turn affect forest cover type (Bigler et al. 2005; Barrett et al. 2011), few measure the extent of climate influence on each of these controlling factors independently while also comparing their relative importance to ecosystem response.  My first hypothesis, that precipitation and temperature influence fire severity and soils, was supported by the model results.  It was reasonable to expect that as precipitation increases, fire severity (measured in this model as % LFH consumption and % woody debris 65  consumption) would decrease because moisture in the forest floor is expected to reduce the consumption of woody debris and the LFH layer during a fire event (Canadian Forestry Service 1984; Macadam 1989).  It was also reasonable to expect a strong positive relationship between temperature and soil %N and %C because higher temperatures are expected to result in increased decomposition rates, leading to greater nutrient availability.  Previous research has found that temperature alone may account for up to 53% of the variation in short-term rates of decomposition (Silver and Miya 2001; Brown et al. 2004).  However, it was unexpected that the relationship between temperature and soils (when mediated by fire severity; 0.59) would be equal but opposite of the relationship between fire severity and soil nutrients (-0.59), where soils represents %N and %C separately.  The positive indirect influence of temperature on soils, and negative direct influence of fire severity on soils, suggests that nitrogen availability increases with temperature (likely due to microbial activity) but decreases with fire (likely due to combustion).  Influences of fire severity and soil nutrients on plant functional types The SEM also supported my second hypothesis, that fire severity and soil %N and %C influence PFTs.  It was expected that as soil %N and %C increased, shade specialists would decline because PFT 6 includes many mixotrophs for which lower soil nutrient availability would not necessarily limit establishment and growth.  However, higher soil %N and %C increases competition with other pioneering species for available space during primary stages of establishment which may explain, in part, why shade specialists are not expected until later stages of succession.  It is also logical that conifers (PFT 1) had a direct positive relationship with shade specialists (PFT 6); with time the conifers develop a canopy, producing enough shade to eventually exclude shade-intolerant species (PFT 2) thus reducing competition in favour of plants adapted to a shade environment. The strong relationship between fire severity and conifers was almost equal but opposite to the relationship between fire severity and semi-shade tolerant gap specialists (PFT 2).  SEM results support 66  the findings reported in Chapter 4, where significantly lower abundance and richness of semi-shade tolerant gap specialists occurred at >30% LFH consumption (Table 4.1; Appendix E).  This indicates that these soil seed-banking species may be eliminated or reduced by high severity burns.  The negative relationship between fire severity and ericaceous shrubs (PFT 9) was weaker than the relationship between fire severity and semi-shade tolerant gap specialists.  Most ericaceous shrubs can resprout from basal portions of the stem or from rhizomes protected beneath the surface layers of soil following low severity burns; however, they may be killed in high severity burns. In contrast to the semi-shade tolerant gap specialists and ericaceous shrubs, conifers (PFT 1) had significantly higher abundance and richness at LFH consumption >30%, providing evidence of how well conifers have adapted to establish following higher severity burns.  The conifer species in the study, interior spruce, is adapted to occupy burned or exposed mineral soil, and to establish canopy dominance in open environments relatively quickly after disturbance.  Other naturally occurring species in the study area, such as western redcedar, western hemlock and subalpine fir, are less competitive for open mineral soil and are more slowly growing than interior spruce, but these species are prolific seeders, enabling regeneration after disturbance and hence greater contribution to ecological resilience relative to other plant species in the communities. Influences of precipitation and temperature on plant functional types My results supported my third hypothesis, that both climate latent variables would have significant relationships with PFT resilience.  The positive relationship between precipitation and conifers (PFT 1) and precipitation and shade specialists (PFT 6) was equal (0.11) while conifers also served to mediate the relationship between precipitation and shade specialists (although the results show only a marginal increase in the strength of the relationship).  Overall these results may be attributed to a general increase in productivity in areas of higher precipitation that would produce greater overstory cover, casting more shade, and benefitting more shade tolerant species.  This highlights the ability of SEM to 67  quantify (at least in relative terms) the importance of the interspecific and species-environment relationships within this complex ecological system. With fire severity as a mediating variable, the total effect of temperature on conifers (PFT 1) was only a third the strength of the direct effect.  This suggests that higher temperatures may be limiting to conifers, but this limitation is offset by more severe burns (which this model indicates may occur in environments with higher temperatures).  Also, by including fire severity as a mediating variable, the total effect of temperature on ericaceous shrubs (PFT 9) was more than 2.5x the strength of the direct effect, illustrating an additive negative influence of temperature and fire severity on ericaceous shrubs.   These results show that fire severity can play a pivotal role in regards to the influence of temperature on certain PFTs; with more severe burns, the strength of the negative effect of temperature on a PFT can either decrease (as for conifers) or increase (as for ericaceous shrubs). A similar situation occurs for semi-shade tolerant gap specialists (PFT 2) as for conifers (PFT 1) and ericaceous shrubs (PFT 9), although opposite in sign.  The strong positive relationship between temperature and semi-shade tolerant gap specialists (0.84) indicates that PFT 2 is more successful in warmer climates.  However, this may depend on the stability agent in question; species richness pre-burn and post-burn was nearly 2x higher in the SBS compared to the ESSF or ICH sites while no such trend was observed for abundance (Table 3.5b).  This result may be due to higher germination rates of the seed-banking species in warmer climates.  It may also be related to additional processes other than those directly affecting rates of plant recovery after fire, such as higher rates of fruit production and more birds distributing the fruits that would produce a greater number of seed-banking species at lower elevations and higher temperatures.  However, when fire severity was included as a mediating variable, the direct effect of temperature on this PFT (0.84) was greatly reduced to a total effect of 0.32.  This means that the benefit of higher temperatures for semi-shade tolerant gap specialists has a threshold; these species are sensitive to higher severity burns (>30% LFH consumption; see Table 4.1 and Appendix 68  E) which, according to this model, occur in environments with higher temperature.  These results apply to mesic sites, but temperature may conversely have adverse influences on many species on drier sites, particularly when combined with high fire severity (Perry et al. 2011). Initially it seems counter-intuitive that conifers (PFT 1) would have a direct positive effect on deciduous trees and shrubs (PFT 4); however, when conifers were included as a mediating variable, the total negative effect of temperature on deciduous trees increased in strength 2x (producing a total effect of -0.42).  While it is expected that the higher elevation (colder ESSF) sites would have less deciduous trees and shrubs than the warmer SBS sites (see Tables 3.5a,b), the negative influence of temperature on deciduous trees and shrubs is only a quarter the strength of the positive influence of temperature on semi-shade tolerant gap specialists (PFT 2).  Although this model found no significant relationship between semi-shade tolerant gap specialists (PFT 2) and deciduous trees and shrubs (PFT 4), it may be possible that the strong positive influence of temperature on semi-shade tolerant gap specialists increases the competitive potential of PFT 4 in warmer climates.  This may also in part explain the dominance of this group of plants over every other group, particularly during earlier stages of secondary succession where sites have large areas of open space after clearcutting and slashburning and prior to the planting of conifer seedlings. Conclusion The SEM indicated that temperature had the greatest influence on resilience in these forest systems, with very strong (>±0.7) positive relationships with fire severity, soil %N and %C, conifers (PFT 1) and gap specialists (PFT 2).  Next, fire severity, which is itself a product of temperature, had strong (>±0.5) positive relationships with soil %N and %C, conifers (PFT 1), gap specialists (PFT 2) and ericaceous shrubs (PFT 9).  Usually temperature and fire severity influences were opposite in sign, with burn influences of slightly less strength, suggesting a balance between how these two variables influence soils, conifers and gap specialists.  Temperature and fire severity also had a negative 69  relationship with ericaceous shrubs (PFT 9), indicating an additive effect.  Precipitation had a strong negative relationship with fire severity, but very weak (<±0.25) relationships with conifers (PFT 1), deciduous trees (PFT 4) and shade specialists (PFT 6).  Conifers had weak (<±0.30) relationships with gap specialists, deciduous trees and shade specialists while soils had a moderate negative relationships with shade specialists.  Therefore, determining which of these two variables, conifers or soils, has more importance depends on which PFT is considered more important to resilience: in other words, it depends on how many components of the system are influenced or how much influence is imposed on a single component.  In summary, the overall relative importance of variables influencing resilience was: temperature > fire severity > precipitation > conifers ≈ soils.70                 Figure 5.1. The initial structural equation meta-model (SEMM) represents major categories of influence on relative dominance of plant functional types (PFTs) and resulting ecosystem types (represented by dashed rectangles) - conifer forest, deciduous forest, and heathland; it illustrates indirect and direct effects of climate on soil N and soil C, fire severity and PFTs. Specifically, the SEMM delineates (i) influence of precipitation and temperature on conifer forest, deciduous forest and heathland ecosystem types, (ii) influence of precipitation and temperature on fire severity and soil %N and %C , (iii) influence of fire severity on soil %N and %C , and (iv) influence of fire severity and soil %N and %C  on conifer forest, deciduous forest and heathland ecosystem types. 71      Figure 5.2. Hypothesis for the structural equation model presented as a path diagram. A “1” indicates parameters that were fixed to “1”. PFT latent variable indictors were represented by the PFT number followed by a C (% cover) or an R (richness). Gamma and beta regression coefficients were represented by single-headed arrows from one latent variable to another; regression coefficients of indicators predicted by latent variables (lam) were represented by single-headed arrows; the covariance of errors for each latent dependent variable (psi) were represented by double-headed arrows; and the covariances among errors associated with indicators predicted from latent variables (the) were represented by double-headed arrows.  72                 Figure 5.3. The results of the SEM presented as the structural model includes the standardized path coefficients indicating the direction and strength of the relationships between latent variables, and the R2 value indicating the amount of variation in each endogenous latent variable that has been accounted for. Beneath the standardized path coefficients (indicating direct effects) the total effects and mediating variables are noted in italics. Model χ2 was 60.06 with 57 degrees of freedom and p = 0.365. The root mean square error of approximation (RMSEA) = 0.006 and the Tucker-Lewis Index (NNFI) = 0.999.   73  Chapter 6. General discussion and conclusions In Chapter 2, I constructed PFT groupings based on plant physical and mechanistic characteristics that simplified a dataset of 183 species into nine PFTs representing vascular plant traits central to ecosystem function: one PFT of plants with seed cones (conifers) and eight without.  