Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of forest roads on the growth of adjacent lodgepole pine trees in the Williams Lake area of B.C. Bowering, Michael Scott 2004

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

Item Metadata

Download

Media
831-ubc_2005-0013.pdf [ 8.76MB ]
Metadata
JSON: 831-1.0074998.json
JSON-LD: 831-1.0074998-ld.json
RDF/XML (Pretty): 831-1.0074998-rdf.xml
RDF/JSON: 831-1.0074998-rdf.json
Turtle: 831-1.0074998-turtle.txt
N-Triples: 831-1.0074998-rdf-ntriples.txt
Original Record: 831-1.0074998-source.json
Full Text
831-1.0074998-fulltext.txt
Citation
831-1.0074998.ris

Full Text

E F F E C T S OF F O R E S T ROADS ON THE G R O W T H O F A D J A C E N T L O D G E P O L E PINE T R E E S IN THE WILLIAMS LAKE A R E A O F B.C. by MICHAEL S C O T T BOWERING B.A., The University of Victoria, 1991 B.S.F., The University of British Columbia, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE D E G R E E OF MASTER OF F O R E S T R Y in THE FACULTY OF G R A D U A T E STUDIES (FORESTRY) THE UNIVERSITY OF BRITISH COLUMBIA November 2004 © Michael Scott Bowering,2004 A B S T R A C T This study examines the effects of roads on the growth of adjacent lodgepole pine (Pinus contorta Dougl. var. latifolia Englm.) trees near Williams Lake, British Columbia. Three questions were addressed in this research: 1) Do roads result in increased diameter or height growth along the road right-of-way edge? 2) If increases occur, how far do they extend from the edge? 3) What are the stand level impacts of any increase in growth? Forty-four plots, ranging from 0.04 to 0.24 ha in size were established and measured in 1999. Plot ages ranged from 37 to 121 years, road widths from 5.3 to 38.5 m, and basal area densities from 14 to 45 m 2 ha" 1. Beginning at the edge, five zones were established at 0 to 5 m, 5 to 10 m, 10 to 20 m, 20 to 30 m, and 30 to 40 m from the edge. Zone 5, at 30 to 40 m from the edge, was used as a baseline for comparison. On average, a 31.3% (using a 95% C.I.; 14.2% to 48.4%) increase in basal area per ha relative to the baseline in Zone 5, occurred within the first 5 m. Mortality was found to be lowest at the edge. As Zone 1 was 5 m wide, this translated into 1.56 m (or 13.6%) of the average 23.4 m road width (using a 95% C.I.; 0.71 to 2.42 m) that is recovered from the increase in basal area for an average stand. If both sides of the road were similarly impacted, then an equivalent of 3.12 m of growing space would be recovered. These results are similar to those by other researchers. Differences in the Zone 1 basal area per ha values relative to Zone 5 were significantly correlated to plot basal area per ha only. Other variables such as road width and edge aspect were not significantly correlated with the relative basal area per ha in Zone 1. The growth rates of trees in the 5 m edge zone, scaled for the 5 years prior to road establishment, differed significantly from growth rates in other zones. The growth rate of trees in the first 5 m increased by 32.1% on average, for the 3 to 15 years after the road opening was established. Significant differences were not found in the bole ellipticity, relative to road orientation. No corresponding increase in average basal area per tree or in average height was noted. The stands sampled in this study were comprised mostly of lodgepole pine, were relatively free of selective cutting and were away from other edges or openings. The results should, therefore, be applicable only to lodgepole pine stands and caution should be used in applying these results to stands with management treatments other than the road, particularly with the cutting of stems at the road edge. Key words: Road edge effects, linear openings, growth and yield, basal area increment, bole ellipticity, edge distance, recovered growing space, Pinus contorta Dougl. var. latifolia Englm., Williams Lake British Columbia. ii T A B L E OF C O N T E N T S ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES vi LIST OF FIGURES vii ACKNOWLEDGEMENTS viii 1 INTRODUCTION 1 2 LITERATURE REVIEW 3 2.1 INTRODUCTION 3 2.2 L O D G E P O L E PINE SILVICS 3 2.2.1 Light and Temperature 4 2.2.2 Moisture 4 2.2.3 Regeneration 5 2.2.4 Damaging Agents to Lodgepole Pine 5 2.2.5 Soils and Roots 6 2.3 DEFINITION OF A F O R E S T E D G E 8 2.4 C H A N G E S IN MICROCLIMATE AND IMPACTS ON VEGETATION 9 2.5 E D G E E F F E C T S ON ROOTS, SEEDLING ESTABLISHMENT, S T E M FORM AND W O O D QUALITY 11 2.5.1 Roots 11 2.5.2 Seedling Establishment 13 2.5.3 Tree Form and Wood Quality 13 2.6 IMPACTS ON T R E E S BY DISTURBANCE T Y P E 15 2.6.1 Impacts of Seismic and Hydro Lines 15 2.6.2 Clearcut Edges 17 2.6.3 Strip Thinning and Thinning Extraction Roads 18 2.6.4 Built Roads 21 2.7 E F F E C T S O F MORTALITY ON STAND YIELD 24 2.7.1 Mortality 2.7.2 Yield 2.8 L ITERATURE REVIEW OVERVIEW 3 METHODS 3.1 INTRODUCTION 3.2 STUDY SITE DESCRIPTION 3.2.1 Sub-Boreal Spruce (SBS) Zone 3.2.2 Sub-Boreal Pine-Spruce (SBPS) Zone 3.2.3 Interior Douglas-fir (IDF) Zone 3.3 FIELD SAMPLING 3.3.1 Map Database 3.3.2 Selection of Polygons for Sampling 3.3.3 Establishing Plots Within Polygons 3.3.4 Plot Measurements Collected 3.3.5 Tree Measurements Collected 3.3.6 Radial Increment Measurements 3.3.7 Determination of Road Age 3.4 PRELIMINARY DATA ANALYSIS 3.4.1 Plot Data 3.4.2 Radial Increment and Tree Basal Area Increment Data 3.5 STATISTICAL ANALYSIS 4 RESULTS 4.1 PLOT DATA 4.1.1 Descriptive Statistics 4.1.2 Zone Level Variables Versus Plot Level Variables by Distance From the Road Edge 50 4.1.2.1 Stems per Hectare 4.1.2.2 Basal Area per Hectare and Relative Basal Area per Hectare 4.1.2.3 Curtis' Relative Density and Relative Curtis' Relative Density. 4.1.2.4 Relative Mean Height 4.1.2.5 Mean Live Crown Ratio and Relative Mean Live Crown Ratio 4.1.3 Mortality 4.1.4 Damage Indicators 4.1.4.1 Mistletoe Damage 4.1.4.2 Scar Damage 4.1.4.3 Mountain Pine Beetle 4.1.4.4 Forks or Crook Damage 4.1.4.5 Dead or Broken Top Damage 4.1.4.6 Frost Crack Damage 4.1.5 Canopy Layers iv 4.2 T R E E B A S A L A R E A INCREMENT DATA 65 4.2.1 Descriptive Statistics 65 4.2.2 Testing for Differences Among Zones 69 4.2.3 Testing for Differences in Bole Basal Area Growth due to Road Orientation 70 4.2.4 Testing for Differences in Mean Tree Basal Area Among Zones 72 4.3 PREDICTING RELATIVE BASAL A R E A P E R H E C T A R E AND RELATIVE T R E E B A S A L A R E A PERIODIC ANNUAL INCREMENT IN Z O N E 1 73 5 DISCUSSION 76 5.1 STAND LEVEL ATTRIBUTES 76 5.1.1 Basal Area Increase at the Edge 76 5.1.1.1 Mortality at the Edge 78 5.1.1.2 Ingrowth at the Edge 79 5.1.2 Height 80 5.1.3 Damage 80 5.2 T R E E BASAL A R E A G R O W T H R E S P O N S E 82 5.2.1 Relative Basal Area Increment 82 5.2.2 Tree Basal Area 84 5.3 C R O W N L E N G T H AND S T E M F O R M 86 5.4 BOLE ELLIPTICITY 87 5.5 YIELD R E C O V E R Y FOR LOST GROWING S P A C E 88 5.5.1 Application of the Yield Adjustment 88 5.5.2 Application of the Growth Response 90 5.5.3 Growth and Yield Applications for the Forest Manager 91 6 CONCLUSION 93 LITERATURE CITED 96 APPENDICES 104 APPENDIX I: Tally cards 105 APPENDIX II: GLMs to test for differences in zonal variables 106 APPENDIX III: Graphs of tree basal area annual increment 135 APPENDIX IV: GLMs to test for differences in tree BA p.a.i. among zones 158 APPENDIX V: Regression to predict Zone 1 tree BA p.a.i 175 APPENDIX VI: GLMs to test for differences in tree BA p.a.i. by core direction.. 181 APPENDIX VII: Regression to predict Zone 1 BA per ha 201 APPENDIX VIII: GLMs to test for differences in tree BA among zones 206 v LIST OF TABLES Table 1: Plot level attributes, based on 44 sample plots 46 Table 2: Zone level attributes for pine trees in the 44 sample plots 48 Table 3: Zone level attributes relative to Zone 5 values for pine trees in the 44 sample plots 49 Table 4: Correlations between zone variables and plot variables. Plot level variables are based on all live trees, whereas zonal variables are based on live pine trees only 50 Table 5: Zone level proportions (by %) of dead standing trees, stumps and live trees of all species based on the 44 sample plots 57 Table 6: Proportions (by %) of zone basal area and stems per hectare for dead standing, stumps and all live trees for three road age classes 58 Table 7: Presence of damage indicators as a percent of all live tree stand basal area across the 44 sample plots 59 Table 8: Damage free lodgepole pine percent basal area and percent stems per hectare, within each zone 60 Table 9: Damage agent proportions, by zone, shown as a percent of lodgepole pine basal area for each damage agent by three stand age classes 61 Table 10: Proportions (by %) of zone lodgepole pine basal area by tree canopy layer, for three road age classes 64 Table 11: Correlations for annual tree basal area increment relative to the five years prior to the road, for 1 to 2 years, 3 to 5 years, and subsequent five-year intervals after the road right-of-way was harvested 66 Table 12: Mean tree basal area periodic annual increment by zone for the 5-year period prior to road establishment 67 Table 13: Relative tree basal area periodic annual increment by zone and period 68 Table 14: Relative tree basal area periodic annual increment for the two radial increment core directions 71 vi LIST OF F IGURES Figure 1: Distribution of the 44 sample plots within the Lingum Ltd. Innovative Forest Practices Agreement area boundaries 31 Figure 2: Plot subdivided into five zones from the road right-of-way edge 38 Figure 3: Pine stems per ha versus plot stems per ha, all species, by distance from the stand/road edge 51 Figure 4: Pine stems per ha versus distance from the stand/road edge 52 Figure 5: Pine basal area per ha versus distance from the stand/road edge 52 Figure 6: Relative density index versus distance from the stand/road edge 53 Figure 7: Mean relative tree basal area periodic annual increment ratio, by zone, for lodgepole pine 69 Figure 8: Mean relative tree basal area periodic increment, by zone and core direction for the 20 years after road establishment. Cores were sampled parallel (W) and perpendicular (X) to the road 70 vii A C K N O W L E D G E M E N T S I am grateful for the funding from the research grant provided by Lignum Ltd., by the National Science and Engineering Research Council, and for the graduate teaching assistantships provided by the UBC Faculty of Forestry. In addition, the encouragement on the part of my employers at the Malcolm Knapp Research Forest and at D.R. Systems Inc. and their flexibility with my work schedule has allowed me to complete this thesis. The very kind assistance from the staff at the Alex Fraser Research Forest was invaluable for completing the field sampling safely, comfortably, and within the allotted time frame. Lignum's employees were most helpful in compiling digital inventory data for the field sampling despite other priorities in their workloads, and Bill Bourgeois was instrumental in initiating this project from the start. The diligence and care taken by the research assistants, Shayne Boelk, Loreen Stevenson and Jodie Krakowski during field sampling and increment core measurements gave me confidence in the accuracy of the data. A special note of thanks goes to my father Lome Bowering for selflessly milling over 1500 wooden mounts for the increment cores, only for the good of the project. Valerie LeMay, as my thesis committee chair, has been extremely helpful in all phases of this research by providing clarity, wisdom and a sound common sense approach to the project's complexity. Committee members Ken Day, Gordon Nigh and Peter Marshall have provided valuable feedback on this research throughout its duration, and their advice has made both the research and this thesis a much better product. Valerie and Peter's unbridled enthusiasm for assisting with the field sampling, undoubtedly improved both the quantity and quality of the data collected. Last but not least, I must express gratitude to my friends and family, in particular Nancy, for their ongoing encouragement, patience and moral support. M.B. November 2004 viii 1 INTRODUCTION Roads and harvesting access corridors are necessary features on a managed forest landscape for timber extraction. A road or corridor creates an edge or a linear gap in the stand structure. Much research has been completed on the effects of road and corridor edges on low lying vegetation, animal habitat, and the ecological community in general (see Kremsater, 1997, Vol lerand Harrison, 1998 and Hourdequin, 2000 for reviews of this literature). However, much less recent research is available concerning the effects of roads on the trees growing within the stands contiguous to the road edge. There are two areas of interest to the land manager, with respect to the yield impact of road right-of-ways: 1) the timber harvested at the time of right-of-way clearing; and 2) the impact that the right-of-way has on the growth of contiguous stands. While the yield from the harvest right-of-way timber is reasonably easy to measure via the scaling of logs or to estimate from an inventory cruise, the question remains as to how new right-of-ways impact the growth in the residual stand (Bella, 1986). Many factors influence tree stem growth along roadsides, including site characteristics, such as slope, aspect, elevation, moisture, soil properties, and microclimatic factors, and stand characteristics, such as species composition, stand age, live crown proportion, crown cover, stocking, and overall health. Construction of roads may alter soil properties, hydrologic regimes and result in physical damage to residual trees. By removing the right-of-way trees, plant competition for light and other resources at the road edge is initially reduced. Light is a critical resource for tree growth (Wales, 1972; Muth and Bazzaz, 2002), and has been argued by some as the most important factor influencing tree growth in temperate and boreal forests (Coates and Burton, 1999). The decrease in competition among trees along the edge may result in larger crowns, increased diameter growth, and, thereby, increased volume (Oliver and Larson, 1996 p. 320). However, any changes in tree growth and yield will likely vary with 1 different site and stand attributes, and the characteristics of the road itself. In developing stands, stand density may become higher at the road edge through seedling establishment and/or reduced mortality whereas in existing stands, overstory edge trees may be prone to sunscald and wind damage (Oliver and Larson, 1996 p.322). This research was implemented to provide insight into the edge effects of forest roads on the growth and yield of road edge stands, within the boundaries of the Lignum Limited's (Lignum) Innovative Forest Practices Agreement (IFPA) area, near Williams Lake, British Columbia. It was determined from findings of the literature review presented in Chapter 2, and in consultation with Lignum and the British Columbia Ministry of Forests (MoF), that the study would focus on lodgepole pine (Pinus contorta Dougl. var. latifolia Engelm.). Although Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) stands were also of interest, only lodgepole pine stands were addressed in this study. Lodgepole pine is a primary succession species, grown naturally in even age stands, whose natural disturbance regime is usually fire (Koch, 1996). The study sought to evaluate any changes in diameter and height for the stands adjacent to the road, the extent of these changes in terms of the distance from the road and the duration of any diameter increment if it occurred. Finally, the study ascertained the stand level impacts of any changes in tree growth attributed to right-of-way openings of forest roads, and developed an appropriate growth and yield model to be applied to stands within the range of the model. The remainder of this thesis is divided into following format: Chapter 2 discusses the relevant background literature pertaining to forest edges, primarily edges from linear openings, and provides background information about the silvics of lodgepole pine. Chapter 3 describes the study site and then details the methodology of the field data collection and subsequent analysis. Chapter 4 reveals the results of the plot attribute and tree growth data, while Chapter 5 covers a discussion of these findings. Chapter 6 provides a summary of the findings. 2 2 L ITERATURE REVIEW 2.1 INTRODUCTION Prior to establishing a field research study, an overview of the relevant literature on the effects of roads on adjacent trees and stands was compiled. Emphasis was placed on commercial tree species similar to those found near Williams Lake, British Columbia (B.C.). Of particular interest were species with similar successional growth patterns and edaphic traits to lodgepole pine and Douglas-fir. Information was gathered from baseline searches using electronic databases containing available literature on the relevant themes surrounding this topic, and by contacting researchers directly. Some literature was translated from the source language (e.g., Finnish), if the research appeared to be particularly related to road impacts in similar sites, and a translator was available. However, the search did not produce a large quantity of directly applicable literature. Chen era/. (1992) stated that "Literature concerning edge effects on tree growth does not exist". Since the impacts of roads are similar, to some extent, to the impacts of other linear openings such as row thinning and oil and gas exploration lines (seismic lines), any literature on these other linear openings was also included. The literature review begins with an overview of the silvics for the species of interest, lodgepole pine. A review of literature on the definition of a forest edge is presented afterwords. Following this, a brief review of literature on edge effects on microclimate is presented. Reviews of literature for roots and regeneration, tree growth, and finally, stand growth are then presented. A summary of the information that is considered most relevant for the Williams Lake study area is presented last. 2.2 L O D G E P O L E PINE SILVICS Two varieties of lodgepole pine exist in B.C. The coastal form, often referred to as shore pine (Pinus contorta var. contorta) is found along the shoreline of the Pacific 3 coast, ranging in elevation from sea level to 610 m (Lotan and Critchfield, 1990). The inland variety, referred to as lodgepole pine (P. contorta var. latifolia) is found from 490 to 3660 m in elevation, and is the variety of interest for this study. Lodgepole pine has a wide ecological amplitude, growing from 31 °N to 64°N latitude, and from the Pacific Ocean to 105° West (Lotan and Critchfield, 1990). 2.2.1 Light and Temperature Lodgepole pine is a very shade intolerant species (Klinka era/ . 1990). As expected with shade intolerant species, lodgepole pine grows best in even-aged stands with access to full sunlight and little influence from overstory trees (Lotan and Critchfield, 1990). Photosynthesis, and therefore plant terminal growth, is related to the length of the photoperiod, temperature, and to the degree of light intensity (Oke, 1987; Stathers era/ . 1990). At light levels of 40 percent, to less than 10 percent of full sunlight, lodgepole pine vigour is significantly reduced (Koch, 1996). The rate of photosynthesis increases with temperature from dormancy to the growing season, with the maximum rate of photosynthesis occurring at a daytime temperature of 21 °C (Koch, 1996). Lodgepole pine is a relatively tolerant species with respect to temperature, provided there are no extreme warm temperatures during dormancy. The species survives in temperatures ranging from -80°C up to over 46°C in the interior of B.C. (Koch, 1996). Lodgepole pine seedlings are also relatively resistant to frost injury compared to other coniferous species, however seedlings are susceptible to frost death a t - 8 °C during the growing season (Koch, 1996 and Lotan and Critchfield, 1990). 2.2.2 Moisture Snowmelt provides much of the water during the rapid growth period of early summer (Lotan and Critchfield, 1990). In the interior, lodgpole pine will grow on areas receiving only 250 mm of mean annual preciptation (Lotan and Critchfield, 1990). The species performs best when the water content of the soils is between 12 and 20 percent of volume, but not exceeding 34 percent (Koch, 1996,). Lodgepole pine tend 4 to be more tolerant of drought than most other species as its water use efficiency is quite high; however drought tends to be a common cause of mortality for first year seedlings (Koch, 1996; Lotan and Critchfield, 1990). 2.2.3 Regeneration Lodgpole pine generally has serotinous cones and regenerates naturally after fire disturbances, warm temperature, or scarification to mineral soils, if there is an abundant seed source (Klinka era/. 1990; Weetman and Vyse, 1990, Lotan and Critchfield, 1990). The species is quite prolific and the seeds within serotinous cones will survive for decades, resulting in very high densities at time of stand establishment (Lotan and Critchfield, 1990). Lodgepole pine seeds will also germinate in very wet soils, however the plant productivity is limited (Dumanski era/., 1973). Lodgepole pine seedlings are tolerant to flooding primarily due to the presence of large gas filled cavities in the roots, where the oxygen is actively transported when the soil is saturated (Coutts and Philipson, 1978; Bassman, 1984). Advanced regeneration will respond to release in dry ecological conditions, however the risk of mistletoe {Arceuthobium americanum Nutt. ex Engelm.) infection is a significant concern (Klinka et al., 1990). Relative to other conifers, lodgepole pine shows the best regeneration response on moisture-deficient, nutrient-poor and frost-prone sites as well as on infertile soils, as it is often the only species that will grow on these sites (Klinka etal., 1990; Lotan and Critchfield, 1990). 2.2.4 Damaging Agents to Lodgepole Pine Lodgepole pine is subject to a number of damaging agents. Fire tends to be the most powerful catastrophic event that affects natural lodgepole pine forests, as the wildfire intervals of 100 years or less are not uncommon (Koch, 1996). The species has a symbiotic relationship with fire at low severity levels, as the serotinous cones require heat or full sunlight for seed release (Lotan and Critchfield, 1990). Lodgepole pine is subject to a number of insect pests that attack cones, seedlings and mature trees. Lodgepole pine terminal weevil {Pissodes terminalis 5 Hopping) is probably the most important insect affecting young stands, while mountain pine beetle {Dendroctonus ponderosae Hopkins) the most significant in older trees (Furniss and Carolin, 1977; Lotan and Critchfield, 1990; Koch, 1996). Lodgepole pine is subject to damage from the root rots Armillari aostoyae, Inonotus tomentosus, Leptographium wageneri and Heterobasidion annosum among others (Allen et al., 1996; Koch, 1996). Lodgepole pine is also subject to stem cankers from Atropellis piniphella, and stem rusts such as Endocronartium harknessii and a number of blister rusts in the Cronartium genus (Allen era/. , 1996; Koch, 1996). Animal damage is also common, primarily in the form of seed predation by birds and small mammals, damage to the cambium from voles, hares, porcupines and bears, seedling damage from the snowshoe hare (Lepus americanus), and trampling by domestic livestock (Koch, 1996). Lodgepole pine is also subject to stem decay, needle casts and needle blights, as well as miscellaneous diseases such as snow mould (Neopeckia coulteri) and Western pine-aster rust (Coleosporium asterum) (Allen era/. , 1996; Koch, 1996). The large broom like features characteristic of the parasitic lodgepole pine dwarf mistletoe {Arceuthobium americanum Nutt. ex Engelm.) is one of the most widespread and damaging agents to affect lodgepole pine trees (Koch, 1996; Allen era/. , 1996). 2.2.5 Soils and Roots Lodgepole pine prefers well-drained, non-compacted soils without restrictive layers, however the species will grow on a wide range of soil structures and on sites with low nutrient availability (Koch, 1996). Lodgepole pine is also found in wet poorly drained soils, in very infertile soils, on shallow dry soils, and in flat areas with cold soils which are subject to radiation frosts (Koch, 1996; Cochran, 1984). The species grows naturally in soils composed of sand, gravel and volcanic pumices with low organic content, but it is also found in wet meadows and peat soils with high organic content (Cochran, 1984). Lodgepole pine often grows in soils that are not optimal for other conifers, primarily due to limited nutrient and water availability. Weetman et al. (1985) found that 6 lodgepole pine will establish and grow on sites with extremely low nutrient availability, as the species has modest nutrient demands, but it will respond quite positively to fertilization if water is not a limiting factor. Soil pH is often a limiting factor, for soils with a pH near 8 in the upper 50cm of the horizon tend to exclude lodgepole pine (Lotan and Perry, 1983). Soils with a pH of 5.5 to 6 are usually most productive for lodgepole pine (Dumanski etal., 1973; Koch, 1996). The productivity of sites is relative to the ability of seeds to germinate and become established seedlings in the soil, the ability of the soil to supply water and nutrients, the degree of soil aeration and the ability of roots to grow in the soil (Cochran, 1984). Root development of lodgepole pine is highly variable, and is somewhat dependent upon soils. Juvenile lodgepole pine trees tend to have a large taproot, which grows downward until a hardpan or water table is encountered, where the taproot may die off or bend, forming the characteristic 'J-root' (Eis, 1970; Lotan and Critchfield, 1990). As the juvenile trees mature, the tap-root becomes less significant, as the trees develop sinker roots near the base of lateral roots (Eis, 1970; Coutts and Nicoll, 1991). The sinker roots (small diameter branches of lateral roots descending vertically) help to provide lateral support to the trees. Lateral roots themselves are often shallow, penetrating deeper into the soil with increasing distance from the rootstock (Eis, 1970). Without establishing sinker roots from the lateral roots, the trees may be more susceptible to windthrow. Furthermore, restrictions in the soil horizons may reduce the growth rates, as the roots have poorer access to water and nutrients. The nature of the rooting characteristics of lodgepole pine leave the trees somewhat susceptible to reduced growth from soil compaction due to ground based harvesting when the soils are wet and subject to compaction (Cochran, 1984). Lodgepole pine is productive across a wide amplitude of different terrain types. The species grows on rough and rocky terrain and on steep slopes, but it grows well on gentle slopes and basins, much like the rolling topography with few high peaks, as found in the study site near Williams Lake, B.C (Lotan and Critchfield, 1990; Valentine and Dawson, 1986) 7 2.3 DEFINITION OF A F O R E S T E D G E The definition of 'edge' varies in the literature. How an edge is defined depends upon the variables of interest. Since the transition between a stand and a linear disturbance is fairly definitive, the edge may be where the tree line starts. However, instead of a thin line, the edge may be better defined as the transition zone between the forested area and non-forested area (Luken etal., 1991). This is similar to the concept of a 'transition width' described by Wales (1972), the 'edge to interior gradient' analyzed by Palik and Murphy (1990), and the 'depth of influence' described by Chen et al. (1992). The transition width edge can be confused with the term 'edge effect'. Leopold (1933) originally defined edge effect as the tendency for boundaries between two habitat types to support a greater variety of species and number of individuals than either adjacent habitat. Edge effect could be defined as the alteration of environmental conditions by the presence of a boundary between a forest and a non-forested area (Cadenasso etal., 1997). In terms of microclimate, edge effect could be defined as the differences in both microclimate and vegetation composition and structure that exist between forest margins and forest interiors (Young and Mitchell, 1994). Bradshaw (1992) argued that there is not one edge effect, but many, and that the magnitudes and significances change with stand characteristics and silviculture treatments. In terms of defining the transition width edge, Chen et al. (1992) suggested that the edge should be defined at the point where a given variable returns to 2/3 the value for the interior forest level. Minko and Hepworth (1990) defined the edge as the point where the branches of the edge tree contributes substantially to forming the gap body. Some studies have assumed that edges are where the tree line interfaces the gap (e.g., Chen etal., 1992; Bella, 1986). Cienciala etal. (2002) defined the edge as a zone being the distance of mean tree height plus one standard deviation from the tree line interface. Others have chosen to be less definitive using a feature such as the area under influence of the crown or root structure (e.g., Isomaki, 1986; Pfister, 1969), even though the gap structure itself (such as a road clearing width) may be easily viewed. Edges which are difficult to measure accurately, such as the limit of the crown 8 drip line, may induce the sampler to use the anthropologic definition of edge, such as the top of a road cut or fill slope (e.g., Pfister, 1969). Since the concentration in this literature review is on linear openings, the edge is considered to be at the tree line. However, the transition zone from this edge to the interior forest is also of interest. 2.4 C H A N G E S IN MICROCLIMATE AND IMPACTS ON VEGETATION Changes in the microclimate from a gap edge to the interior of the forest have been measured, including air and soil temperature, relative humidity, soil moisture, wind speed, and light availability. Many of the microclimatic parameters are influenced by local weather conditions. A sample of the research on changes in microclimate along the gradient from the edge to the interior forest and the impacts on tree growth is presented in this section. In old growth Douglas-fir stands in central Oregon and southern Washington, Chen and Franklin (1990) found that wind speed, relative humidity, and air temperature are very sensitive to 1) distance from a clearcut edge, 2) macroclimate changes, and 3) edge exposure. On south aspects, they found that edge influences on microclimate may extend into the forest for more than 240 m in extreme conditions on windy days, but only a few metres on cool days. Soil temperature and moisture changes were influenced in the zone from 60 to 120 m from the edge. They also found a higher edge influence on the west and east edges, than on the north and south edges. Many other studies examined the effects of aspect on the microclimate changes from the gap to the interior forests including Matlack (1993), Wales (1972), Palik and Murphy (1990), and Luken etal. (1991). Results vary among the studies. Chen era/. (1995) showed that clearcut edge effects on microclimate variables were strongest for southwest facing edges and weakest for northeast facing edges, for old growth Douglas fir stands in Washington and Oregon. They argued that this effect is logical since southwest facing edges receive direct solar radiation in the early afternoon, and this coincides with the maximum local daily temperature and moisture (Chen etal. 1995, and Oke, 1987, for discussion of diurnal temperature and water vapour). In mixed 9 species hardwood stands in northern Kentucky, Luken etal. (1991) noted that the depth of edge influence can be attributed to aspect, as aspect itself may create microclimate changes that extend more than 10 m into adjacent forests. Chen et al. (1995) found that soil moisture tends to increase along the gradient from a clearcut edge to the interior, as does relative humidity. In contrast, Burke and Nol (1998) found no change in relative humidity and air temperature along a gradient from a field to a deciduous forest. Generally, it is expected that air temperature along the edge to interior gradient will decrease during the day, but increase at night, with soil temperature following a similar pattern. Chen era/. (1995) confirmed this expectation, although the oscillations for soil temperature were not as wide as for air temperature. Temperature and moisture gradients at the ground level are influenced by the interception of incoming and outgoing solar radiation by the tall forest canopy (Wales, 1972; Chen etal., 1995). Short wave radiation was found to decrease drastically along the edge to interior gradient (Chen et al., 1995; Chen et al., 1996). Similarly, light intensity was found to decrease rapidly from the edge to the interior gradient (Chen et al., 1996; Burke and Nol, 1998). Macroenvironment factors such as the general conditions of climate and soil have been found to influence the growth characteristics of regeneration (Nyland, 1996). Several researchers have noted the microenvironment factors such as the amount of light, measured as photosynthetically active radiation (PAR), helps to determine the early successional species composition (Young and Mitchell, 1994; Chen et al., 1996). If the light or soil moisture content is altered at the edge, the species that regenerates will depend on the light tolerance of the species if seeds are available (Bradshaw, 1992; Palik and Murphy, 1990; Nyland, 1996). Increased light at the gap edge will likely result in the establishment of less shade tolerant, early successional regeneration (Luken er al., 1991). The amount of available light is expected to influence the physiological response of the forest canopy (Chen and Franklin, 1990; Chen era/., 1992). Several researchers found that the highest canopy density for mixed species of hardwoods was usually found at the gap edge, and decreased towards the interior of the stand (Palik and Murphy, 1990; Luken etal., 1991; Burke and Nol, 1998). Chen etal. (1996) studied the 10 effects of light on moderately shade tolerant Douglas-fir and shade intolerant lodgepole pine near Williams Lake, B. C. Both terminal and lateral growth declined for both species with decreasing light, yet the total leaf area increased for both species. As the new vegetation at the gap edge ages, it may form a vegetative wall which may reduce the degree of edge effects (Burke and Nol, 1998; Matlack, 1993; Wales, 1972; Chen etal., 1992). Also, the age of the gap itself may influence vertical structure of the vegetation at the gap edge. Older gap edges with a more continuous vertical vegetation structure will probably have decreased light and reduced turbidity, therefore reducing the edge effect on the stand. New gaps edges in older stands, such as recently felled road right-of-ways, may have most of the biomass in the upper canopy, with a low number of well-spaced stems. This relatively simple structure may have increased wind speed at great distances from the edge. For an old growth Douglas-fir forest contiguous to a clearcut, Chen et al. (1995) found that a relatively low wind speed at the edge increased the interior wind speed for more than 240 m from the edge. The introduction of linear openings has been shown to influence light, soil moisture, and wind speed. However, the degree of impact varies with the macroclimate, the characteristics of the opening, and the characteristics of the vegetation adjacent to the opening. 2.5 E D G E E F F E C T S ON ROOTS, SEEDLING ESTABLISHMENT, S T E M FORM AND W O O D QUALITY 2.5.1 Roots Kardell and Pettersson (1973) examined root damage resulting from the construction of skid roads in 100-year-old Norway spruce (Picea abies (L.) Karst.) stands. When skid roads were constructed on steep slopes, trees on the uphill side had their roots cut off. For trees within 15 m of the upper side of the skid road, the root damage resulted in a reduction of basal area increment of 15-20% for a period of 20 years. Urban era/. (1994) analyzed ring widths on structural roots and on main trunks of white spruce (Picea glauca (Moench) Voss) edge trees, 16 years after clearing a 11 road right-of-way. The stand was a mixed boreal stand of white spruce, aspen (Populus tremuloides Michx.) and lodgepole pine, 120 years old at the time of sampling. Ten trees in a control area (away from the road) and 10 trees at the edge of the clearing (released trees) were sampled. Three increment cores at breast height were taken from each tree: from the windward, leeward and perpendicular to the wind direction. Also, three to four primary roots were sampled. The roots were elliptical in shape, with lengths of 30 cm and widths of 10 cm. Equations were fit using the pre-release data and used to obtain a predicted ring width series for the roots and for the trunks of the trees for the 16 year post-release period. Annual ring indices were computed by dividing the observed ring widths by the predicted widths; these were then averaged for the 10 control versus the 10 trees at the edge, for both trunk and root increments (four values). A response index for the trunks and for the roots were then produced subtracting the average ring indices of control trees from the average released trees. An allocation index was then calculated for the control and for the trees near the road by subtracting the trunk response from the root response. Finally, the difference in these two allocation indices was calculated (allocation-response index). Both ring indices gave evidence of an increase following the road clearing. The percent increase in roots was more than for the trunk following release (69% versus 31%). The data showed that the roots responded before the trunks did; however the roots also slowed earlier than the trunks. From the allocation index, the control spruce decreased by 21 % per year and the released spruce increased by 17% per year. Urban et al. concluded that the increase was not a gradual rise but an abrupt jump to a new level upon release, that appeared to last for about 15 years. They suggested that the delay in trunk growth supported by their data suggest that increased allocation to roots may account for the observed delay in trunk response to release for three to nine years. While increased tree volume increment may be delayed as the trees are putting more energy into root growth, stand yield may be influenced positively as mortality due to windthrow may decline. Species that are particularly susceptible to windthrow damage may be more stable with increased root mass at the gap edge. Michovski and Ennos (2002) studied the morphology and mechanical stability of suppressed crown Scots pine (Pinus sylvestris L.) root systems growing in clay soils at 12 the University of Manchester in England. The authors investigated the root characteristics on interior and edge trees in a 30 x 20 m stand, and concluded that the asymmetrical nature of root development of edge trees was due to the lack of below ground competition from other trees. The authors found that windthrow was more likely for interior trees, while trees growing at the edge had a slightly greater, but non-significant trend to fail more in their stems than at their roots. 2.5.2 Seedling Establishment Chen etal. (1992) found that the number of Douglas-fir and western hemlock (Tsuga heterophylla (Raf.) Sarg.) seedlings and saplings decreased with distance into the stand from the edge of 10 to 15-year-old clearcuts, but the number of Pacific silver fir (Abies amabilis (Dougl. Ex Loud.)) seedlings increased with distance from the edge. The numbers of smaller seedlings (<30 cm) were affected by distance from the edge, whereas taller seedlings were not. Douglas-fir appeared to be affected for less than 60 m into the forest, whereas western hemlock and Pacific silver fir were affected for less than 120 m; however, these ranges varied by height of the seedling. Luken etal. (1991) determined that greater reproduction of mixed hardwoods was obtained at edges of power line corridors, with a higher mean density and basal area of stems less than 10 cm diameter at breast height (dbh). Species that had low to intermediate shade tolerance increased at the edges, and this effect extended 10 to 15 m inside the residual stand. 