Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The ecology of planted Engelmann spruce (Picea engelmannii Parry) seedlings on subalpine forest cutovers Caza, C. L. (Caroline Louella) 1991

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

Item Metadata

Download

Media
831-UBC_1991_A1 C39.pdf [ 24.04MB ]
Metadata
JSON: 831-1.0100399.json
JSON-LD: 831-1.0100399-ld.json
RDF/XML (Pretty): 831-1.0100399-rdf.xml
RDF/JSON: 831-1.0100399-rdf.json
Turtle: 831-1.0100399-turtle.txt
N-Triples: 831-1.0100399-rdf-ntriples.txt
Original Record: 831-1.0100399-source.json
Full Text
831-1.0100399-fulltext.txt
Citation
831-1.0100399.ris

Full Text

THE ECOLOGY OF PLANTED ENGELMANN SPRUCE (Picea engelmannii Parry) SEEDLINGS ON SUBALPINE FOREST CUTOVERS by C A R O L I N E L . C A Z A B.Sc, The University of Toronto 1980 M.Sc., The University of Toronto 1983 A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Faculty of Forestry) We accept this thesis as conforming to the required standard The University of British Columbia September 1991 © Caroline L . Caza, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of ^ British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of FDABCT SCJEkJCJES The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT At high elevations in south-central British Columbia conifers are slow to re-establish after logging and the vegetation on many harvested sites is dominated by herbs and shrubs. At present, Engelmann spruce (Picea engelmannii Parry) is widely planted on subalpine cutovers to increase stocking levels, but the growth and survival of planted seedlings is often poor and is highly variable. The objectives of this research were to study: 1) variation in the environments within which Engelmann spruce establishes on subalpine cutovers; 2) the performance of planted seedlings in these environments both with and without interference from non-crop species; 3) the relationship between the growth of planted seedlings and naturally established Engelmann spruce and; 4) the responses of non-crop species to post-logging disturbances. Engelmann spruce seedlings were planted in 1987 into patches of herbs and shrubs and on skid trails on two subalpine cutovers that were winter-logged in 1983. Seedlings grew poorly beneath canopies of all non-crop species where low light levels were a dominant factor limiting growth despite differences between patch types in other environmental factors. Differences in light availability between patches were associated with a relatively greater impact of herbs on stem diameter and lateral growth and of shrubs on height growth. The removal of non-crop species increased light levels and soil temperatures and resulted in significantly greater growth and ratios of shoot:root biomass in open-grown seedlings. Increases in the size or number of most components in open-grown seedlings were strongly correlated. Needles, however, responded differently to treatments than other seedling components. Increases in shoot:root ratios were size-related and due to differences in the relative growth rates of roots and shoots. There was no evidence of shifts in carbon allocation within seedlings in response to variation in resource availability. There were significant differences between the characteristics of planted and naturally established Engelmann spruce seedlings. Open-grown planted seedlings were larger than naturals of the same age and had higher relative growth rates, but similar ratios of needle:stem biomass. Shaded planted seedlings were also larger than naturals but had lower relative growth rates and lower ratios of needle:stem biomass. There were also differences between naturals and planted seedlings in the morphology of root systems. After the removal of above-ground vegetation, dominant herbs re-established cover within one season, mainly from persistent below-ground structures. Dominant shrubs recovered more slowly but were not replaced by new species, even after the removal of both above- and below-ground biomass. Shifts to new dominants occurred after the removal of total biomass in herb patches and also in undisturbed herb patches. Species shifts in undisturbed herb patches as well as increases in total cover in both herb and shrub patches over the study period suggested that the plant communities on the study sites were not at equilibrium. The results of this research indicated that shifts in carbon allocation within seedlings are not part of Engelmann spruce's strategy for establishment in heterogeneous subalpine environments. It is suggested that patterns of growth are conservative rather than competitive and that increased levels of resources must be delivered directly to seedlings to improve early performance. The removal of non-crop vegetation is one way to do this but dominant non-crop species, particularly herbs, will re-establish rapidly relative to the rate of spruce establishment, suggesting limited benefits from vegetation management on subalpine cutovers. Any form of harvesting that causes mimimum understorey disturbance on these sites will result in levels of spruce regeneration that fail to meet current stocking standards. Planting large and vigorous seedlings in areas where resources are high and interference is low may alleviate this problem, but changing silvicultural expectations to better reflect the constraints on conifer regeneration in subalpine environments may be a more effective solution. T A B L E O F C O N T E N T S ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES. ; .vii LIST OF FIGURES xi LIST OF PLATES xvii ACKNOWLEDGEMENTS xix CHAPTER 1 IKTRODUCTION 1 The Characteristics of Engelmann spruce - Subalpine fir Forests 4 The Characteristics of Post-logging Plant Communities in the ESSF 12 Research Objectives 23 Ecological Concepts 25 A Review of the Biology of Engelmann spruce 29 CHAPTER 2 THE ENVIRONMENTAL CHARACTERISTICS OF HERB, SHRUB, AND SKID TRAIL PATCHES ON TWO CUTO VERS IN THE ESSFwc2 37 INTRODUCTION 37 Literature Review 38 METHODS 46 Soil profiles . . 46 Air and soil temperatures 47 Soil physical properties 47 Soil chemical properties 48 Light 49 RESULTS 50 The species composition of patches 50 Soil profiles 50 Air and soil temperatures 53 Soil physical properties 58 Soil chemical properties 65 Light 67 DISCUSSION 71 CONCLUSIONS 78 CHAPTER 3 THE PERFORMANCE OF PLANTED ENGELMANN SPRUCE SEEDLINGS IN HERB, SHRUB, AND SKID TRAIL PATCHES ON TWO CUTOVERS IN THE ESSFwc2 80 INTRODUCTION .80 Literature Review 81 V METHODS 91 Experimental design 91 Sampling design and measurements .94 Soil physical and chemical properties 95 Soil temperature 95 Soil moisture •. 96 Nitrogen ion concentrations 96 Litter decomposition rates 97 Data analysis 98 RESULTS 1 100 Seedling survival and condition 100 Seedling size 104 Seedling biomass 117 Allometric relationships 120 Environmental variables 126 Soil and air temperatures 126 Soil moisture content .131 Nitrate and ammonium ion concentrations 131 Litter decomposition 134 DISCUSSION 137 CONCLUSIONS 146 CHAPTER 4 A COMPARISON OF T H E CHARACTERISTICS OF P L A N T E D A N D N A T U R A L L Y ESTABLISHED SEEDLINGS OF E N G E L M A N N SPRUCE O N A CUTOVER A N D IN A N OLD-GROWTH SPRUCE-FIR STAND IN T H E ESSFwc2 147 INTRODUCTION 147 Literature Review 148 METHODS 152 General description of natural regeneration 152 Seedling selection 152 Sampling methods 155 Data analysis 156 RESULTS 158 General description of natural regeneration 158 Seedling size 159 Seedling biomass 165 Environmental variables 170 DISCUSSION 175 CONCLUSIONS 181 CHAPTER 5 C O M M U N I T Y A N D SPECIES RESPONSES TO DISTURBANCE IN HERB A N D SHRUB PATCHES O N CUTOVERS IN T H E ESSFwc2 . . . . 183 INTRODUCTION 183 Literature Review v i . 184 METHODS 190 Experimental design 190 Sampling methods and measurements 191 Data analysis 192 RESULTS . . 195 Community responses 195 Species richness .195 Species cover 199 Invasion indices 205 Species-specific responses 212 Phenology 212 Dominance shifts 213 The characteristics of conifer regeneration 218 DISCUSSION 220 CONCLUSIONS 228 CHAPTER 6 SYNTHESIS AND CONCLUDING REMARKS 229 REFERENCES 240 APPENDIX 1 A LIST OF THE VASCULAR PLANTS ON TWO SIX-YEAR-OLD (IN 1989) CUTOVERS IN THE ESSFwc2 252 APPENDIX 2 DENDROGRAM FROM CLUSTER ANALYSIS OF SPECIES-COVER QUADRATS ON TWO THREE-YEAR-OLD (IN 1986) CUTOVERS IN THEESSFwc2 256 APPENDIX 3 STATISTICS TABLES FOR DATA PRESENTED IN CHAPTER 3 259 APPENDIX 4 STATISTICS TABLES FOR DATA PRESENTED IN CHAPTER 4 282 APPENDIX 5 STATISTICS TABLES FOR DATA PRESENTED IN CHAPTER 5 291 v i i LIST OF TABLES Table 1.1 General characteristics of Engelmann spruce (Picea engelmannii Parry) and subalpine fir (Abies lasiocarpa (Hook.) Nutt.) overstorey vegetation on a young cutover and in an old-growth forest in the ESSFwc2. .7 Table 2.1 Composition of herb-dominated, shrub-dominated, and skid trail vegetation on two cutovers in the ESSFwc2 . .51 Table 2.2 Physical properties of soils beneath herb-dominated, shrub-dominated, and skid trail vegetation on two cutovers in the ESSFwc2. .. i ....... : 60 Table 23 Total carbon in humus and soils beneath herb-dominated, shrub-dominated and skid trail vegetation on study site 1 . . . . . . . . . . . . 66 Table 2.4 Mineralizable nitrogen in humus and soils beneath herb-dominated and shrub-dominated vegetation on study site 2 66 Table 3.1 Statistics on the height of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1. 260 Table 3.2 Nested ANCOVA and orthogonal contrasts for spruce seedling height in seven treatments on site 1 .261 Table 3.3 Statistics on the diameter of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1 . . . . . . . 262 Table 3.4 Nested ANCOVA and orthogonal contrasts for spruce seedling diameter in seven treatments on site 1 263 Table 3.5 Statistics on the relative growth rate (RGR)* in seedling diameter of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1 .264 Table 3.6 Statistics on the number of laterals on cold-planted and hot-lifted spruce seedlings after three growing seasons on site 1 265 Table 3.7 Nested ANCOVA for number of laterals on spruce seedlings in seven treatments on site 1 266 Table 3.8 Statistics on the length of laterals in cold-planted spruce seedlings after three growing seasons in seven treatments on sites 1 and 2 267 Table 3.9 Nested ANCOVA and orthogonal contrasts for length of laterals in cold-planted spruce seedlings in seven treatments on sites 1 and 2 . . . . . . . 268 Table 3.10 Statistics on the length of needles in cold-planted spruce seedlings after three growing seasons in seven treatments on sites 1 and 2 269 Table 3.11 Nested ANCOVA and orthogonal contrasts for length of needles in cold-planted spruce seedlings in seven treatments on sites 1 and 2 270 v i i i Table 3.12 Statistics on total, shoot, and root biomass of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1 271 Table 3.13 Nested ANOVA and orthogonal contrasts for shoot and root biomass in cold-planted spruce seedlings in seven treatments on site 1 272 Table 3.14 Ratio of seedling biomass in clipped treatment to seedling biomass in control treatments for cold and hot stock within each patch type on sites 1 and 2 273 Table 3.15 Statistics on shoot:root biomass ratios in cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1 274 Table 3.16 Nested ANOVA and orthogonal contrasts for shoot:root biomass ratios in seven treatments on site 1 275 Table 3.17 ANOVA and orthogonal contrasts for soil temperatures in seven treatments on site 1 276 Table 3.18 Nested ANOVA and orthogonal contrasts for soil moisture contents in seven treatments on sites 1 and 2 278 Table 3.19 Orthogonal contrasts for nitrate and ammonium ion concentrations in resin bags in seven treatments on site 1 279 Table 3.20 ANOVA and orthogonal contrasts for weight loss in mixed herb and mixed shrub litter bags in seven treatments on site 1 280 Table 3.21 Statistics on weight loss in herb and shrub litter bags in control treatments on site 1 281 Table 4.1 ANOVAs of height and diameter of young planted and naturally established spruce seedlings on a cutover and in an old-growth forest.. 283 Table 4.2 Analysis of relative production rates in young planted and naturally established spruce seedlings on a cutover and in an old-growth forest 284 Table 4.3 Component loadings from a principal components analysis of the correlation matrix of nine variables representing the size characteristics of six-year-old Engelmann spruce seedlings. Two factors had eigenvalues greater than 1.0 285 Table 4.4 ANOVAs of shoot and root biomass of young planted and naturally established spruce seedlings on a cutover land in an old-growth forest 286 Table 4.5 ANOVAs of biomass ratios in young planted and naturally established spruce seedlings on a cutover and in an old-growth forest 287 Table 4.6 Pearson correlation matrix of variables characterizing the growth and environment of Engelmann spruce seedlings .288 ix Table 4.7 Mean and standard deviations for characteristics describing the environments of five groups of Engelmann spruce seedlings 289 Table 4.8 Principal components analysis of the correlation matrix of twelve variables representing the characteristics of the environments of planted and naturally established Engelmann spruce seedlings. Five factors had eigenvalues greater than 1.0 290 Table 5.1 Two-way ANCOVA for total number of species three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 292 Table 5.2 Statistics on total percent cover (summed by species) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 293 Table 5.3 Two-way ANOVA for total percent cover (summed by species) three seasons after disturbance in herb and shrub patches on sites 1 and 2 294 Table 5.4 Statistics on changes in herb cover three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 295 Table 5.5 Two-way ANOVA for logg-transformed changes in herb cover ([final cover/initial cover] x 100) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 296 Table 5.6 Statistics on changes in shrub cover three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 297 Table 5.7 Two-way ANOVA for logg-transformed changes in shrub cover ([final cover / initial cover] x 100) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 298 Table 5.8 Two-way ANOVA for Berger-Parker dominance index three growing seasons after disturbance in herb and shrub patches on sites 1 and 2 299 Table 5.9 Two-way ANOVA for constancy index three seasons after disturbance in herb and shrub patches on sites 1 and 2 300 Table 5.10 Two-way ANOVA and statistics for logg-transformed (distance between sampling point and nearest neighbour) three growing seasons after disturbance in herb and shrub patches on site 1 301 Table 5.11 Species with highest percent cover (in parentheses) in plots before and after disturbance on sites 1 and 2 302 Table 5.12 Species with highest percent cover (in parentheses) before and after repeated clipping in the planted spruce plots on sites 1 and 2. 304 Table 5.13 Growth of clipped and undipped plants of Valeriana sitchensis and Menziesia ferruginea between June 6 and August 6 (1987) on site 1 306 Table 5.14 Characteristics of Engelmann spruce established prior to 1987 in herb and shrub control plots on sites 1 and 2 307 X Table 5.15 Characteristics of Engelmann spruce seedlings invading herb and shrub plots after treatment in 1987.. . .' 308 Table 6.1 Stocking standards for the ESSFwc2 (from Lloyd et al, 1990) 233 x i LIST OF FIGURES Figure 1.1 Location of the study sites in the Engelmann spruce - Subalpine fir biogeoclimatic zone. Source: Meidinger and Pojar, 1991 6 Figure 2.1 Soil profiles beneath herb-dominated and shrub-dominated vegetation on study site 1 .52 Figure 2.2 Average daily air temperatures from July to October 1989 at 15 cm above-ground in a herb patch, a shrub patch, and in the open on study site 1 54 Figure 23 Average daily soil temperatures from July 1989 to June 1990 at 4 depths in a herb patch on study site 1 56 Figure 2.4 Average daily soil temperatures from July to October 1989 at 10 cm in shrub, herb, vegetated skid trail and unvegetated skid trail patches on study site 1 57 Figure 2.5 Average daily soil temperatures from April to June 1990 at 10 cm in shrub, herb, vegetated skid trail and unvegetated skid trail patches on study site 1 59 Figure 2.6 Water retention curves for soils from herb (-4 ), shrub (#), and skid trail (•) patches (both sites combined). Means and standard error of the means are indicated by symbols and error bars. n=6 for herb and shrub patches, n=2 for skid trail patches 64 Figure 2.7 Means and standard error of the means for light levels from 0 to 2 m above the ground in herb (-4 ), shrub (0), and skid trail (•) patches in July 1988 (both sites combined). n=30 .68 Figure 2.8 Average hourly light levels at 15 cm above the ground in herb and shrub patches from July 26th to 28th 1988 on study site 1. PAR = photosynthetically active radiation .69 Figure 3.1 Condition of cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on sites 1 and 2 (both sites combined). Data are presented as percentages of total number of seedlings, n (number of seedlings per treatment) = 150. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid t r a i l . . . . . . . . . . . . 101 Figure 3.2 Treatment means and standard error of the means for height (upper figure) and diameter (lower figure) of two stock-types of Engelmann spruce after three growing seasons on site 1. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 106 Figure 33 Treatment means and standard error of the means for height:diameter ratios in two stock-types of Engelmann spruce after three growing seasons on site 1. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 109 x i i Figure 3.4 Mean annual height (upper figure) and diameter (lower figure) increments in cold stock Engelmann spruce seedlings planted in 1987 in seven treatments on site 1. Standard errors are omitted for clarity but ranged from 0.26 to 0.85 cm for height increments and 0.09 to 0.29 mm for diameter increments. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 110 Figure 3.5 Treatment means and standard error of the means for number of laterals on two stock-types of Engelmann spruce seedlings after three growing seasons on site 1 (upper figure) and length of laterals (cold stock only) after three growing seasons on sites 1 and 2 (lower figure). Lateral lengths were based on measurements of three top-most branches per seedling. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 112 Figure 3.6 Treatment means and standard error of the means for length of needles (upper figure) and number of buds (lower figure) on cold stock Engelmann spruce seedlings after three growing seasons on sites 1 and 2. Needle lengths were based on three measurements per seedling, and bud numbers were based on number of buds on top-most three branches per seedling. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 114 Figure 3.7 Treatment means and standard error of the means for length of leader (upper figure) and number of needles (lower figure) on cold stock Engelmann spruce seedlings after three growing seasons on sites 1 and 2. Needle number was based on the number of needles in one rank along a 1 cm length of the leader. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 116 Figure 3.8 Treatment means and standard error of the means for root and shoot biomass in cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on site 1. Errors are based on total (root + shoot) biomass. Scales are the same for both figures (but ranges are different). n=9. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 118 Figure 3.9 Treatment means and standard error of the means for ratios of shoot:root biomass in two stock-types of Engelmann spruce after three growing seasons on site 1. n=9. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 122 x i i i Figure 3.10 Scatterplots (by treatment) of logp-transformed total seedling biomass and seedling diameter for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings three growing seasons after planting on sites 1 and 2 (both sites combined). /i=18. Treatments: l=clipped herb ( (3 )» 2=herb control ( A ), 3=clipped shrub ( V ), 4=shrub control C )> 5=clipped skid trail ( t> ), 6=skid trail control ( • ), 7=bare skid trail ( <Q> ) 124 Figure 3.11 Scatterplots of logp-transformed shoot:root biomass ratios and total seedling biomass for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on sites 1 and 2 (both sites combined) 125 Figure 3.12 Scatterplots (by treatment) of loge-transformed shoot and root biomass for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings three growing seasons after planting on sites 1 and 2 (both sites combined). n=18. Treatments: l=clipped herb ( O ), 2=herb control ( A )> 3=clipped shrub ( V ), 4=shrub control ( ^ ), 5=clipped skid trail ( t> ), 6=skid trail contro!( • ), 7=bare skid trail ( <Q> ) 127 Figure 3.13 Treatment means and standard error of the means for soil temperatures at 10 cm on June 6 (upper), July 4 (middle), and August 6 (lower) in 1989 on sites 1 and 2. Site 1, n=5. Site 2, n=3. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 128 Figure 3.14 Treatment means and standard error of the means for air temperatures at 25 cm on three days in 1989 on site 1. Treatments marked by a star differed significantly with p <0.05. n=5 (June), n=3 (July,August). Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 130 Figure 3.15 Treatment means and standard error of the means for soil moisture content (expressed as % of dry weight) in August 1988 on site 1 (upper figure) and site 2 (lower figure). n=12. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 132 Figure 3.16 Treatment means and standard error of the means for NO3" and NH44" ion concentrations in resin bags incubated for 14 months on site 1. Due to the destruction of some bags by small mammals n varies from 1 to 8 with an average of 6 per treatment (n=l for N H 4 + in treatment 5). Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 133 Figure 3.17 Treatment means and standard error of the means for weight loss in bags of herb and shrub litter incubated for 14 months on site 1. Due to the destruction of some bags by small mammals n varies from 4 to 9 with an average of 6 per treatment. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail 135 xiv Figure 4.1 Figure 4.2 Figure 43 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Means and standard error of the means for height (upper figure) and diameter (lower figure) of Engelmann spruce seedlings from five different environments. n(number per treatment)=15. Group numbers followed by different letters were significantly different with p<0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted . . .160 Means and standard errorof the means for ratio of height:diameter in Engelmann spruce seedlings from five different environments. n=15. Group numbers followed by different letters were significantly different with p<0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted. 161 Means and standard error of the means for annual height increments between 1987 and 1990 in Engelmann spruce seedlings from five different environments. n=15. Groups: ( O )=forest naturals, ( A)=shaded naturals, ( <] ) =open naturals, ( ^ )=shaded planted, ( >j< )=open planted 162 Means and standard error of the means for needle length (upper figure) and number of needles (lower figure) for Engelmann spruce seedlings from five different environments. n=15. Group numbers followed by different letters were significantly different with p<0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted . . .164 Seedling factor scores on the first two axes of a principal components analysis of the correlation matrix of nine variables representing the size characteristics of Engelmann spruce seedlings from five different environments. The ellipses are centered on the group means for factors 1 and 2 and they encompass a 50% confidence region around the means based on the standard deviations for factor scores. Groups: 1 = forest naturals ( O X 2 = shaded naturals ( o ), 3 = open naturals ( # ), 4 = shaded planted ( Y ), 5 = open planted ( -f ) 166 Means and standard error of the means for root, stem, needle, and bud biomass (bars indicate standard errors for total biomass) for Engelmann spruce seedlings from five different environments. n=15 for Groups 1,2,3. n=6 for Groups 4,5. Group numbers followed by different letters were significantly different with p<0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted 168 Means and standard error of the means for ratios of shoot:root (upper figure) and needle:stem (lower figure) biomass for Engelmann spruce seedlings from five environments. n=15 for Groups 1,2,3. n=6 for Groups 4,5. Group numbers followed by different letters were significantly different with p<0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted 171 XV Figure 4.8 Environmental factor scores for seedlings on the first two axes of a principal components analysis of the correlation matrix of twelve variables characterizing five different environments of Engelmann spruce seedlings. The ellipses are centered on the group means for factors 1 and 2 and they encompass a 50% confidence region around the means based on the standard deviations for factor scores. Groups: 1 = forest naturals ( O ), 2 = shaded naturals ( • ), 3 = open naturals ( # ), 4 = shaded planted ( • ), 5 = open planted ( + ) 173 Figure 5.1 Means and standard error of the means for species richness in clipped, screefed, and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. n(number of plots per treatment)=3 197 Figure 5.2 Means and standard error of the means for the difference between pretreatment species richness and species richness three growing seasons after clipping and screefmg in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). n=3 198 Figure 53 Means and standard error of the means for total percent cover (summed by species) of vegetation in clipped, screefed, and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. n=3 200 Figure 5.4 Means and standard error of the means for differences between pretreatment percent cover of herbs (summed by species) and cover three growing seasons after clipping and screefing in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). Values are expressed as percentages of initial pretreatment cover. n=3 204 Figure 5.5 Means and standard error of the means for differences between pretreatment percent cover of shrubs (summed by species) and cover three growing seasons after clipping and screefing in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). Values are expressed as percentages of initial pretreatment cover. n=3 206 Figure 5.6 Means and standard error of the means for Berger-Parker dominance (D) indices for clipped, screefed, and control plots in herb and shrub patches three growing seasons after treatment on site 1 (upper figure) and site 2 (lower figure). D = Pm a x/P, where P m a x = number of sampling points with individuals of the most abundant species as nearest neighbour and P = total number of sampling points, indicates treatment differing significantly from all other treatments when tested with Tukey's HSD test ata = 0.05.n=3 207 Figure 5.7 Means and standard error of the means for a constancy index ( C ) for clipped, screefed, and control plots in herb and shrub patches three growing seasons after treatment on site 1 (upper figure) and site 2 (lower figure). C = P c o n/P, where P ^ J J = number of sampling points with a nearest neighbour of the same species in 1988 and 1989 and P = total number of sampling points, indicates herb treatments differing significantly from all shrub treatments when tested with Tukey's HSD test at a = 0.05. n=3 209 xvi Figure 5.8 Means and standard error of the means for distances from sampling points to nearest neighbours in clipped, screefed and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. n=90 211 x v i i LIST OF PLATES Plate 1.1 Plate 1.2 Plate 13 Plate 1.4 Plate 2.1 Plate 2.2 Plate 3.1 General view of study site 2 from the landing at the base of the site. A mature Engelmann spruce - subalpine fir stand was winter-logged in 1983 with protection of trees less than 20 cm dbh. The site was classified as not satisfactorily restocked (NSR) with conifers in 1986 and is typical of NSR cutovers in the ESSFwc2. This photo was taken in 1989 .13 Vegetation dominating a skid trail patch six years after logging on study site 1. Major species included Epilobium angustifolium (fireweed) and Valeriana sitchensis (Sitka valerian). Veratrum viride (false hellebore) and Athyrium filix-femina (ladyfern) were also present on this skid trail. The woman kneeling on the ground is holding aim stick which indicates the average height of the vegetation .19 Vegetation dominating a shrub patch six years after logging on study site 1. Major species included Menziesia ferruginea (false azalea) and Rhododendron albiflorum (white-flowered rhododendron). The woman in the photograph is holding aim stick which indicates the average height of the vegetation 20 Vegetation dominating a herb patch 6 years after logging on study site 2. Major species included Valeriana sitchensis (Sitka valerian), Thalictrum occidentale (western meadow-rue) and Arnica latifolia (mountain amica). The woman in the photograph is holding aim stick which indicates the average height of the vegetation 21 This photograph shows a snow-free shrub patch adjacent to a snow-covered herb patch on study site 1 on May 5th, 1988. General observations indicated that snow disappeared earlier in the spring from shrub patches than herb patches .62 This photograph shows a bare skid trail plot on study site 1 which is free of snow on May 19th, 1989 while the surrounding areas with vegetation are still covered with snow. Snow remained on the Vavenby sites approximately a week to ten days longer in 1989 than 1988. In the spring surface runoff from meltwater temporarily flooded spruce seedlings planted into unvegetated areas on skid trails, as illustrated in this photograph. . . 63 Engelmann spruce seedling (cold stock) in poor condition three growing seasons after planting in a herb control plot on site 1. In the herb and shrub controls seedlings in poor condition were spindly with fewer and shorter laterals and needles than seedlings in good condition. These seedlings were not chlorotic, however, unlike seedlings in poor condition on skid trails 102 Plate 3.2 Engelmann spruce seedling (cold stock) in good condition three growing seasons after planting in a clipped shrub plot on site 1. Contrast size and form with seedling in poor condition in Plate 3.1. 103 x v i i i Plate 3.3 Engelmann spruce seedling (hot stock) in fair condition three growing seasons after planting in a shrub control plot on site 1. Contrast size and form with seedlings in poor condition in herb control (Plate 3.1) and in good condition in clipped shrub treatment (Plate 3.2) .105 Plate 3.4 Engelmann spruce seedling (cold stock) in poor condition three growing seasons after planting in a shrub control plot on site 2. Note that most new root growth was from the base of the plug. There had been almost no lateral root growth during the first three years after outplanting 119 Plate 3.5 Engelmann spruce seedling (cold stock) in good condition three growing seasons after planting in a clipped herb plot on site 1. Although there was significantly more root egress from the plug in seedlings with better growth, there was still very little lateral root growth in comparison to the roots of naturally established seedlings (see Plate 4.1). Contrast root size and form with that of seedling in poor condition in Plate 3.4 121 Plate 4.1 Naturally established six-year-old Engelmann spruce seedling from edge of skid trail on study site 2. Seedling was growing in the open and was rooted in rotten wood. Root system was sparse, shallow, and laterally-extended. Contrast with shape of root systems on planted seedlings illustrated in Plates 3.4 and 3.5 169 Plate 5.1 Clipped herb plot three growing seasons after treatment on study site 1. Plot is dominated by Valeriana sitchensis (Sitka valerian), Epilobium angustifolium (fireweed) and Arnica latifolia (mountain arnica). These species dominated prior to treatment and they re-established cover rapidly after clipping. The woman in the photograph is kneeling down and is holding aim stick which indicates the average height of the vegetation 201 Plate 5.2 Clipped shrub plot three growing seasons after treatment on study site 2. The most abundant species is Menziesia ferruginea (false azalea) which dominated before treatment but which has not re-established pretreatment levels of cover within the plot. The woman in the photograph is kneeling down and is holding aim stick. The average height of the vegetation was less than 1 m 202 Plate 53 Screefed herb plot three growing seasons after treatment on study site 2. The new post-treatment dominant is Epilobium angustifolium (fireweed) which is covering 100% of the plot. The woman in the photograph is holding a meter stick which indicates that the average height of the vegetation was less than 1 m 215 ACKNOWLEDGEMENTS I would like to express my gratitude for the advice and assistance given to me by a number of persons during the course of this research. First and foremost I thank Dr. Hamish Kimmins who supervised this work and who's critical and creative thinking has stimulated much growth in my own ideas about ecology. The other members of my committee, Dr. Peter Joliffe, Dr. Tony Kozak, Dr. Dennis Lavender, and Dr. Roy Turkington have been both teachers and advisers to me and their knowledge and experience in related fields has added much depth to my research. I am greatly indebted to Alan Vyse and Dennis Lloyd of the Kamloops Regional Office of the Ministry of Forests for their continual support and encouragement throughout this study and for their role in initiating this research. I am also very grateful to Bob Stathers, Scott Lindeburgh, and Mel Montieth and the personnel of the Clearwater Forest District Office of the Ministry of Forests for their invaluable technical assistance, and the historical and resource information they provided throughout the study. Warmest thanks are given to Theresa Dynstee, Jeff Spruston, Diane Pennington, and Nicole Von Stefenelli for their capable field and laboratory assistance, and their companionship during long days of work on the study sites. I would also like to thank my colleagues in the Faculty of Forestry - fellow graduate students, staff and faculty members for innumerable forms of assistance during the years of this study. And I most gratefully acknowledge the financial support I have received from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canada - British Columbia Forest Resource Development Agreement (FRDA). 1 CHAPTER 1 INTRODUCTION The Engelmann spruce - subalpine fir (ESSF) biogeoclimatic zone is the uppermost forested zone in the southern two-thirds of the province of British Columbia (Meidinger and Pojar, 1991). The zone is characterized by extensive subalpine forests of Engelmann spruce (Picea engelmannii Parry), subalpine fir (Abies lasiocarpa (Hook.) Nutt.) and, under certain conditions, lodgepole pine (Pinus contorta Dougl.). Although there has been commercial logging in the ESSF since the early decades of this century (Stettler, 1958), extensive harvesting of these forests in the province is a relatively recent phenomenon (Ministry of Forests, 1989). Many different cutting systems have been used to harvest subalpine forests. Chief among them are complete clearcutting, clearcutting with protection of advanced regeneration (trees that established naturally under a mature forest canopy), partial cutting to a diameter limit, clearcutting with protection of seed-trees, and selection cutting (sometimes called intermediate utilization) (Butt, 1989,1990; Mather, 1987; Stettler, 1958). In most cases, economic and operational factors have determined the use of a particular cutting system, but regeneration objectives have also influenced the choice of method. Until recently, naturally established trees were relied upon to replace forests after logging in the ESSF (Ministry of Forests, 1989). However, several factors have resulted in poor post-logging performance of advanced regeneration and low rates of establishment of naturals after harvesting. These factors include: 1) high rates of mortality in residual trees from wind and logging damage; 2) poor release growth in 2 suppressed residuals of poor quality, and; 3) post-logging site conditions that are unsuitable for spruce regeneration (Mather, 1987; Stettler, 1958). Also, on many cutovers (cutover is used in this thesis as a general term for any type of harvested site) too few stems have remained after logging to form a new stand, and most residuals have been of subalpine fir which is considered to be a less valuable species than spruce (however, Weetman and Vyse (1990) point out that this assessment of the relative value of fir is questionable as both species have similar wood qualities). Although residual trees, which vary widely in age, do respond to release after logging, their growth rates generally do not satisfy silvicultural expectations (Butt, 1989, 1990; David, 1987; Herring and McMinn, 1980; Mather, 1987; Stettler, 1958). Individual responses are typically highly variable, and despite repeated attempts it is still difficult to predict with any certainty when good release will occur (Ministry of Forests, 1989; Stettler, 1958; Weetman and Vyse, 1990). Several analyses of the problem of release growth have emphasized the importance of assessing prior to harvest the amount and quality of advanced regeneration in a stand, the amount of damage this regeneration is likely to sustain from the logging operations, and the suitability of post-logging site conditions for tree establishment and growth (Herring and McMinn, 1980; Mather, 1987; Stettler, 1958). After harvesting in the ESSF, the rate of invasion and the subsequent growth of Engelmann spruce are very slow relative to rates for residual trees. Herring and McMinn (1980) compared the height growth of naturally established seedlings to that of advanced regeneration 21 years after release and found that increments were much greater in the latter. Insufficient disturbance, particularly on winter-logged sites (where logging on top of the snow pack minimizes the disturbance of understorey vegetation) may be a major factor limiting the invasion of conifers. However, other factors such as low seed availability may be important for Engelmann spruce regeneration because this species is selectively removed from sites during harvesting (Mather, 1987). The slow regeneration . • ~ of spruce after logging in the ESSF has resulted in extensive areas that currently do not meet provincial reforestation objectives for stocking (stems per hectare) or annual height increments. Consequently, there has been a shift to planting seedlings (primarily of Engelmann spruce) on cutovers in the last decade or so, and clearcutting with planting (and under some conditions, the use of fire or mechanical scarification to prepare sites for planting) is currently the recommended method of regenerating old-growth forests in the ESSF (Butt, 1989; Mather, 1987; Ministry of Forests, 1989; Weetman and Vyse, 1990). Although planting is currently the preferred method of re-establishing spruce on logged ESSF sites, the performance of planted seedlings is almost as variable as that of naturals (Butt, 1989; Mather, 1987; Vyse, 1981). Although detailed studies are lacking, general comparisons of the relative performance of naturals and planted seedlings have indicated better early growth in the latter (Butt, 1990; Mather, 1987; Vyse, 1981). Despite this, the available data suggest poor growth and survival of Engelmann spruce in ESSF plantations of the south-central interior. For example, Mather (1987) found that the average survival rate in thirty-five spruce plantations in this region was 45%, and trees were typically growing less than 5 cm per year for five to eight years after planting. One particular factor that has been associated with low plantation success and the large backlog of unsatisfactorily restocked cutovers in the ESSF is the delay (> two years) in planting after harvesting which enables non-crop vegetation to become well-established. The poor performance of planted spruce has been attributed variously to competition, frost damage, cattle trampling, over-winter stress, soil moisture and nutrient deficits, and planting of inappropriate stock-types (Burdett etal., 1984; Butt, 1990; Coates, 1987; Mather, 1987; Vyse, 1981). 4 To date, planting has not solved the difficulties of regenerating cutovers in the ESSF, as several recent problem analyses on this topic indicate (Butt, 1989,1990; Butt et al, 1989; Mather, 1987). While it is already a major component of the harvest in some parts of the province, the forests of the ESSF will become increasingly important as the supply of merchantable timber at lower elevations is progressively reduced (Ministry of Forests, 1989). The provincial Ministry of Forests is currently committed to a planting program (primarily with Engelmann spruce or its hybrid interior spruce (Picea engelmannii x P. glauca (Moench) Voss) on unsatisfactorily restocked cutovers in this zone, but the need for a better understanding of the factors affecting the early performance of planted seedlings has been recognized (Burdett et al, 1984; Butt, 1990; Mather, 1987; Vyse, 1981). This study of the ecology of planted Engelmann spruce seedlings on two cutovers in the ESSF was developed in response to this need. In this introductory chapter some aspects of the dynamics of subalpine forests relevant to spruce regeneration are discussed and the characteristics of post-logging plant communities in the ESSF are described. The research objectives are presented and the major ecological concepts used in the development of this research are briefly introduced. In the last section of the chapter the biology of Engelmann spruce is reviewed, with an emphasis on the information available on factors affecting spruce regeneration and seedling growth and survival. The Characteristics of Engelmann spruce - Subalpine fir Forests The research presented in this thesis was carried out on two adjacent four-year-old (in 1987) cutovers and a nearby old-growth (> 300 years) spruce-fir stand in the ESSFwc2 near Vavenby, British Columbia (Figure 1.1). General characteristics of the overstorey vegetation on the cutovers (which will also be referred to as sites 1 and 2 or the Vavenby sites throughout the thesis) and in the old-growth forest are presented in Table 1.1. The ESSFwc2 (the Monashee Wet Cold Engelmann spruce - Subalpine fir variant of the ESSF) occurs in the wet belt of the Shuswap Highlands in the southern interior of the province. The variant is characterized by a moist, snowy interior climate, with a mean annual precipitation of 1177 cm that includes a 782 cm snowfall, a mean growing season temperature of 11.6 °C, a short frost-free period, an annual average of 672 growing degree days (> 5 °C), and a mean minimum January temperature of -13.8 °C. Most parent materials in this variant are morainal blankets. Overlying soils are typically humo-ferric or ferro-humic podzols, and acidic (with a pH of 4.4) with a hemimor or hemihumimor humus averaging 6.8 cm in depth (Lloyd et al, 1990). The forests of the ESSFwc2 are characterized by mature open stands of Engelmann spruce and subalpine fir. Lodgepole pine and successional stands are rare within this variant. The site index at 100 years for spruce and fir in the ESSFwc2 are 25 m and 22 m, respectively, and it is among the most productive areas for conifer growth in this zone in B.C. The sites selected for this study were in midslope positions, with mesic to submesic moisture regimes and slopes — 30%. The site unit was classified as 01 (zonal) - Bl-Azalea-Oak fern according to Lloyd et al. (1990). Forest understories in the ESSFwc2 are dominated by the ericaceous shrubs Rhododendron albiflorum Hook, (white-flowering rhododendron), Menziesia ferruginea Smith (false azalea), and Vaccinium membranaceum Dougl. (black huckleberry) (scientific nomenclature follows Hitchcock and Cronquist (1973) unless otherwise noted). The well-developed herb layer contains Gymnocarpium dryopteris (L.) Newm. (oak fern), Valeriana sitchensis Bong. (Sitka valerian), Rubuspedatus J.E. Smith (five-6 Figure 1.1 Location of the study sites in the Engelmann spruce - Subalpine fir biogeoclimatic zone. Source: Meidinger and Pojar, 1991. 7 Table 1.1 General characteristics of Engelmann spruce (Picea engebnannii Parry) and subalpine fir (Abies lasiocarpa (Hook.) Nutt.} overstorey on a young cutover and in an old-growth forest in the ESSFwc2. Characteristic Cutover Spruce Fir Forest Spruce Fir Age of dominant trees (years) Height of dominant trees (m) Diameter of dominant trees (cm) Number of trees per ha. > 1.3 m Number of trees per ha. < 1.3 m Total % cover (both species combined) 150-175 100-125 12-14 10-12 20-25 15-20 275-300 200-220 320 390 1350 3420 10 >30 >60 410 860 >30 >40 820 1690 55 * Cutover was site 1 of the Vavenby sites. The data are from five 0 . 0 2 ha sample plots per site. leaved bramble), Streptopus roseus Michx. (rosey twisted stalk), and Tiarella unifoliata Hook, (one-leaved foamflower - nomenclature follows Angove and Bancroft, 1983). The moss layer is less developed, and is dominated by red-stemmed feathermoss (Pleurozium schreberi (Brid.) Mitt. - Vitt et al, 1983) (Lloyd etal, 1990). The understorey species of the ESSFwc2 are common components of Engelmann spruce - subalpine fir forests throughout much of the range of this forest type (Alexander and Sheppard, 1984; David, 1987; Day, 1972; Oosting and Reed, 1952). According to the prelogging statistical data on the cutovers (from the Site History Files of the Clearwater Forest District Office, B.C. Ministry of Forests) harvested stands were mature with dominant spruce 180-305 years old, 28.6-35.4 m in height and 35.3-66.3 cm in diameter. The fir were 143-220 years old, with heights of 25.9-28.9 m and diameters of 32.8-40.9 cm . These values are typical for dominant trees in mature subalpine spruce-fir forests (Barnes, 1937; David, 1987; Oosting and Reed, 1952). Natural forests of Engelmann spruce and subalpine fir are typically all-aged or irregular-aged (Barnes, 1937; Day, 1972; Lowdermilk, 1952; Oosting and Reed, 1952; Veblin, 1986). However, there is great variation in age structures across the range of this forest type (Alexander and Shepperd, 1984; Whipple and Dix, 1979). Spruce frequently dominates the overstorey, contributing approximately 80% of the stand's basal area compared to 20% for subalpine fir (Barnes, 1937; David, 1987; Lowdermilk, 1925). However, this ratio is generally reversed in the understorey, with fir comprising 80% of the regeneration (stems) and spruce only 20% (Day, 1972; Lowdermilk, 1925; Veblin, 1986). The widespread observation of the dominance of fir in the understorey regeneration of spruce-fir stands has led to the prediction that vegetation development over time should 9 result in a shift to extensive fir-dominated forests. However, neither the shifts nor the expected fir stands have been observed in northern forests (Veblin, 1986; Whipple and Dix, 1979). Hypotheses about why such a successional pattern does not occur in spruce-fir forests can be classified into two major groups, which Aplet et al. (1988) have called the nonequilibrium and the equilibrium coexistence models of stand dynamics. In the former, stand dynamics in subalpine spruce-fir forests are considered to be controlled by catastrophic events, typically fire, that periodically re-initiate stand development and thereby disrupt the competitive exclusion of Engelmann spruce by subalpine fir (Bloomberg, 1953; Day, 1972; Romme and Knight, 1981; Stahelin, 1943). Stahelin (1943), for example, argued that fire severity determined the length of time to stand re-establishment after burning. He observed that spruce-fir stands destroyed by a light fire were replaced by shrubs and lodgepole pine, but were re-invaded rapidly (i.e. within a few years) by spruce and fir. In contrast, a severe fire gave rise to a grass and forb cover which was difficult to replace, and which effectively retarded stand development for up to 300 years. Similarly, Romme and Knight (1981) explained differences in stand development in subalpine forests of the central Rocky Mountains by a lower fire frequency and more favourable conditions for growth in drainages, which resulted in a more rapid and persistent development to a spruce-fir stand, in contrast to the adjacent slopes and ridges where development was greatly delayed by a serai lodgepole pine stand. Both Stahelin (1943) and Romme and Knight (1981) considered the spruce-fir forest to be the climatic climax stand in the subalpine forests of the central Rocky Mountains. However, Day (1972) argued that, for subalpine forests of the northern Rocky Mountains, the absence of fire would result in the replacement of a spruce-dominated stand by a fir-dominated one. Johnson and Fryer (1989) reconstructed stand development in northern Rocky Mountain forests of Engelmann spruce and lodgepole pine. In the stands they examined most 10 spruce recruitment took place in the first decade after fire, with low and sporadic understorey recruitment occurring for the next 200 years. The authors concluded that neither spruce nor pine were self-replacing in these forests (mortality was greater than recruitment) and that stand dynamics were controlled by fire frequency. In the second model of stand dynamics (the equilibrium model), spruce-fir stands are considered to be climatic climaxes capable of self-perpetuation without fire, through the colonization of windfall gaps by conifers and the differences in life histories (early survival and adult mortality) of spruce and fir (Barnes, 1937; Oosting and Reed, 1952; Veblin, 1986). Barnes (1937) argued, based on increment rates in naturally released trees as well as the rates of natural regeneration and the higher rates of adult mortality in fir, that a residual spruce-fir stand would develop within 120 years into a stand similar in structure and volume to natural spruce-fir stands of comparable age. Veblin (1986) demonstrated that coexistence of spruce and fir was not dependent on large-scale disturbances such as fire but could also be explained by differences in the life histories of the two species. Subalpine fir has a greater ability to establish on forest litter and at low light levels than does spruce due to its larger taproot and its higher photosynthetic rate (Knapp and Smith, 1982). Because of its greater abundance, fir is recruited into the canopy more frequently than spruce after tree falls (Veblin, 1986). However, higher adult mortality rates in fir (due to lower longevity and greater vulnerability to disease) reduce its frequency to that observed in natural stands. Aplet et al. (1988) recently suggested that stand development in spruce-fir forests in Colorado involved processes of both the equilibrium and nonequilibrium models of stand dynamics. They argued that patterns of stand development were dependent on interactions between a broad range of disturbance intensities and the life histories of spruce and fir. The stands they studied were all initiated by fires (as in the 11 nonequilibrium model) that were followed by the establishment of spruce and fir, with the latter dominating total biomass early in development when spruce recruitment decreased but fir recruitment continued (Aplet et al, 1989). By the third century after disturbance, fir began to die out and spruce began to re-establish in understorey gaps, and dominance in biomass shifted to spruce in the overstorey. According to Aplet et al (1989), a successful spruce cohort late in stand development insured the coexistence of the two species and was consistent with the equilibrium model of spruce-fir dynamics. As the above discussion shows, the conditions under which Engelmann spruce regenerates range from those created by large-scale catastrophic disturbances (e.g. fire) to those that exist within the understories of mature forests (which may or may not involve small-scale disturbances). One other quite different set of conditions under which spruce is recruited occurs at the forest/meadow ecotone at the upper elevations of the subalpine forest zone (Pattern, 1963; Schimpf etal, 1980). Numerous studies have investigated the factors affecting tree invasion into subalpine meadows, which has been observed to occur with a large amount of temporal and spatial variation (Franklin et al, 1971; Schimpf et al, 1980). Franklin et al (1971) suggested that historically, periods of climatic warming that increased temperatures and shortened the length of time the ground was covered with snow favoured the establishment of invading trees. They found that while tree invasion rates were lower within communities dominated by tall herbs such as Valeriana sichensis (a dominant species on the Vavenby sites) than in communities dominated by shrubs or low herbs, the growth rates of established trees were highest in the tall herb communities. Aplet et al's (1988) model of the development of spruce-fir forests is consistent with what is known about the forests of the ESSFwc2. There is a history of fire in these forests as indicated by the presence of charcoal in at least one location on the Vavenby sites and in the old-growth stand. However, the great age of the trees in the undisturbed forests in 12 the study area, the absence of lodgepole pine and successional stands in the variant, and the wet climatic conditions, all suggest that the fire frequency is low, and that stand _ development is also influenced by establishment in, and recruitment of trees from, the forest understorey. The factors controlling regeneration may differ significantly between the different environments within which spruce establishes. Factors influencing seed availability (e.g. production, dispersal, predation) may be more important in controlling the regeneration of spruce after large-scale disturbances, whereas factors influencing early seedling establishment (e.g. light and moisture availability, diseases, herbivory) may be more important in the understorey of spruce-fir stands. Winter-logging with the protection of advanced regeneration (as was done on the Vavenby sites) results in a very heterogeneous environment that is distinctly different from a large-scale burn, a mature forest, or a forest/meadow ecotone, although it shares some characteristics with each of these environments. In the following section these characteristics are described for post-logging communities in the ESSF. The Characteristics of Post-logging Plant Communities in the ESSF The Vavenby sites were clearcut with protection of advanced regeneration in January 1983 (Plate 1.1). Volumes of wood removed by this method range from 207 m-fyha to 460 m3/ha (Barnes, 1937; David, 1987; USDA 1965). The protection of advanced regeneration, defined as trees below 20 cm dbh, was intended to avoid the problem of the slow re-establishment of conifers after logging. However, high post-logging mortality, poor release growth, and slow ingress of conifers resulted in sites which were dominated by non-crop species four years after logging. During harvesting, trees were hand-felled by chainsaws, yarded to landings by tractor skidders, delimbed and loaded onto trucks for 13 Plate 1.1 General view of study site 2 from the landing at the base of the site. A mature Engelmann spruce - subalpine fir stand was winter-logged in 1983 with protection of trees less than 20 cm dbh. The site was classified as not satisfactorily restocked (NSR) with conifers in 1986 and is typical of NSR cutovers in the ESSFwc2. This photo was taken in 1989. 