The eight PFTs without seed cones included four PFTs generally associated with mor humus form (shade specialists, ericaceous shrubs, montane boreal and cool temperate species, and ground surface material generalists), and four that are generally not associated with mor humus form (deciduous trees and tall shrubs, surface water specialists and gap specialists that are either semi-shade tolerant or shade intolerant). Chapter 3 applied these nine PFTs to assess ecosystem stability in response to slashburning after clearcutting in ESSF, SBS and ICH forest of British Columbia.  I found that results varied depending on whether the observation was made at the species, ecosystem, and/or site level.  Resistance measured at the species level was the change in percent cover of each species; and resistance at the ecosystem and site level was measured as the total change of species (+/-) in the plant community.  Species level resistance identified indicator species of low resistance: subalpine fir (Abies lasiocarpa), heart-leaved twayblade (Listera cordata), enchanter's-nightshade (Circaea alpina), devil’s club (Oplopanax horridus), white-flowered rhododendron (Rhododendron albiflorum), and spiny wood fern (Dryopteris expansa) and species favoured by the treatment (grasses (Calamagrostis spp.), willows (Salix spp.) and fireweed (Epilobium angustifolium)).  Results suggested that resistance of the study sites measured by the total number of species change post-burn declined as follows:  Otter Creek (ESSF - cold and wet) > Herron (ESSF - cold and dry) > Goat River (ICH(ESSF)) > Walker Creek (ICH(SBS)) > Mackenzie (SBS).  Resistance measured by ecosystem type measured by the total number of species change post-burn declined as follows: (SBS > ESSF > ICH). 74  Measures of abundance and richness that represent stability agents (matter and information; see Figure 1.1) also provided the data for testing hypotheses of resilience, which identified variations in the expected successional pathway at particular study sites.  In general, regardless of the level of description, successional pathways were as expected, including increased abundance and richness of conifers, gap specialists and deciduous trees and tall shrubs.  However, hypothesis tests at the site level indicated three exceptions: 1. Two decades after slashburning, Otter Creek (the highest elevation site where temperatures were lowest and precipitation was highest) had very little change in the relatively abundant ericaceous shrubs and very little conifer development but significant increases in gap specialists.  These changes indicate a possible threshold shift in stability domain from a conifer ecosystem to an open meadow or possibly heathland; however because the ESSF has an extremely slow recovery process, it may be too early still to draw this conclusion. 2. At Goat River, ericaceous shrubs were nearly eliminated, with significantly lower abundance 20 years post-burn compared to pre-burn.  The planted interior spruce (Picea glauca x engelmanii) dominated the community with an understory abundant in shade specialists as well as significant increases in abundance and richness of gap specialists.  Therefore, while this site shows progression towards conifer forest, whether or not this shift in PFTs is an expression of normal successional processes could only be verified by having control forests for comparison. 3. At Mackenzie River and Walker Creek, planted interior spruce was establishing well.  However, both sites had significant increases in abundance and richness of gap specialists and deciduous trees, with deciduous trees surpassing the conifers at Mackenzie River.  At Walker Creek, the ericaceous shrubs increased significantly, and when combined with the deciduous trees, also surpassed the conifers.  Usually in SBS and ICH forests, it isn’t until after 20 years post-disturbance that the deciduous species would be expected to decline, with conifer dominance 75  not occurring until approximately 50 years post-disturbance (Simard et al. 2004).  Again this demonstrates the usefulness of control forests and the need for long-term research sites. The results of Chapter 4 indicate a strong correlation between abundance and richness of PFTs.  Where this is not true, an inquiry as to the nature of divergence may show us an event or process that occurs with secondary succession.  Therefore, I found the PFT classification useful in the analysis of stability agents (information and matter) by linking PFTs (and the individual species they represent) to ecological processes.  Trends in PFT response to fire severity, including thresholds and limitations following prescribed burning, as well as significant relationships between soil nutrients and the abundance and richness of functional types, may change with time.  This would indicate dynamic vegetation communities, which we have seen assemble, change, and interact with the environment throughout the first two decades after disturbance. The SEM model results of Chapter 5 lead to these general conclusions: 1. precipitation had a negative relationship with fire severity; and temperature had a positive relationship with fire severity and soil %N and %C; 2. each PFT was influenced by either temperature or precipitation except for conifers (PFT 1) which was influenced by both; 3. soil %N and %C had a negative relationship with shade specialists while fire severity had a positive relationship with conifers and a negative relationship with  gap specialists and ericaceous shrubs; and 4. conifers (PFT 1) was the only PFT to mediate relationships detected in the model: between fire severity and gap specialists; between precipitation and shade specialists; and between temperature and deciduous trees and shrubs.  This result shows that planting of conifers clearly had an important influence on relationships between environment and plant functional groups.  They also suggest that any variation in the reforestation regime, such as a shift from planting to 76  natural regeneration, or choice of regenerating species, ought to have a large influence on the composition and resilience of the plant community. Limitations of this study The two decade timeframe of this study is relatively short-term considering the history of disturbance and length of time we had to observe recovery after disturbance.  It has been found that the greatest structural change occurs within the first 100 years after disturbance (Trofymow et al. 2003) but that organizational change can continue even after 100 years post-burn in conifer forests of British Columbia (Brulisauer et al. 1996).  This demonstrates the need for establishment and maintenance of permanent plots and the adoption of consistent methodologies to monitor ecosystem stability over long time periods (Bakker et al. 1996). Another limitation of this study is the lack of controls.  It would have been better to have equivalent unburned control plots for comparison and to better ascertain the role of slashburning alone.  In addition, it would have been helpful to have benchmark mature ecosystems, or preferably benchmark chronosequences of the ecosystems, so that we would have some baseline for comparison or evaluation of resilience. Measuring soil nitrogen and forest floor depth over time would have made it possible to determine how edaphic conditions either influenced or were influenced by vegetation, thus providing a better understanding of the interactions within the vegetation community and parsing out the direct influences of soil nitrogen and fire severity.  Moreover, due to the lack of evidence of the source of nitrogen utilized by vascular plants, this study cannot address the influence of facilitation and mycorrhizae among the members of the existing plant community. Another shortcoming to this study is that it was not originally designed to represent fire severity or nitrogen gradients; wider gradients would better test vegetation response, which in turn may create scenarios of prescribed burn with fire effects more similar to wildfires in North America.  This is 77  particularly important as climatic conditions shift and researchers and forest managers try to make accurate predictions about how these changes will affect the stability of our future forests. As for the SEM approach presented herein, it is essential to be aware, while guiding oneself through the evaluation of the model that this is a study of the relative importance of different processes controlling resilience, and it is only one possible method (that includes very few variables) for investigating these dynamic and complex systems.  Limitations to the model itself are derived from limitations of the dataset (as mentioned above), such as narrow ranges of climates, burn severities, soil nutrients and particularly time, considering the long temporal scale meaningful to the understanding of mature conifer forests.  Furthermore, the model has not yet been verified with an additional dataset, which is necessary for determining the validity of model results and its usefulness in predicting future outcomes in our changing climate (Table 6.1). Another significant limitation of the SEM analysis was that time was not included as a variable.  Alternative methods for addressing this may include either creating the covariance matrix using the average value of each PFT on each plot over the entire measurement period rather than having each time period as a separate data point, or the development of a multi-level model that treats each time interval as an independent level thereby eliminating the problem of temporal autocorrelation. Finally, this study did not address the question of how stability agents relate to one another, but quantifies individual responses of matter and information.  However, this concept could be applied to more holistic measures of the ecosystem that include the much wider array of energy fluxes, structures and species that should be considered in measures of ecological stability.  For example, this approach may also be integrated into studies examining effects of disturbance agents such as the mountain pine beetle (Dendroctonus ponderosae) (Canadian Forest Service 2007) or forest management activities that affect habitat quality on whole biotic communities, such as cavity-nesting birds (Martin et al. 2004), which are key indicators of forest stability.  A more holistic approach that could build on the work of 78  Martin et al. (2004) would examine mountain pine beetle effects on a broader range of ecosystem goods and services than only the cavity nesting bird community, such as animal, plant, fungal and microbial communities (information), concurrently with energy, water, carbon and nutrient fluxes (matter).  My dissertation shows that this can be done successfully by reducing information into functional groups and examining their effects and responses to matter using gradient analysis and structural equation modeling. Suggestions for future research As a culmination of the efforts of many individuals from the provincial government and the University of British Columbia, including substantial monetary investments, these permanent plots have produced one of the very few long-term master datasets from which numerous reports, articles, and theses have been produced and contributed to the understanding of the influence of clearcutting and slashburning in central British Columbia.  By continuing to measure and study these sites, we may be able to answer questions about how manipulated forests might respond during periods of climatic shift and how to best practice adaptive forest management. This study provides a solid benchmark for assessing forest stability and response to disturbance, and demonstrates the need to prioritize long-term research to further assess response and ensure future resilience and complexity of British Columbia’s managed forests.  Lastly, it emphasizes need for care when operating in climatically limiting forest environments, calling for adaptive management strategies that are accountable for the complexity within and range of services provided by ESSF, SBS and ICH ecosystems in central British Columbia.  79  Table 6.1. Summary of SEM results indicating the strength and sign of climate, fire severity and soil nutrient on plant functional types (PFTs); and predictions of how change in temperature and precipitation will influence these relationships.     