2.5.3 Tree Form and Wood Quality Wales (1972) found that trees at the gap edge tend to lean towards light, away from the older overshadowing canopy. This would result in changes to tree form. Muth and Bazzaz (2002) examined the edge tree positions of mature, mixed coniferous and deciduous stands growing near elliptically shaped gaps in the Harvard Forest Long-Term Ecological Research site in central Massachusetts, U.S.A. The authors found that tree canopies were displaced towards gap centres through differential branching, and that canopies had deeper crowns on the open side than on 13 the stand facing side. Early successional trees growing in the sub-canopy generally exhibited a greater magnitude of light foraging than later successional tree species. The two coniferous species in the study, eastern hemlock (Tsuga canadensis (L.) Carriere) and eastern white pine (Pinus strobus L.) had small values of canopy displacement compared to the deciduous species. The authors attribute this to the coniferous trees having excurrent growth and rigid architecture. Harrington and Hendrick (1999) found that the crown areas of slash pine (Pinus elliottii Engelm.) measured in the second year after creating 75 m circular openings, were 22% greater on average for trees growing near the circular gaps than for those in undisturbed portions of the stands. The authors surmised that gap edge trees responded to increased light by increasing their crown areas. Isomaki (1986) began a series of studies on the effects of hydro line corridors on adjacent stands in Southern Finland. He found that tree boles growing at the edge of powerline right-of-way were found to be more elliptical in shape (Isomaki, 1986). These differences were not observed deeper in the stand. In one sample adjacent to a 25 m wide line corridor, the diameters parallel to the line were as much as 3.2% greater than the diameter perpendicular to the line for Norway spruce (Picea abies (L.) Karsten). A similar pattern was found in Scots pine growing adjacent to a 10 m wide line corridor, where the ellipticity was as much as 2.1% greater. The piths were centred within the diameters, indicating that stress on the trees altered the tree shape. Isomaki (1986) noted that wide corridors might increase the wind stress and result in this more elliptical shape. He also noted that the changes in taper, stem form (tree volume/volume of a cylinder), and slenderness (height/diameter at breast height) were relatively stronger and extended further into the adjacent stand for Norway spruce than for Scots pine. Strip thinning roads on Norway spruce and Scots pine stands were also studied by Isomaki (1985 and 1994; Isomaki and Niemisto, 1990). The strips were 4 to 5 m wide and spaced at 30 m intervals. The strips had a similar effect on tree form as line corridors, with a small negative but short-term impact on stem form of Norway spruce. The stem form of edge trees was clearly inferior to the rest of the trees. The tree slenderness values of edge trees were, on average, 0.5% lower than interior trees. Taper increased by 0.8%, but the taper values were on average 2.5% smaller after six 14 to nine years. However, this was due more to operational considerations at time of thinning than the edge effect directly. Trees with larger than average diameters were left as edge trees when the strips were constructed, and these trees had a poorer stem form (larger taper and lower slenderness values) due to their larger base. Kramer (1958) found that the taper was highest at the road edge for Norway spruce in Germany. Edge trees were lower in value due to the increased knots, as well as the increased taper. For European beech (Fagus sylvatica L ) , he found that taper also increased, but the reduction in quality was only noticeable for roads more than 12 m wide. Landbeck (1965) also found that taper increased for 5 m into Scots pine stands adjacent to roads in Germany. The wood quality was poorer for the 5 m nearest the road, and log grades decreased with increasing road width. Berg (1973) looked at the growth and form of radiata pine {Pinus radiata D. Don) in New Zealand. Trees at the stand edge exhibited a response to light from the opening, measured by increased diameter, branch size, and branch formation towards the gap. Berg recommended pruning of the edge trees to yield clear wood. 2.6 IMPACTS ON T R E E S BY DISTURBANCE T Y P E 2.6.1 Impacts of Seismic and Hydro Lines Bella (1986) studied the impacts of 7.3 m wide seismic lines, 10 or more years old, on the growth of adjacent pure and mixed species stands in boreal forests of Alberta. Edge effects on tree growth of lodgepole pine, black spruce (Picea mariana (Mill.) B.S.P.), white spruce, and aspen trees ranging in age from 10 to 100 years of age and growing on a variety of sites were studied. He concluded that the line clearings resulted in a consistent and significant increase in diameter growth for the three coniferous species sampled. In lodgepole pine, the release extended only to a distance from the clearing equal to one half the total tree height. The increase in radial increment declined from 41% for the five to 10-year period after the clearing to 24% for the 16 to 20 year period. White spruce responded to the clearing within five years, with 15 an increase in radial increment of 15 to 20%. Black spruce increased in radial increment by 10% within the first five years and by 20% for the following five-year period. Unlike the coniferous species, aspen showed no consistent response. Bella attributed the diversity of aspen responses to the clonal habits of the species and to difficulties in determining accurate growth increments because of false rings. For lodgepole pine, Bella determined that radial increment decreased with distance from the edge, stand basal area, and stand age at time of clearing and number of years since the clearing. Conversely, the lodgepole pine radial increment increased with mean dominant height over age at the time of survey (similar to site index), diameter at breast height (dbh) at the time of disturbance and relative density, defined from Curtis (1982) as an index of the stand basal area divided the square root of the diameter tree of mean basal area. For all three coniferous species, Bella noted that the most important variables for determining tree growth were the mean dominant height over age at the time of survey, the distance from the edge, the stand age at the time of clearing, and the number of years since the clearing. As well as tree shape changes, Isomaki (1986) examined the growth response of 14 pure species stands in Southern Finland adjacent to hydro line corridors of varying widths. The stands were comprised of Norway spruce, Scots pine, or birch (Betula pendula). Sampling took place on the north aspect of the corridors, at least 10 years after the corridor was established. Isomaki found a clear increase in overall growth for the three stand types over all plots, although in two pine plots the overall growth was not as evident. Height growth for Norway spruce increased at the edge, but this increased height was not evident for Scots pine and birch. Radial growth was most affected, although this effect did not occur until five years after the corridor was cut. The authors attributed this delay in response to the relatively young age (14 years) of the stands. The effect extended for 5 m into the stand interior for Scots pine, but only affected the edge trees for the other two species. This effect lasted for 20 years after the corridor was cut for both Norway spruce and Scots pine. The increased radial increment was larger for Scots pine than for the Norway spruce. For a 22-year-old, 10 m wide hydro corridor, Isomaki (1986) found that the mean volume of trees within 2 m of the edge increased by 27%, while those in the 2 to 5 m zone from the edge 16 increased by 5.3%, compared to the mean volume for trees growing 10 to 20 m from the edge. In contrast, trees growing 5 to 10 m from the edge exhibited a decline in volume growth of 3% compared to the mean volume for trees growing 10 to 20 m from the edge. He attributed the downturn in the 5 to 10 m zone to either the shading from the edge trees, or to normal variations in tree growth. 2.6.2 Clearcut Edges Looking at vegetation responses in an old growth Douglas-fir stand adjacent to a clearcut edge, Chen et al. (1992) found that the growth rates for dominant Douglas-fir and western hemlock trees were higher at the edge than in the interior of the stand. Douglas-fir growth increment was less than that for western hemlock. They calculated a relative growth rate (RGR) by comparing the growth rate of stems for the ten years before the formation of the edge (GR-|0b) to the growth rate for ten years after the formation of the edge ( G R i o a ) , using the following equation: RGR =GR™°~GR™» xlOO [1] G R l 0 b The average R G R for western hemlock was 150% near the edge and showed an exponential decline with distance from the edge. Douglas-fir showed a similar pattern, but the average R G R was only 33%. Variation in R G R for both species was extremely high near the edge, and decreased rapidly within 60 m into the stand. The authors noted that the extreme variability in the R G R s depended not only on levels of light and other resources, but also on other biotic and abiotic factors such as tree position in the stand and current physiological status. The depth of edge influence was arbitrarily defined in Chen etal. (1992) as the point along the edge to interior gradient where the response variable Y returned to a condition representing 2/3 of the interior forest environment found at 240 m from the edge. Using Y 0 and Y 2 4o as the variable of interest at the edge and at 240 m from the edge respectively, Chen et a/.(1992) defined the response equation as: 17 F = y 0 ± | | y 0 - y 2 4 0 | [2] Chen et al. (1992) found the growth rates returned to 2/3 of the original stand within 53 m for Douglas-fir and within 26 m for western hemlock. In contrast, Burton (1999) looked the effects of clearcut edges in the moist and cold Sub-Boreal Spruce (SBS) biogeoclimatic zone of B.C. Based on measurements of 699 increment cores, the mature lodgepole pine trees on the south facing clearcut edges experienced an unexplained 48% decrease in radial growth compared to interior stand levels, for a distance up to 14.5 m from the edge. 2.6.3 Strip Thinning and Thinning Extraction Roads A number of researchers have examined the effects of strip thinning on the growth of edge trees. While comparisons of total stand yield from different types of thinning treatments (i.e., strip thinning versus selection thinning) are outside the scope of this review, the impact of the strip thinned corridors on the adjacent trees is relevant, despite the small corridor widths. Since paths constructed to remove thinned trees also tend to be narrow, the impacts on residual tree growth from thinning extraction roads (strip roads) are also presented in this section. The widths of these strip roads are difficult to determine, since the target or nominal width may differ from the actual width. Isomaki (1994) discusses several measures of widths. For this paper, nominal widths are given for all studies. Bucht (1977) examined the impacts of 7 to 11-year-old strip roads on 13 stands of Scots pine in Central and Northern Sweden, and compared the growth to selectively thinned stands. Using increment cores and height increment measures, he found that the diameter, basal area and volume response extended 4 to 6 m into the residual stand. Relative increments for the 7 to 11 years after thinning were compared to the 10 years prior to thinning. At 3 m from the road, diameter increment was 5% greater, basal area increment was 27% greater and volume increment was 42% greater than 18 the 10-year period prior to cutting the strip roads. Relative increments for diameter, basal area and volume declined after 3 m from the strip road. Bucht and Elfving (1977) looked at growth increment of Scots pine trees growing along an east-west oriented, 4 m wide thinning strip in Southeast Lapland. The site was considered to be a highly productive site, with a site index of 24 m at 100 years. The residual stand was 16 m wide between the strips. Increment was recorded by measuring shoot growth on felled trees and by measuring annual rings on increment cores. Residual trees growing next to the strips increased by an average of 5.9% in increment, and utilized the strip area up to 19.2%. Increment increased steadily over the 13 years since the strips were cut. The trees located along the north edge of the strips reacted more strongly than the trees along the south edge. Similarly, smaller trees reacted much more strongly than the larger trees. However, the increase in increment appeared to extend only 3 m into the residual stand. Isomaki and Niemisto (1990) looked at the growth and yield of even-aged Norway spruce stands adjacent to 6 to 11-year-old, 5.1 m (mean outside width) strips in southern Finland. The stands were sampled after the first commercial thinning. The total sample area covered 1.77 ha. Plot sizes were 25 to 80 m long and 30 m wide. Over all plots, 1734 trees were measured, including dbh, stump diameter, bark thickness, and radial increments. For each tree, neighbouring trees that were closer to the strip were located. The largest angle from the tree to the two nearest neighbours in the direction of the strip was then measured (i.e., the sector angle). Diameter measures were also taken on stumps. In addition, from each plot, 20 trees were destructively sampled, for a total of 420 trees. The total length of the strips involved in the project was 773 m; strips represented 17% of the stand areas. Edge trees showed an increase in growth even during the first season following thinning. A maximum of 25% increase was reached five years after thinning, and this level continued 10 years following thinning. This effect extended only 3 m into the residual stand. Isomaki and Niemisto (1990) noted that the sector angle, as a measure of growing space, was more related to the increments than the distance from the strip edge. The mean proportional increment of the cross sectional area of these classes differed significantly among sector angle classes (> 180 degrees, 121-180 degrees, 46-120 degrees and < 45 19 degrees). Trees classed as edges showed the greatest response, and the response was significantly different than those trees that were 2 to 3 m from the stand edge. Neither distance from the edge nor sector angle opening had any impact on height growth. The diameters of trees growing adjacent to the strips were, on average, 4.7% larger than trees away from the openings. The basal areas of trees growing adjacent to the strips were, on average, 6% larger than trees not near strips. Looking at the edge effects of 3 to 6.5 m wide extraction roads in Germany, Landbeck (1965) found that height growth of 60 to 120-year-old Scots pine was negatively influenced by the extraction road to a distance of 10 m from the edge. The mean diameter was 10% higher in the 10 metres adjacent to the extraction road, relative to the mean diameter farther inside. Van Laar et al. (1990) observed the growth responses of Douglas-fir after a strip thinning in Germany. The authors found that extraction roads (5 m wide) had a statistically significant effect on Douglas-fir tree growth, but that effect was limited to the first row of trees along the extraction road edge. The edge effect increased with tree size. For heavily thinned stands with extraction roads, no edge effect from the extraction road was found. In northwest Oregon, McCreary and Perry (1983) found the edge effects from a 30 foot (9.1 m) wide strip thinning extended no more than 10 feet (3.05 m) into the adjacent 35-year-old Douglas-fir stand. Using two increment cores taken at breast height, they measured the radial increment for five years before and after the strip thinning. The basal area increment for the five years after thinning, relative to the five years prior were calculated as a basal area increment ratio (BAIR). Large ratios were indicative of an accelerated growth rate and the use of this ratio removed the influence of tree size. Significant differences between the BAIR for strip thinned plots versus the control plots (not strip thinned) were found. The authors examined the relationship between the BAIR and the distance from the edge, and determined that the BAIR was negatively related to distance from the edge. However, only 20% of the variation in BAIRs could be explained by distance alone. After stratifying the trees into three zones from the edge (< 10 feet (3 m), 10-16.4 feet (3-5 m) and greater than 16.4 feet (> 5 m)), the authors compared the BAIRs with the control trees. The found a significant 20 difference between control trees and those within the first 3 m from the edge, only. These results of edge influence were similar to those found by Bucht and Elfving (1977), and Isomaki and Niemisto (1990). McCreary and Perry further stratified the trees into two diameter classes of greater or less than the median diameter of 11.8 inches (about 30 cm) and into two location classes from the edge: greater than or less than 3 m from the edge. These four groups of trees were compared with their respective cohorts in the control. Trees in both size classes within 3 m from the strip edges had significantly (p<0.05) larger BAIRs than similar sized trees in the control plot. 2.6.4 Built Roads Much of the literature on the effects of roads relates to the establishment of seedlings on compacted road prisms or bladed trails (e.g., Smith and Wass, 1979; Smith and Wass, 1980; Wert and Thomas, 1981; Lockaby and Vidrine, 1984; Smith and Wass, 1994; Wass and Smith, 1994; Hill etal., 1995; Young etal., 1995; Senyk and Craigdallie, 1996; Guariguata and Dupuy, 1997; Hope, 2001; and Dykstra and Curran, 2002;). There is little recent or comprehensive research available on the growth and yield of stands adjacent to road corridors specifically. Unlike the strip roads used in thinning, skid roads tend to be built with tractors, and result in bladed trails that modify the grounds natural contour. Comparatively, excavated roads are built for more permanent access and for hauling logs and equipment. Unlike skid roads, the excavated roads tend to be wider and are constructed with a more durable road prism and running surface, depending on the nature of the parent material and the design class of the road. In Germany, Kramer (1958) studied the effects of road corridors of varying widths on Norway spruce and European beech. An increase in increment was found for both species, but this did not extend considerably beyond the very edge of the stand. For Norway spruce, an increase of 1 to 3 m in height was noted up to an age of 80 years and reached a maximum for road widths of 9 to 10 m. Height growth was slightly reduced for road widths of more than 20 m. Diameter growth slightly increased in the younger stands, but reached 135% of interior stand growth when the Norway 21 spruce trees were 50 to 60 years of age. For beech, a 1 m increase in height was noted, up to a road width of 23 m. No difference was noticed for heavily thinned stands for road widths less than 10 m. The maximum height increment was noted for a 12 m road width. The diameter increments were similar to those expected for heavy to very heavy thinning. Pfister (1969) expanded on the previous work of Landbeck (1965) and Kramer (1958) by determining how much of the potential timber production was lost due to the permanent road system adjacent to western white pine (Pinus monticola Dougl.) stands in northern Idaho. The stands selected were 25 to 50-year-old western white pine plantations that had permanent roads with an average clearing width of 18.5 feet (5.64 m), constructed at the time of stand establishment. Plots 33 feet wide by 66 feet long (approximately 10 m by 20 m) were established above and below the road at each location. The edge was determined to be the drip line of the tree canopy, corresponding with either at the top of the cut slope on the uphill side, or near the top of the fill slope on the downhill side. Road widths, along with cut and fill dimensions were measured at each plot. Trees greater than 2.5 inches dbh (6.4 cm) were measured for height, diameter, distance from the road edge, and crown class. Twelve sample plots in 40-year-old stands were established on slopes ranging from 20 to 80%, with an average slope of 44%. Roads were out-sloped 1. Heights and diameters of dominant and co-dominant trees were adjusted to remove the between plot variation, and then the data were pooled into six 11 foot (3.35 m) strata, representing six equidistant strips within the plot. For plots below the road plots, Pfister found border tree heights were 113% greater and border tree diameters were 131% greater than trees from plots established within the stand. Edge effects were not evident above the road. Beyond 24 feet (7.3 m) for the plots below the road, the edge effects appeared to be negligible. In order to compare volumes, Pfister grouped the trees into 22 foot (6.7 m) zones from the road edge. For plots below the road, the volume was significantly higher in the 1 An outsloped road has the road surface aligned to direct road surface water towards the fill slope, or shoulder of the road. An insloped road directs the road surface water into the ditch at the toe of the cut slope. 22 zone closest to the road, than in either the second or third 22 foot (6.7 m) zone. The differences in volume between the second and third 22 foot (6.7 m) zones were not significant. For plots above the road, no differences were detected. Pfister attributed the difference between below road and above road growth responses to improved site productivity below the road, primarily due to the increased water availability. Furthermore, he suggested that the expected growth response of edge trees above the road may not have occurred because of poorer site conditions caused by loss of moisture from the cut bank. Wert and Thomas (1981) examined the growth and development of 22-year-old Douglas-fir trees growing on and around skid roads used for tractor logging in northwest Oregon. The main objective was to examine the impacts of soil compaction on Douglas-fir growth. Three types of areas were looked at; the skid road itself, the area unaffected by the skid road, and the transition zone between the two areas. Total volumes were 34.1 m 3 per ha for the skid road, 97.2 m 3 per ha for the transition zone, and 128.9 m 3 per ha for the undisturbed area. The stand densities were 693, 974 and 1180 stems per ha, respectively. Using analysis of covariance, Wert and Thomas found no difference in the rate of older tree height growth between the three areas, however there was a significant difference in the adjusted height means. Trees growing on the skid roads took 4.1 years longer to reach breast height than trees in the other two areas. Further analysis revealed that the trees growing on the skid road itself took longer to become established, but upon reaching certain age-height combinations, the trees on the skid road grew at the same rate as those in the other two zones. Comparing soil bulk densities, significant differences existed between the mean bulk densities at 20 to 30 cm depth for the skid roads, compared to the transition zones and the undisturbed areas. The top 15 cm of the skid roads had recovered from compaction after 32 years. Wert and Thomas suggested that skid road compaction resulted in a significant difference in the overall height of trees growing on the skid roads, and that those same trees are likely to become suppressed. Hope (2000) analysed the effects of construction and use of bladed skid roads on early tree growth in coarse textured soils in the Interior Cedar-Hemlock biogeoclimatic zone near Vernon, B.C. The growth rates of planted lodgepole pine, 23 Douglas-fir and white spruce seedlings were assessed after one, two, three, five and ten growing seasons. Hope (2000) found that the soil properties of water storage capacity and water content did not increase on the skid road surfaces, and that any changes to soil aeration, strength and nutrition did not seem to limit growth. Soil bulk densities were increased by approximately 40% on the skid road surfaces, and were apparent after 10 years, however these high bulk densities did not approach growth-limiting thresholds except on the inner track of the skid road. Early tree height and diameter growth rates were approximately 35 and 55% greater, respectively, on the outer trail position and on the road sidecast, than those in undisturbed positions (Hope, 2001). Tree growth on the inner track of the skid road was comparable to non-track tree growth. Hope (2000) speculated the growth increase on the outer trail and road sidecast positions was attributable to reduced vegetation competition both above and below ground (Smith and Wass, 1976) and to accelerated soil warming due to removal of the organic horizon (Fleming etal. 1998). Mayaka (1994) examined the impacts of roads on ayous {Triplochiton scleroxylon K. Schum) plantations in Cameroon. The road was oriented east to west. Diameters were measured along transects from the road edge. He then modeled this diameter change and found that the road edge effect extended 15 m into the plantations. Tree diameter decreased by nearly 50% at 10 m into the stand, but then increased and levelled off at 58% of edge tree diameter at 15 m into the stand. 2.7 E F F E C T S OF MORTALITY ON STAND YIELD 2.7.1 Mortality Many of the studies on impacts of growth on edge effects ignored any changes in mortality in order to examine gross growth changes only. For example: • Isomaki (1985) did not include stands where damage to trees or soil damage on the strips was evident; • Niemisto (1989) assumed that no tree damage occurred; 24 • Bucht (1977) and Bucht and Elfving (1977) chose sample sites without damage to the stands; and • Pfister (1969) chose against sample sites with obvious stocking irregularities or low survival rates, and eliminated data with high variation in stocking density due to heavy blister rust mortality. Study results in these cases may have overestimated the growth response as a result. Conversely, Chen etal. (1992) chose to study mortality rates, also. They found increased mortality along the edges of a clearcut adjacent to an old growth Douglas-fir forest edge. This was evident in the gradient of stocking densities and canopy cover. They found that the number of stems per ha > 6 cm dbh increased significantly from the edge for 60-120 m into the stand. Canopy cover also increased significantly from 35% to 65% over the first 60 m from the edge towards the stand interior. The variation in stem numbers was higher towards the interior than at the stand edge, although the variation in canopy cover was highest near the edge. This gradient was attributed to the significantly high levels of logs and wind fallen trees found close to the edge, versus very low levels beyond 60 m from the edge. The higher number of logs and wind fallen trees was attributed to the practice of falling snags within 60 m of the clearcut edge, and the higher propensity for windthrow damage near stand edges. Burton (1999) noted in their study that the density of lodgepole pine, subalpine fir (Abies lasiocarpa (Hook.) Nutt.) and hybrid white spruce (Picea engelmanniiX glauca) canopy trees >7.5cm dbh at clearcut edges was significantly reduced by windthrow for the first 10 m from the edge by an average of 25%, as the height-to-diameter ratios in these stands were greater than 80:1. The windthrow resulted in a 3 1 % reduction in stand basal area and a 34% reduction in timber volume. Harrington and Hendrick (1999) investigated the impacts of bark beetle induced mortality and resource availability on slash pine adjacent to 75 m circular gaps in the state of Georgia, U.S.A. Beetle induced mortality was found to increase within 12.5 m of the gap edge, occurring at a rate of 3.1 m 2 ha"1 yr"1, compared to a rate of 0.1 m 2 ha"1 yr"1 for undisturbed stands. The interaction between gap treatment and distance from 25 the edge was not significant, indicating that there was no difference in mortality for trees close to, or further from the gap edges. Bella (1986) attempted to account for mortality by introducing 10% mortality into his estimates of yield due to edges to approximate mortality trends in yield tables. However, he did not account for differential mortality that may have been present at the edges of the seismic lines that he studied. 2.7.2 Yield Bella (1986) also estimated the impacts of seismic lines on stand basal area/ha for lodgepole pine. He used yield tables to assess the gain in basal area over areas that responded to the edge, and then compared the gain with the yield lost on the cleared line. For lodgepole pine on medium high sites, he determined that only about 6% of the loss was recovered for the 20-year-old stand, 6.6% in the 40-year-old stand, and 6.9% in the 60-year-old stand. Since there was no real response in height growth, the values for basal area/ha also indicate the recovery expected for volume/ha. White spruce was expected to regain less than lodgepole pine, since white spruce responded with lower gains in diameter. This recovery of volume per ha would not occur if adjacent stands were cut shortly after cutting the seismic lines; however, salvage of logs from the seismic line would reduce the volume lost. In their study in Lapland on Scots pine, Bucht and Elfving (1977) found that the yield following 4 m wide strip thinning was 8 1 % of the control stand yield. In Germany, Landbeck (1965) found only a 1.8% reduction for Scots pine stand volume per ha, when including half of the 3 to 6.5 m road width with the 10 m treed strips alongside the road. No critical width was established, as the road widths were only 3 to 6.5 m wide. For Norway spruce in Germany, Kramer (1958) found that there was no noticeable reduction in standing volume when road corridors were less than 5 m wide. When roads were 5 to 9 wide, there was an overall reduction in stand volume of 7%. With beech, 12 m was the critical road corridor width where standing volume was reduced. The author attributed the ability of beech to use the wider road openings than spruce to the crown expansion potential of the species (Kramer, 1958). 26 For 10 m wide hydro line clearings in Finland, Isomaki (1986) calculated the increase in volume for edge trees as equivalent to recovering about 0.6 m to 1.1 m of the clearing for Scots pine. In a later paper, Isomaki and Niemisto (1990) estimated losses of 10 m 3 per ha for 15 years from 4 m wide strip roads established at 30 m intervals in Norway spruce stands, in Finland. Pfister (1969) theorized that gains in net yield from road edges can be applied against the loss of timber production growing space used by the road. He calculated that the increased volume for western white pine in Idaho was equivalent to recovering 13.1 feet (3.99 m) of the 18.5 feet (5.64 m) average clearing width in the study, including the road and the cut-bank. With an average interval of 400 feet (121.9 m) between roads, this left only 1.4% of the area as unproductive. Niemisto (1989) used a model to estimate the growth losses caused by strip thinning for Norway spruce in Finland. He assumed no damage occurred during harvesting, and used the research results from the same data as Isomaki (1985) and Isomaki and Niemisto (1990). Niemisto (1989) reconstructed the trees prior to thinning, and then simulated the 4 m wide strip thinning. He noted that improvements in the model were needed to better reflect the results found for the experimental plots. Pukkala (1989) suggested that competition with neighbouring trees must be modeled in order to examine any increases or decreases in growth losses. He used his tree model to examine the impacts of 6 m wide strip roads at 20 m intervals for Scots pine in Finland. He concluded that thinning in the residual stand should not take place at the edge of the roads, if residual volume is to be maximized. 2.8 L ITERATURE REVIEW OVERVIEW The objective of this review was to provide an overview of literature on the growth of trees and stands adjacent to access roads, for species and sites similar to those found in the Williams Lake area of B.C.; however, very little literature was found. Therefore, the review was extended to other linear openings, and to other sites and species. One of the difficulties in reviewing the literature on edge effects is that there are many definitions of edge. For this review, the edge was considered to be at the tree 27 line, and the transition zone from this edge to the interior forest was also considered to be of interest. Another difficulty in examining the impacts of a linear opening is defining the width of the opening. For most of this review, nominal widths were given. Microclimatic changes often produce changes in tree and stand growth. Soil temperature and moisture changes may extend well into the residual forest (up to 120 m or farther), depending on the macroclimate. Changes in light have also been noted. The degree of impact varies with aspect, macroclimate, the type of opening (e.g. clearcut, small openings, linear openings), and the characteristics of the vegetation adjacent to the edge. A number of studies present the impacts of early tree growth on rehabilitated road prisms or bladed trails. Tree growth responses varied between studies, but generally trees growing on the disturbed portion of the trails grew at the same rate, or less, than those growing in an undisturbed section of the stand. Nearly all of the literature reviewed indicated an increase in the diameter increment for trees near the edge of the linear opening relative to trees farther into the residual stand. Only a few studies indicated a change in height increment. The amount of increase declined with increasing distance from the edge and also varied with stand density, aspect, species, and type of opening. For many of the studies, the greatest increase in diameter was within the first 5 m from the linear opening edge. For some species, this increment was delayed for a few years from the disturbance; one study gave evidence that this was due to the time needed to increase the root mass. Many studies noted an increase in taper and number of branches resulting in a corresponding decrease in value. Some studies found trees at the edge tended to be more elliptical in shape also. The reduction in volume caused by linear openings is somewhat compensated by an increase in diameter growth of edge trees. Of the studies included in this review, the most optimistic compensation resulted in a 1.4% decrease in stand volume. However, the increase in diameter will be less or will not occur for older stands, or for more open stands. There can also be a decrease in value for edge trees, and an increase in mortality caused by damage and windthrow. 28 The findings from this literature review led to the establishment of a field study and the subsequent development of a simple model to evaluate the impacts of roads on adjacent tree growth and stand attributes in the Williams Lake area of B.C. The methods for this research project draw upon the studies presented in this literature review that specifically deal with the impacts of linear openings on stands of trees growing near roads. These methods are presented in the following section. 29 3 METHODS 3.1 INTRODUCTION Based on the literature review, the silvics of lodgepole pine, and discussions with personnel from Lignum Ltd. and the Ministry of Forests, a research plan to examine the effects of roads on the growth of adjacent lodgepole pine trees within the boundaries of Lignum's Innovative Forest Practices Agreement Area (IFPA) was developed. Field sampling of predominantly lodgepole pine stands was completed by September 1999, followed by measuring radial increments from sample cores until April 2000. Confirmation of road ages then followed. Construction of the database and analysis of the data proceeded until February 2001, when a summary report was completed for Lignum (LeMay etal. 2001). Additional modelling and analysis of the results continued with the development of this thesis. The methods section begins with a description of the field sampling study site around Williams Lake B.C. followed by presentation of the field sampling and road age estimation methods. Data analysis procedures along with the development of statistical models are presented last. 3.2 STUDY SITE DESCRIPTION The IFPA covers 620,000 ha in the Cariboo-Chilcotin region of B.C. (Lignum, 1998a), and at the time of sampling, was distributed amongst four forest districts: Williams Lake, 100 Mile, Horsefly and Chilcotin (Lignum, 1998b). Figure 1 shows the the distribution of the IFPA area and the sample plot locations, in relation to the communities in the Williams Lake area. The Biogeoclimatic (BEC) zones (see Meidingerand Pojar, 1991 and Klinka et al. 1990 for details on the B E C classification system) within the IFPA area are predominantly Sub-Boreal Spruce (SBS: variants dw1 and dw2), Sub-Boreal Pine-Spruce (SBPS: variants xc and mk) and Interior Douglas-fir (IDF: variants dk3, dk4, and 30 xm). However minor components of Engelmann Spruce - Subapine fir (ESSF), Bunchgrass (BG), and Montane Spruce (MS) zones are also found (Lignum, 1998b). — * = j - . . j j Williams Lake "ST""" " • British Columbia Ukely ...-....-.j • Towns » Plot Locations | | IFPA A 90 0 90 180 Kilometers N Figure 1: Distribution of the 44 sample plots established during the summer of 1999 within the Lingum Ltd. Innovative Forest Practices Agreement area boundaries. Inset map shows the location of Williams Lake in the province of British Columbia. Created in ArcView ver. 3.1 (ESRI, 1996) using digital data from Lignum (1998b). Lignum's allowable annual cut in the IFPA is comprised primarily of two commercial species: lodgepole pine and Douglas-fir (Lignum, 1998a). Lodgepole pine is found predominantly in the S B S and S B P S zones, but also at higher elevations in the IDF zone (Meidinger et al., 1991; Steen and Demarchi, 1991; Hope et al., 1991). 3.2.1 Sub-Boreal Spruce (SBS) Zone The S B S zone is generally found at low elevations in the central interior of British Columbia, below the E S S F . The S B S borders the Boreal White and Black Spruce 31 Zone (BWBS) zone to the north, while the Interior Cedar Hemlock (ICH) zone is found to the northwest and east. The S B P S zone is adjoined in the southwest, and the IDF in the south. The S B S zone has an intermediate size, with a prevailingly continuous distribution (Meidinger etal., 1991). Most of the S B S zone is located on the gently rolling terrain of Nechako and Fraser plateaus. This zone is influenced by the continental sub-boreal climate of B.C., and is under a partial influence of arctic air masses during winter, and relatively warm, moist and short summers (Meidinger et al., 1991). Soils found in the S B S are primarily from the Brunisolic, Podzolic and Luvisolic soil orders. Morainal deposits are abundant, with the most common soils being Podzols and Brunisolic and Orthic Gray Luvisols. (Meidinger et al., 1991) Lodgepole pine is the early serai species, particularly in drier subzones found in Lignum's IFPA (Lignum 1998b; Meidinger etal., 1991). Hybrid white spruce and sub-alpine fir are the main climax species in cooler and wetter subzones. Douglas-fir is a long-lived early serai species on dry and warm sites, while black spruce occurs occasionally with lodgepole pine on nutrient-poor or poorly aerated soils (Meidinger et al., 1991). Ten subzones are recognized in the S B S zone along precipitation and temperature gradients; however, only the dry warm SBSdw subzone is found within Lignum's IFPA (Meidinger etal., 1991; Lignum, 1998b). 3.2.2 Sub-Boreal Pine-Spruce (SBPS) Zone The S B P S zone is located at low elevations in the central interior of British Columbia, below the E S S F and Montane Spruce (MS) zones, above the IDF zone, and it joins the S B S zone in the north and east. It is a small zone, with a prevailingly continuous distribution (Steen and Demarchi, 1991). As with the S B S , the zone is located on the Fraser plateau and the southern portion of the Nechako plateau. The S B P S zone is heavily influenced by the rain shadow of the Coastal Mountains and a relatively high elevation (>1100 m). Climatic characteristics of this zone include low precipitation, dry air, and clear skies at night. Frost is common in all 32 months, and the harsh climate severely limits forest productivity (Steen and Demarchi, 1991). Soils found on zonal sites in the S B P S are primarily of the Brunisolic and Luvisolic order (Lord and Valentine, 1978; Steen and Demarchi, 1991). Orthic Dystric Brunisols and Brunisolic Gray Luvisols are most common, however gleyed subgroups of the Brunisols and Luvisols, as well as Gleysols are also found on imperfectly or poorly drained sites (Steen and Demarchi, 1991). The humus form is relatively thin (<4cm), and has a very slow rate of decomposition (Steen and Demarchi, 1991). Non-forested wetlands are common in the S B P S , and many are managed for agriculture and grazing (Steen and Demarchi, 1991). Even-aged, semi-open canopy lodgepole pine stands extend across nearly all the S B P S landscape. The few stands that reach climatic climax consist of even-aged lodgepole pine, with some white spruce found mainly in the understory. Lodgepole pine is by far the most common species and large areas of the forest contain no tree species other than lodgepole pine. Forest fires are very frequent, creating a mosaic of early and mid-seral stages of stands across the S B P S landscape. Stands more than 120 years old are infrequent in the S B P S (Steen and Demarchi, 1991). Other species found in the S B P S include white spruce, trembling aspen (Populus tremuloides Michx.), sub-alpine fir, black spruce, black cottonwood (Populus tricocarpa Torr. and Gray) and Douglas-fir (Steen and Demarchi, 1991). Four subzones are recognized in the S B P S zone along precipitation and temperature gradients. However, only two are found within Lignum's IFPA: the very dry cold S B P S x c and the moist cool SBPSmk (Lignum, 1998b). 3.2.3 Interior Douglas-fir (IDF) Zone The IDF zone is found at low to middle elevations in south and central B.C., typically located below the MS zone and above the Ponderosa Pine (PP) and Bunchgrass (BG) zones in deep valleys. The IDF zone borders the S B P S and S B S zones in the northern portion of Lignum's IFPA (Lignum, 1998b). This zone is influenced by a dry continental cool temperate climate (Hope etal., 1991). 33 Significant growing-season water deficits and growing-season frost, especially in cooler (northern) subzones, are common. Topography is variable: flat to gently undulating plateaus, valley bottoms and lower to middle sidewalls of the major valleys (Hope etal., 1991). The IDF zone is represented by seven subzones, differentiated along precipitation and temperature gradients. Only two are found within Lignum's IFPA: the very dry mild IDFxm and the dry cool IDFdk (Lignum, 1998b). Climatic climax stands are represented by open-canopy, uneven-aged stands consisting almost entirely of Douglas-fir, which commonly occur as a result of frequent fires (Hope etal., 1991). The very dry subzones commonly have ponderosa pine (Pinus ponderosa Laws.) and lack lodgepole pine except in the higher elevations (Hope era/., 1991). A combination of climatic, edaphic, and topographic conditions in the IDF zone has led to the development of non-forested (grasslands and wetlands) and forested ecosystems (Hope etal., 1991). Soils of the zonal sites within the IDF within Lignum's IFPA are typically Eutric and Dystric Brunisols and Orthic or Dark Gray Luvisols (Lord and Valentine, 1978; Hope etal., 1991; Lignum, 1998b). Soils in the IDF zone are generally medium to rich in their nutrient status (Hope era/., 1991). Grassland phases of subzones in the IDF typically have soils of the Chernozemic order (Hope era/., 1991). 