14 transport to the mill. Landings and skid trails comprise approximately 20% of the total area of the Vavenby sites, a typical value for winter-logged sites that are skidder yarded in the ESSF (Smith and Wass, 1985). Post-harvesting activity on the sites was restricted to the burning of logging residue on landings. Winter-logging in the ESSF results in little disturbance of the forest floor, except on landings and skid trails, where disturbance is usually severe enough to expose mineral soil and compact surface soil layers (Carr, 1987; Smith and Wass, 1985). The lack of disturbance off skid trails restricts opportunities for recruitment from seed banks (Lowdermilk, 1925). The persistence of seeds of most of the major ESSF species in the forest floor or mineral soil is unknown, however Lowdermilk (1925) reported that Engelmann spruce seed was not stored in forest floor material. Strickler and Edgerton (1976) studied seedling emergence from litter and mineral soil in three coniferous stands containing five conifers and numerous shrub and herb species. In two of the stands, Engelmann spruce was a principal overstorey component. After one year of incubation in the greenhouse, no conifer seedlings had emerged from the soil, less than 2% emerged seedlings were woody, and 73% of all emergent individuals belonged to the two species, Cystopteris fragilis (L.) Bernh. in Schrader (fragile fern - nomenclature follows Klinka et al, 1989) and Epilobium watsonii Barbey (Watson's willow-herb). The rate of invasion of Engelmann spruce after logging is slow (Lowdermilk, 1925; Noble and Alexander, 1977; Stahelin, 1943) even after extremely high seedfalls (Ronco and Noble, 1971). In the southern range of the ESSF factors limiting establishment include desiccation, unfavourable seedbeds, exposure to high light intensities, drying winds, clipping by birds, and frost heaving (Noble and Alexander, 1977; Ronco and Noble, 1971). Noble and Alexander (1977) reported a first year survival rate of less than 2% for Engelmann spruce on both northern and southern aspects at high elevations in the 15 central Rocky Mountains. Smith and Clark (1960) reported survival after six years in central British Columbia as varying with substrate type, from <1% on moss and litter to 5-6% on mineral soil, decayed wood and burned plots. Stahelin (1943) also found low stocking rates of both Engelmann spruce and subalpine fir 50 to 60 years after burning at high elevations in Colorado. He reported that stocking levels were higher on northern aspects and under Vaccinium cover than on other aspects or in areas with grass cover. In general, the development of vegetation following logging in northern forests exhibits patterns consistent with Egler's (1954) initial floristic composition model of succession (Bormann and Likens, 1979; Eis, 1981; Halpern, 1988; MacMahon, 1979; Mueggler, 1965; Stickney, 1986). In Egler's model, patterns of succession are most strongly influenced by the propagules (e.g. buried seeds, dormant buds, etc.) of species resident on a site at the time of disturbance. Changes in communities over time are a consequence of shifts in dominance among these initial species, rather than of species replacements (as in Clements' (1928) classical relay floristics model, as discussed by Egler (1954)). MacMahon (1979), in a discussion on succession in subalpine forests in Utah, suggested that patterns of community development in these forests were consistent with Egler's initial floristic composition model. Early in succession, communities were characterized by species exhibiting a broad range of life history strategies (sensu Grime, 1979), whereas in later stages, this range was greatly reduced. Community development, however, did not involve a change in species composition over time but rather a gradual elimination of some species and an increase in the dominance of others, all of which were present from the early stages of development. In contrast to the successional pattern described by MacMahon (1979), Schimpf et al. (1980) concluded that vegetation development from subalpine meadow to spruce-fir forests in northern Utah was facilitated by the initial colonization of the meadows by 16 aspen (Populus tremuloides Michx.) and lodgepole pine. These species caused environmental changes that improved water relations for invading conifers and in the absence of pine or aspen, spruce and fir established extremely slowly. The authors noted that this pattern of vegetation development was not consistent with an initial floristics model of succession. This suggests that one general model of community development is insufficient to describe the range of successional patterns that occurs in subalpine forests. At the start of this study, little information was available on the characteristics of post-logging plant communities in the ESSF in British Columbia, although there was considerable interest in the impacts of post-logging vegetation on conifer regeneration (Conard, 1984; Haeussler and Coates, 1986). David (1987) studied the response of subalpine fir to the increases in available nutrients that occurred after logging on a chronosequence of cutovers in the ESSFwc2. He concluded that the ability of fir to fully utilize the higher levels of nutrients was limited by some factor and speculated that this factor might be competition between non-coniferous species and residual subalpine fir for nitrogen. He suggested that the relatively greater competitive ability of the non-coniferous vegetation for available nutrients may have beee responsible for the increased presence of this vegetation on older sites (8- and 11-years-old) in the chronosequence. Despite a general perception (among operational foresters) that non-coniferous vegetation was severely reducing the performance of planted spruce in the ESSFwc2, prior to this research there had been no studies of the composition of post-logging communities to identify the major species dominating sites after harvesting. Consequently, as an initial step in the development of this study of the impacts of non-coniferous vegetation on planted seedlings, a survey of young cutovers (1 to 19 years) in the ESSFwc2 was carried out in 1986 to describe the characteristics of the major post-logging plant communities. Forty-four cutovers varying in age (i.e. time since logging), slope, elevation, exposure, and logging and site treatment histories were selected in the Clearwater and Salmon Arm Forest Districts of south-central B.C. Most of the cutovers were classified by the Ministry of Forests as not satisfactorily restocked with conifers at the time of the survey. On each cutover, the structure and composition of plant communities were described, noting dominant species (by % cover) and variations in distributions across the site. The study in its entirety was reported in Caza and Kimmins (1987) but some of the results are briefly reviewed here because they influenced the development of this research. The results of the survey showed no clear relationships between plant community characteristics and most site factors with the exception of time-of-year of logging. Winter-logged sites typically had vegetation dominated by several ericaceous shrub species including Rhododendron albiflorum, Menziesia ferruginea, and Vaccinium membranaceum. In contrast, summer- or spring-logged sites typically had vegetation dominated by various herb species, the major ones being Valeriana sitchensis, Epilobium angustifolium L. (fireweed), Veratrum virideAit. (false hellebore), Tiarella unifoliata, Streptopus roseus, and Gymnocarpium dryopteris. These differences between sites were related to the amount of understorey disturbance associated with logging at different times of the year (stump height was a useful general indicator of the intensity of disturbance). Winter-logging minimized the disturbance of the original forest understorey which was typically dominated by the ericaceous shrubs. Logging without a snowpack to protect the understorey vegetation resulted in much more disturbance and the re-establishment of a predominantly herbaceous plant cover. However, the amount of disturbance also varied greatly within sites, as most were harvested with the protection of advanced regeneration. Consequently, the vegetation on many sites was a mosaic of patches dominated by the ericaceous shrub species and the perennial herbs, with numerous skid trails interspersed among the patches. All of the major species on the cutovers were common components of the understories of mature spruce-fir stands in the study area (Lloyd et al, 1990). The similarity in species composition between young _ cutovers and mature spruce-fir stands suggests that the initial floristic composition model is applicable to early vegetation development on subalpine sites in the ESSFwc2. The patchy nature of the vegetation in post-logging communities in the ESSFwc2 was described quantitatively in a more detailed study of three similar and typical three-year-old cutovers, which were selected from the forty-four surveyed sites. Two of these cutovers were the Vavenby sites used for the planted spruce and invasion experiments described in later chapters. The third cutover was dropped from the study after 1986 because neither the time nor the resources were available to replicate the experiments on three sites. Within each site, information on species composition and cover were collected in quadrats along transects and analysed with multivariate cluster analysis to investigate species groupings across sites. The resulting dendrogram is presented in Appendix 2. The analysis indicated the presence of three major types of patches on the sites: 1) patches on or adjacent to skid trails, characterized by low to high covers of various herbs, particularly Epilobium angustifolium (referred to hereafter as skid trail patches, these covered approximately 32% of the sampled area - Plate 1.2); 2) patches off skid trails dominated by medium to high covers of the ericaceous shrubs Menziesia ferruginea and Rhododendron albiflorum (referred to hereafter as shrub patches, these covered 30% of the sampled area - Plate 1.3), and; 3) patches off skid trails dominated by high covers of the herbs Valeriana sitchensis, Gymnocarpium dryopteris, and Thalictrum occidentale Gray (western meadowrue) (referred to hereafter as herb patches, these covered 23% of the sampled area - Plate 1.4). There was also a distinct but infrequent (covering only 9% of the sampled area) patch type dominated by residuals or advanced regeneration of Engelmann spruce and, more commonly, subalpine fir. The vegetation in 19 Plate 1.2 Vegetation dominating a skid trail patch six years after logging on study site 1. Major species included Epilobiwn angustifoliutn (fireweed) and Valeriana sitchensis (Sitka valerian). Veratrum viride (false hellebore) and Athyrium filix-femina (ladyfern) were also present on this skid trail. The woman kneeling on the ground is holding a i m stick which indicates the average height of the vegetation. 20 Plate 1.3 Vegetation dominating a shrub patch six years after logging on study site 1. Major species included Menziesia ferruginea (false azalea) and Rhododendron albiflorum (white-flowered rhododendron). The woman in the photograph is holding a i m stick which indicates the average height of the vegetation. 21 Plate 1.4 Vegetation dominating a herb patch six years after logging on study site 2. Major species included Valeriana sitchensis (Sitka valerian), Thalictrum occidentale (western meadow-rue) and Arnica lalifolia (mountain arnica). The woman in the photograph is holding a i m stick which indicates the average height of the vegetation. approximately 6% of the quadrats contained both herbs and shrubs. The variability in the distribution of the patch types within sites was evident in the different clustering positions of consecutive quadrats (as illustrated in Appendix 2 for a series of quadrats on site 1). It is the patterns of disturbance associated with winter- and diameter-limit logging that have largely determined the pattern of herb, shrub, and skid trail patches across the Vavenby sites, although some of this mosaic is a remnant of the distribution of species in the original stand. The performance of planted or naturally established spruce seedlings will be strongly influenced by the characteristics of these patches, because these characteristics determine both the level of available resources and the biotic environment within which seedlings must compete for these resources. The many ways in which between- and within-patch dynamics can structure plant communities, and the role of disturbance in these dynamics were reviewed by Pickett and White (1985) in a collection of papers on this topic. Among the population and community characteristics that can be significantly influenced by patch variation are nutrient cycling, microclimate, water relations, heterozygosity and fitness, and species coexistence (various papers in Pickett and White, 1985). With regards to species coexistence, Denslow (1985) pointed out that species with the traits necessary to exploit patches that are forming at historic rates will persist within communities when the disturbance pattern resembles that which is historically characteristic of the community. When the characteristics of the disturbance are altered (e.g. by changes in rate, intensity, or impact) the characteristics of patches may also change such that they are no longer suitable for some species. If the alterations are slight (as might be predicted after logging on the Vavenby sites) the result may only be shifts in the relative abundance of species, rather than extinctions from the community. 23 Research Objectives Patches can be described in terms of extent, composition, rates of change, and internal heterogeneity. Differences among patches in these characteristics may cause differences in the level of resources, the survival of residual organisms, or the rate of invasion and establishment of new species (Pickett and White, 1985). In this study, the characteristics of the patch types on the Vavenby sites were described, and the relative performance of planted Engelmann spruce seedlings within these patches was studied and compared to the characteristics of naturally established seedlings in various environments to explore the relationship between within-patch dynamics and the conditions for conifer regeneration on cutovers in the ESSFwc2. The practical objective of the research was to study the growth and survival of planted Engelmann spruce seedlings in relation to the environmental heterogeneity that typically occurs on unsatisfactorily restocked cutovers in the ESSFwc2 to determine whether vegetation management is essential, beneficial, or even appropriate for improving spruce regeneration in these areas. The following four research questions were developed to address this general objective, and for each question a specific research objective was formulated. 1. Do the patch types on the Vavenby sites differ significantly as environments for Engelmann spruce regeneration? Objective 1. To describe the biotic, edaphic, and microclimatic characteristics of the herb, shrub, and skid trail patches on the Vavenby sites (Chapter 2). 2. Does the growth and survival of planted Engelmann spruce seedlings differ significantly among the patch types either with or without interference from non-coniferous vegetation? 24 Objective 2. To compare the early growth and survival of planted spruce seedlings of two contrasting stock-types within and between patch types when vegetation is ~ undisturbed and when it is removed above-ground. (Chapter 3). 3. Do the growth characteristics of planted Engelmann spruce seedlings and naturally established seedlings differ significantly in relation to each other or to their environments? Objective 3. To compare variation in the early growth of planted spruce seedlings within patches to that of naturally established seedlings of a similar age on a cutover and in the understorey of an old-growth spruce-fir stand (Chapter 4). 4. Do the plant species within the patch types differ significantly in their responses to different levels of disturbance? Objective 4. To describe the rates and patterns of re-establishment by plant species within herb and shrub patches in response to the removal of above-ground vegetation and the total removal of vegetation and litter (Chapter 5). In Chapter 6, the results presented in the previous four chapters are integrated in a discussion of the ecology and management of spruce regeneration on logged sites in the ESSF. To address the research objectives stated above, two experiments were set up on the Vavenby sites. These are described in Chapters 3 and 5. Both experiments involved the application of treatments to plots within the three patch types (two in the case of the disturbance experiment). These plots were not located randomly but were located according to plant composition criteria that defined the three patches based on the results of the cluster analysis described earlier in this chapter. Treatments, however, were allocated randomly among the plots in each patch type. Natural densities of neighbours were not manipulated in this study, for several reasons. Firstly, defining and maintaining meaningful levels of competitors within the complex patches would have been very achieve only crudely defined levels of control. Thirdly, the interest in this study was focused primarily on the variation in seedling response to natural variation in cover and composition, with the total removal of vegetation included for the purpose of comparison with responses to this variation. The planted spruce seedlings came from two stock-types grown commercially for outplanting in the study area. The planting density in the plots on the Vavenby sites was higher than that used operationally (the research spacing was 0.5 m whereas operational spacing is approximately 3 m) because the average maximum patch size was only approximately 25 m^  and the research planting distance was determined by the number of replicate seedlings desired in each plot. In summary, the experimental approach used in this research reflected both the operational and scientific objectives of the research, as well as the constraints imposed by the characteristics of the communities being studied. Ecological Concepts The objective of this research was developed from a silvicultural problem - the problem of poor spruce regeneration at high elevations in the southern interior of British Columbia. Because the problem was taken up by an ecologist rather than a silviculturalist, the approach that was adopted was a study of the variation in the early growth and survival of planted spruce in relation to variation in the environmental factors influencing seedling establishment. This approach does not have its origin in any single ecological concept, and it is unsuitable for the testing of any single ecological idea. Like most applied problems in ecology, it touches on many of the concepts underlying current ecological theory. Several of these concepts are described briefly here. They were used in the development of the research objectives stated above, and were also used to explore the results of the research presented in later chapters (and are discussed further in those chapters). The characteristics of the plant communities on the Vavenby sites within which spruce is regenerating are influenced by logging disturbances that remove part of the forest understorey. Natural large-scale disturbances such as fire, and small-scale disturbances such as single tree mortality, are both important and ubiquitous processes in the regeneration of forest species (Oliver and Larson, 1990; West etal, 1981). Disturbance is also an important process in the creation and maintenance of resource and species patchiness within many types of plant communities where differences in species interactions between- and within-patches determine community structure (Pickett and White, 1985). The structure of the communities on the Vavenby sites has already been described in terms of herb, shrub, and skid trail (and residual tree) patches. Spruce seedling establishment is influenced by the processes within these patches, and if establishment differs among patches, spruce regeneration will also be influenced by between-patch processes that determine the relative abundance and distribution of patches across sites. Although many processes may influence species interactions within patches, it was hypothesized that competition for space is the major one affecting patch dynamics on the Vavenby sites. Under such conditions, it is the dominance relationships among species that determine patch structure (Yodzis, 1986). As Grime (1979) discusses for tree seedlings invading herbaceous vegetation, these dominance relationships can have significant impacts on tree regeneration. In this study, the concepts of patch dynamics and dominance-controlled communities were used to interprete the response of patch types to logging and post-logging disturbances, and the implications of these .„ responses for spruce regeneration on ESSF cutovers. In this study it was assumed that seedlings planted within the herb, shrub, and skid trail patches on the Vavenby sites would experience the effects of direct or indirect competition from neighbouring herbs and shrubs. The removal of neighbours causes changes in resource availability and in the relative importance of factors limiting growth and survival. The concept of resource allocation strategies in plants suggests that patterns of carbon allocation within seedlings will change in response to changes in relative resource availability. Individual plant components (e.g. roots, shoots) may experience different intensities of competition for resources, or in the absence of competition may experience different levels of resource availability (Chapin etal., 1987; Wilson, 1988a). According to the current resource allocation paradigm, internal resources will be preferentially allocated to that plant component (e.g. needles, stems, roots) acquiring the limiting external resource (Chapin etal., 1987; Tilman, 1988; Wilson, 1988b). Tilman (1988) used a simulation model ALLOCATE to demonstrate how allocation to roots decreased but allocation to stems increased along a gradient from nutrient-poor to nutrient-rich soils. His results are consistent with observations from other studies of lower shoot:root biomass ratios when water or mineral nutrients are limiting and higher shoot:root ratios when nutrients are high and/or light levels are low (Tilman, 1988). However, there are several reasons to approach the interpretation of patterns of biomass allocation within seedlings cautiously. Patterns of biomass, energy, and nutrient allocation may differ both within and among species (Abrahamson and Caswell, 1982) and higher turnover rates in roots than stems means that standing crop does not necessarily reflect energy investments. In addition, roots function in support and storage as well as in resource acquisition, and at any given time allocation patterns may reflect the dominance of these other functions. Another factor influencing the interpretation of allocation patterns in tree seedlings is species longevity: the resource limiting growth may vary spatially and seasonally, but allocation strategies must reflect the precedence of long term survival over short term production gains (Kurz, 1989). There have been no studies of allocation strategies in Engelmann spruce. In this study, patterns in the allocation of biomass and size in seedlings growing with and without competitors was interpreted in terms of differences in resource availability among environments, and the significance of patterns for the regeneration strategy of spruce. The concept of the regeneration niche (Grubb, 1977) provides a perspective for viewing the problem of planted seedling establishment in terms of the regeneration strategy of spruce. Grubb introduced the concept to help explain how high species diversity could be maintained in plant communities when there was so little obvious differentiation in the habitat niches of most species. Grubb defined the regeneration niche in terms of a species' requirements for the successful replacement of one mature individual by a new mature individual within the community. The responses of planted seedlings can be interpreted in terms of their likely contribution towards this replacement process. However, many factors affecting the performance of planted seedlings may play little part in the natural regeneration process. For this reason, a comparison between naturally established seedlings and planted stock-types of spruce was included in this study. Nevertheless, it seems reasonable to expect that the responses of planted seedlings to environmental variation can be interpreted from the point of view of the regeneration strategy of spruce in subalpine environments. One limitation to the interpretation of seedling responses from the perspective of the regeneration niche is the long period of 29 establishment in Engelmann spruce. The replacement of one mature individual by another is a process spanning decades. However, rates of mortality are high in young Engelmann spruce, supporting the assumption that it is the very early stages of establishment that are most critical to the regeneration process. The final section of this chapter is a review of the information available on factors affecting the growth and survival of young Engelmann spruce in subalpine environments. A Review of the Biology of Engelmann spruce (with emphasis on the factors affecting early growth and survival) Engelmann spruce is a major component of high elevation forests from the Pacific Coast in British Columbia east to the Rocky Mountains of Alberta and south to New Mexico and Arizona (USDA, 1965). The species is monecious, with female strobili borne in the upper crown and male strobili in the lower crown on branchlets (Alexander and Shepperd, 1984). Pollen is wind-dispersed from mid-June to early July at higher elevations and cones mature by August or September of the same year. Cones and seed are shed from September throughout the winter, although most seed is released by the end of October. Seed production can begin when trees are only 15 years old, but it peaks in dominants 150-250 years old (Alexander and Shepperd, 1984). Engelmann spruce is a long-lived species, with decay-free individuals ranging in age up to approximately 350 years in old-growth forests (Barnes, 1937). Some spruce seed is produced in most years, but good crops occur with a reported periodicity of 2-5 years, and year-to-year variation is large (Alexander and Shepperd, 1984; Lowdermilk, 1925). Seeds are small and wind-dispersed, with good viability except at timberline where conditions for growth are harsh (Tranquillini, 1979). Dispersal distance is short (only 120-180 m) and although germination capacity is high relative to other associated conifers, germination under 3 0 natural conditions is typically low (Alexander and Shepperd (1984) reported values of 0 to 28%). Cone and seed insects, and small mammals cause annual losses in seed crops „but little quantitative data are available. Alexander and Shepperd (1984) reported that an average of 28% of the total seed production was lost to insects over a 4-year period in one study in Colorado. Seeds that survive over the winter germinate after snowmelt the following spring. Although Engelmann spruce can germinate and establish on a variety of substrates in undisturbed forests (Alexander and Shepperd, 1984), seedlings establish best where litter layers are thin, or removed to expose mineral soil (Lowdermilk, 1925; Noble and Alexander, 1977), or on decayed wood (Lowdermilk, 1925; Smith and Clark, 1960). Knapp and Smith (1982) found significant microsite differences in the understorey locations of Engelmann spruce and subalpine fir in subalpine forests in southern Wyoming. Spruce seedlings were less frequent in general, but were proportionally more abundant in areas with thinner litter layers. Moisture may be the major factor limiting establishment on other types of substrates, such as moss and thick litter (Noble and Alexander, 1977; Smith and Clark, 1960). Knapp and Smith (1982) showed that under laboratory conditions Engelmann spruce germinants had a lower rate of root growth and a less penetrating tap root than subalpine fir and they suggested that spruce was restricted to understorey microsites that did not dry out early in the summer. Knapp and Smith (1982) also found that subalpine fir seedlings had higher photosynthetic rates than spruce at lower understorey light levels. They concluded, however, that light was not a major factor limiting spruce establishment because seedlings of both species occurred in locations with equally low levels of solar radiation. Alexander and Shepperd (1984) reported that, although spruce germinates'at all light intensities under natural conditions, light levels 40% to 60% of full sunlight are most favourable for establishment of seedlings at high elevations due to the deleterious effects of high light intensities at high altitudes. Ronco's (1970) observations of solarization damage on seedlings at high elevations in the southern Rocky Mountains suggests that this may be the case for more southerly populations of spruce, although interpretation of his results is difficult because of confounding effects from transplanting seedlings acclimated to a low-elevation nursery onto a high-elevation research site. Regardless, it is doubtful that at higher latitudes (where spruce occurs at lower elevations and skies are often cloudy) trees frequently, or for prolonged periods, experience light levels high enough to cause direct damage to seedlings. Spruce seedlings can grow under relatively low light levels, with reported compensation points ranging from 60 p.mol/m^ /s in southern Wyoming (Knapp and Smith, 1982 - temperature not reported) to 75 u,mol/m^ /s in south-central B.C. (Karakatsoulis, 1989 - unpublished data, temperature 20 °C). Saturation points for seedlings in the Wyoming and B.C. studies were 1060 umol/m /^s and 500 umol/m /^s, respectively. This large difference may reflect the much higher light intensities that typically occur at high altitudes in the south. The influence of high levels of solar radiation on other environmental factors, such as relative humidity and soil surface temperatures, has a more significant effect on spruce seedling growth and survival than direct effects of radiation on seedlings (Carter et al, 1988; Kaufmann, 1979; Noble and Alexander, 1977). According to Alexander and Shepperd (1984), Engelmann spruce is restricted to cold humid habitats because of a low tolerance to high temperatures and low soil moisture. Drought is a major cause of seedling mortality in Engelmann spruce in the southern part of its range, preventing establishment after harvesting on southern aspects (Noble and Alexander, 1977). Studies under controlled conditions have shown reductions in both photosynthesis and stomatal conductance in response to soil moisture deficits (Kaufmann, 1976; Roncq, 1970). However, according to Lloyd et al. (1990) growing season moisture deficits are uncommon in the ESSF in the southern interior of B.C. because of high moisture inputs from snowmelt and frequent summer storms. In subalpine environments, low soil and air temperatures present greater constraints to tree growth and survival than high temperatures (Larcher, 1985; Tranquillini, 1979; Tranquillini and Havrarek, 1985; Wardle, 1968). Optimum air and soil temperatures for the growth of Engelmann spruce under laboratory conditions are much higher than values typically found in the subalpine forest (Hellmers etal., 1970; Shepperd, 1981; Tinus, 1984) which led Tranquillini (1979) to suggest that although Engelmann spruce grows naturally in cold climates, it is not well adapted to the temperatures of its natural range. However, the validity of this assumption is challenged by the results of several recent studies on the ecophysiology of spruce (Day et al., 1989, 1990). Day et al. (1990) found that photosynthetic rates in young Engelmann spruce were higher under some conditions in trees growing in cold soils (< 3.5 °C) associated with a snowpack than in trees growing in warmer (> 10 °C) soils. They suggested several factors which might be causing these differences, including greater root growth due to higher nutrient availability during snowmelt, higher rates of photosynthesis due to higher light levels above the snowpack due to the reflection of light off the surface of the snow, and lateral root growth from the colder soils beneath the snowbanks into warmer soils in snow-free areas. They quantified the higher light levels and the lateral root growth, but did not investigate the changes in nutrient levels associated with snowmelt. They did point out, however, that considerable root growth has been found to occur under snow and during snowmelt in other western conifers (Grier et al, 1981). Summer frost is characteristic of the ESSF. Frosts can occur at any time during the growing season (Lloyd et al, 1990) and frost damage to planted seedlings-can be a serious problem on some sites (Stathers, 1989). DeLucia and Smith (1987) found that photosynthesis in Engelmann spruce trees in southern Wyoming was most strongly affected by night-time frosts up until mid-June. Light frosts (-2.5 °Q caused only a slight ~ and reversible reduction in photosynthetic rate and stomatal conductance, whereas a hard frost (-5.0 °C) caused an irreversible reduction in photosynthetic efficiency by increasing internal resistence to CC«2 diffusion. A hard frost in early June caused a 25-35% reduction in photosynthetic rates that persisted throughout the growing season, and a midsummer hard frost reduced rates by 80-90%. Midsummer frosts were rare, however, and after the high risk frost period in the spring photosynthesis was most strongly correlated with soil temperatures. Low soil temperatures have been associated with increased root resistance to water flux in Engelmann spruce, resulting in reduced stomatal conductance (Carter et al, 1988; Kaufmann, 1979). In addition to low soil temperatures, stomatal conductance is also sensitive to leaf-to-air vapour deficits (Day et al, 1989; Kaufmann, 1976). However, other mechanisms in addition to stomatal conductance may be involved in the observed reduction in photosynthesis with low soil temperatures (Turner and Tranquillini, 1985). DeLucia (1986, as cited in DeLucia and Smith, 1987) has suggested that these mechanisms might include the inhibition of electron transportation and the activity of RUBP carboxylase, with effects on nutrient uptake, hormone production and changes in carbohydrate source-sink relationships. Day et al (1990) found no evidence for low soil or air temperature effects on stomatal limitations to photosynthesis, and suggested that such effects in other studies may have been due to the use of unhardened or non-acclimatized seedlings, which are much more sensitive to low temperature effects. Smith et al. (1987) suggested that cloudcover, cold air and soil temperatures, leaf-to-1 air vapour deficits, and soil moisture were the major factors, in decreasing order of importance, limiting photosynthetic carbon assimilation by subalpine conifers in the 34 central Rocky Mountains. Although the slow growth of spruce at high elevations can be at least partially explained by reductions in photosynthesis caused by these factors, the inconclusive results of the few studies that have been done to date indicate that the physiological processes underlying spruce performance in subalpine environments are still not well-understood. According to Alexander (1987), no information is available on the nutrient requirements of Engelmann spruce. The few studies of factors affecting the performance of planted and naturally established seedlings in British Columbia suggest that the poor growth observed under some conditions after burning or mechanical scarification may be due to reductions in the nutritional status of soils caused by the removal of surface organic layers (Burdett et al, 1984; Herring and McMinn, 1980; Mather, 1987). Significant seedling responses (in both growth and mortality) to fertilization have also been observed, suggesting that nutrient availability might be an important factor influencing early seedling establishment (Burdett et al, 1984; Lloyd, 1991 - unpublished data), but this aspect of spruce seedling biology remains uninvestigated. The survival and growth of Engelmann spruce in subalpine environments can be reduced by late-lying snow and winter desiccation or 'frost-drought' (Tranquillini, 1979). Adverse effects on spruce seedlings caused by late-lying snow are a reduction in the short growing season by delayed photosynthesis and needle death or damage from snow moulds (Marr, 1977; Tranquillini, 1979; Wardle, 1968). Wardle (1968) noted extensive damage to spruce foliage by the snow moulds Herpotrichia nigra Hartig and Phoma sp.. Marr (1977) concluded that the extent of snow mould damage was a major determinant of branch mortality, influencing the movement of 'islands' of Engelmann spruce and subalpine fir at treeline. Wardle (1968) found that late-lying snow in gullies or depressions delayed the growth of spruce seedlings for up to a month in comparison with 35 trees in more exposed conditions. With increasing altitude, wind becomes an important determinant of the length of the snow-free period. A Rhododendron ferrugineum community in a sheltered gully remained snow-covered until mid-June in the Austrian Alps, whereas an adjacent heath community on a rise was snow-free for all but a few weeks during the winter (Tranquillini, 1979). Snow-cover can be beneficial by protecting needles from winter desiccation. Frost-drought is a commonly observed phenomenon in subalpine environments and may be a major determinant of treeline (Tranquillini, 1979). Although usually confined to trees at timberline, frost-drought can occur in Engelmann spruce below timberline under certain fall and winter conditions (Wardle, 1968). Winter death of spruce needles can also be caused by cuticle abrasion by windborne ice crystals, and is associated with a reduction of cuticular resistance in wind-exposed needles (Hadley and Smith, 1986). Tranquillini (1979) hypothesized that needle desiccation is a consequence of thin cuticles and inadequate cuticular resistance to transpiration by needles insufficiently matured during the short subalpine summer. In addition, restricted water uptake from cold or frozen soils may limit the plant's ability to meet a high transpiration demand for water caused by the drying effects of wind, resulting in high needle mortality. In concluding this section, it should be noted that most of the information available on Engelmann spruce regeneration is very general in nature, and is based on studies with populations in the southern and central Rocky Mountains. These studies have shown that the production of seed is highly variable, dispersal is limited, and seed survival and germination are generally low, but the factors controlling these stages in spruce regeneration are largely unknown. There are very few studies of requirements for early (< five years) seedling establishment, or of seedling responses to environmental variation. Snow may be a very important factor limiting establishment, particularly because seedlings remain snow-covered for almost as long as snow persists on a subalpine site, but its role in early survival and growth is also unknown. Studies with mostly larger trees „have shown that a number of factors can influence the physiological performance of spruce during the growing season, but not only does the relative importance of these factors vary spatially and temporally in subalpine forests, their overall impacts on spruce growth and survival have not been studied. In this study, variations in the growth of seedlings rather than in physiological responses were related to the characteristics of environments for spruce establishment on subalpine cutovers. In the following chapter patterns of variation in the edaphic and microclimatic characteristics of these environments are described for the herb, shrub, and skid trail patches on the Vavenby sites. CHAPTER 2 THE ENVIRONMENTAL CHARACTERISTICS OF HERB, SHRUB, AND SKID TRAIL PATCHES ON TWO CUTOVERS IN THE ESSFwc2 INTRODUCTION Winter-logging in the ESSFwc2 creates a mosaic of herb-dominated and shrub-dominated patches of vegetation on sites with many skid trails that vary greatly in their properties. Patches of herbs and shrubs in undisturbed areas are remnants from the original stand understorey. Other patches are created by the removal of trees and the destruction of vegetation that Occurs during harvesting operations. The properties of the soils and vegetation that dominate herb, shrub, and skid trail patches affect the level of resources, the survival of remaining plants, and rates of invasion and success of establishment of new plants within each patch type (Pickett and White, 1985). These properties also affect the survival and growth of tree seedlings planted into the patches, and some knowledge of them is necessary to interpret seedling responses to the environments created by logging and silvicultural treatments. From the limited information available (which is reviewed below) there is evidence to suggest that herb-dominated, shrub-dominated, and skid trail patches on ESSFwc2 cutovers differ as environments for tree seedling growth and survival. To characterize the level of resource availability for seedlings after logging but prior to the application of silvicultural treatments, a descriptive study of the microclimatic and edaphic characteristics of the three patch types was undertaken. 38 Literature Review .. w Most of the research on factors affecting the growth and survival of subalpine conifers has focused on the physiological responses of established trees to variations in microclimate (Carter et al, 1988; Day et al, 1989; DeLucia and Smith, 1987; Kaufman, 1975; Tranquillini, 1979; Turner and Tranquillini, 1984). The results of this research were reviewed in Chapter 1. Studies have shown that light intensity, soil and air temperatures, soil moisture, and snowpack properties are important factors influencing the performance of Engelmann spruce in subalpine forests. The relative importance of these and other factors to seedling growth on subalpine cutovers under the conditions of interference from non-coniferous species that exist on the Vavenby sites is not known. Yet all of these factors are influenced by the characteristics of the dominant vegetation on a site - regardless of whether this vegetation is the trees in a forest or the herbs and shrubs on a cutover. Results from other studies suggest that differences in the dominant vegetation and intensity of disturbance between the three patch types on the Vavenby sites affect light availability (Wagner, 1989); soil and air temperatures (Ballard, 1972; Coates, 1987; Stathers, 1989); snowpack depth (Tranquillini, 1979); nutrient availability (Yarie, 1978); and soil physical properties (Smith and Wass, 1985). These factors affect conifer seedling growth and survival, and differences in these variables between the patch types may have significant effects on seedling performance (Carr, 1987; Coates, 1987; Stathers, 1989; Wagner, 1989). Although few studies have focused on the differences between herbaceous and shrubby vegetation as environments for seedling growth, Wagner (1989) noted that both woody and herbaceous vegetation typically occur in various proportions in young forest plantations and understanding their relative competitive impacts on seedlings is important to developing effective management strategies. 39 Wagner (1989) studied the effects of herbs and shrubs on the growth of Douglas-fir ™ seedlings on coastal clearcuts in Oregon. Light levels at 0.5 m and 1.0 m above-ground were lower in shrub canopies than herb canopies, and levels significantly decreased with increasing shrub density, but were unaffected by increasing herb density. Unlike Douglas-fir, Engelmann spruce and its hybrids with white spruce are typically less than 50 cm tall for the first few years after planting on subalpine cutovers (Bassman, 1989; Burdett et al, 1984; Coates, 1987) and seedlings can be overtopped by herb canopies during this period. Light levels within a dense herb canopy that overtops seedlings may be low. Coates (1987) found that the light intensity at ground level beneath a mixed herb-shrub cover on a sunny day on a ESSFwc2 cutover was only 13.5% (135 + 89 umol/m /^s) of full sunlight. Much lower levels have been measured beneath dense canopies of herbs common on subalpine cutovers. Karakatsoulis (1989 - unpublished data) measured light levels between 2 and 5 % of full sunlight on the ground beneath canopies of Epilobium angustifolium (fireweed) and Athyrium filix-femina (L.) Roth. (ladyfern) on a subalpine cutover in the Nelson Forest Region. Wagner (1989) also observed that herbs had a greater effect than shrubs on water potentials down to a depth of 0.6 m, although woody plant density was negatively correlated with soil water potential down to a depth of 1.2 m. Although Engelmann spruce is sensitive to water deficits (Kaufmann, 1975; Ronco, 1970) soil drought is uncommon on submesic and wetter sites in the ESSFwc2 (Lloyd et al, 1990). Coates (1987) found no differences in the moisture content of soils with different levels of vegetation removal on an ESSFwc2 cutover and no differences in predawn or midday . - xylem water potentials in Engelmann spruce seedlings growing with and without competing vegetation on the site. * Wagner (1989) concluded from his study that shrubs had a 10 times greater competitive impact on seedling height than herbs in the first three seasons after planting due to the effects of shrubs on both light and moisture availability. However, shrubs had only 1.6 times the impact of herbs and low-growing shrubs on the basal area growth of seedlings. Wagner suggested that interference from the lower vegetation was just beginning to affect seedling growth after three seasons and that these effects would have an increasingly important impact on subsequent seedling performance. Soil temperatures have been shown to vary with the type of vegetation (Ballard, 1972) and the intensity of disturbance (Coates, 1987) on subalpine sites in British Columbia. Ballard (1972) studied the effect of above-ground vegetation and substrate characteristics on soil temperature on a site in the Mountain Hemlock zone in southwestern B.C. Temperatures at 60 cm (Ballard compared values at this depth to reduce the effects of interference from transient weather conditions on values nearer the surface ) were highest in bare ground where maximum values reached 15 °C in mid-July. Temperatures were lower beneath evergreen shrubs and lower still beneath a herbaceous meadow. The lowest temperatures occurred beneath a clump of subalpine fir trees where maximum values reached 6 °C in late August. Differences in temperatures were associated with differences in the type of ground cover and the amount of organic matter in the soil, both of which influenced soil thermal properties. For example, damping depths (the reduction in amplitude of temperature with increasing soil depth) were greatest in the bare ground and smallest beneath the herb community. The small damping depth in the meadow soil was tentatively attributed to the high organic content of a deep Ah horizon present at this location. On a cutover in the ESSFwc2, Coates (1987) found that removal of a mixed herb-shrub cover without removal of the humus layer resulted in soil temperatures at 10 cm that were 4 to 7 degrees higher than temperatures beneath undisturbed vegetation. Temperatures beneath the undisturbed vegetation ranged from 6 to 8 °C over the growing season and did not exceed 10 °C. Scarification increased mean soil temperatures only slightly above those in areas where vegetation was removed, but greatly increased the diurnal amplitude. Ballard (1972) also found that the highest diurnal amplitude in temperature occurred in bare soil. Coates (1987) found little effect on minimum air temperatures when above-ground vegetation was completely removed on his study site. The presence of vegetation did reduce maximum air temperatures below those over bare soil. Differences in the height and density of vegetation canopies may influence the frequency or intensity of summer frost. Coates (1987) observed frost damage on spruce seedlings during the first growing season on his site in the ESSFwc2. The amount of damage decreased from a treatment with no shrubs and 25% herb cover, to 100% cover of herbs and shrubs, to no herbs and 25% shrub cover, to a treatment with 50% cover of herbs and shrubs. The least damage occurred in the treatments with complete removal of vegetation. Low dense canopies, such as those formed by grasses, pose a significant frost risk to seedlings on frost-prone sites because they trap cold and stagnant air and allow it to cool radiatively to subzero temperatures (Stathers, 1989). The dense herb canopies on the Vavenby sites may behave in a similar manner. In contrast, the taller canopies of shrubs can reduce radiative heat loss from the ground by replacing the cold night sky with warmer foliage. As frosts can occur at any time during the growing season in the ESSFwc2 (Lloyd et al, 1990), the impact of vegetation canopies on air temperature may affect the growth and survival of planted seedlings. Rates of snowmelt may also differ between the three patch types. One hypothesis proposed to explain the characteristic clumping of trees in subalpine meadows is the _ favourable effects of increased rates of snowmelt from canopy drip and the radiation of heat from tree stems on seedling establishment and growth (Kimmins, 1987). Areas adjacent to tree clumps become snow-free earlier than treeless areas. The ericaceous shrubs on ESSF cutovers may have properties similar to these tree clumps; snow-melt may occur sooner in shrub patches than herb patches, and shrubs may thereby provide better environments for establishment than herbs. Tranquillini (1979) noted that small differences in the length of the snow-free period can have a big impact on plants at high elevations. Evans and Fonda (1990) found such differences were strongly associated with the distribution of herb and shrub communities in subalpine meadows in the North Cascades. Differences in the total quantity and relative availability of soil nutrients may exist between the three patch types on the Vavenby sites. Yarie (1978) studied the role of understorey vegetation in the nutrient cycle of three plant associations on xeric, mesic, and hygric sites in Mountain Hemlock forests in southwestern B.C. The dominant understorey species on the xeric and hygric sites were also dominants in the shrub and herb patches on the Vavenby sites: on the xeric site these were the shrubs Rhododendron albiflorum and several Vaccinium species; on the hygric site they were the herbs Valeriana sitchensis, Arnica latifolia Bong, (mountain arnica), Veratrum viride, Rubus pedatus, and several Streptopus species (a list of all the vascular plants on the Vavenby sites is given in Appendix 1). Yarie found significant differences between sites in annual litterfall production and in the nutrient contents of the major species on each site. Understorey litterfall was two-thirds of above-ground production in the herb-dominated (hygric) association, compared to only one-third in the shrub-dominated (xeric) association. Similarly, nutrient concentrations were highest in the herbs Tiarella unifoliata, Rubuspedatus, and Streptopus roseus, and lowest in the shrubs Rhododendron albiflorum and the Vaccinium species. Within individual species values declined from hygric to mesic to xeric sites. Yarie found that the understorey returned a significant proportion of litterfall nutrients to the soil each year, mainly in a single pulse during the first autumn snowfall. His results suggested that cycling rates were more rapid where herb litter inputs were higher and where moisture levels were most favourable for decomposition. Differences in nutrient availability between herb-dominated and shrub-dominated vegetation in the ESSFwc2 may be enhanced during the nutrient flush (called the assart effect) that commonly occurs with changes in site conditions after logging. David (1987) documented this assart effect on a winter-logged site with residuals in the ESSFwc2 and found it was characterized by a general increase in available nutrients, particularly nitrogen, which peaked between three and six years after logging. Factors influencing this flush of nutrients may be increased organic matter inputs to the system from harvesting disturbances, more favourable post-logging conditions for decomposition, and decreased water and nutrient uptake by trees. Differences between the herb, shrub, and skid trail patches in litter characteristics, soil and air temperatures, and in soil moisture may produce significant within-site variation in the assart effect. Soil physical properties, particularly bulk density, may differ significantly between the skid trail and off-skid trail patches. Soil compaction is a common consequence of skidder yarding, even on winter-logged sites where a protective snow cover decreases disturbance of the understorey (Smith and Wass, 1985). Typically, compaction is most severe on landings. Carr (1987) found bulk densities on landings on winter-logged sites in the Sub-Boreal Spruce zone to exceed 1400 kg/m^ (levels above which the root growth of loblolly pine, radiata pine, and Douglas-fir are impaired (Carr, 1987)). On 6-year-old and 11-year-old cutovers, Carr found the height growth of lodgepole pine seedlings on landings was significantly less than that of seedlings in off-landing areas. He attributed the differences to increased resistance to root penetration and lower soil aeration in the compacted soils on landings. Smith and Wass (1985) also found increased bulk densities on winter skid trails on ESSF sites in the Nelson Forest Region, with values at 0-10 cm ranging from 1180 to 1530 kg/m3 in mid-trail locations (values in undisturbed soils ranged from 1010 to 1330 kg/m3 on these sites). Smith and Wass (1985) found no statistically significant differences in the total nitrogen and organic carbon between the top 20 cm of soil on skid trails and the top 20 cm of undisturbed soils on a cutover in the ESSFwc, although levels were consistently higher on the skid trails. Carr (1987) did not find differences in either nitrogen or potassium levels between landing and off-landing soils on his sites, but phosphorous levels were significantly lower on the landings. Both studies reported poorer seedling growth on the highly disturbed substrates, but Carr did not find that the lower height growth on landings was associated with lower foliar nutrient concentrations in seedlings. The studies reviewed above suggested that differences in the vegetation and disturbance history of the three patch types on the Vavenby sites might be associated with differences in light, air and soil temperatures, soil organic matter and nutrient content, and soil bulk density and that these differences would be important to tree seedling growth and survival. Although a detailed study of each was not possible, some description of these properties was required to characterize resource availability within each patch type and to assist in the interpretation of the planted spruce seedling responses to the patch environments, as well as the response of the patches to disturbance. It is recognized that growth is an integrated response to a seedling's total environment and simple growth-resource level correlations are inadequate to understand the relationships among the factors influencing seedling responses to environments. Nevertheless, strong — growth responses in particular directions will be suggested by the levels and relative importance of the environmental factors that determine seedling performance. In this chapter, these physical, chemical, and biological properties are described for undisturbed patches on the Vavenby sites. The effects of post-logging disturbances on these properties are described in later chapters. METHODS Definitions of herb-dominated and shrub-dominated patches were based on the results of the cluster analysis of species data discussed in Chapter 1 and Appendix 1. Prior to the removal of vegetation from and the planting of spruce seedlings into plots in the three patch types for the studies described in following chapters the composition and percent cover (estimated visually) of species in each plot was recorded. Measurements of microclimate and sampling for soil physical and chemical properties were also carried out within plots or in adjacent areas using the methods described below. Soil profiles One shallow pit of 20 cm depth was excavated adjacent to each plot to assess the similarity of the soil profiles beneath the herb, shrub, and skid trail vegetation. Skid trail soils typically lacked well-defined profiles near the surface, and they exhibited the characteristics of disturbed substrates: they were of mixed materials and variable composition. A detailed description of skid trail soils was not undertaken because of the large variability among locations, and the difficulty in interpreting the characteristics of these soils in meaningful terms. However, soils beneath herb-dominated and shrub-dominated vegetation had distinct profile patterns. To describe these in more detail, one soil pit (approximately 80 cm deep) was excavated next to a herb plot and one next to a shrub plot on study site 1. The type, depth, and texture of mineral soil horizons and parent material were described using The Canadian System of Soil Classification (Canada Soil Survey Committee, 1978). Humus forms were identified using the 47 Taxonomic Classification of Humus Forms in Ecosystems of British Columbia (Klinka et al, 1981). Air and soil temperatures Soil temperatures were monitored continuously on site 1 from July 6th 1989 to June 5th 1990 using plastic-shielded combination thermister-thermocouple sensors buried at 5, 10, 20, and 50 cm depths in three herb-dominated locations, three shrub-dominated locations, and at two locations on skid trails (in an area with vegetation, and in an area without it). At one herb and one shrub location, temperatures at the four depths were monitored continuously from July 1987 to June 1990. Daily maxima, minima, and averages were recorded with a CR10 Datalogger (Campbell Scientific Inc., Edmonton, Alberta). Air temperatures were also recorded by a datalogger during 1989 and 1990 with shielded thermisters positioned 15 cm above-ground at one location each within herb and shrub patches, and from 1987 to 1990 by thermisters at 15 cm and 1.2 m in open locations (i.e. unshaded) on a skid trail site 1. Soil physical properties For the determination of humus depth, soil particle size distribution (texture), bulk density, and soil moisture retention curves, soil samples were collected on each site from three herb plots and three shrub plots in the disturbance study (described in Chapter 5), and one skid trail plot in the planted spruce study. Samples were collected 1 m outside of randomly selected corners of plots. The number of samples collected varied with plot but only one sample per plot was analysed for soil physical properties. The depth of the humus layer was measured prior to soil sampling and then this layer was removed from an area aproximately 10 cm x 10 cm. Metal cylinders of two different - sizes were inserted into the soil, removed with a small shovel, and tightly wrapped in plastic until analysis. Samples from the smaller cylinders (diameter 5 cm) were analysed for particle size distribution (Day, in Black, 1965 - pipette method) and moisture retention (Richards, in Black, 1965). Larger cylinder (diameter 9.8 cm) samples were used in the determination of bulk density as described by Blake (in Black, 1965 - core method). Snow depth was measured with aim stick at four locations in two or three plots in each patch type on study site 1 on May 19th, 1989. Soil chemical properties Humus material and soil for the determination of carbon content were collected from the sampling locations described above. Soils were collected at 15 cm and 30 cm depths, dried at 70 °C for four days and hand-ground to a fine powder in a mortar. An approximately 5 cm depth of soil was sampled. Humus material was treated in the same way as mineral soil and was not separated into different layers prior to preparation for analysis. Samples of 0.075 g humus, and 0.05 g soil material were analysed for total carbon with a LECO Analyzer using the procedure described in the Methods Manual of the Pedology Laboratory of the University of British Columbia (Dept. of Soil Science, 1981). As an index of nitrogen availability, gross mineralizable nitrogen (N) was determined for soil and humus samples from herb and shrub plots on study site 2. The plots were established as part of a fertilization study which will not be discussed in the thesis (skid trails were not included in the study). The plots were located using the same criteria as those for the disturbance and planted spruce studies and were similar in composition to plots in these studies. Prior to fertilization, samples of humus and soil (the latter from 0-_10 cm depth) were collected adjacent to three randomly selected planted seedlings in six herb and six shrub plots. Samples were air-dried and sieved through a 2 mm mesh sieve. Samples of 2 g of mineral soil and 0.4 g of humus were mixed with water and anaerobically incubated for 1 week at 40 °G (Keeney in Page et al, 1982). Solutions were filtered and extracted with 2 M KCL, and analysed for ammonium with a Techicon Autoanalyzer (modified from Waring and Bremner, 1964). Light Light levels within herb-dominated, shrub-dominated, and skid trail vegetation were measured in July 1988 at a time of estimated maximum leaf development. Measurements were taken on both sites at five randomly-selected planted spruce seedlings in three control plots within each type of vegetation (see Chapter 3 for a description of the planted spruce study). Instantaneous levels of PAR (photosynthetically active radiation) were recorded in u-mol/m^ /sec with a LI-COR quantum sensor and LI-1000 Datalogger (LI-COR Inc., Lincoln, Nebraska) at 0, 25, 50,100, and 200 cm above-ground. Readings were taken between 1130 h and 1730 h on site 1 under variable but mostly cloudy sky conditions. Readings were taken three days later on site 2 between 1400 h and 1630 h under cloudy conditions. During the following two days the LI-1000 Datalogger was set up on site 1 for continuous readings of PAR. Two LI-COR quantum sensors were fixed at a height of 15 cm at two locations within herb-dominated vegetation. Similarly, two sensors were placed at 15 cm within shrub-dominated vegetation. Light levels were measured every minute and average, maximum, and minimum values were recorded each hour for two consecutive days under variable but mostly cloudy sky conditions. 