MAP2 MAT2Burn severity3Soil nutrients4How will the abundance and/or richness of PFTs be influenced by climate change?PFT 1 Conifers 0.11 -0.74 0.69 increased temperatures directly reduce the abundance and richness of conifers; however, if increased temperatures also result in increased burn severity, conifers will benefitPFT 2 Gap specialists (semi-shade tolerant)0.84 -0.73 increased temperatures directly increase the abundance and richness of gap specialists; however, if increased temperatures also result in increased burn severity, the benefit of higher temperatures will be reducedPFT 4 Deciduous trees and tall shrubs-0.21 increased temperatures may possibly result in negative effects on deciduous trees and tall shrubsPFT 6 Shade specialists 0.12 -0.42 increased precipitation may benefit shade specialists; however, an increase in temperature in turn causing an increase in soil nutrients may have negative effects on shade specialistsPFT 9 Ericaceous shrubs -0.25 -0.59 increased temperatures and increased burn severity will both have direct negative effects on ericaceous shrubs1See Appendix D for a description and list of species for each plant functional type (PFT).Plant functional types12Between 2020 and 2050, the average MAT and MAP for the study sites included in SEM (Brinks Mill, Francis Lake, Genevieve Lake, Goat River, Mackenzie, Otter Creek, Walker Creek, Herron, Walcott, Helene) is expected to increase 0.6˚C and 18 mm respectively (Wang et al. 2006).3Burn severity was a factor used for SEM that included measured indicator varibles (% woody debris consumption and % LFH consumption). SEM results indicated a strong direct effect of temperature on burn severity (0.70) and a stong direct effect of precipitation on burn severity (-0.63) that was mediated by temperature for a reduced total effect (-0.32).4Soil nutrients was a factor in the SEM that included measured indicator variables (total soil N and total soil C). SEM results indicated a strong direct effect of burn severity on soil nutrients (-0.59); and a strong direct effect of temperature on soil nutrients (1.0) that was mediated by burn severity for a reduced total effect (0.59).80  Bibliography Aitken, S.N., Yeaman, S., Holliday,J.A., Wang, T., and Curtis-McLane, S. 2008. Adaptation, migration or extirpation: Climate change outcomes for tree populations. Evolutionary Applications, 1: 95-111. Alavifar, A., Karimimalayer, M., and Anuar, M.K. 2012. Structural equation modeling VS multiple regression. The first and second generation of multivariate techniques. Engineering Science and Technology: An International Journal, 2: 326-329. Allen, R.B., Peet, R.K., and Baker, W.L. 1991. Gradient analysis of latitudinal variation in Southern Rocky Mountains.  Journal of Biogeography, 18: 123-139. Aubert, M., Bureau, F., and Vinceslas-Akpa, M. 2005. Sources of spatial and temporal variability of inorganic nitrogen in pure and mixed deciduous temperate forests. Soil Biology and Biochemistry, 37: 67-79. Aubin, I., Gachet, S., Messier, C., and Bouchard, A. 2007. How resilient are northern hardwood forests to human disturbance? Ecoscience, 14: 259-271. B.C. Ministry of Forests, Mines and Lands. 2010. The State of British Columbia’s Forests, 3rd ed. Forest Practices and Investment Branch, Victoria, B.C. www.for.gov.bc.ca/hfp/sof/index.htm#2010_report Bakker, J.P., Olff, H., Willems, J.H., and Zobel, M. 1996. Why do we need permanent plots in the study of long-term vegetation dynamics? Journal of Vegetation Science, 17: 147-156. Banner, A., MacKenzie, W.H., Haeussler, S., Thomson, S., Pojar, J., and Trowbridge, R.L. 1993. A Field Guide to Site Identification and Interpretation for the Prince Rupert Forest Region. Res. Br., Victoria, B.C. Land Manage. Handb. No. 26. http://www.for.gov.bc.ca/hfd/pubs/docs/Lmh/Lmh26.htm Barrett, K., McGuire, A.D., Hoy, E.E., and Kasischke, E.S. 2011. Potential shifts in dominant forest cover in interior Alaska driven by variations in fire severity. Ecological Applications, 21: 2380-2396. Batalha, M.A. and Martins, F.R. 2004. Floristic, Frequency, and Vegetation Life-Form Spectra of a Cerrado Site. Braz. J. Biol., 64: 203-209. Beaudry, L.J., Coupe, R.A., DeLong, C., and Pojar, J. 1999. Plant Indicator Guide for Northern British Columbia: Boreal, Sub-Boreal, and Subalpine Biogeoclimatic Zones: BWBS, SBS, SBPS, and northern ESSF. Land Management Handbook 46. Research Branch. B.C. Ministry of Forests. Beaudry, L.J., Coupe, R.A., DeLong, C., and Pojar, J. 2003. Plant Indicator Guide for Northern British Columbia: the Northern Portion of the MS and ICH Biogeoclimatic Zones. Technical Report 10. Research Branch. B.C. Ministry of Forests. Bentler, P.M. and Bonett, D.G. 1980. Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88: 588-606. Bentler, P.M. and Dijkstra, T. 1985. Efficient estimation via linearization in structural models. In P. R. Krishnaiah (Ed.), Multivariate analysis VI (p. 9-42). Amsterdam: North-Holland. Bigler, C., Kulakowski, D., and Veblen, T.T. 2005. Multiple disturbance interactions and drought influence fire severity in Rocky Mountain subalpine forests. Ecology, 86: 3018-3029. Boer, M. and Smith M.S. 2003. A plant functional approach to the prediction of changes in Australian rangeland vegetation under grazing and fire.  Journal of Vegetation Science, 14: 333-344. 81  Bollen, K.A. 1986. Sample Size and Bentler and Bonett’s Nonnormed Fit Index. Psychometrika, 51: 375-377. Bollen, K.A. 1989. Structural Equations with Latent Variables. New York, NY. Wiley & Sons. Bollen, K.A. and Davis, W.R. 2009. Two rules of identification for structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 16: 523-536. Boomsma, A. and Hoogland, J.J. 2001. The robustness of LISREL modeling revisited. In Cudeck, R., du Toit, S., and Sörbom D. (Eds.), Structural equation models: Present and future. A Festschrift in honor of Karl Jöreskog, 139-168. Chicago, Scientific Software International. Bouchon, É. and Arseneault, D. 2004. Fire disturbance during climate change: failure of postfire forest recovery on a boreal floodplain. Canadian Journal of Forest Research, 34: 2294-2305. Bradley, R.L. and Fyles, J.W. 1996. Growth of paper birch (Betula papyrifera) seedlings increases soil available C and microbial acquisition of soil-nutrients. Soil Biology & Biochemistry, 27: 1565-1571. Bradstock, R.A. and Kenny, B.J. 2003. An application of plant functional types to fire management in a conservation reserve in southeastern Australia. Journal of Vegetation Science, 14: 345-354. Briske, D.D., Fuhlendorf, S.D., and Smeins, F.E. 2006. A unified framework for assessment and application of ecological thresholds. Rangeland and Ecological Management, 59: 225-236. Brown, J.H., Gillooly, J.F., Allen, A.P., Savage, V.M., and West, G.B. 2004. Toward a metabolic theory of ecology. Ecology, 85: 1771-1789. Brown, C.D. and Johnstone. J.F. 2012. Once burned, twice shy: Repeat fires reduce seed availability and alter substrate constraints on Picea mariana regeneration. Forest Ecology and Management, 266: 34-41. Browne, M.W. 1984. Asymptotically distribution‐free methods for the analysis of covariance structures. British Journal of Mathematical and Statistical Psychology, 37: 62-83. Brulisauer, A.R., Bradfield, G.E., and Maze, J. 1996. Quantifying organizational change after fire in lodgepole pine forest understorey. Can. J. Bot., 74: 1773-1782. Burton, C.M. and Burton, P.J. 2003. A Manual for Growing and Using Seed from Herbaceous Plants Native to the Northern Interior of British Columbia. Symbios Research & Restoration, Smithers, British Columbia. 172 p. Burton, P.J., Messier, C., Smith, D.W., and Adamowicz, W.L. 2003. Towards sustainable management of the boreal forest. Eds. Ottawa, Ontario, Canada, NRC Research Press: 307-368. Burton, P., Taylor, S., and Thandi, G. 2005. Challenges in defining the disturbance regimes of northern British Columbia. Journal of Ecosystems and Management, 6: 119-123. Canadian Forestry Service. 1984. Tables for the Canadian Forest Fire weather Index System. 4th. Ed. Environ. Can., Can Forestry Service. Ottawa, Ontario. For. Tech. Rep. 25. 48 p. Carter, M.R. 1993. Soil Sampling and Methods of Analysis. CRC Press. Cattelino, P.J., Noble, I.R., Slatyer, R.O., and Kessel, S.R. 1979. Predicting the Multiple Pathways of Plant Succession. Environmental Management, 3: 41-50. Chapin III, F.S. 2003. Effects of plant traits on ecosystem and regional processes: a conceptual framework for predicting the consequences of global change. Annals of Botany, 91: 455-463. 82  Chapin III, F.S., Bret-Harte, M.S., Hobbie, S.E., and Zhong, H. 1996a. Plant Functional Types as Predictors of Transient Responses of Arctic Vegetation to Global Change. Journal of Vegetation Science, 7: 347-358. Chapin III, F.S., Matson, P.A., and Mooney, H.A. 2002. Principles of Terrestrial Ecosystem Ecology. New York, NY, Springer. 436 p. Chapin III, F.S., Torn, M.S., and Tateno, M. 1996b. Principles of Ecosystem Sustainability. The American Naturalist, 148: 1016-1037. Chen, H.Y.H., Légaré, S., and Bergeron, Y. 2004. Variation of the understory composition and diversity along a gradient of productivity in Populus tremuloides stands of northern British Columbia, Canada. Canadian Journal of Botany, 82: 1314-1323. Clark, D.F., Antos, J.A., and Bradfield, G.E. 2003. Succession in sub‐boreal forests of West‐Central British Columbia. Journal of Vegetation Science, 14: 721-732. Clark, J.S., Lewis, M., McLachlan, J.S., and HilleRisLambers, J. 2003. Estimating population spread: what can we forecast and how well? Ecology, 84: 1979-1988. Clements, F.E. 1916. Plant succession: analysis of the development of vegetation. Publ. Carnegie Inst., Washington. 242: 1-512. Coates, K.D. 2002. Tree recruitment in gaps of various size, clearcuts and undisturbed mixed forest of interior British Columbia, Canada. Forest Ecology and Management, 155: 387-398. Coates, K.D., Canham, C.D., Beaudet, M., Sachs, D.L., and Messier, C. 2003. Use of a spatially explicit individual-tree model (SORTIE BC) to explore the implications of patchiness in structurally complex forests. Forest Ecology and Management, 186: 297-310. Collins, B.M., Kelly, M., van Wagtendonk, J.W., and Stephens, S.L. 2007. Spatial patterns of large natural fires in Sierra Nevada wilderness areas. Landscape Ecology, 22: 545-557. Collins, D.B., Feller, M.C., Klinka, K., and Montigny, L. 2001. Forest Floor Nutrient Properties in Single-and Mixed-Species, Second Growth Stands of Western Hemlock and Western Redcedar. Northwest Science 75: 407-416. Coulombe, G., Huot, J., Arseneault, J., Bauce, E., Bernard, J.-T., Bouchard, A., Liboiron, M.A., and Szaraz, G. 2004. Commission d’étude sur la gestion de la forêt publique québécoise. Bibliothèque nationale du Québec. Currah, R.S. and Zelmer, C. 1992. A key and notes for the genera of fungi mycorrhizal with orchids and a new species in the genus Epulorhiza. Rept. Tottori Mycol. Inst. 30: 43-59. Dale, V.H., Joyce L.A., McNulty, S., Neilson, R.P., Ayres, M.P., Flannigan, M.D., Hanson, P.J., Irland, L.C., Lugo, A.E., Peterson, C.J., Simberloff, D., Swanson, F.J., Stocks, B.J., and Wotton, B.M. 2001. Climate change and forest disturbances. Bioscience, 51: 723-734. Darwin, C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. London, John Murray.  502 p. DeLong, C. 2003. A Field Guide to Site Identification and Interpretation for the Southeast Portion of the Prince George Forest Region. Res. Br., Victoria, B.C. Land Manage. Handb. No. 51. http://www.for.gov.bc.ca/hfd/pubs/docs/Lmh/Lmh51.htm. Last accessed September 2013. 83  DeLong, C., Tanner, D. and Jull, M.J. 1993. A Field Guide for Site Identification and Interpretation for the Southwest Portion of the Prince George Forest Region. Res. Br., Victoria, B.C. Land Manage. Handb. No. 24. http://www.for.gov.bc.ca/hfd/pubs/docs/Lmh/Lmh24.htm. Last accessed September 2013. DeLong, C. 1998. Natural disturbance rate and patch size distribution of forests in northern British Columbia: Implications for forest management. Northwest Sci. 72 (Special Issue):35-48. Diaz, S., Cabido, M., and Casanoves, F. 1998. Plant functional traits and environmental filters at a regional scale. Journal of Vegetation Science, 9: 113-122. Douglas, G.W., Straley, G.B., Meidinger, D.V., and Pojar, J. (editors). 1998a. Illustrated Flora of British Columbia. Volume 1: Gymnosperms and Dicotyledons (Aceraceae Through Asteraceae). B.C. Ministry of Environment, Lands & Parks and B.C. Ministry of Forests. Victoria. 436 p. Douglas, G.W., Straley, G.B., Meidinger, D.V., and Pojar, J. (editors). 1998b. Illustrated Flora of British Columbia. Volume 2: Dicotyledons (Balsaminaceae Through Cucurbitaceae). B.C. Ministry of Environment, Lands & Parks and B.C. Ministry of Forests. Victoria. 401 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 1999a. Illustrated Flora of British Columbia. Volume 3: Dicotyledons (Diapensiaceae Through Onagraceae). B.C. Ministry of Environment, Lands & Parks and B.C. Ministry of Forests. Victoria. 423 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 1999b. Illustrated Flora of British Columbia. Volume 4: Dicotyledons (Orobanchaceae Through Rubiaceae). B.C. Ministry of Environment, Lands & Parks and B.C. Ministry of Forests. Victoria. 427 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 2000. Illustrated Flora of British Columbia. Volume 5: Dicotyledons (Salicaceae Through Zygophyllaceae) And Pteridophytes. B.C. Ministry of Environment, Lands & Parks and B.C. Ministry of Forests. Victoria. 389 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 2001a. Illustrated Flora of British Columbia, Volume 6: Monocotyledons (Acoraceae Through Najadaceae). B.C. Ministry. Environment, Lands and Parks and B.C. Ministry of Forests. Victoria. 361 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 2001b. Illustrated Flora of British Columbia, Volume 7: Monocotyledons (Orchidaceae Through Zosteraceae). B.C. Ministry of Sustainable Resource Management and B.C. Ministry of Forests. Victoria. 379 p. Douglas, G.W., Meidinger, D.V., and Pojar, J. (editors). 2002. Illustrated Flora of British Columbia, Volume 8: General Summary, Maps and Keys. B.C. Ministry of Sustainable Resource Management and B.C. Ministry of Forests. Victoria. 