3.3 FIELD SAMPLING 3.3.1 Map Database An electronic database of the Forest Cover type information was acquired from Lignum. The database files contained approximately 196,000 forest cover polygons, including those within the Lignum IFPA (Lignum, 1998b). Lignum staff facilitated the transfer of all polygon information to an ArcView (ver. 3.1) Geographic Information System (GIS) platform. Information was added to the GIS database, including: the IFPA boundaries, the forest development plans for the Chilcotin, Williams Lake and Horsefly districts, water features, contour intervals, recent stand history for the previous 34 5 years and four roads coverages, including Terrain Resource Inventory Map (TRIM), IFPA, development plan and Ministry of Forests road files. Additional spatially explicit data included forest district boundaries, and B E C information to the sub-zone and variant level. Current development plan information for the 100 Mile House district was added in August, 1999, when this information became available in digital format. Status information about District Lots was also obtained at that time. The Forest Cover dataset was reduced from 196,000 polygons to 6,055 polygons found within the borders of the IFPA boundary, by linking the sorted dataset with the spatial polygons found within the IFPA. Using ArcView, these 6,055 sorted polygons within the IFPA were spatially viewed using the attributes of interest: site class, age class and crown closure class, as well any of the other attributes found in the headers of the forest cover database. Proximity of the polygon to roads could also be determined. This information allowed more efficient sorting of the polygons to determine whether or not they were eligible for sampling. As well as the electronic database, copies of the 1:20,000 Forest Cover type maps were obtained from the Lignum office in Williams Lake. Information from both the GIS and Forest Cover maps was used, as the GIS provided more recent information, including details on proposed harvesting activities, whereas the forest cover type maps provided information on historic management or disturbance activities. Forest development plan information for companies operating in the IFPA other than Lignum, was not available. 3.3.2 Selection of Polygons for Sampling Using the sorted forest cover data, eligible polygons were selected for viewing. To fill a sample cell matrix, eligible polygons were grouped into site classes, age classes and crown closure classes defined as: 1. Three site classes, derived by dividing the site index (base age 50) range into poor (less than 12.5 m), medium (12.5 to 17.4 m), and good (17.5 to 22.4 m); 35 2. Three age classes: juvenile (15 to 40 years), early mature (40 to 80 years), and late mature (80 to 120 years). Over-mature stands were intentionally not selected for sampling, as no growth impacts were expected; and 3. Three density classes: low, medium, and high based on the Ministry of Forest's Permanent Sample Plot Header Database matrix classes. Ministry of Forest's crown closure classes of 0 to 3 were grouped as low density, 4 to 6 as medium density, and 7 to 9 as high density. The polygons were also grouped into additional road categories: road age, road class and width of the road right-of-way based on the following categories: 1. Two strata for road ages: 10 to 20 years, and greater than 20 years. Previous studies (e.g.: Urban etal., 1994) indicated that the time since disturbance must be 10 years or more in order to assess any growth increases. For sampling, road ages were estimated from development years of adjacent cutblocks, but were later confirmed using sequential air photos; 2. Three road right-of-way width categories of less than 20 m wide, 20 to 30 m wide and 30 m to 40 m wide. Rights-of-way greater than 40 m wide were not sampled; and 3. Three roads types: primary, secondary and tertiary. The dendritic pattern of road placement did not necessarily result in a true classification of road type. For example, a typical tertiary road (i.e., a spur road developing one or two cutblocks, or providing access to a private lot) sometimes joins to a primary road, rather than to a secondary road. Furthermore, predicting road width varied with road types. Therefore, the road type classification and road width were field-checked prior to sampling. Once polygons were classified by the sampling matrix criteria, high site index polygons were initially targeted for sampling, since these were more likely to show a growth response to the road edge. Most of the high sites were relatively close to Williams Lake townsite. The lower site class sites were sampled near the end of the 36 field season, and at more remote locations. Polygons were rejected for sampling for the following reasons: 1. The polygon contained less than 80% pine based on a field reconnaissance, even though the forest typing indicated a primarily pine stand; 2. The polygon had been identified for harvest in the forest development plans, or was already marked for harvest in the field; 3. Trees in the polygon were influenced not only by roads, but also by adjacent or contiguous harvesting; 4. The polygon was selectively cut. Since the removal of trees will affect competition and therefore alter tree growth, stands with excessive stumps were excluded. In some cases, only portions of the stands were ineligible and were excluded from sampling; 5. The road edge was too short to locate a plot without influence from adjacent edges; 6. Other polygons were previously sampled that filled the sampling matrix cell; or 7. The polygon was located in private land, reserves or recreation sites. Prior to August 1999, polygons that fell within a known District Lot were avoided, since the status could not accurately be determined. Once the status information was obtained, District Lot polygons in the IFPA were eligible for sampling if their status was "crown managed forest", and they were not known to be private land, reserves or recreation sites. Polygon selection proved to be quite difficult, primarily due to the number of polygons with stumps near the roadside. This was more prevalent in stands nearer to the Williams Lake town site. Many potential polygons were rejected due to unknown previous harvest activity, such as the removal of dangerous trees for safety reasons, authorized fuel wood or pole use, or unauthorized harvest. Stands recently managed for beetle salvage were not identified in the GIS data, or on the 1:20000 Forest Cover maps. 37 Selected polygons were located in the field by tying to land features found on the 1:20,000 forest cover map sheets, the map viewer from the GIS on a laptop computer, and with the Universal Transverse Mercator (UTM) co-ordinates on a handheld Global Positioning System (GPS). The G P S model used was a Garmin 12XL with an accuracy of ±15 m, but was subject to accuracy degradation of up to ±100 m under the United States Department of Defence-imposed Selective Availability Program (Garmin Corp., 1998). 3.3.3 Establishing Plots Within Polygons Once an appropriate polygon was selected and located, the sample plot was randomly located within the polygon, from the eligible portion of the polygon along the length of the road. The length of the polygon along the road was either determined from the 1:20000 maps, the GIS, or was measured in the field. Difficult polygon edges were found by determining the timber type change, by tying to known natural features on the GIS or 1:20000 maps, and by using the G P S . Offsets of 40 m were made from edges where harvesting activity occurred to remove the affects of the harvest edge. Portions of polygons that were partially cut were also eliminated from the random location of the plot edge to remove the effects of partial cutting. Following previous studies on road edge effects (e.g., Isomaki, 1985; Isomaki and Niemisto, 1990), the fixed area plots were rectangular, and oriented from the stand/road right-of-way edge to 40 m into the stand. Each plot was divided into five zones from the road edge: 0-5 m, 5-10 m, 10-20 m, 20-30 m, and 30-40 m as shown in Figure 2. Road Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Right 0 to 5 m 5 to 10 m 10 to 20 m 20 to 30 m 30 to 40 m Way Figure 2: Plot subdivided into five zones from the road right-of-way edge. 38 Based on the literature, the growth of trees in Zone 5 will not likely be affected, therefore the plots were designed so that Zone 5 trees could used as the basis for assessing trees in the other four zones. The length of the plot edge along the road was varied, based on the density of the adjacent stand. To establish an estimate of stand density, up to three initial circular fixed area reconnaissance plots of 0.04 ha (11.28 m radius) were used to estimate the number of pine stems per ha. These data were recorded, but only used in Equation 3 to set plot width (i.e., the length along the road), in order to have seven to ten pine trees in the 0 to 5 m and 5 to 10 m zones. For example, for 500 pine stems per ha, a plot width (length along the road) of 30 m would result in seven pine trees expected in Zones 1 and 2, and 15 pine trees expected for the remaining zones (i.e. a plotsize of 30 m by 40 m or 0.12 ha in size). Often, some sections of the polygon were ineligible for sampling, for the same reasons as those given for eliminating polygons. , , , ' l ( \ 10 pine stems for Zone 1x10,000 m21 ha r o , length along the road [ni) = [oj #of pine stems I hax5 m width for Zone 1 Once the plot size was determined, the rectangular fixed area plots were randomly established along the roadside by selecting a plot corner location from the total length of the eligible portion of the polygon adjacent to the road. Where plots landed in previously unknown ineligible sites within the polygon, plots were offset by half plot width increments in a systematic manner, until a suitable site was found. If the road right-of-way edge was not straight, the compass bearing and length along the road right-of-way was based on the average road right-of-way edge. All distances were horizontal. Plots were identified with aluminium marking plates on trees near the plot corners at roadside. The plots were divided into the zones shown in Figure 2. Plot zones were identified by orienting the painted tree numbers in opposite direction from the adjacent zone. Plot boundaries were marked by painting a circular ring around trees at the edge of the plot. 39 3.3.4 Plot Measurements Collected For each plot, data recorded on the field cards from the GIS database included the B E C classification to subzone and variant, the location (UTM Northing and Easting), the elevation, road type, and the estimated road age (see Appendix I, plot tally card). The plot slope, aspect, and the plot size from the field were also recorded. Measurements of the road right-of-way clearing width (road width) were taken at each junction of the plot corners with the road; a third measure was taken at the middle, for larger plot sizes, and for more variable road widths. Road grade, road bearing, and road prism measurements, including the orientation of the sample plot to the road were recorded, as was the plot right-of-way edge aspect. Road alignment measurements, including the incoming and outgoing bearings from the stations identifying the plot corners were recorded. At least two road prism measurements were taken for the sample plots at stations along the road identifying the plot corners. A minimum of two site index plots of 0.01 ha circular fixed area plots were located within the plot boundary. In each plot, the largest diameter undamaged pine top height tree was determined, and height and age were measured. 3.3.5 Tree Measurements Collected In each zone, the following measurements were taken for each standing tree of at least 7.0 cm diameter at 1.3 m (breast height) above ground (see Appendix I, tally card): 1. Species; 2. Diameter at breast height (dbh) in cm; 3. Tree crown class (i.e., dominant, co-dominant, intermediate, suppressed, and veteran); 4. Tree class code (refer to the Appendix I tally card for definitions); and 5. Damage, pest and pathogen indicators 40 For a random sample of five live lodgepole pine trees in each zone, total height, and height to live crown were measured using an Impulse Laser Hypsometer. Height trees were marked with aluminium tags at breast height (1.3 m above ground). Out of the five trees measured for height, three were randomly selected and two increment cores at right angles (perpendicular and parallel to the road) were taken at breast height (30 increment cores/plot). The increment cores were placed in tubes for transport, labelled with the stand number, zone, tree number, and position on the tree, and packaged for analysis at the University of British Columbia. Although the measurement of all live trees for height would be desirable, this would have greatly increased the measurement time per plot. Concern was expressed about measuring trees from May through August during the growing season. This was discussed with the thesis advisory committee and the original schedule was retained for the following reasons: 1. Cumulative height growth was measured (height rather than height increment) in the comparison among zones. The final height increment of the measurement year would likely be small, relative to the cumulative height, except for perhaps the young age class. This was not expected to be a large issue because of point 4 below; 2. For dbh, cumulative growth was used to compare among zones. Radial increment for the final year would also likely be small relative to the cumulative dbh, except for the young, low stocking type group. Since this type group may not occur on the landbase to any great degree, this was considered a small issue; 3. For radial increment, the final year could be eliminated from the analysis. Also, the radial increment prior to cutting for the road was compared to that following cutting. The annual radial increment was, therefore, averaged over several years; and 41 4. For all analyses, Zone 5 was used as the baseline for comparison. Since all zones were measured on the same time (or within a two-day period), the season of measurement is not likely an issue. Sub-sampled trees (five per zone) were marked using numbered tags at breast height, and core extraction locations were marked with red paint. 3.3.6 Radial Increment Measurements Cores were permanently mounted and labelled on small wood blocks for re-measurement as required. Core data were measured using the computer program WinDENDRO (Version 6.3a). WinDENDRO records the radial ring attributes in a text based format after the core has been measured as a scanned photo file. The photo file image may be expanded in scale with WinDENDRO to allow for the accurate alignment of the measuring crosshairs. WinDENDRO also has a function which will automatically record the latewood/earlywood boundary; however, the lodgepole pine rings in this project were manually confirmed to minimise measurement error with slow growing rings. As slow growing lodgepole pine rings are often quite difficult to see without aid, ring counts were checked using a microscope as necessary. WinDENDRO was used to compute the breast height ages for all cores, including those obtained for site trees. As each radial increment tree was cored twice, there were two measures of tree age for each cored tree. While great efforts were made in the field to hit the pith of the sample trees, any discrepancies which may have occurred between the ages of the two cores may be the result of missing rings, or from the core alignment being slightly offset from the pith. Increment data were screened, and cores remeasured if any discrepancies were noted. 3.3.7 Determination of Road Age Road right-of-way age estimates were made during the field sampling by using the ages of adjacent operational openings. However, to accurately measure the impact of roads on tree growth, a more reliable estimate of road right-of-way age was required. Aerial photos from the B.C. Ministry of Environment, Lands and Parks photo library in 42 Victoria, B.C. were obtained, and placed in sequence from the most recent photos, until those taken just prior to road construction. This narrowed the road age, but was limited to years when photos were taken before and after the right-of-way was harvested, a period from three to nine years. Road ages were estimated to be at midpoint (rounded up to the next whole number) of the flight interval, unless forest management information on the forest cover maps or GIS indicated an earlier, or later, date of right-of-way harvest. Further interpolation using the year of management activities in the accessed cutblocks provided greater precision and accuracy of road age estimates. At worst, road ages were estimated within ± 5 years of their actual age. Older road openings created greater difficulty narrowing the estimate of age estimates. For the few plots where no pre-harvest aerial photos were found, road age was estimated by examining human settlement or harvest patterns (and therefore access) in adjacent photos of the same flight year and by examining the road surface material to determine existence of a road subgrade (a lack of which might indicate the recent harvest of right-of-way timber). Further estimates were made by reviewing forest management activities on the 1:20000 forest cover maps, the GIS, or on the flight line maps themselves. 3.4 PRELIMINARY DATA ANALYSIS 3.4.1 Plot Data The basal area was calculated of each tree. For height trees, the live crown ratio (crown length/total height) was calculated. Variables obtained for each plot were: basal area per ha, average tree height (based on the sub-sampled 25 trees/plot), average diameter, stems per ha, average live crown ratio (based on the sub-sampled 25 trees/plot). Mean tree basal area, quadratic mean dbh (DBHq), and the Curtis (1982) Relative Density index (Curtis' RD) were calculated using the Equations 4 to 6: [4] 43 DBHq (cm) = ^mean tree basal area x 40,000/n [5] „ . , „ ^  basal area I ha r_. Curtis RD = . — [6] JiDBHq These calculations were performed for lodgepole pine only, and for all live trees. The number of dead stems per ha and the basal area per ha of dead stems were also calculated. Each of these variables were calculated for the five zones within each plot, using the lodgepole pine trees only. The average road width was calculated for the plot. Site index was determined by averaging the ages and the heights for the site trees in the plot, and then using the B C Ministry of Forests site index curves created by Goudie (1984). For the zonal measures (pine only), values relative to Zone 5 as the baseline were also calculated for: basal area/ha, stems/ha, mean basal area per tree, mean dbh, mean height, mean live crown ratio, mean DBHq, and mean Curtis' RD. 3.4.2 Radial Increment and Tree Basal Area Increment Data The annual increments for the sampling year (1999) were measured, but excluded from this analysis, as data were gathered during summer. Radial increment measures were converted to tree basal area increments. Following McCreary and Perry (1983), and Bella (1986), tree basal area periodic annual increments (p.a.i.) were calculated using Equation 7 for the following periods: 1) - 5 to 0 (5 years prior to road); 2) 1 to 2 years following the road; 3) 3 to 5 years following the road; and 4) in subsequent 5 year periods (e.g., 6 to 10, 11 to 15, etc.). n ^ annual increment PAIperiotl=^ [7] n For two of the 44 plots, the randomly selected radial increment trees (three per zone) had all established following the building of the road, so no pre-road p.a.i.'s were 44 available for those plots. For graphing tree basal area increment trends, the p.a.i for the 5 year period prior to the road being built was used to standardize the basal area increments for each tree. 3.5 STATISTICAL ANALYSIS Statistics for the 44 plots, and for each zone in the 44 plots were first calculated. The procedures followed in the analyses were: 1. Simple correlations between zonal measures for pine (e.g., pine basal area/ha for the zone) with distance from the road edge, and plot measures (e.g., plot basal area/ha, all live species) were obtained; 2. Three-dimensional plots of zonal measures versus plot measures and zone (1 to 5) were graphed to examine trends; and 3. General linear modelling (GLM, S A S , Version 6.12) was used to examine differences in zonal measures over the five zones, while controlling for plot measures of site productivity and density. Pairwise comparisons among zones were examined, if differences were detected. A similar approach to analysis was followed for the tree basal area increment data, examining basal area increments and relative basal area increments versus the distance from the road, and with stand level variables. For stepwise regressions, a significance level of 0.10 was used for both entry and exit of variables. For all other analyses, a significance level of 0.05 was used. For pairwise tests, the significance level was divided by the number of comparisons to help preserve the significance level for the overall analysis of covariance (Neter etal., 1996, pp. 1096-1097). The analyses were used to suggest any yield impacts that might occur with road edges, following work by Kramer (1958), Landbeck (1965), Pfister (1969), Bella (1986), Isomaki (1985 and 1986), Isomaki and Niemisto (1990) and others. Impacts were modelled using regression analysis and general linear modelling 45 4 R E S U L T S 4.1 PLOT DATA 4.1.1 Descriptive Statistics Forty-four sample plots were measured in the field, containing 5,929 trees that were at least 7.0 cm dbh. There were 3,998 lodgepole pine trees, 887 live trees of other species, and 1,022 dead standing trees and 22 stumps. The plot sizes ranged from 0.04 ha to 0.24 ha with an average size of 0.10 ha. The average plot age at breast height varied from 37 to 121 years, road widths varied from 5.3 to 38.5 m, and the representation over densities was fairly wide (Table 1). Table 1: Plot level attributes, based on 44 sample plots. Variable Species Minimum Maximum Mean Standard Deviation Mean age at bh (yrs) 37.3 121.3 79.7 22.0 Road age (yrs) 5 51 25 13.2 Mean age at bh at 3.2 102.4 54.7 25.8 time of harvest (yrs) Road width (m) 5.3 38.5 23.4 8.7 Site index at base Pine 9.5 20.8 15.4 3.3 age 50 (m) Curtis' RD Pine 2.1 9.0 6.0 1.6 Curtis' RD All live trees 3.7 10.5 6.8 1.6 DBHq (cm) Pine 9.8 30.7 18.5 4.9 DBHq (cm) All live trees 9.8 27.9 17.8 4.1 Mean dbh (cm) Pine 9.6 30.2 17.6 4.7 Mean dbh (cm) All live trees 9.6 26.7 16.7 3.8 Mean height (m) Pine 9.65 28.30 17.10 4.49 Mean basal area per Pine 0.0076 0.0739 0.0287 0.0156 tree (m2) Mean basal area per All live trees 0.0076 0.0612 0.0261 0.0124 tree (m2) Basal area/ha Pine 11.39 36.91 25.23 6.68 (m2 ha"1) Basal area/ha All live trees 14.86 45.02 28.69 8.02 (m2 ha"1) Stems/ha Pine 154.2 2625.0 1109.5 551.2 Stems/ha All live trees 444.4 2625.0 1266.3 513.2 46 Relative to the road position, the average plot slope was - 1 % , ranging from - 1 8 % below the road to +9% above the road. Plot elevation ranged from 920 to 1235 m. Plots were sampled from four different B E C subzones and variants: 13 plots from the dry cool IDFdk3; 20 plots from the moist cool S B P S m k zone; 8 plots from the dry warm SBSdw2; and 3 plots from the very dry cold S B P S x c (Meidinger and Pojar, 1991). All four cardinal directions were represented in aspect of the plot edge; 18 plot edges faced north; 7 edges faced east, 8 edges faced south, and 11 plot edges had a west aspect. Road directions were perpendicular to the edge aspect. Based on the descriptive statistics for pine trees in each zone, basal area appeared to be higher in Zone 1, as a result of a higher number of stems/ha (Table 2). Curtis' RD also appeared to differ among zones, with Zone 1 (0 to 5 m from the road edge) having the largest relative density. No differences in average diameter, mean basal area per tree, nor in DBHq were evident. On average, values for relative basal area/ha, stems/ha, Curtis' RD, and mean live crown ratio in Zone 1 (0 to 5 m from the road edge) were all more than 1.2 times higher than the Zone 5 values (Table 3). For some plots, however, there was a reduction in these variables relative to the interior zones; in other plots, there was a two-fold increase relative to Zone 5. Some increases in mean values for these variables were also noted in Zone 2 (5 to 10 m from the road edge). Correlations between pine zonal variables with distance and with plot level variables are given in Table 4. The zone basal area/ha decreased with distance from the plot edge (significant negative correlation with distance), as did Curtis' RD, and the mean live crown ratio. For the variables scaled using Zone 5 as a baseline, decreases with distance were significant for the relative basal area/ha, stems/ha, Curtis' RD, and mean live crown ratio. A significant positive relationship was noted for relative mean height versus distance from the road edge. For the variables correlated with distances, significant correlations with plot density measures and with site index were also common. 47 Table 2: Zone level attributes for pine trees in the 44 sample plots. Midpoint distance from the road for each zone is given in brackets. Zone Variable Minimum Maximum Mean Standard deviation 1 Basal area/ha (m2 ha"1) 6.31 56.28 29.95 10.93 (2.5 m) Stems/ha 133.3 3400.0 1289.4 761.8 Curtis' RD 1.3 12.8 7.0 2.5 Mean basal area per tree (m2) 0.0091 0.0853 0.0296 0.0177 Mean dbh (cm) 10.5 32.0 18.0 5.2 DBHq (cm) 10.8 33.0 18.7 5.3 Mean height (m) 8.10 26.05 16.45 4.66 Mean live crown ratio 0.219 0.812 0.438 0.151 Mean age at breast height (yrs) 28.7 132.0 78.8 25.6 2 Basal area/ha (m2 ha"1) 11.53 43.12 25.39 9.28 (7.5 m) Stems/ha 133.3 2600.0 1130.1 611.8 Curtis' RD 2.0 10.1 6.0 2.1 Mean basal area per tree (m2) 0.0078 0.1074 0.0302 0.0205 Mean dbh (cm) 9.7 36.9 18.0 5.8 DBHq (cm) 10.0 37.0 18.7 5.9 Mean height (m) 9.46 31.38 16.97 5.0 Mean live crown ratio 0.178 0.753 0.403 0.133 Mean age at breast height (yrs) 36.3 121.7 78.0 23.6 3 Basal area/ha (m2 ha"1) 10.08 40.46 24.35 7.89 (15 m) Stems/ha 133.3 2900.0 1084.3 574.4 Curtis' RD 2.0 9.3 5.8 1.8 Mean basal area per tree (m2) 0.0075 0.0889 0.0287 0.0173 Mean dbh (cm) 9.6 33.0 17.6 5.0 DBHq (cm) 9.8 33.7 18.4 5.3 Mean height (m) 9.32 28.60 17.23 4.74 Mean live crown ratio 0.172 0.686 0.367 0.134 Mean age at breast height (yrs) 37.7 122.7 81.0 23.3 4 Basal area/ha (m2 ha"1) 13.36 40.34 24.29 7.84 (25 m) Stems/ha 200.0 2500.0 1055.4 566.2 Curtis' RD 2.5 9.9 5.7 1.9 Mean basal area per tree (m2) 0.0062 0.0779 0.0293 0.0158 Mean dbh (cm) 8.7 29.2 17.7 4.8 DBHq (cm) 8.9 31.5 18.7 5.0 Mean height (m) 9.83 27.44 17.31 4.31 Mean live crown ratio 0.181 0.675 0.378 0.122 Mean age at breast height (yrs) 32.0 122.0 79.9 22.6 5 Basal area/ha (m2 ha"1) 10.0604 43.85 24.59 8.36 (35 m) Stems/ha 150.0 3500.0 1088.5 667.0 Curtis' RD 2.0 10.2 5.8 2.0 Mean basal area per tree (m2) 0.0058 0.0729 0.0299 0.0161 Mean dbh (cm) 8.5 30.3 18.0 5.0 DBHq (cm) 8.6 30.5 18.8 5.2 Mean height (m) 8.65 27.98 17.51 4.69 Mean live crown ratio 0.184 0.603 0.366 0.102 Mean age at breast height (yrs) 31.7 121.7 81 22.9 48 Table 3: Zone level attributes relative to Zone 5 values for pine trees in the 44 sample plots. Midpoint distance from the road for each zone is given in brackets. Zone Variable Minimum Maximum Mean Standard deviation 1 Relative basal area ha"1 0.230 2.693 1.313 0.561 (2.5 m) Relative stems ha' 1 0.372 6.250 1.398 0.969 Relative Curtis' RD 0.271 2.632 1.305 0.549 Relative mean tree basal area 0.301 2.516 1.088 0.442 Relative mean dbh 0.534 1.504 1.020 0.212 Relative DBHq 0.549 1.586 1.020 0.218 Relative mean height 0.577 1.318 0.948 0.155 Relative mean live crown ratio 0.704 1.907 1.214 0.309 Relative mean age at breast 0.443 1.721 0.992 0.253 height 2 Relative basal area ha"1 0.522 2.298 1.085 0.372 (7.5 m) Relative stems ha"1 0.364 3.250 1.182 0.626 Relative Curtis' RD 0.481 2.456 1.093 0.383 Relative mean tree basal area 0.395 1.980 1.050 0.386 Relative mean dbh 0.615 1.386 1.010 0.184 Relative DBHq 0.628 1.407 1.008 0.189 Relative mean height 0.673 1.287 0.972 0.130 Relative mean live crown ratio 0.752 2.127 1.126 0.307 Relative mean age at breast 0.542 1.370 0.973 0.160 height 3 Relative basal area ha"1 0.461 1.714 1.040 0.316 (15 m) Relative stems ha"1 0.444 2.625 1.119 0.466 Relative Curtis' RD 0.457 1.724 1.050 0.320 Relative mean tree basal area 0.517 1.674 0.998 0.289 Relative mean dbh 0.710 1.269 0.987 0.143 Relative DBHq 0.719 1.294 0.989 0.144 Relative mean height 0.753 1.316 0.991 0.133 Relative mean live crown ratio 0.506 1.700 1.013 0.265 Relative mean age at breast 0.553 1.495 1.011 0.153 height 4 Relative basal area ha'1 0.443 1.707 1.041 0.304 (25 m) Relative stems ha"1 0.394 2.143 1.068 0.379 Relative Curtis' RD 0.463 1.807 1.043 0.304 Relative mean tree basal area 0.575 1.981 1.018 0.255 Relative mean dbh 0.713 1.224 0.991 0.110 Relative DBHq 0.758 1.407 1.002 0.120 Relative mean height 0.813 1.244 0.998 0.101 Relative mean live crown ratio 0.592 1.615 1.044 0.230 Relative mean age at breast 0.511 1.381 0.994 0.127 height 49 Table 4: Correlations between zone variables and plot variables. Plot level variables are based on all live trees, whereas zonal variables are based on live pine trees only (* indicates correlations significantly different from zero for a = 0.05). Significant correlations with distance from the road edge are highlighted in gray. Zonal variable Distance from Plot Plot basal Plot Curtis' Plot site road edge (m) stems/ha area/ha RD index Basal area/ha -0.16295* 0.18164* 0.47945* 0.51456* 0.26809* (m2 ha"1) Stems/ha -0.09490 0.76839* -0.18049 0.14502* -0.33484* Curtis' RD -0 21983* 0.44285* 0.29997* 0.46580* 0.05886 Mean tree basal -0.00125 -0.58600* 0.50150* 0.16164* 0.59064* area (m2) Mean DBHq (cm) 0.00773 -0.63350* 0.52159* 0.17449* 0.59876* Mean dbh (cm) -0.00181 -0.59937* 0.55848* 0.21917* 0.61252* Mean height (m) 0.07116 -0.41664* 0.73919* 0.46447* 0.69662* Mean live crown -0.17080' -0.30395* -0.49478* -0.57084* -0.10078 ratio Relative basal -0.23218' -0.02522 -0.11578 -0.11268 -0.01717 area/ha Relative stems/ha -0 20675* -0.12314 -0.13792* -0.15303* 0.03198 Relative Curtis' RD -0.23215" -0.06023 -0.11794 -0.12171 0.01085 Relative mean tree -0.08689 0.10787 -0.04629 -0.02253 -0.15876* basal area Relative DBHq -0.03669 0.12031 -0.00459 0.01936 -0.13603* Relative mean dbh -0.04896 0.13293* 0.03430 0.05936 -0.12237 Relative mean 0.14894* 0.16787* -0.01108 0.04685 -0.15799* height Relative mean live -0.25899* -0.09709 -0.16672* -0.17597* -0.15545* crown ratio 4.1.2 Zone Level Variables Versus Plot Level Variables by Distance From the Road Edge Based on the variable summaries by zone (Tables 2 and 3) and the correlations with distance from the edge (Table 4), the main focus for analyzing differences among zones was to: 1) examine differences in stems per ha (SPH), basal area per ha (BA/HA), and live crown ratio, as well as these variables relative to Zone 5; and 2) examine average height relative to Zone 5. 50 4.1.2.1 Stems per Hectare The graph of pine S P H by zone against the overall plot S P H showed the expected direct relationship (Figure 3). The variance tended to increase slightly as plot S P H increased. The two zones closest to the road (2.5 m and 7.5 m) showed more stems than the interior zones, as noted in Table 2. This is also shown in the graph of pine S P H by zone versus distance from the road edge (Figure 4). 2500 J 2000 CO UJ-8 1500 Zone S P H b y , E d g e Distance vs P lo l S P H , Spec ies : ' P INE : * : . * A , o * © X G O k O & ' • * : • ' j . . A A X : ;^ o —I— '500' •DOO . : I. .2030: I ,2500 ,3000 1500 •PLOT.SPH- ' E D G E D I S T A N C E >f E . S X X X 7 . 5 , A A A 1:5 . 0 O « O 2 5 . 0 O O. O 3 5 . 0 Figure 3: Pine stems per ha versus plot stems per ha, all species, by distance from the stand/road edge (Edge Distance). Although there was a declining trend in stems per ha versus distance from the stand edge (Figure 4), the correlation was only significant for stems per ha relative to Zone 5 (Table 4). Significant differences in relative stems per ha between zones were not found (p=0.2683), when plot differences in basal area and site index were accounted for, as shown by the G L M output (Appendix II, Output 4). The logarithm of relative stems per ha was used to linearize the relationship for the G L M analysis. 51 S P H vs Ds tanoe from Stand /Road Edge Spec ies : PINE a 2000 I 25 DISTANCE FROM STAND/ROAD EDGE (m) LINE = Overall mean SPH from stand/road edge over all plots. STAR = SPH from stand/road edge of each plot. Figure 4: Pine stems per ha versus distance from the stand/road edge. 4.1.2.2 Basal Area per Hectare and Relative Basal Area per Hectare An increase in basal area per ha was noted (Table 2). No increase in pine basal area per tree was noted, however. A wide variability of basal area per ha is present over the 44 plots, as shown in Figure 5. i. 30 BA/WK vs Distance from S tand /Road E d g e Spec ies : PINE I 25 I 30 DISTANCE FROM STAND/ROAD EDGE Jn) LINE = Overall mean total BAyHA from stand/road edge over all plots. STAR = Total BA/HA from stand/road edge of each plot. I 35 40 Figure 5: Pine basal area per ha versus distance from the stand/road edge. 52 Controlling for differences in density and using weighted least squares because of unequal variances, differences in zonal basal areas were found (p=0.0001; Appendix II, Output 1). Zone 1 differed from all other zones, with no other pairs of zones being different. Differences in zonal basal area relative to Zone 5 were also detected (p=0.0036; Appendix II, Output 5). Again Zone 1 differed from other zones. 4.1.2.3 Curtis' Relative Density and Relative Curtis' Relative Density A higher mean Curtis' RD across all plots was found in the first zone from the stand/road edge, and the second zone appeared slightly higher than the interior of the plot (Figure 6). Relative Density Index vs Distance from S tand /Road E d g e Spec ies : PINE 40 - 1 1 1 1 I 10 E 20 25 30 OISTANCE FROM STAND/ROAD EDGE (m) LINE = Overall mean RD from stand/road edge over all plots. STAR = RD from stand/road edge of each plot. RD = «BA/HA>DBHq~0.5) Figure 6: Relative density index versus distance from the stand/road edge. Controlling for differences in density and site productivity and using weighted least squares because of unequal variances, differences in zonal Curtis' RD values were found (p=0.0001; Appendix II, Output 2). Zone 1 differed from other zones, with no other pairs of zones being significantly different. Differences in zonal Curtis' RD relative to Zone 5 were also detected (p=0.0059; Appendix II, Output 6). Zone 1 differed from Zones 3 and 4, and was marginally different from Zone 2 (p=0.0128 53 compared to a significance level of 0.0083 for six pairwise comparisons (SAS Institute Inc., 1989, p.943; Neter et. al., 1996, p.1096)). 4.1.2.4 Relative Mean Height Controlling for differences in density and site productivity, differences in zonal mean height relative to Zone 5 were not detected (p=0.2508; Appendix II, Output 7). 4.1.2.5 Mean Live Crown Ratio and Relative Mean Live Crown Ratio Controlling for differences in density and site productivity and using a logarithmic transformation to linearize the relationship, differences in zonal mean live crown ratios were found (p=0.0082; Appendix II, Output 3). Zone 1 differed from Zones 3 and 5, and was marginally different from Zone 4 also (p=0.0121 compared to a significance level of 0.005 for 10 comparisons). Differences in zonal mean live crown ratios relative to Zone 5 were also detected (p=0.0037; Appendix II, Output 8). Zone 1 differed from Zone 3 and Zone 4, with no other pairs of zones being different, indicating a more gradual decline in live crown ratio from the road edge, than with basal area per ha and Curtis' RD. 4.1.3 Mortality Zone level live tree (all species), dead standing tree and stump proportions by percent basal area and percent stems/ha are presented in Table 5. Over all 44 plots, Zone 1 has a lower proportion of dead standing trees than any other zones, using either basal area or number of stems per ha proportions. On average, 6.7% of the basal area is comprised of dead trees in Zone 1, compared to 9.9% in Zone 5. Zone 2 has nearly twice the proportion of standing tree mortality than Zone 1, with 12.8% of the basal area and 19.3% of the stems per ha comprised of dead trees, on average across all plots. Zones 3 and 4 are fairly similar whether comparing by basal area proportion or by the number of trees, with both zones being greater than Zone 5, but less than Zone 2. Zone 3 is more variable than Zone 4, however. Zone 1 and Zone 5 have similar standard deviations for the proportion of dead stems per ha, however Zone 1 has a lower standard deviation for the proportion of 54 dead trees by basal area. Zone 2 has the largest standard deviation for the proportion of dead stems per ha at 14.7%, while Zones 3 and 5 are similar at 11.3% and 11.5% respectively. The coefficient of variation (not shown, but calculated as standard deviation/mean) shows a decline with distance from the edge for the proportion of dead standing trees by stems per ha. A similar relationship exists for the proportion of dead standing tree basal area, although Zone 5 is slightly more variable than Zone 4. Some plots did not have dead trees in all five zones. The proportion of stumps in the plots was negligible, given that stands with excessive stumps were excluded from sampling. As expected, more stumps were present at the road edge than in the interior of the stand. Up to 12.5% of the basal area per ha was comprised of stumps in Zone 1 and up to 10% in Zone 2; however across the 44 plots only 1.6% and 0.6% of the average basal area reflected stump occurrence in these two zones. These mean values are quite variable however, as the coefficient of variation is lowest at the stand edge with a value of 205%. Zones 3 to 5 all had less than 0.1% of the basal area and stem proportions comprised of stumps, on average. For all zones, the mean DBHq for the live trees is greater than the mean DBHq for the dead standing trees, although there was more variability between the zones for the dead trees (coefficient of variation for all zones ranged from 34% in Zone 3 to 62% in Zone 1). In Zone 1, the DBHq for dead trees is lower, on average, then the DBHq of dead trees in the other five zones. The live tree DBHq is similar across all five zones and has similar variability (coefficient of variation from 23 to 26%). To explore the mortality issue further, the plots were classified into three road age classes (roads 20 years old or less, 21 to 40 and 41 to 60 years old) and the mortality proportions by basal area per ha were further examined (Table 6). Using the mean proportion of basal area per ha, the lowest proportion of dead standing basal area per ha was found in Zone 1, across all three road age classes. The variability for mean dead tree basal area proportion was consistently highest at the edge across the three age classes. Variability did not show a definitive trend across the edge interior gradient for all three road age classes, although it can be generalized that the variability was lowest in the middle of the edge interior gradient, then tended to increase by Zone 55 5. The mean proportion of the number of dead stems per ha was also lowest at the edge, and the variability pattern showed similar characteristics with basal area. Both of the mean density proportion statistics for stumps was highest in Zone 1, across all three road age classes. Stumps were not found in any plots in Zones 3 to 5 for the youngest road age class, nor in Zones 4 or 5 for the 21 to 40 year old road class. Stump proportions confound the mortality estimates however, as there was no evidence of the stumps being the result of harvesting live trees or from cutting dead standing trees. In most cases the bole was removed from the plot. The range from lowest to highest mean values of percent basal area for live trees across the five zones differed by 3.7% for the youngest road age class, by 12.4% for the 21 to 40 year old roads and by 8.0% for the roads greater than 40 years of age. Correspondingly, the range from lowest to highest mean values of percent live tree stems per ha across the five zones differed by 6.3% for the youngest road age class, by 14.2% for the 21 to 40 year old roads and by 3.8% for the roads greater than 40 years of age. The live tree mean percent basal area was highest in Zone 1 for the 21 to 40 and 41 to 60 year old road age classes, and these values corresponded with the lowest coefficients of variation for live trees, also. The 20 years old or less road age class had a live tree basal area proportion only 0.5% higher in Zone 5 than Zone 1, on average, and the coefficient of variation in Zone 5 was slightly lower (Zone 5 = 7.35%; Zone 1 = 7.91%). The mean proportion of the number of live trees per ha were highest in Zone 1 for all three road age classes. The lowest proportions of both percent live trees per ha and percent basal area per ha were found in Zone 2 in the youngest and oldest road age classes, and in Zone 4 in the 21 to 40 year road age class. The coefficient of variation for live tree basal area proportion was consistently lower then the coefficient of variation for stems per ha proportion, across all three road age classes and across all five zones. 