50 RESULTS The species composition of patches The average abundances (expressed as percent cover) of the major species in each of the three patch types are presented in Table 2.1. These abundances reflect the criteria used in the selection of plot locations for the planted spruce study. Herb plots were dominated by Valeriana sitchensis (64% cover) with a significant component of Thalictrum occidentale (12% cover). Shrub plots were dominated by Menziesia ferruginea (75% cover) with appreciable cover of Rhododendron albiflorum (22%), and to a lesser extent Gymnocarpium dryopteris (11%). The skid trail plots were also dominated by Valeriana sitchensis (52% cover) with a relatively high cover of Epilobium angustifolium (13%) and several other herbs. The species composition of the three patch types was similar with the exception that fewer species occurred in the shrub plots than in the herb and skid trail plots. However, the species in the shrub patches were also present in the other patch types. The shrubs were approximately twice as tall (209 cm) as the herbs (90 cm). Moss cover was much higher in shrub plots (with an average cover of 28%) than in herb plots (where the average cover was only 2%). Plant community differences between the three patch types are discussed in greater detail in Chapter 5. Soil profiles The soil pit profiles from beneath herb-dominated and shrub-dominated vegetation on site 1 are presented in Figure 2.1. The two profiles were markedly different, particularly in: 1) the presence of an Ahe layer beneath the herb vegetation in contrast to an Ae layer 51 Table 2.1 Composition of herb-dominated, shrub-dominated, and skid trail vegetation on two cutovers in the ESSF wc2. Herb Shrub Skid Trail Major species Mean (and standard deviation) % cover Menziesia ferruginea Smith 2 (1) 75 (15) <1 (<1) Rhododendron albiflorum Hook. <1 (<1) 22 (16) <1 (<1) Vaccinium membranaceum Dougl. 1(<1) 5 (8) <1 (<1) Vaccinium ovalifolium Smith <1 (<1) 4 (7) <1 (<1) Valeriana sitchensis Bong. 64 (12) 5 (3) 52 (12) Thalictrum occidentale Gray 12 (19) <1 (<1) 7 (9) Gymnocarpium dryopteris (L.) Newm. 2 6 (5) 11 (8) 5 (5) Rubus pedatus J.E. Smith 5 (7) 4 (2) 2 (1) Viola glabella Nutt. 5 (5) <1 (<1) 8 (8) Arnica latifolia Bong. 5 (6) <1 (1) 3 (4) Epilobium angustifolium L. 2 3 (6) <1 (<1) 13 (13) Streptopus roseus Michx. 2(<1) 1(<1) 2 (1) Clintonia uniflora (Schult.) Kunth. 2 (3) <1 (<1) <1(<1) Tiarella unifoliata Hook.^  1(<1) 1(<1) <1(<1) Mitella breweri Gray 1 (3) <1 (<1) 2 (3) Ribes lacustre (Pers.) Poir. 2 (2) <1 (1) <1 (<1) Rubus parviflorus Nutt.2 <1(<1) - 1(<1) Picea engelmannii Parry <1 (<1) 1 (1) <1 (<1) Abies lasiocarpa (Hook) Nutt. <1 (<1) <1 (1) 1(<1) Average maximum height4(cm) 90 (10) 209 (18) 75 Range in species number in plots 12-17 9-14 13-1 Moss cover 5 (% mean + S.D.) 2 ± 4 28 ±36 <5 (estimated)  (estimated) Based on n=12 18mz plots per patch type. * Nomenclature follows Hitchcock & Cronquist, 1973. 2 Measurements were made in June 1987, before maximum leaf development in these species. Their cover in mid-summer was approximately twice that in June (see Chapter 5). ' 3 Nomenclature follows Angove & Bancroft, 1983. 4 Height measurements were not taken prior to treatment in the planted spruce plots. These measurements were made on plots in the disturbance experiment. 5 Moss cover values are also taken from the disturbance plots (prior to treatment). The most abundant moss on the study sites was Pleurozium schreberi (Brid.) Mitt, (red-stemmed feathermoss). 52 Figure 2.1 Soil profiles beneath herb-dominated and shrub-dominated vegetation on study site 1. soil depth (cm) HERB SHRUB Horizon L F H Ahe - sandy loam Bf loam IIC - fine sand Horizon L F H *Ae (or Aeg) Bhf - fine sand sandy loam II C - sandy loam Characteristics of profiles Soil depth jParent material Humus type Humus depth Coarse fragment Rooting depth Herb 43 cm glaciofluvial deposit (well sorted) mormoder 8 cm 15% 31 cm Shrub 63 cm morainal till (compacted, unsorted) hemimor 6 cm 30% 45 cm beneath the shrubs; 2) the finer glaciofluvial deposit beneath the herbs in comparison with the coarser morainal till underlying the shrubs, and; 3) a mormoder humus form under the herbs and a hemimor humus form beneath the shrubs. In general, the profiles indicate a higher organic matter content in soils beneath the herbaceous vegetation and a coarser soil texture beneath the shrubs (possibly due to differences in parent materials). Air and soil temperatures Average air temperatures from July to the end of October 1989 at 15 cm in the open and at 15 cm within herb-dominated and shrub-dominated vegetation are presented in Figure 2.2. Within the herb canopy daily averages ranged from 7 to 15 °C between early July and late August. During this period average temperatures within the shrub canopy ranged from 8 to 18 °C. Above 10 °C temperatures were typically 1 to 4 degrees lower beneath the herbs than the shrubs and temperatures in the open were similar to those within the herb canopy. The highest maximum temperature was 27 °C (not shown in Figure 2.2 as values are averages only); it was recorded within the herb canopy on July 27th. During the general decline in air temperatures that began in mid-August, temperatures were highest beneath the shrubs. Daily averages less than 0 °C did not occur at any location until late September. During the period July 6, 1989 to June 5, 1990 minimum temperatures within the herb and shrub canopies dropped below freezing after the first week of September. The lowest recorded mimimum temperature was -11 °C beneath the herbs on October 15th. 18 | II I I I I I I I I I I I | I I I I I I I I I I I I I | II I I I I I I I I I I I | I I I I I I I I I II I I | I I I I I I I I I I I I I | I I I I I I I I I I I M | I I M I I I I I I I I I | I I I I I I I I I I I I I | I I I I I JUL06 JUL20 AUG03 AUG17 AUG31 SEP14 SEP28 0CT12 OCT26 Date Figure 2.2 Average daily air temperatures from July to October 1989 at 15 cm above-ground in a herb patch, a shrub patch, and in the open on study site 1. Snow covered the probes in early November, causing temperatures to rise to slightly above 0 °C where they stayed throughout the winter. In contrast, temperatures at 1.2 m in the open fell to below -25 °C during the 1989-90 winter. The probes 15 cm above-ground were released from snowcover at the end of May in 1990 and on June 5th, the last measurement day, a minimum temperature of -1 °C was recorded. Soil temperatures at depths of 5,10, 20, and 50 cm in one of the herb patches are shown in Figure 2.3. Patterns with depth were similar for all locations and are typified by this profile. During the snow-free period (late May to early November - Figure 2.3 does not show the period between May and July) temperatures decreased with increasing depth; during the winter months, this pattern was reversed. Temperatures at 5,10, and 20 cm closely followed each other as temperatures increased or decreased. Temperature changes were small at 50 cm and over the summer values ranged from 7 and 9 °C. Temperatures at all depths converged in mid-August and dropped rapidly together until diverging again in mid-September. Similarly, they converged in early May and rose rapidly together, diverging again at the end of the month. This pattern was characteristic of all the soil temperature profiles regardless of their location. Another shared characteristic was temperatures above 0 °C at all depths throughout the winter months. Differences in soil temperatures between the herb-dominated, shrub-dominated, and skid trail patches are illustrated by the patterns in average daily temperatures at 10 cm shown in Figures 2.4 and 2.5. Throughout the summer and fall temperatures were highest in the bare skid trail soil and lowest beneath the herb and shrub patches (Figure 2.4). From early July to the end of August, temperatures at 10 cm beneath the herbs and shrubs ranged from 7 to 11 °C: beneath the skid trail vegetation values ranged from 8 to 14 °C and in bare soil (on the skid trail) they ranged from 9 to 15 °C. Temperatures at all V 3 10 V a E V o JUL06 AUG15 SEP24 NOV03 DEC13 JAN22 MAR03 APR12 MAY22 • 5cm + 10cm o 20cm A 50cm Date Figure 23 Average daily soil temperatures from July 1989 to June 1990 at 4 depths in a herb patch on study site 1. tn o V I— D ra w U a E ai • • Shrub 10cm + Herb 10cm A VegSTIOcm o Bare ST 10cm I 'I I I I T J I I n I i - p M i l r i T I I I 11 | I 1 I I I I 1 I I I I I T | 1 I I T I I I I I I I I I J T I I I I I I I I I I I I 11 I I I I I I I I f l I I j I T I r I I I I T'l 1 I I 11 I T I T I I I I T T I I "j JUL06 JUL20 AUG03 AUG17 AUG31 SEP14 SEP28 0CT12 OCT26 Daie Figure 2.4 Average daily soil temperatures from July to October 1989 at 10 cm in shrub, herb, vegetated skid trail and unvegetated skid trail patches on study site 1. locations converged at the end of August, at which time the pattern reversed, with higher temperatures beneath the herb and shrub vegetation during the winter and lower temperatures on the skid trails. The pattern among locations was similar to the pattern among depths shown in Figure 2.3. After snow-melt in the spring soil temperatures rose rapidly at all locations but the rate was much higher on the skid trails than beneath the herbs and shrubs (Figure 2.5). Overall, annual temperatures were low and had a narrow amplitude - averages ranged from 0 to 16 °C on the skid trails and 1 to 12 °C off the skid trails. As with air temperatures, soil temperatures at 10 cm beneath the shrubs were typically higher than beneath the herbs. However, the differences were smaller than those above-ground and except during the spring soil warming were not greater than 1 °C: differences of 1 °C are within the measurement error of the sensors (Stathers, 1991). Soil physical properties The physical properties of the soils beneath the herb, shrub and skid trail vegetation are presented in Table 2.2. There were differences between the three patch types in humus depth, particle size distribution, bulk density and snow depth. The humus layer beneath the herbs was deeper (9.0 cm) than the layer beneath the shrubs (5.1 cm), as it was in the soil profiles presented in Figure 2.1. Also consistent with the profde descriptions, the soils beneath the shrubs had a higher coarse fragment content (31%) and lower silt content (28%) than soils beneath the herbs (which had a coarse fragment content of 10% and a silt content of 43%). Percentages of sand and clay were similar in soils from the two patch types. Differences in bulk densities between the patch types were consistent with differences in particle size distribution - bulk densities were greater o o_ o 3 10 i— V a. e V <** '5 11 -r 10 -9 -8 -7 -6 -5 -4 -3 -2 -- (—i—i—r i i j L i — i — r ~ T — i — i — | — r i i i r- i \ ] \ T i i n i j i r r APR01 APR09 APR17 APR25 MAY03 Date • Shrub 10cm + Herb 10cm A Veg ST 10cm o Baro ST 10cm i i | i i i—i i i i | i i i— r — n— i | i i i i—i i i MAY11 MAY19 MAY27 JUN04 Figure 2.5 Average daily soil temperatures from April to June 1990 at 10 cm in shrub, herb, vegetated skid trail and unvegetated skid trail patches on study site 1. Ul Table 2.2 Physical properties of soils beneath herb-dominated, shrub-dominated, and skid trail vegetation on two cutovers in the ESSF wc2. Herb Shrub Skid Trail Humus depth1 (cm) x±S.D. 9.0 ±1.5 5.1 ±2.5 7.2 ±2.3 n 3 3 3 Soil texture Class silty loam loam silty loam to to loam loam Sample depth x ± S.D. 11.6 ±6.1 7.8 ±3.2 8.0 ±2.0 n 6 6 2 Particle size distribution (%) coarse (>2 mm) 10.0 ±2.1 30.6 ±8.7 20.2 ± 19.2 sand 35.3 ±3.9 34.4 ±3.8 34.8 ±8.4 silt 43.4 ±1.3 28.0 ±2.0 38.0 ±9.7 clay 11.3 ±0.6 7.1 ±2.8 7.0 ±1.7 n 2 2 2 Bulk density2 x ± S.D. 649.8 ± 157.8 795.2 ± 180.3 865.0 ±93.3 (Kg/m3) n 6 6 2 Snow depth1 (cm) May 19/89 x ± S.D. 31.6 ± 20.9 28.2 ± 13.4 15.8 ± 4.0 n 8 8 8 1 Values are based on depth in planted spruce control plots (see Chapter 3) on study site 1. x = sample mean, S.D. = standard deviation, n = number of samples. 2 Values may be underestimated due to underestimation of coarse fragment content. in the shrub soils than in the herb soils, but within-location variation was high. Bulk densities were highest in skid trail soils, a probable consequence of soil compaction from logging. However, because of the small sample size it is not possible to conclude that densities were significantly greater in this location. Observations on snow depth and distribution on the study sites were limited because sites were not accessible until late in the spring. However, photos taken on site 1 on May 5th in 1988 show shrub patches free from snow while adjacent herb patches are still covered with snow (Plate 2.1). Snow remained on site 1 longer in 1989 than in 1988 and was still present in the three patch types on May 19th 1989. On that date levels were lower on the vegetated skid trails than in the herb and shrub patches (Table 2.2) and unvegetated areas on skid trails, which were mostly free of snow, were flooded by meltwater (Plate 2.2). Soil moisture retention curves exhibited patterns consistent with differences in other soil properties among the three patch types and were similar on both sites (Figure 2.6). Soils beneath the shrub vegetation had a lower moisture content at a given soil water tension than soils from the other two locations, reflecting the coarser texture of the shrub soils. At a soil water tension of -6 J/kg the volumetric water content of shrub soils averaged 42%, in comparison to 50% in the herb soils and 57% in the skid trail soils. Although differences between the herb and shrub soils were consistent over the range of tensions tested, within-area variation was large and standard deviations overlapped. Note that the error bars in Figure 2.6 are standard errors of the mean. Throughout the thesis standard errors of the mean are used as error bars in figures. However, in the text and in tables standard deviations are presented with means. 62 Plate 2.1 This photograph shows a snow-free shrub patch adjacent to a snow-covered herb patch on study site 1 on May 5th, 1988. General observations indicated that snow disappeared earlier in the spring from shrub patches than from herb patches. 63 Plate 2.2 This photograph shows a bare skid trail plot on study site 1 which is free of snow on May 19th, 1989 while the surrounding areas with vegetation are still covered with snow. Snow remained on the Vavenby sites approximately one week to ten days longer in 1989 than in 1988. In the spring surface runoff from meltwater temporarily flooded spruce seedlings planted into unvegetated areas on skid trails, as illustrated in this photograph. 6 4 1 10 100 1000 10000 Soil water tension (-J/kg) Figure 2.6 Water retention curves for soils from herb (•< ), shrub (#), and skid trail ( • ) patches (both sites combined). Means and standard error of the means are indicated by symbols and error bars. n=6 for herb and shrub patches, n=2 for skid trail patches. 65 Soil chemical properties Total carbon (%) values for soils from the three patch types are given in Table 2.3. The carbon content of humus varied little among the three areas but levels at 15 cm in mineral soil beneath the herbs were twice as high (8.8%) as levels beneath the shrubs (4.3%), with skid trail values intermediate between the two (5.2%). The carbon content at 30 cm in the herb soils (7.9%) and the skid trail soils (5.2%) were similar to those at 15 cm: in the shrub soil, levels decreased to 2.6%. These results indicated greater accumulation and distribution of organic matter in the soils beneath the herbs than in soils at the other two locations. Mineralizable nitrogen (N) concentrations in the humus and soils beneath the herb and shrub vegetation are presented in Table 2.4 (skid trails were not sampled for mineralizable N). As with total C, levels in the humus material were similar for the two complexes. The major difference between the soils was associated with the presence of a mineral soil horizon (an Ahe layer in some locations) high in organic matter beneath the herbs that was not found beneath the shrubs. However, this horizon was not universally present in herb soils - it was found in only 40% of the sampling locations. Levels of mineralizable N were high in this horizon; the average was 253 ppm, in comparison to 483 ppm in the humus layer, and 58 ppm in mineral soil at 10 cm beneath the herbs. When this horizon was not present beneath the herbs the soil at 5 cm was similar in appearance to the soil at 10 cm and only the greater depth was sampled. The average mineralizable N concentration at 10 cm beneath the herbs (58 ppm) was almost twice the level beneath the shrubs (31 ppm) - a difference similar in magnitude to that for total C at 15 cm beneath the two complexes. 66 Table 23 Total carbon in humus and soils beneath herb-dominated, shrub-dominated and skid trail vegetation on study site 1. Herb Shrub Skid Trail Humus layer1 x ± S.D. 35.7 ±7.5 26.4 ±7.0 29.8 ±11.1 (%) n 5 5 4 Soil 15 cm x ± S.D. 8.8 ± 1.8 4.3 ± 1.3 5.2 ± 2.4 (%) n 5 6 5 Soil 30 cm x±S.D. 7.9 ±2.0 2.6 ±0.7 5.2 ±3.8 (%) n 6 6 5 Table 2.4 Mineralizable nitrogen concentrations in humus and soils beneath herb-dominated and shrub-dominated vegetation on study site 2. Herb Shrub Humus layer (ppm) x ± S.D. n 482.7 ± 168.9 18 477.0 ± 120.44 17 Soil 5 cm2 (ppm) x ± S.D. n 252.8 ± 113.4 7 — Soil 10 cm (ppm) x ± S.D. n 57.7 ± 30.8 18 30.6 ± 13.3 17 x = sample mean, S.D. = standard deviation, n = number of samples. A mineral layer enriched with organic matter was present at this depth in 7/18 herb plots - it was not found in shrub plots. Light Levels of PAR (photosynthetically active radiation) at heights of 0, 25, 50, 100, and 200 cm above-ground in the herb, shrub, and skid trail vegetation are presented in Figure 2.7. Values at 2 m represent values in the open (i.e. full sunlight if conditions were sunny). Data were presented as absolute rather than relative values to facilitate comparisons with published values for compensation and saturation points in spruce. Standard deviations were generally large, and ranged from 23 to 204 % of the means. At heights less than 2 m the variation in light levels was lower in the shrub patches than in the herb or skid trail patches. At heights less than 25 cm light levels were low and similar within the three patch types; at heights above 25 cm levels were much lower beneath the shrubs than beneath the herbs or skid trail vegetation. On the ground and at 25 cm above-ground, light levels were higher in shrub patches that in herb patches, although the differences were not statistically significant (at a=0.05, as tested in a nested A N O V A model using S Y S T A T Version 5.0 (Wilkinson, 1990 -analyses not shown). Means and standard deviations for PAR at ground level were 9.08 + 4.5 umol/m^/sec beneath the herb canopy, 39.29 + 9.09 umol/m^/sec beneath the shrubs, and 74.4 + 151.61 urnol/m^/sec beneath the skid trail vegetation. These values represented 0.84%, 5%, and 6.6% of the levels at 2 m for the herbs, shrubs, and skid trail vegetation, respectively. The average light level at 25 cm within the herb canopy was 116.14 + 52.33 u-mol/m^/sec; within the shrub canopy at this height, the average was 146.55 + 52.62 urnol/m^/sec. The decrease in average light levels from 1 m to 2 m in skid trail vegetation was not significant and was due to variable sky conditions. Average hourly light levels at 15 cm beneath herb and shrub canopies are presented in Figure 2.8 for a two-day period in July 1988. Profiles are presented for one location 68 1500 u u t/3 \ CM a o a 1000 500 0.0 0.5 1.0 1.5 2.0 2.5 Height above ground (cm) Figure 2.7 Means and standard error of the means for light levels from 0 to 2 m above the ground in herb (4 ), shrub and skid trail ( • ) patches in July 1988 (both sites combined). n=30. 400 Herb patch ^r— Shrub patch 1400 2000 0200 0800 1400 2000 0200 0800 1400 Time of day (h) Figure 2.8 Average hourly light levels at 15 cm above the ground in herb and shrub patches from July 26th to 28th 1988 on study site 1. PAR = photosynthetically active radiation. 70 within each vegetation type but profiles at replicate locations were similar. The measurement period was too short to elucidate clear temporal patterns in PAR, but the available data indicated that under cloudy conditions light levels within the shrub 71 DISCUSSION Despite large among-sample variation in most physical and chemical properties, the three patch types on the Vavenby sites have edaphic and microclimatic characteristics that create, in combination with differences in vegetation, three distinctly different environments for plant growth. Below-ground differences in parent materials, and below- and above-ground differences in vegetation have influenced soil development in the herb and shrub patches. The soils beneath the ericaceous shrubs have a higher coarse fragment content, a lower moisture retention capacity under saturated conditions, and a higher bulk density than the soils beneath the herbs. The thinner hemimor humus beneath the shrubs, as well as the lower carbon content in shrub soils, suggests lower above- and below-ground organic matter inputs in the shrub patches compared to the herb patches. This may result from lower root mortality and lower rates of root turnover in shrubs than herbs as well as lower above-ground litter inputs (Grime, 1979). The bulk density of the skid trail soils was higher than that of the other two patch types, most likely as a consequence of the compaction associated with logging activities. The distribution of particle sizes in skid trail soils was intermediate between herb and shrub soils for most classes, reflecting the disturbance history of this substrate (skid trails were created in areas previously occupied by both herb and shrub patches). Light levels, and soil and air temperatures were all higher on skid trails than in the other two patch types, due to less live plant cover and litter cover in the skid trail environment. Are the differences between patch types likely to be significant to the performance of planted spruce seedlings? As the range of values for most variables was large compared to the mean, it is appropriate to question whether the differences discussed above are likely to be significant to the growth or survival of planted spruce seedlings. This question will be considered by reference to what is known about the relationships between spruce performance and environmental conditions. Soil temperatures beneath the shrubs and herbs ranged from about 7 to 11 °C in July and August. These values are similar to those recorded by Coates (1987) beneath mixed herb and shrub vegetation on a young cutover in the ESSFwc2. Although temperatures beneath the herbs and shrubs were generally above the 7 to 8 °C threshold several authors have suggested is important to photosynthesis and conductance in Engelmann spruce (Carter et al, 1988; Kaufmann, 1975), averages were rarely above 10 °C and were below the optimum values reported for growth in this species (Hellmers et al, 1970; Shepperd, 1981). This suggests that low temperatures may limit physiological performance throughout much of year (DeLucia and Smith, 1987). However, Grier et al (1981) found that root growth in amabilis fir (Abies amabilis (Dougl.) Forbes) in subalpine stands occurred at soil temperatures near 0 °C and that root growth was more or less independent of above-ground growth. The recent work of Day et al. (1990) on Engelmann spruce suggests that root growth in this species also occurs at temperatures well below 7 °C. Soil temperatures on the Vavenby sites were above 0 °C throughout the year and therefore may not represent a major limitation to seedling root growth on these sites. However, the possibility of low soil temperature effects on seedling physiology and the apparently poor root growth in planted seedlings during the first three years after outplanting (see Chapter 3) indicate a need for more detailed studies on low temperature effects on seedling growth. The higher soil temperatures beneath the skid trail vegetation (values were > 10 °C for most of the growing season) suggest a more favourable temperature regime for growth in this environment than in off-skid trail areas. Optimum air temperatures for Engelmann spruce growth under controlled conditions have been reported to be between 18 and 25 °C (Hellmers et al, 1970; Tinus, 1984). Although average air temperatures beneath the herbs and shrubs never reached 18 °C, maximum daily temperatures in July and August exceeded 18 °C on 41 days in the shrub canopy and 33 days in the herb canopy, suggesting that low air temperatures are likely not to be a major limitation to growth on the Vavenby sites. Similarly, minimum temperatures during the 1989 and 1990 growing seasons never reached the -2.5 °C threshold below which DeLucia and Smith (1987) observed reductions in photosynthesis in Engelmann spruce in subalpine forests in Wyoming. Temperatures on site 1 did not fall below -1 °C beneath vegetation during the 1989 or 1990 growing seasons, and there was no evidence of frost damage to seedlings during the first four growing seasons after planting. On the Vavenby sites light levels below 25 cm within the herb, shrub, and skid trail vegetation were lower than the 13.5% recorded by Coates (1987) beneath mixed herb and shrub vegetation in the ESSFwc2, but similar to the 2 to 5% measured by Karakatsoulis (1989 - unpublished data) beneath dense canopies of herbs and shrubs on other sites in the ESSF. Karakatsoulis estimated compensation (CP) and saturation (SP) points (at 20 °C) for Engelmann spruce on subalpine sites in the Nelson Region to be 75 p.mol/s/m2 and 500 umol/s/m2, respectively. Draper et al (1985) measured CP and SP (at 15 °Q of 100 u.mol/s/m2 and 600 umol/s/m2 for interior spruce in the Prince George Region. Using these values for comparison with the Vavenby sites indicates that average light levels are below the compensation point for photosynthesis at ground level in the herb 74 and shrub patches, and marginal in the skid trail patches. At 25 cm above-ground, average levels are just above the CP in the three patch types. Differences between the herbs and shrubs in the temporal patterns of light availability at 15 cm within the two canopies may be significant to seedling performance because PAR was at levels near or below CP in both canopies except during periods of highest light availability, which differed between herb and shrub patches. Higher light levels earlier in the day in the shrub patches may be more favourable for seedlings during periods of low soil moisture or atmospheric humidity that reduce stomatal conductance in Engelmann spruce seedlings. Although this may be of little consequence to seedling performance in the ESSFwc2 because soil moisture deficits are uncommon in this variant, little is known about diurnal patterns of physiological activity in Engelmann spruce seedlings in these subalpine environments. Average light levels reach SP within the herb and skid trail vegetation by 50 cm above the ground, but do not do so in the shrubs until > 1 m above the ground. The lower light levels above 25 cm in the shrubs may significantly decrease the performance of planted seedlings in this environment, because seedlings will likely remain within the shrub canopy for at least the first five years of growth. On the other hand, if shrub patches are snow-free earlier in the spring (as the limited data collected in this study suggested), a longer growing season and higher light levels at ground level beneath the shrubs might favour the germination and early growth of naturally established seedlings. Pattern (1963) found that light increased germination in Engelmann spruce seeds at low air temperatures (1.6 to 10 °C), although it had little effect on germination when temperatures were higher (18.3 °C). Carbon levels in the humus and soil layers beneath the herb and shrub vegetation were generally similar to those measured by Carter et al (1985) in humo-ferric podzols in coastal British Columbia. The carbon content of the herb humus layers was also similar to that for forest floor material from residual stands on other sites in the ESSFwc2, a^lthough the shrub humus C content was somewhat lower (David, 1987; Herring and McMinn, 1980). Mineral soil carbon levels beneath the shrubs on the Vavenby sites were similar to levels in the Bfh horizons of soils from these other sites - but the C content of herb mineral soil on the Vavenby sites was much higher. The high carbon contents in the herb soil at 30 cm (7.9%) is consistent with a deep Ah or Ahe horizon beneath this vegetation (as indicated from the soil profile in Table 2.2). There were no analagous soils sampled in either David's (1987) or Herring and McMinn's (1980) studies. Carter et al. (1985) reported a mean and standard deviation of 9.24 + 4.07 % for percent total carbon in the Ah horizon of coastal humo-ferric podzols: in contrast, the percent carbon content of the Ae horizon in that study was 2.7 + 1.2 %, and in the Bfh it was 4.14 + 0.82 %. Differences in the amount and distribution of organic matter (using carbon content as an index of organic matter) beneath the herbs and shrubs may be a consequence of differences in the characteristics of litter input. For example, whereas annual additions from woody vegetation are primarily from leaf-fall, inputs from root mortality may dominate organic matter additions beneath herbaceous vegetation (Armson, 1977). The presence of an Ah layer beneath the herbs, as indicated in the soil and mineralizable N profiles, suggests greater nutrient availability beneath the herb vegetation than the shrubs. Differences in organic matter content may also affect soil structure and thermal conductivity. However, Armson (1977) suggests that under field conditions factors other than organic matter content, such as soil moisture for example, usually have a greater impact on temperature in forest soils. Carbon values were higher in the skid trail soils than in the shrub soils on the Vavenby sites. Smith and Wass (1985) found similar differences on a shrub-dominated site in the 76 ESSFwc in the Nelson Forest Region. The total carbon in the 0-10 cm soil layer on skid trails on their site was greater (3.4 %) than the total carbon in undisturbed soil (2.5 %). The differences between disturbed and undisturbed soils may be caused by the incorporation of humus material into skid trail soils by logging activities, or additions to soil organic matter from the root mortality that follows bole removal and the destruction of non-crop vegetation with harvesting. Mineralizable nitrogen levels in the humus layers beneath the herbs and shrubs are similar to values for the forest floor obtained by David (1987) for a six-year-old cutover in the ESSFwc2, and to levels measured by Carter (1991 - unpublished data) in forest floor material from 56 Douglas-fir sites in coastal British Columbia. However, levels in the mineral soil on the Vavenby sites are higher than levels in the mineral soil on these other sites. Mineral soil mineralizable nitrogen (gross, unadjusted values) at 0-15 cm on David's site were 16 ppm, compared to 58 ppm and 31 ppm at 10 cm beneath the herbs and shrubs on study site 2. Higher levels on the Vavenby site may be due to a much higher herbaceous component in the vegetation on this site, and a higher soil pH (5.15-6.19 oh the Vavenby site, in comparison to 4.4 on David's site). Feller (1990 -unpublished data) measured levels of mineralizable N of 20 to 40 ppm in soils from a subhygric site and 10 ppm in soils from a mesic site in the ESSFwc2. Higher mineralizable N levels in the herb soils on the Vavenby sites were associated with higher moisture contents in these soils (see Chapter 3). The bulk density of mineral soils on David's (1987) chronosequence of ESSFwc2 cutovers ranged from 680 - 980 kg/m ,^ values comparable to those on the Vavenby sites. Skid trail soils on the Vavenby sites had higher bulk densities than soils from beneath the herbs and shrubs. However, differences were small and densities did not exceed 1000 kg/m .^ These values are much lower than those reported by Smith and Wass (1985) for skid trails on winter-logged subalpine sites in British Columbia. Bulk densities on vegetated skid trails on the Vavenby sites fall within the range found in undisturbed soils in the ESSF, and differences between on- and off-skid trail locations are unlikely to significantly affect growth. The relatively low bulk density of the Vavenby skid trails compared to values for skid trails on other sites suggests minimal soil compaction from winter-logging on the Vavenby sites. However, the sample size for bulk density determinations for skid trail soils was very small (n=2), and only vegetated areas were sampled. Densities in soils in areas on skid trails that were still free of vegetation four years after logging may have been much higher. Unvegetated patches constituted approximately 5 to 10% of the area of skid trails. Given the natural variability in the bulk density of soils, as well as the variability in intensity of disturbance that characterizes the skid trail environment, it can not be concluded that soil disturbance in all areas on the Vavenby skid trails was below levels that would affect seedling growth. 78 CONCLUSIONS Low light availability is likely to be the major factor limiting the early growth and survival of seedlings in both herb-dominated and shrub-dominated patches on the Vavenby cutovers. Soil temperatures are also low throughout the year and may reduce the physiological performance of seedlings but the effects of low temperature on root and shoot growth in Engelmann spruce are not well understood. Higher levels of soil organic matter and mineralizable nitrogen in herb patches may provide a more favourable environment for the growth of planted seedlings than that within the shrub patches. However, thinner litter layers, earlier snowmelt, and higher light levels on the ground in the shrub patches may provide a more favourable environment for germination and early survival of naturally established seedlings. Higher soil temperatures and light levels on the skid trails may make these factors less limiting to the growth of planted seedlings in this patch type than in the herb and shrub patches. The physical properties of the skid trail soils were undersampled and mineralizable nitrogen levels were not measured in these soils, but the limited data on texture, bulk density and carbon contents did not suggest major problems for seedling growth in vegetated skid trail patches. Therefore, the growth of seedlings in this patch type is predicted to be as good or better than that of seedlings in the herb and shrub patches. Vegetation, microclimatic and edaphic differences between the three patch types lead to the general prediction of differences in seedling responses, both with and without the presence of above-ground vegetation. Differences in light availability may influence seedling survival and growth within canopies of non-crop species; differences between patches in the physical and chemical characteristics of the soils may influence growth when light is not a limiting factor. To study seedling responses to the presence and absence of the vegetation in the three patch types, and on bare skid trails, an experimental planting of Engelmann spruce was carried out on the Vavenby sites. The results of this e^xperiment are discussed in the following chapter. 8 0 C H A P T E R 3 T H E P E R F O R M A N C E O F P L A N T E D E N G E L M A N N S P R U C E S E E D L I N G S IN H E R B , SHRUB, A N D SKID T R A I L P A T C H E S O N C U T O V E R S IN T H E E S S F w d I N T R O D U C T I O N Poor post-logging regeneration in spruce-fir forests in British Columbia has been the subject of problem analyses for over two decades (Butt, 1988; Dobbs and McMinn, 1971). Planted seedlings of interior spruces (white spruce, Engelmann spruce, and their hybrids) have failed to meet silvicultural expectations - their early survival and growth are generally poor, although individual performance is highly variable (Burdett et al, 1984; Butt, 1988,1989; Coates, 1987; Vyse, 1981). Numerous site factors have been associated with the variation in growth and survival. These include: elevation, slope and aspect (Butt, 1989), soil moisture (Bassman, 1989; Burdett et al, 1984; Butt, 1989), nutrient availability (Burdett et al, 1984), and competing vegetation (Butt, 1989; Eis and Craigdallie, 1983). In Chapter 2, within-site variation in several of these and other factors was discussed in relation to the distribution of herb, shrub, and skid trail patches on the Vavenby sites. Interactions between stock-types and microsites, as well as nursery effects, can also influence the performance of seedlings after outplanting (Burdett et al, 1984; Vyse, 1981). Both abiotic and biotic factors influence seedling establishment on subalpine cutovers in south-central B.C. Tree seedlings are planted onto sites that have both well-developed post-logging plant communities and the climatic constraints that are characterisitic of high-elevation environments (Lloyd et al, 1990). However, no detailed studies of the ™ interactions between these factors and their relative impacts on spruce establishment have been undertaken. Such studies are needed for several reasons: 1) the results from past research have not determined under what conditions particular factors have important impacts on early seedling performance, although this research has suggested that importance can vary with stock-type, site type, and silvicultural treatment (Bassman, 1989; Butt, 1989; Eis and Craigdallie, 1983; Vyse, 1981); 2) many different silvicultural systems have been applied in subalpine forests in B.C., but none has been found to be consistently favourable for spruce regeneration; and 3) management recommendations for the selection of site treatments are being made (Butt, 1989) and implemented with a poor understanding of the responses of spruce and other species to various environmental conditions. Literature Review Although studies have not been done with interior spruces, studies with other conifers have shown that herbs and shrubs can have different impacts on seedling performance (Amaranthus and Perry, 1989; Cole and Newton, 1987; McPherson and Wright, 1989; Wagner, 1989; White and Newton, 1989). Both types of vegetation typically reduce conifer seedling growth and survival, but even under similar site conditions the amount and type of reduction can differ. Several recent studies on the competitive effects of herbs and shrubs on Douglas-fir seedlings on cutovers in Oregon illustrate the complexity of competitor-resource-seedling relationships (Amaranthus and Perry, 1989; Cole and Newton, 1987; White and Newton, 1989; Wagner, 1989). Amaranthus and Perry (1989) found that the survival and growth of Douglas-fir seedlings were lower in a grass-dominated community than in an adjacent Arctostaphylos viscida Parry (whiteleaf manzanita) shrub community. White and Newton (1989) found that the growth of Douglas-fir seedlings under low densities of Arctostaphylos with herbs was poorer than growth beneath high densities of the shrubs without herbs. However, Wagner (1989) concluded that overtopping by a different group of woody neighbours (Rubus spectabilis Pursh, R. parviflorus Nutt., Gaultheria shallon Pursh, Alnus rubra Bong., Acer circinatum Pursh, and Sambucus racemosa L.) had 10 times the competitive impact of herbs on Douglas-fir seedling height growth, but only 1.6 times the impact on basal area growth. Cole and Newton (1987) found that both red alder and grass species reduced the growth of seedlings and that impacts increased with increasing density of competitors. However, whereas grasses had their greatest impacts on the driest sites, the impacts of red alder were greatest on moist sites where light availability was most limiting to growth. A comparison of the results of these studies shows that the impacts of non-crop species on Douglas-fir seedlings depends largely on the relative competitive abilities of non-crop species for the limiting resources on a particular site. The presence of competing vegetation may limit seedling growth or survival by numerous mechanisms, such as light reduction (Chan and Walstad, 1987; Wagner, 1989), low moisture or nutrient availability (Amaranthus and Perry, 1989; Newton and Preest, 1988), or alleopathy (DelMoral and Cates, 1971), among others (Radosevich and Holt, 1984). However, the density of neighbours typically explains little of the observed variation in seedling growth and survival in statistical analyses (Coates, 1987; McPherson and Wright, 1989; Wagner, 1989). The indirect nature of exploitation competition (where competitive effects are through reduction in available resources as opposed to through direct effects of one individual on another) may account for the weak, 83 albeit significant, relationships found between seedling growth and various characteristics of the seedling's biotic neighbourhood. Differences in light, soil moisture, and soil nutrient availability between the herb, shrub, and skid trail patches on the Vavenby sites suggested that there would be differences in seedling performance among the three areas, both in the presence and the absence of above-ground vegetation. However, the precise nature of these differences were difficult to predict, for several reasons. Firstly, studies have shown positive as well as negative impacts of co-occurring vegetation on seedling growth and survival. Positive effects can occur through improved mycorrhizal relations (Amaranthus and Perry, 1989), increased nitrogen availability (Binkley, 1984), or amelioration of radiation, temperature or moisture extremes (Childs and Flint, 1987; Wahlenberg, 1930). Some or all of these factors may be important to successful spruce seedling establishment on interior subalpine cutovers (Bassman, 1989; Burdett etal, 1984; Coates, 1987; Draper etal, 1985; Ronco, 1970), but to what extent and under what conditions they might compensate for negative impacts are not known. Secondly, it is not known whether seedling growth is limited simultaneously by several environmental resources (as suggested by Chapin et al. (1987) for plants in general) or whether control switches among different resources during the growing season (as suggested by DeLucia and Smith (1987) for air and soil limitations on photosynthesis in Engelmann spruce). Consequently, it was difficult to make a priori assumptions about controlling factors. Thirdly, predictions of response require some understanding of the capacity of individuals of a species to respond to environmental heterogeneity. For example, the amount of phenotypic plasticity in a species would be one way to assess such a capacity, v However, there is little information on these characteristics for Engelmann spruce. The selection of appropriate seedling response variables is more problematic when the factors controlling seedling performance are not well-defined. Seedling mortality was -chosen as a variable in this study because seedling survival is the major management objective in plantation establishment. Some evidence from recent studies has suggested that competition thresholds exist for seedling survival, below which there are little gains in survival for further reductions in competitor density (Coates, 1987; Wagner, 1989; Wagner et al, 1989). The results from these studies also suggest that thresholds for survival and growth within a species are different, with that for survival typically higher than that for growth. This implies that it is important to know whether the management objective is improved survival or improved growth (or both) for the selection of appropriate silvicultural treatments. In studies of the effects of competing vegetation on conifer seedlings, height and diameter are the variables typically selected for measurements of seedling responses (Coates, 1987; Cole and Newton, 1987; Eis and Craigdallie, 1983; McPherson and Wright, 1989; Vyse, 1981). Height growth may be very important to establishment when light is the major factor limiting early growth, but with or without competition for light, it is often difficult to interprete height growth in terms of competitive impacts on seedlings. Frequently, height growth appears unresponsive to competition, or less responsive than stem increment (Bassman, 1989; Coates, 1987; Wagner etal, 1989; Zutteref a/., 1986). This may be due to etiolation in response to the low light conditions often associated with intense competition, or the fact that height growth typically occurs early in the season when carbohydrate reserves and light availability are highest (Wagner, 1989). Stem diameter growth often seems more responsive to interference than height growth and is of obvious interest to silviculturalists because of its relationship to wood volume (Coates, 1987; Wagner, 1989). Studies of interference typically correlate current year's growth with the environmental conditions of the same year. As height growth is more strongly affected by the previous year's environment than diameter growth, it is perhaps not surprising that interference seems to have a greater effect on diameter than height. However, plants invest resources in more than just height and diameter growth. Needles and buds also show responses (generally reductions in size and number) to interference from co-occurring vegetation (Burdett et al, 1984; Harrington and Tappeiner, 1991). Schlicting (1986) pointed out in his review of phentoypic plasticity that the plastic responses of individual components in plants may be strongly correlated, although different characters may show very different responses to the same treatment. In this study, all above-ground seedling components (i.e. stems, needles, and buds) were measured to determine both the amount of variation among seedlings in the growth of individual components and the relationship between this variation and environmental heterogeneity. Differences in the size of seedling components will presumably be correlated with differences in biomass - although height has not always been found to correlate well with biomass, perhaps for the reasons discussed above (Burdett et al., 1984; Cole and Newton, 1987). The partitioning of carbon within trees, particularly between above- and below-ground components, in response to site conditions has important implications for long term site productivity (Comeau and Kimmins, 1989; Kurz, 1989) but its role in the early establishment of trees is not well understood. In general, observations of compensatory resource allocation between shoots and roots in plants have led to the concept of a functional balance between above- and below-ground components. Changes in resource availability cause a compensating re-allocation of assimilates resulting in an adjustment in the ratio of shoot:root biomass ( S : R ) and the carbon partitioning pattern, within the plant (Kurz, 1989). Under conditions of resource limitation, plants will shift the 86 allocation of resources to that component acquiring the limiting resource (Chapin et al, 1987; Tilman, 1988; Wilson, 1988b). Although there is evidence that some plant species can shift allocation of net growth between shoots and roots in response to changes in the relative availability of above- and below-ground resources, the capacity to do so and the conditions under which these shifts occur, vary considerably among species (Chapin et al, 1987; Ledig and Perry, 1965; Schlicting, 1986; Wilson, 1988b). Measuring changes in allocation strategies is not as simple as measuring changes in S:R ratios. In wOody seedlings, apparent changes in S:R ratios may be due to size-dependent changes in roots and shoots rather than due to alterations in the relative growth rates of above- and below-ground plant components (Ledig and Perry, 1965; Wilson, 1988b). The equation of simple allometry is an empirical model frequently used in studies of the changes in the relative growth rates of plant components (Gould, 1966; Ledig and Perry, 1965; Wilson, 1988b). The allometric formula is: where a = allometric coefficient b = allometric exponent Y = measure of shoot growth X = measure of root growth Ledig and Perry (1965) suggested that to show shifts in allocation it was necessary to show significant changes in b, the allometric exponent, which is a measure of the differential increase in Y relative to X. According to Ledig and Perry, significant changes in b should not be expected to occur under conditions favouring healthy plant growth. i Although numerous studies since Ledig and Perry's paper have shown significant differences in S:R ratios under different environmental conditions, conclusions of Y = aXb allocation shifts based on these observed changes are not always supported by tests for size-independent effects (e.g. tests for differences in b) (Wilson, 1988b). Wilson (1988b) reported that in woody species S:R ratios typically decrease continually as plants grow. However, evidence from studies with planted conifer seedlings suggests the opposite trend for early growth in some species (Bassman, 1989; Newton and Cole, 1991). Newton and Cole (1991) reported that S:R ratios increased in Douglas-fir seedlings between three and five years after planting. In their study seedlings experiencing different densities of grass, red alder, and intraspecific competition had much lower S:R ratios under high densities of all competitors. They hypothesized that reductions in S:R ratios under severe competition were due to slender stems and the loss of lower foliage and branches that occurred without a corresponding loss in root biomass. After three growing seasons on an ESSF cutover in south-central B.C., the S:R ratios of interior spruce seedlings were lower on mounds than in undisturbed plots, even though the seedlings on mounds had significantly higher total biomass and S:R ratios within both treatments had increased over the three growing seasons (Bassman, 1989 - the author did not calculate S:R ratios; these were derived from treatment means for needle, stem, and root biomass). Measurements of plant and soil water relations indicated that moisture was limiting in the mounds during the latter part of the first two growing seasons which, according to Bassman (1989), suggested that the lower S:R ratios did not represent a shift in allocation but rather a limit on shoot growth. In another study, Lieffers and Titus (1989) found that the S:R ratios in white spruce seedlings increased when soil nutrient limitations were removed through fertilization. However, ratios decreased at higher plant densities where levels of intraspecific competition were more intense. . 