457 p. Downie, D.G. 1943. Source of the Symbiont of Goodyera repens. Trans. Bot. Soc. Edin., 33: 383-390. Driscoll, K.G., Arocena, J.M., and Massicotte, H.B. 1999. Post-fire soil nitrogen content and vegetation composition in Sub-Boreal spruce forests of British Columbia's central interior, Canada. Forest Ecology and Management 121: 227-237. Durall, D.M., Gamiet, S., Simard, S.W., Kudrna, L., and Sakakibara, S.M. 2006. Effects of clearcut logging and tree species composition on the diversity and community composition of epigeous fruit bodies formed by ectomycorrhizal fungi. Canadian Journal of Botany, 84: 966-980. Feller, M.C. 1996. The influence of fire severity, not fire intensity, on understory vegetation biomass in British Columbia. In Proceedings of the 13th Fire and Forest Meteorology Conference. Lorne, Australia. 335-348. 84  Fenton, N., Légaré, S., Bergeron, Y., and Paré, D. 2006. Soil oxygen within boreal forests across an age gradient. Can. J. Soil Sci., 86: 1-9. Flannigan, M.D., Bergeron, Y., Engelmark, O., and Wotton, B.M. 1998. Future wildfire in circumboreal forests in relation to global warming. Jounal of Vegetation Science, 9: 469-476. Fox, J. 1980. Effect analysis in structural equation models. Sociological Methods and Research, 9: 3-28. Fox, J. 2006. Structural-Equation Modeling with the sem Package in R. Structural Equation Modeling, 13: 141-162. Fox, J., Nie, Z., and Byrnes, J. 2013. sem: Structural Equation Models. R package version 3.1-1. http://CRAN.R-project.org/package=sem. Franklin, J. 1995. Predictive vegetation mapping: geographic modelling of biospatial patterns in relation to environmental gradients. Progress in Physical Geography, 19: 474-499. Franklin, J., Spies, T.A., Van Pelt, R., Carey, A.B., Thornburgh, D.A., Berg, D.R., Keeton, W.S., Shaw, D.C., Bible, K., and Chen, J. 2002. Disturbances and structural development of natural forest ecosystems with silvicultural implications, using Douglas-fir forests as an example. Forest Ecology and Management, 155: 399-423. Gebauer, G. and Meyer M. 2003. 15N and 13C natural abundance of autotrophic and mycoheterotrophic orchids provides insight into nitrogen and carbon gain from fungal association. New Phytologist, 160: 209-223. Gillett, N.P., Weaver, A.J., Zwiers, F.W., and Flannigan, M.D. 2004. Detecting the effect of climate change on Canadian forest fires. Geophysical Research Letters, 31: 1-4. Gleason, H.A. 1926. The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club, 53: 7-26. Gonzalez, I., Dejean, S., Martin, P., and Baccini, A. 2008. CCA: An R Package to Extend Canonical Correlation Analysis.  Journal of Statistical Software, 23: 1-14. Grace, J.B. 2006. Structural Equation Modeling and Natural Systems. Cambridge University Press. Grace, J.B. and Pugesek, B.H. 1997. A Structural Equation Model of Plant Species Richness and Its Application to a Coastal Wetland. The American Naturalist, 149: 436-460. Grace, J.B. and Keeley, J.E. 2006. A structural equation model analysis of postfire plant diversity in California shrublands. Ecological Applications, 16: 503-514. Grace, J.B., Anderson, T.M., Olff, H., and Scheiner, S.M. 2010. On the specification of structural equation models for ecological systems. Ecological Monographs, 80: 67-87. Grimm, V. and Wissel, C. 1997. Babel, or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia, 109: 323-334. Guarino, A.J. 2004. A Comparison of First and Second Generation Multivariate Analyses: Canonical Correlation Analysis and Structural Equation Modeling. Florida Journal of Educational Research, 42: 22-40. Gunderson, L.H. 2000. Ecological resilience - In theory and application. Annu. Rev. Ecol. Syst., 31: 425-439. Haeussler, S. 1991. Prescribed fire for forest vegetation management. BC Min. For. and Forestry Canada, Victoria, BC, FRDA Memo 198. 19 p. 85  Haeussler, S. and Kneeshaw, D. 2003. Chapter 9: Comparing forest management to natural process. In Towards sustainable management of the boreal forest.  P. Burton, C. Messier, D. W. Smith, and W. L. Adamowicz, Editors. National Research Council of Canada, Ottawa, Ontario, Canada. 307-368. Hair, J.F., Anderson, R.E., Tatham, R.L., and Black, W.C. 1998. Multivariate Data Analysis. Englewood: Prentice Hall International. p 169-215. Hallett, D.J., Lepofsky, D.S., Mathewes, R.W., and Lertzman, K.P. 2003. 11 000 years of fire history and climate in the mountain hemlock rain forests of southwestern British Columbia based on sedimentary charcoal. Can. J. For. Res. 33: 292-312. Hamann, A. and Wang, T. 2005. Models of climatic normals for genecology and climate change studies in British Columbia. Agric. For. Meteorology, 128: 211–221. Hamilton, E. 2006a. Vegetation Development and Fire Effects at the Walker Creek Site Comparison of Forest Floor and Mineral Soil Plots. Technical Report 26. F.S. Program. Victoria, BC, Ministry of Forests and Range. Hamilton, E. 2006b. Fire Effects and Post-burn Vegetation Development in the Sub-Boreal Spruce Zone: Mackenzie Windy Point Site. Technical Report 33. Victoria, BC, Ministry of Forests and Range. Hamilton, E. 2006c. Vegetation Response, Fire Effects, and Tree Growth after Slashburning in the Engelmann Spruce-Subalpine Fir Zone: Goat River Site. Technical Report 37. Victoria, BC, Ministry of Forests and Range. Hamilton, E. 2007. Post-fire Vegetation Development and Fire Effects in the SBS Zone: Haggen Creek, Francis Lake, Genevieve Lake, Brink, and Indianpoint Sites. Technical Report 41. Victoria, BC, Ministry of Forests and Range. Hamilton, E.H. and Haeussler, S. 2008. Modeling stability and resilience after slashburning across a sub-boreal to subalpine forest gradient in British Columbia. Canadian Journal of Forest Research, 38: 304-316. Hamilton, E. and Peterson, L. 2003. Response of Vegetation to Burning in a Subalpine Forest Cutblock in Central British Columbia: Otter Creek Site. Research Report 23. Victoria, BC, Ministry of Forests and Range. Hamilton, E. and Peterson, L. 2006. Succession after slashburning in an Engelmann Spruce-Subalpine Fir subzone variant: West Twin site. Technical Report 28. Victoria, BC, Ministry of Forests and Range. Hamrick, J.L. 2004. Response of forest trees to global environmental changes. Forest Ecology and Management, 197: 323-335. Haughian, S.R., Burton P.J., Taylor S.W., and Curry C.L. 2012. Expected effects of climate change on forest disturbance regimes in British Columbia. BC Journal of Ecosystems and Management, 13:1-24. Herzog, W. and Boomsma, A. 2009. Small-Sample Robust Estimators of Noncentrality-Based and Incremental Model Fit. Structural Equation Modeling, 16: 1-27. Hiers, J.K., O'Brien, J.J., Will, R.E., and Mitchell, R.J. 2007. Forest floor depth mediates understory vigor in xeric Pinus palustris ecosystems. Ecological Applications, 17: 806-814. 86  Higgins, S.I., Nathan, R., and Cain, M.L. 2003. Are long-distance dispersal events in plants usually caused by nonstandard means of dispersal? Ecology, 84: 1945-1956. Hofmeier, S. 2001. Vegetation response to site preparation treatments in a wet cool subzone of the sub-boreal spruce zone in the central interior of British Columbia: an 11-year assessment. MSc Thesis, University of Northern British Columbia. Holling, C.S. 1973. Resilience and Stability of Ecological Systems. Annual Review of Ecology and Systematics, 4: 1-23. Holling, C.S. 1986. Resilience of ecosystems: Local surprise and global change. Sustainable development and the biosphere. Cambridge, UK, Cambridge University Press: 292-317. Hooper, D.U. and Vitousek, P.M. 1998. Effects of Plant Composition and Diversity on Nutrient Cycling. Ecological Monographs, 68: 121-149. Hooper, D.U., Solan, M., Symstad, A., Diaz, S., Gessner, M.O., Buchmann, N., Degrange, V., Grime, P., Hulot, F., Mermillod-Blondin, F., Roy, J., Spehn, E., and van Peer, L.  2002. Chapter 17: Species diversity, functional diversity, and ecosystem functioning. In Biodiversity and Ecosystem Functioning. Oxford University Press. 195-281. Hu, L. and Bentler, P.M. 1999. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6: 1-55. Hu, L., Bentler, P.M., and Kano, Y. 1992. Can test statistics in covariance structure analysis be trusted? Psychological Bulletin, 112: 351-362. Intergovernmental Panel on Climate Change. 2007. Climate Change 2007: Synthesis Report. Valencia, Spain. 73p. Iriondo, J.M., Albert, M.J., and Escudero, A.  2003. Structural equation modelling: an alternative for assessing causal relationships in threatened plant populations. Biological Conservation, 113: 367-377. Humbert, L., Gagnon, D., Kneeshaw, D., and Messier, C. 2007. A shade tolerance index for common understory species of northeastern North America. Ecological Indicators, 7: 195-207. Johnson, L.M. 2006. Gitksan medicinal plants-cultural choice and efficacy. Journal of ethnobiology and ethnomedicine, 2: 29. Johnstone, J.F. and Chapin III, F.S. 2006. Fire interval effects on successional trajectory in boreal forests of northwest Canada. Ecosystems, 9: 268-277. Johnstone, J.F., Chapin III, F.S., Hollingsworth, T.N., Mack, M.C., Romanovsky, V., and Turetsky, M. 2010. Fire, climate change, and forest resilience in interior Alaksa. Can. J. For. Res., 40: 1302-1312. Jonsson, M. and Wardle, D.A. 2010. Structural equation modelling reveals plant-community drivers of carbon storage in boreal forest ecosystems. Biology Letters, 6: 116-119. Jöreskog, K.G. and Sörbom, D. 1984. LISREL VI user’s guide (3rd ed.). Mooresville, in: Scientific Software. Jöreskog, K.G. and Sörbom, D. 1993. Structural equation modeling with the SIMPLIS command language. Chicago: Scientific Software. Judd, C.M. and Kenny, D.A. 1981. Process Analysis: Estimating mediation in treatment evaluations. Evaluation Review, 5: 602-619. 87  Karim, M.N. and Mallik, A.U. 2008. Roadside revegetation by native plants I. Roadside microhabitats, floristic zonation and species traits. Ecological Engineering, 32: 222-237. Keeley, J.E. 2009. Fire intensity, fire severity and burn severity: a brief review and suggested usage. International Journal of Wildland Fire, 18: 116-126. Keeley, J.E. and McGinnis, T.W. 2007. Impact of prescribed fire and other factors on cheatgrass persistence in a Sierra Nevada ponderosa pine forest. International Journal of Wildland Fire, 16: 96-106. Kessel, S.R. 1979. Gradient modeling, resource and fire management. 433 p. Khamis, F.G. and Hanoon, M.F. 2010. Modeling the Relationship Between Babies’ Mortality and Prosperity Using Fully Latent Models. Journal of Modern Mathematics and Statistics, 4: 89-95. Kimmins, J.P. 2004. Forest Ecology; A foundation for Sustainable Forest Management and Environmental Ethics in Forestry. Third Edition. Pearson Prentice Hall, NJ. 611 p. Kimmins J.P., Blanco J.A., Seely B., Welham C., and Scoullar, K. 2010. Forecasting Forest Futures: A Hybrid Modelling Approach to the Assessment of Sustainability of Forest Ecosystems and their Values. Earthscan Ltd. London, UK. 281 p. Kleyer, M. 1999. Distribution of Plant Functional Types along Gradients of Disturbance Intensity and Resource Supply in an Agricultural Landscape. Journal of Vegetation Science, 10: 697-708. Klinka, K., Krajina, V.J., Ceska, A., and Scagel, A.M. 1989. Indicator Plants of Coastal British Columbia. UBC Press, Vancouver, Canada. 288 p. Kopra, K. 2003. Effects of natural disturbance and harvesting on the landscape and stand level structure of wet, cold Engelmann spruce subalpine fir forests of south-central British Columbia, Canada.  MSc Thesis, University of British Columbia. Krajina, V.J., Klinka, K., and Worrall, J. 1982. Distribution and ecological characteristics of trees and shrubs of British Columbia. Faculty of Forestry, University of British Columbia, Vancouver, Canada. 131 p. Kranabetter, J.M., and Macadam, A.M. 2007. Changes in carbon storage of broadcast burn plantations over 20 years. Canadian Journal of Soil Science, 87: 93-102. Kranabetter, J.M., Dawson, C.R., and Dunn, D.E. 2007. Indices of dissolved organic nitrogen, ammonium and nitrate across productivity gradients of boreal forests. Soil Biology & Biochemistry, 39: 3147-3158. Kranabetter, J.M., Sanborn, P., Chapman, B.K., and Dube, S. 2006. The Contrasting Response to Soil Disturbance between Lodgepole Pine and Hybrid White Spruce in Subboreal Forests. Soil Sci. Soc. Am. J., 70: 1591-1599. Kranabetter, J.M., Simard, S.W., Guy R.D, and Coates, K.D. 2010. Species patterns in foliar nitrogen concentration, nitrogen content and 13C abundance for understory saplings across light gradients. Plant Soil, 327: 389-401. Kurz, W.A., Dymond, C.C., Stinson, G., Rampley, G.J., Neilson, E.T., Carroll, A.L., Ebata, T., and Safranyik, L. 2008. Mountain pine beetle and forest carbon feedback to climate change. Nature, 452: 987-990. Lang, N.L. and Halpern, C.B. 2007. The soil seed bank of a montane meadow: consequences of conifer encroachment and implications for restoration. Canadian Journal of Botany, 85: 557-569. 88  Larsen, J.B. 1995. Ecological stability of forests and sustainable silviculture. Forest Ecology and Management, 73: 85-96. Laughlin, D.C., Abella, S.R., Covington, W.W., and Grace, J.B. 2007. Species richness and soil properties in Pinus ponderosa forests: A structural equation modeling analysis. Journal of Vegetation Science, 18: 231-242. Legendre, P. and Legendre, L. 2012. Numerical ecology. Volume 20 of Developments in Environmental Modelling. Elsevier. 