56 Table 5: Zone level proportions (by %) of dead standing trees, stumps and live trees of all species based on the 44 sample plots. Quadratic mean diameter (DBHq) is calculated for the species of interest only. Stump basal area proportions will be slightly overestimated as the stump diameters were measured at stump height and were not extrapolated to diameters at breast height. Zone Variable Species N Minimum Maximum Mean Standard deviation 1 % of zone basal area/ha Dead 44 0.00 27.49 6.65 6.23 (2.5 m) % of zone basal area/ha Stumps 44 0.00 15.48 1.64 3.71 % of zone basal area/ha All live 44 72.51 100.00 91.71 6.51 % of zone stems/ha Dead 44 0.00 36.36 11.41 9.60 % of zone stems/ha Stumps 44 0.00 12.50 1.37 2.80 % of zone stems/ha All live 44 63.64 100.00 87.23 9.46 DBHq (cm) All live 44 10.77 32.98 18.28 4.76 DBHq (cm) Dead 44 0.00 27.21 10.35 6.39 2 % of zone basal area/ha Dead 44 0.00 39.01 12.78 10.68 (7.5 m) % of zone basal area/ha Stumps 44 0.00 9.53 0.57 1.81 % of zone basal area/ha All live 44 60.99 100.00 86.64 11.00 % of zone stems/ha Dead 44 0.00 50.00 19.31 14.71 % of zone stems/ha Stumps 44 0.00 10.00 0.71 2.10 % of zone stems/ha All live 44 50.00 100.00 79.98 14.94 DBHq (cm) All live 44 9.98 29.67 17.87 4.67 DBHq (cm) Dead 44 0.00 21.19 11.93 5.98 3 % of zone basal area/ha Dead 44 0.00 46.51 11.86 8.69 (15 m) % of zone basal area/ha Stumps 44 0.00 1.84 0.04 0.28 % of zone basal area/ha All live 44 53.49 100.00 88.10 8.71 % of zone stems/ha Dead 44 0.00 60.00 18.58 11.52 % of zone stems/ha Stumps 44 0.00 2.38 0.05 0.36 % of zone stems/ha All live 44 40.00 100.00 81.36 11.55 DBHq (cm) All live 44 9.75 29.01 17.54 4.27 DBHq (cm) Dead 44 0.00 22.99 12.72 4.31 4 % of zone basal area/ha Dead 44 0.00 37.96 12.42 8.12 (25 m) % of zone basal area/ha Stumps 44 0.00 0.65 0.01 0.10 % of zone basal area/ha All live 44 62.04 100.00 87.56 8.11 % of zone stems/ha Dead 44 0.00 45.16 18.83 11.28 % of zone stems/ha Stumps 44 0.00 2.50 0.06 0.38 % of zone stems/ha All live 44 54.84 100.00 81.12 11.26 DBHq (cm) All live 44 8.89 27.43 17.89 4.20 DBHq (cm) Dead 44 0.00 26.81 13.45 5.51 5 % of zone basal area/ha Dead 44 0.00 27.91 9.95 7.40 (35 m) % of zone basal area/ha Stumps 44 0.00 1.33 0.03 0.20 % of zone basal area/ha All live 44 72.09 100.00 90.02 7.45 % of zone stems/ha Dead 44 0.00 37.93 16.04 9.28 % of zone stems/ha Stumps 44 0.00 3.85 0.09 0.58 % of zone stems/ha All live 44 62.07 100.00 83.87 9.44 DBHq (cm) All live 44 8.62 27.81 18.17 4.57 DBHq (cm) Dead 44 0.00 23.64 12.35 4.83 57 Table 6: Proportions (by %) of zone basal area and stems per ha for dead standing, stumps and all live trees for three road age classes. Stump basal area estimates are slightly overestimated due to diameter recorded at stump height. Road Age < 20 years, n=26 Road age 21 to 40 years, n=9 Road age 41 to 60 years, n=9 Zone Variable Species Min Max Mean Std Dev Min Max Mean Std Dev Min Max Mean Std Dev 1 % of zone basal area/ha Dead 0.00 27.49 7.05 7.04 0.00 17.83 5.81 6.00 0.00 14.60 6.33 4.03 (2.5 m) % of zone basal area/ha Stumps 0.00 15.48 1.87 4.37 0.00 8.33 1.97 3.27 0.00 4.36 0.64 1.47 % of zone basal area/ha All live 72.51 100.00 91.07 7.21 75.63 98.63 92.22 7.08 85.40 96.39 93.03 3.42 % of zone stems/ha Dead 0.00 36.36 10.73 9.90 0.00 30.00 11.80 10.21 0.00 31.25 12.99 8.96 % of zone stems/ha Stumps 0.00 7.41 1.12 2.40 0.00 12.50 2.31 4.24 0.00 5.26 1.11 2.21 % of zone stems/ha All live 63.64 100.00 88.15 9.72 65.00 96.67 85.89 9.92 68.75 95.24 85.90 8.97 2 % of zone basal area/ha Dead 0.00 39.01 11.88 11.84 4.45 30.22 14.24 9.87 0.00 24.48 13.94 8.45 (7.5 m) % of zone basal area/ha Stumps 0.00 4.36 0.28 1.00 0.00 4.74 0.95 1.90 0.00 9.53 1.06 3.18 % of zone basal area/ha All live 60.99 100.00 87.84 12.20 69.78 95.55 84.81 10.28 75.52 100.00 85.01 8.23 % of zone stems/ha Dead 0.00 50.00 17.80 16.49 8.70 35.29 22.28 10.06 0.00 38.89 20.72 13.86 % of zone stems/ha Stumps 0.00 5.56 0.41 1.44 0.00 5.56 1.17 2.33 0.00 10.00 1.11 3.33 % of zone stems/ha All live 50.00 100.00 81.79 16.79 61.11 91.30 76.55 10.33 61.11 100.00 78.17 13.59 3 % of zone basal area/ha Dead 0.00 24.28 10.93 7.09 4.58 46.51 16.81 13.63 3.64 19.30 9.60 5.19 (15 m) % of zone basal area/ha Stumps 0.00 0.00 0.00 0.00 0.00 1.84 0.20 0.61 0.00 0.00 0.00 0.00 % of zone basal area/ha All live 75.72 100.00 89.07 7.09 53.49 95.42 82.99 13.61 80.70 96.36 90.40 5.19 % of zone stems/ha Dead 0.00 41.67 17.11 10.79 9.80 60.00 25.41 15.48 4.76 24.14 16.00 6.56 % of zone stems/ha Stumps 0.00 0.00 0.00 0.00 0.00 2.38 0.26 0.79 0.00 0.00 0.00 0.00 % of zone stems/ha All live 58.33 100.00 82.89 10.79 40.00 90.20 74.32 15.47 75.86 95.24 84.00 6.56 4 % of zone basal area/ha Dead 0.50 23.87 11.16 6.51 1.92 37.96 20.14 10.06 0.00 14.61 8.34 5.46 (25 m) % of zone basal area/ha Stumps 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.65 0.07 0.22 % of zone basal area/ha All live 76.13 99.50 88.84 6.51 62.04 98.08 79.86 10.06 85.39 100.00 91.59 5.43 % of zone stems/ha Dead 2.22 38.24 17.12 10.14 6.67 45.16 28.18 12.47 0.00 26.67 14.41 8.80 % of zone stems/ha Stumps 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.50 0.28 0.83 % of zone stems/ha All live 61.76 97.78 82.88 10.14 54.84 93.33 71.82 12.47 73.33 100.00 85.31 8.77 5 % of zone basal area/ha Dead 0.00 25.65 8.47 6.73 1.22 25.17 12.08 7.80 0.38 27.91 12.10 8.63 (35 m) % of zone basal area/ha Stumps 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.33 0.15 0.44 % of zone basal area/ha All live 74.35 100.00 91.53 6.73 74.83 98.78 87.92 7.80 72.09 99.62 87.75 8.81 % of zone stems/ha Dead 0.00 36.00 13.99 8.87 4.76 37.93 20.54 10.11 2.94 30.77 17.49 8.72 % of zone stems/ha Stumps 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.85 0.43 1.28 % of zone stems/ha All live 64.00 100.00 86.01 8.87 62.07 95.24 79.46 10.11 65.38 97.06 82.08 9.51 58 4.1.4 Damage Indicators Damage indicators were prevalent across the 44 sample plots (Table 7). The most prevalent damage agents found on lodgepole pine were mistletoe (Arceuthobium americanum Nutt. ex Engelm.), abiotic scar damage and mountain pine beetle (Dendroctonus ponderosae Hopkins). Dead or broken tops, forks or crooks on the bole and the presence of frost cracks were also common. Many different combinations of these and other low occurrence damage types were often present on a single tree. In total, there were 64 unique combinations of damage indicators found on the live sampled trees of all species. Table 7: Presence of damage indicators as a percent of all live tree stand basal area across the 44 sample plots. Percentages are for occurrence of each single damage indicator. Trees may have more than one damage indicator. Damage indicator N Minimum Maximum Mean Standard deviation Mistletoe 44 0.00 96.65 51.64 34.93 Scar damage 44 3.87 63.30 33.11 15.39 Mountain pine beetle 44 0.00 98.51 23.05 23.85 Fork or crook 44 5.45 39.34 16.64 7.48 Dead or broken top 44 2.57 29.79 12.45 7.09 Frost crack 44 0.00 25.82 7.08 6.67 No damage indicators 44 0.25 77.07 21.52 20.46 Some trees showed no damage types, but their incidence comprised less than 22% of plot level live tree basal area, on average (Table 7). Table 8 shows the proportion across zones of lodgepole pine basal area and stems per ha represented by trees without damage indicators. Zones 1 and 5 have similar proportions, as do Zones 2 and 3. Zone 4 showed the lowest incidence of healthy trees, on average. Zone 3 showed the lowest variability, on average. 59 Table 8: Damage free lodgepole pine percent basal area and percent stems per ha, within each zone. Percents are a proportion of live lodgepole pine total basal area and stems per ha only. Zone Variable N Min Max Mean Std dev 1 % of zone basal area/ha 44 0.00 92.51 16.00 23.66 (2.5 m) % of zone stems/ha 44 0.00 88.89 17.41 22.32 2 % of zone basal area/ha 44 0.00 87.07 19.18 25.28 (7.5 m) % of zone stems/ha 44 0.00 90.91 20.31 25.73 3 % of zone basal area/ha 44 0.00 79.17 19.07 23.06 (15 m) % of zone stems/ha 44 0.00 77.78 19.76 21.73 4 % of zone basal area/ha 44 0.00 89.98 13.82 20.34 (25 m) % of zone stems/ha 44 0.00 90.00 15.47 20.17 5 % of zone basal area/ha 44 0.00 86.17 16.37 22.96 (35 m) % of zone stems/ha 44 0.00 86.05 17.08 23.17 As it is expected that damage indicators are more likely to occur as trees age, the plots were classified into three age class groups based on the average age of the stand; 35 to 65 years (n=16), 65 to 95 years (n=15) and 95 to 125 years (n=13). Table 9 presents the most common damage indicators by percent of basal area per ha, for the three stand age classes. Basal area proportions are for the presence of the single damage indicator on each tree, therefore values cannot be summed across damage agents as many trees had more than one damage indicator. 4.1.4.1 Mistletoe Damage Mistletoe was the most prevalent damage indicator across the 44 sampled plots and across all zones (Table 7 and Table 9). In both the 80 and 110 year old stand age classes, the lowest basal area per ha proportion of mistletoe damage was found at the stand edge, on average. The Zone 1 proportions in these age classes were the most variable, however. In the 50 year stand age class, Zone 2 had the lowest proportion of zone basal area with mistletoe damage indicators followed by Zone 1, and variability in the 50 year stand age class followed the same trend. Generally, the proportion of the stand infected with mistletoe increased with distance from the two zones closest to the edge. Variability was lowest in Zone 4 for the 50 and 80 year old classes, and Zone 5 for the 110 year old class. Across all age classes, while some zones did not have any mistletoe indicators present, the maximum proportion of mistletoe infected zones was at or near 100%. 60 Table 9: Damage agent proportions, by zone, shown as a percent of lodgepole pine basal area for each damage agent by three stand age classes; age class 50 (ages 35 to 65); age class 80 (ages 65.1 to 95) and age class 110 (ages 95.1 to 125). Stand age class 50 (n=16) Stand age class 80 (n=15) Stand age class 110 (n=13) Damage agent Zone Min Max Mean Std dev Min Max Mean Std dev Min Max Mean Std dev Mistletoe 1 (2.5 m) 0.00 100.00 49.71 40.95 0.00 100.00 48.62 45.14 0.00 98.08 46.67 37.82 2 (7.5 m) 0.00 97.17 44.75 42.23 0.00 100.00 56.67 40.84 0.00 100.00 57.65 41.69 3 (15 m) 0.00 98.74 51.69 38.49 0.00 98.14 58.44 37.15 0.00 99.53 60.36 41.56 4 (25 m) 0.00 100.00 60.16 35.33 0.00 100.00 59.97 38.10 0.00 100.00 60.63 42.06 5 (35 m) 0.00 100.00 61.28 40.59 0.00 100.00 55.32 38.21 0.00 100.00 62.70 42.19 Scar damage 1 (2.5 m) 0.00 83.06 40.50 26.34 7.16 93.13 44.61 26.10 4.34 100.00 42.13 25.25 2 (7.5 m) 0.00 55.91 24.40 19.81 7.62 84.89 36.58 24.06 7.52 76.55 34.82 20.39 3 (15 m) 0.00 69.30 35.93 25.90 10.19 90.71 41.43 20.33 4.31 54.98 28.09 16.99 4 (25 m) 4.31 72.46 37.04 20.44 0.00 84.33 40.32 25.12 0.00 84.96 28.50 22.89 5 (35 m) 0.00 65.16 27.99 19.77 6.69 78.76 36.69 18.73 0.00 67.15 28.89 19.58 Mountain pine beetle 1 (2.5 m) 0.00 70.32 18.33 24.01 0.00 87.86 29.25 29.55 0.00 100.00 33.21 37.91 2 (7.5 m) 0.00 84.00 15.64 26.87 0.00 75.92 23.91 26.96 0.00 100.00 35.19 37.47 3 (15 m) 0.00 73.07 13.01 19.30 0.00 76.33 24.31 26.98 0.00 100.00 29.47 30.37 4 (25 m) 0.00 89.81 18.13 23.87 0.00 70.36 25.47 24.81 0.00 95.17 36.00 32.75 5 (35 m) 0.00 92.31 21.92 28.42 0.00 91.41 25.74 25.59 0.00 100.00 31.54 32.13 Dead or broken top 1 (2.5 m) 0.00 43.29 15.08 16.16 0.00 46.84 14.98 14.57 2.09 43.51 19.89 11.64 2 (7.5 m) 0.00 37.77 8.43 10.66 0.00 30.67 12.60 11.36 0.00 34.40 17.44 9.82 3 (15 m) 0.00 25.14 8.79 6.70 0.00 31.25 12.80 11.89 8.17 29.90 17.25 6.30 4 (25 m) 0.00 39.58 10.33 13.11 0.00 36.32 15.36 12.66 0.00 24.92 14.96 7.86 5 (35 m) 0.00 31.78 10.80 8.84 0.00 33.77 14.19 10.59 0.00 21.24 11.47 7.04 Fork or crook 1 (2.5 m) 0.00 46.26 15.52 12.75 0.00 41.71 14.98 12.64 0.00 41.52 11.06 13.47 2 (7.5 m) 0.00 62.48 20.44 18.36 0.00 48.36 17.86 12.74 0.00 36.83 19.46 11.58 3 (15 m) 1.28 36.16 17.94 10.38 0.00 67.29 18.70 17.09 0.00 25.73 10.39 8.52 4 (25 m) 1.10 51.89 23.98 14.87 0.00 62.98 22.56 16.05 3.78 32.94 16.34 10.91 5 (35 m) 1.72 46.51 18.37 10.92 0.00 39.96 15.84 11.84 0.00 24.97 14.68 8.22 Frost crack 1 (2.5 m) 0.00 80.20 11.59 21.19 0.00 25.81 6.23 8.07 0.00 69.02 8.59 19.94 2 (7.5 m) 0.00 33.20 6.27 9.45 0.00 48.59 7.93 13.10 0.00 9.33 1.39 3.40 3 (15 m) 0.00 45.69 12.00 13.50 0.00 12.30 3.71 4.24 0.00 21.82 5.26 7.08 4 (25 m) 0.00 44.05 13.61 14.09 0.00 19.24 6.79 6.64 0.00 27.01 2.60 7.45 5 (35 m) 0.00 33.87 10.52 11.69 0.00 30.58 4.88 8.25 0.00 15.43 2.73 4.68 61 4.1.4.2 Scar Damage Scar damage was most prevalent at the stand edge across all three age classes, on average (Table 9). In the two youngest age classes, Zone 2 showed the lowest proportion of basal area affected by scar damage. In the oldest age class the proportion of the stand affected by scars was lowest, and similar, between Zones 3, 4 and 5 (mean percentages ranged from 28.1 to 28.9%). Stand age classes 50 and 110 had a similar range of variability across their respective zones, while the range of variability was narrower for the zones in age class 80. 4.1.4.3 Mountain Pine Beetle Mountain pine beetle showed the clearest trend in the 80 year old stand age class, with Zone 1 having the highest proportion, dropping to the lowest in Zone 2, and then increasing with distance into the interior of the stand (Table 9). Variability was similar for Zones 1, 4 and 5 in the 80 year age class. In the 50 year stand age class, Zone 1 had a lower mean percentage of the stand displaying beetle damage than Zone 5, but higher than Zones 2 to 4. The 50 year stand age class had high variability across all zones. Stands in the age class 110 showed Zone 1 to have a higher proportion of beetle damaged trees than Zones 3 and 5, but less than Zones 2 and 4. The range of variability across zones was similar for age classes 80 and 110, however the overall variability was lowest for the oldest age class. The range of mean basal area proportions across the five zones for the 50 year age class was 13.0 to 21.9%; for the 80 year class 24.3 to 29.2% and for the oldest age class 29.5 to 36.0%. It appears that beetle damage indicators are more prevalent on older stands. 4.1.4.4 Forks or Crook Damage Fork or crook damage mean percent basal area was lowest in Zone 1 for the two younger stand age classes, and only slightly higher than Zone 3 in the 110 year age class (Table 9). The Zone 1 mean values were highly variable in the oldest age class, however. There was no clear trend across any of the age classes for the mean values, beyond the low mean proportion in Zone 1. 62 4.1.4.5 Dead or Broken Top Damage Dead or broken top trees were most prevalent by percent of zone basal area at the stand edge, on average, in the youngest and oldest age classes (Table 9). In the 110 year stand age class, there was a declining trend with distance from the edge. In the 50 year stand age class, the proportion by basal area with dead or broken tops was highest in Zone 1, lowest in Zone 2 and increased towards the interior of the stand. Variability was highest in the youngest age class and lowest in the oldest age class. Zone 3 was only marginally higher than Zone 2 in the two youngest age classes. In the 80 year age class, Zone 4 had a slightly higher proportion of dead and damaged tops than Zone 1, followed by Zone 5. Overall variability was lowest in the oldest age class and highest in the youngest age class. 4.1.4.6 Frost Crack Damage There was no clear relationship with distance from the edge and frost crack damage. Over all plots, the trees showing indicators of frost crack damage comprised only 7% of the total live tree basal area, and this figure was variable (Table 7). Only the plots in the oldest age class showed the highest basal area proportion of lodgepole pine trees at the stand edge (Table 9), but the highest basal area proportions of frost damaged trees were found in Zone 4 in the 50 year class and Zone 2 in the 80 year class. For all age classes, there was great variability across all zones. 4.1.5 Canopy Layers The proportion of zone lodgepole pine basal area per ha in one of five tree canopy layers is presented in Table 10, under the same road age classes used in Table 6. Plots in the two youngest road age classes (< 20 years; 21 to 40 years) had the largest proportion of stand basal area represented by the codominant canopy layer, on average. In the 41 to 60 year road age class, only Zone 1 had a larger proportion of zone basal area in the codominant layer. The dominant layer comprised a larger proportion of zone basal area than any other layers in Zones 2 to 5. 63 Table 10: Proportions (by %) of zone lodgepole pine basal area by tree canopy layer, for three road age classes: less than 20 years (n=26); 20.1 to 40 years (n=9); and 40.1 to 60 years (n=9). Mean values for each zone and age class sum to 100%. Road Age < 20 years, n=26 Road age 21 to 40 years, n=9 Road age 41 to 60 years, n=9 Zone Canopy Layer Min Max Mean Std Dev Min Max Mean Std Dev Min Max Mean Std Dev 1 Dominant 0.00 81.43 36.49 19.77 0.00 65.09 38.92 21.01 9.93 60.04 33.31 14.14 (2.5 m) Codominant 0.00 73.22 41.78 18.14 29.79 83.95 51.29 15.81 6.79 75.38 40.61 18.56 Intermediate 0.00 43.53 17.91 13.71 0.00 41.61 8.53 13.05 1.90 38.82 19.91 11.84 Suppressed 0.00 18.92 3.83 5.42 0.00 6.27 1.26 2.51 0.00 30.47 6.17 10.24 Veteran 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2 Dominant 0.00 90.88 32.14 27.57 0.00 60.97 26.71 24.83 21.74 100.00 49.79 27.30 (7.5 m) Codominant 5.56 100.00 48.71 23.47 14.08 96.74 59.04 28.74 0.00 76.45 37.07 26.78 Intermediate 0.00 46.17 16.65 14.58 0.00 40.33 13.89 13.50 0.00 20.75 11.52 7.34 Suppressed 0.00 12.57 2.50 3.48 0.00 3.26 0.36 1.09 0.00 9.04 1.62 3.07 Veteran 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 Dominant 0.00 69.89 36.99 18.72 0.00 67.06 34.17 24.73 7.02 82.48 45.09 23.34 (15 m) Codominant 16.89 84.53 42.91 17.76 14.26 90.82 45.93 25.83 14.92 66.85 36.54 16.98 Intermediate 0.00 53.26 16.76 14.88 5.11 29.52 18.71 9.64 0.00 33.79 17.01 10.09 Suppressed 0.00 13.62 3.34 4.22 0.00 5.19 1.19 1.85 0.00 5.04 1.35 1.60 Veteran 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4 Dominant 0.00 69.46 35.04 19.63 5.69 66.73 34.34 22.20 0.00 71.85 39.17 23.66 (25 m) Codominant 14.87 84.92 46.77 18.71 14.53 89.69 48.14 26.22 26.96 48.48 36.40 7.80 Intermediate 0.00 47.34 14.04 9.99 3.35 28.35 15.97 9.29 0.00 42.71 17.10 13.10 Suppressed 0.00 12.11 1.68 3.27 0.00 6.02 1.56 2.19 0.00 13.82 4.23 4.91 Veteran 0.00 63.96 2.46 12.54 0.00 0.00 0.00 0.00 0.00 27.88 3.10 9.29 5 Dominant 0.00 82.80 31.51 23.14 0.00 71.26 33.99 27.21 0.00 76.06 38.33 26.23 (35 m) Codominant 4.77 91.91 50.18 21.71 23.58 87.77 47.26 21.25 20.65 50.23 36.04 10.48 Intermediate 1.27 38.44 14.97 10.11 0.00 40.87 16.28 13.20 0.00 51.60 19.61 17.39 Suppressed 0.00 24.40 2.68 5.03 0.00 7.66 2.47 3.24 0.00 15.94 4.22 4.75 Veteran 0.00 17.15 0.66 3.36 0.00 0.00 0.00 0.00 0.00 16.18 1.80 5.39 64 In the < 20 year road age class, there was a larger proportion of zone basal area comprised of the intermediate layer at the edge, in Zone 1. There was a declining relationship with the proportion of intermediate basal area and distance from the edge. The 21 to 40 year age class had the lowest proportion of intermediate basal area in Zone 1 followed by Zone 2, but the highest proportion in Zone 3. The 41 to 60 year age class showed the highest intermediate canopy layer basal area proportions in Zone 1, with very similar proportions in Zone 5. Zone 2 had the lowest proportion of intermediate basal area, followed by nearly identical mean values in Zones 3 and 4. In the < 20 year road age class, there was little trend in the suppressed canopy layer across the plot edge to interior gradient. Zone 1 had the highest proportion of basal area in the suppressed layer of all other five zones in this road age class. In the 21 to 40 year age class the highest proportion of suppressed layer zone basal area was found in the stand interior, with Zone 2 having the lowest proportion. The 41 to 60 year age class had the highest proportion of zone basal area comprised of suppressed trees in Zone 1, followed by nearly identical average values for Zones 4 and 5 (Zone 1=6.2%; Zone 5 = 4.2%). Variability was high across all three age classes, for the suppressed canopy layer. Veteran trees were only found in the interior of the stand, but this was highly variable. Veteran trees were not found in the 21 to 40 year road age class. 4.2 T R E E B A S A L A R E A INCREMENT DATA 4.2.1 Descriptive Statistics Graphs of the tree basal area increment values relative to the 5 year periodic annual increment (p.a.i.) previous to the road existence averaged over the three trees sampled from each zone/plot indicated that, generally, Zone 1 had higher relative basal area increments (Appendix III). There was much plot to plot variability, however. Three of the plots were included in the graphs shown in Appendix III, but were not included in subsequent analysis, since all randomly selected trees in Zone 1 were either younger 65 than, or very near (< 3 years) the road age. Also, an additional plot was removed from these analyses, as the road age was less than 10 years. Correlations between relative tree basal area p.a.i. values and distance from the road edge were significantly negative for the three periods beginning three years after road building, (3-5, 6-10, 11-15), up to 15 years after building the road (Table 11). Road width showed significant negative correlations for up to 30 years after harvesting, and road age showed significant positive correlations up to 30 years after harvesting. Stand age, estimated for the year the road right-of-way was harvested, showed significant negative correlations with relative tree basal are increment for the 1-2 year period and all periods from 6 to 20 years after road construction. The edge aspect of the stand was not significantly correlated with relative tree basal area increment for any period after building the road. Table 11: Correlations between annual tree basal area increment relative to the five years prior to the road, for 1 to 2 years, 3 to 5 years, and subsequent five-year intervals after the road right-of-way was harvested, with distance from the road edge, road right-of-way width, road age, stand edge aspect and stand age at the time of harvest. Correlations significantly different from zero for a = 0.05 are identified with *. Period after road N Edge Road Road Stand Edge Stand age at right-of-way harvest distance width Age Aspect r/w harvest 1 to 2 years 607 -0.07672 -0.08313* 0.08211* 0.04480 -0.11420* 3 to 5 years 607 -0.15508* -0.10738* 0.10971* 0.02698 -0.03314 6 to 10 years 607 -0.15989* -0.13066* 0.13385* 0.02515 -0.09138* 11 to 15 years 592 -0.11898* -0.21401* 0.20929* -0.04276 -0.14653* 16 to 20 years 517 -0.08540 -0.23891* 0.26438* -0.04068 -0.12032* 21 to 25 years 222 -0.01498 -0.22496* 0.14783* -0.06167 -0.03716 26 to 30 years 147 -0.05943 -0.16970* 0.21442* -0.06031 -0.21539* 31 to 35 years 132 -0.05882 -0.12901 0.15145 -0.02057 -0.18199* 36 to 40 years 88 -0.15800 -0.27814 -0.10622 -0.00231 -0.01782 41 to 45 years 88 -0.17448 -0.14397 -0.05856 -0.00660 -0.01801 46 to 50 years 44 -0.13974 -0.22120 -0.19074 0.15077 -0.11912 Average annual increments of tree basal area for the five year period prior to the road are presented in Table 12. While Zone 1 is more variable than the other zones, all zones were quite variable. Zone 1 had the highest maximum value. 66 Table 12: Mean tree basal area periodic annual increment (shown in square centimetres per year) by zone for the 5-year period prior to road establishment. Up to 3 sub-sampled trees were sampled in each zone within each plot. Zone N Minimum (cm2/year) Maximum (cm2/year) Mean (cm2/year) Standard deviation (cm2/year) 1 (2.5 m) 119 0.33 14.48 2.98 2.40 2 (7.5 m) 123 0.31 12.40 3.00 2.05 3 (15 m) 122 0.36 10.80 2.93 2.23 4 (25 m) 121 0.23 12.97 2.96 2.13 5 (35 m) 122 0.42 12.44 3.28 2.18 Descriptive statistics for relative tree basal area periodic annual increment values by zone and period indicated larger mean and maximum values for Zone 1 in comparison to the other zones (Table 13). For Zones 3 and 4, there was an overall decline in relative tree basal area p.a.i. with increasing time since the road right-of-way was harvested, corresponding with an increase in tree ages. Zone 2 showed a slight increase in relative p.a.i. for the first 2 years, then remained fairly constant until the decline in the 16-20 year period. For Zone 1, the average tree basal area periodic annual increment, relative to the 5 year period prior to the road, was higher than all other zones and was greater than the period prior to the road for all five periods. Zone 1 showed a dramatic increase in the mean relative annual increment until the 11-15 year period, when the relative annual tree basal area increment began to decline. The variability was much higher in Zone 1 for each period than for all other zones, however. 67 Table 13: Relative tree basal area periodic annual increment (p.a.i.) by zone and period. Values are relative to the p.a.i. for the 5-year period prior to road establishment. ZONE Period after road N Minimum Maximum Mean Standard establishment Deviation 1 1 to 2 years 119 0.338 2.589 1.099 0.354 (2.5 m) 3 to 5 years 119 0.360 3.740 1.247 0.634 6 to 10 years 119 0.454 4.765 1.359 0.832 11 to 15 years 116 0.242 5.849 1.329 0.845 16 to 20 years 101 0.212 5.387 1.156 0.892 2 1 to 2 years 123 0.236 1.850 1.044 0.284 (7.5 m) 3 to 5 years 123 0.318 2.507 0.994 0.387 6 to 10 years 123 0.284 3.064 1.010 0.447 11 to 15 years 120 0.232 3.110 0.967 0.439 16 to 20 years 105 0.200 2.434 0.877 0.459 3 1 to 2 years 122 0.253 1.799 0.963 0.243 (15 m) 3 to 5 years 122 0.235 2.252 0.918 0.313 6 to 10 years 122 0.233 3.259 0.960 0.454 11 to 15 years 119 0.268 4.296 0.968 0.531 16 to 20 years 104 0.129 4.136 0.865 0.578 4 1 to 2 years 121 0.438 1.751 0.991 0.274 (25 m) 3 to 5 years 121 0.309 1.595 0.935 0.278 6 to 10 years 121 0.244 2.473 0.930 0.374 11 to 15 years 118 0.239 3.924 0.928 0.477 16 to 20 years 103 0.169 4.098 0.820 0.511 5 1 to 2 years 122 0.458 1.666 1.030 0.256 (35 m) 3 to 5 years 122 0.447 1.823 1.000 0.301 6 to 10 years 122 0.254 2.477 1.029 0.419 11 to 15 years 119 0.207 2.677 1.042 0.493 16 to 20 years 104 0.172 3.434 0.947 0.543 Figure 7 presents the mean values shown in Table 13 for relative tree basal area periodic annual increment, relative to the 5 year period prior to road construction. On average, there is a growth increase in Zone 1 relative to the growth rate prior to road establishment, for the first 20 years. 68 M e a n Re la t i ve Tree Basa l A r e a Per iod ic A n n u a l Increment, b y Z o n e PAI is Relat ive to the PM for the 5 - y e a r P e r i o d Pr ior to R o a d R / W Harves t 15 1.4 -1.3 -CT * O 1.2 -s c 1.1 -* S a ° i -1.0 -0.9 -0.8 -m <» B x A A & ? 0 0 A O X A O I I I I I - 5 0 S 10 15 Period (Midpcint Year) l 20 P _ Z O N E * * * 1 X X X Z A A A 3 O O O H O O N is different between periods due to Road A g e o s Figure 7: Mean relative tree basal area periodic annual increment ratio, by zone, for lodgepole pine. Values are relative to the tree basal area p.a.i. for the 5-year period prior to road establishment. 4.2.2 Testing for Differences Among Zones Controlling for the basal area per ha of the plot and the road age, a multiple analysis of covariance using relative periodic tree basal area increment values for periods 1 to 2, 3 to 5, 6 to 10, 11 to 15, and 16 to 20 years following road building, indicated significant differences among zones (all tests had p=0.0012 or less; Appendix IV, Output 6). This was also confirmed by univariate analysis of covariance for each of the five periods (Appendix IV, Outputs 1 to 5). All periods showed differences among zones, while controlling for basal area per ha and road age differences (p=0.0131 or less). Except for the period from 1 to 2 years following the road, a logarithmic transformation resulted in a linear model with a normal distribution of error terms. For the period from 1 to 2 years, the model was linear, but errors were not normally distributed; however, the number of observations was high, resulting in a more sensitive normality test, and linear models are somewhat robust to non-normality (Neter et al., 1996, p. 54). Basal area per ha and road age were significant variables for all periods. 69 In pairwise comparisons of zones, the relative tree basal area p.a.i for Zone 1 differed from Zone 3, with no other zones being different for the period from 1 to 2 years (Appendix IV, Output 1). After the first two year period, the relative tree basal area p.a.i for Zone 1 was the largest and significantly different from other zones. This continued until the 16 to 20 year period following the road, where Zone 1 differed from Zones 3 and 4 (Appendix IV, Outputs 2 through 5). 4.2.3 Testing for Differences in Bole Basal Area Growth due to Road Orientation Radial increment cores were sampled both perpendicular (X) and parallel (W) to the road to examine any differences in bole shape due to the road opening. Figure 8 shows a slight increase in the relative tree basal area p.a.i. for the direction parallel to the road for Zone 1 for all periods to the end of 20 years after road establishment. Mean Relative Tree Basal Area Periodic Annual Increment, by Zone and Core Direction PAI Ratio is Relative to the PM for the 5—year Period Prior to Road R/W Harvest 1.5 H 1.3 H 15 20 Period (Midpoint Year) Zone & C o r e * * * A A A ltd 3X .* ;t * o o o 1 X H U x x x eu o « o H X x x x ex o o o 5ui A A A 3 Id O O O 5X N is different between periods due to Road Age Figure 8: Mean relative tree basal area periodic increment, by zone and core direction for the 20 years after road establishment. Values are relative to the tree basal area p.a.i. for the 5-year period prior to road establishment. Cores were sampled parallel (W) and perpendicular (X) to the road. 70 While the Zone 1 W relative basal area increments were slightly larger on average, the Zone 1 X relative basal area increments had greater variability and a wider range of values (Table 14) across all periods. Trees in Zones 2 to 5 did not show the same pattern; instead, their range of relative basal area increments, mean values and the standard deviations appear to be fairly consistent between the two radial measure directions for all periods. At most, there was a 9.8% difference between the coefficients of variation for the respective directions for any one period. Both radial directions in Zone 1 showed an increasing trend in the relative tree basal area increment for the first 15 years, followed by a decline during the 16 to 20 year period. Table 14: Relative tree basal area periodic annual increment for the two radial increment core directions: parallel (W) and perpendicular (X) to the road opening. Values are ratios relative to the tree basal area p.a.i. for the 5-year period prior to road establishment. Parallel to road (W) Perpendicular to road (X) Zone Period after road N Min Max Mean SD N Min Max Mean SD 1 1 to 2 years 118 0.237 2.738 1.161 0.425 118 0.187 2.528 1.061 0.427 (2.5 m) 3 to 5 years 118 0.361 3.496 1.294 0.685 118 0.301 6.480 1.272 0.879 6 to 10 years 118 0.327 4.936 1.426 0.883 118 0.312 9.295 1.378 1.177 11 to 15 years 115 0.345 4.587 1.385 0.833 115 0.172 9.924 1.342 1.175 16 to 20 years 100 0.158 6.963 1.208 0.975 100 0.186 7.915 1.133 1.023 2 1 to 2 years 123 0.319 2.185 1.025 0.341 123 0.202 2.421 1.073 0.356 (7.5 m) 3 to 5 years 123 0.238 2.812 1.004 0.424 123 0.267 4.044 1.024 0.525 6 to 10 years 123 0.311 3.742 1.022 0.506 123 0.263 3.207 1.039 0.513 11 to 15 years 120 0.275 3.303 0.983 0.485 120 0.199 4.200 0.989 0.538 16 to 20 years 105 0.116 3.866 0.904 0.556 105 0.136 3.030 0.919 0.525 3 1 to 2 years 122 0.237 1.618 0.992 0.289 121 0.259 2.006 0.959 0.297 (15 m) 3 to 5 years 122 0.206 2.487 0.964 0.375 121 0.270 3.502 0.923 0.415 6 to 10 years 122 0.229 3.197 1.019 0.531 121 0.220 3.624 0.949 0.524 11 to 15 years 119 0.245 4.646 1.026 0.606 118 0.213 3.898 0.955 0.576 16 to 20 years 104 0.134 5.029 0.920 0.705 103 0.124 3.497 0.861 0.575 4 1 to 2 years 121 0.239 2.342 0.989 0.332 121 0.252 2.087 1.008 0.321 (25 m) 3 to 5 years 121 0.256 1.843 0.945 0.355 121 0.287 2.110 0.963 0.349 6 to 10 years 121 0.186 2.606 0.950 0.450 121 0.219 2.375 0.958 0.409 11 to 15 years 118 0.123 4.056 0.948 0.551 118 0.237 3.847 0.957 0.506 16 to 20 years 103 0.079 3.645 0.831 0.561 103 0.183 4.362 0.839 0.529 5 1 to 2 years 121 0.291 2.731 1.058 0.332 122 0.411 2.197 1.037 0.332 (35 m) 3 to 5 years 121 0.384 2.740 1.011 0.347 122 0.233 2.765 1.035 0.413 6 to 10 years 121 0.320 3.515 1.040 0.489 122 0.213 2.328 1.061 0.478 11 to 15 years 118 0.251 4.258 1.032 0.591 119 0.180 3.712 1.112 0.589 16 to 20 years 103 0.175 3.024 0.923 0.560 104 0.170 3.262 0.995 0.589 71 Core direction influence was further explored through modelling. Controlling for zone, plot basal area per ha and road age, the five periods (up to 20 years) after the road were analysed with G L M to test for differences in relative tree basal area increment due to core direction. Using a logarithmic transformation to linearize the relationship, core direction was not found to be a significant variable for any of the five periods (all tests had p=0.4343 or more; Appendix VI, Outputs 1 to 5). The lack of significance for core direction was further confirmed with the least squares means test. The model errors for periods 1 to 2 and 3 to 5 years after the road were not normally distributed but the models were linear, and the number of observations was high. Assumptions of regression were met for all other periods. An additional test was made by including the zone and core direction interaction in the model. Neither the zone and core direction interaction variable, nor the core direction variable were significant in any of the five periods after road building. Pairwise testing of the zone and core direction interaction using the least mean squares procedure confirmed this lack of significance in the difference between the radial directions within each zone. As there were no significant differences in tree basal area increment growth due to the sampled core direction relative to the road opening, all subsequent tree basal area increment analysis was conducted using the average tree basal area increment. 4.2.4 Testing for Differences in Mean Tree Basal Area Among Zones Mean tree basal area calculated from the radial increment cores at 2, 5, 10, 15 and 20 years following road building was not found to be significantly correlated with distance from the edge for any of the five periods (all five periods had positive correlations of r=0.02833 or less and p=0.4861 or greater based on n=607 trees for periods 2, 5 and 10; n=592 trees for period 15 and n=517 trees for period 20). Controlling for the plot age at the time of right-of-way clearing, site index and Curtis' relative density of the plot, and using a multiple analysis of covariance test, tree basal area at the end of 2, 5, 10, 15 and 20 year periods following road building was not found to be significantly different between zones (all tests had p=0.0626 or greater; 72 Appendix VIII, Output 6). This was confirmed with univariate analysis of covariance for each of the five periods (Appendix VIII, Outputs 1 to 5). Using a logarithmic transformation to normalize the error terms, all five periods showed no significant differences between the zones while controlling for plot age at the time of right-of-way clearing, site index and Curtis' relative density of the plot (p=0.2650 or greater). All models were linear and errors were normally distributed, and all variables other than zone were significant for all periods. Zone was retained in the models for each of the periods for consistency, with a loss of 4 degrees of freedom. The pairwise comparison between zones found no significant differences in mean tree basal area between any of the zones, for any of the periods (using a significance level of 0.005 for 10 pairwise tests, all tests had p=0.0405 or greater; Appendix VIII, Outputs 1 to 5). 4.3 PREDICTING RELATIVE BASAL A R E A P E R H E C T A R E AND RELATIVE T R E E BASAL A R E A PERIODIC ANNUAL INCREMENT IN Z O N E 1 The basal area per ha in Zone 1 relative to Zone 5 ranged from 0.230 to 2.693, with an average value of 1.313 (Table 3). A 95% confidence interval for this average based on the 44 plots is 1.142 to 1.483. Given that Zone 1 was 5 m wide, this would translate into 1.56 m of the road (0.71 to 2.42 m using the confidence interval) that is offset from the increase in basal area for an average stand. For the average road width of 23 m sampled in this survey (Table 1), this would translate into 13.6% of the road width (6.2 to 21.0% using the confidence interval), if both sides of the road were similarly impacted. This estimate was based on stands that were relatively free of management activities, and free of other edge impacts, such as adjacent harvesting, however. Also, 15 of the 44 plots had a relative basal area for Zone 1 of less than 1.0. Of the remaining plots, 13 had relative basal areas from 1.0 to 1.5, and 16 had relative basal areas of more than 1.5. In order to better estimate the impacts of roads on the basal area of the first 5 m into the stand, the Zone 1 basal area data relative to Zone 5 was further examined. The correlations with plot variables were -0.33397 with plot basal area per ha, -0.12871 73 with site index, -0.12797 with plot age, 0.02996 with road age, and 0.04181 with road width. Only the correlation with plot basal area per ha was significant (n=44) Stepwise regression was used to model the relative basal area in Zone 1 versus these five plot variables (Appendix VII, Output 1). Equation 8 shows that only plot basal area per ha was retained in the model: predicted relative ba perha(Zone\) - 1.98332-0.023379 x plot basal area per ha [8] This model accounted for 11 percent of the variance in Zone 1 relative basal area. The predicted gain in basal area per ha for Zone 1 relative to Zone 5 reduced to zero when the plot basal area per ha reached 42.0 m 2 per ha. The variability around the fitted line was large. The tree basal area periodic annual increments in Zone 1 were significantly correlated for the period prior to the road and for the five periods after the road with average live crown ratio for the plot (correlations (r) over 0.19), road age ( r > 0.22), plot height (r > 0.22), mean basal area per tree (r > 0.32), mean tree diameter (r > 0.27), quadratic mean diameter (r > 0.30) and negatively correlated with the number of stems per ha (r < -0.22). Tree basal area p.a.i. was significantly correlated with plot basal area per ha only for the 5-year period prior to road construction (r = 0.20), and with site index for the period prior to and for the 2 years immediately following road building (r > 0.21). Elevation was significantly negatively correlated for all pre and post road periods with the exception of the 3 to 5 year period after the road (r < -0.20). Tree basal area periodic annual increment in Zone 1 relative to the 5-year period prior to the road was significantly negatively correlated with site index (r= -0.23) for the 3 to 5 year period and with Curtis' RD for the 16 to 20 year period (r= -0.20). A significant direct correlation was found with elevation for the 3 to 5 year period only (r-0.22). No other plot level variables were correlated significantly in any other periods with relative tree basal area increment. Stepwise multiple linear regression was used to model Zone 1 relative tree basal area increment as a function of the period represented as the midpoint of the years 74 class, and plot level variables for the periods representing the first 15 years after the road (Appendix V, Output 1). The relative tree basal area p.a.i. values are correlated over time, which would likely cause an underestimate of the p values for each variable, but this was not altered in the analysis. A logarithmic transformation was used to linearize the relationship. An r square value of 0.12 was obtained for predicting relative tree basal area p.a.i from six variables: period (midpoint of each class), plot basal area per ha, plot Curtis' RD, plot mean basal area per tree, plot quadratic mean diameter, site index, and stems per ha (Equation 9). All six variables were retained in the model at a significance level of 0.10, and all assumptions of regression were met. [9] LOG A (relative p.ai. for Zone!) - 4.79064 + 0.00782 x period - 0.16984 x plot basal area per ha + 0.96206 x Curtis'RD +161.04339 x meanbasal per tree - 0.464181 xDfl/fy -0.04250xsiteindex-0.00l354xstems per ha Equation 9 presents the model for predicting the relative tree basal area growth response in Zone 1, from the six plot level attributes for up to 15 years after the road right-of-way was established. Midpoint years (shown in brackets) were used for the periods; 1 to 2 years (1); 3 to 5 years (4); 6 to 10 years (8); and 11 to 15 years (13). 