88 Little is known about the resource-acquisition strategies of Engelmann spruce seedlings. If S:R ratios do generally increase with size in the first few years after planting but S:R ratios are lower when soil factors are limiting growth, then seedling responses to the removal of vegetation within the herb, shrub, and skid trail patches on the Vavenby sites will depend on the extent to which light is limiting in the presence of vegetation as well as the impacts of its removal on below-ground resources. Changes will also depend on the plasticity of allocation within individuals. Evidence from recent studies with conifers seedlings suggests that there are significant interspecific differences in the degree to which growth is responsive to environmental variation. Messier (1991) studied differences in the growth of planted seedlings of three conifer species growing on nutrient-poor sites on Vancouver Island. He found that western redcedar (Thuja plicata Donn) was much less responsive than Sitka spruce (Picea sitchensis (Bong.) Carr.) or western hemlock (Tsuga heterophylla (Raf.) Sarg.) to all treatments and site conditions that affected forest floor nutrient availability. Cedar maintained a similarly low level of growth almost independently of the site conditions, suggesting that strong genetic controls limited growth in this species. Lieffers and Titus (1989) found less plasticity but greater size inequality in white spruce seedlings than in lodgepole pine seedlings when both were grown under different levels of soil resources and intraspecific competition. According to Grime (1979), competition in low-resource habitats may be unimportant relative to the ability of plants to conserve resources and resist environmental hazards in these habitats. He suggested that competitive characteristics such as rapid potential growth rates and high phenotypic plasticity might be disadvantageous to species growing under conditions of extreme and continuous environmental stress. Whether spruce is adapted to a low-resource environment, or to one of extreme and continuous stress, is 89 debatable. Subalpine environments are typically characterized as harsh, but mature subalpine forest communities contain large amounts of biomass even though rates of production may be slow relative to lower elevation forests. Interpretation of seedling responses requires consideration of the adaptive significance of these responses to regeneration in subalpine environments. For example, resource conservation rather than resource acquisition may underly the responses of spruce seedlings to variations in resource availability, but this may be difficult to establish with simple growth measurements. Ultimately, an understanding of seedling growth strategies will also require knowledge about the relationship between seedling responses and mature tree performance in subalpine forests. Complex relationships between resource levels, competitive interactions, and seedling responses (which are further complicated in the subalpine by variation in climatic constraints on growth) pose a considerable challenge to attempts to increase our understanding of the conditions necessary for successful establishment of spruce seedlings on subalpine cutovers. Studies comparing conifer seedling growth in environments representing different combinations of competition and resource availability typically suffer from difficulties in interpretation of the complex interactions among biotic and abiotic factors (e.g. Brand, 1991). However, the approach is attractive because it yields useful empirical information for silviculturalists, as well as providing data essential for the development of predictive ecological models. Despite its limitations, this approach was adopted for the study of seedling performance on the Vavenby sites. The herb, shrub, and skid trail patches described in Chapter 2 represent three distinctly different vegetation/resource environments. In this study resource levels were further altered by the removal of above-ground vegetation in some patches. These treatments were designed to eliminate (above-ground) interference from non-crop species and thereby alter the relationship between above- and below-ground resource limitations on seedling growth. Seedlings were also planted in unvegetated areas on skid trails to study growth under conditions of no above- or below-ground interference from non-crop species. The changes in the edaphic and microclimatic conditions in the patch types on the Vavenby sites and their impacts on the survival and growth of planted Engelmann spruce seedlings are described in this chapter. i 91 M E T H O D S Experimental design To study the responses of Engelmann spruce seedlings to the presence and absence of vegetation in the herb, shrub, and skid trail patches, an experiment with a split plot randomized block design was set up on the Vavenby sites in June, 1987. Herb, shrub, and skid trail patches were located on each of the two study sites. The characteristics of the three patch types were described in detail in the previous chapter, but their species composition is repeated here as review. Herb patches had at least 80% cover of any or all oi Valeriana sitchensis, Thalictrum occidentale, and Arnica latifolia (Plate 1.4). Shrub patches had at least 80% cover of Menziesia ferruginea or Rhododendron albiflorum, or both (Plate 1.3). Skid trail patches had at least 50% cover of any or all of Valeriana sitchensis, Thalictrum occidentale, and Epilobium angustifolium (Plate 1.2). Six patches of each type were located on each of the two study sites and a 3 m x 6 m plot was staked out within each patch. Three additional 3 m x 6 m plots were staked out on areas of skid trails that had not revegetated since logging (Plate 2.2). These areas of primarily mineral soil were more severely disturbed than the vegetated patches on skid trails which were described in the last chapter. The bare plots were included in the experiment to study the performance of seedlings in an environment without above- or below-ground interference from non-crop species. Three of the six plots in each patch type were treated by clipping all vegetation (manually) at ground level and removing it from the plots. An additional 1 m border was 92 cleared around plots to minimize edge effects. The other three plots in each patch type were not disturbed. The three bare plots on skid trails were treated differently: their surfaces were broken up with shovels and the mineral soil was cultivated to a depth of approximately 25 cm. Clipped plots were kept free of vegetation until July, 1989 by periodic clipping during the growing season. A l l clipped material was removed from the plots. Vegetation development in the bare skid trail plots was minimal over the course of the experiment, but any invading plants were removed. In summary, on each site there were a total of 21 plots distributed among seven treatments (including controls). These treatments (each replicated in three plots) are defined as follows: 1. Above-ground removal of vegetation in herb patches (hereafter referred to as the clipped herb treatment); 2. Undisturbed herb patches (hereafter referred to as the herb control); 3. Above-ground removal of vegetation in shrub patches (hereafter referred to as clipped shrub); 4. Undisturbed shrub patches (hereafter referred to as shrub control); 5. Above-ground removal of skid trail vegetation (hereafter referred to as clipped skid trail); 6. Undisturbed skid trail vegetation (hereafter referred to as skid trail control); 7. Unvegetated skid trail (hereafter referred to as bare skid trail). Shortly after the initial clipping, 25 seedlings of each of two Engelmann spruce stock-types were planted into each plot (a stock-type was randomly assigned to one half 93 of the plot) with a 50 cm spacing between seedlings. One stock-type was a cold-stored 2-0 Engelmann spruce 313 plug stock (hereafter referred to as the cold stock) grown at the Pelton nursery in Maple Ridge, B.C. The seed for this stock (seedlot 04330) was collected in 1979 at an elevation of 1570 m on Avola Mt., approximately 50 to 60 km north of the Vavenby sites. These seedlings were lifted on Dec. 17th, 1986 and cold-stored until June, 1987. The second stock-type was a hot-planted 2-0 Engelmann spruce 313 plug stock (hereafter referred to as the hot stock) grown at the University of British Columbia nursery in Vancouver. The hot stock seed (seedlot 04332) was collected in 1979 at an elevation of 1670 m near Miledge Cr., approximately 100 to 120 km north of the Vavenby sites. The hot-planted seedlings set their 2nd year buds in the nursery in the spring of 1987 and were lifted on June 20th, just prior to planting. Based on the elevations and locations of the collections, it was assumed that the seeds for both stock-types came from populations of Engelmann spruce rather than of the Engelmann spruce x white spruce hybrids that are widely distributed at lower elevations in the interior. The cold stock was planted with shovels on June 16th and 17th 1987, after screefing down to mineral soil. The hot stock was planted in a similar manner on June 22nd and 23rd, 1987. Air temperatures were cool and the soil was moist on all planting days, with a light snow falling on June 22nd. Differences in treatment responses were expected between the two stock-types because the hot stock, unlike the cold seedlings, did not flush during the first growing season (1987) on the sites. Hot planting is currently viewed as a solution to several subalpine regeneration problems (chief among them a short planting window and poor seedling acclimation during the first year after planting) and is being undertaken on an increasing scale in B.C. (Daniels and Simpson, 1990). However, no comparative studies of the performance of cold-stored and hot-planted stock-types has been carried out. The interpretation of the relative performances of the two stock-types used in this study is limited by their different seed and nursery histories. The two stock-types were not available from a single seedlot or a single nursery when this study was undertaken. Sampling design and measurements Prior to clipping in June 1987, the cover of all species in each plot was recorded. Covers were re-estimated in August 1989 to assess both the changes in control plots between 1987 and 1989, and the ability of the clipped species to re-establish cover after repeated removals between 1987 and 1989. The initial estimates of cover were presented in Chapter 2 and a comparison of initial and final species composition is presented in Chapter 5. After planting in 1987, the following measurements were taken on all seedlings: - height (cm) to the base of the terminal bud; - stem diameter (mm), 5 cm above the ground; - total number of laterals on the main stem; - length (cm) of the top three laterals on the main stem; - length (mm) of three needles on the leader, 1 cm below the apex In the spring of 1988, the following measurements were added to those above: - number of buds on the top three laterals on the main stem; - number of needles in one rank along a 1 cm length of the leader; - leader length (cm) to the base of the terminal bud. 95 Remeasurement of all growth variables after August 1987 was restricted to ten seedlings of each stock-type within each plot. Centrally-situated seedlings were selected for measurement; border seedlings were omitted to reduce the possibility of an edge effect. Measurements were made in August 1987,1988, and 1989. Observations on seedling mortality and condition were made at the beginning (winter damage) and end (summer damage) of each growing season. The condition of seedlings was assessed as poor, fair, or good based on the colour, vigour, relative size, and the extent of physical damage. In August 1989, three trees were randomly selected for harvesting from among the ten measurement trees of each stock-type in each plot. Seedlings were excavated with a shovel, separated into roots and shoots, and air-dried for temporary storage. Root samples were subsequently cleaned of soil, and both roots and shoots were oven-dried at 70 °C for 48 h. Root and shoot samples were weighed on a Mettler PC440 balance with a precision of + 0.01 g. Thirty seedlings of each stock-type selected randomly from the planters' bags at the time of planting were processed in a similar manner in 1987 for the determination of initial biomass. Soil physical and chemical properties Soil temperature Soil temperatures in the planted spruce plots were measured once-monthly from June to August in 1988 and 1989 to determine the effects of vegetation removal (without removal of the humus layer). Thermocouple temperature sensors were prepared by soldering and epoxying 1 m lengths of 20 gauge insulated copper-constantan wire (Omega Engineering, Conn.). In August 1987, thermocouples were buried at depths of 10 and 20 cm in all plots on site 1, and in one randomly selected plot per treatment on site 2. Readings were taken with a WESCOR HR-33T Dew Point Microvoltmeter (WESCOR Inc., Logan, Utah). Soil moisture Soil moisture was determined gravimetrically for samples collected from the planted spruce plots in July of 1987 and 1988. Samples were collected with a Lord Soil Sampler (Hoskin Scientific (Western) Ltd., Vancouver, B.C.) which was inserted to a depth of 15 cm in four locations within each plot on both sites. Samples were bagged and refrigerated until analysis. Soils were weighed wet, dried at 70 °C for seven days and reweighed. Weight loss was expressed as a % of dry weight for the determination of % soil moisture content (Gardner, in Black, 1965). Nitrogen ion concentrations To determine ammonium and nitrate ion movement from the litter to the soil within the planted spruce treatments, resin bags containing anion and cation exchange beads (Rexyn 101(H) and Rexyn 201(OH) Research Grade, Fisher Scientific, Ottawa, Ont.) were placed in plots on site 1 in August 1989, according to DiStefano and Gholz (1986). Four pairs of bags (constructed from pantyhose) containing 15 g of anion beads and 30 g of cation beads were buried at the humus/mineral soil interface in the four corners of two randomly selected plots per treatment. The bags were collected in October 1988 and kept in cold storage until analysis. The method for extraction was modified from that described by David (1987): the resin beads were placed into 100 ml solution of I N K C L , shaken for one hour and left to equilibriate for one to two days. Solutions were filtered through a #41 Whatman filter and the extracts cold-stored until ammonium and nitrate concentrations were measured with a Technicon Autoanalyzer. Litter decomposition rates In July 1987, live individuals of the major herbs and shrubs on site 1 were collected for the preparation of litter bags (Titus and Malcolm, 1987). Plant material was dried at 70 ° C for two days, and 1.5 g samples of mixed shrub leaf litter (from Menziesia ferruginea and Rhododendron albiflorum) and mixed herb litter (from Valeriana sitchensis, Thalictrum occidentale, Gymnocarpium dryopteris, and Epilobium angustifolium) were placed in nylon bags (with a mesh size of 1 mm). Pairs of herb and shrub bags were buried at the humus/mineral soil interface in four locations within two randomly selected plots per treatment on site 1 in late August of 1987. The bags were excavated in October 1988, and after one week in cold storage were dried at 70 °C for two days. Samples were removed from the bags and reweighed, and decomposition expressed as % weight loss over the incubation period. To determine decomposition rates between October and June (i.e. outside the growing season) in the herb and shrub controls only, a second set of litter bags was prepared in September of 1988. Herb bags were prepared using 3 g of dried (70 °C for two days) Valeriana sitchensis litter, and shrub bags using 3 g of dried Menziesia ferruginea litter. Pairs of bags were buried at the humus/mineral soil interface in three locations in each of 98 the herb and shrub control plots on site 1. Bags were collected in June of 1989, and processed as described above for the mixed litter samples. Data analysis For simple growth variables, a nested analysis of covariance (ANCOVA) model was used to test for significant differences among treatments three growing seasons after planting. The covariate was size or number at the time of planting. Variables which were not measured in 1987 were compared graphically but not tested for significant differences. Data were transformed prior to analysis if Bartlett's test for homogeneity of variances was significant. Initial transformations were selected based on either the type of data or the relationship between the mean and measures of variation for a particular dataset (Zar, 1984). If recommended transformations failed to homogenize data, other common transformations were applied. Data for the two stock-types and two sites were analysed separately and in reporting the results the analysis for both stocks or sites is presented only when there were differences between them. Orthogonal contrasts were used for preplanned comparisons between groups of treatments. Contrasts were used to test for differences between herb and shrub treatments, clipped and control treatments, and between on-skid trail and off-skid trail treatments. Tukey's HSD test was used for post hoc pairwise comparisons of treatment means. A l l significance testing was done at the a=0.5 level. SYSTAT Version 5.0 (Wilkinson, 1990) was used for all statistical analyses. Height:diameter ratios and relative growth rates (RGR) were derived from the simple growth measurements. Relative growth rates (in logemm/year) were calculated for each seedling based on diameter measurements, as follows (Hunt, 1982): RGR = log e D 2 - log e Di where D 2 = diameter in 1989 D j = diameter in 1988 Nested analysis of variance (ANOVA), with plots nested within treatments, was used to test for significant differences among treatments in shoot, root, and total biomass. Shoot:root ratios were derived for each seedling and similarly tested. Orthogonal contrasts and Tukey's HSD test were used for preplanned and post hoc comparisons of means, respectively. A l l data were loge-transformed for analysis. Shoot and root data were fitted to the simple allometric equation (see introduction) and the allometric exponent was tested for significant differences among treatments by testing for parallelism among treatment slopes. The allometric relationship between seedling diameter and total biomass was explored in a similar manner. Analysis of variance was used to test for differences among treatments in soil temperatures, soil moisture, nitrogen ion concentrations, and litter bag weight loss. Most variables were transformed before analysis to increase the homogeneity of variances. For most variables within-plot replication was lost through the destruction of samples, or faulty sensors, so that only a simple A N O V A was used for the analysis of the data. Comparisons of means was carried out as described for seedling variables. exponent was tested for significant differences among treatments by testing for parallelism among treatment slopes. The allometric relationship between seedling -diameter and total biomass was explored in a similar manner. Analysis of variance was used to test for differences among treatments in soil temperatures, soil moisture, nitrogen ion concentrations, and litter bag weight loss. Most variables were transformed before analysis to increase the homogeneity of variances. For most variables within-plot replication was lost through the destruction of samples, or faulty sensors, so that only a simple ANOVA was used for the analysis of the data. Comparisons of means was carried out as described for seedling variables. i 100 RESULTS Seedling survival and condition Seedling mortality during the first three years after planting was low in both stock-types, but was slightly higher in the cold stock (1.74% of total) than in the hot stock (1.17% of total). Of the 30 seedlings that died, 2 died shortly after planting, 18 died in the first winter, and the remaining 10 died during the second growing season. In the cold stock, mortality was not distributed evenly across treatments (Figure 3.1 -upper figure). Mortality was highest in the herb control, where 8.2% of the seedlings died (results were similar for the two sites and are combined in Figure 3.1). There were low levels of mortality in the shrub control and the skid trail control, where 1.4% and 2.7% of the cold seedlings died, respectively. There was no mortality in any of the clipped treatments or in the bare skid trail treatment. The highest percentage of cold seedlings in poor condition (20.4%) was also in the herb control treatment (Figure 3.1). Seedlings in poor condition in the herb and shrub controls had fewer, shorter laterals, fewer older needles, and were less sturdy than seedlings in good condition, as illustrated by the two seedlings in Plates 3.1 and 3.2. The majority of seedlings in the clipped shrub and the clipped herb treatments were in good condition - 81.7% and 75.3% of the totals, respectively (Figure 3.1). Despite the lack of mortality in the clipped and bare skid trail treatments, fewer seedlings were in good condition than in the off-skid trail clipped treatments. Most of the seedlings in poor and fair condition on the skid trails were chlorotic - seedlings in off-skid trail treatments did not exhibit symptoms of chlorosis. 101 Figure 3.1 Condition of cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on sites 1 and 2 (both sites combined). Data are presented as percentages of total number of seedlings, n (number of seedlings per treatment) = 150. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 102 Plate 3.1 Engelmann spruce seedling (cold stock) in poor condition three growing seasons after planting in a herb control plot on site 1. In the herb and shrub controls seedlings in poor condition were spindly with fewer and shorter laterals and needles than seedlings in good condition. These seedlings were not chlorotic, however, unlike seedlings in poor condition on skid trails. 103 Plate 3.2 Engelmann spruce seedling (cold stock) in good condition three growing seasons after planting in a clipped shrub plot on site 1. Contrast size and form with seedling in poor condition in Plate 3.1. 104 Some mortality in the hot stock occurred in all treatments except the bare skid trail (Figure 3.1 - lower figure). The greatest number of deaths was in the clipped herb treatment, but the percentage was small (3.3%). With the exception of seedlings in the herb control, the condition of the hot stock three years after planting was poorer than that of the cold stock within the same treatment. Unlike the cold seedlings, which suffered most in the herb control, the hot seedlings did most poorly in the shrub control (Plate 3.3). In general however, the pattern among treatments was similar for the two stock-types. Seedling size The data for sites 1 and 2 and for each of the two stock-types were analysed separately because of interactions among factors. Not all the analyses for each variable are discussed here. For those variables where differences between stock-types were stronger than differences between sites (i.e. height, diameter, and number of laterals) the analyses for the hot and cold stock on site 1 are presented. For variables for which differences between sites were stronger (length of laterals and needles, and number of buds), the analyses for the cold stock on sites 1 and 2 are presented. All the tables of statistics for this chapter are presented in Appendix 3. At the end of the third growing season after planting, treatment means and standard deviations for seedling heights ranged from 29.85 + 7,25 cm to 40.73 + 8.04 cm in the , cold stock, and 23.94 + 4.63 cm to 35.39 + 5.52 cm in the hot stock on site, 1 (Figure 3.2 -upper figure, and Table 3.1 in Appendix 3). Note that standard error of the means are used for error bars in figures but standard deviations are given with means in the text. Engelmann spruce seedling (hot stock) in fair condition three growing seasons after planting in a shrub control plot on site 1. Contrast size and form with seedlings in poor condition in herb control (Plate 3.1) and in good condition in clipped shrub treatment (Plate 3.2). 106 50 40 B 30 o xi to § . 2 0 10 0 15 1. 2. 3. 4. 5. 6. 7. M Hot stock • Cold stock Figure 3.2 Treatment means and standard error of the means for height (upper figure) and diameter (lower figure) of two stock-types of Engelmann spruce after three growing seasons on site 1. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 107 Standard errors were used in the figures to make the trends in means easier to see. A nested analysis of covariance (ANCOVA) indicated nonsignificant differences among treatments in the height of cold seedlings on both sites (the results for site 1 are presented in Table 3.2a). Significance in all tests was tested at the a=0.05 level. In the hot stock differences among treatments were nonsignificant on site 2, but were significant on site 1 (Table 3.2b). Despite the nonsignificance of the ANCOVA results for the cold stock, patterns among treatments were similar to those in the hot stock, and orthogonal contrasts comparing the clipped herb and shrub treatments with the herb and shrub controls showed that seedlings of both stock-types in the vegetation-free treatments were significantly taller than those in the controls (Table 3.2). Differences among treatments in the diameter of seedlings exhibited similar patterns in both stock-types and both sites (Figure 3.2 - lower figure). At the time of planting, the diameter of the cold stock was 25% larger than that of the hot stock. This difference was maintained over the three growing seasons in the clipped treatments, but increased to approximately 50% in the controls. Treatment means and standard deviations for cold seedlings after three growing seasons ranged from 5.75 + 1.22 mm to 10.45 + 1.31 mm on site 1; means for the hot stock on this site ranged from 4.23 + 0.85 mm to 8.62 + 1.38 mm (Table 3.3). Coefficients of variation (CV.) ranged from 12 to 26 % across treatments for both stock-types. There were no patterns in CV. in relation to the presence or absence of above-ground vegetation. Diameters differed significantly among treatments in both stock-types (Table 3.4). Treatment means for seedlings in the clipped and bare treatments were significantly greater than means in the controls when only off-skid trail treatments were contrasted, and when all treatments were compared (Table 3.4). In general, seedlings growing in vegetation-free environments had diameters twice as large as those growing with other 108 species. Differences between the herb and shrub controls were not statistically significant, neither were differences between the clipped herb and shrub treatments (Table 3.3). Height:diameter ratios for cold and hot seedlings on site 1 are presented in Figure 3.3. Means and standard deviations ranged from 38.87 + 6.12 to 51.99 + 6.62 in the cold stock and 41.35 + 5.72 to 60.34 + 8.80 in the hot stock. In both stock-types, ratios were greater in the control treatments than in the clipped and bare treatments, suggesting that in the presence of above-ground vegetation seedlings maintained height growth at the expense of diameter growth. Height:diameter ratios were greater in the hot stock than in the cold stock, and the differences between the two were greatest in the herb and shrub controls. Mechanisms controlling height:diameter ratios are not well-understood and the differences between the stock-types could be related to size differences, to genotypic variation, or to differences between the nursery experiences of the two stocks. Height increments in cold seedlings from 1987, the year of planting, to 1989 are presented for the seven treatments on site 1 in Figure 3.4 (upper figure). Results were similar on both sites. Incremental height growth in 1987 showed the effects of planting shock in the second growing season; in all treatments increments were smaller than in the previous year and there was little difference among treatments. In 1989 treatment responses were much stronger; height growth was greatest in the clipped shrub and herb treatments, and lowest in the shrub control. Variation in the height of seedlings at the time of planting and the lack of treatment responses until the third growing season may explain the nonsignificant differences among treatments in total height in the cold stock (Table 3.2). 109 m Hot • Cold 2. 3. 4. 5. 6. Treatment Figure 3.3 Treatment means and standard error of the means for height:diameter ratios in two stock-types of Engelmann spruce after three growing seasons on site 1. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 110 1990 Treatments * X •it • A O 1 1990 Figure 3.4 Mean annual height (upper figure) and diameter (lower figure) increments in cold stock Engelmann spruce seedlings planted in 1987 in seven treatments on site 1. Standard error of the means were omitted for clarity but ranged from 0.26 to 0.85 cm for height increments and 0.09 to 0.29 mm for diameter increments. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. Diameter increments in cold seedlings from 1987 to 1989 are presented for the seven treatments on site 1 in Figure 3.4 (lower figure). Planting shock did not result in a decrease in increments from 1987 to 1988 (the decrease in increments in the herb control continued from 1988 to 1989, suggesting a treatment response). Treatment responses in 1988 were much stronger than those expressed in height growth. Whereas height increments in 1989 were lowest in the shrub control, diameter increments were lowest in the herb control, suggesting that the two types of vegetation did not have equivalent impacts on seedling growth (even though diameters did not differ significantly between the herb and shrub controls, means were smaller in the herbs). Seedling diameters in 1988 and 1989 were used to calculate the mean relative growth rates (RGR) of cold and hot stock on site 1. A comparison between stock-types showed that the RGR of hot seedlings exceeded that of cold seedlings in all treatments but the shrub control (Table 3.5). In both stock-types, RGR were 4 to 7 times greater in the clipped herb and shrub treatments than in the controls. Rates in the clipped skid trail treatment were only twice as large as those in the skid trail control. Within-treatment variation was very high, as indicated by the coefficients of variation (CV.) in Table 3.5. Variation in RGR, unlike total diameter, exhibited a distinct pattern among treatments; coefficients of variation were higher in the controls (ranging from 1.34 to 4.75 in the cold stock and 0.83 to 3.30 in the hot stock) than in the clipped treatments (0.24 to 0.63 in the cold stock and 0.23 to 0.67 in the hot stock) (Table 3.5). The number of laterals on seedlings after three growing seasons exhibited a pattern similar to that for diameter, both between sites and stock-types and among treatments (Figure 3.5 - upper figure). Treatment means and standard deviations ranged from 19.70 + 8.68 to 39.80 + 8.40 in the cold stock and 18.63 + 5.21 to 34.43 + 8.09 in the hot stock on site 1 (Table 3.6). Differences among means were significant for both stock-types, as Figure 3.5 Treatment means and standard error of the means for number of laterals on two stock-types of Engelmann spruce seedlings after three growing seasons on site 1 (upper figure) and length of laterals (cold stock only) after three growing seasons on sites 1 and 2 (lower figure). Lateral lengths were based on measurements of three top-most branches per seedling. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 113 tested in a nested ANCOVA (Table 3.7). Pairwise comparisons of means indicated significantly more laterals on seedlings in the clipped treatment than in the control within „,each of the three patch types, with the exception of the hot stock on the skid trails (Table 3.6). In addition, in the cold stock only, seedlings in the herb control had significantly fewer laterals than seedlings in the shrub control. This was the only growth variable to differ significantly between these two control treatments. The pattern among treatments in the length of laterals differed between sites 1 and 2. Results are presented for the cold stock in Figure 3.5 (lower figure). Treatment means were greater in the cold stock, ranging from 3.78 + 0.97 cm to 8.11 + 2.14 cm (Table 3.8). Means and standard deviations in the hot stock ranged from 2.68 + 0.93 cm to 6.72 + 1.63 cm (hot stock data are not presented). A nested ANCOVA indicated significant among-treatment differences on both sites (Table 3.9 - only the analysis for the cold stock is presented). Laterals in the clipped herb and shrub treatments were significantly longer than laterals in the herb and shrub controls (Table 3.9). Pairwise comparisons of means indicated nonsignificant differences between treatments within herb patches, but significant differences between treatments within the shrub patches. The major difference between the sites was in the response to the clipped and bare skid trail treatments. On site 1, seedlings in these treatments had significantly shorter laterals than seedlings in the clipped herb and shrub treatments, whereas on site 2 laterals in the clipped and bare skid trail treatments were as long or longer than those in the other clipped treatments. I Treatment means and standard deviations for needle length ranged from 11.16 + 2.13 mm to 16.90 + 2.45 mm in the cold stock, and 10.38 + 3.37 mm to 17.13 + 3.75 mm in V the hot stock. Results for the two stock-types were similar and only the resailts for the cold stock are presented here (Table 3.10, Figure 3.6 - upper figure). Differences among treatments were significant, and needles in the clipped herb and shrub treatments were 114 Treatment Figure 3.6 Treatment means and standard error of the means for length of needles (upper figure) and number of buds (lower figure) on cold stock Engelmann spruce seedlings after three growing seasons on sites 1 and 2, Needle lengths were based on three measurements per seedling, and bud numbers were based on number of buds on top-most three branches per seedling. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 115 longer than those in the herb and shrub controls (Table 3.11). As with lateral length, however, a pairwise comparison of means indicated nonsignificant differences between the clipped herb and herb control treatments, but significant differences between the clipped and control treatments within shrub patches (Table 3.10). Needle number per unit stem, bud number, and leader length were not tested for significant differences among treatments because these variables were not measured at the time of planting and covariates were significant sources of variation in the analyses of most other variables. However, among-treatment differences in two of these variables, bud number and leader length, were very similar to those for lateral length, as illustrated for bud number (Figure 3.6 - lower figure) and leader length (Figure 3.7 - upper figure) in the cold stock. Means and standard deviations for bud number ranged from 7.83 + 3.49 to 22.70 + 7.02 in the cold stock, and 6.37 + 3.06 to 23.27 + 7.62 in the hot stock. Treatment means and standard deviations for leader length ranged from 5.50 + 1.86 cm to 14.73 + 4.10 cm in the cold stock, and 3.31 + 1.29 to 12.85 + 2.96 cm in the hot stock. For both variables, treatment means were greatest in the clipped shrub and herb treatments, intermediate in the clipped and bare skid trail treatments, and lowest in the controls. Needle number per unit stem exhibited a pattern among treatments reversed to that for all other simple growth parameters; the number of needles per cm of leader was greatest in the herb, shrub, and skid trail controls, and least in the clipped treatments (Figure 3.7 -lower figure). Treatment means and standard deviations ranged from 3.27 + 0.74 to 5.23 + 0.93 in the cold stock, and 3.34 + 0.48 to 5.32 + 0.82 in the hot stock. 116 Site 2 Site 1 Figure 3.7 Treatment means and standard error of the means for length of leader (upper figure) and number of needles (lower figure) on cold stock Engelmann spruce seedlings after three growing seasons on sites 1 and 2. Needle number was based on the number of needles in one rank along a 1 cm length of the leader. n=30. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 1 1 7 Differences in needle number may have been largely a function of among-treatment differences in leader and needle lengths (suggesting an allocation trade-off between needle number and size). However, although needle number was significantly correlated with leader length (p <0.001) and needle length (p <0.001), the amount of variation in needle number accounted for by variation in these variables in a simple linear regression model was low (r^  was 0.359 for leader length and 0.376 for needle length in cold stock on site 1 - analyses not shown). Seedling biomass Total seedling biomass after three growing seasons exhibited similar patterns across treatments on both sites for the two stock-types. Treatment means for root and shoot biomass for cold and hot seedlings on site 1 are presented in Figure 3.8. Despite similarities in patterns between the two stock-types, cold seedlings had on average twice the biomass of hot seedlings within treatments. The magnitude of the difference was similar to that between the two stocks at the time of planting. Treatment means and standard deviations for total biomass ranged from 9.07 + 4.09 g to 39.91 + 14.67 g in the cold stock, and 3.48 + 1.09 g to 21.67 + 7.02 g in the hot stock on site 1 (Table 3.12). There were significant differences among treatments in both root and shoot biomass (Table 3.13 - only the results for cold stock on site 1 are presented). Shoot and root biomass were both greater in the clipped and bare treatments than in the controls. In addition to differences in the size of roots in the presence and absence of vegetation, \ , there were also some differences in the form of the root systems. Seedlings in poor condition had little root development outside of the plugs after the first three growing seasons and most new growth was limited to the base of the plug (Plate 3.4). Seedlings in 118 50 40 >S9 I 30 03 s o to .9 20 »—i T ) u u CO 10 1 1 r - r H-1. 2. 3. 4. 5. 6. 7. 30 "1 1—•• 1 1 r 20 d a o bo 5 10 U o CO • Shoot 11 Root 3. 4. 5. Treatment Figure 3.8 Treatment means and standard error of the means for root and shoot biomass in cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on site 1. Errors are based on total (root + shoot) biomass. Scales are the same for both figures but ranges are different. n=9. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 119 Plate 3.4 Engelmann spruce seedling (cold stock) in poor condition three growing seasons after planting in a shrub control plot on site 2. Note that most new root growth was from the base of the plug. There had been almost no lateral root growth during the first three years after outplanting. 120 good condition in clipped treatments also had limited lateral root growth but egress in general was greater and was distributed to a greater extent through the plug (Plate 3.5). In pairwise comparisons of means for shoot and root biomass, differences between clipped and control treatments within patch types were significant in herb and shrub patches, but nonsignificant on the skid trails. The biomass of hot seedlings was generally reduced to a greater extent by above-ground vegetation than the biomass of cold seedlings in herb and shrub patches, and shoot biomass was affected more than root biomass within both stock-types (Table 3.14). Shoot:root (S:R) biomass ratios are presented in Figure 3.9 for cold and hot seedlings in treatments on site 1. Means and standard deviations in the cold stock ranged from 2.66 + 0.53 to 6.09 + 1.37, and ratios in the hot stock ranged from 3.09 + 0.65 to 5.41 + 2.02 (Table 3.15). In both stock-types among-treatment differences were significant (Table 3.16). Contrasts of means indicated that ratios in the clipped and bare treatments were greater than ratios in the controls (Table 3.16). Pairwise comparisons of means indicated nonsignificant differences between clipped and control treatments within each patch type for the hot stock; for the cold stock, differences were significant only in the shrub patch (Table 3.15). Allometric relationships To determine whether treatments affected the relationship between growth and V biomass accumulation, data on seedling diameter and total biomass for each treatment were fitted to the simple allometric equation (loge-transformed): Plate 3.5 Engelmann spruce seedling (cold stock) in good condition three growing seasons after planting in a clipped herb plot on site 1. Although there was significantly more root egress from the plug in seedlings with better growth, there was still very little lateral root growth in comparison to the roots of naturally established seedlings (see Plate 4.1). Contrast root size and form with that of the seedling in poor condition in Plate 3.4. 1 2 2 o o u o o A 09 rt 7 -2 6 0 o 5 -S 4 -3 -0 m Hot • Cold Figure 3.9 Treatment means and.standard error of the means for ratios of shoot:root biomass in two stock-types of Engelmann spruce after three growing seasons on site 1. n=9. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5-clipped skid trail, 6=skid trail control, 7=bare skid trail. logeY = logea + (b*logeX) where Y = seedling biomass X = diameter The ordinary least squares regression procedure for linear models in SYSTAT was used to fit the data (Wilkinson, 1990). Scatterplots of the data points are presented by treatment for each stock-type in Figure 3.10 (both sites combined). Regressions for the skid trail control were nonsignificant in both stock-types, and the regression for the clipped herb treatment was nonsignificant in the cold stock. The slopes (b in the allometric equation) of the regression lines for the remaining treatments did not differ significantly from one another when tested for parallelism in an ANOVA model (using the F-ratio for the interaction between loge(diameter) and treatment). For the hot stock, p=0.190; for the cold stock, p=0.051 (p=0.568 for cold stock when the bare skid trail treatment was omitted from the analysis). There was large variability in r 2 among significant regressions, with values ranging from 0.373 to 0.897 in the cold stock, and 0.237 to 0.901 in the hot stock (the regression analyses are not presented). However, there were no patterns either between stock-types or between treatments in the proportion of variance in diameter explained by seedling biomass. To determine whether variation in S:R ratios among treatments could be accounted for by size differences alone, the allometric equation was used to study the relationship between shoot and root biomass. Regressions of loge-transformed S:R ratios on loge-transformed total seedling biomass for the cold and hot stock (sites and treatments combined) indicated that there was a statistically significant (p <0.001) linear relationship, but the r 2 were very low (0.181 in the cold stock, and 0.146 in the hot stock), suggesting only a weak relationship between the two variables. Scatterplots illustrate the large variability within the two populations (Figure 3.11). * 124 1 2 Seedling diameter (Ln mm) Figure 3.10 Scatterplots (by treatment) of loge-transformed total seedling biomass and seedling diameter for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings three growing seasons after planting on sites 1 and 2 (both sites combined). n=18. Treatments: l=clipped herb (0)> 2=herb control ( A ). 3=clipped shrub ( V ), 4=shrub control ( ^ ) , 5=clipped skid trail ( |>), 6=skid trail control ( Q ) , 7=bare skid trail (<^>). 125 1 2 3 4 Total seedling biomass (Ln g) Figure 3.11 Scatterplots of loge-transformed shoot:root biomass ratios and total seedling biomass for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings after three growing seasons on sites 1 and 2 (both sites combined). Regressions of loge-transformed shoot biomass and loge-transformed root biomass were significant for all treatments in the cold stock, and all but the herb control treatment in the hot stock. When tested for parallelism the slopes of the treatment regressions did not differ significantly in the cold stock (p=0.526), indicating that, despite large variation in the relationship there was no evidence to suggest changes in the relative growth rates of shoots in relation to roots in response to treatment effects (Figure 3.12 - upper figure). In the hot stock, a test for parallel slopes with all treatments except the herb control indicated a significantly steeper slope in the clipped skid trail treatment than in the other five treatments. The F-ratio in the ANOVA test for parallelism was significant when this treatment was included (p=0.025) and nonsignificant when it was excluded (p=0.272) (Figure 3.12 - lower figure, analysis not shown). This indicated that only for the hot stock in the clipped skid trail treatment was the relationship between the relative growth rates of the shoot and root significantly different from that of any other treatment (although the data points for this treatment ( £> ) in Figure 3.12 lie well within the range of values from the other treatments). Similar to the regressions of diameter and total biomass, the coefficient of determination varied widely with little pattern across treatments. In the cold stock r 2 ranged from 0.346 to 0.867, and in the hot stock it ranged from 0.441 to 0.878 (excluding the herb control treatment). Environmental variables Soil and air temperatures Soil temperatures at 10 cm in the seven treatments on sites 1 and 2 for days in early r June, July, and August in 1989 are presented in Figure 3.13. Patterns between sites differed on all three measurement dates. In June and July, soil temperatures were 127 60 a * 3 SO cd a .2 -° 2 o o xi CO -1 0 1 2 3 Root biomass (Ln g) Figure 3.12 Scatterplots (by treatment) of loge-transformed root and shoot biomass for cold stock (upper figure) and hot stock (lower figure) Engelmann spruce seedlings three growing seasons after planting on sites 1 and 2 (both sites combined). n=18. Treatments: l=clipped herb (O) . 2=herb control ( A )> 3=clipped shrub ( V ), 4=shrub control ( ^ ) , 5=clipped skid trail ( [>), 6=skid trail control ( • ) , 7=bare skid trail (<Q>). 128 1. 2. 3. 4. 5. 6. 1. 2. 3. 4. 5. 6. E3 Site 2 @ Site 1 Treatment Figure 3.13 Treatment means and standard error of the means for soil temperatures at 10 cm on June 6 (upper), July 4 (middle), and August 6 (lower) in 1989 on sites 1 and 2. Site 1, n=5. Site 2, n=3. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. higher on site 1 than site 2, with greater differences in June than July. In early August, the relationship was reversed and temperatures were greater on site 2, particularly in the skid trail treatments. There was no apparent reason for the differences between the sites. The range in the means for all treatments on both sites and on all three dates was narrow; values ranged from 7.3 to 13.0 °C on site 1, and 6.2 to 17.0 °C on site 2. On site 1, soil temperatures were highest in June in all but two treatments and lowest in all treatments in July. On site 2, temperatures were highest in all treatments in August while temperatures in June were similar to those in July. The low temperatures in July on site 1 (Figure 3.13 - middle figure) may have been caused by a period of heavy rain in late June. In the previous year temperatures were higher in July than in June on site 1 (data not shown), indicating large year-to-year variation in seasonal patterns. Snow remained on site 2 approximately one week longer than on site 1 in 1989, which may have accounted for the lower soil temperatures in June on site 2. Soil temperatures on site 1 were tested with a one-way ANOVA model for among-treatment differences on each of the three measurement dates (Table 3.17). On each date treatment means differed significantly and orthogonal contrasts indicated that temperatures in the clipped and bare skid trail treatments were significantly higher than in the controls (Table 3.17). In July, temperatures in the controls ranged from 6.50 + 0.25 °C to 7.94 + 1.70 °C across both sites, while in the clipped and bare treatments temperatures ranged from 8.33 + 0.29 °C to 12.15 + 0.86 °C. Air temperatures at 25 cm exhibited much less variation among sites, treatments, and measurement dates than soil temperatures (Figure 3.14 - for site 1 only). Patterns between measurement dates were similar in all treatments, except the bare skid trail. Temperatures were lowest in July, and highest in August (in the bare skid trail, the highest mean temperature was in June). On site 1 treatment means ranged from 17.8 to 130 M August 6 • July 4 §2 June 6 Treatment Figure 3.14 Treatment means and standard error of the means for air temperatures at 25 cm on three days in 1989 on site 1. Treatments marked by a star differed significantly with p <0.05. n=5 (June), n=3 (July.August). Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 131 22.8 °C. Differences among treatments in July and August were nonsignificant (analyses not shown), and in June only the highest (22.1 °C in the bare skid trail treatment) and the „lowest (19.6 °C in the clipped shrub treatment) means differed significantly. Soil moisture content Soil moisture contents (expressed as % of dry weight) in August 1988 are presented in Figure 3.15 for the seven treatments on sites 1 (upper figure) and 2 (lower figure). Patterns across treatments differed between sites, and the within-treatment variation was large. Treatment means and standard deviations ranged from 46.66 + 15.02 % to 100.84 + 25.90 % on site 1 and 47.07 + 13.63 % to 96.37 + 42.23 % on site 2. Across both sites coefficients of variation ranged from 0.13 to 0.84. F-ratios for treatment effects in a nested ANOVA were significant only for site 1. However, on both sites moisture contents in the herb treatments were significantly greater than moisture contents in the shrub treatments (Table 3.18). In contrast to most other variables, moisture contents were more strongly related to patch type than treatment. On neither site were there significant differences between clipped and control treatments within the herb or the shrub patches (Table 3.18). On site 1 only, moisture contents were significantly higher in the skid trail treatments than in the off-skid trail areas. / Nitrate and ammonium ion concentrations V Resin bag nitrate ion concentrations are presented for the seven treatments on site 1 in Figure 3.16 (upper figure). Sample sizes varied considerably among treatments due to the destruction of bags by small mammals in some plots. Among-treatment effects were 132 120 | i i i i i i r 1. 2. 3. 4. 5. 6. 7. 120 I 1 1 1 \ 1—:—i r 100 -1. 2. 3. 4. 5 6. 7. Treatment Figure 3.15 Treatment means and standard error of means for soil moisture content (expressed as % of dry weight) in August 1988 on site 1 (upper figure) and site 2 (lower figure). n=12. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. 133 600 | i 1 1 1 1 1 r 500 -a P. • D. Treatment Figure 3.16 Treatment means and standard error of the means for N O ^ ' and N H 4 + ion concentrations in resin bags incubated for 14 months on site 1. Due to the destruction of some bags by small mammals n varies from 1 to 8 with an average of 6 per treatment (n=l for N H 4 + in treatment 5). Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. nonsignificant when tested with a one-way ANOVA (p=0.096 - analysis not shown). Treatment means and standard deviations ranged from 23.60 + 22.90 ppm in the shrub control to 403.67 + 351.18 ppm in the clipped skid trail treatment. Similar to moisture contents, an orthogonal contrast showed that nitrate concentrations were significantly greater in the herb treatments than in the shrub treatments (Table 3.19a). Despite a high treatment mean in the clipped skid trail treatment, nitrate levels were not significantly greater at this location when compared with the means for all other treatments. The large nitrate values in some skid trail plots were associated with the presence of high Epilobium angustifolium cover and high moisture contents. Resin bag ammonium ion concentrations are presented for the seven treatments on site 1 in Figure 3.16 (lower figure). Patterns across treatments differed from those for nitrate but were also characterized by high variability both within- and among-treatments. Treatment means and standard deviations ranged from 2.90 + 2.07 ppm in the bare skid trail treatment to 56.79 + 54.00 ppm in the clipped shrub treatment. As with nitrate, the F-ratio for treatment effects (excluding the clipped skid trail treatment where n=l) in a one-way ANOVA was nonsignificant (analysis not shown). However, an orthogonal contrast indicated significantly higher ammonium levels in the clipped shrub treatment when compared with all other treatments (Table 3.19b). Litter decomposition Decomposition (expressed as a % loss in initial weight) of mixed herb and mixed shrub litter in litter bags over a period of 14 months is presented for the seven treatments on site 1 in Figure 3.17. In all treatments but the bare skid trail weight loss, was greater in the bags of herb litter. Treatment means and standard deviations for the shrub litter bags 135 to Cfl O 6 0 ••-I O !* t-i 90 80 70 60 50 40 30 20 10 0 El p a Shrub Herb 2. 3. 4. 5. 6. Treatment Figure 3.17 Treatment means and standard error of the means for weight loss in bags of herb and shrub litter incubated for 14 months on site 1. Due to the destruction of some bags by small mammals, n varies from 4 to 9 with an average of 6 per treatment. Treatments: l=clipped herb, 2=herb control, 3=clipped shrub, 4=shrub control, 5=clipped skid trail, 6=skid trail control, 7=bare skid trail. ranged from 45.02 + 12.61 % in the bare skid trail treatment to 60.71 + 14.24 % in the shrub control. Means and standard deviations for the herb litter bags ranged from 46.19 + 21.58 % in the bare skid trail treatment to 73.55 + 5.68 % in the herb control. Although patterns among treatments were similar for both litter types, among-treatment differences were significant only for the herb litter (Table 3.20). However, orthogonal contrasts for both litter types showed that the weight loss in litter bags on bare skid trails was significantly lower than weight loss in all other treatments. Decomposition of Valeriana sitchensis (herb) litter and Menziesia ferruginea (shrub) litter from October to June (overwinter) in the herb control and shrub control treatments only, showed significant differences between litter types but not between treatments (Table 3.21). Weight loss over this period was higher in the Valeriana litter bags, with means and standard deviations of 75.95 + 6.10 % and 72.10 + 7.37 % in the herb and shrub controls, respectively. These values are similar to those for the mixed herb litter bags incubated for 14 months. Rates of decomposition were lower in Menziesia litter bags with weight losses of 46.74 + 7.06 % in the herb control and 42.89 + 2.25 % in the shrub control. These values were approximately 25% lower than those for the mixed shrub litter bags. V i DISCUSSION Seedlings of both stock-types showed strong responses to the presence and absence of vegetation in the three patch types on the Vavenby sites. Differences between treatments in seedling growth and survival were related to differences between patches in various biotic and abiotic characteristics, although the environmental heterogeneity was greater than the variation in seedling responses. The prediction that seedling growth would be greater in herb patches than shrub patches was not supported by the results of this study. Neither was the prediction that the hot-planted stock would outperform the cold-stored stock. After three seasons of growth in the cold stock and two in the hot stock, differences between the two were similar to those at the time of planting. There was little evidence of allocation shifts within seedlings, and most variation was related to differences in size. How did spruce seedlings respond to interference from non-coniferous vegetation? All seedling characteristics with the exception of total height showed large among-treatment differences in both stock-types. The strongest responses were to the presence or absence of above-ground vegetation. No positive effects of vegetation on growth were detected. The size or number of all components were significantly reduced in the presence of herbs, shrubs, and skid trail vegetation. The higher relative growth rates of seedlings in the clipped treatments suggest that differences between the growth of seedlings with and without interference from non-crop species will increase over time. i Better growth in the absence of above-ground vegetation was likely due to the significantly higher light levels and soil temperatures in the clipped treatments. 138 Total height was the only character that showed nonsignificant differences among treatments. Trends in height increment suggest that the reason for this was nursery and planting shock effects on height growth in the first two seasons after outplanting. In addition, etiolation in response to low light availability in the control plots may also have decreased absolute differences among treatments. Results from this and other studies (Coates, 1987; Elliot and White, 1987; Wagner, 1989) indicate that height is not a good measure of treatment response in the first few years after planting. Hot-planted seedlings did not exhibit less of the effects of interference from non-crop species on the Vavenby sites than the cold-stored stock. On average, the differences between the two stock-types at the time of planting were maintained over the three growing seasons. However, there was some evidence that the hot stock responded more negatively to competitive interference, and the cold stock responded more positively to competitive release. This was suggested by the following observations: 1) the better condition, particularly in the clipped treatments, of the cold stock after three growing seasons; 2) consistently higher height:diameter ratios in the hot stock in the controls; 3) generally higher clippedxontrol ratios in root and shoot biomass in the hot stock, and; 4) increased differences between stock-types in the controls when compared to differences at the time of planting, but similar differences in the clipped treatments. Although treatment means for the relative growth rates of diameters were slightly higher in the hot stock, differences between stock-types are unlikely to be significant, given the large / variation in this measurement. The differences that did occur between the two stock-types may have due mostly to the smaller size of the hot stock at the time of planting. V Unfortunately, in this study the effects of other factors such as genotypic variation and nursery experiences that may have influenced the responses of seedlings could not be isolated out from the effects of environmental variation. 