1006 p. LePage, P.T., Canham, C.D., Coates, K.D., and Bartemucci, P. 2000. Seed abundance versus substrate limitation of seedling recruitment in northern temperate forests of British Columbia. Can. J. For. Res., 30: 415-427. Lieffers, V.J., Stewart, J.D., Macmillan, R.B., Macpherson, D., and Branter, K. 1996. Semi-natural and intensive silvicultural systems for the boreal mixedwood forest. For. Chron., 72: 286-292. Lloret, F. and Vilà, M. 2003. Diversity patterns of plant functional types in relation to fire regime and previous land use in Mediterranean woodlands. Journal of Vegetation Science, 14: 387-398. Lloyd, D.A., Angove, K., Hope, G.D., and Thompson, C. 1990. A Guide to Site Identification and Interpretation for the Kamloops Forest Region. Res. Br., Victoria, B.C. Land Manage. Handb. No. 23. http://www.for.gov.bc.ca/hfd/pubs/docs/Lmh/Lmh23.htm. Last accessed September 2013. Macadam, A. (1989). Effects of prescribed fire on forest soils. BC Ministry of Forests. Smithers, Research Report, 89001ΟPR. MacCallum, R.C., Browne, M.W., and Sugawara, H.M. 1996. Power analysis and determination of sample size for covariance structure modeling. Psychological methods, 1: 130-149. MacDougall, A.S. 2005. Responses of diversity and invasibility to burning in a northern oak savanna. Ecology, 86: 3354-3363. MacGillivray, C.W., Grime, J.P., and the Integrated Screening Programme (ISP) Team. 1995. Testing Predictions of the Resistance and Resilience of Vegetation Subjected to Extreme Events. Functional Ecology, 9: 640-649. Mack, M.C., Treseder, K.K., Manies, K.L., Harden, J.W., Schuur, E.A.G., Volgel, J.G., Randerson, J.T., and Chapin III, F.S. 2008. Recovery of aboveground plant biomass and productivity after fire in mesic and dry black spruce forests of interior Alaska. Ecosystems, 11: 209-225. Mackinnon, A., Pojar, R., and Coupé, R. 2005. Plants of Northern British Columbia. Lone Pine Publishing, Vancouver, Canada. 351 p. MacKinnon, D.P., Fairchild, A.J., and Fritz, M.S. 2007. Mediation analysis. Annual Review of Psychology, 58: 593-614. Manly, B. 2005. Multivariate Statistical Methods: A Primer. Chapman and Hall/CRC. 214 p. Martin, K., Aitken, K.E., and Wiebe, K.L. 2004. Nest sites and nest webs for cavity-nesting communities in interior British Columbia, Canada: nest characteristics and niche partitioning. The condor, 106: 5-19. Massicotte, H.B., Melville, L.H., Tackaberry, L.E., and Peterson, R.L. 2008. A comparative study of mycorrhizas in several genera of Pyroleae (Ericaceae) from western Canada. Botany, 86: 610-622. 89  McCune, B. and Grace, J.B. 2002. Analysis of Ecological Communities. MjM Software Design: Gleneden Beach, Oregon, 300 p. McDonald, R.P. and Hartmann, W.M. 1992. A procedure for obtaining initial values of parameters in the RAM model. Multivariate Behavioral Research, 27: 57-76. McMinn, R.G. 1982. Ecology of site preparation to improve performance of planted white spruce in northern latitudes. In Forest regeneration at high latitudes Experiences from British Columbia. USDA Forest Service, Pacific Northwest For. Range Exp. Sta. Report Number 82-1: 23-25. McRae, D.J., Alexander, M.E., and Stocks, B.J. 1979. Measurement and description of fuels and fire behavior on prescribed burns: a handbook. Can. For. Serv. Inf. Rep. O-X-287. Meidinger, D. and Pojar, J. 1991. Ecosystems of British Columbia. Crown Publications Inc. Victoria, Canada 330 p. Mitchell, R.J. 1992. Testing evolutionary and ecological hypotheses using path analysis and structural equation modelling. Functional Ecology, 6: 123-129. Mueller, R.O. and Hancock, G.R. 2008. Chapter 32: Best Practices in Structural Equation Modeling. In Best Practices in Quantitative Methods. Sage Publications, Inc. 488-508.  Newland, J.A. and DeLuca, T.H. 2000. Influence of fire on native nitrogen-fixing plants and soil nitrogen status in ponderosa pine - Douglas-fir forests in western Montana. Canadian Journal of Forest Research, 30: 274-282. Nigh, G.D. 1996. A variable growth intercept model for spruce in the Sub-Boreal spruce and Engelmann Spruce-Subalpine Fir biogeoclimatic zones of British Colombia. Research Report RR-05. Ministry of Forests, Victoria, Canada. 20 p. Noble I, R. and Slatyer, R.O. 1980. The use of vital attributes to predict successional changes in plant communities subject to recurrent disturbances. Vegetatio, 43: 5-21. Oksanen, J., Guillaume, B., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Henry, M., Stevens, H., and Wagner, H. 2013. vegan: Community Ecology Package. R package version 2.0-10. http://CRAN.R-project.org/package=vegan. Olsson, U.H., Foss, T., Troye, S.V., and Howell, R.D. 2000. The performance of ML, GLS, and WLS estimation in structural equation modeling under conditions of misspecification and nonnormality. Structural Equation Modeling, 7: 557-595. Parish, R., Antos, J.A. and Fortin, M.-J. 1999. Stand development in an old-growth subalpine forest in southern interior British Columbia. Can. J. For. Res., 29: 1347-1356. Parminter, J. 1983. Fire-ecological relationships for the biogeoclimatic zones of the Cassiar Timber Supply Area. B.C. Min. For., Prot. Br., Planning, Development and Research Section, Victoria, B.C. 179 p. Perry, D.A., Hessburg, P.F., Skinner, C.N., Spies, T.A., Stephens, S.L., Taylor, A.H., Franklin, J.F., McComb, B., and Riegel, G. 2011. The ecology of mixed severity fire regimes in Washington, Oregon, and Northern California. Forest Ecology and Management, 262: 703-717. Petit, R.J., Aguinagalde, I., de Beaulieu, J.-L., Bittkau, C., Brewer, S., Cheddadi, R., Ennos, R., Fineschi, S., Grivet, D., Lascoux, M., Mohanty, A., Gerhard Müller-Starck, G., Demesure-Musch, B., Palme´, A., Martin, J.P., Rendell, S., and Vendramin, G.G. 2003. Glacial refugia: Hotspots but not melting pots of genetic diversity. Science, 300: 1563-1565. 90  Petit, R.J., Bialozyt, R., Garnier-Géré, P., and Hampe, A. 2004. Ecology and genetics of tree invasions: from recent introductions to Quaternary migrations. Forest Ecology and Management, 197: 117-137. Petraitis, P.S. and Latham, R.E. 1999. The importance of scale in testing the origins of alternative community states. Ecology, 80: 429-442. Pojar, J. and Mackinnon, A. 1994. Plants of the Pacific Northwest Coast. Lone Pine Publishing, Vancouver, Canada. 528 p. Prescott, C.E., Zabek, L.M. Staley, C.L., and Kabzems R. 2000. Decomposition of broadleaf and needle litter in forests of British Columbia: influences of litter type, forest type, and litter mixtures. Can. J. For. Res., 30: 1742-1750. Pugesek, B.H. and Tomer, A. 2003. Structural equation modeling: applications in ecological and evolutionary biology. Cambridge University Press. Read, D.J. and Perez-Moreno, J. 2003. Mycorrhizas and nutrient cycling in ecosystems - a journey towards relevance? New Phytologist, 157: 475-492. Romme, W.H. 1982. Fire and landscape diversity in subalpine forests in Yellowstone national park. Ecological Monographs, 52: 199-221. Rowe, J.S. 1983. Chapter 8: Concepts of Fire Effects on Individuals and on Species. In The Role of Fire in Northern Circumpolar Ecosystems. John Wiley & Sons Ltd. 135-154. Royal Botanic Gardens Kew. 2008. Seed Information Database (SID). Version 7.1. Available from: http://data.kew.org/sid (May 2008). Last accessed September 2013. Safranyik, L. and Wilson, B. 2007. The mountain pine beetle: a synthesis of biology, management and impacts on lodgepole pine. Canadian Forest Service. Victoria, B.C. 299 p. Satorra A. and Bentler P.M. 1994. Corrections to test statistics and standard errors in covariance structure analysis. In: von Eye A, Clogg CC, editors. Latent Variables Analysis: Applications for developmental research. Sage. Thousand Oaks, CA. p. 399-419. Satorra, A. and Bentler, P.M. 2010. Ensuring positiveness of the scaled difference chi-square test statistic. Psychometrika, 75: 243-248. Savage, M. and Mast, J.N. 2005. How resilient are southwestern ponderosa pine forests after crown fires? Canadian Journal of Forest Research, 35: 967-977. Schermelleh-Engel, K., Moosbrugger, H., and Müller, H. 2003. Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of psychological research online, 8: 23-74. Schimel, J.P. and Bennett, J. 2004. Nitrogen mineralization: challenges of a changing paradigm. Ecology, 85: 591-602. Selosse, M.-A. and Roy, M. 2009. Green plants that feed on fungi: facts and questions about mixotrophy. Trends in Plant Science, 14: 64-70. Sera, B. and Sery, M. 2004. Number and weight of seeds and reproductive strategies of herbaceous plants. Folia Geobotanica, 39: 27-40. Shenoy, A., Johnstone, J.F., Kasischke, E.S., and Kielland, K. 2011. Persistent effects of fire severity on early successional forests in interior Alaska. Forest Ecology and Management, 261: 381-390. 91  Silver, W.L. and Miya, R.K. 2001. Global patterns in root decomposition: comparisons of climate and litter quality effects. Oecologia, 129: 407-419. Simard, S.W.  2009.  The foundational role of mycorrhizal networks in self-organization of interior Douglas-fir forests.  Forest Ecology and Management, 258S: S95-S107. Simard, S.W., Beiler, K.J., Bingham, M.A., Deslippe, J.R., Philip, L.J., and Teste, F.P. 2012. Mycorrhizal networks: mechanisms, ecology and modelling. Fungal Biology Reviews, 26: 39-60. Simard, S.W., Sachs, D.L., Vyse, A., and Blevins, L.L. 2004. Paper birch competitive effects vary with conifer tree species and stand age in interior British Columbia forests: implications for reforestation policy and practice. Forest ecology and management, 198: 55-74. Soil Classification Working Group. 1998. The Canadian System of Soil Classification. 3rd ed. Agriculture and Agri-Food Canada Publication 1646, 187 p. Spittlehouse, D.L. 2008. Climate Change, Impacts, and Adaptation Scenarios: Climate Change and Forest and Range Management in British Columbia. Technical Report 45. Victoria, BC, Ministry of Forests and Range. Starfield, A.M. and Chapin III, F.S. 1996. Model of transient changes in arctic and boreal vegetation in response to climate and land use change. Ecological Applications, 6: 842-864. Stark, K.E., Arsenault, A., and Bradfield, G.E. 2006. Soil seed banks and plant community assembly following disturbance by fire and logging in interior Douglas-fir forests of south-central British Columbia. Canadian Journal of Botany, 84: 1548-1560. Steiger, J.H. and Lind, J.C. 1980. Statistically-based tests for the number of common factors. Paper presented at the annual Spring meeting of the Psychometric Society, Iowa City, IA. Stocks, B.J., Mason, J.A., Todd, J.B., Bosch, E.M., Wotton, B.M., Amiro, B.D., Flannigan, M.D., Hirsch, K.G., Logan, K.A., Martell, D.L., and Skinner, W.R. 2003. Large forest fires in Canada, 1959-1997. J. Geophys. Res., 108 (D1) Article No. 8149. Swift, K. 2008. Summary of the 2008 SISCO Winter Workshop, Part 1: Defining, designing, and planning for the resilient forest in everyday silviculture. Link, 10: 18. Swain, A.J. 1975. Analysis of parametric structures for variance matrices. Unpublished doctoral dissertation, Department of Statistics, University of Adelaide, Australia. Tabachnik, B.G. and L.S. Fidell 2007. Using Multivariate Statistics. Boston, MA, Pearson Education, Inc. 980 p. Taylor, D.L., Bruns, T.D., Leake, J.R., and Read, D.J. 2002. Mycorrhizal Specificity and Function in Myco-heterotrophic Plants. Ecological Studies, 157: 375-413. Taylor, S.W. 1987. Initial effects of slashburning on the nutrient status of two sub-boreal spruce zone ecosystems. MSc Thesis, University of British Columbia. Tendersoo, L., Pellet, P., Koljalg, U., and Selosse, M.-A. 2007. Parallel evolutionary paths to mycoheterotrophy in understorey Ericaceae and Orchidaceae: ecological evidence for mixotrophy in Pyroleae. Oecologia, 151: 206-217. ter Braak, C.J.F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179. 92  Thanos, C.A. and Rundel, P.W. 1995. Fire-followers in chaparral: nitrogenous compounds trigger seed germination. Journal of Ecology, 83: 207-216. The R Foundation for Statistical Computing. 2013. R -version 3.0.1. Vienna, Austria. Thompson, K., Bakker, J.P., and Bekker, R.M. 1997. The soil seed banks of North West Europe: methodology, density and longevity. Cambridge University Press, Cambridge, UK. 288 p. Tilman, D. 1987. Secondary succession and the pattern of plant dominance along experimental nitrogen gradients. Ecological monographs, 57: 198-214. Treseder, K.K., Mack, M.C., and Cross, A. 2004. Relationships among fires, fungi, and soil dynamics on Alaskan Boreal Forests. Ecological Applications, 14: 1826-1838. Trofymow, J.A., Addison, J., Blackwell, B.A., He, F., Preston, C.A., and Marshall, V.G. 2003. Attributes and indicators of old-growth and successional Douglas-fir forests on Vancouver Island. Environ. Rev. 11: S187-S204. Trowbridge, R., Hawkes, B., Macadam, A., and Parminter, J. 1987. Field handbook for prescribed fire assessments in British Columbia: logging slash fuels [online]. Victoria, BC. Available from: www.for.gov.bc.ca/hfd/pubs/Docs/FRH/Frh001.htm. Last accessed September 2013. Tucker, L.R. and Lewis, C. 1973. A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38: 1-10. USDA Forest Service. 2011. Fire Effects Information System (FEIS). Rocky Mountain Research Station, Fire Sciences Laboratory, Missoula MT USA. Available from: http://www.fs.fed.us/database/feis. Last accessed September 2013. USDA Natural Resources Conservation Service. 2011. The PLANTS Database. National Plant Data Team, Greensboro NC USA. Available from: http://plants.usda.gov. Last accessed Septermber 2013. Van Wagner, C.E. 1987. Development and structure of the Canadian forest fire weather index system. Ottowa, ON, Canadian Forest Service. For. Tech. Rep. 35. 