75 5 DISCUSSION 5.1 STAND LEVEL ATTRIBUTES 5.1.1 Basal Area Increase at the Edge On average, a 31% increase in basal area per ha, relative to the baseline, occurred within Zone 1, the first 5 m from the edge. The confidence interval for the relative basal area for Zone 1 had values all greater than 1.0, indicating a response in basal area, on average. These results are similar to those by Landbeck (1965), Pfister (1969), Bucht (1977), Bucht and Elfving (1977), Bella (1986), and Isomaki (1985 and 1986) for linear edges near pine stands. The studies on Douglas-fir and spruce also showed similar responses in stand basal area (Kramer, 1958; Van Laar etal., 1990; Isomaki and Niemisto, 1990; McCreary and Perry, 1993; Bella, 1986). The relative basal area in Zone 1 showed the largest variability among plots, compared to the other zones. Corresponding with basal area per ha, stand density, expressed as the relative density index from Curtis (1982), was also found to be higher at the edge. Differences were also found in Curtis' RD, relative to the baseline zone found in the 30 to 40 m interval from the edge. The relative Curtis' RD for the first 5 m from the edge was significantly different than the 10 to 20 m and 20 to 30 m intervals from the edge, but only marginally and not significantly different than the 5 to 10 m interval. As with basal area, the variability around the mean was high for Zone 1. The Zone 1 relative basal area per ha was significantly correlated to plot basal area per ha only and Curtis' RD only. Plot basal area per ha accounted for only 11 percent of the Zone 1 relative basal area per ha variability. Other variables such as road width, did not have a significant correlation with the relative basal area per ha in Zone 1. Likely, this was due to the limited range of widths in the sample, from 5.3 to 38.5 m, with most road widths more than one tree height wide. Surprisingly, the Zone 1 76 relative basal area per ha was not correlated with the age of the road opening nor with stand age or edge aspect. Other researchers found road width to be a significant variable in explaining the edge effects on adjacent stands (Kramer, 1958; Landbeck, 1965; Isomaki 1985 and 1986). The time since the seismic line clearing (equivalent to road age in this study) was important for explaining the variability around radial increment regressions for lodgepole pine, black spruce and aspen in Bella (1986). Edge aspect has been widely used as an explanatory variable in a number of studies analyzing edge effects (Wales,1972; Chen and Franklin, 1990; Palik and Murphy, 1990; Luken etal., 1991; Matlack, 1993; Chen etal. ,1995). It is expected that a larger basal area per ha at the stand edge would be attributed to some combination of a greater number of trees and larger diameters of the edge trees. While Figure 4 shows a declining trend in the number of stems per ha with distance from the edge, the difference between zones was not found to be significant, and there was high variability among plots (Table 3). Relative mean tree size, whether measured as the average mean diameter, as the quadratic mean diameter, or as the mean basal area per tree, was not correlated with the distance from the edge (Table 4), had similar mean values between zones, and did not appear to vary to any great degree between plots (Table 3). The question is, if average tree size was similar across the zones and the number of trees only appeared to be higher at the edge, but was not found to significantly higher, why was the basal area significantly higher at the edge? Quadratic mean diameter (DBHq) is strongly correlated with stand density (Clutter etal., 1983). Due to the relationship of the mean tree basal area and DBHq (Equations 4 and 5, respectively), an increase in basal area per ha with a corresponding increase in the number of stems per ha would result in no real difference in average tree size. This would not occur if the same number of additional tallied trees contributed negligibly to basal area (i.e., understory trees), or appreciably (i.e., veteran trees), for in either of these cases the quadratic mean diameter would change. If DBHq does not differ, but basal area per ha is higher, then an increase in the number of trees of average diameter, must correspond to the increase in basal area per ha. If 77 there was only a growth response in edge trees, and either no increase or decline in the number of trees, then the average diameter would have increased. For Zone 1 to have higher basal area on average, an appearance of, but not significantly higher numbers of stems on average, and an average tree size that was similar between zones, then any incremental response in Zone 1 basal area combined with the small difference in the number of stems must be, on average, representative of the mean diameter. As the expected relationship is for tree diameter to increase with time, neither ingrowth, nor small tree retention could be the only contributor to the basal area increase at the edge, otherwise both mean tree diameter, DBHq, and the mean basal area per tree would be lower at the edge as more small trees were tallied. Due to the 7.0 cm diameter limit, tallying of ingrowth is also more likely to be associated with older road openings than younger roads, which helps to explain some of the variation in the numbers of trees in the edge zone between plots. Across all 44 plots, the mean DBHq for the dead standing trees was lower in Zone 1 than any other zone, although this statistic was also more variable than similar means in the other zones (Table 5). The edge zone had a higher mean proportion of the stand basal area comprised of live trees than any other zone, and a lower proportion of dead standing and stumps combined, indicating less mortality at the stand edge. 5.1.1.1 Mortality at the Edge Across the 44 plots, mortality was lower at the stand edge, using either basal area proportions or stems per ha proportions (Table 5). Although the mortality statistics are slightly confounded by the negligible proportions and distribution (mostly in Zones 1 and 2) of stumps in the plots, the live tree proportions are greater at the road edge and the dead standing tree proportions are lower at the road edge. The dead standing proportion by basal area was always lower than the dead standing proportion by stems per ha (Table 5), indicating that on average, most of the dead standing trees occur in the smaller diameter classes. The DBHq of dead trees in Zone 1 also appeared to be lower than the DBHq of dead trees in the other zones, on average (Table 5). 78 The number of years since the road was constructed did not drastically change the mortality relationships between zones, although there was a greater range between the mean live tree proportions across the zones for the 21 to 40 year road age class. In the < 20 year road age class, the proportion of live trees by basal area at the edge was only slightly less than the interior of the plot, but the proportion by stems per ha was higher. In the 21 to 40 year, and 41 to 60 year road age classes, the proportion of live trees by basal area in Zone 1 was 4.3% and 5.3% greater than the baseline zone, respectively. Previous studies on clearcut edges found excessive mortality, usually in the form of windthrow or broken stems at the stand edge (Burton, 1999; Chen etal., 1992). The literature on linear edges have not discussed stand mortality in relation to distance from the stand edge (Kramer, 1958; Landbeck, 1965; Pfister, 1969; Bella, 1986; Isomaki 1985 and 1986; Van Laar etal., 1990; Isomaki and Niemisto, 1990; McCreary and Perry, 1993). Only one study found snowpress damage reduced the number of pine stems by 7% , or 3% of volume, in strip thinned stands (Bucht and Elfving, 1977). On average, the highest proportion of mortality was found in Zone 2 across all 44 plots (Table 5), and it is likely the higher mortality could be attributed to shading from the edge zone for some aspects. Isomaki (1985) and (1986) noted that reduced growth in the portion of the stand 5 to 10 m from the edge to was attributed to shading from trees found in the first 5 m. Other researchers have discussed the concept of a vegetative wall that forms at the gap edge as the opening ages (Wales, 1972; Chet et al., 1992; Matlack, 1993; and Burke and Nol, 1998). It seems logical to extend the shading argument for reduced growth to higher proportions of mortality in the 5 to 10 m zone also. 5.1.1.2 Ingrowth at the Edge Evidence in the literature points to the establishment of less shade tolerant species at the edge, with decreasing densities with distance from the edge (Luken et al., 1991; Chen etal., 1992). Ingrowth appeared to occur at the edge in the older road openings (Table 10), as plots near roads greater than 40 years of age had a larger proportion of the edge zone basal area comprised of trees in the suppressed canopy 79 layer. It is likely that these trees were suppressed only with respect to their relative canopy position, and in fact, either became established or were released after the road right-of-way was harvested. If small trees were established after the road in the younger road age classes at the edge, it is likely they were not tallied as they may not have grown into the 7.0 cm diameter limit. There was evidence of younger trees at the road edge, particularly in plots near older roads, as three plots were discarded from the basal area increment analysis as all randomly selected sub-sampled trees in the edge zone were younger than road right-of-way opening. 5.1.2 Height While relative basal area per ha was significantly different among zones, the mean heights for lodgepole pine were not found to be significantly different across the edge to interior gradient (Section 4.1.1.4). This was consistent with the findings in the literature, as most linear edge studies in pine found either no response, or a reduction in height at the road edge due to linear openings (Landbeck, 1965; Isomaki, 1985 and1986; Niemisto, 1989; Bucht 1977, Bucht and Elfving 1977; Bella, 1986). Only Pfister (1969), in a study on sloped terrain in Oregon, found an increase in the height of white pine at the road edge of the plots located below the road, but not on the paired plot located above the road. Pfister attributed the below road height growth to increased water from the out-sloping roads. As most of the sample plots in this study were located on relatively flat ground, the favorable impact of water runoff from the road would likely be quite negligible across all 44 plots. 5.1.3 Damage Generally, there were larger proportions of the edge zone afflicted with dead or broken tops than in the interior of the stand (Table 9). Trees growing near a new edge may be more susceptible to broken tops when they were not grown under exposed conditions and when they have high crowns (Oliver and Larson, 1996 p. 322). Other studies have found edge trees with large height to diameter ratios are more susceptible 80 to wind damage (Burton, 1999; Chen etal., 1992). Broken tops at the edge may also have been the result of felling trees for the road right-of-way. Scar damage was much more prevalent at the stand edge across all stand age classes than in the interior of the plots (Table 9), indicating that in many cases the scars were the result of residual damage from harvesting, road construction, or road use and maintenance such as mechanical limbing and snow clearing. Scar damage leaves trees more susceptible to mortality, and if severe enough, will reduce log quality due to wound recovery and the potential for decay (Nyland, 1996, pp 468). Table 9 shows the plots in the more mature stands have a higher incidence of mountain pine beetle damage than younger stands, as expected (Furniss and Carolin, 1977). Stands that are overstocked are also more susceptible to mountain pine beetle damage (Furniss and Carolin, 1977). In the 65 to 95 year old stands, mountain pine beetle was most prominent at the edge, compared to the other zones (Table 9). There did not appear to be a clear relationship with distance from the edge in the lowest age class, and these same plots showed little difference in the proportions of the stand with beetle damage indicators between the zones. Similarly, Harrington and Hendrick (1999) found no significant relationship between distance from the edge and beetle mortality in their research on 44 year old slash pine stands near circular openings. Trees infested with the mountain pine beetle are prone to blue stain fungus Ceratocystis montia (Rumb.) Hunt., which changes the wood appearance (Furniss and Carolin, 1977). Dwarf mistletoe presence was the most prevalent damage indicator found in the survey, infecting an average of 51.6% of the plot basal area (Table 7). While this statistic may seem high, Lotan and Critchfield (1990) noted that more than 50% of lodgepole pine forests are infected in many areas. While some plots were free of mistletoe, up to 100% of zone basal area was common across all stand age classes. Mistletoe indicators were generally less frequent at or near the edge, and were more prevalent in the interior of the stands (Table 9). The mistletoe proportions at the stand edges were also the most variable, however. Heavy infestations of dwarf mistletoe will reduce wood quality as well as height and diameter growth (Allen era/., 1996). 81 5.2 T R E E B A S A L A R E A G R O W T H R E S P O N S E 5.2.1 Relative Basal Area Increment Trees in Zone 1 grew faster than those in other zones, and faster than they were growing before the road was built, for up to 20 years after the road (Table 13 and Figure 7). All other zones showed no significant response. Annual radial increments were converted to tree basal area increments in order to standardize the influence of tree size on increment growth, as there is an expected indirect relationship with annual radial increment and tree size. Standardizing the basal area growth to a period prior to the disturbance also reduces the impact of tree size, for the growth response in each tree is relative to the growth response prior to the disturbance (McCreary and Perry, 1983 ). Basal area increment (not standardized to pre-road growth) was significantly correlated (r=0.32 or greater) with mean tree basal area and DBHq for up to 30 years after the road right-of-way harvest. Relative basal area increment (standardized to the 5-year period prior to the road) was not significantly correlated (r=0.06 or less) with either of these tree size statistics for any period after the right-of-way harvest. Tree basal area p.a.i. were calculated for periods of 5 years prior to the road, 1 to 2 years after the road, then 3 to 5 years and subsequent 5 year intervals as described in Bella (1986). The tree basal area p.a.i. values were standardized at the tree level to the p.a.i. for the 5-year period prior to harvest of the road right-of-way. A positive growth response was found in Zone 1 tree basal area increment when scaled with he 5-year period prop to the road, for up to 20 years after the road (Table 13 and Figure 7). In Zone 1, an average increase in the annual growth rate of 9.9% appeared immediately in the 1 to 2 years following road construction, followed by a 24.7% increase in the 3 to 5 year period after the road. The growth response was highest in the 6 to 10 year period at 35.9%, and began to decline, to an annual average of 32.9% for 11 to 15 year period. By the 16 to 20 year period, the response in the growth rate had declined to 15.6%. The variability around the mean growth rate 82 responses was high for Zone 1, across all periods. High variability in the growth rates at the edge was also noted by Chen etal. (1992) for clearcut edges. Significant differences in relative tree basal area p.a.i. were found between the edge zone and all other zones in the plot for the periods 3 to 5 years, 6 to 10 years and 11 to 15 years after the road right-of-way cutting. For the first 1 to 2 years after the harvest, the 5 m edge zone only differed from the portion of the plot 10 to 20 m from the edge, while no other zones within the plot differed. For the 16 to 20 year period, the first 5 m edge zone significantly differed from both the 10 to 20 m and 20 to 30 m zones (p=0.009 or less) on average, and was marginally but not significantly different from the 5 to 10 m zone (p=0.0069). The most interior zone at 30 to 40 m did not differ significantly from the first 5 m edge zone after the 15 to 20 year period. A significant tree basal area increment response occurred in the first 5 m from the edge, for the periods from 3 to 15 years after the road right-of way was harvested. Other studies have found similar responses in tree growth at the stand edge. In a study on lodgepole pine near seismic lines in Alberta, Bella (1986) found a 4 1 % increase in radial increment at the edge for the 5 to 10 years after clearing, relative to the increment in the control portions of the plots. Bella's study indicated the radial increment response declined to 33% for the 11 to 15 year period and to 24% for the 16 to 20 year period after clearing the openings. The response portions of Bella's plots were equivalent to a distance from the edge of one half of the average tree height. Bucht and Elfving (1977) found the diameter increment response was 5.9% greater for Scots pine trees within 3 m of the edge, and the response lasted for 13 years. Relative to 10 years prior to disturbance, Bucht (1977) showed a mean diameter increment response of 5% in Scots pine, which only occurred within the first 3 m from the edge. McCreary and Perry (1983) found a 12% increase in Douglas-fir basal area increment relative to the 5 years prior to strip thinning that occurred within the first 3 m from the edge. Chen etal., (1992) determined that the growth rates of Douglas-fir edge trees relative to the 10 years prior to establishing a clearcut were 33% higher at the edge than in the interior of the stand, although the effect decreased quickly within 60 m from the edge. 83 Kramer (1958) found Norway spruce diameter growth for the first 10 m from the edge had increased by 135% over the interior portions of the stand. Isomaki and Niemisto (1990) found that by the fifth growing season, the Norway spruce diameter growth response was 25% higher at the edge. The lack of significant difference in the relative basal area increment response between the edge zone and all other zones for the two year period following harvesting may be attributed to two factors. The first is that there was some potential for error around the road age estimates, as the road ages were best determined by interpolation between sequential air photos and available stand history information. The flight intervals were between three and nine years apart, and road ages were estimated at the midpoint of the intervals, unless more precise estimates of the road age were confirmed from harvest history information. At worst, road ages may be out by 5 years. The second factor which may actually delay the growth response after the road may be attributed to increased allocation to roots. Urban etal., (1994) found that after clearing road right-of-ways, white spruce roots grew at a greater percent than the tree trunks (69% versus 31%). The authors suggested that root growth allocation could help explain a delay in trunk response for three to nine years. The stage of stand development may be another factor, as Isomaki, (1986) attributed a five year delay in growth response to the 14 year age of the Scots pine stands in their study. Young stands were not sampled in this study, however (Table 1). Many of the other studies found a similar tree growth response time in the edge trees, as this study did. Both Bella (1986) and Isomaki (1985 and 1986) found a 20 year radial growth response in lodgepole pine and Scots pine respectively. Bucht (1977) and Bucht and Elfving (1977) had a 9 and 13 year response in Scots pine increments respectively, and; Isomaki and Niemisto (1990) found a 10 year growth response in Norway spruce. 5.2.2 Tree Basal Area Other studies found significant differences in tree size, usually expressed as mean tree diameter at the edge (Landbeck, 1965; Pfister, 1969; Bucht, 1977; Isomaki 84 1985 and 1986; Isomaki and Niemisto, 1990; Van Laar era/., 1990). While the first 5 m edge zone of this study had a significantly higher tree basal area growth response relative to the five year period prior to the road, there were no significant differences found in the mean tree basal area of the sub-sampled increment core trees at the edge. As a significant difference in total stand basal area was found at the edge relative to the interior of the plot, and a response in the growth rate of the edge trees was determined, the lack of a higher average tree basal area in the core trees at the edge is likely attributable to the sub-sampling procedure, and to the fact that the within zone variation around sub-sampled tree mean basal areas was high. The coefficient of variation was greater than 75% for all zones, over all periods from the time of right-of-way harvest to 20 years after the road. For each of the five zones within each plot, three trees were randomly selected for sub-sampling for the increment core measures, and no preference was made for trees with respect to size or relative position in the canopy. Due to the small number of sub-sampled trees, there is a probability that the mean size of the sub-sampled trees may have differed from the mean size of all pine trees within the zone. It would be expected that any bias between the sub-sample means and zone means would be reduced, given that there were 44 plots with five zones, each with their own tree size distribution and sub-sampled tree size distribution. However, if basal area per ha is higher at the edge, yet the numbers of stems and the DBHq are not significantly different between zones, then the distribution of tree diameters in the edge zone will likely differ from the interior zones in the plot. Clearly, with both higher relative densities and higher basal area per ha values at the edge relative to the plot interiors, it appears the plots were more heavily stocked at or near the edge due to reduced mortality. Ingrowth may also have been a factor in higher densities, but those trees will likely have been eliminated from increment analysis due to their age, relative to the road. If the interior zones in the plots have fewer trees in the smaller diameter classes due to expected higher mortality, yet those same sizes of trees were retained at the edge and were eligible for core measurements, then the probability of randomly selecting smaller trees for sub-sampling at the edge zone would be higher at the edge than in the other zones. 85 Variability was higher around the tree basal area mean statistics of edge trees, than in any other portion of the plot. The high variability at the edge and the higher probability of selecting the few sub-sampled trees from smaller trees in the edge zone would confound any observation of increased tree basal area at the edge, even though there was a response in tree basal area increment and in stand basal area per ha. While efforts were made to reduce the impacts of tree size on increment response by standardizing post-road increment with pre-road increment, the mean basal area per tree statistic does not address the inherent variation between tree sizes within the zones. Certainly, it is shown in the basal area increment analysis that there is a favourable response and significant difference at the stand edge for the periods from 3 to 15 years after the road. Given that there is an increase in the total stand basal area per ha at the edge, the increment response due to the road clearing treatment is likely more useful for growth and yield modeling, than simple tree level basal area differences between the zones. 5.3 C R O W N LENGTH AND S T E M FORM Live crown ratio and live crown ratio relative to the baseline zone were negatively correlated with distance from the edge (Table 4), and showed a gradual decline across the edge to interior gradient, on average. For trees in the first 5 m zone, only site index was significantly correlated with relative live crown ratio (r=-0.34). Tree crown lengths relative to the baseline zone in the plot interior were significantly longer, on average at or near the edge than in the interior of the plot. There were significant differences in the live crown ratios relative to the interior baseline between the first 5 m zone and the two zones between 10 and 30 m from the edge. Significant differences were not found between the first 5 m zone and the second 5 to 10 m zone, however. Edge tree crowns were on average 21% larger than the interior zone crowns, while live crown in the 5 to 10 m zone were only 12% larger than interior crowns. Average crown proportions found in the most interior zone of the plots in this study were 36% of the total tree height. Larger crowns at the edge corresponded with the increased basal area increment and increased basal area per ha at the edge. 86 While Harrrington and Hendrick (1999) suggested that slash pine edge trees increased their crown area very soon after cutting circular openings, Chen etal. (1996) found lodgepole pine trees studied near Williams Lake, B.C. to have less plasticity in their crowns for adapting to decreasing light conditions. Similarly, Muth and Bazzaz (2002) attributed the reluctance of eastern white pine to forage for light at edge to be due to the species excurrent growth and rigid architecture. Isomaki (1985 and 1986) found a more dramatic change in stem form for Norway spruce at the edge than for Scots pine, and the stem form change extended further into the spruce stands than it did for the pine stands. Kramer, (1958) noted that while there was increased volume in the edge trees, the quality was reduced through greater knottiness and taper. Presumably, greater knottiness was due to branch sizes and longer crowns. Landbeck (1965) found unfavorable changes in taper to a depth of 10 m from the edge that was influenced by the width of the road opening. Isomaki and Niemisto (1990) considered the sector angle of edge trees to be more relevant in explaining the poorer form and lower crowns of edge trees than distance from the edge. As all but one road opening in this study were older than 10 years, it is likely that the increased crown lengths on the edge trees are due to a lower propensity to shed the lower branches while the trees continued to grow in height. Damage agents, in particular scarring, mistletoe and mountain pine beetle, are probably of greater to concern with regards to the wood quality of edge trees than the small, but significant, relative increase in overall crown length. 5.4 BOLE ELLIPTICITY Following the results of Isomaki (1985 and 1986), differences in bole ellipticity were tested. Despite the appearance of slight differences in the tree basal area increment for the direction parallel with the edge (Figure 8 and Table 14), significant differences in bole ellipticity due to the road orientation were not found. The variability around the mean was high for the edge trees, however. Studying lodgepole pine, white spruce, black spruce and hardwood species, Bella (1986) used radial measures from four quadrants on stem disks to check for bole ellipticity relative to seismic openings. Even with the precision of measurements from stem disks, Bella's study did not 87 determine any significant difference in tree ellipticity due to the linear openings, either. For Scots pine and Norway spruce edge trees, Isomaki (1985 and 1986) attributed the slightly larger allocation of stem wood growth in the direction parallel to the hydro line openings, to be the result of stress from prevalent winds blowing through the line corridors. 5.5 YIELD R E C O V E R Y FOR LOST GROWING S P A C E In his study on seismic lines in Alberta, Bella (1986) estimated that 6% of the lost volume due to seismic line clearing was recovered, Isomaki (1985 and 1986) estimated 0.58 to 1.10 m of a hydro line was recovered, and Pfister (1969) estimated 3.99 to 5.64 m of road was recovered. Based on the fact that the edge zone in this study was 5 m wide and that it showed an average increase of 31.3% in basal area per ha over the plot interior, an average of 1.56 m of the unstocked growing space lost to the road opening is recovered through increased basal area per ha at the edge. If both sides are impacted similarly, then 3.12 m of growing space is recovered . Using the 95% confidence interval for the basal area increase relative to the plot interior, 1.42 to 4.83 m of the road opening if both sides of the road are similarly impacted. 5.5.1 Application of the Yield Adjustment As relative basal area per ha in Zone 1 was not correlated with road width, nor with road age, and neither of these variables were significant enough to be retained in final the model shown in Equation 8, the predicted response in basal area per ha from Equation 8 could apply to all widths sampled in this survey (5.3 to 38.5 m) and to all road ages (5 to 51 years). The predicted relative basal area per ha would depend on the stand density, however. Based on the work of Pfister (1969), Isomaki (1985 and 1986), Bella (1986) and Isomaki and Niemisto (1990), Equation 8 was modified to determine the recovered growing space (RGS) in hectares, for each side of the road based on stand basal area, the 5 m width of the edge zone, and the length of the road segment contiguous to the stand of interest (Equation 10). As with the plot measures, 88 all distances in Equation 10 are in horizontal distance units; therefore, lengths are easily determined from GIS overlays without conversion from slope distance. RGS(ha) = [(1.98332- 0.023379 x plot basal area per ha)-l]x5(m) , ^ l k a [ 1 0 ] x road segment (m)x -10000m2 Treating each side of the road individually, the stand basal area along with the stand's road segment length are applied in Equation 10 to calculate the attributable recovered growing space. Stand basal area per ha from the applicable stand is applied to the calculated R G S , and the total basal area for the R G S is determined. Total basal area is subsequently summed with the attributable stand total basal area, resulting in an adjusted total basal area for the stand. It is suggested that both the adjusted and unadjusted total basal area values be explicit in the inventory to identify those stands which have been adjusted, and equally important, those stands which may not be eligible for adjustment. Summing the total basal area values for all applicable stands on both sides of the road will give the total recovered yield, in basal area per ha, for the total length of the road. There would be no need to split the polygons contiguous to the road edge because the adjustment would be applied on a stand by stand basis, and would be a function of the length of the linear opening. Theoretically, volume adjustments could be applied to the stand through an appropriate localized volume and basal area relationshi as there were no significant differences in tree heights across the edge to interior gradient of the sample plots. It is suggested that measures be taken to apply the basal area adjustment to eligible stands that fall within the ranges for the plot attributes of the 44 sample plots. Table 1 provides the range of all live tree statistics for the stands that were used to construct the models. Stands should be predominantly lodgepole pine stands. Furthermore, because stands were deliberately excluded from sampling if they had been subject to partial harvesting or were adjacent to other openings, the same criteria should be used when selecting suitable stands for applying the model. 89 Other studies have shown differences in growth and yield between selection cutting and strip thinning. Selection cutting, with the increased growing space dispersed amongst the entire stand, and strip thinning, with increased growing space only affecting trees at the strip thin edge, have resulted in differences in both total stand yield and in the growth response of individual trees, for similar stands (Bucht and Elfving, 1977; McCreary and Perry, 1983). As partially cut stands were not eligible for sampling in this study, it is suggested that appropriate yield adjustments be applied to stands with partial cutting treatments, other than the adjustments presented in this thesis for linear edges. GIS stand attribute and history data as well as appropriate photogrammetry methods with a skilled interpreter can readily be employed to assess suitable road openings and stand conditions. After selecting appropriate stands from the air photos, field review of questionable stands would be relatively inexpensive given that access to road edge stands is relatively easy. 5.5.2 Application of the Growth Response The tree basal area growth response was only found within the first 5 m of the stand. The tree basal area growth response was limited in its duration also, as there were only significant differences between the edge and the plot interior for the periods between 3 and 15 years after creating the road right-of-way. This duration of the tree growth response is consistent with the diameter increment responses found in other research (Bucht, 1977; Bucht and Elfving, 1977; Isomaki, 1985 and 1986; Bella, 1986; Isomaki and Niemisto, 1990) To apply the model for the response in tree basal increment, the same criteria for stand selection should be applied as for stand yield. In particular, the range of stand level attributes for the live trees as presented in Table 1, and the physical attributes such as the width and age of the roads. As with the yield adjustment, selection against stands with partial cutting should be employed when applying the growth response adjustment. Bucht and Elfving (1977) and McCreary and Perry (1983) provided results of the growth response differences between selection cutting versus strip thinning. 90 The plot level attributes of basal area per ha, Curtis' RD, mean basal area per tree, DBHq, site index and stems per ha were all necessary attributes to determine the tree basal area increment in the first 5 m from the edge, along with the midpoint of the period class. Using Equation 9, the first 5 m edge zone tree basal area relative p.a.i. can be calculated for the respective stand for the periods up to 15 years after the road. The growth responses are then applied in a locally calibrated tree level growth model for simulation. The yield difference between applying the growth response, and not applying the growth response is determined. This difference in yield is applied to the total yield of the stand to create an adjusted total yield. 5.5.3 Growth and Yield Applications for the Forest Manager The ease of applying an adjustment to stand yield after determining the recovered growing space from using Equation 10, make the yield adjustments more appropriate for established roads. As significant differences in tree heights were not found between the zones, localized volume and basal area relationships could be used to convert the additional total basal area into total volume. Furthermore, inherent to the results of higher basal area at the edge is the reduced mortality, a factor which may, or may not be simulated as precisely when modeling stand growth with a tree level model. The growth model approach is more appropriate for determining yield responses on recent or on future roads. The growth model can also provide information to the forest manager to justify reducing some of the lost growing space from roads where appropriate, while still meeting the necessary design class and safety requirements for the intended road use. Appropriate widths for road right-of-ways (and running surfaces) are usually based on the design class of the road for the intended use, season (snow clearing will require wider road openings), traffic, harvesting and hauling equipment size, and speed. Safety is of utmost concern for log hauling operations, and should be a preface to any discussion on reducing logging road right-of-way widths. While road right-of-way width was not a significant variable for determining the growth of the edge trees or the yield in the edge zone for that matter, larger road widths do result in greater 91 losses of growing space. This study demonstrated that some of the growing space is recovered for lodgepole pine stands through increased growth and reduced mortality at the edge. If the forest manager is able to reduce the lost growing space from road openings, yet still meet the design class and safety requirements for the intended road use, a larger proportion of the land base will be in production. While there were many variations of what constitutes an edge in the literature (e.g., Wales, 1972; Minko and Hepworth, 1990; Palik and Murphy, 1990; Luken etal., 1991; Bradshaw, 1992; Chen etal., 1992; Young and Mitchell, 1994; Bella, 1986; and Cadenasso et al., 1997) this study defined the tree line as the edge. Appropriate road right-of-way widths must also consider the impact of edge tree lateral branches on road safety and design use; therefore, any measures taken by the forest manager to recover growing space should not result in associated higher maintenance costs and potential damage to edge trees from mechanical limbing. As the incidence of scar damage was higher at the edge than in the interior of the sample plots established in this study, appropriate planning by the forest manager to recover growing space should not come at the cost of increased damage to edge trees. 92 6 CONCLUSION This study found a significant difference in stand level basal area per ha for the first 5 m from the edge of the road, relative to the plot interior. On average, stand basal area of the sampled plots was 31.3% higher at the edge than in the interior of the stands but variability was higher, also. These results are similar to the findings in other research on pine stands (Landbeck,1965; Pfister, 1969; Bucht, 1977; Bucht and Elfving, 1977; Bella, 1986; Isomaki, 1985 and 1986). Higher relative basal area per ha at the edge corresponded with higher values of Curtis' RD, although the RD declined more gradually than basal area per ha did across the plot edge to interior gradient. Tree heights were not significantly different across the edge to interior gradient. This may have been attributable to the range of widths in this sample, from 5.3 to 38.5 m, with most road widths more than one tree height wide. Only Pfister (1969) found an increase in pine tree heights at the edge but the height increases in that study were attributed to increased water flow to the plots located below the road. While the number of trees per ha appeared to be slightly higher at the edge, the difference was not significant. Surprisingly, no significant differences were found in tree size, whether measured as mean basal area per tree, DBHq or as average diameter. The 31.3% increase in basal area per ha at the edge can be attributed to reduced mortality in the edge zone. Ingrowth also appeared to be a factor to some degree in the older roads. While there was less mortality at the edge, there was a higher incidence of abiotic damage in the form of scars. Broken or dead tops were more common at the edge but the incidence of dwarf mistletoe was lower. Mountain pine beetle appeared to be more a function of stand age and density then just distance from the edge of the road. Live crowns were 21% longer at the edge than the interior of the plot, and were 12% longer than the plot interior for the 5 to 10 m zone. The larger crowns at the edge were relative to the interior crown lengths, which were only 36% of the total tree height, on average. Longer crowns on the edge trees will allow the tree to capture more light 93 (Oliver and Larson, 1996), but will also tend to reduce the wood quality through larger diameter branches and loose knots (Berg, 1973). Growth on the lower bole was tested for ellipticity relative to the road right-of-way orientation. Significant differences were not found in radial growth due to the road orientation. These findings were consistent with Bella's (1986) study on lodgepole pine in Alberta, where stem disks were used to check ellipticity instead of increment cores. Across the range of road ages, the tree basal area increment, scaled for the 5 years prior to the road origin, showed an increase in the first 5 m zone near the edge. On average, trees in the first 5 m edge zone grew 32.1% more than the 5 year period prior to the road, for a duration of 15 years after the road was established. The higher growth response in the first 5 m zone was significantly different than the growth rate of all other zones from the third year following the road right-of-way harvest, up to 15 years after the road was built. It was suggested that the results of the growth response be applied as a treatment response in a locally calibrated tree level growth model, to determine stand yield for the first 5 m from the edge. Given that roads will require minimum clearing widths for road design class, equipment transport and safety, the growth and yield information of this study could be applied by forest managers to help determine appropriate widths of future road openings in lodgepole pine stands to reduce production loses on the landscape. Based on the work of Pfister (1969), Bella (1986), Isomaki (1985 and 1986), and Isomaki and Niemisto (1990), an equation was derived from the generalized linear modeling to determine the theoretical recovered growing space (RGS) due to the increased basal area per ha at the edge. The stand basal area per ha is subsequently applied to the R G S , and then added to the attributable stand total basal area. This applies an adjusted total stand basal area based on the eligible portion of the stand near the road opening. Only plot level basal area per ha was retained in the final model as a dependent variable; however, application of the model should be limited to stands within the range of live tree attributes sampled in this study. Similar studies could be initiated for other stand types. Bella (1986) noted increases in diameter growth for white and black spruce, but no consistent results with 94 aspen. Chen et al., 1992 noted only a 33% increase in the growth response of Douglas-fir at the edge of a clearcut, but a 150% response in western hemlock. As most of the studies have been conducted on the pine and spruce species, it would be of interest to study the response of a more shade tolerant and commercially valuable species like western redcedar {Thuja plicata Donn ex D. Don.). Certainly in the locale of this study, the interior variety of Douglas-fir would also be of interest. The plot size of 40 m into the stand from the road edge appeared appropriate for this study. One modification to the plot that could be considered for future studies would be to divide the 40 m into eight zones, instead of five, at 5 m intervals from the road edge. This would keep the sampling unit equal, resulting in similar precision for all zones. This would also allow for finer analysis of the relationships with distance, without the cost of stem mapping all of the trees. Based on the reduced mortality and higher density at the edge, it is suggested that the number of increment core trees be increased to address the issue of high variability in mean tree size at the edge than in the other zones. As there was no significant difference in tree ellipticity due to road orientation in this study or in Bella's (1986) work, it may be more appropriate to increase the number of sample trees for increment coring, yet decrease the number of cores per tree. The stands sampled in this study were comprised mostly of lodgepole pine. To avoid confounding with other changes to the stand composition, stands with some partial cutting, and other edges were avoided. Other impacts would further alter the basal area per ha in the first zone. Partial cutting of edge trees was commonly noted during reconnaissance and sampling in the Lignum IFPA area. The results should therefore be applicable only to predominantly lodgepole pine stands within the range of stand attributes found in this study. Caution should be used in applying these results to stands with changes other than the road. 95 LITERATURE CITED Allen, E.A., D.J. Morrison and G.W. Wallis. 