139 In general, seedling responses to the herb and shrub vegetation were very similar. ~ Among-treatment differences in height increments and diameter increments did suggest that height growth was more adversely impacted by the shrubs than the herbs, whereas the opposite was true for diameter growth. These differences, due perhaps to the lower light levels above 25 cm beneath shrub canopies, were not statistically significant after three growing seasons although they may become more pronounced with time. Similarities in height:diameter ratios in seedlings in both the herb and shrub controls on the Vavenby sites suggest that low light availability was the dominant factor limiting early spruce growth. After three growing seasons, the spruce seedlings were still within the influence of both herb and shrub canopies. Higher levels of mortality and the lower number of laterals on cold-stored seedlings in the herb vegetation indicate the adverse effects of the low, dense herb cover on early seedling growth and survival. The lower number of laterals on the cold stock was probably due to the death of laterals after flushing rather than to a reduction in the number of laterals produced by the seedling. Soon after flushing laterals would have been particularly vulnerable to the effects of shading and physical damage from the dense herbaceous vegetation. Seedling growth in patches of herbs and shrubs may diverge markedly once seedlings grow above the herb canopy, provided that they can survive to this stage. Ideally a study should monitor seedlings to this point to determine the recruitment of trees from within herb and shrub patches. } How did spruce seedlings respond to the removal of interference from non-V coniferous vegetation? t i Although differences between the clipped herb and clipped shrub treatments were not statistically significant for any of the characters measured, there was a distinct trend towards greater growth in the clipped shrub treatment. Of the environmental variables measured, only resin bag ammonium ion concentrations were significantly greater in this treatment than in all others. Several authors have found that seedlings of other conifer species preferentially take up ammonium nitrogen over nitrate nitrogen in wet soils (Gijsman, 1991 and references therein). The preference of ectomycorrhizae for ammonium nitrogen may also be responsible for the differences among treatments (Richards, 1987). It is interesting that ammonium concentrations were much lower in the presence of shrubs, suggesting that the ericaceous shrubs (or their mycorrhizae) were very competitive for soil nitrogen. Removal of the shrubs may have increased the availability of nitrogen to soil microbes as suggested by increases in both ammonium and nitrate ion concentrations in the clipped shrub treatment. In contrast, the presence or absence of vegetation had no impact on the ammonium concentrations in the herb patches, which were nonetheless higher than levels in the shrub controls. Higher concentrations of nitrate and ammonium nitrogen in herb treatments than shrub treatments (with the exception of ammonium in the clipped shrub plots) were not associated with relatively greater seedling growth in herb patches. The higher potential mineralizable nitrogen levels in the herb patches relative to the shrub patches (as described in Chapter 2) were also not associated with greater seedling growth. Differences between skid trail and off-skid trail treatments were more strongly manifested in the absence of vegetation. In general, the growth responses of seedlings in the skid trail control were indistinguishable from those in the herb and shrub controls, supporting the role of light as a dominant limiting factor in these environments (the significantly higher soil temperatures in the skid trail control compared to \ht other controls were not associated with greater growth or biomass in this treatment). The poorer general condition of seedlings in the clipped and bare skid trail treatments than in the clipped herb and shrub treatments suggests that the skid trails were a poorer —environment for growth. This was reflected to a much greater extent in differences in biomass than in differences in size, and also to a greater extent on site 1 than site 2. Most environmental measurements were made only on site 1, and these measurements do not provide an explanation for the poorer growth in the clipped treatment on the skid trails on this site, unless differences are related to the much higher soil moisture contents at this location than in the clipped treatments off the skid trails. Engelmann spruce does not grow well in saturated soils (Alexander, 1987) and the poorest growth among clipped treatments on both sites was associated with the highest soil moisture contents. What was the relationship between growth and mortality in the planted spruce seedlings? Interference from non-coniferous vegetation increased mortality in Engelmann spruce seedlings in the same way that interference reduced growth. However, few seedlings died over the first three growing seasons and most mortality occurred in the herb patches. Very low light levels near the ground beneath the dense, herb vegetation may have been particularly detrimental to the cold-stored seedlings, which flushed late in the year of planting and had only a short time in which to complete their growth in the first season. In late summer, the die-back of herbs in the herb and skid trail controls crushed many seedlings. Similarly, the weight of snow on the stems of shrubs also crushed seedlings in the shrub control treatment. However, seedlings in the shrub patches straightened up the following spring when shrub stems were released from snow cover. In contrast, some seedlings smothered beneath the stems of dead herbs never released and subsequently died. 142 Coates (1987) also found high survival rates in planted Engelmann spruce seedlings " after two growing seasons on a cutover in the ESSFwc2. The lowest survival rate (81.5%) on his site was beneath a mixed canopy of herbs and shrubs (composed of Valeriana sitchensis, Rhododendron albiflorum, and Menziesia ferruginea), but the causes of mortality were not identified. Summer frost damage to seedlings the year of planting did not seem to be associated with subsequent mortality although it did affect growth the following year. It is unlikely that frost damage contributed to mortality in any treatments on the Vavenby sites: no symptoms of damage were visible on the seedlings prior to death and air temperatures below 0 °C were not recorded during the 1988 and 1989 growing seasons. What are the patterns of early growth in Engelmann spruce? The growth of the Engelmann spruce seedlings on the Vavenby sites was good compared to the growth of seedlings of comparable age on other sites in the ESSF (Bassman, 1989; Burdett etal., 1984; Coates, 1987; Vyse, 1981). The height and diameter of seedlings growing with herbs and shrubs (control treatments) on the Vavenby sites were as great or greater than that of seedlings of several stock-types growing under various conditions in these other studies, including mounding and scarification treatments (Bassman, 1989), different levels of vegetation removal (Coates, 2987); plug and ^ bareroot plantations (Vyse, 1981), and fertilization (Burdett et al., 1984 - although the diameters of the seedlings in their study were greater than those of the Vavenby stock-^ types). Seedlings growing without above-ground vegetation (clipped treatments) on the Vavenby sites were considerably larger than seedlings in any of the treatments in these other studies. Third-year height increments in the clipped treatments on sites 1 and 2 143 were similar to third-year height increments in fertilized seedlings in Burdett et aVs (1984) study; increments in the controls were similar to the unfertilized seedlings (which ~ were growing without competitive interference in their study). Factors which may have contributed to the good seedling growth on the Vavenby sites include an insulating winter snow cover, no summer frost events, and beneficial effects of post-logging changes in microclimatic and edaphic factors on nutrient availability. Cole and Newton (1987) suggested that height:diameter ratios were good indicators of both overtopping and encroaching vegetation. They found similar increases in ratios in Douglas-fir seedlings with increasing densities of grass, red alder, and Douglas-fir. They found that trees with ratios above 100 had poor vigour, and that ratios above 70 were associated with decreased height growth. Maximum height growth occurred at ratios of 65 to 70. On the Vavenby sites, average maximum height:diameter ratios did not reach 65, and the best height growth was associated with ratios of approximately 40. This suggests that ratios differ for different species, and they may also differ with other factors, such as site conditions or genotype. Neither Burdett et al. (1984) nor Coates (1987) found differences in heighttdiameter ratios in Engelmann spruce seedlings in response to differences in nutrient availability (Burdett et al.) or levels of competition (Coates). Cole and Newton (1987) suggest that ratios may be used as a guide for the selection of plantation treatments, but the results from the studies mentioned above indicate that this will not be possible without a better understanding of the factors controlling ratios, and variation in ratios among and within species. The results of this study do not support the idea that competitive release allows greater * expression of inherent genetic variability among individuals and increases In size inequality of populations of Engelmann spruce. Lieffers and Titus (1989) observed such an increase in white spruce seedlings when intraspecific competition for soil nutrients was alleviated by fertilization. Several other authors have observed increases in variation among conifer seedlings in response to competitive release (Coates, 1987; Daniels et al., 1986; Wagner, 1988). In the treatments on the Vavenby sites, within-treatment variation (as measured by coefficients of variation) varied widely across treatments but, with one notable exception, showed no consistent patterns between clipped and control treatments (or among treatments in general). This exception was in the relative growth rates (RGR) calculated from stem diameters. Coefficients of variation for RGR in the control treatments were consistently greater than those in the clipped treatments - indicating, in contrast to observations in the studies cited above, a decrease in variation in response to competitive release. That RGR was the only variable to exhibit this pattern suggests that size-related effects obscure or confound treatment effects on variation. Increased variability in the controls could be explained by the large point-to-point variation in light availability within plant canopies, as illustrated by the light profiles presented in Figure 2.6. The other environmental variables measured in this study did not exhibit consistent patterns of variation between clipped and control treatments. The results from this study indicate that biomass allocation strategies in planted Engelmann spruce seedlings do not change in response to changes in resource availability. In the presence of vegetation, low light availability resulted in reduced growth of all plant parts. The higher light availability (and other resource changes) resulting from the removal of above-ground vegetation increased the growth of all components. The increases in S:R ratios in clipped treatments were a consequence of increases in seedling size and differences between the relative growth rates of shoots and roots. The predicted shift in allocation to below-ground components in response to the removal of above-ground interference from non-crop species did not occur. However, rates of biomass turnover were not measured and, although they are assumed to have been low because of limited root growth and needle loss during the first three growing seasons after planting, nevertheless it is important to note that allocation to mycorrhizae and fine roots can not be determined from estimates of total biomass. Assuming that most allocation was captured in measurements of total biomass, there are a number of possible reasons why shifts did not occur between root and shoot biomass. Among them are: 1) seedling growth after three seasons was still heavily dependent on nutrients in the root plug; 2) below-ground resources (outside of the plug) were not limiting to growth after the removal of above-ground vegetation, or; 3) root growth was limited by soil physical factors such as low temperatures. However, perhaps shifts in allocation are simply not a plastic response to environmental variation in a species such as Engelmann spruce which normally grows under conditions where microclimatic factors strongly influence rates of production. The evidence from this study does suggest that allocation in Engelmann spruce is controlled, but not tightly controlled, as there was large scatter in the relationship between seedling components. CONCLUSIONS The results from this study suggested that, in the presence of non-coniferous vegetation low light availability was the dominant factor limiting the growth of planted Engelmann spruce seedlings. Total growth was much greater when light was removed as a limiting factor, but patterns of allocation among above-ground components and between above- and below-ground components were not altered. This suggests that growth in Engelmann spruce seedlings is not plastic, and that improving the performance of seedlings after outplanting will depend strongly on the growth capacity of the planting stock and maximizing the level of resources available to seedlings. When light availability was low seedling responses were very similar although herbs had a somewhat greater impact on diameter and lateral growth whereas shrubs had a greater impact on height growth. When light levels were high, variation in substrate characteristics such as moisture contents and ammonium concentrations were associated with differences in seedling responses, although differences were not statistically significant after three growing seasons. It is suggested that strategies for resource conservation rather than resource acquisition are more consistent with the observed responses of Engelmann spruce seedlings to environmental variation, including that associated with competing vegetation. However, interpretations of the responses of seedlings after outplanting are difficult without some understanding of the effects of the characteristics of planted seedlings on field performance. To explore these effects, the growth of planted seedlings in the clipped and control treatments on site 2 were compared with those of naturally established seedlings in several environments on the cutover and in the understorey of an old-growth spruce stand. The results of this study are presented in the following chapter. CHAPTER 4 A COMPARISON OF THE CHARACTERISTICS OF PLANTED AND NATURALLY ESTABLISHED SEEDLINGS OF ENGELMANN SPRUCE ON A CUTOVER AND IN AN OLD-GROWTH SPRUCE-FIR STAND IN THE ESSFwc2 INTRODUCTION Planting Engelmann spruce is currently the preferred method of regenerating high elevation forests in the interior of British Columbia, but in the United States spruce-fir forests are managed for the natural regeneration of spruce (Alexander, 1987). The poor performance of seedlings in plantations in B.C. suggests that planting may not be the most appropriate stategy for reforestation. Nevertheless, general observations have shown that young planted seedlings grow faster than naturals (Butt, 1989; Mather, 1987; Vyse, 1981). On the Vavenby sites, seedlings planted in vegetation-free plots grew well after recovering from planting shock, but those planted in plots with herbs and shrubs grew poorly and their survival is uncertain. In both groups of seedlings poor root development during the first three years after planting may significantly limit establishment on the cutovers. The effects of nursery, storage, and planting experiences can negatively affect seedling performance after outplanting, with adverse impacts on both survival and growth under some conditions (Lavender et al., 1990). If it is assumed that naturally established seedlings are well-adapted to the environments in which they are growing, then a comparison between the characteristics of naturally established and 148 planted seedlings may indicate the degree to which the latter are also well-adapted to sites. Literature Review In the planted spruce experiment described in Chapter 3, patterns of allocation within seedlings were found to be similar regardless of environmental conditions, although total growth increased with increasing resource availability. There are no studies of variation in resource allocation in naturally established Engelmann spruce with which to compare the results from the planted spruce experiment. However, Scagel (1984) studied variation in individual morphological and anatomical characters of Engelmann spruce trees from southern B.C. In a comparison of levels of variation within and among individuals, and among populations, he found the highest variation to be associated with individual trees: variation decreased from within individuals to among individuals to among populations. Patterns within and among individuals were similar, and there was a high degree of correlation among characters. There was only a weak overall correlation between seedling and environmental variation at the population level, although the relationship between the two sets of variables was stronger for nursery grown populations than for natural populations of Sitka spruce (Picea sitchensis (Bong.) Carr) (Engelmann spruce populations were not tested for this source of variation). Higher within-population than among-population variation in morphological variables . has been found in other conifer species (references in Funck et al., 1990 and Scagel, 1984). Funck et al. (1990) found little variation among five-year-old seedlings from four ^ provenances of Pacific silver fir (Abies amabilis (Dougl.) Forbes) grown at one location on Vancouver Island. Average internodal length (mean distance between needles over a unit of stem) was the only variable to differ among provenances, with the largest length associated with the provenance from the location most similar to the test site. This result would be expected if, as Funck et al. (1990) suggested, internodal length assesses the efficiency of expression of predetermined growth potential and it is assumed that the closest provenance comes nearest to expressing its full potential on the test site. The correlations among morphological variables differed within each provenance, although there were some consistent relationships. Particularly, needle lengths on lateral branches showed a similar relationship in correlation coefficients among provenances and one that was different from most other variables. Funck et al (1990) suggested that in trees under stress, environmental constraints limited the expression of innate genetic variation. High within- and among-individual variation in Engelmann spruce trees might be an adaptive response to heterogeneity in microsite conditions (as suggested by Scagel, 1984). The relationship between variation in the growth of Engelmann spruce seedlings and environmental conditions in northern forests is unknown. Knapp and Smith (1982) showed that the distribution of Engelmann spruce seedlings in the understorey of a mature spruce-fir forest in Wyoming was associated with thin litter layers. Their study indicated that early root growth in Engelmann spruce seedlings under controlled conditions was very slow, and they hypothesized that seedlings were restricted to rooting substrates that did not dry out early in the summer. The distributions of seedlings of other conifer species have also been shown to be correlated with thin litter or moss layers in forest understories (Collins and Good, 1987; Harmon and Franklin, 1989; Shirley, 1945). Collins (1990) studied the relationships between tree seedlings of different ages and their microhabitats in an eastern hemlock-hardwood forest and found that older seedlings of some species were associated with significantly different microhabitats;than germinants or very young seedlings. He suggested that studies of older see'dlings could not completely incorporate the regeneration strategy of a species because they could not detect the effects of environmental heterogeneity on the survival of younger trees. Similarly, Scott and Murphy (1987) found that the microenvironments suitable for the regeneration of northern white cedar (Thuja occidentalis L.) in forests established on sand dunes changed as the stands matured. Early in stand development seedlings established on open sandy soils, whereas later in development seedlings occurred only on fallen logs. Late in stand development larger seedlings (> 25 cm) were found only on logs in canopy gaps where light levels were higher. In many mature and old-growth forests tree seedlings are most commonly found rooted in or on fallen logs (Harmon et al, 1986). In the Pacific northwest the phenomenon has been observed in coastal Douglas-fir, cedar-hemlock, and Sitka spruce forests, as well as high-elevation Engelmann spruce-subalpine fir forests, and northern white spruce forests (Christy and Mack, 1984; Harmon etal, 1987; Harmon and Franklin, 1989; McCullough, 1948). Engelmann spruce seedlings are not restricted to logs in mature forests and do establish on other substrates such as dead wood, mineral soil, and leaf litter (Alexander, 1984; Griffith, 1931; Lowdermilk, 1925; Stahelin, 1943). This is also true for other conifer species, although logs do appear to provide the largest source of safe sites (sensu Harper, 1977) for early establishment in many forests. There have been few detailed studies of the factors involved in seedling establishment on logs: among those factors which have been associated with establishment in these microsites are reduced competition from forest floor vegetation (Harmon and Franklin, 1989), increased light availability in elevated microsites (Scott and Murphy, 1987), vigorous mycorrhizal growth in decayed wood (Eide, 1926 - cited in Baldwin, 1927), and higher moisture contents in rotten wood (Amaranthus et al, 1989). Although establishment may be favoured on fallen logs in mature forests, low nutrient availability in dead wood might limit seedling growth in these microsites. However, the 151 nutrient relations of seedlings growing on logs remains to be studied. As Engelmann spruce can establish in the understorey of mature stands (Johnson and Fryer, 1990; Veblin, 1986), seedlings may be adapted to growth under low nutrient conditions. Studies on several other northern spruces have shown such adaptations in growth and physiological characteristics. Munson and Timmer (1990) found increases in nutrient-use efficiency in black spruce (Picea mariana (Mill.) B.S.P.) seedlings in response to lower nutrient availability on sites in northern Ontario. Under conditions of greater nutrient stress seedlings also allocated more growth to stems and roots, and relatively less to current growth than seedlings growing on richer sites. The higher allocation to stems under low nutrient conditions was consistent with the results of Sheppard and Cannell (1985), who studied nutrient-use efficiency in eight-year-old clones of Sitka spruce. They found that higher efficiencies were strongly related to the relative proportion of dry matter allocated to stems as opposed to branches and needles. Other characteristics associated with high nutrient-use efficiency were long needle retention and low nutrient concentrations in the needles. The presence of these characteristics in Engelmann spruce seedlings would be consistent with the conservation strategy hypothesized for this species in Chapter 3. However, only the first characteristic (that of within-seedling allocation) could be examined with the data collected in this study. The objective of this study was to compare six-year-old planted and naturally established seedlings for similarities and differences in their growth and allocation and in the characteristics of the environments in which they were growing. The study was intended to provide greater insight into the patterns of growth and allocation in planted spruce seedlings that were described in Chapter 3, although the results are also discussed in terms of the regeneration strategy of Engelmann spruce. From a silvicultural viewpoint these results will enable inferences to be made about the suitability of container planting stock for ESSF cutovers. 152 METHODS General description of natural regeneration In July 1990, naturally established seedlings of Engelmann spruce were selected for sampling from shaded and unshaded microsites on site 2 and from the understorey of a nearby old-growth spruce-fir stand. Planted seedlings from the clipped and undipped treatments on site 2 (these treatments were described in Chapter 3) were collected at the same time for a comparison of the growth characteristics of planted and naturally established seedlings. Prior to sampling on the cutover and in the old-growth stand, a general survey of the amount, size, and species composition of the natural conifer regeneration was undertaken on both sites. The following information was collected in five 0.02 ha plots which were systematically located at the centre and the four corners of a randomly situated 50 m x 50 m quadrat on each site: - species, height, diameter (at 1.3 m), condition of all trees > 1.3 m dbh; - species, height, condition of all trees between 0.05 m and 1.3 m; - percent cover of shrubs and herbs by species, and total precent cover of moss; - age (based on cores taken with an increment borer) of three dominant Engelmann spruce or subalpine fir in each plot. Seedling selection The three locations from which naturally established seedlings were collected were: 1. The understorey of the open, old-growth spruce-fir stand (hereafter the seedlings from this location are referred to as the forest naturals). In the understorey of the forest, seedlings occurred only on fallen logs in locations where overstorey shading was high and understorey shading was low. Consequently all seedlings from the forest were treated as one group. 2. Areas with herb and shrub vegetation adjacent to skid trails on the cutover (hereafter seedlings from this location are referred to as the shaded naturals). On the cutover, most post-logging spruce regeneration had established in disturbed areas on or adjacent to skid trails so sampling was carried out in these areas. 3. Open areas of mineral soil exposure adjacent to skid trails on the cutover (hereafter seedlings from this location are referred to as the open naturals). Planted seedlings were collected from the following two locations: 4. The herb, shrub, and skid trail control plots on site 2 in the planted spruce experiment described in Chapter 3 (this group is hereafter referred to as the shaded planted seedlings). 5. Clipped herb, shrub, and skid trail treatment plots in the planted spruce experiment on site 2 (this group is hereafter referred to as the open planted seedlings). Fifteen seedlings were selected from each location in the following manner. 154 a. Forest naturals In the centre of the spruce-fir stand, a compass line was set perpendicularly to the slope of the site and at 5 m intervals the acceptable seedling nearest to the line was tagged for sampling. Acceptable seedlings were those six years of age (from seed) and free from obvious disease, insect or physical damage. Six-year-old seedlings were chosen because the planted spruce were this age at the time of sampling. Several seedlings estimated to be six years old in the field were found to be seven or eight years old after re-examination in the lab but were included in the analysis to maintain equal sample sizes. b. Cutover naturals A skid trail situated approximately mid-site was traversed in a manner similar to that for the compass line in the forest. Acceptable seedlings were located at 2 m intervals rather than 5 m intervals due to a higher density of naturals on or adjacent to the skid trail. In addition to the age and condition criteria, acceptable seedlings were those either in fully exposed locations on the skid trail (open naturals), or beneath a canopy of herbs or shrubs on the edge of the skid trail (shaded naturals). c. Planted seedlings The details of the planted spruce experiment are described in Chapter 3. For this study, five seedlings were selected (randomly) from one plot (also randomly selected) in each of the three control treatments (i.e. treatments with undisturbed herb, shrub and skid trail vegetation) on site 2. These formed the shaded planted seedling group. Similarly, 155 five seedlings were collected from one plot in each of the clipped herb, shrub and skid trail treatments on site 2. These constituted the open planted seedling group. Sampling methods The following measurements were made on each seedling: - height to the base of the terminal bud; - diameter 5 em from the base of the tree; - total number of laterals; - length of the topmost three laterals; - number of buds on the topmost three laterals; - number of needles in one rank along a 1 cm length of the leader; - length of three needles 1 cm below the tip of the leader; - length of the 1990 leader; - length of height increments for the previous three years. Instantaneous light readings (umol/m^/sec) were recorded with a LI-COR quantum sensor and LI-1000 Datalogger (LI-COR Inc., Lincoln, Nebraska) at the top of the seedling, 2 m above the seedling and if necessary, in full sunlight. Readings were taken between 1000 and 1600 h under mainly sunny conditions. Soil temperatures were measured at a depth of 10 cm beneath each seedling with an EL504-026 Portable Dial Gauge Thermometer (ELE International, Hertfordshire, England). A circular quadrat (wire ring) of 15 cm radius was centered on each seedling. Litter depth was measured and litter and rooting substrate composition were described within the quadrat. The percent 156 cover of each plant species was recorded, along with the amount of dead wood in the quadrat. After excavation with a hand trowel, seedlings were separated into roots and shoots and stored for biomass determinations. Roots were cleaned, dried at 70 °C for 48 h and weighed on a Mettler PC440 balance (Mettler Instruments, Hightstown, New Jersey). Shoots were also dried at 70 °C for 48 h, sorted into needles, stems and buds, and redried and weighed on the Mettler balance. A l l of the naturals were harvested but due to time constraints only two planted seedlings (randomly selected) were harvested in each plot, for a total of six seedlings per group. Data analysis Bar graphs were used to investigate differences among groups in the means of the simple and derived size and biomass variables. One-way A N O V A was carried out on group means to test for significant differences among groups. In all cases, significance was tested at a=0.05. Data transformations were carried out prior to the A N O V A on those variables for which Bartlett's test for homogeneity of variance yielded significant results. Based on the data analysis discussed in Chapter 3, a square-root transformation was applied to size variables and a logarithmic transformation to biomass variables. Generally, derived variables did not require transformation. Tukey's test for HSD was used for pairwise comparisons of means. Relative production rates (RPR) were calculated for each seedling from the equation (Brand etal, 1987): [loge(HIj) - loge(HIj_i)] / (timej - timej.j) where HIj = 1990 height increment HL i = 1989 height increment time: = 1990 timej.! = 1989 This is different from the approach used in Chapter 3, where relative growth rates were calculated based on 1988 and 1989 seedling diameters. Relative production rates were not derived for the seedling data analysed in Chapter 3 because it was felt that seedling responses in 1987 and 1988 were confounded by nursery and planting shock effects, and 1990 measurements were not available for all trees. A Pearson product-moment correlation matrix was constructed to describe the relationships among environmental, size, and biomass variables. Bonferroni-adjusted probabilities were used to examine the significance of the correlations in the matrix. Principal components analysis was carried out on the correlation matrices of growth and environmental variables to describe and summarize patterns of variation in the original data sets. In PCA, the observations on the original variables are transformed to observations on a new set of independent variables that are linear additive functions of the original ones. The number of components derived was defined by an eigenvalue greater than 1. Scatterplots of the component scores for each seedling on the first two P C A axes were used to illustrate among-group differences in growth and environmental characteristics. A l l statistical analyses were carried out using the SYSTAT Version 5.0 statistical package (Wilkinson, 1990). RESULTS General description of natural regeneration The old-growth forest was situated at the same elevation as the Vavenby sites (1550 m) and had a north-northwestern aspect and a 30% slope. Stand characteristics were summarized in Table 1.1. The understorey vegetation consisted of patches of herbs and shrubs that were similar in structure and composition to those on the cutovers (as described in Chapter 2) although in the forest, shrub patches were more abundant than herb patches. The forest stand was open and the canopy cover was approximately 55%. There were many dead standing trees and over-mature individuals with broken tops scattered throughout the stand. As on the cutovers there was a large amount of coarse woody debris, mainly in the form of large fallen trees on the forest floor. There were approximately twice as many stems per hectare of subalpine fir as there were of Engelmann spruce in the old-growth spruce-fir stand. There were more trees > 1.3 m in the forest than on the cutover, and on the logged site there were equal numbers of subalpine fir and Engelmann spruce (spruce was selectively removed during logging but there was extensive post-logging mortality of suppressed and damaged fir trees). Subalpine fir individuals less than 1.3 m were much more abundant than spruce on the cutover and many of these were less than 10 cm, suggesting higher post-logging recruitment of fir. In the forest, advanced regeneration of spruce and fir had a very patchy distribution. Some clumps of young trees were in canopy gaps, while others were clustered around mature trees. Engelmann spruce less than 50 cm were uncommon in the forest, and many seedlings of this species were in poor condition. 159 Seedling size Seedling height and diameter ranged from means and standard deviations of 17.8 + 5.1 cm and 2.8 + 0.8 mm, respectively, in the forest naturals to 46.1 + 6.3 cm and 10.6 + 1.9 mm in the open planted seedlings (Figure 4.1). Standard error of the means are used for error bars in figures but standard deviations are given with means in the text. An ANOVA indicated statistically significant differences among groups in both height and diameter (Table 4.1 - all tables for this chapter are presented in Appendix 4). Means for the forest naturals were significantly lower than those for each of the other groups (Figure 4.1 - group numbers followed by different letters differed significantly with p <0.05). On the cutover, mean height did not differ significantly between the two groups of naturals or between the two groups of planted seedlings. In both cases, however, diameter was significantly greater in the open-grown seedlings than in the shaded ones. The ratio of seedling height to diameter also differed significantly among groups (Figure 4.2). Ratio means and standard error of the means ranged from 44.0 + 5.0 in the open planted seedlings to 66.2 + 12.4 in the forest naturals, and were lower for seedlings growing in locations with higher light availability, i.e. for the open naturals and the open planted seedlings. A comparison among groups of the trends in annual height increment between 1987 and 1990 suggested that the shaded planted seedlings were declining in performance in comparison to the other four groups (Figure 4.3). Although increments in the forest naturals were low, they were also very similar from year-to-year,with a mean annual average of 2.87 + 0.71 cm. Increments in the cutover naturals increased annually over the 1987-1990 period, although the increases were notably greater in 1987 and 1988 than in 1989. The planted seedlings had larger height increments relative to all three groups of Figure 4.1 Means and standard error of the means for height (upper figure) and diameter (lower figure) of Engelmann spruce seedlings from five different environments. «(number per treatment)=15. Group numbers followed by different letters were significantly different with p <0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted. 1 6 1 l.a 2.ab 3.be 4.ab 5.c Group Figure 4.2 Means and standard error of the means for ratio of height:diameter in Engelmann spruce seedlings from five different environments. n=15. Group numbers followed by different letters were significantly different with p <0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted. 1 6 2 15 6 o S i o 6* o M 0 bO ~ 5 rt rt 1986 1987 1988 1989 1990 1991 Year Figure 43 Means and standard error of the means for annual height increments between 1987 and 1990 in Engelmann spruce seedlings from five different environments. n=\5. Groups: ( Q )=forest naturals, ( / \ )=shaded naturals, ( <] ) =open naturals, (-^j-)=shaded planted, ( >j< )=open planted. 163 naturals in 1987, the year of planting (in Chapter 3 this was explained as a carryover effect from the nursery - Figure 3.6) but smaller increments in 1988 (a "planting shock" effect). In both 1989 and 1990, the leader lengths of open planted seedlings were much greater than in the previous year and were higher than any other seedling group. The 1989 height increment in the shaded planted seedlings was greater than the 1988 increment, but height growth decreased from 1989 to 1990 in these seedlings. This was the only group to exhibit a decrease in increment between 1989 and 1990. Relative production rates (RPR) were very low in all groups (Table 4.2a). An ANOVA indicated that among-group differences were statistically significant (Table 4.2b). Rates for the forest naturals, the open naturals and the shaded planted seedlings ranged from -0.144 + 0.335 to 0.011 + 0.177 loge cm/year (Table 4.2a). Growth rates for the shaded naturals and the open planted seedlings were greater than rates for the other three groups of seedlings at 0.108 and 0.243 loge cm/year, respectively. Lateral number and lateral length exhibited the same differences among groups as seedling diameter (results for lateral variables are not presented), but the pattern for needle variables was different. Needles ranged in length from 12.4 + 1.8 mm to 16.6 + 1.8 mm and were significantly shorter in both the forest naturals and the shaded planted seedlings than in the other three groups (in contrast, the diameter of the shaded planted seedlings did not differ from the diameter of the cutover naturals) (Figure 4.4 - upper figure). The number of needles per cm stem length (hereafter referred to as needle number) * exhibited a pattern reversed to that of the other variables, as it did among the treatments in the planted spruce experiment (Figure 4.4 - lower figure). The forest naturals had a significantly higher needle number (7.3 + 0.8) and the open planted seedlings had a 164 Group Figure 4.4 Means and standard error of the means for needle length (upper figure) and number of needles (lower figure) for Engelmann spruce seedlings from five different environments. n=15. Group numbers followed by different letters were significantly different with p <0.05 in a pairwise comparison of means using Tukey's H S D test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted. 165 significantly lower needle number (3.1 + 0.5) than each of the other groups. Unlike needle length, however, differences between the shaded naturals, the open naturals, and the shaded planted seedlings were nonsignificant. Principal components analysis was used to summarize and describe the differences in the size variables among the five groups of seedlings (Figure 4.5). With the exception of needle length, growth variables were strongly correlated with one another. Factor scores on the first two PCA axes separated out all the groups except the shaded and open naturals on the cutover. This is illustrated by the bivariate ellipses which have been plotted for each group in Figure 4.5. The ellipse is centered on the group means for factors 1 and 2, and it encompasses a 50% confidence region based on the standard deviations for factor scores. The first two components of the PCA accounted for 81% of the variance in the data set (Table 4.3). Seedlings with high scores on the first axis (the open planted seedlings) were taller, had larger diameters, more and longer laterals, more buds, and fewer needles (per cm leader) than seedlings with low scores (the forest naturals). The second axis was heavily loaded on needle length, and the shaded planted seedlings were separated out from the cutover naturals along a gradient of increasing needle length. The two groups of naturals on the cutover did not separate strongly along either axis despite present differences in their environments (see next section). Seedling biomass Differences among groups in total seedling biomass were similar to differences in height. The mean and standard deviation for total biomass in the open planted seedlings 166 u O •M o cd 2 -0 --1 --2 --3 -2 -1 0 Factor 2 Figure 4.5 Seedling factor scores on the first two axes of a principal components analysis of the correlation matrix of nine variables representing the size characteristics of Engelmann spruce seedlings from five different environments. The ellipses are centered on the group means for factors 1 and 2 and they encompass a 50% confidence region around the means based on the standard deviations for factor scores. Groups: 1 = forest naturals ( O ) , 2 = shaded naturals ( • ) , 3 = open naturals ( # ), 4 = shaded planted ( V ) , 5 = open planted ( + )• was 52.9 + 24.5 g, a value significantly greater than all other groups (Figure 4.6). The biomass of the shaded planted seedlings (20.4 + 10.4 g) was also significantly greater than that of the shaded and forest naturals. The forest naturals had the lowest biomass, and with a mean of 2.2 + 1.7 g, were significantly smaller than all other groups. Coefficients of variation (CV.) were high for total biomass and ranged from 0.46 for the open planted seedlings to 0.75 for the forest naturals. Coefficients were much higher for this variable than for any of the simple size variables (CV. were also very large for RPR - Table 4.2a). Coefficients for all variables tended to be largest in the forest naturals and lowest in the planted seedlings. Stem, needle, bud and root biomass each exhibited a pattern across groups that was similar to that for total biomass (Figure 4.6). Both shoot and root biomass differed significantly among groups (Table 4.4). In addition to differences in root biomass between the planted and naturally established seedlings, there were also striking differences in the morphology of their root systems. All naturals were shallowly rooted in comparison to the planted seedlings. Root systems on naturals were characterized by extensive horizontal extension with little vertical development of roots. This was most pronounced in the forest naturals, which were all rooted in fallen logs (at an average height above the ground of 33.9 + 13.9 cm) but it was also evident in the roots of cutover naturals growing in both mineral soil and rotten wood (Plate 4.1). The naturals in the open locations on the cutover were all rooted in mineral soil. Those in the shaded locations on the edge of the skid trail were mostly rooted in mineral soil. However, a few were growing in decayed wood on top of mineral soil and had roots in both substrates. In contrast to the naturals, most of the root growth on the planted seedlings (which were all growing in mineral soil) was at the base of the root plug, particularly on seedlings in poor condition (Plates 3.4 and 3.5). 168 G r o u p Figure 4.6 Means and standard error of the means for root, stem, needle, and bud biomass (bars indicate standard errors for total biomass) for Engelmann spruce seedlings from five different environments. n=l5 for Groups 1,2,3. n=6 for Groups 4,5. Group numbers followed by different letters were significantly different with p <0.05 in a pairwise comparison of means using Tukey's HSD test. Groups: l=forest naturals, 2=shaded naturals, 3=open naturals, 4=shaded planted, 5=open planted. 169 Plate 4.1 Naturally established six-year-old Engelmann spruce seedling from the edge of a skid trail on study site 2. The seedling was growing in the open and was rooted in rotten wood. The root system was sparse, shallow, and laterally-extended. Contrast this with the shape of the root systems on the planted seedlings in Plates 3.4 and 3.5. Shoot:root biomass ratios were very similar for the naturals on the cutover and the open planted seedlings (Figure 4.7 - upper figure). Means and standard deviations in these groups ranged from 5.5 + 1.3 to 5.6 + 1.7 and did not reflect the significant difference in light levels between the open and shaded locations from which naturals were collected on the cutover. (Average light levels at the height of the seedling were 26% of full sunlight in the shaded area and 74% of full sunlight in the open area.) However, differences among groups in shoot:root ratios were statistically significant (Table 4.5a) and ratios were lower for the seedlings from the two locations with the lowest light levels. The mean shoot:root ratio for the forest naturals was 3.7 +1.2 and for the shaded planted seedlings it was 4.2 + 0.7. (Average light levels were 15% of full sunlight in the forest and 12% in the control plots from which the planted seedlings were collected.) Four of the five groups of seedlings had very similar needlerstem biomass ratios, which ranged from 1.14 + 0.26 in the open planted seedlings to 1.29 + 0.26 in both the forest naturals and the shaded naturals on the cutover (Figure 4.7 - lower figure). In the fifth group, the shaded planted seedlings, the ratio was significantly lower (0.72 + 0.09) than in all other groups, indicating a different partitioning of resources between stemwood and foliage b iomass in this group of seedlings. Environmental variables The relationships between environmental and growth variables were explored using a Pearson product-moment correlation matrix and Bonferroni-adjusted probabilities to test for significant correlations between variables (correlation coefficients are presented in Table 4.6). Light levels (measured at the height of the seedling and expressed as % of full 171 8 [— r—-—~i T 1 r 7 -Group Figure 4.7 Means and standard error of the means for ratios of shoot:root (upper figure) and needle:stem (lower figure) biomass for Engelmann spruce seedlings from five environments. n=15 for Groups 1,2,3. n=6 for Groups 4,5. Group numbers followed by different letters were significantly different with p <0.05 in a pairwise comparison of means using Tukey's HSD test. sunlight) were significantly correlated with all size variables except needle length when all seedlings were treated as one group. Light levels were also significantly correlated with total seedling biomass, although the coefficient of determination was low (r2=0.228, p=0.021) as it generally was for all pairs of variables. Soil temperature and seedling biomass were not significantly correlated when seedlings from all groups were analysed together. However, when seedlings from the two groups of naturals on the cutover were analysed by themselves, there was a significant linear correlation between root biomass and soil temperature (r2=0.206, p=0.012), and between total biomass and soil temperature (r2=0.177, p=0.021). These relationships were nonsignificant for the planted seedlings and the forest naturals. Soil temperature and light level were not significantly correlated for the cutover naturals. Light level was correlated with root biomass (r2=0.192, p=0.015) but not with total biomass (r2=0.095, p=0.097) for these two groups of seedlings. There was a large amount of variation at each location in the environmental characteristics (Table 4.7). Although most variables differed significantly between at least two of the five locations, an exploratory analysis of the relationships between variables with a Pearson correlation matrix and Bonferroni-adjusted probabilities showed that most correlations were weak (correlation coefficients ranged from -0.461 to 0.388) and not statistically significant (results not presented). Principal components analysis was used to explore the differences between the environments from which the seedlings were collected. The first five components of the PCA had eigenvalues greater than 1 and accounted for 75% of the variation in the data set (Table 4.8). Only the first two components are discussed here. Despite large variation, there were distinct differences among the five seedling environments (Figure 4.8). The 17 3 o -t-» o c3 fe 0 --1 --2 --3 - 2 -1 0 Factor 2 Figure 4.8 Environmental factor scores for seedlings on the first two axes of a principal components analysis of the correlation matrix of twelve variables characterizing five different environments of Engelmann spruce seedlings. The ellipses are centered on the group means for factors 1 and 2 and they encompass a 50% confidence region around the means based on the standard deviations for factor scores. Groups: 1 = forest naturals ( O )> 2 = shaded naturals ( • \ 3 = open naturals ( # ), 4 = shaded planted ( v ), 5 = open planted (• + ). planted seedling locations were separated out from those of the naturals along the first PCA axis. Locations with high scores on the first axis, i.e. the control and clipped plots in which the spruce seedlings were planted, were characterized by low moss cover, deeper litter layers, wood and deciduous leaves in the litter, and a higher cover of herbs. Locations with high scores on the second PCA axis had relatively higher light levels and soil temperatures than other locations (Table 4.8). This axis separated out the open planted seedlings and the open naturals on the cutover from the other three groups. DISCUSSION There were large differences in the size and biomass of planted and naturally established seedlings from the mature forest and the cutover environments. The patterns among groups for different growth components varied more for the five seedling populations than they did for the seven treatments in the planted spruce experiment described in Chapter 3, but ranges in values for variables were similar in both studies. How do the characteristics of planted and naturally established seedlings differ? The six-year-old seedlings in the forest understorey were much smaller and had much less biomass than the naturally established and planted seedlings on the cutover. Low annual height increments and a RPR nonsignificantly different from 0 indicate how slowly these seedlings are growing. In contrast, the open planted seedlings were much larger and were growing much faster than all three groups of naturals. The shaded planted seedlings were somewhat larger than the naturals but were growing poorly, with a negative RPR in 1989-1990 and an annual height increment greater only than that of the forest naturals. The larger size of the shaded planted seedlings is a legacy from the size at the time of planting. Significantly lower needle:stem biomass ratios in the seedlings planted in the herb and shrub vegetation may be indicative of the severe resource limitations in these environments. Scarce resources may be preferentially allocated to stem growth in response to metabolic adjustments in nutrient-use efficiency, as suggested by Munson and Timmer (1990), or the lower needle:stem biomass ratios may be due to insufficient resources for needle production in the post-planting environment. If the relationship between the amount of needle and stem biomass is significant to survival, then the differences observed among the five groups of seedlings suggest that creating nursery environments that will produce seedlings with a good chance for surival after outplanting requires more than setting standards for height or diameter. The distinct differences in the morphology of the root systems between naturally established and planted seedlings could not be related to differences in growth in this study. The range in shoot:root biomass ratios were the same for both types of seedlings and seemed to be more strongly related to total resource availability than seedling type. Nevertheless, it is difficult to believe that in cold subalpine soils root characteristics such as depth and distribution are unrelated to seedling performance. Slow seedling root growth is a characteristic of Engelmann spruce (Alexander, 1987), which may be why the root growth on the planted seedlings was so poor (rather than because of nursery or planting effects). However, there may still be important consequences for the growth of planted seedlings due to the location within the soil profile where early root growth is occurring, as well as from mutual root competition within the dense root plug. How did growth vary among locations in the naturally established seedlings? The relationships among size variables within naturally established seedlings were similar to those described in Chapter 3 for the planted seedlings. Similar to the results for the planted spruce experiment, the number of needles exhibited a pattern among groups reversed to that for other growth variables. The number of needles per unit stem was much higher in the forest naturals than in any of the other groups, and was lowest in the open planted seedlings. If this variable is an expression of the seedling's growth potential on a site (with lower needle numbers indicating greater fulfillment of this potential), then one would predict lower needle numbers on the cutover naturals than on the planted seedlings, assuming that the former are better adapted to this environment However, this was not observed and differences were more closely related to height increments, suggesting an inverse relationship controlled by the factors influencing height growth. Differences between the groups of naturals and the planted seedlings in needle variables versus stem variables does imply differences in the factors influencing the development of the two seedling components, as suggested by Funck et al. (1990). The significantly lower shoot:root (S:R) biomass ratios in the forest naturals than the naturals on the cutovers is consistent with observations on the planted spruce seedlings. Increases in S:R ratios have been observed with increasing age and size in other conifer seedlings (Cole and Newton, 1987), despite some general references to opposite trends in woody plants (Wilson, 1988b). What was the relationship between environmental factors and the characteristics of seedling growth? The PCA of environmental variables showed differences in the environments of the five groups of seedlings that were related to differences in seedling growth. As was the case for the spruce seedlings planted into the herb, shrub, and skid trail patches, light was the factor most strongly associated with seedling growth. Nevertheless, the amount of variation in seedlings accounted for by any environmental variable was low, as Scagel (1984) found in his study of Engelmann spruce. The significant correlation between growth variables and soil temperature only among the naturals on the skid trails (and the weaker relationship between light and growth in this area) suggests that there is variation among locations in the relative importance of factors controlling seedling growth. As suggested by DeLucia and Smith (1987), the relative importance of the impacts of factors such as soil temperature may also vary seasonally. Most of the variance in the PCA of growth variables was accounted for by the first PCA axis. All growth variables but needle length loaded strongly on the first axis which separated the groups along a gradient of increasing size. This pattern provides little evidence of changes in allocation among groups and is consistent with the results from the planted spruce experiment where greater growth involved increases in all seedling components, except needle number per unit stem. There was some evidence that needle variables behaved differently from other seedling components. Needle length was significantly correlated only with needle number (negatively) and current height increment (positively), and differences among groups accounted for the second major axis of variation in the PCA of the growth data. It is not clear what significance this may have to seedling survival and growth. Perhaps allocation to needles is more plastic that allocation to other components in Engelmann spruce seedlings. The lower needle:stem ratios in the shaded planted seedlings suggests that proportional allocation to needles does reflect the status of the seedlings. Under natural conditions, large expenditures of photosynthate on woody tissue can result in a slow rate of expansion of leaf area in tree seedlings, reducing their ability to compete with non-coniferous species (Grime, 1979). However, the actual amount of biomass allocated to stems over the study period was not measured, and neither was tje needle:stem biomass ratios in seedlings at the time of planting. Therefore, it is not possible to determine whether the lower ratios in the shaded planted seedlings were the result of a shift in allocation to woody tissue in these seedlings, or a lack of resources for needle production. What are the silvicultural implications of these results with respect to regenerating Engelmann spruce? In the forest environment, the naturals were largely restricted to microsites in which growth and rates of biomass accumulation were very slow. On the cutover, naturals were most abundant in environments that do not occur in a forest, i.e. those on or adjacent to skid trails. Rates of biomass accumulation were greater in these areas, but growth was still very slow. Planted seedlings were able to survive and grow in locations in which spruce rarely established naturally, i.e. in areas with deep litter or dense vegetation (see next chapter also). When planted seedlings were growing without overtopping vegetation rates of growth and biomass accumulation were much higher than, but patterns of allocation within seedlings were similar to, those in naturally established seedlings. However, rates of growth in seedlings planted beneath canopies of non-coniferous vegetation were very low, and despite a larger size (due to size at planting) their performance was declining relative