37 p. Vile, D., Shipley, B., and Garnier, E. 2006. A structural equation model to integrate changes in functional strategies during old-field succession. Ecology, 87: 504-517. Voller, J. and S. Harrison. 1998. Conservation biology principles for forested landscapes. University of British Columbia Press, 256 p. Vyse, A. and Muraro, S.J. 1973. Reduced planting cost - a prescribed fire benefit. Can. For. Service Information report. BC-X-84. Wang, B. and Qi, Y.-L. 2006. Phylogenetic distribution and evolution of mycorrhizas in land plants. Mycorrhiza, 16: 299-363. Wang, T., Hamann, A., Spittlehouse, D.L., and Aitken, S.N. 2006. Development of scale-free climate data for western Canada for use in resource management. Int. J. Climatol., 26: 383–397. Whitham, T.G., Bailey, J.K., Schweitzer, J.A., Shuster, S.M., Bangert, R.K., LeRoy, C.J., Lonsdorf, E., Allan, G.J., DiFazio, S.P., Potts, B.M., Fischer, D.G.,Gehring, C.A., Lindroth, R.L., and Marks, J. 2006. Community and ecosystem genetics: a framework for integrating from genes to ecosystems. Nature Reviews. Genetics, 7: 510-523. Wilcoxon, F. 1945. Individual comparisons by ranking methods. Biometrics Bulletin 1: 80–83. Wright, S. 1918. On the nature of size factors. Genetics 3: 367-374. 93  Zak, S.K., Beven, K., and Reynolds, B. 1997. Uncertainty in the estimation of critical loads: a practical methodology. Water, Soil and Air Pollution, 98: 297-316. Zak, D.R., Holmes, W.E., White, D.C., Peacock, A.D., and Tilman, D. 2003. Plant Diversity, Soil Microbial Communities, and Ecosystem Function: Are There any Links? Ecology, 84: 2042-2050. 94  Appendix A. Species list of 183 vascular plants from 16 study sites in central British Columbia Code Species Authority Common NameABIEAMA Abies amabilis (Dougl. ex Loud.) Dougl. ex Forbes amabilis firABIELAS Abies lasiocarpa (Hook.) Nutt. subalpine firACERGLA Acer glabrum Torr. Douglas mapleACHIMIL Achillea millefolium L. yarrowACTARUB Actaea rubra (Ait.) Willd. baneberryAGOSAUR Agoseris aurantiaca (Hook.) Greene orange agoserisAGROEXA Agrostis exarata Trin. spike bentgrassAGROSCA Agrostis scabra Willd. hair bentgrassALNUCRI Alnus crispa (Ait.) Pursh green or Sitka alderALNUTEN Alnus tenuifolia Nutt. mountain alderAMELALN Amelanchier alnifolia Nutt. saskatoonANAPMAR Anaphalis margaritacea (L.) Benth. & Hook. f. ex C.B. Clarke pearly everlastingANTENEG Antennaria neglecta Greene field pussytoesANTEPUL Antennaria pulcherrima (Hook.) Greene showy pussytoesAQUIFOR Aquilegia formosa Fisch. ex DC. Sitka columbineARALNUD Aralia nudicaulis L. wild sarsaparillaARNICOR Arnica cordifolia Hook. heart-leaved arnicaARNILAT Arnica latifolia Bong. mountain arnicaARUNDIO Aruncus dioicus (Walt.) Fern. goatsbeardATHYFIL Athyrium filix-femina (L.) Roth lady fernBETUPAP Betula papyrifera Marsh. paper birchBOTRLUN Botrychium lunaria (L.) Sw. common moonwortBOTRVIR Botrychium virginianum (L.) Sw. rattlesnake fernBROMCIL Bromus ciliatus L. fringed bromeCALACAN Calamagrostis canadensis (Michx.) Beauv. bluejoint reedgrassCALARUB Calamagrostis rubescens Buckl. pinegrassCALASTR Calamagrostis stricta (Timm) Koel. slimstem reedgrassCALTLEP Caltha leptosepala DC. white mountain marsh-marigoldCARECOI Carex concinna R. Br. low northern sedgeCAREDEW Carex deweyana Schwein. Dewey's sedgeCAREFOE Carex foenea Willd. bronze sedgeCAREMAC Carex macloviana d'Urv. Falkland Island sedgeCAREMER Carex mertensii Prescott ex Bong. Mertens' sedgeCAREPAC Carex pachystachya Cham. ex Steud. thick-headed sedgeCAREPEC Carex peck ii Howe Peck's sedgeCAREROS Carex rossii Boott Ross' sedgeCASTMIN Castilleja miniata Dougl. ex Hook. scarlet paintbrushCASTPAR Castilleja parviflora Bong. small-flowered paintbrushCERA_SP Cerastium sp. unidentified chickweedCHIMUMB Chimaphila umbellata (L.) Bart. prince's pineCHRYTET Chrysosplenium tetrandrum (Lund) T. Fries northern golden-saxifrageCINNLAT Cinna latifolia (Trev. ex Goepp.) Griseb. nodding wood-reedCIRCALP Circaea alpina L. enchanter's-nightshadeCIRS_SP Cirsium sp. unidentified thistleCLINUNI Clintonia uniflora (Menzies ex J.A. & J.H. Schult.) Kunth queen's cupCORNCAN Cornus canadensis L. bunchberryCORNSTO Cornus stolonifera Michx. red-osier dogwoodCORYSEM Corydalis sempervirens (L.) Pers. pink corydalisCREP_SP Crepis sp. unidentified hawksbeardDESCCES Deschampsia cespitosa (L.) Beauv. tufted hairgrassDIPHCOM Diphasiastrum complanatum (L.) Holub ground-cedarDRYOEXP Dryopteris expansa (K.B. Presl) Fraser-Jenkins & Jermy spiny wood fernELYMGLA Elymus glaucus Buckl. blue wildryeEPILANG Epilobium angustifolium L. fireweedEPILCIL Epilobium ciliatum Raf. purple-leaved willowherbEPILGLA Epilobium glaberrimum Barbey smooth willowherbEQUIARV Equisetum arvense L. common horsetailEQUIPRA Equisetum pratense Ehrh. meadow horsetailEQUISCI Equisetum scirpoides Michx. dwarf scouring-rushEQUISYL Equisetum sylvaticum L. wood horsetailEURYCON Eurybia conspicua (Lindl.) Á. Löve & D. Löve showy aster95 Appendix A continued…   Code Species Authority Common NameFESTOCC Festuca occidentalis Hook. western fescueFRAGVIR Fragaria virginiana Duchesne wild strawberryGALIBOR Galium boreale L. northern bedstrawGALIKAM Galium kamtschaticum Steller ex Schult. & Schult. boreal bedstrawGALITRF Galium triflorum Michx. sweet-scented bedstrawGEOCLIV Geocaulon lividum (Richards.) Fern. false toad-flaxGERABIC Geranium bicknellii Britt. Bicknell's geraniumGEUMMAC Geum macrophyllum Willd. large-leaved avensGOODOBL Goodyera oblongifolia Raf. rattlesnake-plantainGOODREP Goodyera repens (L.) R. Br. dwarf rattlesnake orchidGYMNDRY Gymnocarpium dryopteris (L.) Newman oak fernHERAMAX Heracleum maximum Bartr. cow-parsnipHEUCCHL Heuchera chlorantha Piper meadow alumrootHIERALB Hieracium albiflorum Hook. white hawkweedHIERAUR Hieracium aurantiacum L. orange-red king devilHIERGRA Hieracium gracile Hook. slender hawkweedHIERSCO Hieracium scouleri Hook. Scouler's hawkweedHIERUMB Hieracium umbellatum L. narrow-leaved hawkweedHUPEOCC Huperzia occidentalis (Clute) Kartesz & Gandhi western fir clubmossJUNC_SP Juncus sp. unidentified rushLEPTPYR Leptarrhena pyrolifolia (D. Don) R. Br. ex Ser. leatherleaf saxifrageLEUCVUL Leucanthemum vulgare Lam. oxeye daisyLINNBOR Linnaea borealis L. twinflowerLISTCOR Listera cordata (L.) R. Br. heart-leaved twaybladeLONIINV Lonicera involucrata (Richards.) Banks ex Spreng. black twinberryLUPIARC Lupinus arcticus S. Wats. Arctic lupineLUZUPAR Luzula parviflora (Ehrh.) Desv. small-flowered wood-rushLYCOANN Lycopodium annotinum L. stiff club-mossLYCOCLA Lycopodium clavatum L. running club-mossLYCODEN Lycopodium dendroideum Michx. ground-pineMAIACAN Maianthemum canadense Desf. wild lily-of-the-valleyMAIARAC Maianthemum racemosum (L.) Link false Solomon's-sealMAIASTE Maianthemum stellatum (L.) Link star-flowered false Solomon's-sealMELALIN Melampyrum lineare Desr. cow-wheatMENZFER Menziesia ferruginea Sm. false azaleaMITEBRE Mitella breweri A.Gray brewer's mitrewortMITENUD Mitella nuda L. common mitrewortMITEPEN Mitella pentandra Hook. five-stamened mitrewortMONEUNI Moneses uniflora (L.) A. Gray single delightONOPACA Onopordum acanthium L. Scotch thistleOPLOHOR Oplopanax horridus (Smith) Miq. devil's clubORTHSEC Orthilia secunda (L.) House one-sided wintergreenORYZASP Oryzopsis asperifolia Michx. rough-leaved ricegrassOSMOBER Osmorhiza berteroi DC. mountain sweet-cicelyPAXIMYR Paxistima myrsinites (Pursh) Raf. falseboxPEDIBRA Pedicularis bracteosa Benth. bracted lousewortPETAFRI Petasites frigidus (L.) Fries sweet coltsfootPETAPAL Petasites palmatus (L.) Fries palmate coltsfootPETASAG Petasites sagittatus (Banks ex Pursh) A. Gray arrow-leaved coltsfootPHLEALP Phleum alpinum L. alpine timothyPHLEPRA Phleum pratense L. common timothyPICEENE Picea engelmannii x glauca Parry ex Engelm. hybrid white sprucePICEGLA Picea glauca (Moench) Voss white sprucePINUCON Pinus contorta Dougl. ex Loud. lodgepole pinePLATAQU Platanthera aquilonis Sheviak northern green rein orchidPLATOBT Platanthera obtusata (Banks ex Pursh) Lindl. one-leaved rein orchidPLATORB Platanthera orbiculata (Pursh) Lindl. large round-leaved rein orchidPLATSTR Platanthera stricta Lindl. slender rein orchidPOAPAL Poa palustris L. fowl bluegrassPOAPRA Poa pratensis L. Kentucky bluegrassPOPUBAL Populus balsamifera L. balsam poplar96 Appendix A continued…  Code Species Authority Common NamePOPUTRE Populus tremuloides Michx. trembling aspenPROSHOO Prosartes hookeri Torr. Hooker's fairybellsPRUNPEN Prunus pensylvanica L. f. pin cherryPRUNVUL Prunella vulgaris L. self-healPSEUMEG Pseudotsuga menziesii var. glauca (Mirbel) Franco interior Douglas-firPTERAQU Pteridium aquilinum (L.) Kuhn bracken fernPYROASA Pyrola asarifolia Michx. pink wintergreenPYROCHL Pyrola chlorantha Sw. green wintergreenPYROMIN Pyrola minor L. lesser wintergreenRANU_SP Ranunculus sp. unidentified buttercupRHINMIN Rhinanthus minor L. yellow rattleRHODALB Rhododendron albiflorum Hook. white-flowered rhododendronRIBEGLA Ribes glandulosum Grauer skunk currantRIBEHUD Ribes hudsonianum Richards. northern blackcurrantRIBELAC Ribes lacustre (Pers.) Poir. black gooseberryRIBELAX Ribes laxiflorum Pursh trailing black currantRIBEOXY Ribes oxyacanthoides L. northern gooseberryROSAACI Rosa acicularis Lindl. prickly roseROSAWOO Rosa woodsii Lindl. prairie roseRUBUIDA Rubus idaeus L. red raspberryRUBUPAR Rubus parviflorus Nutt. thimbleberryRUBUPED Rubus pedatus J.E. Sm. five-leaved brambleRUBUPUB Rubus pubescens Raf. dwarf red raspberrySALIBAC Salix barclayi Andersson Barclay's willowSALIBEB Salix bebbiana Sarg. Bebb's willowSALISCO Salix scouleriana J. Barratt ex Hook. Scouler's  willowSAMBRAC Sambucus racemosa L. red elderberrySENEPAU Senecio pauperculus Michx. Canadian butterweedSENETRI Senecio triangularis Hook. arrow-leaved groundselSORBSCO Sorbus scopulina Greene western mountain-ashSORBSIT Sorbus sitchensis Roemer Sitka mountain-ashSPIRBET Spiraea betulifolia Pall. birch-leaved spireaSPIRDOU Spiraea douglasii Hook. hardhackSPIRPYR Spiraea pyramidata Greene pyramid spireaSTREAMP Streptopus amplexifolius (L.) DC. clasping twistedstalkSTRELAN Streptopus lanceolatus (Ait.) Reveal rosy twistedstalkSYMPCIL Symphyotrichum ciliolatum (Lindl.) Á. Löve & D. Löve Lindley's asterSYMPFOL Symphyotrichum foliaceum Lindl. leafy asterTARAOFF Taraxacum officinale G.H. Weber ex Wiggers common dandelionTHALOCC Thalictrum occidentale A. Gray western meadowrueTHUJPLI Thuja plicata Donn ex D. Don western redcedarTIARTRI1 Tiarella trifoliata var. trifoliata L. three-leaved foamflowerTIARTRI2 Tiarella trifoliata var. unifoliata (Hook.) Kurtz one-leaved foamflowerTRIFHYB Trifolium hybridum L. Alsike cloverTRISCER Trisetum cernuum Trin. nodding trisetumTSUGHET Tsuga heterophylla (Raf.) Sarg. western hemlockTSUGMER Tsuga mertensiana (Bong.) Carr. mountain hemlockURTIDIO Urtica dioica L. stinging nettleVACCALA Vaccinium alaskaense Howell Alaskan blueberryVACCCAE Vaccinium caespitosum Michx. dwarf blueberryVACCMEM Vaccinium membranaceum Dougl. ex Hook. black huckleberryVACCMYR Vaccinium myrtilloides Michx. velvet-leaved blueberryVACCOVA Vaccinium ovalifolium Sm. oval-leaved blueberryVACCPAR Vaccinium parvifolium Sm. red huckleberryVAHLATR Vahlodea atropurpurea (Wahlenb.) Fries mountain hairgrassVALESIT Valeriana sitchensis Bong. Sitka valerianVERAVIR Veratrum viride W. Ait. Indian helleboreVIBUEDU Viburnum edule (Michx.) Raf. highbush-cranberryVIOLADU Viola adunca J.E. Smith early blue violetVIOLAGLA Viola glabella Nutt. stream violetVIOLORB Viola orbiculata Geyer ex Holz. round-leaved violet97 Appendix B continued… Appendix B. Traits used to develop plant functional types Traits related to biotic properties:  1. Plant height*: 0.0 = graminoids, forbs, ferns and dwarf woody plants; 0.2 = medium shrub; 0.6 = tall shrub; 1.0 = tree 2. Plant duration*: 0 = annual; 0.5 = annual and/or biennial and/or perennial; 1 = perennial 3. Leaf duration*: 0 = deciduous (one growing season); 0.5 = semi-deciduous to semi-evergreen (typically some basal leaves overwinter under snowpack); 1 = evergreen 4. Nitrogen-fixing bacteria: 0 = absent; 1 = present Source: Beaudry et al. 1999, 2003. 5. Depth of rooting*: 0.0 = no roots; 0.1 = very shallow-on rock or rooting in duff only; 0.3 = shallow; 0.6 = moderate; 1.0 = deep 6. Vigour of sprouting*: 0 = nonsprouting; 0.3 = weak; 0.6 = moderate; 1.0 = strong 7. Rate of lateral spread: 0.0 = no spread; 0.3 = weak -tufted plants able to get larger; 0.6 = moderate-rhizomatous plants that typically become mat-forming; 1.0 = rapid, wide spreading, from suckers, underground stems etc. Sources: USDA FS 2011; Douglas et al. 1998a, 1998b, 1999a, 1999b, 2000, 2001a, 2001b, 2002; Haeussler 1991; Haeussler pers. obs. 8. Dominant mycorrhizal guild: 1 = primary; 0.5 = secondary; 0 = not applicable (categories: arbuscular mycorrhizal or non-mycorrhizal; ectomycorrhizal; ericoid mycorrhizae; mixotrophic and mycoheterotrophic species with orchid mycorrhizae, ect-endo, arbutoid or a variable or uncertain combination) Source: Wang and Qi 2006.  9. Seed size: 0 = spores and orchids very small; 1 = small; 2 = medium; 3 = large Sources: Burton and Burton 2003; Royal Botanic Gardens Kew 2008; Sera and Sery 2004; USDA NRCS 2011. 