1996. Common tree diseases of British Columbia. Can. For. Serv., Pacific Forestry Centre, Victoria. 178pp Bassman, J .H. 1984. Selected physiological characteristics of lodgepole pine. In Lodgepole pine: the species and its management: symposium proceedings; 1984 May 8-10 Spokane WA, 1984 May 14-16 Vancouver, BC. Office of Conferences and Institutes, Cooperative Extension, Pullman, WA. 27-44. Bella, I. 1986. Tree growth response along seismic lines in Alberta. Forestry Chronicle 62(1): 29-34. Berg, P. 1973. Silviculture of Pinus radiata stand edge trees at Woodhill Forest. New Zealand Journal of Forestry 18(1): 115-123. Bradshaw, F.J. 1992. Quantifying edge effect and patch size for multiple use silviculture: a discussion paper. Forest Ecology and Management 48(3-4): 249-264. Bucht, S. 1977. Increment losses caused by strip roads [at the first thinning of Scots pine stands]. Skogen 64(6): 218-222. Bucht, S. and B. Elfving. 1977. Thinning response and increment in a strip thinned stand. Sveriges Skogsvardsforbunds Tidskrift 75(4): 323-345. Burton, P.J. 1999. Effects of block edges and patch retention on vegetation in the SBSmc Final Report. Report produced for Forest Renewal British Columbia. Project Number SB96029-RE. Burke, D.M. and E. Nol. 1998. Edge and fragment size effects on the vegetation of deciduous forests in Ontario, Canada. Natural Areas Journal 18(1): 45-53. Cadenasso, M.L., M.M. Traynor, and S.T.A. Pickett. 1997. Functional location of forest edges: gradients of multiple physical factors. Canadian Journal of Forest Research 27(5): 774-781. Chen, H. Y. H., K. Klinka and G.J . Kayahara. 1996. Effects of light on growth, crown architecture, and specific leaf area for naturally established Pinus contorta var. latifolia and Pseudotsuga menziesii var. glauca saplings. Canadian Journal of Forest Research 26(7): 1149-1157. Chen, J . Q., and J.F. Franklin. 1990. Microclimatic pattern and basic biological responses at the clearcut edges of old-growth Douglas-fir stands. Northwest Environmental Journal 6(2): 424-425. 96 Chen, J.Q., J . F. Franklin, and T.A. Spies. 1992. Vegetation responses to edge environments in old growth Douglas fir forests. Ecological Applications 2(4): 387-396. Chen, J.Q., J .F. Franklin, and T.A. Spies. 1995. Growing season microclimatic gradients from clearcut edges into old growth Douglas fir forests. Ecological Applications 5(1): 74-86. Cienciala, E., P.E. Mellander, J . Kucera, M. Oplustilova, M. Ottosson-Lofvenius and K. Bishop. 2002. The effect of a north-facing forest edge on tree water use in a boreal Scots pine stand. Canadian Journal of Forest Research 32(4): 693-702. Clutter, J.L., J .C . Fortson, L.V. Pienaar, G.H. Bristerand R .L Bailey. 1983. Timber management: a quantitative approach. Krieger Publishing Co. , Malabar, Florida. 333 pp. Coates, D.K. and P.J. Burton. 1999. Growth of planted tree seedlings in response to ambient light levels in northwestern interior cedar-hemlock forests of British Columbia. Canadian Journal of Forest Research 29(9): 1374-1382. Cochran, P. H. (1984). Soils and productivity of lodgepole pine. In Lodgepole pine: the species and its management: symposium proceedings; 1984 May 8-10 Spokane WA, 1984 May 14-16 Vancouver, BC. Office of Conferences and Institutes, Cooperative Extension, Pullman, WA. 89-94. Coutts, M. P. and B. C. Nicoll. 1991. Orientation of the lateral roots of trees. I. Upward growth of surface roots and deflection near the soil surface. New Phytologist 119(2): 227-234. Coutts, M. P. and J . J . Philipson 1978. Tolerance of tree roots to waterlogging. II. Adaptation of Sitka spruce [Picea sitchensis] and lodgepole pine [Pinus contorta] to waterlogged soil. New Phytol 80(1): 71-77. Curtis R.O. 1982. A simple index of stand density for Douglas-fir. Forest Science 28 (1): 92-94. Dumanski, J . , J .C . Wright and J.D Lindsay. 1973. Evaluating the productivity of pine forest in the Hinton-Edson area, Alberta, from soil survey maps. Canadian Journal of Soil Science 53(4): 405-419. Dykstra, P. and M. Curran. 2002. Skid road recontouring in southeastern British Columbia: 7-year tree growh results. Res. Br., B.C. Min. For., Victoria, B.C. Tech. Rep. 001. 20 pp Eis, S. 1970. Root growth relationships of juvenile white spruce, Alpine fir, and lodgepole pine on the three soils in the interior of British Columbia. [Picea glauca, Abies lasiocarpa, Pinus contorta]. Can. Dept. Agr. Publ. 1276. 10pp. ESRI 1996. ArcView GIS, ver. 3.1. Environmental Systems Research Institute, Inc., Redlands, O A . 97 Fleming, R.L., T.A. Black, R.S. Adams and R.J. Stathers. 1998. Silvicultural, treatments, microclimatic conditions and seedling response in Southern Interior clearcuts. Can. J . Soil Sci . 78: 115-126. Cited in Hope, G. D. 2000 Furniss, R.L. and V.M. Carolin. 1977. Western Forest Insects. Misc. Pub. 1339. USDA For. Serv. Washington, D.C. 654pp Garmin Corp. 1998. G P S 12XL personal navigator owners manual and reference. Part No. 190-00134-10 Rev. A. Garmin Corporation, Olathe, K.S. 66pp Goudie, J . W. 1984. Height growth and site index curves for lodgepole pine and white spruce and interim managed stand yield tables for lodgepole pine in British Columbia. B.C. Min. For., Res. Branch., F.P.D.S. Section. Unpubl. Rep. 75 pp. Guariguata, M.R. and J .M. Dupuy. 1997. Forest regeneration in abandoned logging roads in lowland Costa Rica. Biotropica 29(1): 15-28. Harrington, T. B. and R.L. Hendrick. 1999. Tree growth and resource availability in response to simulated canopy gaps in mature slash pine forest. Proceedings of the tenth Biennial Southern Silviculture Research Conference, Shreveport, Louisiana, February 16-18, 1999. U.S.D.A. For. Serv. Gen. Tech. Rep. SRS-30: 374-376 Hill, J.D., C D . Canham, and D.M. Wood. 1995. Patterns and causes of resistance to tree invasion in rights of way. Ecological Applications 5(2): 459-470. Hope, G. D. 2001. Effects of bladed skid trails on soil properties and early tree growth on two steep slopes in the southern interior of British Columbia. B.C. Min. For., Res. Branch., B.C. Min. for., Victoria, B.C. Res. Rep; 21 pp. Hope, G.D., W.R. Mitchell, D.A. Lloyd, W.R. Erickson, W.L. Harper, and B.M. Wikeem. 1991. Interior Douglas-fir Zone. In Ecosystems of British Columbia (6th ed.). D. Meidinger and J . Pojar (eds). B.C. Min. For. Res. Br. Chap 10, pp. 153-165 Hourdequin, M. 2000. Introduction: special section: ecological effects of roads. Conservation Biology 14(1): 16-18. Isomaki, A. 1985. Edge effects of strip roads in coniferous stands. Paper presented at IUFRO congress, Moscow. 10 pp. Unpublished. Isomaki, A. 1986. Effects of line corridors on the development of edge trees. Folia Forestalia 678. 30 pp. Isomaki, A. 1994. Determination of strip road width. Metsantutkimuslaitoksen tiedonantoja 501. 66 pp. Isomaki, A. and P. Niemisto. 1990. Effect of strip roads on the growth and yield of young spruce stands in Southern Finland. Folia Forestalia 756. 36 pp. 98 Kardell, L. and W. Pettersson. 1973. The effect of root damage on increment in Norway spruce. Sveriges Skogsvardforbunds Tidskrift 71(5): 423-447. Klinka, K. M.C. Feller, R.N. Green, D.V. Meidinger, J . Pojar, and J . Worrall. 1990. Ecological principles: applications. In Regenerating British Columbia's Forests. D.P. Lavender, R. Parish, C M . Johnson, G. Montgomery, A.Vyse, R.A. Willis and D. Winston (eds.) Govt, of Canada and Prov. of B.C., FRDA. U B C Press, Vancouver. Chap 6, pp 55-73 Koch, P. 1996. Lodgepole pine in North America. Part 1: Background. Forest Products Society, Madison Wl . 343 pp. Kramer, H. 1958. The effect of width of forest roads on increment in adjacent stands. Allgemeine Forst und Jagdzeitung 129(6): 121-134. Kremsater, L.L. 1997. A review of edge effects: theory, evidence, and recommendations for managers. Report for MacMillan Bloedel Ltd., funded by Forest Renewal B.C. 69 pp. Landbeck, H. 1965. Road width and edge effects in Scots pine. Arch. Forstw. 14(1): 41-59. Lavender, D.P. R. Parish, C M . Johnson, G. Montgomery, A.Vyse, R.A. Willis and D. Winston (eds.). 1990. Regenerating British Columbia's Forests. Govt, of Canada and Prov. of B.C., FRDA. UBC Press, Vancouver. 372pp. LeMay V., M. Bowering and P. Marshall. 1999. Effects of roads on the growth of adjacent lodgepole pine trees in the Williams Lake are of B.C.: research plan. Report produced for Lignum Ltd. and Forest Renewal B.C. University of British Columbia, Dept. of Forest Resources Mgmt. 11pp. LeMay V., M. Bowering and P. Marshall. 2001. Effects of roads on the growth of adjacent lodgepole pine trees in the Williams Lake are of B.C.: Final Report. Unpublished report produced for Lignum Ltd. University of British Columbia, Dept. of Forest Resources Mgmt. 114pp. Leopold, A. 1933. Game management. Charles Scribner's Sons, New York. Lignum. 1998a. 1996/97 stewardship update: toward sustainability. Lignum Ltd., Vancouver, B.C. <http://www.lignum.com/pdf/stewarts.pdf> Lignum. 1998b. Lignum Ltd., Innovative Forest Practices Agreement Area (IFPA) electronic database (June, 1999 ed.,). Lignum Ltd., Williams Lake, B.C. Lockaby, B.G. and C G . Vidrine. 1984. Effect of logging equipment traffic on soil density and growth and survival of young loblolly pine. Southern Journal of Applied Forestry 8(2): 109-112. 99 Lord, T. M. and K. W. G. Valentine. 1978. The soil landscapes of British Columbia - the soil map of British Columbia. In: The soil landscapes of British Columbia. K. W. G. Valentine, P. N. Sprout, T. E. Baker and L. M. Lavkulich (eds). B.C. Min of Env., Victoria, B.C. Chap 3.2, pp 99-100 Lotan, J . E. and W. B. Critchfield. 1990. Lodgepole pine (Pinus contorta Dougl. ex. Loud.) In: Silvics of North America Volume 1; Conifers. R.M. Burns and B.H. Honkala (tech. Coordinators) Agriculture handbook 654. USDA Forest Serv. Washington, DC pp 302-314 Lotan, J . E. and D. A. Perry (1983). Ecology and regeneration of lodgepole pine. Washington, D.C., USDA Forest Service; Agriculture Handbook 606. 52 pp. Luken, J.O., A .C . Hinton and D.G. Baker. 1991. Forest edges associated with power line corridors and implications for corridor siting. Landscape and Urban Planning 20(4): 315-324. Marshall, P., T. Szikszai, V. LeMay and A. Kozak. 1995. Testing the distributional assumptions of least squares linear regression. Forestry Chronicle 71 (2): 213-219. Matlack, G.R. 1993. Microenvironment variation within and among forest edge sites in the Eastern United States. Biological Conservation 66(3): 185-194. Mayaka, T.B. 1994. A family of segmented polynomial functions for modelling the border effect on the diameter growth of ayous (Triplochiton scleroxylon K. Schum). Forest Ecology and Management 70(1-3): 275-283. McCreary, D. D. and D.A. Perry. 1983. Strip thinning and selective thinning in Douglas fir. Journal of Forestry 81(6): 375-377. Meidinger, D. and J . Pojar (eds.). 1991. Ecosystems of British Columbia (6th ed.). B.C. Min. For. Res. Br. Victoria, B.C. 330pp Meidinger, D, J . Pojar and W.L. Harper. 1991. Sub-Boreal Spruce Zone. In Ecosystems of British Columbia (6th ed.). D. Meidinger and J . Pojar (ed). B.C. Min. For. Res. Br. Chap 14, pp. 209-222 Mickovski, S.B and A.R. Ennos. 2002. A morphological and mechanical study of the root systems of suppressed crown Scots pine Pinus sylvestris. Trees (2002) 16: 274-280 Minko, G . and G. Hepworth. 1990. Growth effects of large gaps in Pinus radiata plantations. New Zealand Journal of Forestry Science 20(1): 22-28. Muth, C .C. and F.A. Bazzaz. 2002. Tree canopy displacement at forest gap edges. Canadian Journal of Forest Research 32(2): 247-254 100 Neter, J . , M.H. Kutner, C . J . Nachtsheim and W. Wasserman. 1996. Applied linear statistical models, fourth edition. Ridhard D. Irwin, Toronto. 1408 pp. Niemisto, P. 1989. A simulation method for estimating growth losses caused by strip roads. Scandinavian Journal of Forest Research 4(1): 203-214. Nyland, R.D. 1996. Silviculture concepts and applications. McGraw-Hill, New York, pg.68-70 of 633pp. Oke, T.R. 1987. Boundary Layer Climates - 2 n d ed. Routlege, London & New York. pg.66 of 455pp. Oliver, C D . and B.C. Larson. 1996. Forest stand dynamics: update edition. John Wiley and Sons, Inc. New York. 509 pp. Palik, B.J. and P.G. Murphy. 1990. Disturbance versus edge effects in sugar maple/beech forest fragments. Forest Ecology and Management 32:2-4. Pfister, R.D. 1969. Effect of roads on growth of western white pine plantations in Northern Idaho. U.S.D.A. For. Serv. Res. Pap. INT 65. Pukkala, T. 1989. Methods to describe the competition process in a tree stand. Scandinavian Journal of Forest Research 4(2): 187-202. S A S Institute Inc. 1989. SAS /STAT user's guide, version 6, fourth edition, volumes 1 & 2 . Cary, NC. 1739 pp. Senyk, J . and D. Craigdallie. 1996. Effects of harvesting methods on soil properties and forest productivity in Interior British Columbia. Canada Forest Service, Pacific Forestry Centre, Victoria, B.C. Information Report BCX365E. Smith, R.B. and E.F. Wass. 1979. Tree growth on andadjacent to contour skidroads in the subalpine zone, Southeastern British Columbia. Canada Forest Service, Pacific Forestry Centre, Victoria, B.C. BC R 2. Smith, R.B. and E.F. Wass. 1980. Tree growth on skidroads on steep slopes logged after wildfires in Central and Southeastern British Columbia. Canada Forest Service, Pacific Forestry Centre, Victoria, B.C. B C R 6. Smith, R.B. and E.F. Wass. 1994. Impact of skidroads on properties of a calcareous loamy soil and on planted seedling performance. Columbia. Canada Forest Service, Pacific Forestry Centre, Victoria, B.C. B C X 346. Stage, A.R. 1976. An expression for the effect of aspect, slope, and habitat type on tree growth. Forest Science 22 (4): 457-460. Steen, O. and D.A. Demarchi. 1991. Sub-Boreal Pine-Spruce Zone. In Ecosystems of British Columbia (6th ed.). D. Meidinger and J . Pojar (eds). B.C. Min. For. Res. Br. Chap 13, pp. 195-207 101 Stathers, R.J., R. Trowbridge, D.L. Spittlehouse, A Macadam and J .P . Kimmins. 1990. Ecological principles: basic concepts. In: Regenerating British Columbia's Forests. D.P. Lavender, R. Parish, C M . Johnson, G . Montgomery, A.Vyse, R.A. Willis and D. Winston (eds.) Govt, of Canada and Prov. of B.C., FRDA. UBC Press, Vancouver. Chap 5, pp 45-54 Urban, S T . , V . J . Lieffers, and S .E . MacDonald. 1994. Release in radial growth in the trunk and structural roots of white spruce as measured by dendrochronology. Canadian Journal of Forest Research 24(8): 1550-1556. Valentine, K.W.G. and A.B. Dawson. 1986. The interior plateau. In: The soil landscapes of British Columbia. K. W. G. Valentine, P. N. Sprout, T. E. Baker and L. M. Lavkulich (eds). B.C. Min of Env., Victoria, B.C. Chap 3.4, pp121-134 Van Laar, A., H. Kramer, H.J. Schmidt. 1990. Individual tree investigations in the Douglas fir thinning trial at Manderscheid. Allgemeine Forst und Jagdzeitung 161: 6-7. Voller, J . and S. Harrison (eds.). 1998. Conservation biology principles for forested landscapes. UBC Press, Vancouver, B.C. Wales, B. A. 1972. Vegetation analysis of north and south edges in a mature oak-hickory forest. Ecological Monographs 42(4): 451-471. Wass, E.F. and R.B. Smith. 1994. Impacts of soil disturbance on root systems of Douglas-fir and lodgepole pine seedlings. Canadian Forest Service, Pacific Forestry Centre, Victoria, B.C. Information Report BC-X-348. Weetman, G.F. and A. Vyse. 1990. Natural regeneration. In Regenerating British Columbia's Forests. D.P. Lavender, R. Parish, C M . Johnson, G. Montgomery, A.Vyse, R.A. Willis and D. Winston (eds.) Govt, of Canada and Prov. of B.C., FRDA. UBC Press, Vancouver. Chap 10, pp 118-130 Weetman, G.F., R . C Yang, R . C and I.E. Bella. 1985. Nutrition and fertilization of lodgepole pine. In Lodgepole pine: the species and its management: symposium proceedings; 1984 May 8-10 Spokane WA, 1984 May 14-16 Vancouver, BC . Office of Conferences and Institutes, Cooperative Extension, Pullman, WA. 225-232. Wert, S. and B.R. Thomas. 1981. Effects of skid roads on diameter, height, and volume growth in Douglas fir. Soil Science Society of America Journal 45(3): 629-632. Young, A. and N. Mitchell. 1994. Microclimate and vegetation edge effects in a fragmented podocarp broadleaf forest in New Zealand. Biological Conservation 67(1): 63-72. 102 Young, P.J. , B.D. Keeland, and R.R. Sharitz. 1995. Growth response of baldcypress {Taxodium disticum (L.) Rich.) to an altered hydrologic regime. American Midland Naturalist 133(2): 206-212. 103 A P P E N D I C E S 104 APPENDIX I Tally cards as used in the field sampling. 105 UBC TALLY CARD-- ROAD EDGE STUDY 1999 POLY LABEL: ROAD R/W WIDTH (m) 1 2 3 PLOT ORIENTATION: C R E W : P A G E OF Date: UTM: f romGPS NAD: N: E: POLY PINE STEMS/HA FROM S W E E P 1 2 3 UTM:from GIS NAD: N: E POLY TOT. STEMS/HA F R O M S W E E P 1 2 3 STEMS/HA PLOT SIZE 40 m X m B E C Z O N E PLOT SLOPE(%) S W E E P PLOT RADIUS (m): CONFIRMED ROAD A G E , F R O M : PLOT A S P E C T E D G E A S P E C T ROAD T Y P E : EST. ROAD A G E , FROM: ROAD B R G ROAD G R A D E ELEV (m) C O M M E N T S : PLOT Z O N E T R E E NO S P E C I E S C O D E DBH (cm) S A M P L E HT /CORE? HEIGHT (m) BASE LIVE CROWN(m) C R O W N C L A S S T R E E C L A S S DAMAGE C O D E S C O M M E N T S : 106 UBC TALLY CARD- ROAD EDGE STUDY 1999 POLY LABEL: ROAD R/W WIDTH (m) 1 2 3 PLOT ORIENTATION: C R E W : P A G E O F Date: PLOT Z O N E T R E E NO S P E C I E S C O D E DBH (cm) S A M P L E HT/CORE? HEIGHT (m) B A S E LIVE CROWN(m) C R O W N C L A S S T R E E C L A S S DAMAGE C O D E S C O M M E N T S : C O M M O N S P E C I E S C O D E S C R O W N C L A S S C O D E S T R E E C L A S S C O D E S Lodgepole pine PI Subalpine fir Bl White spruce Sw Black spruce Sb Englemann spruce Se Spruce hybrid Sx Aspen At Douglas fir Fd Paper birch Ep Black cottonwood Ac Willow W Dominant 1 Codominant 2 Intermediate 3 Suppressed 4 Veteran 5 Understory 6 No decay indicators present 1 One or more decay indicators present 2 Dead potential @ remeasure 3 Dead useless @ remeasure 4 Veteran 5 DAMAGE AND DISEASE C O D E S Conks C Scars S Dead or Broken top DT Rotton Branches RB Frost Crack FC Fork or Crook FK Mountain Pine Beetle IBM Mistletoe M 107 UBC TALLY CARD-- ROAD EDGE STUDY 1999 ROAD PRISM AND SITE INDEX POLY LABEL ROAD R/W WIDTH (m) PLOT ORIENTATION C R E W : P A G E O F 1 2 3 Date: UTM:from G P S NAD: N: E: POLY PINE STEMS/HA FROM S W E E P 1 2 3 UTM:from GIS NAD: N: E POLY TOT. STEMS/HA FROM S W E E P 1 2 3 S T E M S / H A PLOT SIZE B E C ZONE PLOT SLOPE(%) S W E E P PLOT RADIUS (m): CONFIRMED ROAD A G E , F R O M : 40 m X m PLOT A S P E C T E D G E A S P E C T ROAD T Y P E : EST. ROAD A G E , F R O M : ROAD B R G ROAD G R A D E ELEV (m) COMMENTS ROAD PRISM MEASUREMENTS STA BRG(deg) % SD (m) HD (m) SIDE SLOPES FROM CL TO R/W EDGE (% SLO / SD) SSL % / m SSR % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m SSL % / m SSR % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m SSL % / m SSR % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m SSL % / m SSR % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m % / m INITIAL B R G : SITE T R E E S TREE NO. HT(m) BHAGE 108 APPENDIX II General Linear Models to test for differences in zonal variables among zones, while controlling for density and site productivity changes. 109 O u t p u t 1 . Z o n a l b a s a l a r e a / h a ( B A _ H A _ Z N ) v e r s u s z o n e ( P _ Z 0 N E ) a n d C u r t i s ' RD ( C R D _ P T ) . ( B a s e d o n w e i g h t e d l e a s t s q u a r e s u s i n g 1 / C R D _ P T s q u a r e d ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 2 2 0 D e p e n d e n t V a r i a b l e : B A _ H A _ Z N W e i g h t : WT2 S o u r c e M o d e l E r r o r C o r r e c t e d T o t a l DF 5 2 1 4 2 1 9 R - S q u a r e 0 . 4 2 0 7 5 3 Sum o f S q u a r e s 177 . 6 2 7 0 0 1 7 9 2 4 4 . 5 3 7 1 0 2 7 7 4 2 2 . 1 6 4 1 0 4 5 7 C . V . 4 . 6 1 9 7 9 1 F V a l u e 3 1 . 0 9 P r > F 0 . 0 0 0 1 B A _ H A _ Z N M e a n 23 . 1 3 8 9 2 3 5 S o u r c e P _ Z 0 N E C R D _ P T S o u r c e P _ Z 0 N E C R D _ P T DF T y p e I SS 4 3 4 . 2 9 6 8 9 6 8 6 1 1 4 3 . 3 3 0 1 0 4 9 3 DF T y p e I I I SS 4 3 4 . 2 9 6 8 9 6 8 6 1 1 4 3 . 3 3 0 1 0 4 9 3 F V a l u e 7 . 5 0 1 2 5 . 4 3 F V a l u e 7 . 50 1 2 5 . 4 3 P r > F 0 . 0 0 0 1 0 . 0 0 0 1 P r > F 0 . 0 0 0 1 0 . 0 0 0 1 L e a s t S q u a r e s M e a n s P _ Z 0 N E B A _ H A _ Z N L S M E A N L S M E A N N u m b e r 1 3 0 . 6 3 4 1 6 7 7 1 2 25 . 0 3 8 8 0 0 5 2 3 24 . 3 4 4 3 6 0 9 3 4 2 4 . 3 9 4 0 4 3 0 4 5 2 4 . 4 1 0 3 2 1 9 5 T f o r HO L S M E A N ( i ) = L S M E A N ( j / P r > T | i / j 1 2 3 4 5 1 3 . 9 5 9 4 8 8 4 . 4 5 0 8 9 9 4 . 4 1 5 7 4 2 4 . 4 0 4 2 2 3 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 2 - 3 . 9 5 9 4 9 0 . 4 9 1 4 1 1 0 . 4 5 6 2 5 4 0 . 4 4 4 7 3 5 0 . 0 0 0 1 0 . 6 2 3 6 0 . 6 4 8 7 0 . 6 5 7 0 3 - 4 . 4 5 0 9 - 0 . 4 9 1 4 1 - 0 . 0 3 5 1 6 - 0 . 0 4 6 6 8 0 . 0 0 0 1 0 . 6 2 3 6 0 . 9 7 2 0 0 . 9 6 2 8 4 - 4 . 4 1 5 7 4 - 0 . 4 5 6 2 5 0 . 0 3 5 1 5 7 - 0 . 0 1 1 5 2 0 . 0 0 0 1 0 . 6 4 8 7 0 . 9 7 2 0 0 . 9 9 0 8 5 - 4 . 4 0 4 2 2 - 0 . 4 4 4 7 3 0 . 0 4 6 6 7 6 0 . 0 1 1 5 2 0 . 0 0 0 1 0 . 6 5 7 0 0 . 9 6 2 8 0 . 9 9 0 8 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 110 P l o t o f W T R E S I D * W T Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . W T R E S I D 4 - 4 54 2 3 2 5 24 5 4 2 2 5 2 5 4 2 3 4 3 5 4 4 3 34 3 3 3 2 2 2 5 5 4 3 3 5 3 22 3 2 4 5 2 2 2 4 2 3 5 3 3 2 4 3 4 3 44 5 53 5 4 2 4 2 4 2 322 3 2 3 3 52 52 1 1 2 12 11 2 11 2 52 1 j 1 1 1 1 1 1 1 1 s~fffffffff"fffffffff"fffffffff"fffffffff"fffffffff~fffffffff' 3 . 0 3 . 5 4 . 0 4 . 5 5 . 0 5 . 5 6 . 0 N O T E : 87 o b s h i d d e n . V a r i a b l e = W T R E S I D WTYHAT U n i v a r i a t e P r o c e d u r e M o m e n t s N 2 2 0 Sum W g t s 2 2 0 M e a n 0 Sum 0 S t d D e v 1 . 0 5 6 6 9 7 V a r i a n c e 1 . 1 1 6 6 0 8 S k e w n e s s - 0 . 2 8 2 6 K u r t o s i s 0 . 5 1 4 0 0 4 U S S 2 4 4 . 5 3 7 1 C S S 2 4 4 . 5 3 7 1 C V S t d M e a n 0 . 0 7 1 2 4 2 T : M e a n = 0 0 Pr>1T1 1 . 0 0 0 0 Num A = 0 2 2 0 Num > 0 1 1 0 M ( S i g n ) 0 P r > = | M 1 . 0 0 0 0 S g n R a n k 2 9 2 Pr>= j S 0 . 7 5 8 2 W : N o r m a l 0 . 9 8 2 3 1 1 Pr<W 0 . 4 5 7 6 111 U n i v a r i a t e P r o c e d u r e V a r i a b l e = W T R E S I D S t e m L e a f # 2 567 3 2 004 3 1 5 5 5 5 6 6 7 7 7 7 7 9 12 1 0 0 0 0 0 0 0 0 0 0 1 1 2 2 2 3 3 3 3 4 4 21 0 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 9 9 9 9 36 + 0 0 0 0 0 1 1 1 1 1 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 35 | - 0 4 4 4 4 4 4 4 4 4 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 44 * - 0 9 9 9 9 8 8 7 7 7 7 7 7 6 6 6 5 5 5 5 5 5 5 22 + - 1 4 4 4 3 3 3 3 2 2 2 2 2 2 1 1 1 1 1 0 0 0 0 0 0 0 0 0 27 - 1 9 9 7 6 6 6 5 5 5 9 - 2 2 2 2 1 4 - 2 86 2 - 3 4 1 - 3 5 1 N o r m a l P r o b a b i l i t y P l o t 2 . 7 5 + + * *++ ****** * * * * * * * * * * * * * * * * * * * * * * * * * * * * _|_ * * * * * * * * * * * . * * * -3 . 7 5 + + + -- 2 - - + + + -+ 1 +2 112 O u t p u t 2 . Z o n a l C u r t i s ' RD (CRD_ZN) v e r s u s zone ( P _ Z 0 N E ) , C u r t i s ' RD ( C R D _ P T ) , and s i t e i n d e x ( A V _ S I _ 5 0 ) . (Based on w e i g h t e d l e a s t s q u a r e s u s i n g 1 /CRD_PT s q u a r e d ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 2 2 0 D e p e n d e n t V a r i a b l e : C R D _ Z N W e i g h t : WT2 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 6 1 0 . 0 2 2 8 7 4 1 2 27 . 93 0 . 0 0 0 1 E r r o r 213 12 . 7 3 9 4 5 8 8 9 C o r r e c t e d T o t a l 2 1 9 22 . 7 6 2 3 3 3 0 1 R- S q u a r e C . V . C R D _ Z N M e a n 0 . 4 4 0 3 2 7 4 . 4 4 0 4 1 8 5 . 5 0 7 5 9 2 3 2 S o u r c e DF T y p e I SS F V a l u e P r > F P _ Z 0 N E 4 1 . 9 8 2 2 4 4 9 8 8 . 2 9 0 . 0 0 0 1 C R D _ P T 1 6 . 9 8 7 2 6 5 4 8 1 1 6 . 8 3 0 . 0 0 0 1 A V _ S I _ 5 0 1 1 . 0 5 3 3 6 3 6 6 17 . 61 0 . 0 0 0 1 S o u r c e DF T y p e I I I SS F V a l u e P r > F P _ Z 0 N E 4 1 . 9 8 2 2 4 4 9 8 8 . 2 9 0 . 0 0 0 1 C R D _ P T 1 7 . 9 1 2 0 3 9 1 7 1 3 2 . 2 9 0 . 0 0 0 1 A V _ S I _ 5 0 1 1 . 0 5 3 3 6 3 6 6 17 . 61 0 . 0 0 0 1 L e a s t S q u a r e s M e a n s P _ Z O N E C R D _ Z N L S M E A N L S M E A N N u m b e r 1 7 . 2 4 8 4 6 2 2 3 1 2 5 . 9 4 5 6 1 1 2 4 2 3 5 . 7 5 9 2 6 0 8 7 3 4 5 . 7 1 0 6 6 9 0 6 4 5 5 . 7 4 9 0 9 0 5 9 5 T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T i / j 1 2 3 4 5 1 4 .029814 4 .606209 4 756507 4 . 637667 0 .0001 0 .0001 0 .0001 0 . 0001 2 -4 . 02981 0 .576395 0 726693 0 . 607853 0 . 0001 0 .5650 0 .4682 0 .5439 3 - 4 . 6 0 6 2 1 - 0 . 5 7 6 4 0 150298 0 . 031457 0 . 0001 0 . 5650 0 .8807 0 .9749 4 -4 .75651 -0 .72669 - 0 . 1 5 0 3 -0 .11884 0 . 0001 0 .4682 0 .8807 0 .9055 5 - 4 . 6 3 7 6 7 - 0 . 6 0 7 8 5 -0 .03146 0 118841 0 . 0001 0 .5439 0 .9749 0 . 9055 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 113 P l o t o f W T R E S I D * W T Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . W T R E S I D 1 . 0 0 . 5 0 . 0 -0 . 5 - 1 . 0 4 5 1 4 5 5 12 35 5 4 5 4 2 1 44 3 2 4 3 2 2 3 2 4 5 3 2 2 2 5 4 4 3 3 3 552 11 3 3 4 4 5 4 3 4 4 4 3 4 2 3 1 5 5 5 4 5 3 5 53 2 2 5 2 1 2 4 5 4 5 2 3 3 2 2 1 4 2 1 1 5 1 5 4 3 2 2 5 4 3 4 3 4 2 3 5 4 5 5 4 5 2 4 3 3 2 24 32 1 2 4 14 2 1 1 553 1 1 1 3 £ff~fffffffffff~fffffffffff"fffffffffff~fffffffffff"ff 0 . 6 0 . 8 1 . 0 1 . 2 1 . 4 N O T E : 60 o b s h i d d e n . V a r i a b l e = W T R E S I D WTYHAT U n i v a r i a t e P r o c e d u r e M o m e n t s N 2 2 0 Sum W g t s 2 2 0 M e a n 0 Sum 0 S t d D e v 0 . 2 4 1 1 8 7 V a r i a n c e 0 . 0 5 8 1 7 1 S k e w n e s s - 0 . 3 5 3 4 2 K u r t o s i s 0 . 3 9 0 9 8 9 U S S 12 . 7 3 9 4 6 C S S 12 . 7 3 9 4 6 C V S t d M e a n 0 . 0 1 6 2 6 1 T : M e a n = 0 0 P r > | T | 1 . 0 0 0 0 Num ~= 0 2 2 0 Num > 0 117 M ( S i g n ) 7 P r > = | M | 0 . 3 8 0 8 S g n R a n k 396 Pr>= j S | 0 . 6 7 6 2 W : N o r m a l 0 . 9 8 5 5 1 4 Pr<W 0 . 7 0 5 3 114 U n i v a r i a t e P r o c e d u r e V a r i a b l e = W T R E S I D S t e m L e a f 6 8 5 2 4 2 4 5 8 8 3 0 1 1 2 4 5 5 5 6 7 8 8 9 2 0 0 0 0 0 0 1 1 1 3 3 3 3 3 3 4 4 4 5 5 6 7 7 7 8 8 8 9 9 1 0 0 1 2 2 3 3 3 3 4 4 4 5 5 5 5 6 6 6 6 8 8 8 8 9 9 9 0 0 0 0 1 1 1 2 2 3 3 3 3 3 4 4 4 5 5 5 5 6 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 9 - 0 9 9 9 9 9 9 8 8 7 7 7 7 7 6 6 6 5 5 5 5 4 4 3 3 2 2 2 2 2 1 1 1 - 1 9 9 9 8 8 8 7 7 7 7 6 6 6 5 5 4 4 4 3 3 2 2 2 1 1 1 1 0 0 - 2 9 5 5 5 5 5 4 4 3 3 3 2 1 0 0 - 3 8 6 6 4 3 3 3 2 2 2 2 1 0 0 - 4 8 7 6 3 3 2 - 5 7 5 3 2 0 - 6 - 7 8 - 8 0 M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 # 1 1 5 13 29 27 41 32 29 15 14 6 B o x p l o t 0 + + * | * V a r i a b l e = W T R E S I D 0 . 65 + 0 . 3 5 + 0 . 05 + - 0 . 2 5 + U n i v a r i a t e P r o c e d u r e N o r m a l P r o b a b i l i t y P l o t + + + ' _ * * * * _ * * * * * * * * * * * ***** * * * * * * ***** ***** * * * * + * * * * . * * * * - 0 . 5 5 + +*+** 1 + - 0 . 85 + * + -- 1 - - + + + -+ 1 +2 115 O u t p u t 3 . Z o n a l mean l i v e crown r a t i o ( l o g a r i t h m ; InLCR) v e r s u s zone ( P _ Z 0 N E ) , b a s a l a r e a / h a ( B A _ H A _ P T ) , and s i t e i n d e x ( A V _ S I _ 5 0 ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t D e p e n d e n t V a r i a b l e : L N L C R S o u r c e DF M o d e l 6 E r r o r 213 C o r r e c t e d T o t a l 2 1 9 R - S q u a r e 0 . 4 1 4 2 9 0 Sum o f S q u a r e s 1 0 . 3 9 6 6 8 2 3 4 1 4 . 6 9 8 5 1 6 8 8 25 . 0 9 5 1 9 9 2 2 C . V . - 2 6 . 3 4 3 1 8 = 2 2 0 F V a l u e 2 5 . 1 1 P r > F 0 . 0 0 0 1 L N L C R M e a n - 0 . 9 9 7 1 9 1 7 7 S o u r c e DF T y p e I SS F V a l u e P _ Z O N E 4 0 . 9 7 4 3 3 1 9 1 3 . 53 B A _ H A _ P T 1 6 . 3 3 1 7 8 6 8 1 9 1 . 7 6 A V _ S I _ 5 0 1 3 . 0 9 0 5 6 3 6 2 4 4 . 7 9 S o u r c e DF T y p e I I I SS F V a l u e P _ Z O N E 4 0 . 9 7 4 3 3 1 9 1 3 . 53 B A _ H A _ P T 1 9 . 1 1 7 9 6 0 8 5 1 3 2 . 1 3 A V _ S I _ 5 0 1 3 . 0 9 0 5 6 3 6 2 44 . 7 9 L e a s t S q u a r e s M e a n s P _ ZONE L N L C R L S M E A N L S M E A N N u m b e r 1 - 0 8 8 5 0 1 4 7 5 1 2 - 0 9 6 0 4 8 0 7 7 2 3 - 1 0 6 7 2 8 6 6 6 3 4 - 1 0 2 6 7 1 2 6 9 4 5 - 1 0 4 6 4 6 3 9 6 5 T f o r HO : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 1 1 . 3 4 7 4 6 3 . 2 5 4 4 9 9 2 . 5 3 0 0 4 3 2 . 8 8 2 7 0 6 0 . 1 7 9 3 0 . 0 0 1 3 0 . 0 1 2 1 0 . 0 0 4 3 2 - 1 . 3 4 7 4 6 1 . 9 0 7 0 3 9 1 . 1 8 2 5 8 3 1 . 5 3 5 2 4 6 0 . 1 7 9 3 0 . 0 5 7 9 0 . 2 3 8 3 0 . 1 2 6 2 3 - 3 . 2 5 4 5 - 1 . 9 0 7 0 4 - 0 . 7 2 4 4 6 - 0 . 3 7 1 7 9 0 . 0 0 1 3 0 . 0 5 7 9 0 . 4 6 9 6 0 . 7 1 0 4 4 - 2 . 5 3 0 0 4 - 1 . 1 8 2 5 8 0 . 7 2 4 4 5 6 0 . 3 5 2 6 6 3 0 . 0 1 2 1 0 . 2 3 8 3 0 . 4 6 9 6 0 . 7 2 4 7 5 - 2 . 8 8 2 7 1 - 1 . 5 3 5 2 5 0 . 3 7 1 7 9 3 - 0 . 3 5 2 6 6 0 . 0 0 4 3 0 . 1 2 6 2 0 . 7 1 0 4 0 . 7 2 4 7 P r > F 0 . 0 0 8 2 0 . 0 0 0 1 0 . 0 0 0 1 P r > F 0 . 0 0 8 2 0 . 0 0 0 1 0 . 0 0 0 1 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 116 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 1 . 0 0 . 5 0 . 0 -0 . 5 - 1 . 0 3 2 2 3 14 5 5 2 42 5 54 1 35 4 5 4 4 15 3 5 4 4 5 3 2 1 3 2 1 1 2 5 4 3 2 4 4 4 2 3 15 45 5 51 54 3 3 1 33 42 3 2 1 1 514 1 5 3 5 5 1 4 5 4 22 45 3 3 2 2 3 14 4 2 4 2 5 1 4 3 3 5 1 23 5 2 3 544 4 2 33 1 5 2 5 33 3 1 5 2 4 54 132 3 4 3 5 1 2 4 2 2 5 33 1 3 2 4 5 3 4 22 3 1 4 3 i. 2 4 1 3 35 1 45 5 5 2 32 3 44 22 2 5 1 1 5 1 2 2 Sf"fffffffffff"fffffffffff"fffffffffff"fffffffffff"fffffffffff~f - 1 . 4 - 1 . 2 - 1 . 0 - 0 . 8 - 0 . 6 - 0 . 4 N O T E : 3 0 o b s h i d d e n . V a r i a b l e = R E S I D YHAT U n i v a r i a t e P r o c e d u r e M o m e n t s N 2 2 0 Sum W g t s 2 2 0 M e a n 0 Sum 0 S t d D e v 0 . 2 5 9 0 6 9 V a r i a n c e 0 . 0 6 7 1 1 7 S k e w n e s s - 0 . 0 1 8 7 5 K u r t o s i s -0 . 2 7 0 4 U S S 14 . 6 9 8 5 2 C S S 14 . 6 9 8 5 2 CV S t d M e a n 0 . 0 1 7 4 6 6 T : M e a n = 0 0 P r > | T | 1 . 0 0 0 0 Num ~= 0 2 2 0 Num > 0. 1 1 2 M ( S i g n ) 2 P r > = | M | 0 . 8 3 9 8 S g n R a n k - 1 7 P r>= j S j 0 . 9 8 5 7 W : N o r m a l 0 . 9 8 2 2 5 2 Pr<W 0 . 4 5 3 0 117 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D S t e m L e a f 6 027 0 0 0 1 3 3 3 3 3 4 4 6 6 6 9 1 1 3 4 7 9 0 1 1 1 2 3 4 4 4 4 5 5 5 6 6 6 6 7 7 7 7 8 8 8 8 1 0 0 0 1 1 1 1 1 1 1 1 2 3 3 4 4 4 4 4 6 6 6 6 6 7 7 8 9 9 9 9 0 0 0 1 1 1 2 2 2 2 2 2 3 3 3 3 3 4 4 6 6 6 6 6 7 7 7 8 9 9 9 9 - 0 9 9 9 8 8 8 8 7 7 7 6 6 6 5 5 4 4 4 3 3 3 3 2 2 2 2 2 2 2 2 2 1 1 - 1 9 8 7 6 6 5 5 4 3 3 2 2 2 1 1 1 0 0 0 - 2 9 9 9 8 8 8 8 7 7 6 6 6 5 5 4 4 3 3 1 1 0 0 0 0 - 3 8 7 6 6 5 5 5 4 3 3 2 2 1 1 0 0 0 0 0 - 4 9 7 7 6 6 4 2 1 0 - 5 54 - 6 8 - 7 1 M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 # 3 1 15 6 25 31 31 33 19 24 19 9 2 1 1 B o x p l o t + + * i * U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 0 . 65 + - 0 . 0 5 + N o r m a l P r o b a b i l i t y P l o t * * * + *++ * * * * * * * * * * _j_ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * + + * - 0 . 7 5 + * + + + + + -- 2 - 1 0 +1 +2 118 O u t p u t 4 . Z o n a l s t e m s / h a r e l a t i v e t o Zone 5 (LNRELSPH) v e r s u s zone (P_Z0NE) and b a s a l a r e a / h a ( B A _ H A _ P T ) , and s i t e i n d e x ( A V _ S I _ 5 0 ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 17 6 D e p e n d e n t V a r i a b l e : L N R L S P H S o u r c e M o d e l E r r o r C o r r e c t e d T o t a l DF 5 1 7 0 1 7 5 R - S q u a r e 0 . 0 9 6 9 7 4 Sum o f S q u a r e s 3 . 5 9 8 7 5 7 3 5 3 3 . 5 1 1 6 5 4 2 2 37 . 1 1 0 4 1 1 5 7 C . V . 6 8 8 . 4 6 4 8 F V a l u e 3 . 65 P r > F 0 . 0 0 3 7 L N R L S P H M e a n 0 . 0 6 4 4 8 9 9 0 S o u r c e DF T y p e I S S F V a l u e P r > F P _ Z 0 N E A V _ S I _ 5 0 BA H A _ P T 0 .78292811 0 .28402082 2 .53180842 1.32 1.44 12 .84 0 .2683 0 .2317 0 .0004 S o u r c e DF T y p e I I I SS F V a l u e P r > F P _ Z 0 N E A V _ S I _ 5 0 BA HA PT 0 . 7 8 2 9 2 8 1 1 2 . 2 6 7 4 0 1 7 8 2 . 5 3 1 8 0 8 4 2 1 . 3 2 1 1 . 5 0 12 . 84 0 . 2 6 8 3 0 . 0 0 0 9 0 . 0 0 0 4 L e a s t S q u a r e s M e a n s P _ Z O N E L N R L S P H T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T L S M E A N i / j 1 2 3 4 1 0 1 7 6 6 5 2 3 2 1 1 . 3 8 4 0 8 6 1 . 5 0 9 9 5 7 1 8 4 5 5 9 5 0 . 1 6 8 1 0 . 1 3 2 9 0 . 0 6 6 7 2 0 0 4 5 6 3 6 0 4 2 - 1 . 3 8 4 0 9 0 . 1 2 5 8 7 1 0 4 6 1 5 0 9 0 . 1 6 8 1 0 . 9 0 0 0 0 . 6 4 5 0 3 0 0 3 3 7 2 1 2 1 3 - 1 . 5 0 9 9 6 - 0 . 1 2 5 8 7 0 3 3 5 6 3 8 0 . 1 3 2 9 0 . 9 0 0 0 0 . 7 3 7 6 4 0 0 0 1 9 5 0 0 5 4 - 1 . 8 4 5 5 9 - 0 . 4 6 1 5 1 -0 . 3 3 5 6 4 0 . 0 6 6 7 0 . 6 4 5 0 0 . 7 3 7 6 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 119 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 2 2 1 2 4 3 3 3 1 3 4 2 1 4 2 2 4 4 4 3 2 42 1 2 4 4 4 2 24 2 13 4 4 4 2 4 3 3 4 32 21 4 3 3 4 3 32 4 3 2 2 2 3 3 1 1 1 2 1 21 2 3 3 2 3 2 13 4 1 1 2 4 2 1 1 34 33 4 2 4 1 1 3 4 33 2 1 4 1 1 4 1 32 4 1 4 1 2 2 32 41 1 2 1 11 1 13 43 1 1 1 2 §//"fffffffffff"fffffffffff~ fffffffffff"fffffffffff"// - 0 . 2 0 . 0 0 . 2 0 . 4 0 . 6 YHAT N O T E : 33 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D M o m e n t s N M e a n S t d D e v S k e w n e s s U S S C V T : M e a n = 0 Num A = 0 M ( S i g n ) S g n R a n k W : N o r m a l 176 0 0 . 4 3 7 6 0 2 0 . 0 7 8 2 2 3 3 . 5 1 1 6 5 0 176 4 73 0 . 9 8 3 2 6 3 Sum W g t s Sum V a r i a n c e K u r t o s i s C S S S t d M e a n P r > | T | Num > 0 Pr>= | M Pr>= j S Pr<W 1 7 6 0 0 . 1 9 1 4 9 5 0 . 6 4 3 7 9 3 3 3 . 5 1 1 6 5 0 . 0 3 2 9 8 5 1 . 0 0 0 0 92 0 . 5 9 7 9 0 . 9 1 4 5 0 . 5 8 2 1 120 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D S t e m L e a f 14 3 12 10 19 8 0 0 5 5 6 8 9 3 5 6 4 0 1 4 6 6 6 7 7 9 4 4 6 6 7 7 8 2 0 0 0 2 3 3 3 5 5 7 8 0 2 3 3 4 5 7 7 8 8 9 0 1 1 1 2 2 2 3 3 4 4 5 5 5 7 7 7 8 8 9 9 9 9 9 0 0 1 2 2 2 3 3 4 4 5 5 5 5 7 8 9 9 9 - 0 8 8 7 7 6 6 5 5 5 4 1 1 0 9 8 8 8 7 7 6 6 6 5 5 4 4 1 1 1 1 1 - 2 9 6 6 6 6 4 4 2 2 2 9 8 8 7 7 3 3 3 3 3 2 0 0 0 0 - 4 7 6 3 7 6 3 3 2 1 0 - 6 7 3 2 1 0 5 4 3 2 - 8 7 6 0 0 8 6 3 - 1 0 00 M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D N o r m a l P r o b a b i l i t y P l o t 1 . 5 + * + * + ** + + _i_ * * * * * * * * * * + * * * * * * * * * * * * * * * * * * * * * * + * * * + + * * * * * * _j_ * * * * - 1 . 1 + * + + + + + + + + + + + + + - 2 - 1 0 +1 +2 # B o x p l o t 1 0 2 0 4 0 5 I 16 I 22 + + 42 * — + — * 3 1 I I 25 + + 10 | 9 I 7 0 2 0 121 O u t p u t 5 . Z o n a l b a s a l a r e a / h a r e l a t i v e t o Zone 5 (RLBAHAZ) v e r s u s zone (P_Z0NE) and b a s a l a r e a / h a ( B A _ H A _ P T ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 17 6 D e p e n d e n t V a r i a b l e : R L B A H A Z S o u r c e M o d e l E r r o r C o r r e c t e d T o t a l DF 4 1 7 1 1 7 5 R - S q u a r e 0 . 0 9 1 6 5 4 Sum o f S q u a r e s 2 . 7 5 2 6 2 6 3 0 27 . 2 8 0 2 6 1 4 4 30 . 0 3 2 8 8 7 7 4 C . V . 35 . 6 7 4 0 0 F V a l u e 4 . 3 1 P r > F 0 . 0 0 2 4 R L B A H A Z M e a n 1 . 1 1 9 6 2 9 6 4 S o u r c e P _ Z 0 N E B A _ H A _ P T S o u r c e P _ Z 0 N E BA HA PT DF T y p e I SS 3 2 . 2 4 0 9 7 7 8 3 1 0 . 5 1 1 6 4 8 4 7 DF T y p e I I I SS 3 2 . 2 4 0 9 7 7 8 3 1 0 . 5 1 1 6 4 8 4 7 F V a l u e 4 . 6 8 3 . 2 1 F V a l u e 4 . 68 3 . 2 1 P r > F 0 . 0 0 3 6 0 . 0 7 5 1 P r > F 0 . 0 0 3 6 0 . 0 7 5 1 P _ Z 0 N E 1 2 3 4 R L B A H A Z L S M E A N 1 . 3 1 2 5 2 4 5 8 1 . 0 8 4 9 8 8 6 9 1 . 0 3 9 9 1 2 5 8 1 . 0 4 1 0 9 2 7 0 L e a s t S q u a r e s M e a n s T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 1 - 2 . 6 7 1 9 9 0 . 0 0 8 3 - 3 . 2 0 1 3 3 0 . 0 0 1 6 - 3 . 1 8 7 4 7 0 . 0 0 1 7 2 . 6 7 1 9 9 1 3 . 2 0 1 3 2 8 3 . 1 8 7 4 6 9 0 . 0 0 8 3 0 . 0 0 1 6 0 . 0 0 1 7 0 . 5 2 9 3 3 6 0 . 5 1 5 4 7 8 0 . 5 9 7 3 - 0 . 5 2 9 3 4 0 . 5 9 7 3 - 0 . 5 1 5 4 8 0 . 0 1 3 8 5 8 0 . 6 0 6 9 0 . 0 1 3 8 6 0 . 9 8 9 0 0 . 6 0 6 9 0 . 9 8 9 0 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 122 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 1 . 5 1 . 0 0 . 5 0 . 0 - 0 . 5 - 1 . 0 3 3 2 4 2 4 3 2 2 4 3 3 4 3 3 2 3 3 2 3 3 4 2 4 4 4 3 3 3 2 2 2 4 4 243 3 2 2 4 2 2 4 3 3 4 2 2 2 4 4 4 2 2 2 2 4 3 4 3 3 3 3 2 3 3 2 2 3 34 4 4 3 4 22 2 4 2 3 2 4 3 3 2 3 2 3 2 2 4 43 22 3 4 4 2 4 3 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 11 1 1 1 1 ] 1 1 1 1 1 1 11 11 $f"fffffffff"fffffffff~fffffffff~ffffffffffffffffff"fffffffffff 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 1 . 4 1 . 5 N O T E : 2 6 o b s h i d d e n . V a r i a b l e = R E S I D Y H A T U n i v a r i a t e P r o c e d u r e M o m e n t s N 176 Sum W g t s 176 M e a n 0 Sum 0 S t d D e v 0 . 3 9 4 8 2 6 V a r i a n c e 0 . 1 5 5 8 8 7 S k e w n e s s 0 . 3 9 0 5 2 2 K u r t o s i s 0 . 5 2 5 6 1 6 U S S 27 . 2 8 0 2 6 C S S 27 . 2 8 0 2 6 C V S t d M e a n 0 . 0 2 9 7 6 1 T : M e a n = 0 0 P r>1T1 1 . 0 0 0 0 Num A = 0 176 Num > 0 87 M ( S i g n ) - 1 P r > = | M 0 . 9 3 9 9 S g n R a n k - 1 9 5 P r > = | S 0 . 7 7 4 2 W : N o r m a l 0 . 9 8 0 1 7 5 Pr<W 0 . 3 7 1 6 123 V a r i a b l e = R E S I D U n i v a r i a t e P r o c e d u r e S t e m L e a f 13 0 12 11 10 9 1 18 8 5 7 0 6 237 5 0 0 2 2 2 3 5 5 8 9 4 0 1 1 3 4 7 9 3 0 2 2 4 8 9 2 0 0 1 2 2 4 5 5 6 6 7 7 8 9 1 0 0 0 1 1 1 2 2 3 4 4 4 5 5 5 5 6 6 6 6 7 8 9 9 9 0 0 2 2 3 4 4 4 5 6 6 7 7 8 8 9 - 0 9 9 8 7 5 5 4 4 4 4 4 3 3 2 2 1 1 0 0 - 1 8 7 6 6 4 3 2 2 2 1 0 0 - 2 8 8 7 5 5 4 3 2 2 2 1 0 0 0 - 3 9 9 7 6 6 6 6 6 5 3 2 2 1 0 - 4 9 8 8 8 7 7 7 6 6 5 5 5 4 0 - 5 8 4 3 2 1 1 0 - 6 9 5 3 2 1 0 - 7 9 - 8 7 - 9 - 1 0 3 M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 # 1 1 1 2 1 1 3 10 7 6 14 25 15 19 12 14 14 14 7 6 1 1 B o x p l o t 0 0 0 0 V a r i a b l e = R E S I D 1 . 35 + U n i v a r i a t e P r o c e d u r e N o r m a l P r o b a b i l i t y P l o t * * * * * * * * 0 . 1 5 + * * * * * * * * * * * * * * * * * * * * * * * * * * + + + - 1 . 05 + * - 1 + 1 +2 124 O u t p u t 6. Z o n a l C u r t i s ' RD r e l a t i v e t o Zone 5 (RLCRD_ZN) v e r s u s zone ( P _ Z 0 N E ) , s i t e i n d e x (AV_SI_50) and b a s a l a r e a / h a ( B A _ H A _ P T ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 17 6 D e p e n d e n t V a r i a b l e : R L C R D _ Z N S o u r c e DF Sum o f S q u a r e s M o d e l E r r o r C o r r e c t e d T o t a l 5 1 7 0 1 7 5 R - S q u a r e 0 . 1 0 8 3 9 0 3 . 2 1 1 1 7 7 0 5 2 6 . 4 1 5 0 8 2 6 9 29 . 6 2 6 2 5 9 7 4 C . V . 3 5 . 1 1 1 3 3 F V a l u e 4 . 1 3 P r > F 0 . 0 0 1 4 R L C R D _ Z N M e a n 1 . 1 2 2 6 7 5 4 1 S o u r c e P _ Z O N E A V _ S I _ 5 0 B A _ H A _ P T S o u r c e P _ Z O N E A V _ S I _ 5 0 BA HA PT DF T y p e I SS 3 2 . 0 0 4 1 1 8 0 8 1 0 . 0 0 4 4 3 9 5 6 1 1 . 2 0 2 6 1 9 4 1 DF T y p e I I I SS 3 2 . 0 0 4 1 1 8 0 8 1 0 . 6 8 2 7 4 7 4 3 1 1 . 2 0 2 6 1 9 4 1 F V a l u e 4 . 3 0 0 . 03 7 . 7 4 F V a l u e 4 . 3 0 4 . 3 9 7 . 7 4 P r > F 0 . 0 0 5 9 0 . 8 6 6 0 0 . 0 0 6 0 P r > F 0 . 0 0 5 9 0 . 0 3 7 5 0 . 0 0 6 0 P _ Z O N E 1 2 3 4 R L C R D _ Z N L S M E A N 1 . 3 0 4 5 0 8 2 6 1 . 0 9 2 9 8 7 0 3 1 . 0 5 0 4 4 8 8 4 1 . 0 4 2 7 5 7 4 9 L e a s t S q u a r e s M e a n s T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 1 2 . 5 1 6 8 8 7 3 . 0 2 3 0 4 9 3 . 1 1 4 5 6 8 0 . 0 1 2 8 0 . 0 0 2 9 0 . 0 0 2 2 0 . 5 0 6 1 6 1 0 . 6 1 3 4 2 - 2 . 5 1 6 8 9 0 . 0 1 2 8 3 - 3 . 0 2 3 0 5 - 0 . 5 0 6 1 6 0 . 0 0 2 9 0 . 6 1 3 4 4 - 3 . 1 1 4 5 7 - 0 . 5 9 7 6 8 - 0 . 0 9 1 5 2 0 . 0 0 2 2 0 . 5 5 0 8 0 . 9 2 7 2 0 . 5 9 7 6 8 0 . 5 5 0 8 0 . 0 9 1 5 1 9 0 . 9 2 7 2 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 125 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 1 . 5 1 . 0 0 . 5 0 . 0 - 0 . 5 - 1 . 0 4 23 3 2 33 1 3 2 44 4 3 3 44 34 3 4 2 2 3 1 24 42 2 2 2 3 3 2 2 4 4 4 2 4 3 33 4 34 2 1 3 3 32 2 4 3 4 22 4 2 4 33 3 4 4 3 3 4 2 2 4 4 3 2 4 3 4 42 2 4 14 3 3 234 4 21 3 2 32 3 3 2 1 3 4 4 2 4 22 4 2 1 21 4 4 3 1 2 11 11 1 1 11 2 §/"/////////"/////////"fffffffirfffffffff~ fffffffff"fffffffff~/ 0 . 9 1 . 0 1 . 1 1 . 2 YHAT 1 . 3 1 . 4 1 . 5 N O T E : 23 o b s h i d d e n . V a r i a b l e = R E S I D U n i v a r i a t e P r o c e d u r e M o m e n t s N 176 Sum W g t s 176 M e a n 0 Sum 0 S t d D e v 0 . 3 8 8 5 1 4 V a r i a n c e 0 . 1 5 0 9 4 3 S k e w n e s s 0 . 4 9 6 8 7 8 K u r t o s i s 0 . 7 9 0 2 6 4 U S S 26 . 4 1 5 0 8 C S S 26 . 4 1 5 0 8 C V S t d M e a n 0 . 0 2 9 2 8 5 T : M e a n = 0 0 Pr>1T1 1 . 0 0 0 0 Num A = 0 176 Num > 0 80 M ( S i g n ) - 8 P r > = M 0 . 2 5 8 1 S g n R a n k - 2 9 8 P r > = S 0 . 6 6 1 0 W : N o r m a l 0 . 9 7 8 2 2 9 Pr<W 0 . 2 5 9 5 126 V a r i a b l e = R E S I D U n i v a r i a t e P r o c e d u r e S t e m L e a f # 13 8 1 12 2 1 11 4 1 10 9 8 9 1 7 1 2 6 9 4 6 56 2 5 0 0 1 1 2 2 7 7 8 4 0 1 2 3 4 8 9 7 3 0 0 1 1 3 4 6 8 8 8 8 9 12 2 1 2 2 3 3 4 6 7 9 9 9 11 1 0 0 1 2 2 4 4 4 4 4 4 4 5 6 7 7 7 8 8 19 0 0 2 2 2 3 3 4 5 5 6 8 8 8 13 - 0 9 9 8 8 8 8 7 7 7 6 6 5 5 4 4 3 3 2 2 2 1 1 1 1 1 1 0 27 - 1 9 8 7 7 7 7 3 3 2 1 1 1 1 0 0 15 - 2 9 9 6 4 4 3 3 3 2 1 1 0 12 - 3 7 7 6 6 6 5 4 3 3 3 2 1 0 13 - 4 9 9 9 8 8 7 7 6 6 6 5 3 3 3 2 1 16 - 5 9 6 3 3 1 5 - 6 9 8 2 0 4 - 7 40 2 - 8 4 1 - 9 3 1 M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 B o x p l o t 0 0 0 V a r i a b l e = R E S I D N o r m a l P r o b a b i l i t y P l o t 1 . 3 5 + * *. * * * * * * * * * * * * * * * * * * + * * * * * * * * * * * * * * * * * * * * * * * * * * + + * * *++ + + -0 . 9 5 + * + + - - - - + -- 2 - - + -+ 1 - - + -+2 127 O u t p u t 7 . Z o n a l mean h e i g h t r e l a t i v e t o Zone 5 (RLHT_ZN) v e r s u s zone ( P _ Z 0 N E ) , s i t e i n d e x (AV_SI_50) and b a s a l a r e a / h a ( B A _ H A _ P T ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 17 6 D e p e n d e n t V a r i a b l e : R L H T _ Z N S o u r c e M o d e l E r r o r C o r r e c t e d T o t a l DF 5 1 7 0 1 7 5 R - S q u a r e 0 . 0 7 9 8 1 7 Sum o f S q u a r e s 0 . 2 4 2 1 2 6 8 6 2 . 7 9 1 3 8 4 0 1 3 . 0 3 3 5 1 0 8 7 C . V . 13 . 1 1 2 3 0 F V a l u e 2 . 95 P r > F 0 . 0 1 4 0 R L H T _ Z N M e a n 0 . 9 7 7 2 5 1 7 1 S o u r c e P _ Z 0 N E A V _ S I _ 5 0 B A _ H A _ P T S o u r c e P _ Z 0 N E A V _ S I _ 5 0 BA HA PT DF T y p e I SS 3 0 . 