to naturals, particularly in height growth and needle production. These results suggest that harvesting systems in the ESSF that preserve preharvest forest conditions (i.e. selection cutting) will result in extremely slow natural regeneration of Engelmann spruce due to slow rates of seedling growth (rates of invasion may also be 180 slow but these were not measured in this study). Systems that create small disturbances analagous to those on the edges of skid trails (i.e. patch clearcutting with understorey disturbance) will improve performance, but rates of growth will still be slow. Planting seedlings to increase early rates of regeneration is an option that will be effective only where competition from non-crop species does not severely limit the availability of resources to seedlings. C O N C L U S I O N S A comparison of the characteristics of planted and naturally established Engelmann spruce seedlings revealed significant differences between the growth and the environments of the two groups of seedlings. The growth of seedlings planted in the open was much better than that of the naturals, which were growing very slowly, particularly in the understorey of the old-growth stand. In contrast, shaded planted seedlings were growing very poorly and there was some evidence that needle production was particularly limited in these seedlings. A large stem biomass relative to needle biomass in seedlings at the time of outplanting might be especially disadvantageous for seedlings planted into competitive environments (assuming the limiting resource was light). This is speculation however, as the changes in the ratio of stem biomass:needle biomass during the first three growing seasons were not measured in this study. There were also distinct differences in the morphology of the root systems between the planted and naturally established seedlings. There was no evidence from this study that this reduced the performance of planted seedlings but further study is warranted, particularly for seedlings growing under conditions of interference from non-crop species. The effectiveness of planting as a management strategy for regenerating spruce seems to depend on delivering a high level of resources to seedlings. Past harvesting in the ESSF has left a legacy of sites with well-developed non-coniferous vegetation within which both natural and planted spruce perform poorly. The results from this study suggest that to improve early growth and survival at least partial removal of this vegetation must be undertaken to increase the availability of resources for which spruce cannot compete. Prescribed burning is one silvicultural treatment which may do this but for both sociological and ecological reasons it is not always an option for site preparation. Mechanical scarification can also be used to create disturbances of different scales and intensities. The efficacy of these treatments will depend on the ability of the species that commonly occur in the ESSF to re-establish cover after the removal of vegetation. An experiment was carried out on the Vavenby sites to determine the response of the major species on these cutovers to the removal of vegetation and to determine how effective small scale disturbances could be in providing a high resource environment for planted spruce trees. The results of this experiment are presented in the following chapter. CHAPTER 5 183 COMMUNITY AND SPECIES RESPONSES TO DISTURBANCE IN HERB AND SHRUB PATCHES ON CUTOVERS IN T H E ESSFwc2 INTRODUCTION Knowledge of the patterns and processes of early secondary succession is now recognized as essential for effectively managing sites for conifer regeneration after logging (Conard, 1984; Hamilton and Watts, 1988; Walstad and Kuch, 1987). In addition to assessing the impacts of co-occurring vegetation on early seedling performance, much effort has been spent in recent years in characterizing the temporal patterns in community development after logging and related disturbances (Bormann and Likens, 1979; Day, 1972; Hamilton and Yearsley, 1988 - see references therein). From a silvicultural viewpoint, the major objective of these efforts has been to develop the ability to: 1) predict which successional pathways will retard conifer regeneration; 2) recognize the conditions under which these pathways will develop, and; 3) prescribe effective vegetation management treatments to achieve specified silvicultural objectives (Lavender et al, 1990; Walstad and Kuch, 1987). Post-logging plant communities have a major impact on spruce regeneration in the v ESSFwc2. The results presented in Chapters 3 and 4 indicated that planted tree seedlings can outperform naturals in early growth when above-ground interference from herb and shrub species is low. However, most cutovers in the ESSFwc2 have well-developed vegetation and levels of interference are typically high. The extent to which these levels can be reduced depends on the species responses to disturbance. Two components of these responses of interest for management are the speed of re-establishment of plant cover after disturbance, and the pattern of shifts in dominance to new species with different impacts on seedlings. An experiment was carried out in the herb and shrub patches on the Vavenby sites to study these two characteristics of species responses to disturbance. Literature Review On the Vavenby sites, most species occurred in both the herb and the shrub patches but with significant differences in abundance between the two patch types (Table 2.1). Apparent differences in competitive dominance between the patches leads to the prediction of different changes in composition or structure in patch vegetation in response to disturbance. Keddy (1989a) studied the effects of competitive release from shrubs on populations of herbs along a fertility/disturbance gradient on a lakeshore. He found that herb release from shrubs was greater at the high fertility, low disturbance end of the gradient where competition was most intense. In contrast, there was no evidence for release under low fertility, high disturbance conditions. Where release did occur, however, relatively few species were involved (in any year, less than one-fourth of the species showed evidence of release) and the rate of colonization of clearings was slow after the removal of vegetation. Keddy hypothesized that the limited response to shrub removal was due to insufficient propagules for colonization and soils in the shrub zone that were too wet for herbs. He suggested that competitive interactions controlled the distributional limit of 185 herbs at the high fertility, low disturbance end of the gradient and that abiotic constraints (i.e. tolerance to high water levels) controlled their distribution at the other end. Keddy's results demonstrate how the control of relative competitive abilities by environmental factors can influence community structure in a lakeshore community. The effects of environmental factors on competitive interactions can also structure subalpine herb and shrub communities. Del Moral (1983,1985) studied the relative importance of competitive interactions on the structure of several subalpine herbaceous communities distributed along a soil moisture/disturbance gradient, using an approach similar to that of Keddy. He found that competitive intensity was well correlated with productivity and, like Keddy, concluded that community structure was controlled by interactions among competitive relationships, microscale heterogeneity, and levels of disturbance. However, he observed competitive effects over the entire environmental/productivity gradient, although the rate and intensity of these effects were greatly diminished when abiotic factors severely limited productivity. He suggested that both intense competition and intense stress (defined as the sum of external constraints upon plant production, sensu Grime (i979)) resulted in strong dominance within communities at both ends of the environmental/productivity gradient. Evans and Fonda (1990) also studied factors controlling the distribution of plant communities in subalpine meadows. They found that the distribution of six herb and shrub communities (one of which was dominated by Valeriana sitchensis) in the North Cascades was explained by differences among topographical locations in the time of snowmelt, which in turn significantly influenced soil temperatures and the length of the growing season. There is no evidence from the data presented in Chapter 2 to suggest that the distribution of patch types on the Vavenby sites was related to an environmental gradient across these areas but the differences in edaphic properties between the patches suggest that their distribution is influenced by environmental heterogeneity. The presence of both herb and shrub patches in the understories of old-growth spruce-fir stands in the study area also suggests that these patches are not transient features of the post-logging environment. The distribution of the two patch types after logging likely results from a combination of the persistence of some patches through disturbance and the post-logging development of new patches. The importance of disturbance in patch dynamics can be assessed to some extent by the stability (in the sense of unchanging composition and location) of patches, and by determining whether the factors controlling within-patch structure are the same as those controlling between-patch structure. For example, differences in dominants between the two patch types, in combination with the differences in environmental factors, suggests that abiotic factors (e.g. resource heterogeneity) may be more important determinants of patch size and location, and biotic processes more important in structuring the vegetation within patches. Del Moral (1985) concluded from his study of competitive effects in subalpine meadow communities that the distribution of competitive dominants was more directly influenced by environmental factors whereas that of weak competitors was more influenced by the degree of competitive intensity within a community. The importance of distinguishing within- from between-patch processes in community structuring was emphasized by Bowers and Dooley (1991), who used a simulation model to explore the factors affecting patch dynamics at a landscape level. They suggested that predictive theories of community structure in patchy environments should be hierarchically based with local, within-patch processes nested within those occurring over patch networks. At a landscape level they found that the dispersal distance of species, relative to patch size and separation, determined the extent to which competition occurred locally or regionally. When dispersal distance was short, population dynamics were dominated by processes operating locally. In their simulation model the competitive 187 ability of a species varied in proportion to the abundance of optimal habitat for that species, as well as its dispersal ability. Viewing patch dynamics on the Vavenby sites from this perspective, the relative ability of a species to invade an area (an unoccupied patch) after disturbance, or its relative ability to occupy that area (e.g. through rapid growth or clonal expansion) to the exclusion of other species, may be more important in determining dominance than within-site variation in resources or environmental factors. Colonizing ability and the ability to hold on to space are well-recognized as major factors influencing patterns of community development in early secondary succession (Keddy, 1989b; Yodzis, 1986). Some of the empirical knowledge underlying this recognition has come from studies of western coniferous forests (Halpern, 1988,1989; MacMahon, 1980). Halpern, for example, did a retrospective study of 21 years of succession following logging and burning in Douglas-fir forests in the western Cascades. He found that species that dominated early in succession had either survived logging or colonized sites within a year or two after the disturbance. After a temporary convergence in the composition of several post-logging communites, which was due to rapid colonization and dominance by invading species, further community development was characterized by gradual and uni-directional shifts in the abundance of generally persistent species towards their predisturbance compositions. This pattern was influenced by the intensity of soil disturbance associated with logging and burning such that there were increasing compositional changes in communities with increasing soil disturbance. Results from studies such as Halpern's suggest that there is both a deterministic component and a stochastic component to early community development after disturbance. The deterministic component is based on the life history characteristics of local species and their relative competitive abilities under different environmental conditions. Variations in these conditions, in species availability, and in the 188 characteristics of disturbance (e.g. timing, intensity) introduce stochastic elements that limit predictive capabilities. It has been argued that knowledge of competitive abilities among species is inadequate to understand community organization because of the obfuscating effects of other factors such as environmental heterogeneity and noncompetitive biotic relationships (Hoekstra et al, 1991; Naeem, 1990). However, much of this problem may lie in traditional approaches to studying competition in communities, and the potential for alternative approaches for detecting competitive effects at the community level has been demonstrated by Keddy (1989). Knowledge of species-specific responses to disturbances are necessary, if not sufficient, for predicting early community development Halpern (1989) attributed the community stability he observed in Douglas-fir forests to the properties of the component species rather than to emergent properties of the system itself. Knowledge of species-specific responses, however, would appear to be more useful for predicting early successional patterns in communities of persistent species rather than in communities dominated by species which invade after disturbance (Halpern, 1989; Hamilton and Yearsley, 1988; Stickney, 1986). For example, during early succession following logging and site preparation in the Sub-boreal Spruce zone in the northern interior of British Columbia, invading species exhibited the largest variation in abundance among communities (Hamilton and Yearsley, 1988). Probabilistic factors likely play a greater role in determining patterns of distribution in invading species than they do in determining the patterns of persistent species. Nevertheless, it is useful to know to what extent, and if possible, under what conditions, invading species may dominate early community development. Some of these species can have a severe impact on conifer regeneration and their predominance in post-logging communities may present a significant vegetation management problem. For example, 189 both Halpern (1988) and Hamilton and Yearsley (1988) identified Epilobium angustifolium (fireweed) and Rubus parviflorus (thimbleberry) as invading species with early dominance on some sites after logging. Both of these species can reduce the growth and survival of planted conifer seedlings (Haeussler and Coates, 1986), and both were present on the Vavenby sites (Table 2.1). An experiment was set up on the Vavenby sites to study the response of herbs and shrubs to two levels of disturbance. The major objectives of the experiment were to determine the rate of re-establishment of plant cover and to determine whether the composition of patches changed in response to disturbance. These objectives are both of practical significance to improving conifer regeneration, and will enable inferences to be made about the predictability of early community development after disturbance in the ESSFwc2. The complexity of the post-logging communities on the Vavenby sites would make the determination of the relative importance of competitive relationships, invasion capabilities, and habitat heterogeneity to patterns of early community development a study unto itself, and this was not an explicit objective of the research. Nevertheless, some general hypotheses about patch stability and the factors controlling patch dynamics on these sites can be derived from a comparison of spatial and temporal variation in species abundances within patches (competitive relationships), between herb and shrub patches (habitat heterogeneity), and among the control, clipped, and screefed treatments (invasion potential). 190 METHODS Experimental design To investigate plant species responses to disturbance within the herb and shrub patches a randomized block experiment was set up on sites 1 and 2 in June 1987. Nine 4 m x 4 m plots were staked out in herb patches and nine in shrub patches on each site. The composition of the patches was described in the methods section of Chapter 3. Each of three treatments was applied to three plots randomly selected from the nine plots in each patch type. The treatments were: 1. No disturbance (herb control and shrub control); 2. Removal of above-ground vegetation by clipping (manually), with minimal disturbance of the forest floor (clipped herb and clipped shrub); 3. Removal of both above- and below-ground vegetation by screefing to a depth of approximately 20 cm (which included humus and mineral soil) to remove the forest floor and expose mineral soil or buried wood (some mixing of surface organic material and mineral soil occurred during screefing) and removal of whole plants by ripping out the root systems. In June 1988 four permanent transects were established within each plot by placing nails at 0.25 m intervals along lines from the 1 m to 3 m points on opposite sites of the X plot. This resulted in a total of 52 fixed sampling points (nails) per plot. 191 Sampling methods and measurements Prior to the application of the treatments in 1987, the composition and percent cover of all species within each plot in the clipped and screefed treatments were recorded. A list of the species present in the control plots was made in 1987 but cover estimates were not taken until the spring of 1988. The percent cover and average maximum height of each species were subsequently remeasured in August 1987, June 1988, and July 1989. General phenological observations were made on all species in the control plots at monthly intervals from June to August in 1988 and 1989. In August 1988 and 1989 the following data were collected from all fixed sampling points within each plot: - distance from the point to its nearest neighbour; - species of nearest neighbour; - distance from this individual to its nearest neighbour; - species of this neighbour. To assess the regrowth capacity of the herb dominant, Valeriana sitchensis, the height and width of twenty plants in herb patches and twenty plants in shrub patches were measured on June 6th 1987. Ten randomly selected plants at each location were then clipped at ground level, and the height and width of all plants were remeasured on August 6th 1987. Similarly, ten mature plants and twenty juvenile (defined as stems which had originated since logging) of the shrub dominant, Menziesia ferruginea, were also selected for measurement on June 6th. Mature plants were in shrub patches (by v definition), and juvenile plants were generally in disturbed areas, frequently adjacent to skid trails. The lengths of five randomly selected stems were measured on each mature plant and the length of each juvenile stem was also recorded. All the stems of five 192 randomly selected mature plants were then clipped off at ground level. On August 6th the lengths of five new shoots per plant were measured on each clipped plant, and the total number of new shoots per plant was recorded. The lengths of the stems on the undipped plants and the juvenile plants (none of which were clipped) were remeasured at this time. To assess the natural regeneration of Engelmann spruce within the herb and shrub patches, the number, height, condition (see Chapter 3 for criteria used in the evaluation of condition), and rooting substrate of trees in the control plots which had established prior to treatment were recorded in August 1989. To assess post-treatment invasion by spruce the number, year of germination, and rooting substrate of germinants were recorded in two randomly placed 1 m2 quadrats in each plot in all treatments. The number of seed-bearing Engelmann spruce trees on the study sites was less than one per hectare. Data analysis Two-way analysis of covariance (ANCOVA) was used to test for significant vegetation type (herb, shrub) and treatment (control, clipping, screefing) effects on the number of species in plots three growing seasons after treatment. Initial species number was used as the covariate. Total plant cover, and changes in the cover of herb and shrub species three growing seasons after treatment were tested for significant differences in a two-way analysis of variance (ANOVA). Tukey's HSD test was used for pairwise comparisons of means. SYSTAT Version 5.0 (Wilkinson, 1990) was used for all statistical tests. Total plant cover was derived by summing the percent cover of each v species in the plot. Changes in the cover of herb species were derived as follows: ([1989 total percent cover of herb species/1987 total percent cover of herb species] x 100). The change in shrub cover was derived in a similar manner. Cover values were loge-transformed to satisfy the A N O V A assumption of homogeneity of variances. Raw values for species richness and total plant cover yielded non-significant Bartlett's statistics and consequently were not transformed. Random distributions of residuals from the A N O V A s suggested no significant deviations from test assumptions. Two indices were derived from the permanent transect data. A modified Berger-Parker dominance index (Magurran, 1988) was calculated for each plot as follows: D = Pmax/P where P m a x = the number of fixed sampling points with individuals of the most abundant species as nearest neighbour, and P = the total number of sampling points (i.e. 52). A constancy index, based on changes in species adjacent to fixed sampling points was calculated for each plot as follows: C = PCon/P where P C O n = the number of sampling points with a nearest neighbour of the same species in 1988 and 1989. The nails were not placed into the plots prior to clipping and screefing as the treatments would have removed them, and consequently predisturbance data were not available to use for the index. The nails were placed into the plots in the spring of 1988 when the first measurements of nearest neighbours were taken. Bartlett's statistic was non-significant for both indices when untransformed data were tested for homogeneity of variances. Treatment means were subsequently tested for significant differences in a two-way A N O V A and Tukey's HSD test was used for pairwise comparisons of means. Average distances between fixed sampling points and their nearest neighbours were analysed for significant vegetation and treatment effects in a two-way nested A N O V A . Analysis was based on thirty randomly selected points per plot. Loge-transformed values yielded a non-significant Bartlett's statistic and residuals were randomly distributed for the transformed data. Species shifts were assessed on the basis of changes in percent cover of individual species before and after treatment. Data on species-specific responses are presented in tabular form only. 195 RESULTS Community responses Species richness The vascular plants present in the invasion plots three growing seasons after treatment are given in Appendix 1. There was a total of 48 species in the plots on sites 1 and 2 in July, 1989. Most of the Engelmann spruce and subalpine fir trees present in the clipped and screefed plots had established prior to treatment. These individuals were not removed during clipping and screefing with the original intention of measuring their response to release. All individuals that established following treatment were recorded separately and this data is presented in Table 5.15 (all of the tables for this chapter are presented in Appendix 5). There were nine species on site 1 that were not found in the plots on site 2. The most abundant of these were Cornus canadensis (bunchberry) and Athyrium filix-femina (lady-fern). Rubusparviflorus (thimbleberry) and Galium triflorum (bedstraw) were also common on site 1 and rare on site 2. Two species, Senecio triangularis (ragwort) and Ranunculus uncinatus (little-flowered buttercup), were found only in the plots on site 2 and occurred occasionally in wetter areas on that site. The only species which occurred in all 36 plots were the herbs Valeriana sitchensis (Sitka valerian), Streptopus roseus (twisted-stalk), and Tiarella unifoliata (foamflower) and the fem, Gymnocarpium dryopteris (oak fern). Ribes lacustre (gooseberry) occurred in all herb plots. Species restricted mostly to herb plots included Actea rubra (baneberry), Thalictrum occidentale (meadow-rue), Viola glabella (violet), Galium triflorum, and Athyrium filix-femina. Menziesia ferruginea (false azalea), Vaccinium membranaceum (huckleberry), and Rubuspedatus (creeping raspberry) were found in all shrub plots. Rhododendron albiflorum (rhododendron) was the only species restricted mostly to shrub plots. The two Epilobium species, E. angustifolium (fireweed) and£. glandulosum (willow-herb), were more frequent in screefed plots than in the other treatments. Several species highly characteristic of disturbed environments appeared with low frequency and cover in some plots after the removal of vegetation. Among these were Taraxacum officinale (dandelion), Cirsium vulgare (bull thistle), and Anaphalis margaritacea (pearly everlasting). Total number of species (species richness) per plot three growing seasons after treatment is presented by treatment and vegetation type for sites 1 and 2 in Figure 5.1. Standard error of the means are used for error bars in figures but standard deviations are given with means in the text. The maximum average number of species was 23 in the clipped herb treatment on site 1 and the minimum number of species was 13 in the clipped shrub treatment on site 2. There were no consistent differences between sites in treatment means. On average, there were fewer species in the shrub treatments than in the herb treatments, but a two-way analysis of covariance (using the initial number of species as the covariate) indicated that differences between the two vegetation types were nonsignificant on both sites (Table 5.1 - Appendix 5). The covariate was a significant source of variation - the number of species in the herb plots prior to treatment ranged from 14 to 22, and in the shrub plots from 7 to 18. The ANCOVA indicated no significant differences among treatment means for species richness on either site (Table 5.1). Differences between the number of species in plots prior to treatment in 1987 and the number of species three growing seasons after treatment are presented for sites 1 and 2 in Figure 5.2. There was an increase in species richness in all treatments between 1987 and 28 197 21 -S Screef • Clip M Control Herb Shrub Patch type Figure 5.1 Means and standard error of the means for species richness in clipped, screefed, and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. /i(number of plots per treatment)=3. 198 te U •FH u o o. O u X I 0 .a 3 u a rt u u • rH u w p. Wi V X I 0 a .3 3 u BO a td XI U 4 -2 -1 -Herb Shrub S Screef • Clip M Control Herb Shrub Patch type Figure 5.2 Means and standard error of the means for the difference between pretreatment species richness and species richness three growing seasons after clipping and screefing in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). n=3. 1989. Increases ranged from a minimum (treatment mean) of 0.33 in the shrub control on site 1 to a maximum of 6.67 in the screefed shrub treatment on site 2. In all treatments but the clipped shrub, increases were greater on site 2 than site 1. Within each vegetation type increases were also greater in the clipped and screefed treatments than in the controls. However, differences among treatments and between vegetation types were not statistically significant when analysed in a two-way ANOVA (results not presented). Species cover Total percent plant cover (summed by species) three seasons after treatment is presented for sites 1 and 2 in Figure 5.3. Patterns among treatments and between vegetation types were similar on the two sites. Total cover was lowest (28%) in the screefed shrub treatments on both sites, and highest (162%) in the shrub control on site 2 (Table 5.2). A two-way ANOVA indicated significant differences between vegetation types and among treatments, and a significant interaction between the two main effects on both sites (Table 5.3). Within the herb patches, screefing reduced cover by approximately 50% in comparison to the controls, but total cover in the clipped treatment was as high or higher than that in undisturbed plots. On both sites covers were significantly higher in the clipped herb treatment than in the screefed herb treatment (Table 5.2). In contrast, within the shrub patches, cover in the clipped treatment was significantly less than in the controls and nonsignificantly different from cover in the sceefed treatment (total cover in the disturbance treatments was less than one-third that in the controls). The results indicated that the removal of above-ground vegetation by clipping had no effect on plant cover in the herb patches (Plate 5.1), but significantly reduced cover in the shrub patches (Plate 5.2). 200 0 Screef • Clip M Control Herb Shrub Patch type Figure 53 Means and standard error of the means for total percent cover (summed by species) of vegetation in clipped, screefed, and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. n=3. 201 Plate 5 .1 Clipped herb plot three growing seasons after treatment on study site 1. The plot is dominated by Valeriana sitchensis (Sitka valerian), Epilobium angustifolium (fireweed) and Arnica latifolia (mountain arnica). These species dominated prior to treatment and they re-established cover rapidly after clipping. The woman in the photograph is kneeling down and is holding a i m stick which indicates the average height of the vegetation. Plate 5.2 Clipped shrub plot three growing seasons after treatment on study site 2. The most abundant species is Menziesia ferruginea (false azalea) which dominated the plot before treatment but which has not re-established pretreatment levels of cover. The woman in the photograph is kneeling down and is holding a i m stick. The average height of the vegetation was less than 1 m. The change in the cover of herb species three growing seasons after treatment is presented for sites 1 and 2 in Figure 5.4. Changes were calculated by summing the percent cover of all herbs in 1989 and expressing the total as a percentage of the total cover in 1987 before treatment. Vegetation was not removed in the controls, so increases in cover since 1987 are indicated by values over 100% (e.g. a value of 100% in a control treatment indicates no change in cover since 1987). Patterns among treatments and between vegetation types were similar on the two sites. Changes in cover were lowest (18%) in the screefed shrub treatment on site 2, and highest (281%) in the shrub control on this site (Table 5.4). The very high values for the controls were a result of increases over the study period in species that had very low covers (generally < 5%) in 1987. A two-way ANOVA indicated significant treatment effects on site 1 and significant vegetation and treatment effects on site 2 (Table 5.5). Means over 100% in the controls on both sites (as well as increases in species numbers in these areas) suggest that the vegetation on the Vavenby sites was not at equilibrium. However, the relatively large increases in herb cover within the shrub controls were to some extent a consequence of low initial covers for these species in the shrub patches. Herbs re-established approximately 100% of their original cover after clipping in herb patches, but recovery of herbs was less than 100% in the shrub patches. During this time, herb cover also increased in the controls, although differences between the clipped and control treatments were statistically significant only in the shrub patches on site 2 (Table 5.4 - note that differences in the shrub treatments on site 2 would still have been significant if only increases above initial cover values were considered). Three growing seasons after screefing, herbs had not recovered to predisturbance levels within either the herb or shrub patches, but recovery was greater in the herb patches than in the shrub patches (Figure 5.4). The significant interaction between treatment and vegetation type 204 300 00 Xl S Screef • Clip W Control Herb Shrub Patch type Figure 5.4 Means and standard error of the means for differences between pretreatment percent cover of herbs (summed by species) and cover three growing seasons after clipping and screefing in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). Values are expressed as percentages of initial pretreatment cover. n=3. 205 on site 2 arose from the higher cover values in the shrub control and lower cover values in the shrub screefed and clipped treatments, relative to the herb treatments (Table 5.4). Changes in the cover of shrub species (calculated in the same manner as herb recovery values) are presented for sites 1 and 2 in Figure 5.5. Patterns among treatments and between vegetation types were similar on the two sites, and were similar to those for herbs. Changes in cover were lowest (9%) in the screefed shrub treatment on site 2, and highest (493%) in the herb control on site 1 (Table 5.6). The very high value in the herb control on site 1 was due to a large increase in the cover of Rubus parviflorus in the herb control plots: the average cover per plot of this shrub increased from a mean and standard deviation of 1.7 + 0.6 % in 1987 to 23.3 + 7.6 % in 1989. A two-way ANOVA indicated significant treatment differences on site 2 and significant vegetation type and treatment differences on site 1 (Table 5.7). In contrast to the herbs, shrub species did not recover to predisturbance levels after clipping, and means at most locations were significantly lower in the clipped treatment than in the controls (Table 5.6). Changes in shrub cover were lowest in the screefed treatments, and values were only one-half those in the clipped treatments although differences between the two treatments were not statistically significant (Table 5.6). Invasion and dominance indices The Berger-Parker dominance index is presented by treatment and vegetation type for v sites 1 and 2 in Figure 5.6. In this index, dominance (D) is defined by the number of individuals of the most abundant species rather than by size or biomass. Values of D 206 500 oo 3 400 a u o Wi U 300 200 S> 100 a cd X l u 500 Herb Shrub 00 X) 3 Wl X l 00 400 300 S 200 u CV s> ioo a cd XI U 0 Screef • Clip M Control Herb Shrub Patch type Figure 5.5 Means and standard error of the means for differences between pretreatment percent cover of shrubs (summed by species) and cover three growing seasons after clipping and screefing in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure). Values are expressed as percentages of initial pretreatment cover. n=3. 207 m Scrcef • Clip M Control Herb Shrub Patch type Figure 5.6 Means and standard error of the means for Berger-Parker dominance (D) indices for clipped, screefed, and control plots in herb and shrub patches three growing seasons after treatment on site 1 (upper figure) and site 2 (lower figure). D = P m a x / P , where P m a x = number of sampling points with individuals of the most abundant species as nearest neighbour and P = total number of sampling points, "^indicates treatment differing significantly from all other treatments when tested with Tukey's HSD test at a = 0.05. n=3. 208 ranged from a mean of 0.20 in the herb control on site 2 to 0.63 in the screefed herb treatment on site 2, with a higher value of D indicating greater proportional abundance of a single species. Patterns were similar on both sites, but treatment means were significantly different only on site 2 where D was significantly higher in the screefed herb treatment than in all other treatments (Figure 5.6 and Table 5.8). Within herb patches, clipping did not affect D because vegetation re-established quickly after cutting. In both the clipped and control treatments the most abundant individual in herb patches composed approximately 20% of the sampled population. Screefing increased values of D in herb patches on both sites, although the increase was only significant on site 2 (Figure 5.6). On site 2, one species (Epilobium angustifolium) composed on average over 60% of the sampled population. In contrast, in the shrub patches there were no significant changes in D in response to clipping or screefing on either site, although values were slightly higher in the controls and slightly lower in the clipped treatments on both sites. Values ranged from 20% to 40% for the three treatments in the shrub patches. The constancy index (C), which was based on the proportion of fixed sampling points with a nearest neighbour of the same species in 1988 and 1989, is presented for sites 1 and 2 in Figure 5.7. Patterns among treatments and between vegetation types were similar on the two sites. Means were relatively high in all treatments with values ranging from 0.40 in the herb control on site 2 to 0.69 in the clipped shrub treatment on site 2. Within the herb patches patterns were similar to those for D. Values of C were lower in the control and clipped treatments than in the screefed treatments, although differences were not statistically significant (Table 5.9). In patches but differences between the vegetation types were significant only on site 2 v (Table 5.9). Similarly, although the trend on both sites was for higher values in the clipped shrub treatment than in the screefed treatment or the control, differences between the three treatments were not statistically significant. 209 @ Screef • Clip m Control Herb Shrub Patch type Figure 5.7 Means and standard error of the means for a constancy index ( Q for clipped, screefed, and control plots in herb and shrub patches three growing seasons after treatment on site 1 (upper figure) and site 2 (lower figure). C = P c o n / P , where P c o n = number of sampling points with a nearest neighbour of the same species in 1988 and 1989 and P = total number of sampling points, indicates herb treatments differing significantly from all shrub treatments when tested with Tukey's HSD test at a = 0.05. n=3. 210 The average distance between a fixed sampling point and its nearest neighbour is presented by treatment and vegetation type for site 1 in Figure 5.8. Average distances ranged from a mean of 1.59 cm in the clipped herb treatment to 7.20 cm in the screefed shrub treatment. There were significant differences among treatments and between vegetation types, as well as significant plot-to-plot variation in means (Table 5.10 -ANOVA). Within each of the three treatments distances were greater in the shrub patches than in the herb patches. Among treatments, distances were greater in the screefed plots that in either the clipped or control plots, but these differences were statistically significant only between the screefed and control treatments in shrub patches (Table 5.10). Distances between a fixed point and its nearest neighbour were significantly greater in the screefed shrub treatment than in most other treatments, reflecting the slow rate of invasion and expansion of populations after severe disturbance in this vegetation type. After three growing seasons, the average maximum height of the non-crop vegetation exceeded that of the planted spruce seedlings (Chapter 3) in all three treatments in both patch types. The height of the hot and cold stock (averaged over all treatments on each site) ranged from 27.2 cm for the hot stock in shrub patches on site 2 to 36.2 cm for the cold stock in shrub patches on site 1. In comparison, the average maximum height of the vegetation in the clipped and screefed treatments (values were similar in magnitude for the two treatments) ranged from 48.7 cm in the clipped shrub treatment on site 2 to 87.0 cm in the screefed shrub treatment on site 1. The height of vegetation in the controls was four to five times greater than that of the planted seedlings: average maximum values v ranged from 108.7 cm to 133.0 cm in the herbs and 181.7 cm to 187.3 cm in the shrubs. 211 Figure 5.8 Means and standard error of the means for distances from sampling points to nearest neighbours in clipped, screefed and control plots in herb and shrub patches on site 1 (upper figure) and site 2 (lower figure) three growing seasons after disturbance. n=90. 212 Species-specific responses Phenology Observations on the phenology of the major species on the Vavenby sites indicated that growth began earlier in the season on site 1 than on site 2. By early June, the leaves of most species had unfolded on site 1, although some snow remained on the site until the end of May. On May 19th 1989, 4 cm tall Valeriana sitchensis plants were found beneath a snow cover 49 cm deep on site 1. In contrast, only a few leaves were partially unfolded by early June in most species on site 2, where snow persisted for approximately one week longer. However, by early July plants on both sites were at the same stage of development. At this time, vegetative plants were fully developed and most flowering individuals were at or past full bloom. Individuals of most species were beginning to yellow (i.e. senesce) by early August, and common and abundant species such as Gymnocarpium dryopteris, Valeriana sitchensis, and Streptopus roseus were up to 50% yellow by this time. Most flowering individuals (Epilobium angustifolium being the most notable exception - it released seed in early to mid-August) had formed or released their seed by early August. Some species exhibited differences in phenology between herb and shrub patches. Vaccinium ovalifolium and Rubuspedatus, species which were more abundant within shrub patches than herb patches, did not flower where they occurred in herb plots. Generally, herb species in shrub patches had faster rates of development in June and slower rates of development in August than the same species in herb patches. Species exhibiting this pattern, which was more pronounced on site 2 than site 1, included Streptopus roseus, Tiarella unifoliata, Gymnocarpium dryopteris and, to a lesser extent, Valeriana sitchensis and Rubus pedatus. 213 Dominance shifts (in cover) The species with the highest percent covers before and after treatment in the plots on sites 1 and 2 are presented in Table 5.11. The herb and shrub control plots, although established in 1987, were not measured for cover until the following summer and predisturbance and two month values are missing for these treatments. Results showed changes in dominance in four out of six of the herb control plots (both sites combined) between 1988 and 1989. A shift, when it did occur, was most commonly from a single dominant to several (two to three) co-dominants. New dominants were, in all cases, herb or shrub species present in the plots at the time of the first measurements. Dramatic increases in the cover of Rubusparviflorus in some herb control plots suggests a shift to shrub dominance is occurring in these areas. Shrub control plots showed no changes in species and little change in the cover of dominants between 1988 and 1989. In the herb patches clipping had no effect on the species and little effect on the cover of dominants. In five out of six plots, Valeriana sitchensis, the predisturbance dominant, had re-established 18-75% of its original cover within two months of clipping, and at 25 months cover was 62-125% of the original estimates. Epilobium angustifolium was the dominant in the other plot and, like Valeriana, it maintained its dominance after disturbance. It had re-established 70% of its original cover by the third growing season after disturbance (Table 5.11). In the shrub patches, Menziesia ferruginea and Rhododendron albiflorum, the ^ dominant species prior to clipping, generally retained dominance after treatment, but their percent cover was low. In one plot only was there a shift to the herb Valeriana sitchensis, although this species also had a low cover (10%) at 25 months in this plot. The 214 percent cover of shrubs in the clipped plots they dominated at this time was —25% (Table 5.11). Screefing resulted in species changes by August 1989 in all herb plots - on site 1 dominance shifted from Valeriana sitchensis to Epilobium angustifolium in all three plots, and on site 2 it shifted from V. sitchensis or Thalictrum occidental to Epilobium glandulosum in all plots (Plate 5.3). The Epilobium species did not dominate cover in any of the plots 12 months after screefing, at which time the predisturbance dominants still retained the highest covers (although these were —15%). The dominance of Epilobium in plots in which they were absent prior to treatment, as well as ones in which they were present before treatment indicates the colonization potential of species with both rapid invasion and population expansion strategies. The overall low rates of invasion into the screefed plots suggests that there are few species possessing such strategies in the area of the Vavenby sites. Screefing in the shrub patches did not result in shifts to Epilobium species in any of the plots. In fact, there were no shifts to herbs at all, and in 4 out of 6 plots Menziesia ferruginea, the predisturbance dominant, was still the species with the highest cover (values ranged from 5-15%) 25 months after treatment. In one plot dominance shifted to Vaccinium membranaceum, another ericaeous shrub, and in one plot it shifted to the shrub Rubusparviflorus. Covers for all species 12 months after screefing were — 5%, indicating slow recovery and invasion rates within shrub patches. It appears that the eventual recovery of the ericaceous shrubs was possible because species did not invade the screefed plots (shrub regrowth was primarily from root systems not completely removed during screefing). 215 Plate 5.3 Screefed herb plot three growing seasons after treatment on study site 2. The new post-treatment dominant is Epilobium angustifolium (fireweed) which is covering 1 0 0 % of the plot. The woman in the photograph is holding a meter stick which indicates that the average height of the vegetation was less than 1 m. Of the four species that occurred in all plots, none responded to clipping or screefing with increases in cover in either patch type. Within herb patches, the cover of Valeriana sitchensis (the most abundant herb before treatment) in control plots decreased over the study period while that of other herb species increased. The percent cover of Valeriana in clipped plots after three growing seasons was similar to that prior to treatment but cover was severely reduced in the screefed herb treatment. Within the shrub patches, Valeriana cover was low before clipping and screefing (< 5%) and remained low three seasons later in all treatments. The cover of Gymnocarpium dryopteris (the second most abundant herb prior to treatment) was severely reduced by both clipping and screefing, but cover in the controls remained unchanged over the study period. The final percent covers (< 5%) of Tiarella unifoliata and Streptopus roseus were unrelated to either clipping or screefing, and were similar to initial values in all three treatments. Unlike Valeriana (and most other species on the Vavenby sites) the cover of the other three species did not differ between the herb and shrub patches. Changes in species dominance after repeated clipping are presented in Table 5.12. These data come from the planted spruce experiment described in Chapter 3. Plots were established according to the same criteria as the invasion plots, but vegetation in the clipped treatments was removed several times during the growing season in 1988 and 1989 to provide a vegetation-free environment for the spruce seedlings. Vegetation removal was discontinued after June 1989 and percent cover of species was recorded in August to assess the impacts of repeated removal on the species composition in plots. Whereas a single removal (by clipping above-ground vegetation) did not affect species dominance in the herb patches, repeated removals resulted in species shifts in all plots in this patch type (Table 5.12). The new post-disturbance dominants were different in each of the clipped plots. Between June and August of 1989, these species developed covers of 217 10-30%. Two of the species, Rubus parviflorus and Epilobium glandulosum, were not present in the plots prior to clipping. Repeated clipping on the skid trails also caused species shifts in all but one of the plots and, as in the herb plots, post-disturbance dominance was distributed among several different species. Repeated clipping in the shrub patches resulted in a change in the dominant species in only one plot and a shift to co-dominance in another plot In the other plots, Menziesia ferruginea had the highest cover both before and 26 months after the initial treatment. However, repeated clipping greatly reduced the cover of this species, with values after 26 months ranging from 7-15% across plots. Epilobium angustifolium invaded one plot in the clipped shrub treatment, and after 26 months it was co-dominant (with a cover of 10%) with Rubus pedatus . The ability of the dominant herb Valeriana sitchensis and the dominant shrub Menziesia ferruginea to regrow after clipping was demonstrated by the rapid recovery of individual plants removed on June 6 and remeasured on August 6 in 1987 (Table 5.13). Clipped Valeriana individuals in herb patches recovered approximately 148% of their initial height and 102% of their initial width (plants were multi-stemmed) by August, whereas control plants increased to 228% of their initial height and 160% of their initial width between June and August (Table 5.13). The initial plant size and the recovery of clipped individuals were less for plants in the shrub patches than for those in the herb patches. Mature plants of Menziesia ferruginea exhibited little new growth on uncut stems, but v there was considerable growth of new shoots after cutting (Table 5.13). Clipped plants recovered 19% of their original cover in one season, whereas stems on mature plants increased only 1.5% beyond their initial length. The new growth on shoots of juvenile 218 plants (stems that had originated since logging) was proportionally greater (138% of initial height) than that of the new shoots on uncut mature plants. The average height of new stems from clipped plants of both Valeriana and Menziesia after one growing season was approximately the same as that of the planted spruce seedlings after three growing seasons. The characteristics of conifer regeneration The number, average height, condition, and rooting substrate of Engelmann spruce naturals which had established prior to 1987 in the control plots on sites 1 and 2 are given in Table 5.14. There was large between-plot variation in the amount of conifer regeneration, although seedlings were generally more abundant in shrub plots than in herb plots. The numbers of spruce in shrub plots ranged from 0 to 46 (plot size = 16 m2). In herb plots, numbers ranged from 1 to 7. Most seedlings were in good condition, and most were rooted in dead or rotten wood. The height of the trees ranged from 3.0 to 137.0 cm and did not differ between the patch types. The characteristics of spruce seedlings which established after treatment in 1987 are given in Table 5.15. Rates of invasion during the two seasons following disturbance were low but were higher in the screefed treatment than in the clipped or control treatments. Only one seedling was found in the control plots in 1989 - it was a two-year-old germinant in a herb plot on site 2. Although most seedlings occurred in the screefed plots, there was large between-plot variation within this treatment. For example, the 13 germinants in the screefed shrub treatment on site 2 were all in one plot. The rate of establishment of seedlings was similar in both years of the study -12 of the total of 29 seedlings germinated in 1988 and 17 germinated in 1989. Approximately one-half (14) of 219 the seedlings were rooted in rotten wood, 10 were in mineral soil, and the remaining 4 were rooted in humus. DISCUSSION There were differences between the herb and shrub patches in the rate at which vegetation re-established after clipping and screefing, and in the extent to which shifts in dominance occurred after treatments. Plant cover was re-established rapidly only in the clipped herb treatment where cover was —100% of the control by the end of the study. The development of cover was much slower after screefing in the herb patches, with total percent cover values after three growing seasons only 50% of those in the controls. Re-establishment was slowest after clipping and screefing in the shrub patches - total cover was — 30% of that in the controls after three growing seasons. Shifts in dominance (defined by maximum cover values) after single treatments occurred only in the herb control and screefed herb treatments. Shifts in dominant species in the controls were the result of small changes in the relative covers of species well-represented in the plots prior to disturbance - as indicated by the co-dominance of two or three species in plots in this treatment after three growing seasons. Changes in the screefed plots were generally to species which had either invaded or greatly increased in cover in plots after disturbance. Species shifts did not occur in the clipped herb treatment or in any of the shrub treatments. Why did species shifts occur only in the herb patches? The absence of species shifts in the clipped and screefed shrub treatments is somewhat surprising given the slow rates of re-establishment of the predisturbance dominant, Menziesia ferruginea. In particular, significant increases in Valeriana sitchensis might 221 have been expected, given the capacity of this species for vegetative growth. Its rapid re-establishment in the clipped herb treatment was likely made possible by carbohydrate reserves in well-developed root systems on stout rhizomes. These reserves may not exist in herbs growing beneath shrubs, as indicated by the small size of plants and their reduced capacity for regrowth after clipping in this location. Other species present before disturbance in both the herb and shrub patches also failed to show competitive release, in the sense of increases in cover after clipping or screefing. The other three species which occurred in all plots (Gymnocarpium dryopteris, Tiarella unifoliata, and Streptopus roseus) either decreased in cover or, more commonly, exhibited no change in cover in response to either type of disturbance. Keddy (1989a) also observed that herbaceous individuals beneath shrubs in lakeshore communities were much less vigorous than individuals of the same species growing without the influence of shrubs. He suggested that although the herbs could survive beneath the shrubs they could not expand their populations. He also observed slow rates of colonization after the clearing of shrubs, and suggested that insufficient propagules limited the rate of invasion into the openings. Establishment rates were very low on the Vavenby sites, as indicated by the relatively greater increases in cover in the control treatment than in either of the other two treatments over the study period, suggesting that some aspect of regeneration (e.g. production of viable seed, dispersal, germination, and establishment of seedlings) was limiting invasion. Most species produced copious seed on the sites and, with the possible exception of Engelmann spruce which was selectively removed from the area during logging, it is unlikely that the dispersal of any species was seed-limited. In addition, the sites were open and wind-dispersed seed was well-distributed across the sites. It is possible that seed and seedling predation may have limited early survival, as small mammals were very active on the study sites. Unfavourable environmental conditions may also have reduced seedling survival, although no obvious stresses (e.g. frost, drought, high surface temperatures) existed during the growing season. Finally, the slow development of vegetation after disturbance may also be due to the very slow early growth rates of most subalpine plants establishing from seed. There is little life history or autecological information on the species on the Vavenby sites with which to select among the alternative explanations presented above. However, based on the limited results from this study, it is hypothesized that seed and seedling predation and slow early growth rates are major factors limiting invasion. What processes control the structure of vegetation within patches? The results from the disturbance experiment indicate that species dominance is a major characteristic of the vegetation within the herb and shrub patches on the Vavenby sites. Yodzis (1986) has argued that dominance relationships are important determinants of community structure when competition for space is the underlying process driving community dynamics. In patchy communities of this nature, competition structures the vegetation within patches and the dispersal of species controls the structure between