10. Seed quantity*: 0 = no sexual reproduction or very few; 1 = few; 2 = moderate; 3 = abundant/frequent; 4 = very large quantities 11. Seed dispersal*: 0 = does not travel; 1 = limited distance (includes ant dispersal); 2 = moderate (includes seed wings and transport by larger animals - berries, hooks, etc.); 3 = adapted for long distance travel (small seed with plumes, pappus); 4 = unlimited (spores and minute seed) 12. Seed longevity: 1 = short (1 yr or less); 2 = medium (transient seedbank >1 yr); 3 = long (persistent seedbanking in soil or canopy, can appear even if plant was not present prior to disturbance) Sources: Lang and Halpern 2007; Thompson et al. 1997; S. Haeussler, pers. comm.   *Sources: Douglas et al. 1998a, 1998b, 1999a, 1999b, 2000, 2001a, 2001b, 2002; Haeussler 1991; S. Haeussler, pers. comm. 98 Appendix B continued…  Traits related to abiotic properties:  1. Ground surface material: 1 = primary; 0.5 = secondary; 0 = not applicable (categories: mor; moder and mull; exposed mineral soil; very shallow soils; surface water) Sources: Douglas et al. 1998a, 1998b, 1999a, 1999b, 2000, 2001a, 2001b, 2002; Mackinnon et al. 1992; Pojar and MacKinnon 1994; S. Haeussler, pers. comm. 2. Climate: 1 = primary; 0.5 = secondary; 0 = not applicable (categories: alpine tundra; montane boreal; subalpine boreal; cool temperate; cool mesothermal; semi-arid) Sources: Douglas et al. 1998a, 1998b, 1999a, 1999b, 2000, 2001a, 2001b, 2002; Klinka et al. 1989; S. Haeussler, pers. comm. 3. Light index: 0 = tolerates deep shade, intolerant of full sun to 1 = requires full sun, does not tolerate shade Sources: Beaudry et al. 1999, 2003; Humbert et al. 2007; Krajina et al. 1982. 99 Appendix B continued…  100 Appendix B continued… 101 Appendix B continued…  102  Appendix C. Dendrogram of cluster analysis results for 183 vascular plants      103 Appendix D. Plant functional type descriptions  Description Species n1 Conifers tall Abies amabilis 9evergreen Abies lasiocarpaseed cones Picea engelmannii x glaucaPicea glaucaPinus contortaPseudotsuga menziesii var. glaucaThuja plicataTsuga heterophyllaTsuga mertensiana2 Gap specialists understory Asteraceae spp. 79semi-shade tolerant deciduous Poaceae spp.mull/moder humus form Rosaceae spp.3 Gap specialists understory Asteraceae spp. 42shade intolerant deciduous Cyperaceae spp.persistent soil seedbanking Poaceae spp.exposed mineral soil 4 Deciduous trees tall Alnus crispa 8semi-shade tolerant Alnus tenuifoliaBetula papyriferaPopulus balsamiferaPopulus tremuloidesSalix barclayiSalix bebbianaSalix scouleriana5 Surface water specialists understory Caltha leptosepala 7generally deciduous Chrysosplenium tetrandrumshade tolerant Equisetum scirpoidessmall seeds/spores Leptarrhena pyrolifolialate snow-melt sites Petasites frigidusPetasites sagittatusPlatanthera aquilonis6 Shade specialists understory Lycopodiaceae 20evergreen Orchidaceaesmallest seeds dwarf woody members of:mor humus form Caprifoliaceae, Cornaceae and Ericaceaemixotrophic7 Ground surface material understory Clintonia uniflora 8generalists deciduous Festuca occidentalisshade tolerant Melampyrum linearemor, moder, mull humus form Oryzopsis asperifoliaPaxistima myrsinitesSorbus scopulinaSorbus sitchensisViola orbiculata8 Montane boreal/ understory Geocaulon lividum 3cool temperate climates deciduous Maianthemum canadenseshade tolerant Vaccinium myrtilloidesmor humus formrevegetate via rhizomes9 Subalpine ericaceous shrubs moderately tall Menziesia ferruginea 7mor humus form Rhododendron albiflorumericoid mycorrhizal Vaccinium spp.late snow-melt sitesPlant Functional Type104 Appendix E. Plant functional type response to gradients of fire severity and soil N Mean and standard deviation (in parentheses) of abundance (mean % cover of species in each functional group) at three levels of forest floor consumption due to slashburning (% LFH) 1, 3, 5, 10 and 20 years post-burn.  Different letters within columns indicate significantly different means (p<0.05) determined by Wilcoxon rank sum tests with Bonferroni correction.  See Appendix D for PFT descriptions.     Year 1LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.03 a 36.29 a 0.85 a 0.03 a 0.00 na 2.07 a 0.39 a 0.00 na 5.91 a(n =30) (0.18) (39.05) (2.01) (0.13) (0.00) (3.20) (0.75) (0.00) (11.13)20-30 0.12 a 48.90 b 8.43 b 0.02 a 0.00 na 0.35 b 0.08 b 0.00 na 1.11 b(n =168) (0.95) (33.89) (11.59) (0.14) (0.00) (0.89) (0.43) (0.00) (3.03)>30 0.36 b 10.23 a 11.03 b 0.66 b 0.00 na 3.24 c 0.40 a 0.00 na 2.01 a(n =21) (0.28) (11.30) (13.48) (1.41) (0.00) (2.91) (0.53) (0.00) (1.81)Year 3LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.53 a 72.42 a 3.20 a 0.02 a 0.00 na 4.68 a 1.33 a 0.00 ab 8.96 a(n =30) (1.02) (48.87) (6.64) (0.09) (0.00) (7.33) (2.95) (0.00) (8.20)20-30 0.11 b 60.18 a 16.46 b 0.50 a 0.00 na 1.65 b 0.13 b 0.00 a 1.36 b(n =168) (0.42) (32.17) (17.25) (5.06) (0.00) (4.40) (0.53) (0.00) (3.24)>30 0.64 a 20.16 b 7.89 c 0.90 b 0.00 na 4.50 a 0.60 a 0.02 b 0.31 b(n =21) (0.45) (21.06) (11.69) (1.15) (0.00) (4.81) (0.65) (0.11) (0.54)Year 5LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.63 a 65.66 a 3.95 a 0.03 a 0.00 na 4.90 a 2.77 a 0.00 na 9.02 a(n =30) (1.20) (50.64) (6.61) (0.18) (0.00) (6.38) (5.51) (0.00) (8.53)20-30 0.39 b 82.01 a 16.91 b 1.24 a 0.00 na 4.59 b 0.68 b 0.00 na 2.37 b(n =168) (1.34) (43.71) (18.44) (8.80) (0.00) (13.24) (5.47) (0.00) (6.79)>30 1.74 c 27.32 b 7.46 c 1.88 b 0.00 na 6.24 a 0.81 a 0.00 na 0.73 b(n =21) (1.74) (28.99) (7.29) (2.60) (0.00) (5.83) (1.03) (0.00) (1.26)Year 10LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 2.07 a 46.18 a 5.79 a 0.17 a 0.00 a 4.13 a 2.00 a 0.00 ab 15.21 a(n =30) (3.73) (39.74) (6.78) (0.65) (0.00) (6.28) (3.79) (0.00) (16.69)20-30 4.63 a 78.59 b 6.82 a 2.90 a 0.03 a 8.13 a 0.59 b 0.00 a 4.26 b(n =168) (9.13) (38.71) (10.00) (11.06) (0.39) (17.54) (2.15) (0.00) (8.88)>30 12.56 b 20.55 a 4.22 a 6.22 b 0.44 b 9.47 b 1.40 a 0.05 b 2.93 c(n =21) (10.99) (16.78) (3.74) (8.75) (1.20) (6.35) (1.34) (0.22) (4.65)Year 20LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 17.29 a 58.90 a 8.39 a 1.43 a 0.00 ab 19.22 ab 2.28 a 0.00 ab 33.25 a(n =30) (19.54) (61.11) (6.46) (4.93) (0.00) (21.11) (3.71) (0.00) (30.00)20-30 26.23 a 109.36 b 8.65 b 13.98 b 0.03 a 10.47 a 0.98 b 0.00 a 16.42 b(n =168) (28.44) (60.14) (15.07) (25.55) (0.23) (16.13) (3.39) (0.00) (25.72)>30 50.96 b 29.95 c 3.57 b 14.31 b 0.15 b 27.90 b 4.73 c 0.14 b 11.60 b(n =21) (21.17) (33.84) (5.23) (16.91) (0.48) (22.06) (5.77) (0.36) (12.77)105 Appendix E continued… Mean and standard deviation (in parentheses) of richness (mean number of species in each functional group) at three levels of forest floor consumption due to slashburning (% LFH) 1, 3, 5, 10 and 20 years post-burn.  Different letters within columns indicate significantly different means (p<0.05) determined by Wilcoxon rank sum tests with Bonferroni correction.  See Appendix D for PFT descriptions.     Year 1LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.03 a 5.77 ab 0.87 a 0.07 a 0.00 na 1.17 a 0.37 a 0.00 na 1.70 a(n =30) (0.18) (4.70) (1.22) (0.25) (0.00) (1.09) (0.49) (0.00) (0.99)20-30 0.09 a 6.44 a 2.22 b 0.04 a 0.00 na 0.29 b 0.09 b 0.00 na 0.45 b(n =168) (0.33) (2.44) (1.32) (0.19) (0.00) (0.53) (0.29) (0.00) (0.93)>30 0.67 b 4.76 b 2.14 b 0.48 b 0.00 na 1.33 a 0.76 a 0.00 na 0.95 c(n =21) (0.48) (3.30) (1.24) (0.68) (0.00) (0.73) (0.70) (0.00) (0.38)Year 3LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.33 a 6.83 ab 1.17 a 0.03 a 0.00 na 1.20 a 0.40 a 0.00 ab 2.07 a(n =30) (0.48) (4.62) (1.26) (0.18) (0.00) (1.13) (0.50) (0.00) (1.17)20-30 0.13 b 7.53 a 2.58 b 0.08 a 0.00 na 0.33 b 0.11 b 0.00 a 0.54 b(n =168) (0.37) (2.54) (1.27) (0.28) (0.00) (0.54) (0.32) (0.00) (1.10)>30 0.71 c 5.00 b 1.52 a 0.62 b 0.00 na 1.57 a 0.81 a 0.05 b 0.38 b(n =21) (0.46) (3.08) (1.21) (0.74) (0.00) (0.68) (0.75) (0.22) (0.59)Year 5LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.37 a 6.73 ab 1.70 a 0.03 a 0.00 na 1.23 a 0.50 a 0.00 na 2.07 a(n =30) (0.49) (4.31) (1.51) (0.18) (0.00) (1.01) (0.63) (0.00) (1.17)20-30 0.15 b 7.79 a 2.38 b 0.10 a 0.00 na 0.40 b 0.14 b 0.00 na 0.56 b(n =168) (0.40) (2.53) (1.30) (0.32) (0.00) (0.60) (0.38) (0.00) (1.12)>30 0.71 c 5.67 b 1.48 a 0.57 b 0.00 na 1.48 a 0.81 c 0.00 na 0.67 b(n =21) (0.46) (4.21) (0.81) (0.68) (0.00) (0.51) (0.60) (0.00) (0.66)Year 10LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 0.70 a 6.77 a 2.40 a 0.13 a 0.00 a 1.30 a 0.47 a 0.00 ab 2.00 a(n =30) (0.53) (4.29) (1.54) (0.35) (0.00) (1.06) (0.57) (0.00) (1.31)20-30 0.52 a 8.92 b 2.77 a 0.29 a 0.02 a 0.73 b 0.18 b 0.00 a 0.65 b(n =168) (0.60) (2.88) (1.66) (0.62) (0.13) (0.80) (0.45) (0.00) (1.12)>30 1.29 b 6.14 a 2.57 a 0.90 b 0.24 b 2.14 a 1.10 c 0.05 b 1.14 a(n =21) (0.72) (4.93) (1.80) (0.94) (0.44) (1.01) (0.54) (0.22) (0.48)Year 20LFH (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft910-20 1.13 ab 7.30 ab 3.23 a 0.27 a 0.00 ab 1.67 ab 0.47 a 0.00 ab 2.10 a(n =30) (0.68) (3.77) (2.01) (0.58) (0.00) (1.42) (0.57) (0.00) (1.18)20-30 0.96 a 8.60 a 1.46 b 0.54 ab 0.02 a 1.11 a 0.21 b 0.00 a 0.88 b(n =168) (0.70) (3.15) (1.38) (0.75) (0.13) (1.07) (0.45) (0.00) (1.18)>30 1.33 b 6.33 b 1.52 b 0.90 b 0.14 b 2.62 b 1.05 c 0.14 b 1.14 c(n =21) (0.66) (6.22) (1.03) (1.00) (0.36) (1.20) (0.59) (0.36) (0.57)106 Appendix E continued… Mean and standard deviation (in parentheses) of abundance (mean % cover of species in each functional group) at four levels of % soil nitrogen 1, 3, 5, 10 and 20 years post-burn.  Different letters within columns indicate significantly different means (p<0.05) determined by Wilcoxon rank sum tests with Bonferroni correction.  See Appendix D for PFT descriptions.     Year 1Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 0.44 a 13.56 a 9.01 ab 0.44 a 0.00 na 2.71 a 0.32 a 0.00 na 1.74 a(n =36) (0.44) (15.57) (14.50) (1.12) (0.00) (2.45) (0.55) (0.00) (1.66)0.1-0.2 0.00 b 18.88 a 8.99 ab 0.16 a 0.00 na 1.11 b 0.09 a 0.00 na 1.28 a(n =20) (0.00) (17.93) (10.57) (0.28) (0.00) (1.39) (0.18) (0.00) (2.34)0.2-0.3 0.01 b 54.90 b 8.34 a 0.00 b 0.00 na 0.47 c 0.02 b 0.00 na 0.31 b(n =126) (0.09) (34.49) (11.41) (0.00) (0.00) (1.67) (0.11) (0.00) (1.51)0.3-0.4 0.32 b 46.91 b 3.23 b 0.00 b 0.00 na 0.26 c 0.48 a 0.00 na 7.55 c(n =37) (1.97) (34.86) (6.01) (0.00) (0.00) (1.19) (1.01) (0.00) (10.25)Year 3Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 0.63 a 23.93 a 8.71 a 0.57 a 0.00 na 3.66 a 0.57 a 0.01 a 0.97 a(n =36) (0.50) (28.24) (13.97) (0.97) (0.00) (4.38) (0.93) (0.08) (1.43)0.1-0.2 0.18 b 47.53 b 16.00 ab 0.58 a 0.00 na 2.63 ab 0.21 a 0.00 a 2.00 a(n =20) (0.29) (34.43) (17.36) (1.77) (0.00) (3.31) (0.49) (0.00) (2.97)0.2-0.3 0.15 bc 63.44 bc 18.23 b 0.56 b 0.00 na 2.52 bc 0.05 b 0.00 a 1.10 b(n =126) (0.61) (32.43) (17.35) (5.80) (0.00) (5.97) (0.33) (0.00) (3.60)0.3-0.4 0.08 c 78.38 c 2.62 a 0.00 b 0.00 na 0.26 c 1.20 a 0.00 a 7.84 c(n =37) (0.49) (34.60) (4.41) (0.02) (0.00) (0.91) (2.67) (0.00) (7.49)Year 5Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 1.65 a 25.06 a 9.74 a 1.25 a 0.00 na 4.94 a 0.72 ac 0.00 na 0.86 a(n =36) (1.50) (24.77) (14.55) (2.26) (0.00) (5.19) (1.21) (0.00) (1.16)0.1-0.2 0.46 b 41.95 a 14.09 ab 1.40 a 0.00 na 3.08 ab 0.11 a 0.00 na 1.75 a(n =20) (0.99) (29.12) (14.90) (3.50) (0.00) (3.08) (0.26) (0.00) (2.55)0.2-0.3 0.37 bc 90.38 b 18.68 b 1.38 b 0.00 na 6.29 bc 0.58 b 0.00 na 1.81 b(n =126) (1.44) (44.13) (18.93) (10.07) (0.00) (15.21) (6.24) (0.00) (7.31)0.3-0.4 0.14 c 86.26 b 3.55 a 0.05 b 0.00 na 0.47 c 3.09 c 0.00 na 10.53 c(n =37) (0.82) (34.07) (3.94) (0.33) (0.00) (1.73) (4.95) (0.00) (7.03)Year 10Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 10.22 a 26.46 a 10.02 a 3.79 a 0.01 ab 8.30 ab 1.48 a 0.05 a 2.50 a(n =36) (7.81) (15.05) (15.91) (5.47) (0.03) (8.42) (2.65) (0.22) (2.82)0.1-0.2 10.69 a 24.81 a 5.15 a 4.82 a 0.26 a 6.25 a 0.99 a 0.00 a 2.86 a(n =20) (10.50) (22.13) (6.25) (9.64) (0.95) (6.07) (1.50) (0.00) (4.11)0.2-0.3 3.64 b 89.05 b 6.83 a 2.89 b 0.04 b 10.10 b 0.18 b 0.00 b 1.85 b(n =126) (9.10) (37.80) (9.35) (12.08) (0.45) (19.66) (1.79) (0.00) (6.49)0.3-0.4 1.51 b 64.19 c 4.21 a 0.28 b 0.00 ab 0.60 c 2.69 a 0.00 a 22.89 c(n =37) (3.72) (25.98) (4.08) (1.15) (0.02) (1.84) (3.62) (0.00) (12.47)Year 20Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 43.48 a 28.29 a 2.69 a 11.88 a 0.14 a 28.77 a 3.24 a 0.08 a 10.43 a(n =36) (20.72) (27.88) (4.22) (15.08) (0.49) (24.68) (5.04) (0.28) (11.39)0.1-0.2 31.10 ab 43.50 a 15.72 b 15.58 a 0.00 ab 21.95 a 3.40 a 0.00 ab 7.45 ab(n =20) (16.66) (32.89) (20.14) (24.41) (0.00) (15.86) (7.08) (0.00) (9.23)0.2-0.3 27.36 bc 117.12 b 8.38 ac 14.37 a 0.02 b 11.14 b 0.16 b 0.00 b 11.67 b(n =126) (30.74) (53.48) (14.62) (26.45) (0.18) (15.62) (0.85) (0.00) (22.22)0.3-0.4 9.78 c 111.42 b 8.44 bc 3.84 b 0.03 ab 1.15 c 3.45 a 0.00 ab 54.21 c(n =37) (17.14) (75.95) (8.99) (16.18) (0.16) (2.45) (4.67) (0.00) (24.90)107 Appendix E continued… Mean and standard deviation (in parentheses) of richness (number of species in each functional group) at four levels of % soil nitrogen 1, 3, 5, 10 and 20 years post-burn.  Different letters within columns indicate significantly different means (p<0.05) determined by Wilcoxon rank sum tests with Bonferroni correction.  See Appendix D for PFT descriptions.  