0 6 7 9 5 3 3 1 1 0 . 0 9 5 2 2 0 0 8 1 0 . 0 7 8 9 5 3 4 6 DF T y p e I I I SS 3 0 . 0 6 7 9 5 3 3 1 1 0 . 1 7 3 7 0 4 8 2 1 0 . 0 7 8 9 5 3 4 6 F V a l u e 1 . 3 8 5 . 80 4 . 81 F V a l u e 1 . 3 8 10 . 58 4 . 81 P r > F 0 . 2 5 0 8 0 . 0 1 7 1 0 . 0 2 9 7 P r > F 0 . 2 5 0 8 0 . 0 0 1 4 0 . 0 2 9 7 L e a s t S q u a r e s M e a n s R L H T _ Z N T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T L S M E A N i / j 1 2 3 4 0 9 4 7 6 8 8 5 2 1 - 0 . 8 8 2 6 5 - 1 . 5 8 6 5 5 - 1 . 8 5 9 3 0 . 3 7 8 7 0 . 1 1 4 5 0 . 0 6 4 7 0 9 7 1 8 0 2 0 4 2 0 . 8 8 2 6 4 6 - 0 . 7 0 3 9 - 0 . 9 7 6 6 6 0 . 3 7 8 7 0 . 4 8 2 5 0 . 3 3 0 1 0 9 9 1 0 3 2 4 2 3 1 . 5 8 6 5 5 1 0 . 7 0 3 9 0 4 - 0 . 2 7 2 7 5 0 . 1 1 4 5 0 . 4 8 2 5 0 . 7 8 5 4 0 9 9 8 4 8 3 8 7 4 1 . 8 5 9 3 0 2 0 . 9 7 6 6 5 6 0 . 2 7 2 7 5 1 0 . 0 6 4 7 0 . 3 3 0 1 0 . 7 8 5 4 P _ Z 0 N E 1 2 3 4 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 128 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 0 . 4 0 . 2 0 . 0 - 0 . 2 - 0 . 4 4 3 2 2 31 1 1 3 4 2 4 1 2 3 1 2 1 3 2 4 4 24 1 4 2 2 21 21 3 12 4 3 23 3 1 1 1 1 1 2 3 1 4 21 2 4 3 4 2 4 4 2 4 4 3 4 12 3 2 4 1 3 42 4 1 1 4 4 2 3 4 2 3 3 2 4 3 3 4 4 3 4 4 4 32 21 3 3 12 23 2 1 3 43 3 SfTffffffffffff fffffffffffffffffffffff ffffffffff" ff 0 . 8 5 0 . 9 0 0 . 9 5 1 . 0 0 1 . 0 5 YHAT N O T E : 23 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D M o m e n t s N 176 Sum W g t s 176 M e a n 0 Sum 0 S t d D e v 0 1 2 6 2 9 6 V a r i a n c e 0 . 0 1 5 9 5 1 S k e w n e s s 0 2 0 0 7 7 2 K u r t o s i s 0 . 3 9 8 2 9 U S S 2 7 9 1 3 8 4 C S S 2 . 7 9 1 3 8 4 CV S t d M e a n 0 . 0 0 9 5 2 T : M e a n = 0 0 P r > | T | 1 . 0 0 0 0 Num A = 0 176 Num > 0 85 M ( S i g n ) - 3 P r > = M | 0 . 7 0 6 4 S g n R a n k - 1 2 0 Pr>= s| 0 . 8 5 9 9 W : N o r m a l 0 9 8 1 8 9 7 Pr<W 0 . 4 8 6 2 129 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D S t e m L e a f # 3 67 2 3 34 2 2 6 1 2 1 3 4 3 1 5 5 5 6 6 6 6 7 7 7 8 8 8 9 9 15 1 0 0 0 0 0 0 1 1 1 2 2 2 3 3 4 4 16 0 6 6 6 6 7 7 7 7 7 8 8 8 8 9 9 9 16 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 3 3 3 4 4 4 4 4 4 4 30 - 0 4 4 4 4 4 4 3 3 3 3 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 32 - 0 9 9 9 9 9 8 7 7 7 7 7 7 7 6 6 6 5 5 5 5 5 21 - 1 4 4 4 3 3 3 3 2 2 2 2 2 1 0 0 0 0 17 - 1 9 8 7 7 6 5 5 5 8 - 2 4 3 2 2 0 0 0 0 8 - 2 7 6 5 5 4 - 3 4 1 B o x p l o t 0 0 + + I + I + + M u l t i p l y S t e m . L e a f b y 1 0 * * - 1 V a r i a b l e = R E S I D U n i v a r i a t e P r o c e d u r e 0 . 3 7 5 + 0 . 0 2 5 + N o r m a l P r o b a b i l i t y P l o t * * * * + + t++ + + * * * * * * * * * * * * * * * * * * * * * * * * * * * j ****_!_ * * * * * * * * * * + * + * * - 0 . 3 2 5 + * + - 2 - 1 ._ + + + + + 1 +2 130 O u t p u t 8. Z o n a l mean l i v e crown r a t i o r e l a t i v e t o Zone 5 (RLLCR_ZN) v e r s u s zone (P_Z0NE) and b a s a l a r e a / h a ( B A _ H A _ P T ) . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 17 6 D e p e n d e n t V a r i a b l e : L N R E L L C R S o u r c e M o d e l E r r o r C o r r e c t e d T o t a l DF 4 1 7 1 1 7 5 R - S g u a r e 0 . 1 0 7 7 6 4 Sum o f S q u a r e s 1 . 2 3 9 7 0 5 3 6 1 0 . 2 6 4 1 9 2 3 8 1 1 . 5 0 3 8 9 7 7 4 C . V . 397 . 7 8 8 7 F V a l u e P r > . F 5 . 1 6 0 . 0 0 0 6 L N R E L L C R M e a n 0 . 0 6 1 5 9 0 2 4 S o u r c e P _ Z 0 N E B A _ H A _ P T S o u r c e P _ Z 0 N E BA HA PT DF T y p e I SS 3 0 . 8 4 0 8 0 5 7 2 1 0 . 3 9 8 8 9 9 6 4 DF T y p e I I I SS 3 0 . 8 4 0 8 0 5 7 2 1 0 . 3 9 8 8 9 9 6 4 F V a l u e 4 . 6 7 6 . 65 F V a l u e 4 . 67 6 . 65 P r > F 0 . 0 0 3 7 0 . 0 1 0 8 P r > F 0 . 0 0 3 7 0 . 0 1 0 8 P _ Z 0 N E 1 2 3 4 L N R E L L C R L S M E A N 0 . 1 6 1 4 4 9 2 1 0 . 0 8 5 9 8 3 1 8 - 0 . 0 2 0 8 2 2 7 0 0 . 0 1 9 7 5 1 2 7 L e a s t S q u a r e s M e a n s T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 - 1 . 4 4 4 7 7 0 . 1 5 0 4 -3 . 4 8 9 5 3 0 . 0 0 0 6 1 . 4 4 4 7 6 9 3 . 4 8 9 5 2 8 2 . 7 1 2 7 5 5 0 . 1 5 0 4 0 . 0 0 0 6 0 . 0 0 7 4 2 . 0 4 4 7 5 9 1 . 2 6 7 9 8 6 0 . 0 4 2 4 2 . 0 4 4 7 6 0 . 0 4 2 4 2 . 7 1 2 7 5 - 1 . 2 6 7 9 9 0 . 7 7 6 7 7 4 0 . 0 0 7 4 0 . 2 0 6 5 0 . 4 3 8 4 0 . 2 0 6 5 - 0 . 7 7 6 7 7 0 . 4 3 8 4 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 131 P l o t o f R E S I D * Y H A T . S y m b o l i s v a l u e o f P _ Z O N E . R E S I D 0 . 7 5 3 2 2 3 1 4 1 33 4 1 2 1 3 33 3 4 3 2 1 1 3 4 4 2 4 14 1. 1 1 4 3 3 2 4 2 4 2 1 1 4 3 4 2 2 1 2 1 1 2 1 3 3 3 4 2 2 1 1 3 3 443 24 4 2 4 4 1 2 2 1 1 1 3 3 3 3 4 3 4 4 4 4 3 3 4 3 2 2 4 1 12 1 1 3 4 4 4 32 3 4 3 2 3 4 2 2 1 11 3 3 1 4 4 2 2 1 1 1 4 3 2 2 2 1 3 3 2 2 4 1 2 2 2 2 3 4 1 1 1 2 4 1 4 1 - 0 . 7 5 £rffffffffffrffffffffffrfffffffffff"fffffffffff"ffffffffffrf - 0 . 2 - 0 . 1 0 . 0 0 . 1 0 . 2 0 . 3 YHAT N O T E : 2 0 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D M o m e n t s N 176 Sum W g t s 1 7 6 M e a n 0 Sum 0 S t d D e v 0 . 2 4 2 1 8 3 V a r i a n c e 0 . 0 5 8 6 5 3 S k e w n e s s - 0 . 0 1 1 3 1 K u r t o s i s 0 . 1 0 8 4 0 8 U S S 10 . 2 6 4 1 9 C S S 10 . 2 6 4 1 9 C V S t d M e a n 0 . 0 1 8 2 5 5 T : M e a n = 0 0 P r > | T | 1 . 0 0 0 0 Num A = 0 176 Num > 0 82 M ( S i g n ) - 6 P r > = | M | 0 . 4 0 7 1 S g n R a n k - 1 3 6 P r > = | S | 0 . 8 4 1 4 W : N o r m a l 0 . 9 8 2 3 3 2 Pr<W 0 . 5 1 6 5 132 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D Stem L e a f # 6 2 1 5 7 1 5 04 2 4 5689 4 4 3 567888 6 3 00122344 8 2 56677799 8 2 00123444 8 1 577777899 9 1 00122223444 11 0 566677899 9 0 011111111222234 15 -0 4444433333322211111 19 -0 988877777666665 15 -1 44433332211110000 17 -1 9999887766655 13 -2 4311 4 -2 988755 6 -3 32211110 8 -3 55 2 -4 3200 4 -4 99 2 -5 4 1 -5 5 1 -6 -6 86 2 B o x p l o t M u l t i p l y S t e m . L e a f by 10**-1 133 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 0 . 625 + -0 . 025 + N o r m a l P r o b a b i l i t y P l o t + + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + ' + + -0 . 675 + ** + + --2 - - + + + -+ 1 +2 134 APPENDIX III Graphs of tree basal area annual increment, relative to the 5 year pre-road periodic annual increment, by plot (polygon). 135 TrQQ Basal Area Relates Average Annual Increment by Year For Lodgepole Pine Por/gon= 92P042 343 Road Year= 1984 136 Tree Basal AreaRelatwe Average Annual Increment by Yea; Fot Lodgepole Pine Po!/gon= 92P061J40 Road Year =1979 137 Trea Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Por/gon= 92P061.336 Road Year =1980 Tree Basal AreaRelatoe Average Annual Increment by Year For Lodgepole Pine Polygon:92P061.2S7 Road Year =1982 u-M-© n-10->• as-10 • at E <v zo-o c 1S-w-CQ as-— T " US* S33 190 S92 1996 B95 Year Zone 138 Tree Basal AreaRelatwe Average Annual Increment by Yea; For Lodgepole Pine PoVgon= 92P061_77 Road Year= «48 139 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon= 93A002_24S Road Year=1982 140 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon^ 93A002_256 Road Year= 1932 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon=93Au02_2« Road Year=1982 &o-sis--< a - S.0-o (.«• a . i d -1S-IS} 3z v>-C ai E ai zo-o 1S-10-CO as-ao -1 1 1 1333 est use us: Year Zone " 6 141 Tree Basal Area Relatve Average Annual Increment by Year For Lodgepole Pine Por/gon=93A002_263 Road Year= 1987 142 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Po^ gon=93A002_470 Road Year=19S4 6.0-_ 6.8-o_ 6.0" m I 0.6 H 1950 1960 WO SB) 1990 2000 Year Zone 1 S Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Polygon= 93A002 321 Road Year=1981 60 -66-Q_ 60-O 16" 0_ <,o->. 26-== 20-tz Of 26" E ai 20" 16-10-CQ 0.6-00" i 1 1 1 1 1 1 1 r 19S0 1982 6EI «56 SSS *90 »92 (#1 1996 Sffi Year Zone 1 — 5 143 Tree Basal Area Relatwe Average Annual Increment by Year Far Lodgepole Pine Polygon^93A003_608 Road Year= 1994 S0-ss--< so-Q_ O i s -10-is • 1^ 10-c ai ZS" e a) 20-o c 1S--<c 10-CO as-ao -89S 1 Year - r -Zone Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Polygon: 93A003J389 R o a d Year: 19*9 6.0-S.S -T 1 1 1 1 1 r SW B8 «60 870 CEO M0 2000 Year 144 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Poygon= 93A003_9t Road Year=1982 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine P or/go n= 93A0 03_883 R oad Year: 1982 &0-SS-Cl_ so-o is-u-is-^: I d -c at as-e in -o c= 1S • -<c 10-CO as-a.o-—i r~ «S2 SSI «SS -r~ too Year Zona 145 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93A0HJ51 Road Year: 1382 Tree Basal AreaRelat've Average Annual Increment by Year For Lodgepole Pine Poygon: 93A003J17 Road Year: 1982 so-ss-so-<,s-10 -1S-10-1S" io-"1 10-aoi T 1 r-1SS2 ttSl ISO MO «S2 TO - r use Year Zone 146 Tree Basal AreaRelatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93A0ti_172 Road Year= TO2 Tree Basal AreaRelatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93A0fl_154 Road Year= m 6.0-S.S-SL0-is-u-2S-2 0 -2 S -2 0 -1«H 10-<xs-ao-r VS2 wo S92 Year Zone 147 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Poygon= 93A013.233 Road Year=1987 6.0-68-a - 60-o 18-Cl_ M -16 • 10 -ai Z.6-E at 20-o a 18--< 10- t CO 08-oo-i — i — i — i — i — i — i — i — i — i — i — r es7 ess «so mn M I W2 m wt «os m «o? ese Year Zone 1 5 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Polygon: 93A013_231 Road Year: fl82 60-s.s-80 " o 1.8-o_ 10-28-=J= 10-ai 28" E ai 20" o c 18--oC 10-CQ 06-oo-est —r— *90 Year I 892 esc Zone 1 5 148 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon= 93A013J3 Road Year: "663 Tree Basal AreaRelatr/e Average Annual Increment by Year For Lodgepole Pine Poygon: 93A013_585 Road Year: 1963 6.0-S.6-1970 BBC t90 2000 Year Zone 1 5 149 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Poygon= 93A014_288 Road Year=1977 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Poygon: 93A013.706 Road Year=1948 6.0-6.6-T 1 1 1 1 i r 1910 £60 1960 1970 ISO *90 2000 Year Zone 1 S 150 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Polygon= 93AO23.672 Road Year= 1963 Tree Basal Area Relati/e Average Annual Increment by Year For Lodgepole Pine Polygon: 93A021.89 Road Yea;: "982 Year Zone 1 5 151 Tree Basal AreaRelat'we Average Annual Increment by Year For Lodgepole Pine . Polygon^  938006 411 Road Year: «77 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Po^gon: 93A031.271 Road Year: -054 6.0-6.6-1960 1960 (70 *3) 1990 2000 Year Zone 1 5 152 Tree Basal Area Relative Average Annual Increment by Year For Lodgepole Pine Polygon: 93B006_417 Road Year=1977 153 Tree Basal AreaRelatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93B024.189 Road Year= 1988 S.0-s.s-so-o «.*-a _ 10-J S -so-C at E at z.o-o 1S-CO as-ao • i — i — i — i — i — i — i — i — i — r «ss vm «s« «91 m m ssi m w> ss? «SG Year Zone Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93B01*_83 Road Year: -077 s.o _ s.s o_ S.O .2 t.o ^ . 2 .H ~ !0 <" 2.S E a. 2.0 _ c IS CO o.s 0.0 2000 Year Zone 154 Tree Basal AreaRelatwe Average Annual Increment by Year For Lodgepole Pine Polygon^93B024.337 Road Year=1979 155 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon^  93B04CM06 Road Year=1954 Tree Basal A tea Relative Average Annual Increment by Year For Lodgepole Pine Polygon:93B025_626 R o ad Year: 1977 6.0 -6.6-1970 SSO fOO 2000 Year Zone 1 S 156 Tree Basal AreaRetoe Average Annual Increment by Year For Lodgepole Pine Pol/gon= 93B0S0_73 Road Year= "fiSS 6.0-1 8.8 A 1960 1960 (70 1930 1990 2000 Year Zone 1 5 Tree Basal Area Relatwe Average Annual Increment by Year For Lodgepole Pine Polygon: 93B0S0S4 Road Year: -fiSO 6.0-S.S -«60 W0 VS) 1990 2000 Year 157 APPENDIX IV General Linear Models to test for differences in relative tree basal area increment among zones, while controlling for density and road age changes. 158 O u t p u t 1. G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 1 t o 2 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI2) w i t h zone ( P _ Z 0 N E ) , b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t D e p e n d e n t V a r i a b l e : L N R B A I 2 S o u r c e DF M o d e l 6 E r r o r 600 C o r r e c t e d T o t a l 6 0 6 R - S g u a r e 0 . 0 4 7 7 3 4 Sum o f S q u a r e s 2 . 4 8 2 6 6 1 6 4 49 . 5 2 8 0 0 4 5 8 52 . 0 1 0 6 6 6 2 2 C . V . - 1 8 0 4 . 5 7 7 607 F V a l u e P r > F 5 . 0 1 0 . 0 0 0 1 L N R B A I 2 M e a n - 0 . 0 1 5 9 2 1 1 5 S o u r c e DF T y p e I SS F V a l u e P r > F P _ Z 0 N E B A _ H A _ P T C _ R D _ A G E S o u r c e P _ Z 0 N E B A _ H A _ P T C RD A G E 1 . 0 3 8 1 2 1 2 0 0 . 4 6 9 1 1 3 1 9 0 . 9 7 5 4 2 7 2 6 DF T y p e I I I SS 4 1 . 0 5 4 4 9 0 1 0 1 0 . 9 7 1 4 0 9 1 4 1 0 . 9 7 5 4 2 7 2 6 3 . 1 4 5 . 68 1 1 . 82 F V a l u e 3 . 1 9 1 1 . 7 7 1 1 . 82 0 . 0 1 4 2 0 . 0 1 7 4 0 . 0 0 0 6 P r > F 0 . 0 1 3 1 0 . 0 0 0 6 0 . 0 0 0 6 L e a s t S q u a r e s M e a n s P _ Z 0 N E 1 2 3 4 5 L N R B A I 2 L S M E A N 0 . 0 4 3 5 9 6 3 2 0 . 0 0 6 1 5 6 6 6 - 0 . 0 7 5 7 4 0 5 8 -0 . 0 4 7 6 4 7 7 5 -0 . 0 0 4 9 4 7 8 5 L S M E A N N u m b e r 1 2 3 4 5 T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 - 1 . 0 1 3 4 3 0 . 3 1 1 3 -3 . 2 2 3 7 1 0 . 0 0 1 3 -2 . 4 5 9 8 5 0 . 0 1 4 2 - 1 . 3 1 1 3 5 0 . 1 9 0 2 1 . 0 1 3 4 2 8 0 . 3 1 1 3 - 2 . 2 3 0 8 4 0 . 0 2 6 1 - 1 . 4 6 2 5 6 0 . 1 4 4 1 - 0 . 3 0 2 4 8 0 . 7 6 2 4 3 . 2 2 3 7 0 5 0 . 0 0 1 3 2 . 2 3 0 8 3 5 0 . 0 2 6 1 0 . 7 6 2 0 9 6 0 . 4 4 6 3 1 . 9 2 4 4 3 7 0 . 0 5 4 8 2 . 4 5 9 8 4 5 0 . 0 1 4 2 1 . 4 6 2 5 6 0 . 1 4 4 1 - 0 . 7 6 2 1 0 . 4 4 6 3 1.158353 0 . 2 4 7 2 1 . 3 1 1 3 4 7 0 . 1 9 0 2 0 . 3 0 2 4 8 1 0 . 7 6 2 4 - 1 . 9 2 4 4 4 0 . 0 5 4 8 - 1 . 1 5 8 3 5 0 . 2 4 7 2 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 159 P l o t o f R E S I D 2 * P R E D 2 . S y m b o l u s e d i s ' * ' R E S I D 2 1 . 0 0 . 5 0 . 0 - 0 . 5 - 1 . 0 - 1 . 5 * * * * * * * * * * * * * * * * * * * i ** * * ** * * * * * * ** ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** •> * * * * * * * * * * * * * * * * * * * ** * ** * * ** * ** s/f - ffffffffffriffffffffff ~ fffffffff/rfffffffffff ~ // - 0 . 2 - 0 .1 0 . 0 PRED2 0 . 1 0 . 2 N O T E : 3 4 4 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 M o m e n t s N 607 Sum W g t s 607 M e a n 0 Sum 0 S t d D e v 0 . 2 8 5 8 8 4 V a r i a n c e 0 0 8 1 7 2 9 S k e w n e s s -0 . 7 7 6 5 3 K u r t o s i s 2 4 1 7 7 1 4 U S S 4 9 . 5 2 8 C S S 4 9 . 5 2 8 C V S t d M e a n 0 0 1 1 6 0 4 T : M e a n = 0 0 P r>1T1 1 . 0 0 0 0 Num ~= 0 607 Num > 0 3 1 6 M ( S i g n ) 12 . 5 P r > = M 0 . 3 3 0 0 S g n R a n k 5 1 2 6 P r > = S 0 . 2 3 6 0 W : N o r m a l • 0 . 9 6 6 1 9 7 Pr<W 0 . 0 0 0 1 160 V a r i a b l e = R E S I D 2 U n i v a r i a t e P r o c e d u r e H i s t o g r a m 0.85+* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * # i i 4 11 21 27 77 81 93 101 62 57 27 11 13 8 4 4 B o x p l o t 0 0 0 + + * | * - 1 . 4 5 + ' may r e p r e s e n t up t o 3 c o u n t s V a r i a b l e = R E S I D 2 N o r m a l P r o b a b i l i t y P l o t 0 . 85 + _ * * * * * _ * * * * * _i_ * * * * * * * * * * * * * * * * * * * * * * * * ****_{. * * * * * _|_ * * * * + + * * + -1.45+* + - - - + --2 - - + -+ 1 - - + -+ 2 161 O u t p u t 2 . G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 3 t o 5 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI3) w i t h zone ( P _ Z O N E ) , b a s a l a r e a p e r h a (BA_HT^PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t = 607 D e p e n d e n t V a r i a b l e : L N R B A I 3 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 6 1 0 . 9 2 1 2 6 2 4 4 1 4 . 4 6 0 . 0 0 0 1 E r r o r 600 75 . 5 0 5 5 7 5 2 3 C o r r e c t e d T o t a l 606 8 6 . 4 2 6 8 3 7 6 7 R- S q u a r e C V . L N R B A I 3 M e a n 0 . 1 2 6 3 6 4 - 6 4 7 . 0 2 0 2 - 0 . 0 5 4 8 2 7 1 9 S o u r c e DF T y p e I SS F V a l u e P r > F P _ Z 0 N E 4 4 . 7 5 1 2 6 1 0 2 9 . 4 4 0 . 0 0 0 1 B A _ H A _ P T 1 2 . 3 3 6 7 6 3 3 1 18 . 57 0 . 0 0 0 1 C _ R D _ A G E 1 3 . 8 3 3 2 3 8 1 2 3 0 . 4 6 0 . 0 0 0 1 S o u r c e DF T y p e I I I SS F V a l u e P r > F P _ Z 0 N E 4 4 . 8 3 3 0 2 3 5 5 9 . 6 0 0 . 0 0 0 1 B A _ H A _ P T 1 4 . 4 6 9 2 0 3 3 9 3 5 . 5 1 0 . 0 0 0 1 C _ R D _ A G E 1 3 . 8 3 3 2 3 8 1 2 3 0 . 4 6 0 . 0 0 0 1 L e a s t S q u a r e s M e a n s P _ ZONE L N R B A I 3 L S M E A N L S M E A N N u m b e r 1 0 1 1 2 7 5 3 7 2 1 2 - 0 0 7 6 0 9 1 9 8 2 3 - 0 1 4 3 7 5 7 0 6 3 4 - 0 1 1 6 6 6 2 6 9 4 5 - 0 0 4 6 5 8 9 6 7 5 T f o r HO : L S M E A N ( i ) = L S M E A N ( j ) / P r > T | i / j 1 2 3 - 4 5 1 4 . 1 4 0 0 3 6 5 . 6 1 2 0 5 7 5 . 0 0 9 1 4 5 3 . 4 8 6 1 8 6 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 5 2 - 4 . 1 4 0 0 4 1 . 4 9 2 7 9 0 . 8 9 3 1 9 1 - 0 . 6 5 0 8 6 0 . 0 0 0 1 0 . 1 3 6 0 0 . 3 7 2 1 0 . 5 1 5 4 3 - 5 . 6 1 2 0 6 - 1 . 4 9 2 7 9 - 0 . 5 9 5 2 9 - 2 . 1 3 9 3 0 . 0 0 0 1 0 . 1 3 6 0 0 . 5 5 1 9 0 . 0 3 2 8 4 - 5 . 0 0 9 1 4 - 0 . 8 9 3 1 9 0 . 5 9 5 2 9 1 - 1 . 5 3 9 5 8 0 . 0 0 0 1 0 . 3 7 2 1 0 . 5 5 1 9 0 . 1 2 4 2 5 - 3 . 4 8 6 1 9 0 . 6 5 0 8 6 4 2 . 1 3 9 2 9 9 1 . 5 3 9 5 7 5 0 . 0 0 0 5 0 . 5 1 5 4 0 . 0 3 2 8 0 . 1 2 4 2 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 162 P l o t o f R E S I D 3 * P R E D 3 . S y m b o l u s e d i s ' * ' . R E S I D 3 2 - 1 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * $frfffffffffff~fffffffffff~ffffffffffrffffffffffrff - 0 . 4 - 0 . 2 0 . 0 0 . 2 0 . 4 PRED3 N O T E : 3 5 4 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 3 M o m e n t s N 607 Sum W g t s 607 M e a n 0 Sum 0 S t d D e v 0 . 3 5 2 9 8 3 V a r i a n c e 0 . 1 2 4 5 9 7 S k e w n e s s -0 . 2 2 6 2 K u r t o s i s 1 . 2 7 9 1 2 5 U S S 75 . 5 0 5 5 8 C S S 75 . 5 0 5 5 8 cv. S t d M e a n 0 . 0 1 4 3 2 7 T : M e a n = 0 0 P r > | T | 1 . 0 0 0 0 Num 0 607 Num > 0 3 1 4 M ( S i g n ) 10 . 5 Pr>= M| 0 . 4 1 6 9 S g n R a n k 2 7 4 6 P r > = s| 0 . 5 2 5 7 W : N o r m a l 0 . 9 8 4 7 3 3 Pr<W 0 . 4 0 2 6 163 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 3 1 . 3 + J H i s t o g r a m - 0 . 1 + * r * * r * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * c * * * * * * * * * c * * * * * - 1 . 5 + v - + + + + may r e p r e s e n t u p t o 4 c o u n t s U n i v a r i a t e P r o c e d u r e # 1 2 7 10 46 94 1 5 4 1 4 5 74 40 23 4 5 1 1 B o x p l o t 0 0 0 + + * i * + + V a r i a b l e = R E S I D 3 1 . 3 + - 0 . 1 + N o r m a l P r o b a b i l i t y P l o t * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * * * * * _^ * * * * * * + + +* * * * * - 1 . 5 + * + + + + + + + + + + + - 2 - 1 0 +1 +2 164 O u t p u t 3 . G e n e r a l l i n e a r model o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 6 t o 10 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI4) w i t h zone (P_ZONE) , b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . N u m b e r o f o b s e r v a t i o n s i n d a t a s e t 607 D e p e n d e n t V a r i a b l e : L N R B A I 4 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 6 1 4 . 6 0 5 7 4 2 0 6 13 . 63 0 . 0 0 0 1 E r r o r 600 1 0 7 . 1 3 4 3 1 1 1 1 C o r r e c t e d T o t a l 606 1 2 1 . 7 4 0 0 5 3 1 6 R-0 . S q u a r e 1 1 9 9 7 5 C V . - 8 3 7 . 5 3 3 7 L N R B A I 4 M e a n - 0 . 0 5 0 4 5 2 9 3 S o u r c e DF T y p e I SS F V a l u e P r > F P _ Z 0 N E B A _ H A _ P T C _ R D _ A G E 4 1 1 7 . 6 1 7 9 0 7 4 5 1 . 2 1 4 3 5 2 1 6 5 . 7 7 3 4 8 2 4 5 10 . 67 6 . 80 32 . 3 3 0 . 0 0 0 1 0 . 0 0 9 3 0 . 0 0 0 1 S o u r c e DF T y p e I I I S S F V a l u e P r > F P _ Z 0 N E B A _ H A _ P T C _ R D _ A G E 4 1 1 7 . 7 2 9 6 2 8 2 6 3 . 4 9 2 1 1 0 9 2 5 . 7 7 3 4 8 2 4 5 1 0 . 8 2 1 9 . 5 6 32 . 3 3 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 G e n e r a l L i n e a r M o d e l s P r o c e d u r e L e a s t S q u a r e s M e a n s P _ Z 0 N E 1 2 3 4 5 L N R B A I 4 L S M E A N 0 . 1 6 5 6 5 8 5 4 - 0 . 0 7 4 8 5 9 9 6 - 0 . 1 3 5 2 3 9 2 9 - 0 . 1 5 0 9 3 1 9 2 - 0 . 0 5 2 2 0 1 3 6 L S M E A N N u m b e r 1 2 3 4 5 T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 1 4 . 4 2 6 6 0 4 5 . 5 2 6 6 3 5 5 8 0 3 1 2 8 4 . 0 0 1 4 6 5 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 2 - 4 . 4 2 6 6 1 . 1 1 8 2 7 2 1 4 0 5 9 8 8 - 0 . 4 1 9 6 5 0 . 0 0 0 1 0 . 2 6 3 9 0 . 1 6 0 2 0 . 6 7 4 9 3 - 5 . 5 2 6 6 3 - 1 . 1 1 8 2 7 0 2 8 9 4 4 8 - 1 . 5 3 4 8 0 . 0 0 0 1 0 . 2 6 3 9 0 . 7 7 2 3 0 . 1 2 5 4 4 - 5 . 8 0 3 1 3 - 1 . 4 0 5 9 9 -0 . 2 8 9 4 5 - 1 . 8 2 1 0 7 0 . 0 0 0 1 0 . 1 6 0 2 0 . 7 7 2 3 0 . 0 6 9 1 5 - 4 . 0 0 1 4 7 0 . 4 1 9 6 5 5 1 . 5 3 4 8 0 3 1 8 2 1 0 7 1 0 . 0 0 0 1 0 . 6 7 4 9 0 . 1 2 5 4 0 . 0 6 9 1 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 165 P l o t o f R E S I D 4 * P R E D 4 . S y m b o l u s e d i s ' * ' R E S I D 4 2 • * * * * * * * * * * * ** * * * ** ** * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * ** ** * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * ** * * * * ** * * * * * * * * * * * * * * * * * * * * S"fffffffffff"jfjffffffff"fffffffffff"fffffffffff"ffjffffffffff - 0 . 4 - 0 . 2 0 . 0 0 . 2 0 . 4 0 . 6 PRED4 N O T E : 3 04 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 M o m e n t s N 607 Sum W g t s 607 M e a n 0 Sum 0 S t d D e v 0 4 2 0 4 6 3 V a r i a n c e 0 1 7 6 7 8 9 S k e w n e s s 0 0 8 6 5 8 8 K u r t o s i s 0 5 3 4 0 0 5 U S S 1 0 7 . 1 3 4 3 C S S 107 . 1 3 4 3 CV S t d M e a n 0 0 1 7 0 6 6 T : M e a n = 0 0 Pr>1T1 1 . 0 0 0 0 Num A = 0 607 Num > 0 3 1 2 M ( S i g n ) 8 . 5 P r > = | M 0 . 5 1 6 1 S g n R a n k - 1 3 2 Pr>= jS 0 . 9 7 5 7 W : N o r m a l 0 9 8 4 9 1 8 Pr<W 0 . 4 2 6 7 166 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 1.5 + -* r * * r * * * r * * * r * * * r * * * r * * * r * * * T * * * : * * -k ; * * * r * * H i s t o g r a m 0 . 1 + *- 1 . 3 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - + — + — + — + -may r e p r e s e n t u p t o 3 c o u n t s # 1 2 8 12 23 40 102 1 2 4 113 81 55 33 8 2 3 B o x p l o t 0 0 0 + + * | * U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 1 . 5 + 0 . 1 + N o r m a l P r o b a b i l i t y P l o t * * * * * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 3 + J - 1 + 1 +2 167 O u t p u t 4. G e n e r a l l i n e a r model o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 11 t o 15 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI5) w i t h zone (P_ZONE) , b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . N O T E : Due t o m i s s i n g v a l u e s , o n l y 592 o b s e r v a t i o n s c a n b e u s e d i n t h i s a n a l y s i s . D e p e n d e n t V a r i a b l e : L N R B A I 5 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 6 2 0 . 3 0 1 7 7 0 0 4 1 6 . 0 7 0 . 0 0 0 1 E r r o r C o r r e c t e d T o t a l 5 8 5 1 2 3 . 1 6 8 0 2 8 3 9 5 9 1 1 4 3 . 4 6 9 7 9 8 4 4 R - S q u a r e 0 . 1 4 1 5 0 6 C . V . - 5 7 4 . 4 5 1 2 L N R B A I 5 M e a n - 0 . 0 7 9 8 7 6 3 0 S o u r c e P _ Z 0 N E B A _ H A _ P T C RD A G E DF 4 1 1 T y p e I S S 7 . 5 9 0 1 0 6 3 4 0 . 3 5 1 3 8 8 2 1 12 . 3 6 0 2 7 5 4 9 F V a l u e 9 . 01 1 . 67 5 8 . 7 1 P r > F 0 . 0 0 0 1 0 . 1 9 6 9 0 . 0 0 0 1 S o u r c e P _ Z 0 N E B A _ H A _ P T C RD A G E DF T y p e I I I SS 4 7 . 7 2 5 4 2 6 2 3 1 2 . 9 3 4 9 0 6 7 3 1 1 2 . 3 6 0 2 7 5 4 9 F V a l u e 9 . 1 7 13 . 9 4 5 8 . 7 1 P r > F 0 . 0 0 0 1 0 . 0 0 0 2 0 . 0 0 0 1 L e a s t S q u a r e s M e a n s P _ Z O N E L N R B A I 5 L S M E A N L S M E A N N u m b e r 1 0 . 1 3 9 6 0 2 3 0 1 2 - 0 . 1 2 8 1 7 7 4 3 2 3 - 0 . 1 4 8 7 0 5 0 3 3 4 - 0 . 1 8 4 3 5 8 8 5 4 5 - 0 . 0 7 2 6 8 1 5 1 5 L e a s t S q u a r e s M e a n s T f o r H O : L S M E A N ( i ) = L S M E A N ( j ) / P r > i / j 1 2 3 4 5 1 4 . 4 8 1 9 0 8 4 . 8 1 5 4 6 9 5 . 3 9 9 7 8 3 . 5 4 5 6 8 2 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 0 0 4 2 - 4 . 4 8 1 9 1 0 . 3 4 5 8 0 5 0 . 9 4 4 4 0 9 - 0 . 9 3 4 8 8 0 . 0 0 0 1 0 . 7 2 9 6 0 . 3 4 5 4 0 . 3 5 0 2 3 - 4 . 8 1 5 4 7 - 0 . 3 4 5 8 0 . 5 9 8 0 9 4 - 1 . 2 7 8 0 1 0 . 0 0 0 1 0 . 7 2 9 6 0 . 5 5 0 0 0 . 2 0 1 8 4 - 5 . 3 9 9 7 8 -0 . 9 4 4 4 1 - 0 . 5 9 8 0 9 - 1 . 8 7 3 3 9 0 . 0 0 0 1 0 . 3 4 5 4 0 . 5 5 0 0 0 . 0 6 1 5 5 - 3 . 5 4 5 6 8 0 . 9 3 4 8 7 5 1 . 2 7 8 0 1 3 1 . 8 7 3 3 9 2 0 . 0 0 0 4 0 . 3 5 0 2 0 . 2 0 1 8 0 . 0 6 1 5 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 168 P l o t o f R E S I D 5 * P R E D 5 . S y m b o l u s e d i s R E S I D 5 2 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * i * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * S"fffffffffff"fffffffffff"fffffffffff"fffffffffff"fffffffffff~f - 0 . 4 - 0 . 2 0 . 0 0 . 2 0 . 4 0 . 6 PRED5 N O T E : 15 o b s h a d m i s s i n g v a l u e s . 2 8 1 o b s h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 5 M o m e n t s N 592 Sum W g t s 5 9 2 M e a n 0 Sum 0 S t d D e v 0 . 4 5 6 5 1 5 V a r i a n c e 0 . 2 0 8 4 0 6 S k e w n e s s 0 . 0 0 9 2 6 9 K u r t o s i s 0 . 7 0 8 3 5 3 U S S 1 2 3 . 1 6 8 C S S 123 . 1 6 8 C V S t d M e a n 0 . 0 1 8 7 6 3 T : M e a n = 0 0 P r > 1 T 1 . 0 0 0 0 Num 0 592 Num > 0 3 0 2 M ( S i g n ) 6 P r > = M 0 . 6 5 1 2 S g n R a n k 665 Pr>= S 0 . 8 7 3 3 W : N o r m a l 0 . 9 8 4 7 2 Pr<W 0 . 4 0 9 8 169 V a r i a b l e = R E S I D 5 H i s t o g r a m 1 . 5 + J 0 . 9 + 0 . 3 - 0 . 3 - 0 . 9 - 1 . 5 + ' * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * may r e p r e s e n t u p t o 3 c o u n t s # 2 4 5 13 24 42 97 1 1 5 1 1 0 70 55 31 13 6 4 1 B o x p l o t 0 0 0 V a r i a b l e = R E S I D 5 1 . 5 + 0 . 9 + 0 . 3 + - 0 . 3 + -0 . 9 + U n i v a r i a t e P r o c e d u r e N o r m a l P r o b a b i l i t y P l o t * * * * * * 4-4. * * * * 4. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 5 + * - 1 + 1 +2 170 O u t p u t 5. G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 16 t o 20 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI6) w i t h zone ( P _ Z O N E ) , b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , o n l y 517 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNRBAI6 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 6 25.89710278 15.55 0.0001 E r r o r 510 141.52283571 C o r r e c t e d T o t a l 516 167.41993849 R- Square c . v . LNRBAI6 Mean 0 . 154684 -221 . 6818 -0.23762833 S o u r c e DF Type I SS F V a l u e P r > F P_Z0NE 4 4.53548033 4.09 0 . 0029 BA_HA_PT 1 0.00619076 0 . 02 0.8813 C_RD_AGE 1 21.35543169 76 . 96 0.0001 S o u r c e DF Type I I I SS F V a l u e P r > F P_Z0NE 4 4.64496553 4 .18 0.0024 BA_HA_PT 1 2.20137336 7 . 93 0.0050 C_RD_AGE 1 21.35543169 76 . 96 0.0001 L e a s t S q u a r e s Means P_ ZONE LNRBAI6 LSMEAN LSMEAN Number 1 -0 06449998 1 2 -0 26382722 2 3 -0 31025731 3 4 -0 33272963 4 5 -0 21249593 5 T f o r HO : L S M E A N ( i ) = L S M E A N ( j ) / P r > T | i / j 1 2 3 4 5 1 2 .714899 3.339359 3 . 636073 2 . 010974 0.0069 0 . 0009 0.0003 0.0449 2 -2 .7149 0.6371 0 .943155 -0.70435 0 . 0069 0.5243 0.3460 0.4815 3 -3 .33936 -0.6371 0 .306876 -1.33826 0 . 0009 0.5243 0.7591 0.1814 4 -3 . 63607 -0.94316 -0.30688 - 1 . 64188 0.0003 0.3460 0.7591 0.1012 5 -2.01097 0.704353 1.338261 1 . 641881 0.0449 0.4815 0.1814 0.1012 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 171 P l o t o f R E S I D 6 * P R E D 6 . S y m b o l u s e d i s R E S I D 6 2 * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * i * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * i * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * ** ** ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * s " f f f f f f f f f " f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f - 0 . 8 - 0 . 6 - 0 . 4 - 0 . 2 0 . 0 0 . 2 0 . 4 N O T E : 90 o b s h a d m i s s i n g v a l u e s . PRED6 198 o b s h i d d e n . 172 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 6 1 . 7 + * H i s t o g r a m r * * c * * r * * r * * c * * r * * r * * F * * r * * r * * r * * * * * * * * * * * * * * -k * * * * * k * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 7 + * may r e p r e s e n t u p t o 2 c o u n t s U n i v a r i a t e P r o c e d u r e # 1 5 4 5 14 2 5 52 73 78 68 87 43 36 10 8 5. 2 . 1 B o x p l o t 0 0 0 + + I + I * * + + V a r i a b l e = R E S I D 6 N o r m a l P r o b a b i l i t y P l o t 1 . 7 + _i_ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 7 + ' + + + + + -- 2 - 1 - + + + -0 +1 - + + + +2 173 O u t p u t 6 . M u l t i p l e a n a l y s i s o f c o v a r i a n c e f o r r e l a t i v e t r e e b a s a l a r e a p . a . f o r f i v e p e r i o d s ( l o g a r i t h m i c t r a n s f o r m a t i o n s ) j o i n t l y v e r s u s zone (P_ZONE) , b a s a l a r e a p e r ha ( B A _ H A _ P T ) , and r o a d age ( C _ R D _ A G E ) . E = E r r o r SS&CP M a t r i x LNRBAI2 LNRBAI3 LNRBAI4 LNRBAI5 LNRBAI6 LNRBAI2 42 .2005552 34 .1685782 22 .5152052 23 . 6582382 22 .3125144 LNRBAI3 34 .1685782 63 .3914392 56 .8155009 46 .7357919 45 . 8793471 LNRBAI4 22 .5152052 56 . 8155009 92 .4007759 77 .3008616 69 .3948295 LNRBAI5 23 . 6582382 4 6 . 7 3 5 7 9 1 9 7 7 . 3 0 0 8 6 1 6 1 0 5 . 9 7 1 9 1 5 97 .4513121 LNRBAI6 22 .3125144 4 5 . 8 7 9 3 4 7 1 69 .3948295 97 .4513121 141 .522836 P a r t i a l C o r r e l a t i o n C o e f f i c i e n t s f rom t h e E r r o r SS&CP M a t r i x / P r o b > I r I DF = 510 LNRBAI2 LNRBAI2 1 .000000 0 . 0001 LNRBAI3 0 . 660621 0 .0001 LNRBAI4 0 .360561 0 . 0001 LNRBAI5 0 .353776 0 .0001 LNRBAI6 0 .288720 0 .0001 LNRBAI3 LNRBAI4 LNRBAI5 LNRBAI6 0 .660621 0 . 0001 0 .360561 0 .0001 0 .353776 0 .0001 0 .288720 0 .0001 1.000000 0 .0001 0 .742359 0 .0001 0 .570215 0 .0001 0 .484383 0 .0001 0 .742359 0 .0001 1 .000000 0 .0001 0 .781180 0 .0001 0 .606843 0 .0001 0 .570215 0 . 0 0 0 1 0 .781180 0 .0001 1 .000000 0 . 0 0 0 1 0 .795754 0 .0001 0 .484383 0 .0001 0 .606843 0 .0001 0 .795754 0 .0001 1 .000000 0 .0001 H = Type I I I SS&CP M a t r i x f o r P_ZONE LNRBAI2 LNRBAI3 LNRBAI4 LNRBAI5 LNRBAI6 LNRBAI2 0 . 83692318 1 .33095049 1 .56668304 1 .45971225 1 .37159088 LNRBAI3 1 .33095049 3 .29734905 4 .15202752 4 .14622154 3 .78891548 LNRBAI4 1 .56668304 4 .15202752 5 .43493349 5 .43074663 4 . 9989985 LNRBAI5 1 .45971225 4 . 1 4 6 2 2 1 5 4 5 .43074663 5 . 58084271 5 . 07288302 LNRBAI6 1 .37159088 3 .78891548 4 .9989985 5 . 07288302 4 .64496553 H = Manova T e s t C r i t e r i a and F A p p r o x i m a t i o n s f o r t h e H y p o t h e s i s o f no O v e r a l l P_ZONE E f f e c t Type I I I SS&CP M a t r i x f o r P_Z0NE E = E r r o r SS&CP M a t r i x S=4 M=0 N=252 S t a t i s t i c V a l u e Num DF Den DF P r > F W i l k s ' Lambda 0 .915026 2 .27835 P i l l a i ' s T r a c e 0 .08645931 2 . 2 4 9 H o t e l l i n g - L a w l e y T r a c e 0 .09124955 2 .30177 R o y ' s G r e a t e s t Roo t 0 .06973617 7 .09914 20 20 20 5 1 6 7 9 . 1 6 2036 2018 509 0 .0010 0 . 0012 0 .0009 0 . 0001 NOTE: F S t a t i s t i c f o r R o y ' s G r e a t e s t Roo t i s an u p p e r b o u n d . 174 APPENDIX V Stepwise backwards regression for predicting relative tree basal area periodic annual . increment for Zone 1 from plot level variables for the first 15 years after the road was built. 175 Output 1. R e g r e s s i o n o f r e l a t i v e t r e e b a s a l a r e a p e r i o d i c a n n u a l i n c r e m e n t ( l o g a r t h m i c t r a n s f o r m a t i o n ; LNRLPAI) as a f u n c t i o n o f p l o t l e v e l v a r i a b l e s and p e r i o d f o r t h e f i r s t 15 y e a r s a f t e r t h e r o a d was b u i l t . (PERIOD i s t h e number o f y e a r s s i n c e t h e r o a d was b u i l t ; BA_HA_PT i s t h e p l o t b a s a l a r e a p e r h a ; CRD_PT i s C u r t i s ' RD; HT_PL0T i s t h e p l o t a v e r a g e h e i g h t ; LCR_PT i s t h e p l o t a v e r a g e l i v e crown r a t i o ; MNBATRPT i s t h e p l o t mean b a s a l a r e a p e r t r e e ; AVG_RW i s t h e r o a d w i d t h ; C_RD_AGE i s t h e number o f y e a r s s i n c e t h e r o a d was b u i l t ; QDBH_PT i s t h e p l o t q u a d r a t i c mean d i a m e t e r ; A V _ S I _ 5 0 i s t h e s i t e i n d e x ; SPH_PT i s t h e p l o t stems p e r h a ; EDGE_ASP i s t h e s t a n d edge a s p e c t ) . FOR THE F I R S T 15 YEARS FROM ROAD ESTABLISHMENT (PERIODS 2-5) B a c k w a r d E l i m i n a t i o n P r o c e d u r e f o r D e p e n d e n t V a r i a b l e LNRLPAI S t e p 0 A l l V a r i a b l e s E n t e r e d R - s q u a r e = 0 .13545977 C ( p ) = 13 .00000000 DF Sum o f S q u a r e s Mean S q u a r e F P r o k » F R e g r e s s i o n 12 13 . 93192945 1. 16099412 6 01 0 0001 E r r o r 460 88 .91727778 0. 19329843 T o t a l 472 102 .84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum o f S q u a r e s F Prob>F INTERCEP 5 .09296406 1 . 13006145 3 . 92613680 20 31 0 0001 PERIOD 0 .00772201 0 . 00451171 0 . 56624715 2 93 0 0877 BA_HA_PT -0 .16377303 0 . 05292143 1. 85118536 9 58 0 0021 CRD_PT 0 .94415127 0 .28136029 2 . 17663638 11 26 0 0009 H T _ P L 0 T -0 .02798295 0 . 02208506 0 . 31032600 1 61 0 2058 L C R _ P T -0 .27682001 0 .41005305 0 . 08809337 0 46 0 5000 MNBATRPT 164 .91480109 32 .05758785 5 . 11547439 26 46 0 0001 AVG_RW -0 .00045587 0 .00327465 0 . 00374610 0 02 0 8893 C_RD_AGE 0 .00213807 0 .00277640 0. 11463233 0 59 0 4416 QDBH_PT -0 .45797728 0 .10366497 3 . 77269900 19 52 0 0001 A V _ S I _ 5 0 -0 .04000525 0 . 01206210 2 . 12626171 11 00 0 0010 SPH_PT -0 .00135851 0 .00037893 2 . 48453626 12 85 0 0004 EDGE_ASP -0 .00026515 0 .00020940 0 . 30992660 1 60 0 2061 Bounds on c o n d i t i o n number: 5 2 0 . 4 4 4 , 23655 .62 S t e p 1 V a r i a b l e AVG_RW Removed R - s q u a r e = 0 .13542334 C ( p ) = 11 .01937987 DF Sum o f S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 11 13 . 92818335 1. 26619849 6 56 0 0001 E r r o r 461 88. 92102388 0 . 19288725 T o t a l 472 102 . 84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum of S q u a r e s F Prob>F INTERCEP 5 .06671283 1. 11303150 3 . 99706485 20 72 0 0001 PERIOD 0 .00771974 0. 00450688 0. 56592201 2 93 0 0874 BA_HA_PT -0 .16341686 0. 05280330 1. 84746022 9 58 0 0021 CRD_PT 0 .93978737 0 . 27931104 2 . 18366757 11 32 0 0008 H T J ? L 0 T -0 .02847242 0 . 02178019 0 . 32963168 1 71 0 1918 L C R _ P T -0 .29143130 0 . 39597122 0 . 10448373 0 54 0 4621 MNBATRPT 164 .69411593 3 1 . 98429630 5 . 11429878 26 51 0 0001 C_RD_AGE 0 .00238233 0 . 00214940 0 . 23695712 1 23 0 2683 QDBH_PT -0 .45618330 0. 10275141 3 . 80195258 19 71 0 0001 A V _ S I _ 5 0 -0 .03971031 0. 01186194 2 . 16171748 11 21 0 0009 SPH_PT -0 .00134800 0. 00037094 2 . 54734394 13 21 0 0003 EDGE_ASP -0 .00026737 0. 00020857 0. 31698040 1 64 0 2005 Bounds on c o n d i t i o n number: 5 1 3 . 9 8 3 8 , 21427 .63 176 S t e p 2 V a r i a b l e L C R _ P T Removed R - s q u a r e = 0 .13440745 C ( p ) = 9 .55991052 DF Sum o f S q u a r e s Mean S q u a r e F P r o k » F R e g r e s s i o n 10 13 82369962 1. 38236996 7 17 0 0001 E r r o r 462 89 02550760 0. 19269590 T o t a l 472 102 84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum o f S q u a r e s F Prob>F INTERCEP 4 .86971300 1 07983153 3 . 91892733 20 34 0 0001 PERIOD 0 .00770887 0 00450462 0. 56433435 2 93 0 0877 BA_HA_PT -0 .15360942 0 05106888 1. 74339405 9 05 0 0028 CRD_PT 0 .89823711 0 27341046 2 . 07981269 10 79 0 0011 H T _ P L O T -0 .01868332 0 01723881 0 . 22634252 1 17 0 2790 MNBATRPT 159 .45961945 31 16805312 5 . 04376192 26 17 0 0001 C_RD_AGE 0 .00172209 0 00195229 0 . 14993323 0 78 0 3782 QDBH_PT -0 .45166656 0 10251709 3 . 74038061 19 41 0 0001 A V _ S I _ 5 0 -0 .04367139 0 01056546 3 . 29222460 17 09 0 0001 SPH_PT -0 .00130682 0 00036651 2 . 44981869 12 71 0 0004 EDGE_ASP -0 .00026621 0 00020846 0 . 31426679 1 63 0 2022 Bounds on c o n d i t i o n number: 4 9 2 . 9 8 5 9 , 18611 .4 S t e p 3 V a r i a b l e C_RD_AGE Removed R - s q u a r e = 0 .13294965 C ( p ) = 8 .33556728 DF Sum of S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 9 13 . 67376639 1. 51930738 7 89 0 0001 E r r o r 463 89 . 17544084 0 . 19260354 T o t a l 472 102 . 84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum of S q u a r e s F Prob>F INTERCEP 4 .87912683 1. 07951998 3 . 93447795 20 43 0 0001 PERIOD 0 .00773948 0. 00450341 0 . 56885987 2 95 0 0864 B A _ H A _ P T -0 . 14986287 0 . 05087975 1. 67094616 8 68 0 0034 CRD_PT 0 . 88131185 0 . 27267097 2 . 01208206 10 45 0 0013 HT_PLOT -0 .01631273 0 . 01702395 0 . 17684670 0 92 0 3385 MNBATRPT 158 .06787775 3 1 . 12063184 4. 96883651 25 80 0 0001 QDBH_PT -0 . 45087471 0 . 10248859 3 . 72756287 19 35 0 0001 A V _ S I _ 5 0 -0 . 04336676 0 . 01055729 3 . 24992609 16 87 0 0001 SPH_PT -0 . 00129657 0 . 00036624 2 . 41396520 12 53 0 0004 EDGE_ASP -0 . 00025500 0. 00020802 0. 28942722 1 50 0 2209 Bounds on c o n d i t i o n number: 4 9 0 . 5 5 7 9 , 16678 .56 S t e p 4 V a r i a b l e H T _ P L O T Removed R - s q u a r e = 0 .13123018 C ( p ) = 7 .25045676 DF Sum of S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 8 13 . 49691969 1. 68711496 8 76 0 0001 E r r o r 464 89 . 35228754 0. 19256959 T o t a l 472 102 . 84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum of S q u a r e s F Prob>F INTERCEP 4 .77263933 1. 07368998 3 . 80493399 19 76 0 0001 PERIOD 0 .00777898 0 . 00450282 0 . 57472861 2 98 0 0847 B A _ H A _ P T -0 .14699824 0 . 05078737 1. 61324593 8 38 0 0040 CRD_PT 0 . 84034804 0. 26927519 1. 87548417 9 74 0 0019 MNBATRPT 152 .58157173 30 . 58671863 4 . 79210238 24 89 0 0001 QDBH_PT - 0 . 4 4 6 2 6 0 9 0 0 . 10236640 3 . 65974235 19 00 0 0001 A V _ S I _ 5 0 -0 . 04212435 0. 01047644 3 . 11333896 16 17 0 0001 SPH_PT - 0 . 0 0 1 2 3 9 0 2 0. 00036125 2 . 26535016 11 76 0 0007 EDGE_ASP - 0 . 0 0 0 2 8 9 0 9 0. 00020494 0. 38320085 1 99 0 1590 Bounds on c o n d i t i o n number: 4 7 8 . 4 9 9 8 , 14470 .2 177 S t e p 5 V a r i a b l e EDGE_ASP Removed R - s q u a r e = 0 .12750433 C(p) = 7 .23288802 DF Sum of S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 7 13 . 11371884 1. 87338841 9 71 0 0001 E r r o r 465 89 . 73548839 0. 19297954 T o t a l 472 102 . 84920722 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum of S q u a r e s F Prob>F INTERCEP 4 .79064174 1. 07475633 3 . 83423424 19 87 0 0001 PERIOD 0 .00781713 0 . 00450753 0 . 58040106 3 01 0 0835 B A _ H A _ P T -0 .16984454 0 . 04818706 2 . 39747223 12 42 0 0005 CRD_PT 0 .96205523 0 . 25535046 2 . 73928984 14 19 0 0002 MNBATRPT 161 .04338962 30. 02465921 5. 55189074 28 77 0 0001 QDBH_PT -0 .46418166 0 . 10168311 4 . 02151390 20 84 0 0001 A V _ S I _ 5 0 -0 .04250191 0 . 01048416 3 . 17146985 16 43 0 0001 SPH_PT -0 .00135457 0 . 00035221 2 . 85430756 14 79 0 0001 Bounds on c o n d i t i o n number: 4 5 2 . 2 4 1 8 , 11837 .57 A l l v a r i a b l e s l e f t i n t h e m o d e l a r e s i g n i f i c a n t a t t h e 0 .1000 l e v e l . Summary o f B a c k w a r d E l i m i n a t i o n P r o c e d u r e f o r D e p e n d e n t V a r i a b l e LNRLPAI V a r i a b l e Number P a r t i a l M o d e l S t e p Removed In R**2 R**2 C ( p ) F Prob>F 1 AVG_RW 11 0 .0000 0 .1354 11 0194 0 0194 0 .8893 2 L C R _ P T 10 0 .0010 0 .1344 9 5599 0 5417 0 .4621 3 C_RD_AGE 9 0 .0015 0 .1329 8 3356 0 7781 0 .3782 4 H T _ P L 0 T 8 0 .0017 0 .1312 7 2505 0 9182 0 .3385 5 EDGE_ASP 7 0 .0037 0 .1275 7 2329 1 9899 0 .1590 178 P l o t o f R E S I D 2 * P R E D 2 . Symbol u s e d i s R e s i d u a 1 * * * * * * * * * i * * * * * * * * * 2 . 0 1.5 1 .0 0 .5 0 .0 -0 . 5 -1 .0 - 1 . 5 ~ &ffffffffffffrffffffffffffnffffffffffff"fffffffffffff"fffffffffffffff - 0 . 2 0 .0 0 .2 0 .4 0 .6 0 .8 P r e d i c t e d V a l u e o f LNRLPAI NOTE: 3 obs h a d m i s s i n g v a l u e s . 164 obs h i d d e n . U n i v a r i a t e P r o c e d u r e * * * * * * * * * • * * * * k ***** ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * t * * * * * * * * * * * * * * * * k * * * * * * * * * * * * * * * * * * * * * * * * * * i * * * * * * * * * * * i * * * * * * V a r i a b l e = R E S I D 2 R e s i d u a l Moments N 473 Sum Wgts 473 Mean 0 Sum 0 S t d Dev 0 .436025 V a r i a n c e 0. 190118 Skewness 0 .340039 K u r t o s i s 1. 123325 USS 8 9 .73549 CSS 89 .73549 CV S t d Mean 0. 020048 T:Mean=0 0 P r > | T | 1 .0000 Num A = 0 473 Num > 0 231 M ( S i g n ) -5 . 5 Pr> = M 0 .6457 Sgn Rank - 1 6 2 7 . 5 Pr>= S 0 .5848 W:Normal 0 .980601 Pr<W 0 .1088 179 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 R e s i d u a l H i s t o g r a m # B o x p l o t 1.7 + * 1 0 . * 2 0 . * * 3 0 . * * * 5 0 * * * * * * | * * * * * * * * j * * * * * * * * * * * * * * * * * 24 j * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * g4 _i ^ 0 ]^_|_************************************************ | .j. J * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * cjg * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 7 g _l )_ * * * * * * * * * * * * * * * * * * * 28 | * * * * * * * * * * * * * * 28 j . * * * 5 | 2 | . * 2 0 - 1 . 5 + * 1 0 * may r e p r e s e n t up t o 2 c o u n t s N o r m a l P r o b a b i l i t y P l o t 1.7+ * * * * * * * +++ * * * * _|__|_ * * * * _|_ * * * * * + * * * * * Q -|- * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * _|_ | _|_* * - 1 . 5 + * + + + + + + + + + + + -2 -1 0 +1 +2 180 APPENDIX VI General Linear Models to test for differences in relative tree basal area increment between core directions, among zones, while controlling for density and road age changes. 181 O u t p u t 1. G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 1 t o 2 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI2) w i t h zone (P_Z0NE) and i n c r e m e n t c o r e d i r e c t i o n ( C 0 R J D I R ) , w h i l e c o n t r o l l i n g f o r b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t and p a i r w i s e t e s t between c o r e d i r e c t i o n s ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , o n l y 1210 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNRBAI2 S o u r c e M o d e l DF Sum o f S q u a r e s F V a l u e P r > F 7 4 .97008860 5 .76 0 .0001 E r r o r C o r r e c t e d T o t a l 1202 1209 148 .13821854 153 .10830713 R - S q u a r e 0 .032461 C V . - 1 4 7 0 . 2 3 6 LNRBAI2 Mean -0 . 02387779 S o u r c e P_Z0NE C0R_DIR BA_HA_PT C_RD_AGE DF 4 1 1 1 Type I SS 1.77122542 0 .07419496 0 . 89955473 2 .22511350 F V a l u e 3 .59 0 . 60 7 . 3 0 18 . 05 P r > F 0 . 0064 0 .4380 0 .0070 0 . 0001 S o u r c e P_Z0NE C0R_DIR BA_HA_PT C_RD_AGE DF 4 1 1 1 Type I I I SS 1. 81186399 0 . 07539656 1. 97966047 2 .22511350 F V a l u e 3 . 68 0 . 61 16 . 06 18 . 05 P r > F 0 . 0055 0 .4343 0 . 0001 0 . 0001 L e a s t Squa re s Means P_Z0NE 1 2 3 4 5 LNRBAI2 LSMEAN 0 .02961930 -0 .00664491 -0 .07732091 -0 . 05834538 -0 . 00551059 LSMEAN Number 1 2 3 4 5 T f o r HO: LSMEAN( i )=LSMEAN( j ) / P r > i / j 1 - 1 . 1 3 3 6 6 0 .2572 -3 .33291 0 .0009 -2 .73884 0 .0063 - 1 . 0 9 4 8 7 0 .2738 1 .133661 0 .2572 -2 .22589 0 . 0262 - 1 . 6 2 6 5 7 0 .1041 0 . 035725 0 . 9715 3 .332908 0 . 0009 ,225888 0 .0262 .595176 0 .5518 .254716 0 .0243 2 .738837 0 .0063 1. 626573 0 .1041 - 0 . 