patches (Bowers and Dooley, 1991; Yodzis, 1986). Yodzis' concept of a dominance-controlled community can be applied to the pattern of vegetation on the Vavenby sites, where patch structure may be largely determined by competition for space and the distribution of herb and shrub patches can be explained by the effects of colonization (dispersal) events and environmental heterogeneity on the relative competitive abilities of the dominant species. If within-patch structure is generally dominance-controlled, the following results of the disturbance experiment suggest that control is stronger within the shrub patches than the herb patches: 1) there were more species in the herb patches; 2) shifts in dominance were restricted to this patch type; 3) rates of re-establishment of vegetation were faster in herb plots; 4) the constancy of neighbours from year-to-year was lower in the herb patches, and; 5) there were greater changes in species cover in the herb control plots than in the shrub control plots. According to Yodzis (1986), niche differentiation and colonization processes are two other ways (in addition to dominance) in which competition can control community structure. Perhaps the relative importance of colonization events or consumptive competition (the major process underlying niche control) in structuring communities is greater within herb patches than it is within the shrub patches, where the small changes in cover in control plots over the study period indicated strong competitive dominance by Menziesia ferruginea. If consumptive competition (defined as competition among individuals gathering a fraction of the resources from a shared space) is relatively more important in herb patches, the flush of nutrients that characteristically follows logging (and which peaks on sites in the ESSFwc2 between three and six years after logging (David, 1987)) may increase competitive intensity and result in greater community flux within herb patches. Under these conditions, changes in relative species abundance may continue for several years as the flush of nutrients gradually diminishes over time. Are patches stable and persistent? The results of the disturbance study suggest that the shrub patches on the Vavenby sites are very stable (i.e. unchanging in composition and location), and that this stability is based primarily on a high degree of persistence, rather than high resilience (the speed with which a community returns to its original condition following disturbance). Herb patches, in contrast, are less stable than the shrubs under all conditions. The patterns of patch response to treatments suggest that the factors controlling between-patch structure and within-patch structure are different, and that the relative importance of different competitive processes differs between the shrub patches and the herb patches. Generally, disturbance did not cause shifts from shrubs to herbs under any conditions, suggesting that the development of patches is influenced to some extent by within-site environmental heterogeneity, which would increase patch stability. In can be argued, however, that three years is too short a period of time to characterize early patterns of succession in subalpine environments, as was evidenced by the slow rates of vegetation development in some treatments on the Vavenby sites. Nevertheless, some comparisons of patterns in responses to disturbance can be made with other studies of succession on forest cutovers. What is the general pattern of early community development after disturbance on subalpine cutovers in the ESSFwc2? Changes following disturbance on the Vavenby sites were similar in character to those observed by Halpern (1988,1989) after logging and burning in Douglas-fir forests. Communities on the Vavenby sites were dominated largely by residual species which persisted through disturbance and by a few opportunistic species which expanded their populations either by invasion or clonal growth after disturbance. After three growing seasons, the abundance and distribution of species in all treatments except the screefed herb were most strongly related to initial species distributions. In the screefed herb treatment, they were most strongly related to colonization events. The deterministic (and hence predictable) component of community development, as discussed by Halpern (1988) was evident in the role that life history traits, specifically a species capacity for clonal growth and its dispersal ability, played in establishing dominance following disturbance. Results from this study are consistent with others which have shown the colonizing ability of a species such as Epilobium angustifolium, which produces numerous wind-dispersed seed, has seedlings with rapid early growth rates, and is capable of vegetative reproduction through vigorous rhizomatous growth (Haeussler and Coates, 1986; Halpern, 1988; Hamilton and Yearsley, 1988). Interestingly, Valeriana sitchensis also produces copious wind-dispersed seed, and can produce multiple plants from its stout rhizome. However, it lacks the capacity for lateral expansion of populations by rhizomatous growth that Epilobium has, and seedling growth rates are slow. These differences in life history traits seem to reduce the ability of Valeriana to colonize openings as aggressively as Epilobium after disturbance. Stochasticity was also evident in the early stages of community development by the variation in colonization by invading species, particularly within the herb patches. The extent to which this variation might influence future community development depends on the length of time the invading species (primarily Epilobium angustifolium, E. glandulosum, and Rubus parviflorus) can maintain dominance. Although this period may be long on other sites (Halpern, 1988; Hamilton and Yearsley, 1988; Stickney, 1986), its length on subalpine sites is not known. Stochasticity was also evident in the variation in species distributions between the two study sites. Although the two species found only on site 2 may have been associated with the higher moisture levels in some areas on this site (Figure 3.16), differences were most easily explained by stochastic variation in invasion events. These events are probably still occurring on the Vavenby sites, given that logging occurred relatively recently and that the rate of development of vegetation in subalpine environments appear to be slow. What is the pattern of Engelmann spruce regeneration within the post-disturbance community? Several characteristics of spruce regeneration were apparent from the distribution of trees which had established prior to treatment in the herb and shrub control plots. There was more regeneration in the shrub patches and Engelmann spruce was locally abundant within these patches. Light levels on the ground beneath the shrubs were higher than levels beneath the herbs (Figure 2.7) and there was some evidence that shrub patches were snow-free earlier earlier in the spring than herb patches (Plate 2.1). Both of these conditions may favour greater germination and early establishment of Engelmann spruce in shrub patches, as seeds of this species germinate in the early spring following snowmelt (Alexander and Shepperd, 1984). Another factor which may favour establishment in the shrub patches is the thinner litter layer in this patch type (Table 2.2). Knapp and Smith (1982) suggested that the establishment of spruce in mature spruce-fir forests in southern Wyoming was limited to microsites without a thick litter layer, such as decomposing wood, because of the low rates of root growth and penetration in germinants and young seedlings. On the Vavenby sites, as in the old-growth spruce-fir stand, naturals were most commonly rooted in dead wood. However, these microsites were not elevated on the cutovers as they were in the forest (Chapter 4). This suggests that association of seedlings with this substrate may be related to an inability to root on humus or the avoidance of root competition from other species that do not utilize dead wood rather than a response to light competition: the trees on the cutover were growing on the ground beneath herb and shrub canopies (unlike seedlings in the forest). The patchy occurrence of seedlings among the plots on the Vavenby sites indicates the importance of both the autecological characteristics of the species plus the stochastic events that affect seed availability (as influenced by the 227 proximity of trees, patterns of seed predation, etc.) on the distribution of naturally established seedlings of Engelmann spruce. C O N C L U S I O N S The predictability of patterns of early community development on cutovers in the ESSFwc2 will be related to the extent to which sites are dominated by herbs and shrubs, and to the amount of disturbance caused by logging or related activities. Shrub patches appear to be stable, and increasing the intensity of disturbance will decrease the rate of re-establishment of vegetation but will not result in major shifts in dominance - at least not in the first few years after disturbance. During this period, the growth of planted spruce seedlings was significantly improved by the removal of shrubs by clipping (growth in screefed treatments was not studied), suggesting that the treatment of sites dominated by shrubs has a high potential to provide better growing environments for improved planted spruce performance. In contrast, early patterns of community development will be more difficult to predict on herb-dominated sites, particularly as the intensity of disturbance increases. Regardless of the degree of disturbance, however, it is unlikely that single vegetation removals in herb-dominated communities will create conditions for improved conifer seedling growth for more than a short time (one season perhaps), because of rapid rates of regrowth in predisturbance dominants or shifts to new dominants with high rates of seedling growth. These conclusions are based on only three seasons of post-disturbance observations, however, and a longer study period is necessary to establish the relationship between patterns of vegetation development and windows for improved performance of planted spruce. 229 CHAPTER 6 SYNTHESIS AND CONCLUDING REMARKS In this concluding section the research presented in previous chapters is discussed with respect to its contribution to our understanding of the ecology of Engelmann spruce and the management of spruce regeneration on subalpine cutovers. The major results are first summarized briefly by reference to the research questions presented in Chapter 1. 1. Do the patch types on the Vavenby sites differ significantly as environments for spruce regeneration? There was large variation between the biotic and abiotic characteristics of the herb, shrub, and skid trail patches on the Vavenby sites. There were differences among the three patch types in the composition of vegetation, soil temperature and moisture, soil nitrogen and carbon, depth of litter, and light availability, and there was some evidence for differences in snow duration and parent materials. Many of these factors have been shown to influence both seed and seedling stages of regeneration in Engelmann spruce. It was speculated that the greater number of naturally established seedlings in the shrub patches than the herb patches was due to the thinner litter layers and higher light availability on the ground beneath the shrub patches. Earlier snowmelt in shrub patches may also favour regeneration in this patch type, particularly with respect to germination and early development. Despite the possibility of factors favouring early establishment in the shrubs, it was predicted that the growth of seedlings would be higher in the herb V patches because of higher levels of soil moisture and mineralizable nitrogen in this patch type. However, as discussed below, the growth of planted seedlings did not exhibit the predicted differences between patch types. 2. Does the growth and survival of planted spruce seedlings differ among the patch types? Interference from non-coniferous vegetation significantly reduced the growth of planted seedlings. Low light availability was likely a dominating limiting factor for seedlings growing within herb, shrub, and skid trail vegetation. Under these conditions the size or number of all seedling components were reduced and there were few differences between the three patch types. When vegetation was removed variations in total growth among patch types were associated with differences in below-ground factors. All seedling components exhibited increases in growth in response to the higher resource availability in vegetation-free treatments. When vegetation was removed, seedling growth in the skid trail treatments was lower than in the off-skid trail treatments. Assuming that below-ground factors were more limiting to growth than above-ground factors in vegetation-free treatments, this suggests that there were differences between the two locations in the below-ground factors limiting growth. There was little evidence of shifts in allocation among seedling components in either stock-type in response to environmental variation, and allometric changes were size-related. Differences in size between the two stock-types were related to differences at the time of planting. Survival in both stocks was high but mortality was greater in the herbs than in the other two patch types. 3. Do the characteristics of planted spruce seedlings and naturally established seedlings differ in relation to each other and their environments? In general, the characteristics of the planted and naturally established seedlings were quite similar and variation among the three groups of naturals and the two groups of planted seedlings exhibited patterns similar to those for the planted spruce treatments. Overall, the planted seedlings were larger than naturals of the same age, although height increments indicated that the growth of the shaded planted seedlings was declining relative to the naturals, whereas that of the open planted seedlings was increasing. The 231 rate of growth of naturals in the understorey of the old-growth stand was very slow, and seedlings at this location shared many characteristics with the shaded planted seedlings. One important way in which they differed, however, was in the significantly lower needle:stem biomass ratios in the planted seedlings. This lower ratio suggests an imbalance in the planted seedlings that may be related to insufficient resources for needle production under dense vegetation (the needlerstem ratios for the other three groups were the same as for the forest naturals). Light was the factor most strongly associated with the growth characteristics of the forest naturals and the two groups of planted seedlings. However, for the naturals growing in disturbed areas on the cutover, growth was more strongly correlated with soil temperatures than with light levels. One striking way in which naturals and planted seedlings differed was in the morphology of root systems. Naturals had sparse and shallow laterally-extended roots, whereas planted seedlings had very dense, nearly vertical root plugs. Weak and superficial lateral root systems are characteristic of Engelmann spruce seedlings (Alexander and Shepperd, 1984) but the significance of this for early survival and growth is unknown. Perhaps growth at greater depths is limited by low soil temperatures or low nutrient availability. There was no evidence in this study to suggest that the characteristics of the root systems on the planted seedlings had adverse effects on growth or survival, but the possibility of such effects merits further study. 4. Do the patch types differ significantly in their responses to disturbance? There were differences both within- and between-patches in the responses of vegetation to disturbance on the Vavenby sites. Vegetation in the shrub patches re-established slowly after disturbance but shifts to dominance by new species did not occur, regardless of the type of disturbance. Three growing seasons following clipping and screefing, plants in shrub plots were taller than planted spruce seedlings, although in most cases the cover of non-crop species was still sparse. Dominance within undisturbed shrub patches did not change over the study period but herbs beneath shrub canopies increased in cover, suggesting that communities were not in equilibrium. Vegetation in the herb patches re-established much more quickly, and shifts in dominance (in percent cover) to new species of herbs and shrubs occurred when disturbance was of high frequency or intensity. Dominance shifts also occurred within undisturbed herb patches and vegetation development was more dynamic and less predictable in these patches than in undisturbed patches of shrubs. What can this research contribute to improving management strategies for spruce regeneration on subalpine cutovers? The results of this research indicate that the presence of well-developed non-coniferous vegetation on winter-logged sites in the ESSF reduces rates of establishment of both artifically and naturally regenerated spruce. Site preparation treatments that remove vegetation may increase the window for early growth and survival but vegetation will re-establish dominance over the seedling environment quickly compared to rates of seedling growth, particularly in herb-dominated communities. Vegetation control will increase early survival more than early growth and the relative costs and benefits of treatments will likely be different depending on whether improved survival or improved growth is the management objective. Both new (post-logging) and advanced regeneration were present on the Vavenby sites (Table 1.1). However, a review of the current stocking standards for the ESSFwc2 (Table 6.1 - Lloyd et al, 1990) suggests that the regeneration would fail to meet the current Table 6.1 Stocking Standards for the ESSFwc2 (from Lloyd et al, 1990). 233 TARGETS STANDARD Silvicultural system (in order of preference) 1. Clearcut 2. Shelterwood 3. Selective Acceptable regeneration height (cm) 15 Target stocking at free growing (stems/ha) 1200 Target stocking at regeneration stage (stems/ha) 1300 Leader growth at free growing (cm/year) 25 MNTMUMS Preferred species Engelmann spruce Acceptable species Subalpine fir (Lodgepole pine) Minimum stocking at free growing (stems/ha) 700 Minimum inter tree spacing (m) 1.6 Regeneration delay (years) 4 Earliest to latest free growing period (years) 12-20 Conifer/brush ratio (%) 125 Based on standards for the 01 Site unit (Bl - Azalea - Oak fern). Acceptable regeneration height = minimum acceptable height of a naturally regenerated tree at the regeneration stage. Minimum stocking standard = lowest number of acceptable, well-spaced stems per hectare required to consider an area satisfactorily restocked at free-growing stage. Free-growing trees are of preferred or acceptable species, meet minimum stocking standards, include natural regeneration at least 5 years old, planted trees that have been in the ground at least 5 years, have good form, are free of insects, disease, and damage, and are free of competing vegetation within a im radius of the crop tree. Conifer/brush ratio - minimum height that a well-spaced crop tree must attain in relation to competing brush with a im radius. A ratio of 125% means that the crop tree must be 25% taller than competing brush. standards for this variant and that the Vavenby sites would be classified as unsatisfactorily restocked at present (in fact, the sites were assessed in 1986 using slightly different standards and were classified as understocked). The standards presented in Table 6.1 would be very difficult to meet even on good sites in a relatively productive part of the ESSF and comparisons of seedling growth and soil nutrient contents on the Vavenby sites with other ESSF cutovers (Chapter 3) indicated that the Vavenby sites were good growing sites for Engelmann spruce. In particular, the standards for conifer/brush ratio, intertree spacing distances, and target leader growth at free-growing are unlikely to be met by enough trees to satisfy the minimum stocking level of 700 stems/ha. Using the Vavenby sites as an example, to achieve the recommended conifer/brush ratio for trees in the herb patches (the lower vegetation) trees must attain an average height of 1.12 m to be considered free-growing (based on an average maximum height of 90 cm for the herbs - Table 2.1). The number of stems > 1.3 m on the Vavenby sites was only 710 (Table 1.1) and many of these would not have met the other criteria for free-growing. By the free-growing period (i.e. the time by which stocking standards and growth criteria must be satisfied, which is 12 to 20 years after logging in the ESSFwc2) many of the post-logging naturals may have entered this size class. However, regeneration assessments are initially carried out four years after harvesting in the ESSF to predict whether future stocking targets will be met. Four years after logging most new regeneration would not meet the 15 cm minimum height criterion to be included in the survey. The stocking standards in Table 6.1 are biased against natural regeneration because they do not reflect several common characteristics of conifer regeneration in subalpine environments, particularly the clumped distribution of regeneration, the slow invasion and early growth rates of seedlings, and the establishment of seedlings beneath non-coniferous vegetation. The standards are biased towards planting as the method of re-establishing conifers on these subalpine cutovers to ensure stocking targets are satisfied. If planting is to be the method of choice, then improving the performance of planted Engelmann spruce will depend on increasing the level of resources available to the seedlings by removing interfering vegetation and ensuring that seedlings can utilize the resources as fully as possible by planting stock-types with a good growth capacity. Spruce is a weak competitor and resources must be delivered directly to the seedling to maximize growth. On the Vavenby sites, there was no evidence that hot-planting resulted in better growth or survival than planting cold-stored seedlings. On other types of sites, such as those where summer frost frequently damages first year growth on cold-stored stock, differences between the two stock-types might be significant. However, Lavender (1990) warns against the use of hot stock on northern sites where the risk of exposure to summer frost is great. Although these seedlings appear dormant, they are not sufficiently cold-hardy to resist damage from below-freezing temperatures. Although disturbance will increase conifer establishment on subalpine cutovers, clearcuts of any size in the ESSF are unlikely to meet current regeneration standards because rates of invasion and early growth of Engelmann spruce are too far below silvicultural expectations. However, this does not mean that conifers do not invade most clearcuts. Partial cutting with minimal understorey disturbance will be satisfactory only if advanced regeneration meets the standards and a good assessment of this can be done in a preharvest evaluation of the amount and condition of regeneration present in a stand, as well as the probable effects of logging on these trees. The satisfactory restocking of older cutovers that have failed to satisfy stocking standards in the southern interior will require a large input of resources in planting and site treatments. Because natural regeneration is establishing on these sites, a change in the stocking standards or a lengthier period of time before standards must be met might reduce the need for treatments of dubious cost-effectiveness. What do the results of this research contribute to our scientific understanding of subalpine ecology in general, and the regeneration strategy of spruce in particular? In earlier chapters, the concepts of differentiation in the regeneration niche and the role of resource allocation in competitive strategies were introduced as part of a theoretical framework around which to build an understanding of spruce seedling responses to environmental variation. However, neither concept was very useful for interpreting the adaptive significance of seedling responses in terms of the regeneration strategy of Engelmann spruce. The characteristics of the planted and natural spruce seedlings suggest that differentiation in the seedling establishment stage (of the regeneration niche) is not the key to spruce persistence in subalpine environments. As Yodzis (1986) pointed out, niche differentiation is not necessary to explain species coexistence within communities when communities are controlled by competition for space. Spruce establishes under many different environmental conditions (as discussed in Chapter 1) and while safe sites (sensu Harper, 1977) may be extremely limited in number and type in one environment (as in the old-growth stand in this study), the sites suitable for establishment under other conditions (e.g. on the cutovers) may be very different. However, in this study it was the environment of older seedlings that were found to differ from each other: it is not known whether the same is true for germinants and very young seedlings. There was no evidence to suggest that spruce was competitively superior in any of the environments within which it was growing either naturally or after planting. Growth under low but variable resource conditions was slow and seedlings did not shift the allocation of carbon in response to differences in resource availability. 237 Spruce seedlings did not exhibit any of the characteristics one might assoicate with a high competitive ability, such as phenotypic plasticity, high invasion rates, or rapid early growth rates. Low rates of carbon accumulation controlled by either intrinsic or extrinsic factors may necessitate a conservative rather than a competitive strategy in spruce seedlings. The role of disturbance in patch dynamics and the competition for space were useful concepts for examining the structure of post-logging communities in the ESSF. Disturbance is important in creating persistent patterns that influence vegetation development within these communities. The dominant understorey species, which persist through disturbance, are not as affected by overstorey removal as species like spruce, which must re-enter the community after logging and which are strongly influenced by the patterns of disturbance. If the understorey dominants control space in post-logging communities, how can spruce re-enter these environments? If environmental factors significantly limit seed abundance (i.e. climatic conditions limit seed production and biotic and abiotic factors result in high seed mortality), then the the successful establishment of the relatively few seedlings that do survive will be critical to spruce regeneration. In subalpine environments, however, there are many biotic and abiotic factors with large spatial and temporal variation that can limit seedling establishment. A strategy of conservation (rather than competition) in a system with slow dynamics seems most consistent with the observed responses of spruce seedlings to environmental variation. Such a strategy may limit a population or an individual's ability to respond competitively to higher resource v levels (such as occur after logging) but may ensure persistence in environments where resource availability is low for long periods of time (such as in a stand understorey). 238 The results of this research suggest a number of areas for future study on both the silviculture and ecology of spruce regeneration. More detailed studies of natural regeneration on cutovers in the ESSF are required to understand the relationship between characteristics of early regeneration (particularly growth rates, density, and proportions of spruce and fir) and later stages of individual tree and stand development. This might result in a re-evaluation of current stocking standards for ESSF sites and reduce the amount of post-logging treatment currently required to successfully regenerate cutovers. Finally, information on the long-term recruitment of spruce seedlings from beneath canopies of herbs and shrubs would enable a better assessment of the relative costs and benefits of vegetation management in the ESSFwc2. Where planting is selected as the regeneration option, the use of stock-types with a good growth capacity will be critical to improving early seedling performance. Differences between planted seedlings and naturally established spruce were discussed earlier in this chapter. There was no evidence from this study to suggest that these differences significantly affected the performance of planted trees. However, there is still little known about the relationship between nursery conditioning and outplanting performance in Engelmann spruce. The optimum seedling morphology for subalpine environments is still by no means evident, and until it is it will be difficult to design a nursery enironment to produce optimum planting stock for high elevation cutovers. More research on such things as root characteristics, size criteria, and genetic variability is required if planting is to continue as the recommended method of regeneration. In this study, the effects of individual factors were not isolated for their effects on seedling growth and survival. Nevertheless, the results suggest that there are several factors which have a significant impact on spruce regeneration in the ESSFwc2 but about which very little is known. These include snow cover (depth, time of arrival and departure, effects on seedling physiology, etc.), seed predation and seedling damage from insects and small mammals, and the role of nutrients in early seedling growth. In conclusion, a better understanding of the natural patterns of early regeneration and long-term recruitment after disturbance in subalpine forests will not only benefit our understanding of the ecology of long-lived species such as Engelmann spruce but will also assist managing these species as a sustainable resource. 240 REFERENCES Abrahamson, W.G. and Caswell, H. 1982. On the comparative allocation of biomass, energy, and nutrients in plants. Ecology 63:982-991. Alexander, R.R. 1984. Natural regeneration of Engelmann spruce after clearcutting in the central Rocky Mountains in relation to environmental factors. USDA For. Ser. Res. Pap. RM-254. Fort Collins, Colorado. 15 pp. Alexander, R.R. 1987. Ecology, silviculture, and management of the Engelmann spruce -subalpine fir type in the central and southern Rocky Mountains. USDA For.Ser. Agricultural Handbook No. 659. Fort Collins, Colorado. 144 pp. Alexander, R.R. and Shepperd, W.D. 1984. Silvical characteristics of Engelmann spruce. USDA For. Ser. Gen. Tech. Rep. RM-114. Fort Collins, Colorado. 38 pp. Amaranthus, M.P. and Perry, D.A. 1989. Interaction effects of vegetation type and Pacific madrone soil inocula on survival, growth, and mycorrhiza formation of Douglas-fir. Can. J. For. Res. 19:550-556. Amaranthus, M.P., Parrish, D.S. and Perry, D.A. 1989. Decaying logs as moisture reservoirs after drought and wildfire.'/n: Proceedings of Watershed '89: Conference on the Stewardship of Soil, Air and Water Resources. Alexander,E.B. (ed.). March 21-23 1989. Juneau, Alaska. USDA For. Ser. RIO-MB-77. pp. 191-194. Angove, K. and Bancroft, B. 1983. A Guide to Some Common Plants of the Southern Interior of British Columbia. Land Management Handbook No. 7. British Columbia Ministry of Forests. Victoria, B.C. 225 pp. Aplet, G.H., Laven, R.D. and Smith, F.W. 1988. Patterns of community dynamics in Colorado Engelmann spruce - subalpine fir forests. Ecology 69:312-319. Aplet, G.H., Smith, F.W. and Laven, R.D. 1989. Stemwood biomass and production during spruce-fir stand development. J. Ecology 77:70-77. Armson, K.A. 1977. Forest Soils: Properties and Processes. University of Toronto Press. Toronto, Ont. 390 pp. Baldwin, H.1.1927. A humus study in Norway. Ecology 8:380-383. Ballard, T.M. 1972. Subalpine soil temperature regimes in southwestern British Columbia. Arctic and Alpine Research 4:139-146. Barnes, G.H. 1937. The development of unevenaged stands of Engelmann spruce, and probable development of residual stands after logging. For. Chron. 13:417-457. Bassman, J.H. 1989. Influence of two site preparation treatments on ecophysiology of planted Picea engelmannii x glauca seedlings. Can. J. For. Res. 19:1359-1370. Binkley, D. 1984. Douglas-fir stem growth per unit leaf area increased by interplanted sitka alder and red alder. For. Sci. 30:259-263. 241 Blake, G.R. 1965. Bulk density. In: Methods of Soil Analysis. Part 1. First edition. Black,C.A. (ed.). American Society of Agronomy. Madison, Wisconsin, pp. 374-390. Bloomberg, W.J. 1950. Fire and spruce. For. Chron. 26:157-161. Bormann, F.H. and Likens, G.E. 1979. Pattern and Process in a Forested Ecosystem. Springer-Verlag. New York. 253 pp. Bowers, M.A and Dooley, J.L.,Jr. 1991. Landscape composition and the intensity and outcome of two-species competition. Oikos 60:180-186. Brand, D.G. 1991. The establishment of boreal and sub-boreal conifer plantations: an integrated analysis of environmental conditions and seedling growth. For. Sci. 37:68-100. Brand, D.G., Weetman, G.F. and Rehsler, P. 1987. Growth analysis of perennial plants: the relative production rate and its yield components. Ann. Bot. 59:45-53. Burdett, A.N., Herring, L.J. and Thompson, C.F. 1984. Early growth of planted spruce. Can. J. For. Res. 14:644-651. Butt, G. 1989. Forest Regeneration in the ESSFf Subzone of the Merritt Forest District: An Historical Analysis. Unpublished report to Merritt Forest District, British Columbia Ministry of Forests. Merritt, B.C. 59 pp. Butt, G. 1990. Forest Regeneration in the ESSFa Subzone: A Problem Analysis. FRDA Report 118. British Columbia Ministry of Forests. Victoria, B.C. 47 pp. Butt, G., Bancroft, B. and Folk, R. 1989. Ingress of Engelmann spruce and Subalpine Fir in the Southern Interior, ESSF. Unpublished report to FRDA Southern Interior Technical Advisory Committee. British Columbia Ministry of Forests. Kamloops, B.C. 32 pp. Canada Soil Survey Committee. 1978. The Canadian System of Soil Classification. Canada Department of Agriculture Publication 1646. Supply and Services. Ottawa, Ont. 164 pp. Carr, W.W. 1987. The Effect of Landing Construction on Some Forest Soil Properties: A Case Study. FRDA Report 003. British Columbia Ministry of Forests. Victoria, B.C. 20 pp. Carter, G.A and Smith, W.K. 1988. Microhabitat comparisons of transpiration and photosynthesis in three subalpine conifers. Can. J. Bot. 66:963-969. Carter, G.A., Smith, W.K. and Hadley, J.L. 1988. Stomatal conductance in three conifer species at different elevations during the summer in Wyoming. Can. J. For. Res. 18:242-246. Carter, R.E. 1991. Unpublished data from a study of the physical and chemical properties of soils in Douglas-fir plantations in southwestern B.C. Faculty of Forestry, University of British Columbia. Vancouver, B.C. 242 Carter, R.E., Klinka, K., Schrier, H. and Lavkulich, L.M. 1985. Ecological Approach to Differentiating Podzols in Coastal Southwestern British Columbia. Research Note No. 95. British Columbia Ministry of Forests. Victoria, B.C. 88 pp. Caza, CL. and Kimmins, J.P. 1987. Composition and structure of plant communities on NSR cutovers in the ESSF, Clearwater Forest District Unpublished report to FRDA Southern Technical Advisory Committee. British Columbia Ministry of Forests. Kamloops, B.C. 96 pp. Chan, S.S. and Walstad, J.D. 1987. Correlations between overtopping vegetation and development of Douglas-fir seedlings in the Oregon Coast Range. West. J. Appl. For. 2:117-119. Chapin, F.S.JU, Bloom, A.J., Field, CB. and Waring, R.H. 1987. Plant responses to multiple environmental factors. BioScience 37:49-57. Childs, S.W. and Flint, L.E. 1987. Effect of shadecards, shelterwoods, and clearcuts on temperature and moisture environments. For. Ecol. Management 18: 205-217. Christy, E.J. and Mack, R.N. 1984. Variation in demography of juvenile Tsuga heterophylla across the substratum mosaic. J. Ecology 72:75-91. Clements; F.E. 1928. Plant Succession and Indicators. H.W. Wilson Comp. New York. 453 pp. Coates, K.D. 1987. Effects of Shrubs and Herbs on Conifer Regeneration and Microclimate in the Rhododendron-Vaccinium-Menziesia Community of South-central British Columbia. M.Sc. Thesis. Oregon State University. Corvallis, Oregon. 93 pp. Cole, E.C. and Newton, M. 1987. Fifth-year responses of Douglas-fir to crowding and non-coniferous competition. Can. J. For. Res. 17:181-186. Collins, S.L. 1990. Habitat relationships and survivorship of tree seedlings in hemlock-hardwood forest. Can. J. Bot. 68:790-797. Collins, S.L. and Good, R.E. 1987. The seedling regeneration niche: habitat structure of tree seedlings in an oak-pine forest. Oikos 48:89-98. Comeau, P.G. and Kimmins, J.P. 1989. Above- and below-ground biomass and production of lodgepole pine on sites with differing soil moisture regimes. Can. J. For. Res. 19:447-454. Conard, S. 1984. Forest vegetation management in British Columbia: problem analysis. Research Paper RR84001-HQ. British Columbia Ministry of Forests. Victoria, B.C. 119 pp. Daniels, R.F., Burkhart, H.E. and Clason, T.E. 1986. A comparison of competition measures for predicting growth of loblolly pine trees. Can. J. For. Res. 16:1230-1237. 243 Daniels, T.G. and Simpson, D.G. 1990. Seedling production and processing: bareroot. In: Regenerating British Columbia's Forests. Lavender,D.P., Parish,R., Johnson,CM., Montgomery,G., Vyse,A, Willis,R.A and Winston,D. (eds.). University of British Columbia Press. Vancouver, B.C. pp. 206-225. David, CA. 1987. Post-harvesting forest floor changes and nitrogen mobilization in an Engelmann Spruce - Subalpine Fir Forest. PhD thesis. Faculty of Forestry, University of British Columbia. Vancouver, B.C. 325 pp. Day, P.R. 1965. Particle fractionation and particle-size analysis. In: Methods of Soil Analysis. Part 1. First edition. Black,C. A (ed.). American Society of Agronomy. Madison, Wisconsin, pp. 545-567. Day, R.J. 1972. Stand structure, succession, and use of southern Alberta's Rocky Mountain Forest. Ecology 53:472-478. Day, T.A, DeLucia, E.H. and Smith, W.K. 1989. Influence of cold soil and snowcover on photosynthesis and leaf conductance in two Rocky Mountain conifers. Oecologia 80:546-552. Day, T.A., DeLucia, E.H. and Smith, W.K. 1990. Effect of soil temperature on stem sap flow, shoot gas exchange and water potential oiPicea engelmannii (Parry) during snowmelt. Oecologia 84:474-481. Del Moral, R. 1983. Competition as a control mechanism in subalpine meadows. Amer. J. Bot. 70:232-245. Del Moral, R. 1985. Competitive effects on the structure of subalpine meadow communities. Can. J. Bot. 63:1444-1452. Del Moral, R. and Cates, R.G. 1971. Alleopathic potential of the dominant vegetation of western Washington. Ecology 52:1030-1037. DeLucia, E.H. and Smith, W.K. 1987. Air and soil temperature limitations on photosynthesis in Engelmann spruce during summer. Can. J. For. Res. 17:527-533. Denslow, J.S. 1985. Disturbance-mediated coexistence of species. In: The Ecology of Natural Disturbance and Patch Dynamics. Pickett, S.T.A. and White, P.S. (eds.). Academic Press. Toronto, Ont. pp. 307-323. DiStefano, J. and Gholz, H.L. 1986. A proposed use of ion exchange resin to measure nitrogen mineralization and nitrification in intact soil cores. Comm. Soil Sci. PI. Anal. 17:989-998. Dobbs, R.C. and McMinn, R.G. 1971. Regeneration of Spruce/Alpine Fir Types in British Columbia. Internal Report BC-24. Canadian Forest Service. Victoria, B.C 26 pp. Draper, D., Binder, W., Fahlman, R. and Spittlehouse, D. 1985. Post-planting ecophysiology of interior spruce. In: Interior Spruce Seedling Performance. February 5-6th 1985 Workshop Proceedings. Northern Silviculture Committee. British Columbia Ministry of Forests. Prince George, B.C. 20 pp. Egler, F.E. 1954. Vegetation science concepts I. Initial floristic composition, a factor in old-field vegetation development. Vegetatio 4:412-417. Eis, S. 1981. Effect of vegetative competition on regeneration of white spruce. Can. J. For. Res. 11:1-8. Eis, S. and Craigdallie, D. 1983. Regeneration and growth of white spruce and alpine fir in the interior of British Columbia. Can. J. For. Res. 13:339-343. Elliot, KJ. and White, AS. 1987. Competitive effects of various grasses and forbs on ponderosa pine seedlings. For. Sci. 33:356-366. Evans, R.D. and Fonda, R.W. 1990. The influence of snow on subalpine meadow community pattern, North Cascades, Washington. Can. J. Bot. 68:212-220. Feller, M. 1990. Unpublished data from a study of the effects of prescribed burning on nutrients in soils on ESSF cutovers in the Clearwater Forest District. Faculty of Forestry, University of British Columbia. Vancouver, B.C. Franklin, J.F., Moir, W.H., Douglas, G.W. and Wiberg, C. 1971. Invasion of subalpine meadows by trees in the Cascade Range, Washington and Oregon. Arctic and Alpine Research 3:215-224. Fox, D.J. and Guire, K.E. 1976. Documentation of MIDAS. Statistical Research Laboratory. University of Michigan. Illinois. 203 pp. Funck, P., Banerjee, S. and Maze, J. 1990. The structure of variation and correlations in Abies amabilis from southwestern British Columbia as assessed through a provenance test. Can. J. Bot. 68:1796-1802. Gardner, W.H. 1965. Water content. In: Methods of Soil Analysis. Part 1. First edition. Black,CA. (ed.). American Society of Agronomy. Madison, Wisconsin, pp. 82-127. Gijsman, A.J. 1991. Soil water content as a key factor determining the source of nitrogen (NrLi+ or NO3") absorbed by Douglas-fir (Pseudotsuga menziesii) and the pattern of rhizosphere pH along its roots. Can. J. For. Res. 21:616-625. Gould, S.J. 1966. Allometry and size in ontogeny and phylogeny. Biol. Rev. 41:587-640. Grier, C C , Vogt, K.A, Keyes, M.R. and Edmonds, R.L. 1981. Biomass distribution and above- and below-ground production in young and mature Abies amabilis zone ecosystems of the Washington Cascades. Can. J. For. Res. 11:155-167. Griffith, B.G. 1931. The natural regeneration of spruce in central British Columbia. For. Chron. 7:204-219. Grime, J.P. 1979. Plant Strategies and Vegetation Processes. John Wiley & Sons. New York. 222 pp. Grubb, P.J. 1977. The maintenance of species richness in plant communities: the importance of the regeneration niche. Biol. Rev. 52:107-145. Hadley, J.L. and Smith, W.K. 1986. Wind effects on needles of timberline conifers: seasonal influence on mortality. Ecology 67:12-19. 245 Haeussler, S. and Coates, D. 1986. Autecological characteristics of selected species that compete with conifers in British Columbia: A literature review. Land Management Report No. 33. British Columbia Ministry of Forests. Victoria, B.C. 180 pp. Halpern, CB. 1988. Early successional pathways and the resistence and resilience of forest communities. Ecology 69:1703-1715. Halpern, CB. 1989. Early successional patterns of forest species: interactions of life history traits and disturbance. Ecology 70:704-720. Hamilton, E. and Yearsley, K. 1988. Vegetation development after clearcutting and site preparation in the SBS zone. FRDA Report 018. British Columbia Ministry of Forests. Victoria, B.C 66 pp. Harmon, M.E. 1989. Retention of needles and seeds on logs of Picea sitchensis - Tsuga heterophylla forests of coastal Oregon and Washington. Can. J. Bot. 67:1833-1837. Harmon, M.E. and Franklin, J.F. 1989. Tree seedlings on logs in Picea-Tsuga forests of Oregon and Washington. Ecology 70:48-59. Harmon, M.E., Franklin, J.F., Swanson, F.J., Sollins, P., Gregory, S.V., Lattin, J.D., Anderson, N.H., Cline, S.P., Aumen, N.G., Sedell, J.R., Lienkaemper, G.W., Cromack, K.,Jr. and Cummins, K.W. 1986. Ecology of coarse woody debris in temperate ecosystems. Adv. Ecol. Res. 15:133-302. Harper, J.L. 1977. Population Biology of Plants. Academic Press. New York. 892 pp. Harrington, T.B. and Tappeiner, J.C.,IL 1991. Competition affects shoot morphology, growth duration, and relative growth rates of Douglas-fir saplings. Can. J. For. Res. 21:474-481. Hellmers, H, Genthe, M.K. and Ronco, F. 1970. Temperature affects growth and development of Engelmann spruce. For. Sci. 16:447-452. Herring, L.J. and McMinn, R.G. 1980. Natural and advance regeneration of Engelmann spruce and subalpine fir compared 21 years after site treatment. For. Chron. 56:55-57. Hilborn, R. and Stearns, S.C. 1982. On inference in ecology and evolutionary biology: the problem of multiple causes. Acta Biotheoretica 31:145-164. Hitchcock, CL. and Cronquist, A. 1973. Flora of the Pacific Northwest. University of Washington Press. Seattle, Washington. 730 pp. Hoekstra, T.W., Allen, T.F.H. and Flather, CH. 1991. Implicit scaling in ecological research. Bioscience 41:148-154. Hunt, R. 1982. Plant Growth Curves: The Functional Approach to Plant Growth Analysis. Edward Arnold. London. 248 pp. Johnson, EA. and Fryer, G.I. 1989. Population dynamics in lodgepole pine - Engelmann spruce forests. Ecology 70:1335-1345. 246 Karakatsoulis, J. 1989. Unpublished data from a study of the effects of vegetation on the ecophysiology of planted spruce in the Nelson Forest Region in B.C. Faculty of Forestry, University of British Columbia. Vancouver, B.C. Kaufmann, M.R. 1975. Leaf water stress in Engelmann spruce. Plant Physiol. 56:841-844. Kaufmann, M.R. 1976. Stomatal response of Engelmann spruce to humidity, light, and water stress. Plant Physiol. 57:898-901. Kaufmann, M.R. 1979. Stomatal control and the development of water deficit in Engelmann spruce seedlings during drought. Can. J. For. Res. 9:297-304. Keddy, P.A. 1989a. Effects of competition from shrubs on herbaceous wetland plants: a 4-year field experiment. Can. J. Bot. 67:708-716. Keddy, P.A. 1989b. Competition. Chapman and Hall. New York. 202 pp. Keeney, D.R. 1982. Nitrogen - availability indices. In: Methods of Soil Analysis. Part 2. Second edition. Page, AL. (ed.). American Society of Agronomy. Madison, Wisconsin, pp. 711-733. Kimmins, J.P. 1987. Forest Ecology. MacMillan Publ. Comp. New York. 531 pp. Klinka, K., Green, R.N., Trowbridge, R.L. and Lowe, L.E. 1981. Taxonomic Classification of Humus Forms in Ecosystems of British Columbia. Land Management Report No. 8. British Columbia Ministry of Forests. Victoria, B.C. 54 pp. Klinka, K., Krajina, V.J., Ceska, A. and Scagel, AM. 1989. Indicator Plants of Coastal British Columbia. University of British Columbia Press. Vancouver, B.C. 288 pp. Knapp, A.K. and Smith, W.K. 1982. Factors influencing understory seedling establishment of Engelmann spruce (Picea engelmannii) and subalpine fir (Abies lasiocarpa) in southeast Wyoming. Can. J. Bot. 60:2753-2761. Knapp, AK. and Smith, W.K. 1989. Influence of growth form on ecophysiological responses to variable sunlight in subalpine plants. Ecology 70:1069-1082. Kurz,W.A. 1989. The significance of shifts in carbon allocation to long-term site procuctivity research. In: Research Strategies for Long-term Site Productivity. Proc. International Energy Agency BE A3 Workshop, Seattle, WA August 1988. Dyck,W.J. and Mees,C.A (eds.). IE A/BE A3 Report No. 8.16 pp. Larcher, W. 1985. Winter stress in high mountains. In: Establishment and Tending of Subalpine Forests: Research and Management. Proceedings of the 3rd IUFRO Workshop. Turner,H. and Tranquillini,W. (eds.). Report 270. Swiss Fed. Inst. For. Res. Birmensdorf, Switzerland, pp 11-19. Lavender, D.P. 1990. Physiological principles of regeneration. In: Regenerating British Columbia's Forests. Lavender,D.P., Parish,R., Johnson,C.M., Montgomery,G., Vyse, A, Willis,R.A. and Winston,D. (eds.). University of British Columbia Press. Vancouver, B.C. pp. 30-44. 247 Lavender, D.P., Parish, R., Johnson, C M . , Montgomery, G., Vyse, A., Willis, R.A. and Winston, D. (eds.). 1990. Regenerating British Columbia's Forests. University of British Columbia Press. Vancouver, B.C. 371 pp. Ledig, F.T. and Perry, T.0.1965. Physiological genetics of the shoot-root ratio. Proc. Soc. Am. For. 19:39-43. Lieffers, V J . and Titus, S.J. 1989. The effects of stem density and nutrient status on size inequality and resource allocation in lodgepole pine and white spruce seedlings. Can. J. Bot. 67:2900-2903. Lloyd, D. 1991. Unpublished data from a study of the effects of site preparation treatments on the early growth and survival of planted spruce and pine in the Salmon Arm Forest District of B.C. Kamloops Forest Region, B.C. Ministry of Forests. Kamloops, B.C. Lloyd, D., Angove, K., Hope, G. and Thompson, C. 1990. A Guide to Site Identification and Interpretation for the Kamloops Forest Region. Land Management Handbook No. 23. British Colmbia Ministry of Forests. Victoria, B.C. 398 pp. Lowdermilk, W.C. 1925. Factors affecting reproduction of Engelmann spruce. J. Agric. Res. 30:995-1009. MacMahon, J.A. 1979. Ecosystems over time: succession and other types of change. In: Forests: Fresh Perspectives from Ecosystem Analysis. Proceedings of the 40th Annual Biology Colloquium. Waring,R.H. (ed.). Oregon State University Press. Corvallis, Oregon, pp. 27-58. Magurran, A.E. 1988. Ecological Diversity and Its Measurement. Princeton University Press. Princeton, N.J. 179 pp. Marr, J.W. 1977. The development and movement of tree islands near the upper limit of tree growth in the southern Rocky Mountains. Ecology 58:1159-1164. Mather, J. 1987. Assessment of Silvicultural Practices in the Engelmann spruce -Subalpine fir Zone in the Kamloops Forest Region. Unpublished report to Kamloops Forest Region, British Columbia Ministry of Forests. Kamloops, B.C. 158 pp. McCullough, H.A 1948. Plant succession on fallen logs in a virgin spruce-fir forest. Ecology 29:508-513. McPherson, G.R. and Wright, H.A. 1989. Direct effects of competition on individual juniper plants: a field study. J. Appl. Ecol. 26:979-988. Meidinger, D. and Pojar, J. 1991. Ecosystems of British Columbia. Special Report Series No. 6. British Columbia Ministry of Forests. Victoria, B.C. 330 pp. Ministry of Forests. 1990. The Challenge of Regenerating the ESSF. Unpublished course notes. Southern Interior Extension Program. British Columbia Ministry of Forests. Kamloops, B.C. 248 Messier, C 1991. Factors Limiting Early Conifer Growth in Salal-dominated Cutovers on Northern Vancouver Island, British Columbia. Ph.D. Thesis. Faculty of Forestry, University of British Columbia. Vancouver, B.C 160 pp. Mueggler, W.F. 1965. Ecology of serai shrub communities in the cedar-hemlock zone of northern Idaho. Ecol. Monog. 35:165-185. Munson, A.D. and Timmer, V.R. 1990. Site-specific growth and nutrition of planted Picea mariana in the Ontario Clay Belt. HI. Biomass and nutrient allocation. Can. J. For. Res. 20:1165-1171. Naeem, S. 1990. Patterns of the distribution and abundance of competing species when resources are heterogeneous. Ecology 71:1422-1429. Newton, M. and Cole, E.C. 1991. Root development in planted Douglas-fir under varying competitive stress. Can. J. For. Res. 21:25-31. Newton, M. and Preest, D.S. 1988. Growth and water relations of Douglas-fir (Pseudotsuga menziesii) seedlings under different weed control regimes. Weed Sci. 36:653-662. Noble, D.L. and Alexander, R.R. 1977. Environmental factors affecting natural regeneration of Engelmann spruce in the central Rocky Mountains. For. Sci. 23:420-429. Oliver, CD. and Larson, B.C. 1990. Forest Stand Dynamics. McGraw-Hill. Toronto, Ont. 467. Oosting, HJ. and Reed, J.F. 1952. Virgin spruce-fir of the Medicine Bow Mountains, Wyoming. Ecol. Monog. 22:69-91. Pattern, D.T. 1963. Light and temperature influence on Engelmann spruce seed germination and subalpine forest advance. Ecology 44:817-818. Pickett, S.T.A. and White, P.S. (eds.). 1985. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press. Toronto, Ont. 472 pp. Pielou, E.C. 1984. The Interpretation of Ecological Data. John Wiley & Sons. Toronto, Ont. 263 pp. Popper, K.R. 1959. The Logic of Scientific Discovery. Harper Torchbooks, London. 480 pp. Radosevich, S.R. 1987. Methods to study crop and weed interactions. In: Altieri.M.A. and M.Z.Liepman (eds.). Weed Management in Agroecosystems: Ecological Approaches. CRC Press. (Manuscript), pp. 1-50. Radosevich, S.R. and Holt, J.S. 1984. Weed Ecology: Implications for Vegetation Management. John Wiley and Sons. Toronto, Ontario. 265 pp. Richards, B.N. 1987. The Microbiology of Terrestrial Ecosystems. Longman Scientific & Technical. Harlow, England. 399 pp. 249 Richards, L.A. 1965. Physical condition of water in soil. In: Methods of Soil Analysis. Part 1. First edition. Black,CA (ed.). American Society of Agronomy. Madison, Wisconsin, pp. 128-152. Romme, W.H. and Knight, D.H. 1981. Fire frequency and subalpine forest succession along a topographic gradient in Wyoming. Ecology 62:319-326. Ronco, F. 1970. Influence of high light intensity on survival of planted Engelmann spruce. For. Sci. 16:331-339. Ronco, F. and Noble, D.L. 1971. Engelmann spruce regeneration in clearcut not insured by record seed crop. J. Forestry 69:578-579. Scagel, R.K. 1984. Morphological and Anatomical Variation of Picea in Southwestern British Columbia. M.Sc. Thesis. Department of Botany, University of British Columbia. Vancouver, B.C. 313 pp. Schimpf, D.J., Henderson, J.A and MacMahon, J.A 1980. Some aspects of succession in the spruce-fir forest zone of northern Utah. The Great Basin Naturalist 40:1-26. Schlichting, CD. 1986. The evolution of phenotypic plasticity in plants. Ann. Rev. Ecol. Syst. 17:667-693. Scott, M.L. and Murphy, P.G. 1987. Regeneration patterns of northern white cedar, an old-growth forest dominant. Am. Mid. Nat. 117:10-16. Sheppard, W. and Cannell, M.G.R. 1985. Nutrient use efficiency of clones of Picea sitchensis and Pinus contorta. Silvae Genetica 34:126-132. Shepperd, W.D. 1981. Variation in growth of Engelmann spruce seedlings under selected temperature environments. USDA For. Ser. Res. Note RM-404. Fort Collins, Colorado. 3 pp. Shirley, H.L. 1945. Reproduction of upland conifers in the Lake States as affected by root competition and light. Am. Mid. Nat. 33:537-612. Smith, J.H. and Clark, M.B. 1960. Growth and survival of Engelmann spruce and alpine fir on seed spots at Bolean Lake, B.C. 1954-59. For. Chron. 36:46-49. Smith, R.B. and Wass, E.F. 1985. Some Chemical and Physical Characteristics of Skidroads and Adjacent Undisturbed Soils. BC-X-261. Canadian Forestry Service. Pacific Forest Research Centre. Victoria, B.C. 28 pp. Smith, W.K., DeLucia, E.H. and Knapp, AK. 1987. Environmental limitations to photosynthesis in subalpine plants of the central Rocky Mountains, USA In: Management of Subalpine Forests: Building on 50 Years of Research. Proceedings of a Technical Conference. USDA For. Ser. Gen. Tech. Rep. RM-149. Fort Collins, Colorado, pp. 199-206. Stahelin, R. 1943. Factors influencing the natural restocking of high altitude burns by coniferous trees in the central Rocky Mountains. Ecology 24:19-30. Slathers, R.J. 1989. Summer frost in young forest plantations. FRDA Report 073. British Columbia Ministry of Forests. Victoria, B.C. 24 pp. 250 Stettler, R.F. 1958. Development of a residual stand of interior spruce - alpine fir during the first 28 years following cutting. Unpublished report to Nelson Forest Region. British Columbia Forest Service. Nelson, B.C. 65 pp. Stickney, P.F. 1986. Data base for early postfire succession on the Sundance Burn, northern Idaho. USDA For. Ser. Gen. Tech. Rep. LNT-189. Ogden, Utah. Strickler, G.S. and Edgerton, P.J. 1976. Emergent seedlings from coniferous litter and soil in eastern Oregon. Ecology 57:801-807. Tilman, D. 1988. Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton University Press. Princeton, New Jersey. 360 pp. Tinus, R.W. 1984. Optimum temperatures for growth of southern Rocky Mountain Engelmann spruce and Douglas-fir seedlings. USDA For.Ser.Res.Note RM-442. Fort Collins, Colorado. 5 pp. Titus, B.D. and Malcolm, D.C. 1987. The effect of fertilization on litter decompostion in clearfelled spruce stands. Plant and Soil 100:297-322. Tranquillini, W. 1979. Physiological Ecology of Alpine Timberline. Springer-Verlag. New York. 137 pp. Tranquillini, W. and Havranek, W.M. 1985. Influence of temperature on photosynthesis in spruce provenances from different altitudes. In: Establishment and Tending of Subalpine Forests: Research and Management. Proceedings of the 3rd IUFRO Workshop. Turner,H. and Tranquillini,W. (eds.). Report 270. Swiss Fed. Inst. For. Res. Birmensdorf, Switzerland, pp 41-51. Turner, H. and Tranquillini, W. (eds.). 1985. Establishment and Tending of Subalpine Forests: Research and Management. Proceedings of the 3rd IUFRO Workshop. Report 270. Swiss Fed. Inst. For. Res. Birmensdorf, Switzerland. 270 pp. USDA. 1965. Engelmann spruce (Picea engelmannii Parry). In: Silvics of Forest Trees of the United States. USDA For.Ser. Agricultural Handbook No. 271. Washington, D.C. pp. 299-304. Veblin, T.T. 1986. Treefalls and the coexistense of conifers in subalpine forests of the central Rockies. Ecology 67:644-649. Vitt, D.H, Marsh, J.E. and Bovey, R.B. 1988. Mosses, Lichens and Ferns of Northwest North America. Lone Pine Publ. Edmonton, Alberta. 296 pp. Vyse, A. 1981. Growth of young spruce plantations in interior British Columbia. For. Chron. 57:174-180. Wagner, R.G. 1989. Interspecific competition in young Douglas-fir plantations of the Oregon Coast Range. PhD thesis. Oregon State University. Corvallis, Oregon. 200 pp. Wagner, R.G., Petersen, T.D., Ross, D.W. and Radosevich, S.R. 1989. Competition thresholds for the survival and growth for ponderosa pine seedlings associated with woody and herbaceous vegetation. New Forests 3:151-170. 