Year 1Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 0.75 a 4.19 a 1.81 ab 0.33 a 0.00 na 1.33 a 0.50 a 0.00 na 0.86 a(n =36) (0.50) (2.79) (1.21) (0.59) (0.00) (0.68) (0.65) (0.00) (0.42)0.1-0.2 0.00 b 5.00 ab 2.40 ab 0.30 a 0.00 na 1.10 a 0.30 a 0.00 na 0.60 a(n =20) (0.00) (4.47) (2.52) (0.47) (0.00) (0.97) (0.47) (0.00) (0.50)0.2-0.3 0.01 b 6.25 b 2.25 a 0.00 b 0.00 na 0.26 b 0.02 b 0.00 na 0.13 b(n =126) (0.09) (2.29) (1.01) (0.00) (0.00) (0.54) (0.15) (0.00) (0.41)0.3-0.4 0.05 b 8.54 c 1.30 b 0.00 b 0.00 na 0.24 b 0.41 a 0.00 na 2.32 c(n =37) (0.33) (2.49) (1.51) (0.00) (0.00) (0.64) (0.50) (0.00) (1.13)Year 3Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 0.75 a 4.61 a 1.28 a 0.42 a 0.00 na 1.33 a 0.56 a 0.03 a 0.56 a(n =36) (0.50) (2.54) (1.11) (0.65) (0.00) (0.76) (0.69) (0.17) (0.56)0.1-0.2 0.30 b 6.10 ab 2.20 ab 0.40 a 0.00 na 1.05 a 0.35 a 0.00 a 0.45 a(n =20) (0.47) (4.22) (2.24) (0.50) (0.00) (1.00) (0.49) (0.00) (0.51)0.2-0.3 0.09 c 7.27 b 2.57 b 0.03 b 0.00 na 0.36 b 0.02 b 0.00 a 0.19 b(n =126) (0.28) (2.38) (0.94) (0.18) (0.00) (0.64) (0.15) (0.00) (0.52)0.3-0.4 0.05 c 10.03 c 2.32 ab 0.03 b 0.00 na 0.30 b 0.49 a 0.00 a 2.92 c(n =37) (0.33) (2.19) (1.76) (0.16) (0.00) (0.62) (0.51) (0.00) (1.14)Year 5Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 0.75 a 4.94 a 1.31 a 0.39 a 0.00 na 1.36 a 0.56 ab 0.00 na 0.69 a(n =36) (0.50) (3.46) (0.86) (0.60) (0.00) (0.68) (0.61) (0.00) (0.58)0.1-0.2 0.40 a 6.45 ab 1.90 ab 0.35 a 0.00 na 1.05 a 0.20 b 0.00 na 0.50 a(n =20) (0.60) (4.55) (1.92) (0.49) (0.00) (0.89) (0.41) (0.00) (0.51)0.2-0.3 0.11 b 7.47 b 2.29 b 0.06 b 0.00 na 0.40 b 0.02 c 0.00 na 0.18 b(n =126) (0.32) (2.10) (0.95) (0.28) (0.00) (0.65) (0.15) (0.00) (0.41)0.3-0.4 0.05 b 10.32 c 2.92 b 0.03 b 0.00 na 0.43 b 0.76 a 0.00 na 3.03 c(n =37) (0.33) (2.01) (1.88) (0.16) (0.00) (0.65) (0.64) (0.00) (1.12)Year 10Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 1.33 a 6.62 a 2.57 ab 0.90 a 0.10 ab 1.62 a 0.57 a 0.05 a 0.86 a(n =36) (0.66) (4.70) (2.01) (1.00) (0.30) (1.32) (0.68) (0.22) (0.36)0.1-0.2 1.11 a 6.03 a 2.91 ab 0.80 a 0.11 a 1.54 a 0.69 a 0.00 a 0.86 a(n =20) (0.58) (4.97) (2.41) (0.93) (0.32) (0.82) (0.68) (0.00) (0.60)0.2-0.3 0.34 b 8.55 b 2.47 a 0.17 b 0.01 b 0.80 b 0.04 b 0.00 b 0.25 b(n =126) (0.49) (2.41) (1.20) (0.41) (0.09) (0.86) (0.20) (0.00) (0.47)0.3-0.4 0.70 c 10.89 c 3.35 b 0.11 b 0.03 ab 0.46 b 0.73 a 0.00 a 3.05 c(n =37) (0.52) (1.95) (1.81) (0.31) (0.16) (0.73) (0.61) (0.00) (1.10)Year 20Soil N (%) pft1 pft2 pft3 pft4 pft5 pft6 pft7 pft8 pft90-0.1 1.33 ac 5.61 a 1.44 a 1.00 a 0.11 a 2.42 a 0.78 a 0.08 a 1.00 a(n =36) (0.59) (4.96) (1.03) (0.93) (0.32) (1.13) (0.64) (0.28) (0.53)0.1-0.2 1.75 a 9.40 abc 4.10 b 1.05 a 0.00 ab 2.85 a 0.75 a 0.00 ab 0.50 b(n =20) (0.72) (6.40) (1.37) (1.00) (0.00) (0.81) (0.72) (0.00) (0.51)0.2-0.3 0.77 b 8.05 b 1.20 a 0.42 b 0.01 b 1.03 b 0.05 b 0.00 b 0.59 b(n =126) (0.55) (2.12) (1.36) (0.58) (0.09) (1.00) (0.21) (0.00) (0.74)0.3-0.4 1.16 c 10.59 c 2.43 c 0.19 c 0.03 ab 0.49 c 0.59 a 0.00 ab 3.11 c(n =37) (0.80) (2.79) (1.34) (0.70) (0.16) (0.69) (0.55) (0.00) (1.13)108  Appendix F. Structural equation modeling (SEM) R code library(sem) data <- read.table("C:/R/sem/data.csv", sep = ",", head=T) attach(data) semdata<-data[,c(3, 4, 6, 11, 183:228)] names(semdata) nrow(semdata) sem.cov<-cov(semdata)  model <- specifyModel() R1->pft1r,NA,1 R1->pft1,lam1 R2->pft2r,NA,1 R2->pft2,lam2 R4->pft4r,NA,1 R4->pft4,lam4 R6->pft6r,NA,1 R6->pft6,lam6 R9->pft9r,NA,1 R9->pft9,lam9 precipitation->MAP,lam21 temperature->MAT,lam22 soils->soil_C,NA,1 soils->soil_N,lam32 burn->LFH_consumed,NA,1 burn->WD_consumed,lam42  temperature<->temperature,NA,1 precipitation<->precipitation,NA,1 soils<->soils,psi30 burn<->burn,psi40 R1<->R1,psi1 R2<->R2,psi2 R4<->R4,psi4 R6<->R6,psi6 R9<->R9,psi9  pft1<->pft1,the11 pft1r<->pft1r,the21 pft2<->pft2,the12 pft2r<->pft2r,the22 pft4<->pft4,the14 pft4r<->pft4r,the24 pft6<->pft6,the16 pft6r<->pft6r,the26 pft9<->pft9,the19 pft9r<->pft9r,the29 MAT<->MAT,NA,1 MAP<->MAP,NA,1 soil_C<->soil_C,the31 soil_N<->soil_N,the32 LFH_consumed<->LFH_consumed,the41 WD_consumed<->WD_consumed,the42  precipitation->temperature,gam10 precipitation->burn,NA,-1 precipitation->R1,gam11 precipitation->R6,gam16109 Appendix F continued… temperature->burn,NA,1 temperature->soils,bet10 temperature->R1,bet21 temperature->R2,bet22 temperature->R4,bet24 temperature->R9,bet29 burn->soils,bet60 burn->R1,bet61 burn->R2,bet62 burn->R9,bet69 soils->R6,bet76 R1->R2,bet82 R1->R4,bet84 R1->R6,bet86  soil_C<-temperature,the101 LFH_consumed<-temperature,the102 MAT<-MAP,the103 R1<-pft9,the104 pft1<-MAP,the105 pft1<-WD_consumed,the106 pft6<-pft1,the107 LFH_consumed<-MAT,the108 R1<-pft2r,the109 pft1r<-MAT,the110 R4<-WD_consumed,the111 R6<-LFH_consumed,the112 pft9<-MAT,the113 soils<-WD_consumed,the114 R4<-LFH_consumed,the115 pft6<-MAT,the116 pft4<-pft2,the117 pft2r<-R6,the118  pft4r<->pft2r,the201 burn<->pft9,the202 burn<->pft2,the203 soil_N<->temperature,the204 precipitation<->pft4,the205 WD_consumed<->pft6r,the206 WD_consumed<->pft2r,the207 R2<->MAT,the208 soil_N<->pft6,the209 LFH_consumed<->pft2r,the210 precipitation<->pft4r,the211 burn<->pft6,the212 R1<->LFH_consumed,the213 burn<->WD_consumed,the214 precipitation<->pft9,the215  sem.out<- sem(model, sem.cov, N=1314) summary(sem.out, standardized = TRUE, fit.measures=TRUE, fit.indices=c("GFI", "AGFI", "CFI", "RMSEA", "AIC", "AICc", "BIC", "NFI", "NNFI", "RNI", "IFI", "SRMR")) modIndices(sem.out) standardizedCoefficients(sem.out) effects(sem.out) residuals(sem.out) pathDiagram(sem.out)110  Appendix G. Covariance matrix for the variables included in the structural equation model  pft1 pft2 pft4 pft6 pft9pft1 242.1533825 62.35921419 42.2032672 55.93519678 27.58405919pft2 62.35921419 2304.615127 66.77521632 -44.2508578 113.0083218pft4 42.2032672 66.77521632 140.3031273 13.42958537 2.337766093pft6 55.93519678 -44.2508578 13.42958537 147.4012229 6.834734206pft9 27.58405919 113.0083218 2.337766093 6.834734206 188.9065202pft1r 4.804324858 -3.332650662 1.125810737 1.981892699 2.058498958pft2r 2.349351874 82.82697408 2.944464151 -2.523397972 6.715443272pft4r 2.338443628 -2.354722358 3.363539873 1.235382564 -0.325221268pft6r 4.325514785 -10.12812723 1.211648183 6.448436256 0.534611037pft9r 0.129895982 -0.941596678 -1.05711901 -0.141460005 9.376497407WD_consumed 17.08717983 -40.89831088 14.88523511 13.49506098 -56.92246346LFH_consumed 9.655413202 -87.98336342 2.472068566 2.125208401 -3.660098697soil_C -3.741068722 32.56845261 -1.263087268 -4.175330693 7.36253711soil_N -0.192745543 2.328038381 -0.025241986 -0.215836948 0.245236215MAT -1.762347981 -9.147423013 -0.641830657 -2.103505218 -3.069861484MAP 342.8208366 6941.77739 536.7243108 -143.477612 -29.90271538pft1r pft2r pft4r pft6r pft9rpft1 4.804324858 2.349351874 2.338443628 4.325514785 0.129895982pft2 -3.332650662 82.82697408 -2.354722358 -10.12812723 -0.941596678pft4 1.125810737 2.944464151 3.363539873 1.211648183 -1.05711901pft6 1.981892699 -2.523397972 1.235382564 6.448436256 -0.141460005pft9 2.058498958 6.715443272 -0.325221268 0.534611037 9.376497407pft1r 0.365449822 -0.087321377 0.113084122 0.263977715 0.102254588pft2r -0.087321377 11.03810102 0.187136364 -0.456753157 0.496238876pft4r 0.113084122 0.187136364 0.275083725 0.143621738 -0.063474841pft6r 0.263977715 -0.456753157 0.143621738 0.885851704 0.0515707pft9r 0.102254588 0.496238876 -0.063474841 0.0515707 1.358470094WD_consumed 0.250800739 -4.151707373 1.476228234 1.819163476 -9.130726455LFH_consumed 1.05862346 -2.320764953 0.735145907 0.874156225 0.759997496soil_C -0.32920152 2.255391597 -0.305059902 -0.710142855 0.885245784soil_N -0.024423383 0.140674586 -0.017134851 -0.046272295 0.018304501MAT -0.116385147 -0.571133936 0.00333076 -0.117148849 -0.470496301MAP -56.82896014 369.7053456 -25.09305261 -86.73808803 31.77424908WD_consumed LFH_consumed soil_C soil_N MAT MAPpft1 17.08717983 9.655413202 -3.741068722 -0.192745543 -1.762347981 342.8208366pft2 -40.89831088 -87.98336342 32.56845261 2.328038381 -9.147423013 6941.77739pft4 14.88523511 2.472068566 -1.263087268 -0.025241986 -0.641830657 536.7243108pft6 13.49506098 2.125208401 -4.175330693 -0.215836948 -2.103505218 -143.477612pft9 -56.92246346 -3.660098697 7.36253711 0.245236215 -3.069861484 -29.90271538pft1r 0.250800739 1.05862346 -0.32920152 -0.024423383 -0.116385147 -56.82896014pft2r -4.151707373 -2.320764953 2.255391597 0.140674586 -0.571133936 369.7053456pft4r 1.476228234 0.735145907 -0.305059902 -0.017134851 0.00333076 -25.09305261pft6r 1.819163476 0.874156225 -0.710142855 -0.046272295 -0.117148849 -86.73808803pft9r -9.130726455 0.759997496 0.885245784 0.018304501 -0.470496301 31.77424908WD_consumed 105.3890181 0.531899133 -11.26242771 -0.412425683 3.728173481 -973.0373608LFH_consumed 0.531899133 35.51586349 -3.278092727 -0.258667744 0.796164801 -707.8925393soil_C -11.26242771 -3.278092727 2.653223901 0.143499786 -0.37086263 320.8651038soil_N -0.412425683 -0.258667744 0.143499786 0.009105182 -0.011422909 21.58483266MAT 3.728173481 0.796164801 -0.37086263 -0.011422909 1.441406848 -104.2192323MAP -973.0373608 -707.8925393 320.8651038 21.58483266 -104.2192323 95564.3907111  Appendix H. Structural equation modeling (SEM) results Parameter Estimate SE z value Plam1 21.8735 1.1283 19.3859 0.0000 pft1 <--- R1lam2 19.1512 1.0596 18.0745 0.0000 pft2 <--- R2lam4 15.1036 1.0755 14.0430 0.0000 pft4 <--- R4lam6 3.7627 0.6079 6.1900 0.0000 pft6 <--- R6lam9 2.9466 0.4429 6.6522 0.0000 pft9 <--- R9lam21 310.2765 5.6206 55.2035 0.0000 MAP <--- precipitationlam22 0.4877 0.0309 15.7995 0.0000 MAT <--- temperaturelam32 0.0458 0.0007 66.9771 0.0000 soil_N <--- soilslam42 6.1736 0.2098 29.4299 0.0000 WD_consumed <--- burnpsi30 0.2622 0.0416 6.3096 0.0000 soils <--> soilspsi40 1.2299 0.1050 11.7118 0.0000 burn <--> burnpsi1 0.1424 0.0122 11.6624 0.0000 R1 <--> R1psi2 1.9068 0.1990 9.5800 0.0000 R2 <--> R2psi4 0.2007 0.0185 10.8678 0.0000 R4 <--> R4psi6 1.0325 0.1836 5.6223 0.0000 R6 <--> R6psi9 0.9629 0.1837 5.2409 0.0000 R9 <--> R9the11 127.8433 6.8021 18.7946 0.0000 pft1 <--> pft1the21 0.0887 0.0099 8.9163 0.0000 pft1r <--> pft1rthe12 925.3996 73.5834 12.5762 0.0000 pft2 <--> pft2the22 6.1126 0.3202 19.0876 0.0000 pft2r <--> pft2rthe14 95.4717 4.7386 20.1476 0.0000 pft4 <--> pft4the24 0.0138 0.0160 0.8599 0.3898 pft4r <--> pft4rthe16 111.4872 5.2660 21.1713 0.0000 pft6 <--> pft6the26 -0.5152 0.1796 -2.8691 0.0041 pft6r <--> pft6rthe19 149.2573 6.8615 21.7528 0.0000 pft9 <--> pft9the29 -0.7591 0.1787 -4.2472 0.0000 pft9r <--> pft9rthe31 -0.0291 0.0134 -2.1770 0.0295 soil_C <--> soil_Cthe32 0.0006 0.0001 6.2580 0.0000 soil_N <--> soil_Nthe41 24.7118 1.0153 24.3387 0.0000 LFH_consumed <--> LFH_consumedthe42 27.6836 2.2583 12.2589 0.0000 WD_consumed <--> WD_consumedgam10 0.4944 0.0342 14.4709 0.0000 temperature <--- precipitationgam11 0.0586 0.0239 2.4480 0.0144 R1 <--- precipitationgam16 0.1403 0.0293 4.7821 0.0000 R6 <--- precipitationbet10 1.9685 0.0680 28.9657 0.0000 soils <--- temperaturebet21 -0.3435 0.0258 -13.2888 0.0000 R1 <--- temperaturebet22 1.6739 0.1143 14.6481 0.0000 R2 <--- temperaturebet24 -0.1051 0.0213 -4.9268 0.0000 R4 <--- temperaturebet29 -0.3320 0.0225 -14.7364 0.0000 R9 <--- temperaturebet60 -0.8213 0.0589 -13.9543 0.0000 soils <--- burnbet61 0.2270 0.0195 11.6321 0.0000 R1 <--- burnbet62 -1.0324 0.0765 -13.5009 0.0000 R2 <--- burnbet69 -0.5459 0.0223 -24.4695 0.0000 R9 <--- burnbet76 -0.2295 0.0163 -14.0410 0.0000 R6 <--- soilsbet82 0.6378 0.1891 3.3730 0.0007 R2 <--- R1bet84 0.3094 0.0349 8.8721 0.0000 R4 <--- R1bet86 0.6415 0.0582 11.0281 0.0000 R6 <--- R1Path112 Appendix H continued…   Parameter Estimate SE z value Pthe101 -0.9789 0.0532 -18.3844 0.0000 soil_C <--- temperaturethe102 -3.2489 0.1455 -22.3230 0.0000 LFH_consumed <--- temperaturethe103 -0.0019 0.0001 -17.5398 0.0000 MAT <--- MAPthe104 0.0156 0.0011 13.6946 0.0000 R1 <--- pft9the105 0.0206 0.0015 13.9799 0.0000 pft1 <--- MAPthe106 0.2625 0.0370 7.0955 0.0000 pft1 <--- WD_consumedthe107 0.1225 0.0199 6.1492 0.0000 pft6 <--- pft1the108 0.8251 0.1251 6.5927 0.0000 LFH_consumed <--- MATthe109 0.0207 0.0060 3.4555 0.0005 R1 <--- pft2rthe110 -0.0533 0.0116 -4.6044 0.0000 pft1r <--- MATthe111 0.0213 0.0018 11.5414 0.0000 R4 <--- WD_consumedthe112 -0.0250 0.0039 -6.4105 0.0000 R6 <--- LFH_consumedthe113 -1.3566 0.2976 -4.5587 0.0000 pft9 <--- MATthe114 -0.0411 0.0079 -5.2309 0.0000 soils <--- WD_consumedthe115 0.0092 0.0025 3.6485 0.0003 R4 <--- LFH_consumedthe116 -0.8595 0.2335 -3.6808 0.0002 pft6 <--- MATthe117 0.0193 0.0063 3.0766 0.0021 pft4 <--- pft2the118 0.1556 0.0558 2.7875 0.0053 pft2r <--- R6the201 0.3026 0.0297 10.1855 0.0000 pft2r <--> pft4rthe202 -4.7413 0.5050 -9.3894 0.0000 pft9 <--> burnthe203 12.2627 1.6570 7.4007 0.0000 pft2 <--> burnthe204 -0.0167 0.0022 -7.5309 0.0000 temperature <--> soil_Nthe205 4.4156 0.4414 10.0044 0.0000 pft4 <--> precipitationthe206 0.5724 0.1080 5.3012 0.0000 pft6r <--> WD_consumedthe207 3.1549 0.4340 7.2700 0.0000 pft2r <--> WD_consumedthe208 -0.3098 0.0552 -5.6122 0.0000 MAT <--> R2the209 0.0310 0.0051 6.0592 0.0000 pft6 <--> soil_Nthe210 2.4968 0.3817 6.5414 0.0000 pft2r <--> LFH_consumedthe211 0.1202 0.0183 6.5804 0.0000 pft4r <--> precipitationthe212 1.6059 0.3300 4.8663 0.0000 pft6 <--> burnthe213 0.2101 0.0698 3.0083 0.0026 LFH_consumed <--> R1the214 -1.3154 0.2934 -4.4826 0.0000 WD_consumed <--> burnthe215 -0.9029 0.3080 -2.9318 0.0034 pft9 <--> precipitationPath

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0167539/manifest

Comment

Related Items