5 9 5 1 8 0 .5518 657184 0 .0977 -0 -1 5 094869 0 .2738 . 03572 0 .9715 .25472 0 .0243 .65718 0 . 0977 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 182 L e a s t Squa re s Means LNRBAI2 LSMEAN -0 . 01574671 T / P r > | T | HO: L SMEAN1=LSMEAN2 0 .782158 0 .4343 RESID2 1 COR_DIR W X - 0 . 0 3 1 5 3 4 2 8 P l o t o f RESID2*PRED2. Symbol u s e d i s ' * ' * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Sff"ffffffffffffffffffffff"ffffffffffffffffffffff"// - 0 . 2 - 0 . 1 0 .0 0 .1 0 .2 PRED2 NOTE: 1 obs had m i s s i n g v a l u e s . 867 obs h i d d e n . U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 Moments N 1210 Sum Wgts 1210 Mean 0 Sum 0 S t d Dev 0 .350042 V a r i a n c e 0 .12253 Skewness - 0 . 7 2 7 8 K u r t o s i s 2 . 006042 USS 148 .1382 CSS 1 4 8 . 1 3 8 2 CV S t d Mean 0 .010063 T:Mean=0 0 Pr>1T1 1.0000 Num A= 0 1210 Num > 0 644 M ( S i g n ) 39 Pr>= M 0 . 0268 Sgn Rank 20645 .5 Pr>= S 0 .0895 W:Normal 0 .963454 Pr<W 0 .0001 183 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 H i s t o g r a m 0 . 9 + * r * * f * * ; * * f * * r * * r * * - * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 7 + ' - + + + # 10 22 91 2 1 1 3 1 0 2 6 6 1 7 4 67 28 12 9 6 3 1 B o x p l o t 0 - + - - * + may r e p r e s e n t u p t o 7 c o u n t s U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 N o r m a l P r o b a b i l i t y P l o t 0 . 9 + _ * * * * * * . * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *******_!_ *****_)_ . * * * * * * * * • 1 . 7 + * + + + + + + + + + + + - 2 - 1 0 +1 +2 184 O u t p u t 2 . G e n e r a l l i n e a r model o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 3 t o 5 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI3) w i t h zone (P_Z0NE) and i n c r e m e n t c o r e d i r e c t i o n (COR_DIR) , w h i l e c o n t r o l l i n g f o r b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t and p a i r w i s e t e s t between c o r e d i r e c t i o n s ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , t h i s a n a l y s i s . o n l y 1210 o b s e r v a t i o n s c a n be u s e d i n Dependent V a r i a b l e : LNRBAI3 S o u r c e M o d e l DF Sum o f S q u a r e s F V a l u e P r > F 7 21 .76550146 1 7 . 2 8 0 .0001 E r r o r 1202 216 .30082175 C o r r e c t e d T o t a l 1209 238 . 06632321 R - S q u a r e 0 . 091426 C V . -735 .2080 LNRBAI3 Mean -0 . 05769878 S o u r c e P_Z0NE C0R_DIR BA_HA_PT C RD AGE DF 4 1 1 1 Type I SS 9 .17513849 0 . 04847019 4 .76556165 7 .77633113 F V a l u e 12 .75 0 .27 26 .48 43 .21 P r > F 0 .0001 0 . 6039 0 . 0001 0 . 0001 S o u r c e P_Z0NE C0R_DIR BA_HA_PT C_RD_AGE DF Type I I I SS F V a l u e P r > F 4 9 .38326523 1 3 . 0 4 0 .0001 1 0 .05021021 0 .28 0 .5974 1 9 .08903550 5 0 . 5 1 0 .0001 1 7 .77633113 4 3 . 2 1 0 .0001 L e a s t Squa re s Means P_Z0NE 1 2 3 4 5 LNRBAI3 LSMEAN 0 .10767492 - 0 . 0 8 0 0 9 4 9 4 -0 .14342646 - 0 . 1 2 1 6 2 4 1 0 -0 . 04624605 LSMEAN Number 1 2 3 4 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > |T i / j 1 -4 .85775 0 . 0001 -6 .47643 0 . 0001 -5 . 90833 0 . 0001 -3 .96997 0 . 0001 2 857753 0 .0001 - 1 . 6 5 0 6 5 0 .0991 - 1 . 08128 0 .2798 0 . 88223 0 .3778 6 .476433 0 . 0001 1.650653 0 .0991 . 565925 0 . 5716 , 525155 0 .0117 4 i . 908327 0 .0001 . . 081277 0 .2798 -0 . 56593 0 .5716 1 .956591 0 . 0506 3 . 969974 0 . 0 0 0 1 - 0 . 8 8 2 2 3 0 .3778 -2 . 52516 0 . 0117 - 1 . 95659 0 . 0506 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 185 L e a s t Squa re s Means COR_DIR LNRBAI3 T / P r > | T | HO: LSMEAN LSMEAN1=LSMEAN2 W - 0 . 0 5 0 3 0 1 5 5 0 .528225 0 .5974 X - 0 . 0 6 3 1 8 5 1 0 P l o t o f RESID3*PRED3. Symbol u s e d i s ' * ' . RESID3 2 * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * ** * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * ** •* * * * * * * * : * * * * * * §//" ffffffffffffffffffffff ~ ffffffffffr/ffffffffff" // - 0 . 4 - 0 . 2 0 .0 0 .2 0 .4 1 obs had m i s s i n g v a l u e s . PRED3 82 9 obs h i d d e n . 186 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 3 M o m e n t s N 1 2 1 0 Sum W g t s 1 2 1 0 M e a n 0 Sum 0 S t d D e v 0 . 4 2 2 9 7 6 V a r i a n c e 0 . 1 7 8 9 0 9 S k e w n e s s - 0 . 2 0 5 6 1 K u r t o s i s 1 . 1 8 9 2 0 1 U S S 2 1 6 . 3 0 0 8 C S S 2 1 6 . 3 0 0 8 C V S t d M e a n 0 . 0 1 2 1 6 T : M e a n = 0 0 Pr>1T1 1 . 0 0 0 0 Num / s = 0 1 2 1 0 Num > 0 6 3 5 M ( S i g n ) 30 P r > = M 0 . 0 8 9 8 S g n R a n k 1 3 2 3 7 . 5 Pr>= S 0 . 2 7 6 4 W : N o r m a l 0 . 9 7 9 8 2 7 Pr<W 0 . 0 0 0 1 V a r i a b l e = R E S I D 3 H i s t o g r a m 1 . 9 + * 0 .1 + : * * : * * * ; * * * r * * * r * * * r * * * r * * * r * * * t * * * t * * * t * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 7 + J # 1 1 2 1 10 18 32 1 0 9 194 2 6 7 2 4 5 1 3 9 98 45 19 17 10 1 1 B o x p l o t + + * -| * may r e p r e s e n t u p t o 6 c o u n t s 187 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 3 1.9 + 0 .1 + N o r m a l P r o b a b i l i t y P l o t _j_ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * * * * * -1.7+i -2 -1 0 +1 +2 188 O u t p u t 3 . G e n e r a l l i n e a r model o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 6 t o 10 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI4) w i t h zone (P_Z0NE) and i n c r e m e n t c o r e d i r e c t i o n (C0R_DIR) , w h i l e c o n t r o l l i n g f o r b a s a l a r e a p e r ha (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t and p a i r w i s e t e s t between c o r e d i r e c t i o n s ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . N O T E : D u e t o m i s s i n g v a l u e s , o n l y 1210 o b s e r v a t i o n s c a n b e u s e d i n t h i s a n a l y s i s . D e p e n d e n t V a r i a b l e : L N R B A I 4 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 7 28 .30224925 17.39 0 . 0001 E r r o r 1202 279.49974291 C o r r e c t e d T o t a l 1209 307 . 80199216 R- S q u a r e c . V . L N R B A I 4 M e a n 0 . 091950 -904.7193 -0 . 05329968 S o u r c e DF T y p e I S S F V a l u e P r > F P _ Z 0 N E 4 14.41395486 15 . 50 0 . 0001 C 0 R _ D I R 1 0.09494335 0.41 0.5230 B A _ H A _ P T 1 2.28224803 9 . 81 0 . 0018 C _ R D _ A G E 1 11.51110301 49.50 0.0001 S o u r c e DF T y p e I I I SS F V a l u e P r > F P _ Z 0 N E 4 14.70221332 15 . 81 0 . 0001 C 0 R _ D I R 1 0 . 09823603 0 .42 0.5158 B A _ H A _ P T 1 6 .72796363 28 . 93 0.0001 C _ R D _ A G E 1 11.51110301 49 . 50 0 . 0001 L e a s t S q u a r e s M e a n s P _ ZONE L N R B A I 4 L S M E A N L S M E A N N u m b e r 1 0 15667930 1 2 -0 07607516 2 3 -0 14372970 3 4 -0 14820003 4 5 -0 04923342 5 T f o r HO : L S M E A N ( i ) = L S M E A N ( j ) / P r > T | i / j 1 2 3 4 5 1 5.297195 6.816134 6. 910805 4.672092 0 . 0001 0.0001 0.0001 0 . 0001 2 -5 .2972 1. 551212 1. 651989 -0 . 61544 0 .0001 0 .1211 0.0988 0 . 5384 3 -6 . 81613 -1.55121 0 . 102078 -2.16004 0 . 0001 0.1211 0.9187 0.0310 4 -6 . 91081 - 1 . 65199 -0 .10208 -2.25986 0 . 0001 0.0988 0.9187 0.0240 5 - 4 . 67209 0.615441 2.160043 2 . 259864 0 . 0001 0.5384 0.0310 0.0240 N O T E : T o e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d b e u s e d . 189 L e a s t S q u a r e s M e a n s C O R _ D I R L N R B A I 4 T / P r > | T | H O : L S M E A N L S M E A N 1 = L S M E A N 2 W - 0 . 0 4 3 1 0 1 3 8 0 . 6 4 9 9 7 5 0 . 5 1 5 8 X - 0 . 0 6 1 1 2 2 2 2 P l o t o f R E S I D 4 * P R E D 4 . S y m b o l u s e d i s R E S I D 4 2 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * ** ** * s " f f f f f f f f f f f " f f f f f f f f f f f " f f f f f f f f f f f " f f f f f f f f f f f f f f f f f f f f f f f f - 0 . 4 - 0 . 2 0 .0 0 .2 0 .4 0 .6 PRED4 N O T E : 1 o b s h a d m i s s i n g v a l u e s . 7 5 9 o b s h i d d e n . 190 V a r i a b l e = R E S I D 4 U n i v a r i a t e P r o c e d u r e Moments N Mean S t d Dev Skewness USS CV T:Mean=0 Num A = 0 M ( S i g n ) Sgn Rank W:Normal 1210 0 0 .480814 0 .032586 279 .4997 0 1210 17 1586 . 5 0 .988775 U n i v a r i a t e Sum Wgts Sum V a r i a n c e K u r t o s i s CSS S t d Mean P r > | T | Num > 0 Pr>= | M Pr>= | S Pr<W P r o c e d u r e 1210 0 0 .231183 0 .546321 2 7 9 . 4 9 9 7 0 .013822 1 .0000 622 0 .3428 0 .8962 0 .7995 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 2 .1 + J H i s t o g r a m : * * r * * : * * ; * * ; * * r * * r * * r * * t * * r * * t * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 7 + ' # 1 1 1 9 15 27 62 112 172 222 187 166 113 59 37 14 10 1 1 B o x p l o t 0 0 0 0 + + * | * may r e p r e s e n t up t o 5 c o u n t s 191 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 N o r m a l P r o b a b i l i t y P l o t 2 .1 + * * * * * * * * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * * * * * * * * - 1 . 7 + * + + + + + -2 -1 0 + 1 + 2 192 O u t p u t 4 . G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 11 t o 15 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI5) w i t h zone (P_Z0NE) and i n c r e m e n t c o r e d i r e c t i o n (C0R_DIR) , w h i l e c o n t r o l l i n g f o r b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t and p a i r w i s e t e s t between c o r e d i r e c t i o n s ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , o n l y 1180 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNRBAI5 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 7 39 . 68983634 2 1 . 1 3 0 ..0001 E r r o r 1172 314 .56193351 C o r r e c t e d T o t a l 1179 354 .25176985 R- Squa re C V . LNRBAI5 Mean 0 . 112038 - 6 1 8 . 8 1 1 8 -0 . 08372028 S o u r c e DF Type I SS F V a l u e P r > F P_Z0NE 4 14 .55529370 13 .56 0 . 0001 C0R_DIR 1 0 .03157154 0 .12 0 .7317 BA_HA_PT 1 0 . 63543950 2 .37 0 .1242 C_RD_AGE 1 24 .46753159 9 1 . 1 6 0 . 0001 S o u r c e DF Type I I I SS F V a l u e P r > F P_Z0NE 4 14 .91972463 13 . 90 0 .0001 C0R_DIR 1 0 . 03470166 0 .13 0 .7192 BA_HA_PT 1 5 .62318901 20 . 95 0 .0001 C_RD_AGE 1 24 .46753159 9 1 . 1 6 0 . 0001 L e a s t S q u a r e s Means P_Z0NE LNRBAI5 LSMEAN LSMEAN Number 1 0. 13152485 1 2 - 0 . 13084297 2 3 -0 . 15473281 3 4 - 0 . 18666225 4 5 -0 . 07136863 5 T f o r HO L S M E A N ( i )=LSMEAN(j) / P r > T | i / j 1 2 3 4 5 1 5 .488184 5 .969246 6 . 628411 4 . 2 3 0 9 0 8 0 . 0001 0 .0001 0 .0001 0 . 0001 2 -5 . 48818 0 .503548 1 .175299 - 1 . 2 5 3 6 0 . 0001 0 .6147 0 .2401 0 .2102 3 -5 . 96925 - 0 . 5 0 3 5 5 0 .670184 - 1 . 7 5 1 6 5 0 . 0001 0 .6147 0 .5029 0 .0801 4 - 6 . 62841 - 1 . 1 7 5 3 - 0 . 6 7 0 1 8 -2 .41996 0 . 0001 0 .2401 0 .5029 0 .0157 5 -4 . 23091 1.253599 1.75165 2 .419955 0 . 0001 0 .2102 0 .0801 0 .0157 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 193 L e a s t Squa re s Means COR_DIR LNRBAI5 T / P r > | T | HO: LSMEAN LSMEAN1=LSMEAN2 W - 0 . 0 7 6 9 9 3 4 0 0 .359572 0 .7192 X - 0 . 0 8 7 8 3 9 3 2 P l o t o f R E S I D 5 * P R E D 5 . Symbol u s e d i s R E S I D 5 4 * * * * * * * * ** * * * * * ** * * * * ** * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * ** * ** ** ** ** * * * * * 1 * * * * * § ~fffffffff"fffffffff"fffffffff"fffffffff"ffffffffffffffffff"/ -0 .6 - 0 . 4 - 0 . 2 0 .0 0 .2 0 .4 0 .6 PRED5 NOTE-: 31 obs had m i s s i n g v a l u e s . 836 obs h i d d e n . 194 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 5 Moments N 1180 Sum Wgts 1180 Mean 0 Sum 0 S t d Dev 0 . 516531 V a r i a n c e 0 .266804 Skewness - 0 . 0 3 6 6 2 K u r t o s i s 0 . 635726 USS 314 .5619 CSS 3 1 4 . 5 6 1 9 CV S t d Mean 0 . 015037 T:Mean=0 0 Pr> 1 T 1 .0000 Num ~= 0 1180 Num > 0 611 M ( S i g n ) 21 Pr> = M 0 .2326 Sgn Rank 4589 Pr> = S 0 .6953 W:Normal 0 .988019 Pr<W 0 .6866 E x t r e m e s L o w e s t Obs H i g h e s t Obs -1 90819( 93) 1. 574995( 852) -1 80768( 523) 1. 587977( 975) -1 51023 ( 667) 1. 621801( 1073) -1 44025( 256) 1. 696446( 801) -1 41614( 211) 2 .07273 ( 709) U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 5 H i s t o g r a m # B o x p l o t 2.1+* 1 0 1.7+* 2 0 5 0 1.3+*** 11 0 . * * * 15 0.9+****** 30 * * * * * * * * * * * 5^ 0.5+************************ 120 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 168 H i-0.1 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 208 * i * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 185 | J —0 3+**************************** 137 H H * * * * * * * * * * * * * * * * * * * * * 103 — 0 7 + * * * * * * * * * * * * * 5 5 * * * * * * * * * ^2 -1.1+**** 20 . * * * 11 0 -1.5+* 4 0 -1.9+* 2 0 * may r e p r e s e n t up t o 5 c o u n t s 195 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 5 2 .1 + 1 . 7 + I 1 . 3 + I 0 . 9 + 0 . 5 + - 0 . 7 + N o r m a l P r o b a b i l i t y P l o t * * * * * * * + 4. * * * * _ ( _ * * * * * * * * * * * * * * * Q 2 _ - f * * * * * I ****** - 0 . 3 + * * * * + I * * * * * * * * * * * * * * - 1 . 1 + + * * * * j * * * * - 1 . 5 + * I - 1 . 9 + * + + + + + + + + + + -- 2 - 1 0 +1 +2 196 O u t p u t 5 . G e n e r a l l i n e a r mode l o f r e l a t i v e t r e e b a s a l a r e a p . a . i . f o r 16 t o 20 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNRBAI6) w i t h zone (P_Z0NE) and i n c r e m e n t c o r e d i r e c t i o n (C0R_DIR) , w h i l e c o n t r o l l i n g f o r b a s a l a r e a p e r h a (BA_HT_PT) and r o a d age ( C _ R D _ A G E ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t and p a i r w i s e t e s t between c o r e d i r e c t i o n s ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , o n l y 103 0 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNRBAI6 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 7 50 . 73559099 2 1 . 2 3 0 . 0001 E r r o r 1022 348 .87795194 C o r r e c t e d T o t a l 1029 399 .61354293 R- Squa re C V . LNRBAI6 Mean 0 . 126962 - 2 3 8 . 7 2 0 5 - 0 . 2 4 4 7 4 9 3 3 S o u r c e DF Type I SS F V a l u e P r > F P_Z0NE 4 8 .26585924 6 . 05 0 .0001 C0R_DIR 1 0 . 00322620 0 . 01 0 . 9226 BA_HA_PT 1 0 . 00265276 0 . 01 0 .9298 C_RD_AGE 1 42 .46385279 1 2 4 . 3 9 0 . 0001 S o u r c e DF Type I I I SS F V a l u e P r > F P_Z0NE 4 8 .55793563 6 .27 0 . 0001 C0R_DIR 1 0 .00187741 0 . 01 0 .9409 BA_HA_PT 1 3 .72723098 10 . 92 0 .0010 C_RD_AGE 1 42 .46385279 1 2 4 . 3 9 0 .0001 L e a s t Squa re s Means P_Z0NE LNRBAI6 LSMEAN LSMEAN Number 1 - 0 . 0 8 2 6 9 7 9 1 1 2 - 0 . 2 5 7 6 9 5 1 7 2 3 - 0 . 3 1 6 9 2 3 6 6 3 4 - 0 . 3 4 4 1 3 8 5 0 4 5 - 0 . 2 1 7 1 0 3 9 5 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 1 3 . 031385 4 . 042925 4 507487 2 .319979 0 . 0025 0 . 0001 0 .0001 0 .0205 2 -3 . 03138 1 . 035001 1 508732 - 0 . 7 0 9 3 2 0 . 0025 0 .3009 0 .1317 0 .4783 3 - 4 . 0 4 2 9 3 - 1 . 035 0 473292 - 1 . 7 3 8 0 9 0 . 0001 0 .3009 0 . 6 3 6 1 0 . 0825 4 - 4 . 5 0 7 4 9 -1.50873 - 0 .47329 -2 .20925 0 .0001 0 .1317 0 . 6361 0 . 0274 5 -2 .31998 0 .709323 1 .738091 2 209246 0 . 0205 0 .4783 0 .0825 0 .0274 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 197 L e a s t Squa re s Means COR_DIR LNRBAI6 T / P r > | T | HO: LSMEAN LSMEAN1=LSMEAN2 W - 0 . 2 4 5 0 6 1 9 3 - 0 . 0 7 4 1 6 0 .9409 X - 0 . 2 4 2 3 6 1 7 4 P l o t o f RESID6*PRED6. Symbol u s e d i s RESID6 2 * * * * * * * * * * * * * * * * * * * * i * * * * * * * * * * * * * * * * * * * * * * * * •> * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * : * * * * * * * * * * * : * * * * * * * * * * * * * * * * * * * * * * * S"fffffffff~ fffffffff~ fffffffff~fffffffff~fffffffff~ffffffffff - 0 . 8 - 0 . 6 - 0 . 4 - 0 . 2 0 .0 0 .2 0 .4 NOTE: 181 obs had m i s s i n g v a l u e s . PRED6 554 obs h i d d e n . 198 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 6 Moments N 1030 Sum Wgts 1030 Mean 0 Sum 0 S t d Dev 0 .582276 V a r i a n c e 0 .339046 Skewness - 0 . 0 0 8 4 7 K u r t o s i s 0 .464095 USS 348 . 878 CSS 3 4 8 . 8 7 8 CV S t d Mean 0 .018143 T:Mean=0 0 Pr>1T1 1 .0000 Num ~= 0 1030 Num > 0 514 M ( S i g n ) -1 Pr>=|M 0 . 9751 Sgn Rank 8 6 5 . 5 Pr>=|S 0 . 9278 W:Normal 0 . 987425 Pr<W 0 . 6279 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 6 H i s t o g r a m 2 .1 + " 0 .7 + - 0 . 7 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * -2 .1+* may r e p r e s e n t up t o 4 c o u n t s # 2 4 7 11 18 33 69 97 134 139 152 125 91 67 34 22 10 11 3 B o x p l o t 0 0 0 + + I + I * * + + 199 V a r i a b l e = R E S I D 6 2 . 1 + U n i v a r i a t e P r o c e d u r e N o r m a l P r o b a b i l i t y P l o t 0 .7 + -0 .7 + _i_ * * * ***** ***** ***** * * * * ***** * * * * ***** * * * * * * * * * * * * + + * * * * * * * -2 . ! + •> + 1 + 2 200 APPENDIX VII Stepwise backwards regression for predicting relative tree basal area per hectare for Zone 1 from plot level variables. 201 O u t p u t 1. R e g r e s s i o n o f Zone 1 b a s a l a r e a p e r h a r e l a t i v e t o Zone 5 (RLBAHAZ) as a f u n c t i o n o f p l o t l e v e l v a r i a b l e s . BA_HA_PT i s t h e p l o t b a s a l a r e a p e r h a ; A V _ S I _ 5 0 i s t h e s i t e i n d e x ; AGE_PLOT i s t h e p l o t age a t t i m e o f s a m p l i n g ; C_RD_AGE i s t h e number o f y e a r s s i n c e t h e r o a d was b u i l t ; and AVG_RW i s t h e r o a d w i d t h . B a c k w a r d E l i m i n a t i o n P r o c e d u r e f o r D e p e n d e n t V a r i a b l e RLBAHAZ S t e p 0 A l l V a r i a b l e s E n t e r e d R - s q u a r e = 0 15778713 C ( p ) = 6. 00000000 DF Sum o f S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 5 2 . 13859454 0 .42771891 1 42 0 .2379 E r r o r 38 11 . 41507454 0 .30039670 T o t a l 43 13 . 55366907 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum o f S q u a r e s F Prob>F INTERCEP 1 .28389351 0 . 63510881 1 .22760001 4 09 0 .0503 BA_HA_PT - 0 . 0 4 0 0 6 2 9 8 0. 01708977 1 .65085655 5 50 0 .0244 A V _ S I _ 5 0 0 .04464335 0. 03807537 0 .41297170 1 37 0 .2483 A G E _ P L O T 0 .00186117 0 . 00433402 0 .05539688 0 18 0 .6700 C_RD_AGE 0 .00703162 0 . 00749617 0 .26431819 0 88 0 .3542 AVG_RW 0 .00707337 0 . 01113531 0 .12121129 0 40 0 .5291 Bounds on c o n d i t i o n number: 2 . 688933, 44 72503 S t e p 1 V a r i a b l e A G E _ P L 0 T Removed R - s q u a r e = 0 .15369991 C(p) = 4 .18441242 DF Sum o f S q u a r e s Mean S q u a r e F Prob>F R e g r e s s i o n 4 2 . 08319765 0 .52079941 1 77 0 .1543 E r r o r 39 11 . 47047142 0 .29411465 T o t a l 43 13 . 55366907 P a r a m e t e r S t a n d a r d T y p e I I V a r i a b l e E s t i m a t e E r r o r Sum o f S q u a r e s F Prob>F INTERCEP 1 .42610878 0. 53623481 2 .08023293 7 07 0 .0113 B A _ H A _ P T - 0 . 0 3 6 5 9 5 5 6 0 . 01490367 1. 77331867 6 03 0 .0186 A V _ S I _ 5 0 0 .03978755 0. 03597543 0 .35974858 1 22 0 .2755 C_RD_AGE 0 .00659717 0. 00734951 0 .23698206 0 81 0 .3749 AVG_RW 0 .00674868 0. 01099283 0 .11084976 0 38 0 .5428 Bounds on c o n d i t i o n number: 2 . 0 8 8 6 8 4 , 27 .27273 S t e p 2 V a r i a b l e AVG_RW Removed R - s q u a r e = 0 .14552133 C(p) 2 . 55342368 R e g r e s s i o n E r r o r T o t a l DF 3 40 43 Sum o f S q u a r e s 1 ..97234789 11 .58132118 13 . 55366907 Mean S q u a r e 0 . 65744930 0 .28953303 F 2 .27 Prob>F 0 . 0950 V a r i a b l e P a r a m e t e r E s t i m a t e S t a n d a r d E r r o r T y p e I I Sum o f S q u a r e s Prob>F INTERCEP B A _ H A _ P T A V _ S I _ 5 0 C RD AGE 63307070 03602711 03917737 00436418 0 .41375182 0 .01475857 0 .03568050 0 .00633640 4 .51053709 1 .72531504 0 .34906538 0 .13734627 15 .58 5 .96 1 .21 0 .47 0 .0003 0 . 0192 0 . 2788 0 .4950 Bounds on c o n d i t i o n number: 2 . 0 8 0 6 2 2 , 15 .43841 202 S t e p 3 V a r i a b l e C_RD_AGE Removed R - s q u a r e = 0 .13538781 C ( p ) = 1 .01063998 R e g r e s s i o n E r r o r T o t a l DF 2 41 43 Sum o f S q u a r e s 1. 83500162 11 .71866745 13 . 55366907 Mean S q u a r e 0 .91750081 0 .28582116 F 3 . 2 1 Prob>F 0 .0507 V a r i a b l e P a r a m e t e r E s t i m a t e S t a n d a r d E r r o r T y p e I I Sum o f S q u a r e s Prob>F INTERCEP B A _ H A _ P T AV SI_50 1 .71657503 -0 .03430283 0 .03762899 0 .39304571 0 .01445115 0 .03538061 The SAS S y s t e m 5 .45172358 1 .61045815 0 .32330244 19 .07 5 . 63 1 .13 0 .0001 0 .0224 0 .2938 Bounds on c o n d i t i o n number: 2 . 020753, 8 . 083012 S t e p 4 V a r i a b l e A V _ S I _ 5 0 Removed R - s q u a r e = 0 .11153431 C ( p ) = 0 .08689162 R e g r e s s i o n E r r o r T o t a l DF 1 42 43 Sum o f S q u a r e s 1 .51169918 12 . 04196989 13 .55366907 Mean S q u a r e 1 .51169918 0 .28671357 F 5 .27 Prob>F 0 .0267 V a r i a b l e P a r a m e t e r E s t i m a t e S t a n d a r d E r r o r T y p e I I Sum o f S q u a r e s Prob>F INTERCEP BA HA PT 1 .98333191 -0 . 02337929 0 .30308628 0 .01018176 12 .27740241 1 .51169918 42 . 82 5 .27 0 .0001 0 .0267 Bounds on c o n d i t i o n number: 1, A l l v a r i a b l e s l e f t i n t h e m o d e l a r e s i g n i f i c a n t a t t h e 0 .0500 l e v e l . Summary o f B a c k w a r d E l i m i n a t i o n P r o c e d u r e f o r D e p e n d e n t V a r i a b l e RLBAHAZ V a r i a b l e Number P a r t i a l M o d e l S t e p Removed I n R**2 R**2 C ( p ) F Prob>F 1 A G E _ P L O T 4 0 . 0041 0 .1537 4 1844 . 0 1844 0 .6700 2 AVG_RW 3 0 .0082 0 .1455 2 5534 0 3769 0 .5428 3 C_RD_AGE 2 0 .0101 0 .1354 1 0106 0 4744 0 .4950 4 A V _ S I _ 5 0 1 0 .0239 0 .1115 0 0869 1 1311 0 .2938 203 P l o t o f R E S I D * Y H A T . Symbol u s e d i s 1.5 - 1 . 5 * * * * * sfffffffff-fffffff~fffffff"fffffff~fffffff"ffffffffffffff"fffffff"ff 0 .9 1.0 1 .1 1.2 1.3 1.4 1.5 1.6 1.7 P r e d i c t e d V a l u e o f RLBAHAZ U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D R e s i d u a l Moments N 44 Sum Wgts 44 Mean 0 Sum 0 S t d Dev 0. 529194 V a r i a n c e 0 .280046 Skewness 0 . 191491 K u r t o s i s -0 . 47122 USS 12 .04197 CSS 12 .04197 CV S t d Mean 0 .079779 T:Mean=0 0 P r > | T | 1 .0000 Num A = 0 44 Num > 0 19 M ( S i g n ) -3 P r > = | M | 0 .4514 Sgn Rank -15 Pr>= S j 0 .8634 W:Normal 0. 970322 Pr<W 0.4393 204 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D R e s i d u a l Stem L e a f # 10 0 1 8 758 3 6 266 3 4 783 3 2 16737 5 0 1678 4 -0 919543 6 -2 854429885554 12 -4 8 1 -6 55 2 -8 043 3 -10 6 1 B o x p l o t M u l t i p l y S t e m . L e a f b y 10**-1 V a r i a b l e = R E S I D R e s i d u a l N o r m a l P r o b a b i l i t y P l o t 1.1 + * * + * **+*++ **++ * * * * * * * + + * * * * * * * * * * * + *++ ++* + * * _i_ * * -1 .1 + "++ + — i — -2 -1 ._ + + + + + 1 +2 205 APPENDIX VIII General Linear Models to test for differences in tree basal area among zones, while controlling for density, plot age at the time of road right of way and site productivity changes. 206 O u t p u t 1. G e n e r a l l i n e a r model o f t r e e b a s a l a r e a 2 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNBA2) w i t h zone ( P _ Z 0 N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t - o f - w a y h a r v e s t (STDAGEAC) , and s i t e i n d e x ( A V _ S I _ 5 0 ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . Number o f o b s e r v a t i o n s i n d a t a s e t 607 Dependent V a r i a b l e : LNBA2 S o u r c e DF M o d e l 7 E r r o r 599 C o r r e c t e d T o t a l 606 R - S q u a r e 0 .446996 Sum o f S q u a r e s 267 .89629023 331 .42922080 599 .32551103 C . V . 15 .75174 F V a l u e 69 .17 P r > F 0 . 0001 LNBA2 Mean 4 . 7 2 2 2 9 9 5 1 S o u r c e DF Type I SS F V a l u e P r > F P_Z0NE 4 1. 82770906 0 .83 0 .5090 CRD_PT 1 4 . 00359129 7 .24 0 .0073 STDAGEAC 1 181 .71002118 3 2 8 . 4 1 0 .0001 A V _ S I _ 5 0 1 80 .35496869 145 .23 0 .0001 S o u r c e DF Type I I I SS F V a l u e P r > F P_Z0NE 4 1 .72986576 0 .78 0 .5374 CRD_PT 1 7 . 00605138 12 . 66 0 .0004 STDAGEAC 1 194 .26211137 3 5 1 . 09 0 . 0001 A V _ S I _ 5 0 1 80 .35496869 1 4 5 . 2 3 0 .0001 L e a s t Squa re s Means P_Z0NE 1 2 3 4 5 LNBA2 LSMEAN 76283706 71419414 62670072 72889436 .77998859 LSMEAN Number 1 2 3 4 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 - 0 . 5 0 8 5 6 0 .6112 - 1 . 4 2 0 4 4 0 .1560 - 0 . 3 5 3 4 3 0 .7239 0 .178958 0 .8580 0 .508562 0 .6112 -0 . 92053 0 .3577 0 .154342 0 . 8774 0 .692233 0 .4891 1.420438 0 .1560 0 . 920531 0 .3577 1 .070805 0 .2847 1 .609499 0 .1080 0 .353431 0 .7239 -0 .15434 0 . 8774 - 1 . 0 7 0 8 0 .2847 0 .535375 0 . 5926 - 0 . 1 7 8 9 6 0 . 8580 -0 . 69223 0 .4891 - 1 . 6 0 9 5 0 .1080 -0 . 53538 0 .5926 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 207 P l o t o f R E S I D 2 * P R E D 2 . S y m b o l u s e d i s * * * * * * * * * * * • * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * i * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * £ff~fffffff"fffffff"fffffff"fffffff"fffffff"fffffff"fffffff"ff 3 .5 4 . 0 4 . 5 5 .0 5 . 5 6 . 0 6 .5 7 . 0 PRED2 NOTE: 230 obs h i d d e n . V a r i a b l e = R E S I D 2 Moments N 607 Sum Wgts 607 Mean 0 Sum 0 S t d Dev 0 .739536 V a r i a n c e 0 546913 Skewness -0 . 08058 K u r t o s i s 0 201737 USS 331 .4292 CSS 3 3 1 . 4 2 9 2 CV S t d Mean 0 030017 T:Mean=0 0 Pr>1T1 1.0000 Num A= 0 607. Num > 0 293 M ( S i g n ) -10 . 5 Pr> = M 0 .4169 Sgn Rank 42 Pr> = S 0 .9923 W:Normal 0 .982562 Pr<W 0 .1662 208 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 2 H i s t o g r a m 1.9 + 0 .5 -0 . 9 * * * * * * * * * k k k k k k * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * k k k k k k k k k * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * -2 .3 + * # 6 5 6 12 21 35 49 44 50 65 66 76 64 32 24 20 12 7 3 6 1 3 B o x p l o t 0 may r e p r e s e n t up t o 2 c o u n t s V a r i a b l e = R E S I D 2 1.9 + 0 .5 + N o r m a l P r o b a b i l i t y P l o t * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + -0 .9 + * * * _^  * * * •k -k * -2 .3 + * -- + -+ 1 - - + + +2 209 O u t p u t 2 . G e n e r a l l i n e a r mode l o f t r e e b a s a l a r e a 5 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNBA3) w i t h zone ( P _ Z 0 N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t - o f - w a y h a r v e s t (STDAGEAC) , a n d s i t e i n d e x ( A V _ S I _ 5 0 ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among zones . Number o f o b s e r v a t i o n s i n d a t a s e t = 607 Dependent V a r i a b l e : L N B A 3 S o u r c e DF M o d e l 7 E r r o r 599 C o r r e c t e d T o t a l 606 Sum o f Squa re s 236 .15569863 305 .47481660 541 .63051523 F V a l u e 66 .15 P r > F 0 . 0001 R - S q u a r e 0 .436009 C V . 1 4 . 8 6 8 5 1 LNBA3 Mean 4 . 80293722 S o u r c e P_Z0NE CRD_PT STDAGEAC AV S I 50 DF Type I SS F V a l u e P r > F 4 2 .04313692 1.00 0 .4061 1 3 .40022988 6 .67 0 .0101 1 151 .33890158 2 9 6 . 7 6 0 .0001 1 79 .37343025 1 5 5 . 6 4 0 .0001 S o u r c e P_ZONE CRD_PT STDAGEAC A V _ S I _ 5 0 DF Type I I I SS F V a l u e P r > F 4 1 .95011942 0 .96 0 .4312 1 7 .82569134 1 5 . 3 5 0 .0001 1 162 .78041305 3 1 9 . 1 9 0 .0001 1 79 .37343025 1 5 5 . 6 4 0 .0001 L e a s t S q u a r e s M e a n s P_Z0NE 1 2 3 4 5 LNBA3 LSMEAN 4 .85041715 4 .78961251 4 .70452022 4 .80595457 4 .86548313 LSMEAN Number 1 2 3 4 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 -0 . 66217 0 . 5081 - 1 . 5 8 5 6 3 0 .1134 -0 .48224 0 .6298 0 .163739 0 . 8700 0 . 662169 0 .5081 - 0 . 9 3 2 5 3 0 .3514 0 .17872 0 .8582 0 .831465 0 .4060 1 .585631 0 .1134 0 .932526 0 .3514 1 .107081 0 .2687 1 .760421 0 . 0788 0 .482238 0 . 6298 - 0 . 1 7 8 7 2 0 .8582 - 1 . 1 0 7 0 8 0 .2687 0 . 64971 0 . 5161 -0 .16374 0 .8700 -0 . 83147 0 .4060 - 1 . 7 6 0 4 2 0 .0788 - 0 . 6 4 9 7 1 0 . 5 1 6 1 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 210 P l o t o f RESID3*PRED3. Symbol u s e d i s RESID3 2 -1 * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * ** ** * * * * * * * * * * * * * * * * * * * * * * * * * * s/rfffffff"fffffff"fffffff~ fffffff~ fffffff~fffffff~fffffff*ff 3 .5 4 . 0 4 . 5 5 . 0 5 .5 6 . 0 6 .5 7 . 0 PRED3 NOTE: 241 obs h i d d e n . V a r i a b l e = R E S I D 3 Moments N 607 Sum Wgts 607 Mean 0 Sum 0 S t d Dev 0 .709989 V a r i a n c e 0 504084 Skewness -0 . 04328 K u r t o s i s 0 085126 USS 305 .4748 CSS 3 0 5 . 4 7 4 8 CV S t d Mean 0 028818 T:Mean=0 0 P r > | T | 1. 0000 Num A= 0 607 Num > 0 287 M ( S i g n ) - 1 6 . 5 Pr> = M | 0 .1940 Sgn Rank -93 Pr>= s 1 0 .9829 W:Norma l 0 . 981788 Pr<W 0 .1102 211 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 3 H i s t o g r a m # 1 . 9 + * * 4 * * * * 7 1 . 5 + * * 3 * * * * * * * 23 \ ]_ + * * * * * * * * * 2_8 * * * * * * * * * * * * * * * * * * 35 0 7 + * * * * * * * * * * * * * * * * * * * * * * * 4g * * * * * * * * * * * * * * * * * * * * * * * * * 49 + 0 3+**************************** 55 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 57 _o.1 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 73 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * | _o_5 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * g2 + * * * * * * * * * * * * * * * * 32 _0 9+************* 26 * * * * * * * * * _]_3 + * * * * * 9 * * * * * ^ - 1 . 7 + * * 4 . * * 3 - 2 . 1 + * * 3 * may r e p r e s e n t u p t o 2 c o u n t s V a r i a b l e = R E S I D 3 1 . 9 + 1 . 1 + N o r m a l P r o b a b i l i t y P l o t * * * * 1 .5+ **+ I * * * * * * * * * * * 0.7+ * * * * | * * * 0 3+ * * * * + * * * - 0 . 1 + * * * * j * * * * * _ Q 5 + * * * * | * * * + - 0 . 9 + * * * | * * * * - 1 . 3 + +** | + * * * - 1 . 7 + + + * * * j * * - 2 . 1 + * - 2 - 1 0 +1 +2 212 O u t p u t 3 . G e n e r a l l i n e a r mode l o f t r e e b a s a l a r e a 10 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNBA4) w i t h zone ( P _ Z O N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t - o f - w a y h a r v e s t (STDAGEAC), and s i t e i n d e x ( A V _ S I _ 5 0 ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among zones . Number o f o b s e r v a t i o n s i n d a t a s e t 607 Dependent V a r i a b l e : LNBA4 S o u r c e DF M o d e l 7 E r r o r 599 C o r r e c t e d T o t a l 606 R - S q u a r e 0 .420811 Sum o f Squa re s 198 .76260730 273 . 56963563 472 .33224293 C V . 13 .73056 F V a l u e 62 .17 P r > F 0 . 0001 LNBA4 Mean 4 .92189503 S o u r c e P_Z0NE CRD_PT STDAGEAC AV S I 5 0 DF Type I SS F V a l u e P r > F 4 2 .45523308 1.34 0 .2522 1 2 .58628726 5 .66 0 .0176 1 115 .01359441 2 5 1 . 8 3 0 .0001 1 78 .70749255 1 7 2 . 3 4 0 .0001 S o u r c e P_Z0NE CRD_PT STDAGEAC AV S I 50 DF Type I I I SS F V a l u e P r > F 4 2 .36862627 1.30 0 .2700 1 9 .24269512 2 0 . 2 4 0 .0001 1 125 .02003259 2 7 3 . 7 4 0 .0001 1 78 .70749255 1 7 2 . 3 4 0 .0001 L e a s t S q u a r e s Means P_Z0NE 1 2 3 4 5 LNBA4 LSMEAN ,98242840 ,89969665 , 81904875 , 91906737 , 99088129 LSMEAN Number 1 2 3 4 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 -0 . 95205 0 .3415 - 1 . 8 7 6 3 2 0 . 0611 - 0 . 7 2 6 1 8 0 .4680 0 .097077 0 . 9227 0 .952046 0 .3415 - 0 . 9 3 3 9 4 0 .3507 0. 223855 0 . 8229 1. 055956 0 .2914 1 .876324 0 . 0611 0 .933936 0 .3507 1.15353 0 .2492 1 .985865 0 .0475 0 .726177 0 .4680 - 0 . 2 2 3 8 5 0 . 8229 - 1 . 1 5 3 5 3 0 .2492 0 . 828241 0 .4079 -0 . 09708 0 . 9227 -1 .05596 0 .2914 -1 .98587 0 . 0475 -0 .82824 0 .4079 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 213 P l o t o f RESID4*PRED4. Symbol u s e d i s RESID4 -1 * * * * * ** * * * * * * * * * ** ** * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * ** * * * * * * * * * * * * * * * * ** * ** * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * ** *** * * * * * * * * * * * * * * * *** * ***** * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * •> * * * * * * * * * * * * * * * * * * i r * * * * * * * * * * * * * i * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Sff"fffffff"fffffff"fffffff"fffffff"fffffff"fffffff"fffffff"ff 3 .5 4 . 0 4 . 5 5 .0 5 .5 6 .0 6 .5 7 .0 PRED4 NOTE: 253 obs h i d d e n . V a r i a b l e = R E S I D 4 Moments N 607 Sum Wgts 607 Mean 0 Sum 0 S t d Dev 0 671889 V a r i a n c e 0 .451435 Skewness 0 005871 K u r t o s i s - 0 . 0 8 8 4 4 USS 273 .5696 CSS 273 .5696 CV S t d Mean 0 . 027271 T:Mean=0 0 Pr>1T1 1. 0000 Num ~= 0 607 Num > 0 292 M ( S i g n ) - 1 1 . 5 Pr>=|M 0 .3719 Sgn Rank -311 Pr>= jS 0 .9427 W:Norma l 0 980883 Pr<W 0 .0638 214 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 4 H i s t o g r a m 1. 9 + * : * * : * * * * * r * * * * * r * * * * * : * * * * * r * * * * * r * * * * * r * * * * * r * * * * * : * * * * * r * * * * * r * * * * * r * * * * * r * * * * * r * * * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * - 1 . 9 + ' # 1 5 6 13 17 32 46 55 56 61 67 87 63 32 26 15 12 8 3 2 B o x p l o t may r e p r e s e n t u p t o 2 c o u n t s U n i v a r i a t e P r o c e d u r e N o r m a l P r o b a b i l i t y P l o t * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ***** ***** * * * _|_ * * * * * * * * * * * ***** * * - 1 . 9 + * - 2 - 1 0 +1 +2 V a r i a b l e = R E S I D 4 1 . 9 + 215 O u t p u t 4. G e n e r a l l i n e a r model o f t r e e b a s a l a r e a 15 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNBA5) w i t h zone ( P _ Z O N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t - o f - w a y h a r v e s t (STDAGEAC), and s i t e i n d e x ( A V _ S I _ 5 0 ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among z o n e s . NOTE: Due t o m i s s i n g v a l u e s , o n l y 592 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNBA5 S o u r c e DF Sum o f Squa re s F V a l u e P r > F M o d e l 7 160 .19981911 52 .92 0 . 0001 E r r o r 584 252 . 54850190 C o r r e c t e d T o t a l 591 412 .74832101 R -Squa re C V . LNBA5 Mean 0 .388130 13 .03356 5 . 04548489 S o u r c e DF Type I SS F V a l u e P r > F P_Z0NE 4 2 .32190287 1.34 0 .2529 CRD_PT 1 1 .96601973 4 . 5 5 0 . 0334 STDAGEAC 1 8 1 . 57352408 1 8 8 . 6 3 0 . 0001 A V _ S I _ 5 0 1 74 .33837243 1 7 1 . 9 0 0 .0001 S o u r c e DF Type I I I SS F V a l u e P r > F P_Z0NE 4 2 .26566175 1.31 0 .2650 CRD_PT 1 9 .51273552 22 . 00 0 . 0001 STDAGEAC 1 92 .98278847 2 1 5 . 0 2 0 . 0001 A V _ S I _ 5 0 1 74 .33837243 1 7 1 . 9 0 0 .0001 L e a s t S q u a r e s Means P_Z0NE LNBA5 LSMEAN LSMEAN Number 1 5 .09843305 1 2 5 .02027806 2 3 4 .94468116 3 4 5 .04590601 4 5 5 .11967635 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / ' j 1 2 3 4 5 1 0 .912729 1 791873 0 .610882 - 0 . 2 4 7 5 8 0 .3618 0 .0737 0 .5415 0 .8045 2 -0 . 91273 0 888584 - 0 . 3 0 0 6 - 1 . 1 6 8 3 5 0 .3618 0 .3746 0 .7638 0 .2431 3 - 1 . 7 9 1 8 7 -0.88858 - 1 . 1 8 4 8 4 -2 . 05267 0 .0737 0 .3746 0 .2366 0 . 0405 4 - 0 . 6 1 0 8 8 0 .300595 1 184842 - 0 . 8 6 3 4 9 0 .5415 0 .7638 0 .2366 0 .3882 5 0 .247576 1 .168351 2 052668 0 .863486 0 .8045 0 .2431 0 .0405 0 .3882 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 216 P l o t o f RESID5*PRED5. Symbol u s e d i s * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * -2 . 0 S//"7 f f f f f f f f f f f f f f ffffffffffffffffffffffffffffffffffffff 3 .5 4 . 0 4 . 5 5 .0 5 .5 6 .0 6 .5 7 .0 PRED5 NOTE: 15 obs had m i s s i n g v a l u e s . 227 obs h i d d e n . V a r i a b l e = R E S I D 5 Moments N 592 Sum Wgts 592 Mean 0 Sum 0 S t d Dev 0 .6537 V a r i a n c e 0 .427324 Skewness - 0 . 0 2 4 0 7 K u r t o s i s -0 .13881 USS 252 .5485 CSS 2 5 2 . 5 4 8 5 CV S t d Mean 0 .026867 T:Mean=0 0 Pr>1T1 1 .0000 Num A= 0 592 Num > 0 290 M ( S i g n ) -6 Pr> = M 0 .6512 Sgn Rank 82 Pr>= S 0 .9843 W:Normal 0 . 981022 Pr<W 0 .0754 217 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 5 H i s t o g r a m # B o x p l o t * * 3 * * * 6 * * * * * * 11 * * * * * * * * * 17 * * * * * * * * * * * * * * * * 31 * * * * * * * * * * * * * * * * * * * * * 41 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 60 + + * * * * * * * * * * * * * * * * * * * * * * * * * * * 54 1 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 67 1 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 70 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 80 | | * * * * * * * * * * * * * * * * * * * * * * * * * * * * * 58 + + * * * * * * * * * * * * * * 28 * * * * * * * * * * * * * * 27 * * * * * * * * * * 19 * * * * * * 11 * * 4 * * * 5 0 + + + + + + + + * may r e p r e s e n t up t o 2 c o u n t s N o r m a l P r o b a b i l i t y P l o t 1.7+ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * -k -k -k -k -k •k * * * * * * * * * * * * * * * * * * _|_ * * - 1 . 7 + * * + + + + + + + + + + + -2 -1 0 +1 +2 218 O u t p u t 5 . G e n e r a l l i n e a r model o f t r e e b a s a l a r e a 20 y e a r s a f t e r t h e r o a d ( l o g a r i t h m i c t r a n s f o r m a t i o n ; LNBA6) w i t h zone ( P _ Z 0 N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t - o f - w a y h a r v e s t (STDAGEAC), and s i t e i n d e x ( A V _ S I _ 5 0 ) . S i g n i f i c a n c e l e v e l o f 0 .05 u s e d f o r o v e r a l l t e s t ; 0 .005 u s e d f o r e a c h o f t h e p a i r w i s e t e s t s among zones . NOTE: Due t o m i s s i n g v a l u e s , o n l y 517 o b s e r v a t i o n s c a n be u s e d i n t h i s a n a l y s i s . Dependent V a r i a b l e : LNBA6 S o u r c e DF Sum o f S q u a r e s F V a l u e P r > F M o d e l 7 110 .52723988 3 6 . 9 1 0 .0001 E r r o r 509 217 .77178705 C o r r e c t e d T o t a l 516 328 .29902692 R - S q u a r e 0 .336666 C V . 12 . 74978 L N B A 6 Mean 5 .13025748 S o u r c e P_Z0NE CRD_PT STDAGEAC AV S I 50 DF 4 1 1 1 Type I SS 1 .28339531 3 . 03598875 54 .77834002 51 .42951580 F V a l u e 0 . 7 5 7 .10 128 . 03 1 2 0 . 2 1 P r > F 0 . 5584 0 .0080 0 . 0001 0 . 0001 S o u r c e P_Z0NE CRD_PT STDAGEAC A V _ S I _ 5 0 DF 4 1 1 1 Type I I I SS 1.22049617 8 .37559495 71 .35160640 51 .42951580 F V a l u e 0 . 7 1 19 . 58 1 6 6 . 7 7 1 2 0 . 2 1 P r > F 0 . 5832 0 . 0001 0 .0001 0 . 0001 L e a s t Squa re s Means P_Z0NE L N B A 6 L S M E A N 5 . 1 7 8 9 0 8 7 2 5 . 1 2 0 5 2 3 5 9 5 . 0 4 3 5 2 3 8 4 5 . 1 3 7 6 2 9 5 4 5 . 1 7 2 2 6 9 6 1 LSMEAN Number 1 2 3 4 5 T f o r HO: L S M E A N ( i ) = L S M E A N ( j ) / P r > | T | i / j 1 2 3 4 5 - 0 . 6 4 0 4 2 0 .5222 - 1 . 4 8 1 5 3 0 .1391 - 0 . 4 5 0 6 4 0 .6524 -0 . 07265 0 . 9421 0 . 64042 0 .5222 -0 . 8509 0 .3952 0 .18857 0 .8505 0 .571827 0 .5677 1 .481525 0 .1391 0 . 850898 0 .3952 1.034958 0 .3012 1.41936 0 .1564 0 .450636 0 . 6524 - 0 . 1 8 8 5 7 0 .8505 - 1 . 0 3 4 9 6 0 .3012 0 .380965 0 .7034 0 . 072652 0 . 9 4 2 1 - 0 . 5 7 1 8 3 0 .5677 - 1 . 4 1 9 3 6 0 .1564 -0 .38097 0 .7034 NOTE: To e n s u r e o v e r a l l p r o t e c t i o n l e v e l , o n l y p r o b a b i l i t i e s a s s o c i a t e d w i t h p r e - p l a n n e d c o m p a r i s o n s s h o u l d be u s e d . 219 P l o t o f R E S I D 6 * P R E D 6 . S y m b o l u s e d i s " * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * ** * * ** * * * * * * * ** * ** ** * * ** ** * * * * * * * * * * * * * ** * * * * ** * * * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ** ** ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * r * * * * * k-ffffffffrfffffffirfffffffffffffffffrffffffffrffffffffrf 4 . 0 4 . 5 5 .0 5 .5 6 .0 6 .5 7 . 0 PRED6 NOTE: 90 obs had m i s s i n g v a l u e s . 206 obs h i d d e n . V a r i a b l e = R E S I D 6 Moments N 517 Sum Wgts 517 Mean 0 Sum 0 S t d Dev 0 . 649645 V a r i a n c e 0 .422038 Skewness -0 . 08699 K u r t o s i s - 0 . 1 4 9 0 4 USS 217 .7718 CSS 2 1 7 . 7 7 1 8 CV S t d Mean 0 . 028571 T:Mean=0 0 P r > | T | 1 .0000 Num "= 0 517 Num > 0 257 M ( S i g n ) - 1 . 5 Pr> = M 0 . 9299 Sgn Rank 616 .5 Pr> = S 0 . 8563 W:Norma l 0 . 982598 Pr<W 0 .2232 220 U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 6 H i s t o g r a m 1. 9 + * r * * r * * * * r * * * * r * * * * : * * * * : * * * * : * * * * r * * * * r * * * * r * * * * r * * * * r * * * * r * * * * r * * * * r * * * r * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * -1 .9+ ' # 1 1 5 10 13 28 37 58 43 61 52 71 50 31 22 16 8 5 4 1 B o x p l o t * may r e p r e s e n t up t o 2 c o u n t s U n i v a r i a t e P r o c e d u r e V a r i a b l e = R E S I D 6 N o r m a l P r o b a b i l i t y P l o t 1. 9 + -1 .9+" * * * * * * * * * * * ***** * * * * ***** * * * _|_ * * * * + * * * ***** ***** * * * * * * * * * * * + * * * + * * * * * + + + + + + + + + + + -2 -1 0 +1 +2 221 O u t p u t 6. M u l t i p l e a n a l y s i s o f c o v a r i a n c e f o r t r e e b a s a l a r e a f o r f i v e p e r i o d s ( l o g a r i t h m i c t r a n s f o r m a t i o n s ) j o i n t l y v e r s u s zone ( P _ Z O N E ) , C u r t i s ' r e l a t i v e d e n s i t y i n d e x o f t h e p l o t ( C R D _ P T ) , p l o t age a t t i m e o f r o a d r i g h t -o f - w a y h a r v e s t (STDAGEAC), and s i t e i n d e x ( A V _ S I _ 5 0 ) . E = E r r o r SS&CP M a t r i x L N B A 2 LNBA2 294 .5170004 LNBA3 2 8 1 . 8 7 6 3 6 4 1 LNBA4 263 .0628334 LNBA5 250 .0174503 LNBA3 2 8 1 . 8 7 6 3 6 4 1 271 .0189969 254 .4770792 243 .0239376 LNBA4 263 . 0628334 254 .4770792 241 .5927467 232 .576394 LNBA5 250 . 0174503 243 . 0239376 232 . 576394 225 . 8000234 LNBA6 241 .0502544 235 .1647048 226 .3664973 2 2 1 . 0 9 2 4 0 2 5 LNBA6 241 .0502544 235 .1647048 226 .3664973 2 2 1 . 0 9 2 4 0 2 5 217 .771787 P a r t i a l C o r r e l a t i o n C o e f f i c i e n t s f rom t h e E r r o r SS&CP M a t r i x / P r o b > I r I DF = 509 LNBA2 L N B A 2 1. 000000 0 . 0001 L N B A 3 0 . 997708 0 . 0001 L N B A 4 0 . 986194 0 .0001 L N B A 5 0 . 969512 0 . 0001 LNBA6 0 . 951813 0 . 0001 LNBA3 0 . 997708 0 .0001 1. 000000 0 . 0001 0 .994505 0 . 0001 0 . 982397 0 . 0001 0 .967991 0 . 0001 LNBA4 0 .986194 0 . 0001 0 .994505 0 . 0001 1. 000000 0 . 0001 0 . 995776 0 . 0001 0 . 986891 0 . 0001 L N B A 5 0 .969512 0 . 0001 0 .982397 0 . 0001 0 . 995776 0 . 0001 1 .000000 0 . 0001 0 .997036 0 .0001 LNBA6 0 .951813 0 .0001 0 .967991 0 . 0001 0 . 986891 0 . 0001 0 . 997036 0 . 0001 1. 000000 0 .0001 H = Type I I I SS&CP M a t r i x f o r P_Z0NE LNBA2 LNBA3 LNBA4 LNBA5 LNBA6 LNBA2 0 .758663519 0 .78281296 0 . 828477495 0 .831782492 0 . 840637154 LNBA3 0 .78281296 0 .820560359 0 .888722249 0 . 908830876 0 . 927868385 LNBA4 0 .828477495 0. 888722249 0 . 996515661 1. 045085461 1 .080699331 LNBA5 0 .831782492 0 . 908830876 1 .045085461 1 .116462071 1 .164429132 LNBA6 0 . 840637154 0 . 927868385 1 .080699331 1 .164429132 1 .22049617 H = Manova T e s t C r i t e r i a and F A p p r o x i m a t i o n s f o r t h e H y p o t h e s i s o f no O v e r a l l P_ZONE E f f e c t Type I I I SS&CP M a t r i x f o r P_Z0NE E = E r r o r SS&CP M a t r i x S=4 M=0 N=251.5 S t a t i s t i c V a l u e Num DF Den DF P r > F W i l k s ' Lambda 0 .96795352 0 .82694 P i l l a i ' s T r a c e 0 .03232745 0 .82781 H o t e l l i n g - L a w l e y T r a c e 0 .03281802 0 .82619 R o y ' s G r e a t e s t Roo t 0 .02078948 2 . 1 1 2 2 1 20 20 20 5 1 6 7 5 . 8 5 2032 2014 508 0 .6822 0 . 6812 0 .6833 0 .0626 NOTE: F S t a t i s t i c f o r R o y ' s G r e a t e s t Roo t i s an u p p e r b o u n d . 222 

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.831.1-0074998/manifest

Comment

Related Items