251 Wahlenberg, W.G. 1930. Effect of Ceanothus brush on western yellow pine plantations in the northern Rocky Mountains. J. Agric. Res. 41:601-612. Walstad, J.D. and Kuch, P.J. 1987. Forest Vegetation Management for Conifer Production. John Wiley & Sons. Toronto, Ont. 523 pp. Wardle, P. 1968. Engelmann spruce (Picea engelmannii Engel.) at its upper limits on the Front Range, Colorado. Ecology 49:483-495. Waring, S.A. and Bremner, J .M. 1964. Ammonium production in soil under waterlogged conditions as an index of nitrogen availability. Nature 201:951-952. Weetman, G. and Vyse, A . 1990. Natural regeneration. In: Regenerating British Columbia's Forests. Lavender, D.P., Parish, R., Johnson, C M . , Montgomery, G., Vyse, A. , Willis, R.A. and Winston, D. (eds.). University of British Columbia Press. Vancouver, B.C. pp. 118-130. West, D . C , Shugart, H.H. and Botkin, D.B. (eds.). 1981. Forest Succession: Concepts and Application. Springer-Verlag. New York. 517 pp. Whipple, S.A. and Dix, R.L. 1979. Age structure and successional dynamics of a Colorado subalpine forest. Am. Mid. Nat. 101:142-158. White, D.E. and Newton,M. 1989. Competitive interactions of whiteleaf manzanita, herbs, Douglas-fir, and ponderosa pine in southwest Oregon. Can. J. For. Res. 19:232-238. Wilkinson, L . 1990. SYSTAT: The System for Statistics. Version 5. SYSTAT Inc. Evanston, 111. Wilson, J.B. 1988a. Shoot competition and root competition. J. Appl. Ecol. 25:279-296. Wilson, J.B. 1988b. A review of evidence on the control of shoot:root ratio, in relation to models. Ann. Bot. 61:433-449. Yarie, J. 1978. The role of understory vegetation in the nutrient cycle of forested ecosystems in the Mountain Hemlock Biogeoclimatic Zone. PhD thesis. Faculty of Forestry, University of British Columbia. Vancouver, B.C. 176 pp. Yodzis, P. 1986. Competition, mortality, and community structure. In: Community Ecology. Diamond, J. and Case, T.J. (eds.). Harper & Row. New York. pp. 480-491. Zar, J.H. 1984. Biostatistical Analysis. Second edition. Prentice-Hall, Inc. Englewood Cliffs, N.J. 718 pp. Zutter, B.R., Glover, G.R. and Gjerstad, D-H. 1986. Effects of herbaceous weed control using herbicides on a young loblolly pine plantation. For. Sci. 32:882-899. APPENDIX 1 A LIST OF THE VASCULAR PLANTS ON TWO SIX-YEAR-OLD (IN 1989) CUTOVERS IN THE ESSFwc2 Species present in clipped, screefed and control plots in herb and shrub patches on two cutovers in the ESSFwc2 six years after logging and three growing seasons after treatment. The numbers represent the number of plots per treatment (total = 3) in which a species was present. Trees Picea engelmannii Parry2 Abies lasiocarpa (Hook.) Nutt.2 Shrubs Menziesia ferruginea Smith Rhododendron albiflorum Hook. Vaccinium membranaceum Dougl. V. ovalifolium Smith Ribes lacustre (Pers.) Poir. R. viscosissimum Pursh Rubus parviflorus Nutt. Sambucus racemosa L . Viburnum edule (Michx.) Raf. Lonicera involucrata (Rich.)Banks Oplopanax horridus (Smith) Miq. Salix sp. L . Sorbus sitchensis Roemer Herbs (including ferns and grasses) Valeriana sitchensis Bong. Thalictrum occidentale Gray Arnica latifolia Bong. Streptopus roseus Michx. Rubus pedatus J.E. Smith Herb patches Study Site 1 Study Site 2 Treatments1 Treatments :R C L SC C R C L SC 3 3 1 3 3 1 'I 3 2 3 J) 1 2 3 2 3 1 1 0 0 0 0 0 1 2 2 1 2 1 2 3 1 3 3 0 3 3 3 3 3 3 1 0 0 0 0 0 3 2 2 0 0 0 2 0 2 0 2- 1 0 0 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 3 3 3 3 3 3 2 2 1 3 1 2 2 3 0 3 3 3 3 3 3 3 3 3 3 3 3 3 3 1 Shrub patches Study Site 1 Study Site 2 Treatments Treatments ?R C L SC CR C L SC 2 1 1 3 2 2 3 2 J 3 . j 3 3 3 3 . 3 3 1 3 1 3 1 1 3 3 3 3 3 j -3 2 2 3 2 2 3 3 3 0 . 2 2 0 0 0 0 0 0 2 3 3 0 0 1 0 0 2 0 0 1 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 3 3 3 3 3 3 0 0 1 0 1 3 1 1 0 3 2 3 3 3 3 3 3 3 3 3 3 3 3 3 Herb patches Study Site 1 Study Site 2 Shrub patches Study Site 1 Study Site 2 Treatments1 Treatments Treatments Treatments CR CL SC CR CL SC CR CL SC CR CL SC Herbs (cont'd.) Tiarella unifoliata Hook. 3 3 3 3 3 3 3 3 3 3 3 3 Gymnocarpium dryopteris (L.) Newm. 3 3 3 3 3 3 3 3 3 3 3 3 Athyrium filix-femina (L.) Roth3 1 2 0 0 0 0 0 1 1 0 0 0 Dryopteris expansa (Presl) 0 0 0 T 1 0 1 1 0 0 0 0 Fraser-Jenkins & Jermy Epilobium angustifolium L. 2 2 3 3 3 3 1 1 3 0 1 3 E. glandulosum Lehm. 0 2 ' 2 - 2 3 3 0 0 1 0 0 3 Osmorhiza chilensis H. & A. 2 3 2 3 3 3 1 1 0 1 0 .3 Actaea rubra (Ait.) Willd. 2 2 2 3 2 1 0 1 0 0 0 1 Viola glabella Nutt. 1 3 1 3 2 3 0 1 0 0 0 1 Galium triflorum Michx. 2 3 1 0 0 1 1 0 1 0 0 0 Cornus canadensis L? 3 3 2 0 0 0 2 3 2 0 0 0 Mitella breweri Gray 1 2 0 3 3 3 6 1 0 2 1 3 Clintonia uniflora (Schult.) Kunth. 2 1 2 1 1 1 2 3 1 2 1 2 Lycopodium annotinum L. 0 0 0 1 0 0 1 0 0 1 0 0 Veratrum viride Ait. 0 1 1 0 1 0 1 1 1 0 0 0 Equisetum arvense L. 0 1 0 0 0 1 0 0 0 0 1 0 Lister a cordata (L.) R. Br. 2 0 0 1 1 0 2 0 1 1 1. 1 Senecio triangularis Hook. 0 0 0 1 0 0 0 0 0 0 0 0 Hieracium albiflorum Hook. 0 2 1 0 1 2 1 1 2 0 0 2 Streptopus amplexifolius (L.) DC. 3 1 1 1 0 0 2 3 0 1 0 0 Anaphalis margaritacea (L.) B.& H. 0 0 1 0 0 0 0 1 1 0 0 0 Goody era oblongifolia Raf. 0 1 0 0 0 0 0 0 0 0 0 0 Smilacina stellata (L.) Desf. 0 0 2 0 0 0 0 1 0 0 0 0 Taraxacum officinale Web. 0 2 0 0 1 1 1 0 0 0 0 0 Cirsium vulgare (Savi) Tenore 0 0 1 0 0 1 0 0 0 0 0 0 Luzula hitchcockii Hamet-Ahti 1 0 0 1 0 1 0 0 0 0 1 1 Calamagrostis canadensis 2 2 1 1 0 2 0 0 0 0 0 0 (Michx.) Beauv. Ranunculus uncinatus D. Don 0 0 0 1 1 2 0 0 0 0 0 0 The species listed below were also present on the study sites but did not occur in any of the plots in the disturbance experiment. Trees Tsuga heterophylla (Raf.) Sarg. (site 1 only) Shrubs Sorbus scopulina Greene (site 1 only) Rubus idaeus L. (site 1 only) Herbs Smilacina racemosa (L.)Dest (site 1 only) Mitella pentandra Hook. (site 2 only) Viola macloskeyi Lloyd (both sites) Melica smithii (Porter) Vasey (site 1 only) 1 CR = control, CL = clipped, SR = screefed. 2 Picea and Abies were present in most plots but were not removed during treatment, therefore their presence after three seasons did not necessarily represent reinvasion of plots. 3 These species were also found only on site 1 when the sites were surveyed three years after logging. Note: Moss cover in CL & SR plots was < 5% after 3 seasons, values in CR plots ranged from 1-70%. The most common species were Pleurozium schreberi (Brid.) Mitt., Ptilium crista-castrensis (Hedw.) De Not., Polytrichum juniperinum Hedw., Rhyttdiopsis robusta (Hedw.) Broth., and several Dicranum species including Dicranum fuscescens Turn, and D. scopariwn Hedw. References for species names: Angove & Bancroft, 1983; Hitchcock & Cronquist, 1973; Klinka et al, 1989; Vitt et al, 1988. cn tn APPENDIX 2 D E N D R O G R A M F R O M CLUSTER A N A L Y S I S OF SPECIES-COVER Q U A D R A T S O N TWO T H R E E - Y E A R - O L D (TN 1986) CUTOVERS IN T H E ESSFwc2 257 Information on species composition and cover was collected in 2 m z quadrats at 5 m intervals along 50 m transects on 3 three-year-old cutovers in the ESSFwc2. Multivariate cluster analysis was used to determine whether quadrats clustered together . on the basis of similarities in species composition and cover. The analysis used was part of the MIDAS statistical package (Fox and Guire, 1976). A minimum variance clustering algorithm was used as the presence of several classes of quadrats was anticipated based on preliminary observations (mimimum variance clustering is described by Pielou (1984) who suggests that this approach is useful for descriptive classification). In mimimum variance clustering the two clusters whose fusion yields the least increase in within-cluster dispersion are united at each step in the analysis. The within-cluster dispersion is the sum of the squares of the Euclidean distances between every point and the centroid of the cluster. The cophenetic correlation for the dendrogram (correlation between input dissimilarities and output dissimilarities implied by the dendrogram) = 0.5425. In the dendrogram the numbers along the vertical axis represent the quadrat codes. The lower case letters on the left indicate the positions in the dendrogram of 8 consecutive quadrats along a transect on study site 1. They illustrate the large variation in vegetation over short distances on the cutovers. The upper case letters on the right indicate seven major groups of quadrats. The positions of the horizontal lines separating the groups were chosen arbitrarily. Quadrats within each group shared the following similarities in species cover and composition: Group A - 20 to 40% cover of the herbs Valeriana sitchensis and Gymnocarpium dryopteris; Group B - 20 to 40% cover of any or all of the woody species Menziesia ferruginea, Rhododendron albiflorum or Abies lasiocarpa; Group C - low cover (< 20%) of any or all species; Groups D and E - 30 to 90% cover of residual trees of Abies lasiocarpa or Picea engelmannii and high (> 20%) cover of the shrubs Menziesia ferruginea or Rhododendron albiflorum; Group F - 20 to 50% cover of Thalictrum occidentale, Gymnocarpium dryopteris and Valeriana sitchensis; Group G - 50 to 80% cover of Menziesia ferruginea. Four patch types were identified on the cutovers on the basis of similarities in the life-forms of the major species in the groups described above. Herb patches were dominated by high covers of the herbs Valeriana sitchensis, Gymnocarpium dryopteris, and Thalictrum occidentale (groups A and F). Shrub patches were dominated by the ericaceous species Menziesia ferruginea and Rhododendron albiflorum (groups B and G). Skid trail patches were on or adajacent to skid trails and had high to low covers of various herb species, particularly Epilobium angustifolium (some quadrats in group C). Residual tree patches were dominated by advanced regeneration of Abies lasiocarpa and Picea engelmannii (groups D and E). Some quadrats in group C contained both herb and shrub species with similar cover. APPENDIX 3 STATISTICS TABLES FOR DATA PRESENTED IN CHAPTER 3 Table 3.1 Statistics on the height of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1. 260 a) Cold Stock Treatment1 Mean(cm) S.D. CV. n Tukey2 1 36.10 5.79 0.16 30 a 2 29.85 7.25 0.24 30 a 3 40.73 8.04 0.20 30 a 4 31.77 5.02 0.16 30 a 5 31.21 6.73 0.21 30 a 6 31.77 5.65 0.18 30 a 7 30.99 6.88 0.22 30 a b) Hot Stock Treatment Mean(cm) S.D. CV. n Tukey 1 35.10 7.52 0.21 30 ab 2 25.14 4.74 0.19 30 be 3 35.29 5.52 0.16 30 a 4 25.07 3.64 0.14 30 c 5 27.75 8.25 0.30 30 c 6 23.94 4.63 0.19 30 c 7 27.47 4.87 0.18 30 c S.D. = Standard deviation. CV. = Coefficient of variation, n = number of measurements per treatment. 1 Treatments: 1 = clipped herb; 2 = herb control; 3 = clipped shrub; 4 = shrub control; 5 = clipped skid trail; 6 = skid trail control; 7 = bare skid trail. 2 Treatment means differing significantly with p < 0.05 are represented by different letters. Tukey's HSD test based on PLOT MS from nested ANCOVA of height (Table 3.2). Table 3.2 Nested ANCOVA and orthogonal contrasts for spruce seedling height in seven treatments on site 1. a) Nested ANCOVA for seedling height (untransformed) in cold stock.1 SOURCE SS DF MS F-ratio Treatment Plot Covariate3 1895.00 1840.75 2498.42 6 14 1 315.83 131.48 2498.42 2.40 ns2 5.64 * 19.00 * Error 4384.05 188 23.31 Orthogonal contrast Test of hypothesis HO: (herb control + shrub control) -(clipped herb + clipped shrub) = 0. SOURCE SS DF MS F-ratio HO 1011.78 1 1011.78 7.69 * Error 1840.75 14 131.48 b) Nested ANCOVA for seedling height (untransformed) in hot stock. SOURCE SS DF MS F-ratio Treatment Plot Covariate 3868.89 1226.96 1782.70 6 14 1 644.81 87.64 1782.70 7.36* 5.21* 20.34* Error 3161.20 188 16.81 Orthogonal contrast Test of hypothesis HO: (herb control + shrub control) -(clipped herb + clipped shrub) = 0. SOURCE SS DF MS F-ratio HO 2828.01 1 2828.01 32.27* Error 1226.96 14 87.64 1 Numbers are rounded to two decimal places. 2 ns = nonsignificant with p > 0.05, * = significant with p < 0.05, critical F^ 6 1 4j = 2.85. 3 Covariate was height at time of planting. 262 Table 3.3 Statistics on the diameter of cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1. a) Cold Stock Treatment1 Mean(mm) S.D. C.V. n Tukey2 1 9.15 1.46 0.16 30 ab 2 5.75 1.22 0.21 30 d 3 10.45 1.31 0.12 30 a 4 6.31 1.32 0.21 30 cd 5 8.08 1.53 0.19 30 be 6 6.42 1.16 0.18 30 cd 7 7.86 1.28 0.16 30 be b) Hot Stock Treatment Mean(mm) S.D. C.V. n Tukey 1 8.01 1.51 0.19 30 a 2 4.25 0.70 0.17 30 c 3 8.62 1.38 0.16 30 a 4 4.23 0.85 0.20 30 c 5 6.51 1.71 0.26 30 b 6 4.44 0.66 0.15 30 c 7 6.32 1.15 0.18 30 b H.l). = Standard deviation. CJ.V. = (Joetticient of variation, n = number of measurements per treatment. 1 Treatments: 1 = clipped herb; 2 = herb control; 3 = clipped shrub; 4 = shrub control; 5 = clipped skid trail; 6 = skid trail control; 7 = bare skid trail. 2 Treatment means differing significantly with p < 0.05 are represented by different letters. Tukey's HSD test based on PLOT MS from nested ANCOVA of diameter (Table 3.4). 263 Table 3.4 Nested ANCOVA and orthogonal contrasts for spruce seedling diameter in seven treatments on site 1. a) Nested ANCOVA for seedling diameter (untransformed) in cold stock1. SOURCE SS DF MS F-ratio Treatment Plot Covariate3 509.06 62.02 31.57 6 14 1 84.84 4.43 31.57 19.15*2 3.11* 7.13* Error 267.57 188 1.42 Orthogonal contrasts Test of hypotheses HO^ (herb control + shrub control) - (clipped herb + clipped shrub) = 0. HO?: (clipped + bare treatments) - (control treatments) = 0. SOURCE SS DF MS F-ratio HO, H0 2 426.67 387.01 1 1 426.67 387.01 96.31* 87.36* Error 62.02 14 4.43 b) Nested ANCOVA for seedling diameter (untransformed) in hot stock. SOURCE SS DF MS F-ratio Treatment Plot Covariate 583.18 50.76 37.03 6 14 1 97.20 3.63 37.03 26.78* 3.63* 10.20* Error 187.287 188 1.00 Orthogonal contrasts Test of hypotheses HO!: HO?: (herb control + shrub control) - (clipped herb + cl (clipped + bare treatments) - (control treatments) lipped shrub) = 0. = 0. SOURCE SS DF MS F-ratio HO, H02 499.68 453.83 1 1 499.68 453.83 137.80* 125.16* Error 50.76 14 3.63 1 Numbers are rounded to two decimal places. 2 * = significant with p < 0.05, critical F ^ 6 = 2.85. 3 Covariate was seedling diameter at time of planting. 264 Table 3.5 S t a t i s t i c s o n t h e r e l a t i v e g r o w t h r a t e (RGR)1 i n s e e d l i n g d i a m e t e r o f c o l d - p l a n t e d a n d h o t - l i f t e d s p r u c e s e e d l i n g s a f t e r t h r e e g r o w i n g s e a s o n s i n s e v e n t r e a t m e n t s o n s i t e 1. a ) C o l d S t o c k T r e a t m e n t 2 M e a n S . D . C V . n ( l o g e ( m m V y r ) 1 0 . 3 0 9 0 . 1 9 6 0 . 6 3 4 3 0 2 0 . 0 4 4 0 . 2 0 9 4 . 7 5 0 3 0 3 0 . 4 2 5 0 . 1 0 1 0 . 2 3 8 3 0 4 0 . 1 0 2 0 . 2 2 0 2 . 1 5 7 3 0 5 0 . 2 6 0 0 . 1 5 8 0 . 6 0 8 3 0 6 0 . 1 2 5 0 . 1 6 8 1 . 3 4 4 3 0 7 0 . 2 3 9 0 . 1 2 7 0 . 5 3 1 3 0 b ) H o t S t o c k T r e a t m e n t M e a n S . D . C V . n aoge(mm)/yr) 1 0 . 4 0 5 0 . 1 3 7 0 . 3 3 8 3 0 2 0 . 0 7 3 0 . 2 0 6 2 . 8 2 2 3 0 3 0 . 4 6 0 0 . 1 0 7 0 . 2 3 3 3 0 4 0 . 0 6 7 0 . 2 2 1 3 . 2 9 8 3 0 5 0 . 2 9 5 0 . 1 9 8 0 . 6 7 1 3 0 6 0 . 1 9 1 0 . 1 5 9 0 . 8 3 2 3 0 7 0 . 2 9 3 0 . 1 4 9 0 . 5 0 8 3 0 S . D . = S t a n d a r d d e v i a t i o n . C V . = C o e f f i c i e n t o f v a r i a t i o n , n = n u m b e r o f m e a s u r e m e n t s p e r t r e a t m e n t . 1 C a l c u l a t e d a s R G R ( d i a m e t e r ) = l o g g ( d i a m e t e r a t time2 ) - l o g g ( d i a m e t e r a t t i m e j ^ ) t i m e - t i m e 2 1 w h e r e d i a m e t e r a t t i m e j = d i a m e t e r i n A u g u s t 1 9 8 8 d i a m e t e r a t t i m e ? = d i a m e t e r i n A u g u s t 1 9 8 9 t i m e 2 - timej = l y e a r . 2 T r e a t m e n t m e a n s d i f f e r i n g s i g n i f i c a n t l y w i t h p < 0 . 0 5 a r e r e p r e s e n t e d b y d i f f e r e n t l e t t e r s . T u k e y ' s H S D t e s t b a s e d o n P L O T M S f r o m n e s t e d A N C O V A o f h e i g h t ( T a b l e 3 . 2 ) . 265 Table 3.6 S t a t i s t i c s o n t h e n u m b e r o f l a t e r a l s o n c o l d - p l a n t e d a n d h o t - l i f t e d s p r u c e s e e d l i n g s a f t e r t h r e e g r o w i n g s e a s o n s o n s i t e 1. a ) C o l d S t o c k T r e a t m e n t 1 M e a n S . D . C V . n T u k e y 2 1 3 7 . 6 0 8 . 0 6 0 . 2 1 3 0 a 2 1 9 . 7 0 8 . 6 8 0 . 4 4 3 0 c 3 3 9 . 8 0 8 . 4 0 0 . 2 1 3 0 a 4 2 8 . 1 7 7 . 1 9 0 . 2 5 3 0 b 5 3 4 . 1 3 8 . 1 4 0 . 2 4 3 0 ab 6 2 3 . 8 0 6 . 6 2 0 . 2 8 3 0 c 7 3 2 . 3 7 8 . 9 3 0 . 2 8 3 0 ab b ) H o t S t o c k T r e a t m e n t M e a n S . D . C V . n T u k e y 1 3 1 . 5 7 9 . 1 1 0 . 2 9 3 0 ab 2 1 8 . 6 3 5 . 2 1 0 . 2 8 3 0 d 3 3 4 . 4 3 8 . 0 9 0 . 2 3 3 0 a 4 1 9 . 7 0 6 . 0 5 0 . 3 1 3 0 cd 5 2 8 . 6 3 8 . 3 2 0 . 2 9 3 0 ab 6 2 1 . 0 7 6 . 1 0 0 . 2 9 3 0 be 7 3 1 . 8 3 7 . 7 2 0 . 2 4 3 0 abc S . D . = S t a n d a r d d e v i a t i o n . C V . = C o e f f i c i e n t o f v a r i a t i o n , n = n u m b e r o f m e a s u r e m e n t s p e r t r e a t m e n t . 1 T r e a t m e n t s : 1 = c l i p p e d h e r b ; 2 = h e r b c o n t r o l ; 3 = c l i p p e d s h r u b ; 4 = s h r u b c o n t r o l ; 5 = c l i p p e d s k i d t r a i l ; 6 = s k i d t r a i l c o n t r o l ; 7 = b a r e s k i d t r a i l . 2 T u k e y ' s H S D t e s t b a s e d o n P L O T M S i n n e s t e d A N C O V A o f s q u a r e r o o t ( l a t e r a l n u m b e r ) ( T a b l e 3 . 7 ) . T r e a t m e n t m e a n s d i f f e r i n g s i g n i f i c a n t l y w i t h p < 0 . 0 5 a r e r e p r e s e n t e d b y d i f f e r e n t l e t t e r s . 266 Table 3.7 N e s t e d A N C O V A f o r n u m b e r o f l a t e r a l s o n s p r u c e s e e d l i n g s i n s e v e n t r e a t m e n t s o n s i t e 1. a ) N e s t e d A N C O V A f o r n u m b e r o f l a t e r a l s ( s q u a r e r o o t - t r a n s f o r m e d ) i n c o l d s t o c k 1 . S O U R C E S S D F M S F - r a t i o T r e a t m e n t 8 1 . 6 9 6 1 3 . 6 1 1 6 . 3 4 * 2 P l o t 1 1 . 6 6 1 4 0 . 8 3 1 . 6 8 n s C o v a r i a t e 3 7 . 3 1 1 7 . 3 1 8 . 7 8 * E r r o r 9 3 . 1 2 1 8 8 0 . 4 9 b ) N e s t e d A N C O V A f o r n u m b e r o f l a t e r a l s ( s q u a r e r o o t - t r a n s f o r m e d ) i n h o t s t o c k . S O U R C E S S D F M S F - r a t i o T r e a t m e n t 7 0 . 1 2 6 1 1 . 6 9 7 . 3 1 * P l o t 2 2 . 3 9 1 4 1 . 6 0 3 . 9 2 * C o v a r i a t e 6 . 9 7 1 6 . 9 7 4 . 3 6 * E r r o r 7 6 . 6 9 1 8 8 0 . 4 1 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . * = s i g n i f i c a n t w i t h p < 0 . 0 5 , c r i t i c a l F(6,i4) = 2 . 8 5 , n s = n o n s i g n i f i c a n t w i t h p > 0 . 0 5 . C o v a r i a t e w a s s q u a r e r o o t - t r a n s f o r m e d l a t e r a l n u m b e r a t t h e t i m e o f p l a n t i n g . 2 Table 3.8 Statistics on the length of laterals in cold-planted spruce seedlings after three growing seasons in seven treatments on sites 1 and 2. a) Cold Stock - site 1 Treatment1 Mean(cm) S.D. CV. n Tukey2 1 6.51 1.61 0.25 30 ab 2 4.06 1.29 0.32 30 b 3 8.11 2.14 0.26 30 a 4 4.08 1.19 0.29 30 b 5 4.34 1.55 0.36 30 b 6 4.32 1.31 0.30 30 b 7 4.59 1.57 0.34 30 b b) Cold Stock - site 2 Treatment Mean(cm) S.D. CV. n Tukey 1 5.34 1.85 0.35 30 ab 2 3.84 1.56 0.41 30 b 3 6.76 2.12 0.31 30 a 4 3.78 0.97 0.26 30 b 5 6.11 2.01 0.33 30 ab 6 4.32 1.65 0.38 30 ab 7 5.75 1.52 0.26 30 ab S.D. = Standard deviation. CV. = Coefficient of variation, n = number of measurements per treatment. 1 Treatments: 1 = clipped herb; 2 = herb control; 3 = clipped shrub; 4 = shrub control; 5 = clipped skid trail; 6 = skid trail control; 7 = bare skid trail. 2 Tukey's HSD test based on PLOT MS from nested ANCOVA of log-transformed (lateral length) (Table 3.9). Treatment means differing significantly with p < 0.05 are represented by different letters. 2 6 8 Table 3.9 Nested ANCOVA and orthogonal contrasts for length of laterals in cold-planted spruce seedlings in seven treatments on sites 1 and 2. a) Nested ANCOVA for length of laterals (loge-transformed) in cold stock on site 11. SOURCE SS DF MS F-ratio Treatment Plot Covariate3 13.61 4.79 0.05 6 14 1 2.27 0.34 0.05 6.63 * 2 3.98 * 0.15 ns Error 16.14 188 0.086 Orthogonal contrasts Test of hypotheses HO j: (herb control + shrub control) - (clipped herb + clipped shrub) = 0. HO?: (clipped + bare skid trail) - (clipped herb + clipped shrub) = 0. SOURCE SS DF MS F-ratio HOi H0 2 10.06 8.14 1 1 10.06 8.14 29.41 * 23.81* Error 4.79 14 0.34 b) Nested ANCOVA for length of laterals (logg-tranformed) in cold stock on site 2. SOURCE SS DF MS F-ratio Treatment Plot Covariate 10.12 4.78 0.05 6 14 1 1.69 0.34 0.05 4.95 * 2.92 * 0.14 ns Error 21.93 188 0.12 Orthogonal contrasts Test of hypotheses HOi: (herb control + shrub control) - (clipped herb + clipped shrub) = 0. HO?: (clipped + bare skid trail) - (clipped herb + clipped shrub) = 0. SOURCE SS DF MS F-ratio HOi H0 2 6.18 0.001 1 1 6.18 0.0010 18.12 * 0.003 ns Error 4.78 14 0.34 1 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . 2 * = s i g n i f i c a n t w i t h p < 0 . 0 5 , c r i t i c a l F ( 6 1 4 ) = 2 . 8 5 , n s = n o n s i g n i f i c a n t w i t h p > 0 . 0 5 . 3 C o v a r i a t e w a s l o g e - t r a n s f o r m e d l a t e r a l l e n g t h a t t h e t i m e o f p l a n t i n g . 269 Table 3.10 S t a t i s t i c s o n t h e l e n g t h o f n e e d l e s i n c o l d - p l a n t e d s p r u c e s e e d l i n g s a f t e r t h r e e g r o w i n g s e a s o n s i n s e v e n t r e a t m e n t s o n s i t e s 1 a n d 2 . a ) C o l d S t o c k - s i t e 1 T r e a t m e n t 1 Mean(mm) S . D . C V . n T u k e y 2 1 1 6 . 9 0 2 . 4 5 0 . 1 4 3 0 a 2 1 3 . 9 3 3 . 4 7 0 . 2 5 3 0 ab 3 1 5 . 5 6 2 . 8 1 0 . 1 8 3 0 a 4 1 1 . 1 6 2 . 1 3 0 . 1 9 3 0 b 5 1 4 . 6 8 3 . 6 6 0 . 2 5 3 0 a 6 1 3 . 4 4 2 . 2 2 0 . 1 6 3 0 ab 7 1 3 - 6 4 3 . 8 1 0 . 2 8 3 0 ab b ) C o l d S t o c k - s i t e 2 T r e a t m e n t Mean(mm) S . D . C V . n T u k e y 1 1 5 . 9 2 3 . 3 4 0 . 2 1 3 0 ab 2 1 3 . 0 9 3 . 1 2 0 . 1 6 3 0 ab 3 1 6 . 5 0 2 . 4 2 0 . 1 5 3 0 a 4 1 1 . 6 3 2 . 1 4 0 . 1 8 3 0 b 5 1 5 . 4 7 3 . 7 0 0 . 2 4 3 0 ab 6 1 3 . 2 3 3 . 6 3 0 . 2 7 3 0 ab 7 1 6 . 2 4 2 . 4 7 0 . 1 5 3 0 a S . D . = S t a n d a r d d e v i a t i o n . C V . = C o e f f i c i e n t o f v a r i a t i o n , n = n u m b e r o f m e a s u r e m e n t s p e r t r e a t m e n t . 1 T r e a t m e n t s : 1 = c l i p p e d h e r b ; 2 = h e r b c o n t r o l ; 3 = c l i p p e d s h r u b ; 4 = s h r u b c o n t r o l ; 5 = c l i p p e d s k i d t r a i l ; 6 = s k i d t r a i l c o n t r o l ; 7 = b a r e s k i d t r a i l . 2 T u k e y ' s H S D t e s t b a s e d o n P L O T M S f r o m n e s t e d A N C O V A o n l o g e - t r a n s f o r m e d ( n e e d l e l e n g t h ) f o r s i t e 1 a n d u n t r a n s f o r m e d ( n e e d l e l e n g t h ) f o r s i t e 2 ( T a b I e 3 . 1 1 ) T r e a t m e n t m e a n s d i f f e r i n g s i g n i f i c a n t l y w i t h p < 0 . 0 5 a r e r e p r e s e n t e d b y d i f f e r e n t l e t t e r s . 270 Table 3.11 N e s t e d A N C O V A a n d o r t h o g o n a l c o n t r a s t s f o r l e n g t h o f n e e d l e s i n c o l d -p l a n t e d s p r u c e s e e d l i n g s i n s e v e n t r e a t m e n t s o n s i t e s 1 a n d 2 . a ) N e s t e d A N C O V A f o r l e n g t h o f n e e d l e s ( l o g e - t r a n s f o r m e d ) i n c o l d s e e d l i n g s o n s i t e l 1 . S O U R C E S S D F M S F - r a t i o T r e a t m e n t P l o t C o v a r i a t e 3 3 . 3 8 1 . 5 3 0 . 5 6 6 1 4 1 0 . 5 6 5 . 1 7 * 2 0 . 1 1 3 . 9 6 * 0 . 5 6 5 . 1 2 * E r r o r 5 . 0 9 1 8 5 4 0 . 0 3 Orthogonal contrast T e s t o f h y p o t h e s i s H O : ( h e r b c o n t r o l + s h r u b c o n t r o l ) -( c l i p p e d h e r b + c l i p p e d s h r u b ) = 0 . S O U R C E S S D F M S F - r a t i o H O 2 . 4 6 1 2 . 4 6 2 2 . 6 0 * E r r o r 1 . 5 3 1 4 0 . 1 1 b ) N e s t e d A N C O V A f o r l e n g t h o f n e e d l e s ( u n t r a n s f o r m e d ) i n c o l d s t o c k o n s i t e 2 . S O U R C E S S D F M S F - r a t i o T r e a t m e n t P l o t C o v a r i a t e - * 6 1 5 . 6 4 3 3 6 . 0 8 6 3 . 0 9 6 1 4 1 1 0 2 . 6 1 4 . 2 7 * 2 4 . 0 1 3 . 4 4 * 6 3 . 0 9 2 . 6 3 n s E r r o r 1 2 8 9 . 5 1 1 8 5 6 . 9 7 Orthogonal contrast T e s t o f h y p o t h e s i s H O : ( h e r b c o n t r o l + s h r u b c o n t r o l ) -( c l i p p e d h e r b + c l i p p e d s h r u b ) = 0 . S O U R C E S S D F M S F - r a t i o H O 3 8 4 . 0 4 1 3 8 4 . 0 4 1 6 . 0 0 * E r r o r 3 3 6 . 0 8 1 4 2 4 . 0 1 1 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . 2 * = s i g n i f i c a n t w i t h p < 0 . 0 5 , c r i t i c a l F ^ 6 1 4 ) = 2 . 8 5 , n s = n o n s i g n i f i c a n t w i t h p > 0 . 0 5 . 3 C o v a r i a t e w a s l o g ~ - t r a n s f o r m e d n e e d l e l e n g t h a t t h e t i m e o f p l a n t i n g . * T h r e e c a s e s w e r e a e l e t e d d u e t o m i s s i n g 1 9 8 7 d a t a . ^ C o v a r i a t e w a s u n t r a n s f o r m e d n e e d l e l e n g t h a t t h e t i m e o f p l a n t i n g . 271 Table 3.12 S t a t i s t i c s o n t o t a l , s h o o t , a n d r o o t b i o m a s s o f c o l d - p l a n t e d a n d h o t - l i f t e d s p r u c e s e e d l i n g s a f t e r t h r e e g r o w i n g s e a s o n s i n s e v e n t r e a t m e n t s o n s i t e 1. a) C o l d S t o c k - s i t e 1 T o t a l b i o m a s s S h o o t b i o m a s s R o o t b i o m a s s n T r e a t m e n t 1 M e a n Q j ) S . D . T u k e y 2 M e a n ( g ) S . D . T u k e y M e a n ( g ) S . D . T u k e y 1 2 3 . 5 6 5 . 7 7 ab 1 9 . 3 0 4 . 6 0 ab 4 . 2 6 1 . 2 5 abc 9 2 9 . 0 7 4 . 0 9 d 6 . 8 4 3 . 1 3 d 2 . 2 3 1.0.1 be 9 3 3 9 . 9 1 1 4 . 6 7 a 3 4 . 6 6 1 3 . 2 2 a 5 . 2 5 2 . 0 0 a 9 4 1 3 . 4 8 4 . 6 1 bed 9 . 7 3 3 . 3 7 bed 3 . 7 4 1 . 3 4 abc 9 5 1 9 . 7 6 7 . 2 5 abc 1 5 . 9 4 6 . 4 1 abc 3 . 8 2 1 . 1 7 abc 9 6 9 . 9 0 2 . 6 4 cd 7 . 6 8 2 . 1 4 cd 2 . 2 2 0 . 5 7 c 9 7 2 3 . 1 4 8 . 6 1 ab 1 7 . 9 5 7 . 3 0 abc 4 . 9 6 2 . 1 2 a 9 b ) H o t S t o c k - s i t e 1. T o t a l S h o o t R o o t n T r e a t m e n t M e a n ( g ) S . D . T u k e y M e a n ( g ) S . D . T u k e y M e a n ( g ) S . D . T u k e y 1 1 9 . 5 4 7 . 5 1 ab 1 6 . 2 1 6 . 1 6 ab 3 . 3 3 1 . 9 2 ab 9 . 2 4 . 7 9 1 . 1 5 d 3 . 5 6 0 . 8 9 cd 1 . 2 3 0 . 4 3 cd 9 3 2 1 . 6 7 7 . 0 2 a 1 7 . 6 7 5 . 7 6 a 4 . 0 0 1 . 3 6 a 9 4 3 . 4 8 1 . 0 9 d 2 . 8 2 0 . 8 5 d 0 . 7 6 0 . 2 9 d 93 5 9 . 8 8 5 . 3 2 bed 7 . 1 8 4 . 7 4 bed 1 . 9 8 0 . 8 8 abc 9 3 6 7 . 3 8 5 . 1 4 cd 5 . 6 1 4 . 1 8 cd 1 . 7 7 0 . 9 8 be 9 7 1 1 . 6 8 3 . 3 5 abc 8 . 9 6 3 . 0 3 abc 2 . 7 2 0 . 7 8 ab 9 S . D . = S t a n d a r d d e v i a t i o n , n = n u m b e r o f m e a s u r e m e n t s p e r t r e a t m e n t . 1 T r e a t m e n t s : 1 = c l i p p e d h e r b ; 2 = h e r b c o n t r o l ; 3 = c l i p p e d s h r u b ; 4 = s h r u b c o n t r o l ; 5 - c l i p p e d s k i d t r a i l ; 6 = s k i d t r a i l c o n t r o l ; 7 = b a r e s k i d t r a i l . 2 T u k e y ' s H S D t e s t b a s e d o n P L O T M S f r o m n e s t e d A N O V A o n l o g e - t r a n s f o r m e d s h o o t , r o o t , a n d t o t a l b i o m a s s ( T a b l e 3 . 1 3 ) . T r e a t m e n t m e a n s d i f f e r i n g s i g n i f i c a n t l y w i t h p < 0 . 0 5 a r e r e p r e s e n t e d b y d i f f e r e n t l e t t e r s . 3 n = 8 f o r r o o t b i o m a s s . 272 Table 3.13 Nested ANOVA and orthogonal contrasts for shoot and root biomass in cold-planted spruce seedlings in seven treatments on site 1. a) Nested ANOVA for shoot biomass (logg-transformed) in cold seedlings1. SOURCE SS DF MS F-ratio Treatment 18.49 6 3.08 10.48 *2 Plot 4.12 14 0.29 2.42 ns Error 4.99 41 0.12 Orthogonal contrasts Test of hypotheses HO^ (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H02: (clipped + bare treatments) - (control treatments) = 0. SOURCE SS DF MS F-ratio HOi H02 12.90 14.46 1 1 12.90 14.46 43.87 * 49.19 * Error 4.12 14 0.29 b) Nested ANOVA for root biomass (logg-transformed) in cold seedlings. SOURCE SS DF MS F-ratio Treatment 6.74 6 1.12 6.24* Plot 2.51 14 0.18 1.34 ns Error 5.61 42 0.13 Orthogonal contrasts Test of hypotheses HO^ (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H02: (clipped + bare treatments) - (control treatments) = 0. SOURCE SS DF MS F-ratio HO-! H02 2.45 4.68 1 1 2.45 4.68 13.63 * 26.00 * Error 2.51 14 0.18 1 Numbers are rounded to two decimal places. 2 * = significant with p < 0.05, critical F(6>i4) = 2.85, ns = nonsignificant with p > 0.05. 273 Table 3.14 R a t i o o f s e e d l i n g b i o m a s s i n c l i p p e d t r e a t m e n t t o s e e d l i n g b i o m a s s i n c o n t r o l t r e a t m e n t s f o r c o l d a n d h o t s t o c k w i t h i n e a c h p a t c h t y p e o n s i t e s 1 a n d 2 1 . a ) S h o o t b i o m a s s C o l d S t o c k H o t S t o c k P a t c h t y p e S i t e 1 S i t e 2 S i t e 1 S i t e 2 H e r b ( 1 : 2 ) 2 2 . 8 2 2 . 1 7 4 . 5 5 3 . 0 5 S h r u b ( 3 : 4 ) 3 . 5 6 4 . 0 6 6 . 2 7 3 . 0 1 S k i d t r a i l ( 5 : 6 ) 2 . 0 7 2 . 7 3 1 . 2 8 3 . 1 9 b ) R o o t b i o m a s s C o l d S t o c k H o t S t o c k P a t c h t y p e S i t e 1 S i t e 2 S i t e 1 S i t e 2 H e r b ( 1 : 2 ) 1 . 9 1 1 . 4 4 2 . 7 1 2 . 9 3 S h r u b ( 3 : 4 ) 1 . 4 0 2 . 9 5 5 . 2 4 2 . 5 1 S k i d t r a i l ( 5 : 6 ) 1 . 7 2 1 . 8 3 1 . 1 2 2 . 4 1 B a s e d o n t r e a t m e n t m e a n s . N u m b e r s i n p a r e n t h e s e s c o r r e s p o n d t o t r e a t m e n t c o d e s i n f i g u r e s a n d t a b l e s . T r e a t m e n t s : 1 = c l i p p e d h e r b ; 2 = h e r b c o n t r o l ; 3 = c l i p p e d s h r u b ; 4 = s h r u b c o n t r o l ; 5 = c l i p p e d s k i d t r a i l ; 6 = s k i d t r a i l c o n t r o l ; 7 = b a r e s k i d t r a i l . 274 Table 3.15 Statistics on shoot:root biomass ratios in cold-planted and hot-lifted spruce seedlings after three growing seasons in seven treatments on site 1. a) Cold Stock Treatment1 Mean S.D. CV. n Tukey2 1 4.68 0.91 0.19 9 ab 2 3.06 0.54 0.18 9 be 3 6.09 1.37 0.23 8 a 4 2.66 0.53 0.20 9 c 5 4.20 1.31 0.31 9 abc 6 3.50 0.53 0.15 9 be 7 3.60 1.32 0.37 8 b b) Hot Stock Treatment Mean S.D. CV. n Tukey 1 5.41 2.02 0.37 9 a3 2 3.12 1.01 0.32 9 a? 3 4.50 0.65 0.14 9 a 4 4.03 1.83 0.45 8 a 5 3.83 1.24 0.32 8 a 6 3.09 0.65 0.21 9 a 7 3.45 1.17 0.34 9 a S.D. = Standard deviation. CV. = Coefficient of variation, n = number of measurements per treatment. 1 Treatments: 1 = clipped herb; 2 = herb control; 3 = clipped shrub; 4 = shrub control; 5 = clipped skid trail; 6 = skid trail control; 7 = bare skid trail. 2 Tukey's HSD test based on PLOT MS in nested ANOVA on loge-transformed (shoot: root biomass ratios) different letters. 3 Pairwise comparison probability for treatments 1 and 2 was 0.055. 275 Table 3.16 Nested ANOVA and orthogonal contrasts for shoot: root biomass ratios in seven treatments on site 1. a) Nested ANOVA for shoot: root biomass ratios (loge-transformed) in cold stock1. SOURCE SS DF MS F-ratio Treatment 3.98 Plot 1.02 6 14 0.66 0.07 9.09 * 2 1.60 ns Error 1.82 40 0.04 Orthogonal contrasts Test of hypotheses: HO^ (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H0 2: (clipped + bare treatment) - (control treatments) = 0. SOURCE SS DF MS F-ratio HOx 3.48 H0 2 2.18 1 1 3.48 2.18 47.70* 29.92 * Error 1.02 14 0.07 b) Nested ANOVA for shoot: root biomass ratios (logg-transformed) in hot stock. SOURCE SS DF MS F-ratio Treatment 2.15 Plot 1.59 6 14 0.36 0.11 3.15* 1.33 ns Error 3.43 40 0.09 Orthogonal contrasts Test of hypotheses: HO^ (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H02: (clipped + bare treatment) - (control treatments) = 0. SOURCE SS DF MS F-ratio HOI 1.07 H02 0.71 1 1 1.07 0.71 9.43* 6.23* Error 1.59 14 0.11 1 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . 2 * = s i g n i f i c a n t w i t h p < 0 . 0 5 , c r i t i c a l F ( 6 1 4 ) = 2 . 8 5 , n s = n o n s i g n i f i c a n t w i t h p > 0 . 0 5 . Table 3.17 276 ANOVA and orthogonal contrasts for soil temperatures in seven treatments on site 1. a) ANOVA for soil temperatures at 10 cm (untransformed) on June 6, 1989 on site l 1 . SOURCE SS DF MS F-ratio Treatment 86.03 6 14.34 3.18 * 2 Error 112.69 25 4.51 Orthogonal contrast Test of hypoth esis: HO: (herb control + shrub control) -(clipped herb + clipped shrub) = 0. SOURCE SS DF MS F-ratio HO 45.22 1 45.22 10.03* Error 112.69 25 4.51 b) ANOVA for soil temperatures at 10 cm (untransformed) on July 4, 1989 on site 1. SOURCE SS DF MS F-ratio Treatment 76.39 6 12.73 3.18 * Error 92.00 23 4.00 Orthogonal contrast Test of hypothesis: HO: (clipped + bare treatments) -(control treatments) = 0. SOURCE SS DF MS F-ratio HO 66.83 1 66.83 16.71 * Error 92.00 23 4.00 Table 3.17 Continued c) ANOVA for soil temperatures at 10 cm (untransformed) on August 6,1989 on site 1. SOURCE SS DF MS F-ratio Treatment 41.89 6 6.98 7.43 * Error 25.37 27 0.94 Orthogonal contrast Test of hypothesis: HO: (clipped + bare treatment) -(control treatments) = 0. SOURCE SS DF MS F-ratio HO 37.29 1 37.29 39.68 * Error 25.37 27 0.94 1 Numbers rounded to two decimals. 2 * = significant with p < 0.05. 278 Table 3.18 Nested ANOVA and orthogonal contrasts for soil moisture contents in seven treatments on sites 1 and 2. a) Nested ANOVA for soil moisture content (loge-transformed) on site l 1 . SOURCE SS DF MS F-ratio Treatment Plot 4.35 1.85 6 14 0.72 0.13 5.48 * 2 2.78 ns Error 2.89 61 0.05 Orthogonal contrasts Test of hypotheses: HOi: H02: H0 3: (herb control + shrub control) - (clipped herb + clipped shrub) = 0. (herb control + clipped herb) - (shrub control + clipped shrub) = 0. (skid trail treatments) - (off-skid trail treatments) = 0. SOURCE SS DF MS F-ratio HO, H0 2 H0 3 0.008 0.67 1.35 1 1 1 0.008 0.67 1.35 0.06 ns 5.08* 10.22 * Error 1.85 14 0.13 b) Nested ANOVA for soil moisture content (logg-transformed) on site 2. SOURCE SS DF MS F-ratio Treatment 3.13 6 0.52 2.02 ns Plot 3.35 13 0.26 1.68 ns Error 8.92 58 0.15 Orthogonal contrasts Test of hypotheses: HOj: (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H0 2: (herb control + clipped herb) - (shrub control + clipped shrub) = 0. H0 3: (skid trail treatments) - (off-skid trail treatments) = 0. SOURCE SS DF MS F-ratio HO, H0 2 H0 3 0.86 1.61 0.29 1 1 1 0.86 1.61 0.29 3.32 ns 6.23* 1.11ns Error 1 M u m V \ £ » r c r m i n ^ o H 1r\ 3.35 f>i rr-\ nAPimol r \ l i f OP 13 0.26 2 * -= significant with p < 0.05, ns = nonsignificant with p > 0.05. 279 Table 3.19 Orthogonal contrasts for nitrate and ammonium ion concentrations in resin bags in seven treatments on site 1. a) Test of hypotheses: HOj: (herb control + shrub control) - (clipped herb + clipped shrub) = 0. H0 2 : (herb control + clipped herb) - (shrub control + clipped shrub) = 0. H0 3 : (clipped skid trail) - (all other treatments) = 0. for nitrate ion concentrations (Iogg-transformed)1. SOURCE 2 SS DF MS F-ratio HOi 4.15 1 4.15 2.54 ns3 H 0 2 12.00 1 12.00 7.33 * H 0 3 3.56 1 3.56 2.18 ns Error 62.20 38 1.64 b) Test of hypothesis: HO: (clipped shrub) 1 - (all other treatments except clipped skid trail) = 0. for ammonium ion concentrations (logg-transformed). SOURCE 4 SS DF MS F-ratio HO 11.16 1 11.16 7.03* Error 50.84 32 1.59 Numbers rounded to two decimal places. ANOVA on logg-transformed values for all treatments yielded a nonsignificant (p = 0.096) F-value. ns = nonsignificant with p > 0.05, * = significant with p < 0.05. ANOVA on logg-transforraed values for all treatments except 5 (where n=l), yielded a nonsignificant (p = 0.091) F-value. 280 Table 3.20 ANOVA and orthogonal contrasts for weight loss in mixed herb and mixed shrub litter bags in seven treatments on site 1. a) ANOVA on weight loss (squared) in mixed herb litter bags. SOURCE SS DF MS F-ratio Treatment 3.76E + 08 2 Error 3.84E + 08 37 1038284.97 ' Orthogonal contrast Test of hypoth esis: HO: (bare skid trail) -(clipped + control treatments) = 0. SOURCE SS DF MS F-ratio HO 3.46E + 08 1 3.46E + 08 33.29 * Error 3.84E + 08 37 1038284.97 b) ANOVA on weight loss (square root-transformed) in mixed shrub litter bags. SOURCE SS DF MS F-ratio Treatment 5.07 6 0.84 2.04 ns Error 15.34 37 0.41 Orthogonal contrast Test of hypoth esis: HO: (bare skid trail)-(clipped + control treatments) = 0. SOURCE SS DF MS F-ratio HO 4.08 1 4.08 9.83 * Error 15.34 37 0.41 1 Numbers rounded to two decimal places. Herb litter came from Valeriana sitchensis (Sitka valerian), Thalictrum occidentale (western meadow-rue, Gymnocarpiun dryopteris (oak fern), and Epilobium angustifolium (fireweed). 2 * = significant with p < 0.05, ns = nonsignificant with p > 0.05. Tukey's test for HSD indicated that treatment 7 was significantly different (with p < 0.05) than each of the other treatments - no other treatment pairs differed significantly. 3 Shrub litter came from Menziesia ferruginea (false azalea) and Rhododendron albiflorum (white-flowered rhododendron). 281 Table 3.21 Statistics on weight loss in herb and shrub litter bags in control treatments on site 1. Litter/Patch Type Mean S.D. CV. n Tukey2 (% weight loss) 1. Herb litter/ 75.95 6.10 0.08 9 a herb control 2. Herb litter/ 72.10 7.37 0.10 8 a shrub control 3. Shrub litter/ 46.74 7.06 0.15 9 b herb control 4. Shrub litter/ 42.89 2.25 0.05 9 b shrub control 1 H e r b l i t t e r c a m e f r o m Valeriana sitchensis ( S i t k a v a l e r i a n ) . S h r u b l i t t e r c a m e f r o m Menziesia ferruginea ( f a l s e a z a l e a ) . 2 T u k e y ' s H S D t e s t b a s e d o n P L O T M S ( 0 . 2 0 8 ( 8 D F ) ) f r o m n e s t e d A N O V A o f s q u a r e r o o t -t r a n s f o r m e d w e i g h t l o s s v a l u e s . M e a n s d i f f e r i n g s i g n i f i c a n t l y w i t h p < 0 . 0 5 a r e r e p r e s e n t e d b y d i f f e r e n t l e t t e r s . APPENDIX 4 STATISTICS T A B L E S FOR D A T A PRESENTED IN CHAPTER 4 Table 4.1 ANOVAs of height and diameter of young planted and naturally established spruce seedlings on a cutover and in an old-growth forest. a) ANOVA for seedling height (untransformed) SOURCE SS DF MS F-ratio Groups 6819.73 4 1704.93 34.91* Error 3418.68 70 48.84 b) ANOVA for seedling diameter (square-root transformed) SOURCE SS DF MS F-ratio Groups 20.66 4 5.16 52.43* Error 6.89 70 0.10 Numbers are rounded to two decimal places. * = significant with p < 0.05. 284 Table 4.2 Analysis of relative production rates (RPR) in young planted and naturally established spruce seedlings on a cutover and in an old-growth forest. a) Statistics on RPR Group Mean (loggCm/yr) S.D. CV. DF Forest naturals 0.003 0.164 55.33 14 Shaded naturals 0.108 0.217 2.01 14 Open naturals 0.011 0.177 15.69 14 Shaded planted -0.144 0.335 2.33 14 Open planted 0.243 0.287 1.18 14 b) ANOVA for RPR (untransformed)2 SOURCE SS DF MS F-ratio Groups 1.23 4 0.31 5.11* Error 4.20 70 0.06 S.D. = Standard deviation. CV. = Coefficient of variation. 1 RPR = loggOeader^  - loge(leader1) where leader2 = leader length in 1990 time2 - timej leade^  = leader length in 1989 time2 - time1 = 1 year. 2 Numbers are rounded to two decimal places. 285 Table 4.3 Component loadings from a principal components analysis of the correlation matrix of nine variables representing the size characteristics of six-year-old Engelmann spruce seedlings. Two factors had eigenvalues greater than 1.0. Variable Factor i n Diameter 0.936 -0.224 Leader length 0.927 0.188 Lateral number 0.892 -0.230 Lateral length 0.892 0.061 Height 0.883 -0.008 Needle number -0.882 -0.148 Bud number 0.871 0.023 Height:diameter -0.642 0.483 Needle length 0.452 0.782 Percent of total variance 69.56 11.23 Table 4.4 ANOVAs of shoot and root biomass of young planted and naturally established spruce seedlings on a cutover and in an old-growth forest. 286 a) ANOVA on shoot biomass (loge-transformed) SOURCE SS DF MS F-ratio Groups 64.03 4 16.01 43.86* Error 18.98 52 0.36 b) ANOVA on root biomass (loge-transformed) SOURCE SS DF MS F-ratio Groups 49.84 4 12.46 41.01* Error 15.80 52 0.30 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . * = s i g n i f i c a n t w i t h p < 0 . 0 5 . 2 8 7 Table 4.5 ANOVAs of biomass ratios in young planted and naturally established spruce seedlings on a cutover and in a mature forest. a) ANOVA of ratio of shoot:root biomass (untransformed) SOURCE SS DF MS F-ratio Groups 38.82 4 9.71 5.90* Error 85.49 52 1.64 b) ANOVA of ratio of needle:stem biomass (untransformed) SOURCE SS DF MS F-ratio Groups 1.70 4 0.42 6.22* Error 3.55 52 0.07 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . * = s i g n i f i c a n t w i t h p < 0 . 0 5 . Table 4.6 Pearson correlation matrix of variables characterizing the growth and environment of Engelmann spruce seedlings. Values are correlation coefficients.1 288 Light Soil temp. Litter depth Moss cover Shrub cover Diameter 0.582* 0-087 0.155 -0.392 -0.343 Height 0.423* 0.039 0.239 -0.498* -0.216 Ht:dm -0.561* -0.165 -0.079 0.168 0.357 Lateral number 0.612* 0.146 0.070 -0.276 -0.262 Lateral length 0.471* 0.207 0.087 -0.402* -0.258 Needle number -0.551* -0.129 -0.275 0.448* 0.393 Needle length 0.355 0.309 -0.121 -0.047 -0.248 Leader length 0.612* 0.300 0.042 -0.365 -3.348 Total biomass 0.478* 0.035 0.097 -0.415 -0.295 Root biomass 0.444 -0.037 0.118 -0.396 -0.312 1 Significance based on Bonferroni-adjusted probabilities: * = significant with p<0.05, n=75 for size variables, n=57 for biomass variables. Table 4.7 Mean and standard deviations for characteristics describing the environments of five groups of Engelmann spruce seedlings. 289 Forest Cutover Characteristic Naturals Mean SD Shaded natural Mean SD Open natural Mean SD Shaded planted Mean SD Open planted Mean SD Moss cover (%) 73.20 27.23 38.53 34.96 51.67 33.79 19.27 32.14 20.60 27.14 Litter depth (cm) 0.10 0.20 2.07 1.97 0.77 1.13 3.50 2.31 1.77 1.05 Woody Utter (%) 0.13 035 0.67 0.49 0.60 0.51 0.67 0.49 8.33 20.41 Herb cover (%) 19.33 26.05 6033 29.19 58.4 41.42 96.67 46.88 23.60 29.02 Light level (% of full) 14.75 34.61 26.15 17.56 73.80 34.12 11.97 14.97 100.00 0 Soil temperature 10 cm (1°C) 7.92 1.53 10.13 1.88 11.27 2.79 7.00 1.00 8.27 2.05 Coniferous litter (%) 0.67 0.49 0.47 0.52 0.07 0.26 0.40 031 0.80 0.41 Moss litter (%) 0.13 035 0.73 0.46 0.73 0.46 0.07 0.26 0 0 Decayed wood (%) 17.07 21.72 17.33 23.13 15.6 18.74 7.00 13.86 833 20.41 Deciduous litter (%) 0.20 0.41 0.73 0.46 0.47 0.52 0.73 0.46 0.73 0.46 Shrub cover (%) 25.33 23.44 44.67 34.66 1333 15.12 32.07 4637 2.47 5.15 290 Table 4.8 Principal components analysis of the correlation matrix of twelve variables representing the characteristics of the environments of planted and naturally established Engelmann spruce seedlings. Five factors had eigenvalues greater than 1.0. Variable Factor i n ni rv v Moss cover -0.72 -0.04 0.28 0.14 -0.36 Litter depth 0.70 -0.37 0.27 -0.01 0.09 Seedling diameter 0.63 0.54 -0.06 0.32 -0.04 Woody litter 0.62 0.25 0.05 -0.24 0.33 Herb cover 0.56 -0.27 0.56 -0.22 -0.12 Light level 0.23 0.80 -0.23 0.28 -0.08 Soil temperature -0.31 0.66 0.26 0.11 0.35 Coniferous litter 0.03 -0.25 -0.74 0.25 0.13 Moss litter -0.37 0.25 0.68 0.11 0.30 Decayed wood cover -0.02 0.19 -0.23 -0.79 0.29 Deciduous litter 0.49 -0.27 0.18 0.55 0.09 Shrub cover -0.34 -0.43 -0.07 0.31 0.70 Percent of total variance 22.87 17.21 14.16 11.83 8.99 APPENDIX 5 STATISTICS TABLES FOR DATA PRESENTED IN CHAPTER 5 292 Table 5.1 Two-way ANCOVA for total number of species three growing seasons after disturbance in herb and shrub patches on sites 1 and 2. a) Site 1 SOURCE1. SS DF MS F-ratio Type 0.13 1 0.13 0.02 ns2 Treat 10.84 2 5.42 1.00 ns Type * Treat 2.73 2 1.37 0.25 ns Covariate3 101.65 1 101.65 18.73* Error 59.69 11 5.43 b) Site 2 SOURCE SS DF MS F-ratio Type 1.81 1 1.81 0.42 ns Treat 30.99 2 15.49 3.58 ns Type * Treat 20.14 2 10.07 2.32 ns Covariate 27.02 1 27.02 6.24* Error 47.65 11 4.33 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . T y p e = H e r b , S h r u b , T r e a t = C o n t r o l , C l i p p e d , S c r e e f e d . n s = n o n s i g n i f i c a n t w i t h p > 0 . 0 5 , * = s i g n i f i c a n t w i t h p < 0 . 0 5 . C o v a r i a t e w a s n u m b e r o f s p e c i e s b e f o r e t r e a t m e n t . 293 Table 5.2 Statistics on total percent cover (summed by species) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2. a) Site 1 Treatment Mean (%) S.D. C V . n Tukey1 Herb control 128.67 28.59 0.22 3 ab Herb clip 153.67 13.20 0.09 3 a Herb screef 76.67 43.94 0.57 3 be Shrub control 152.33 22.37 0.15 3 a Shrub clip 51.00 7.00 0.14 3 c Shrub screef 28.67 6.66 0.23 3 c b) Site 2 Treatment Mean (%) S.D. C V . n Tukey Herb control 151.67 22.37 0.15 3 a Herb clip 152.67 23.69 0.15 3 a Herb screef 62.00 17.44 0.28 3 b Shrub control 161.67 17.01 0.10 3 a Shrub clip 41.00 25.16 0.61 3 b Shrub screef 27.00 2.00 0.07 3 b 1 Treatment means differing significantly with p < 0.05 are represented by different letters. S.D. = Standard deviation. CV. = Coefficient of variation, n = number of plots. Numbers are rounded to two decimal places. 294 Table 5.3 Two-way ANOVA for total percent cover (summed by species) three seasons after disturbance in herb and shrub patches on sites 1 and 2. a) Site 1 SOURCE1 SS DF MS F-ratio Type 8064.50 1 8064.50 13.76 * 2 Treat 23276.33 2 11638.17 19.86* Type* Treat 12042.33 2 6021.17 10.28* Error 7031.33 12 585.94 b) Site 2 SOURCE SS DF MS F-ratio Type 9338.89 1 9338.89 24.45* Treat 37800.33 2 18900.18 49.48* Type* Treat 11352.78 2 5676.39 14.86* Error 4584.00 12 382.00 1 N u m b e r s a r e r o u n d e d t o t w o d e c i m a l p l a c e s . U n t r a n s f o r m e d t o t a l % c o v e r v a l u e s w e r e u s e d i n t h e a n a l y s i s . T y p e = H e r b , S h r u b , T r e a t = C o n t r o l , Q i p p e d , S c r e e f e d . 2 * = s i g n i f i c a n t w i t h p < 0 . 0 5 . Table 5.4 Statistics on changes in herb cover three growing seasons after disturbance in herb and shrub patches on sites 1 and 21. a) Site 1 Treatment Mean(%) S.D. CV. n Tukey2 Herb control 159.43 46.35 0.29 3 ab Herb clip 94.27 6.90 0.07 3 abc Herb screef 54.30 37.97 0.70 3 be Shrub control 195.03 154.50 0.79 3 a Shrub clip 73.83 5.86 0.08 3 abc Shrub screef 35.10 13.38 0.38 3 c b) Site 2 Treatment Mean (%) S.D. CV. n Tukey Herb control 122.70 7.43 0.06 3 ab Herb clip 93.93 22.18 0.24 3 be Herb screef 42.00 24.59 0.58 3 cd Shrub control 280.77 39.45 0.14 3 a Shrub clip 22.50 13.23 0.59 3 d Shrub screef 18.30 3.05 0.17 3 d 1 Numbers are rounded to two decimal places. Changes in herb cover calculated as: ([final cover / initial cover] x 100). Cover was summed by species. 2 Treatment means differing significantly with p < 0.05 are represented by different letters. S.D. = Standard deviation. CV. = Coefficient of variation, n = number of plots. 296 Table 5.5 Two-way ANOVA for logg-transformed changes in herb cover ([final cover/initial cover] x 100) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2. a) Site 1 SOURCE1 SS DF MS F-ratio Type 0.15 1 0.15 0.76 ns2 Treat 5.76 2 2.88 14.75* Type* Treat 0.11 2 0.06 0.29 ns Error 2.34 12 0.19 b) Site 2 SOURCE SS DF MS F-ratio Type 1.04 1 1.04 7.73* Treat 12.42 2 6.21 46.16* Type* Treat 4.30 2 2.15 15.99* Error 1.61 12 0.13 1 Numbers are rounded to two decimal places. Type = Herb, Shrub, Treat = Control, Clipped, Screefed. 2 ns = nonsignificant with p > 0.05, * = significant with p < 0.05. 297 Table 5.6 Statistics on changes in shrub cover three growing seasons after disturbance in herb and shrub patches on sites 1 and 21. a) Site 1 Treatment Mean (%) S.D. CV. n Tukey2 Herb control 492.83 72.12 0.15 3 a Herb clip 83.50 47.85 0.57 3 be Herb screef 44.57 41.29 0.93 3 bed Shrub control 107.13 10.56 0.10 3 b Shrub clip 21.43 11.02 0.51 3 cd Shrub screef 12.67 6.46 0.51 3 d b) Site 2 Treatment Mean (%) S.D. CV. n Tukey Herb control 104.43 39.05 0.37 3 a Herb clip 50.17 29.33 0.58 3 ab Herb screef 13.43 7.74 0.58 3 be Shrub control 114.20 5.72 0.05 3 a Shrub clip 19.53 11.73 0.60 3 be Shrub screef 8.60 2.09 0.24 3 c 1 Numbers are rounded to two decimal places. Changes in shrub cover calculated as: ([final cover / initial cover] x 100). Cover was summed by species. 2 Treatment means differing significantly with p < 0.05 are represented by different letters. S.D. = Standard deviation. CV. = Coefficient of variation, n = number of plots. Table 5.7 Two-way ANOVA for loge-transformed changes in shrub cover ([final cover /initial cover] x 100) three growing seasons after disturbance in herb and shrub patches on sites 1 and 2. 298 SS DF MS F-ratio a) Site 1 SOURCE1 Type 7.66 Treat 19.25 Type * Treat 0.16 Error 3.56 b) Site 2 SOURCE Type 0.56 Treat 17.17 Type * Treat 0.84 Error 3.48 1 7.66 25.82 * 2 2 9.62 32.42 * 2 0.08 0.27 ns 12 0.30 1 0.56 1.92 ns 2 8.59 29.57 * 2 0.42 1.45 ns 12 0.29 SS DF MS F-ratio 1 Numbers are rounded to two decimal places. Type = Herb, Shrub, Treat = Control, Clipped, Screefed. 2 * = significant with p < 0.05, ns = nonsignificant with p > 0.05. Table 5.8 Two-way ANOVA for Berger-Parker dominance index three growing seasons after disturbance in herb and shrub patches on sites 1 and 21. 299 a) Site 1 SOURCE2 SS DF MS F-ratio Type 0.000 1 0.000 0.001ns3 Treat 0.101 2 0.051 3.290 ns Type* Treat 0.041 2 0.021 1.338 ns Error 0.184 12 0.015 b) Site 2 SOURCE SS DF MS F-ratio Type 0.007 1 0.007 1.760 ns Treat 0.163 2 0.081 21.039* Type* Treat 0.205 2 0.103 26.527* Error 0.046 12 0.004 1 Berger-Parker index calculated as: number of individuals of most abundant species / total number of individuals. 2 Type = Herb, Shrub, Treat = Control, Clipped, Screefed. 3 ns = nonsignificant with p > 0.05, * = significant with p < 0.05. Table 5.9 Two-way ANOVA for constancy index three seasons after disturbance in herb and shrub patches on sites 1 and 21. 300 a) Site 1 SOURCE2 SS Type 0.024 Treat 0.025 Type * Treat 0.060 Error 0.203 DF MS F-ratio 1 0.024 1.429 ns3 2 0.013 0.746 ns 2 0.030 1.762 ns 12 0.017 b) Site 2 SOURCE SS DF MS F-ratio Type 0.120 1 0.120 25.068* Treat 0.009 2 0.004 0.901ns Type* Treat 0.071 2 0.035 7.403 * Error 0.057 12 0.005 1 Constancy index calculated as: proportion of total number of sampling points with same species as nearest neighbour in 1988 and 1989. 2 Type = Herb, Shrub, Treat = Control, Clipped, Screefed. 3 ns = nonsignificant with p > 0.05, * = significant with p < 0.05. 301 Table 5.10 Two-way ANOVA and statistics for logg-transformed (distance between sampling point and nearest neighbour) three growing seasons after disturbance in herb and shrub patches on site 1. a) ANOVA SOURCE1 SS DF MS F-ratio Type 26.49 1 26.49 62.15 * 2 Treat 34.07 2 17.03 39.97 * Type * Treat 2.70 2 1.35 3.17* Plot 7.19 6 1.20 2.81 * Error 225.03 528 0.43 1 Numbers are rounded to two decimal places. Type = Herb, Shrub. Treat = Control, Clipped, Screefed. 2 * = significant with p < 0.05. b) Statistics (for untransformed values) Treatment Mean (cm) S.D. CV. n Tukey Herb control 2.20 2.41 1.09 90 b Herb clip 1.59 1.88 1.19 90 b Herb screef 4.03 4.78 1.19 90 ab Shrub control 3.17 3.25 1.02 90 b Shrub clip 3.61 3.42 0.95 90 ab Shrub screef 7.20 5.56 0.77 90 a S.D. = Standard deviation. CV. = Coefficient of variation, n = number of plots. Table 5.11 Species with highest percent cover (in parentheses) in plots before and after disturbance on sites 1 and 2. 302 a) Site 1 Time since disturbance (months) Treatment-Plot Herb control Herb clipped 0 2 12 25 1 nm 1 nm 1 VS(15) 1 RV(30) 2 nm nm VS(30) VS(20) EA(20) 3 nm nm VS(25)TO(25) TO(25) RV(25) GD(25) 1 EA(70) EA(70) EA(18) VS(18) EA(50) 2 VS(40) VS(10) DA(10) VS(20) VS(50) 3 VS(80) VS(15) RV(15) VS(20) VS(50) 1 VS(95) RP(2) VS(5) EA(25)» 2 VS(80) VS(2) VS(2) EA(25)» 3 VS(40) RP(2) VS(15) RV(15) EA(65) Shrub control -1 nm nm MF(96) MF(90) - 2 nm nm RA(70) RA(50) - 3 nm nm MF(80) MF(80) Shrub clipped -1 MF(60) nv CU(4) MF(7) - 2 MF(90) VS(3) VS(6) VS(10) - 3 MF(60) RA(60) nv W(2) RA(25) Shrub screefed -1 MF(100) RV(4) RV(1) RV(6) - 2 MF(100) MF(8) MF(5) MF(15) - 3 MF(90) MF(2) MF(1) RV(1) MF(5) 3 0 3 Table 5.11 (Cont'd) b) Site 2 Time since disturbance (months) Treatment-Plot Herb control Herb clipped 0 2 12 25 1 1 nm 1 nm 1 VS(40) 1 VS(40) 2 nm nm TO(85) TO(80) 3 nm nm VS(25) VS(30) TO(30) AL(30) 1 VS(60) VS(30) VS(35) VS(60) 2 VS(80) VS(60) VS(85) VS(90) 3 VS(85) VS(50) VS(40) VS(80) 1 VS(90) VS(2) VS(2) EG(30) 2 VS(90) VS(1)AL(1) VS(3) EG(50)» 3 TO(70) TO(2) TO(3) EG(10) Shrub control -1 nm nm MF(40) MF(50) -2 nm nm RA(50) RA(50) -3 nm nm MF(80) MF(80) Shrub clipped -1 MF(95) MF(4) VS(8) MF(12) -2 MF(100) MF(3) GD(3) VS(8) MF(25) -3 MF(50) RA(50) MF(2) VS(1) SR(1) MF(8) Shrub screefed -1 MF(80) nv uh(l)* MF(5) -2 MF(70) MF(2) MF(1) MF(6) -3 MF(90) MF(1) VS(1) SR(1) VM(5) Species Codes: VS = Valeriana sitchensis DA = Dryopteris assimilis W = Veratrum viride RV = Rubus parviflorus RP = Rubus pedatus AL = Arnica latifolia EA = Epilobium angustifolium MF = Menziesia ferruginea EG = Epilobium glandulosum TO = Thalictrum occidentale RA = Rhododendron albiflorum SR = Streptopus roseus GD = Gymnocarpium dryopteris CU = Clintonia uniflora VM = Vaccinium membranaceum Scientific nomenclature follows Angove and Bancroft (1983) & Hitchcock and Cronquist (1973). nm = not measured: nv = no vegetation: uh = unidentified herb: * = not present in plot prior to disturbance. 304 Table 5.12 Species with highest percent cover (in parentheses) before and after repeated clipping in the planted spruce plots on sites 1 and 2. a) Site 1 Time since disturbance (months) Treatment-Plot 0 26 i !— r Herb control -1 VS(80) VS(50) - 2 VS(80) VS(80) - 3 VS(60) TO(50) Herb clipped -1 VS(75) RV(15)* - 2 VS(50) RP(15) -3 VS(60) VG(30) Shrub control -1 MF(50)RA(50) MF(70) - 2 MF(80) MF(75) - 3 MF(80) MF(75) Shrub clipped -1 MF(85) MF(10) -2 MF(95) RP(10) EA(10)* -3 MF(80) MF(8)TU(8) Skid trail control -1 VS(60) VS(50) -2 VS(40) EA(40) -3 VS(50) VS(60) Skid trail clipped -1 VS(40) RV(10)* - 2 VS(40) RV(30) - 3 VS(40) VG(10) 305 Table 5.12 (Cont'd) b) Site 2 Time since disturbance (months) Treatment-Plot 0 26 I [ Herb control -1 VS(80) VS(60) - 2 VS(60) GD(55) -3 VS(60) GD(40) Herb clipped -1 VS(60) AL(10) - 2 VS(50) TO(15) EG(10)* -3 VS(50) TO(10) Shrub control -1 MF(50)RA(50) RA(50) - 2 MF(80) MF(80) - 3 MF(55) MF(90) Shrub clipped -1 MF(90) MF(7) -2 MF(75) MF(10) -3 MF(85) MF(15) Skid trail control -1 VS(60) VS(60) - 2 VS(40) EA(40) EA(85) -3 VS(40) EA(80) Skid trail clipped -1 VS(60) VS(20) - 2 VS(70) GS(5) - 3 VS(70) TU(15) Species Codes: VS = Valeriana sitchensis RV = Rubusparviflorus RP = Rubuspedatus AL = Arnica latifolia EA = Epilobium angustifolium MF = Menziesia ferruginea EG = Epilobium glandulosum TO = Thalictrum occidentale RA = Rhododendron albiflorum GD = Gymnocarpium dryopteris TU = Tiarella unifoliata VG = Viola glabella GS = grass species (unidentified) Scientific nomenclature follows Angove and Bancroft (1983) & Hitchcock and Cronquist (1973). * = not present in plot prior to disturbance. 306 Table 5.13 Growth of clipped and undipped plants of Valeriana sitchensis and Menziesia ferruginea between June 6 and August 6 (1987) on site 1. a) Valeriana sitchensis (Sitka valarian) Month Treatment1 Herb patches2 Shrub patches3 Height(cm) Diameter(cm) n Height(cm) n June Control 20.4 ± 4.6 29.0 ± 6.4 10 13.6 ± 7.1 10 Clipped 20.8 ± 3.4 34.4 ± 7.2 10 15.1 ± 5.5 10 August Control 44.3 ± 13.3 46.5 ± 13.2 10 27.6 ± 7.1 8 CUpped 30.3 ± 6.3 34.6 ± 7.6 10 15.5 ± 10.94 10 August as Control 227.9 ± 95.0 159.8 ± 25.1 10 275.6 ± 90.3 8 % ofJune Clipped 147.9 ± 33.7 102.5 ± 22.9 10 104.0 ±61.1 10 b) Menziesia ferruginea (false azalea) Month Treatment1 Height Mature plants Number of stems n Juvenile plants Height(cm) n June Control 134.4 ± 27.4 18.6 ± 11.3 5 6 35.1 ± 6.7 20 Clipped 144.0 ± 23.4 11.6 ± 1.8 5 -August Control 2.01 ± 1.04 7 18.6 ± 11.3 5 17.4 ± 4.07 20 Clipped 30.0 ±24.4 16.0 ± 7.1 5 -August as Control 101.5 ± 0.8 _ 5 137.8 ± 48.9 20 % ofJune Clipped 19.1 ± 13.6 5 Values are treatment means and standard deviations. Height was measured to top of leaf stems, not inflorescence. Diameter was the diameter of the clump (i.e., entire plant), not individual stems. Plants beneath shrubs were single-stemmed, diameter was not measured. One plant died. Generally, shrubs were not common in herb patches, mature plants were in shrub patches (by definition), and juveniles were disturbed environments, typically adjacent to skid trails. Each n was the mean of 5 stems/plant. These values represent only the length of new growth, not the length the entire stem. However, for the calculation of %, the entire stem length (new and old growth) was used. 307 Table 5.14 Characteristics of Engelmann spruce established prior to 1987 in herb and shrub control plots on sites 1 and 2. . a) Site 1 Characteristics Plotl Herb Patch Plot 2 Plot 3 Plotl Shrub Patch Plot 2 Plot 3 Number of trees per plot (plot size = 16mz) 2 1 7 10 20 0 Average height (cm) 3.0 137.0 30.8 33.0 22.0 — Seedling condition FAIR GOOD GOOD POOR GOOD — Rooting substrate ROTTEN WOOD DEAD WOOD ROTTEN WOOD ROTTEN WOOD ROTTEN WOOD b) Site 2 Characteristics Plotl Herb Patch Plot 2 Plot 3 Plotl Shrub Patch Plot 2 Plot 3 Number of trees per plot (plot size = 16mz) 1 7 5 1 5 46 Average height (cm) 36.0 19.4 19.8 10.0 4.8 27.9 Seedling condition GOOD GOOD GOOD GOOD GOOD FAIR Rooting substrate DEAD WOOD ROTTEN WOOD, HUMUS ROTTEN WOOD HUMUS DEAD WOOD ROTTEN WOOD 308 Table 5.15 Characteristics of Engelmann spruce seedlings invading herb and shrub plots after treatment in 19871. a) Site 1 Characteristics Control Herb Patch Clipped Screefed Control Shrub Patch Clipped Screefed Number of seedlings per m2 0 2 4 0 0 2 Year of germination - 1988 1988(1)2 1989(3) - - 1989 Rooting substrate — ROTTEN WOOD MINERAL SOIL - - ROTTEN WOOD b) Site 2 Characteristics Control Herb Patch Clipped Screefed Control Shrub Patch Clipped Screefed Number of seedlings per 1 2 5 0 0 13 Year of germination 1988 1989 1988 (2) 1989 (3) - - 1988 (6) 1989(7) Rooting substrate HUMUS ROTTEN MINERAL SOIL, WOOD HUMUS, ROTTEN WOOD MINERAL SOIL, HUMUS, ROTTEN WOOD 1 2 Numbers of seedlings are presented as totals for the three plots in each treatment. Numbers in parentheses indicate the number of seedlings that germinated in that year. 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100399/manifest

Comment

Related Items