UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Host responses in Douglas-fir, western hemlock and western redcedar to infection by Armillaria ostoyae… Cleary, Michelle R. 2007

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

Item Metadata

Download

Media
831-ubc_2007-266960.pdf [ 39.29MB ]
Metadata
JSON: 831-1.0074967.json
JSON-LD: 831-1.0074967-ld.json
RDF/XML (Pretty): 831-1.0074967-rdf.xml
RDF/JSON: 831-1.0074967-rdf.json
Turtle: 831-1.0074967-turtle.txt
N-Triples: 831-1.0074967-rdf-ntriples.txt
Original Record: 831-1.0074967-source.json
Full Text
831-1.0074967-fulltext.txt
Citation
831-1.0074967.ris

Full Text

HOST RESPONSES IN DOUGLAS-FIR, WESTERN HEMLOCK AND WESTERN REDCEDAR TO INFECTION BY ARMILLARIA OSTOYAE AND ARMILLARIA SINAPINA by MICHELLE R. CLEARY HBES Forest Conservation, Lakehead University, 2001 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY m THE FACULTY OF GRADUATE STUDIES (Forest Sciences) THE UNIVERSITY OF BRITISH COLUMBIA April 2007 © Michelle R. Cleary, 2007 A B S T R A C T Necrophylactic periderm (NP) formation and compartmentalization of infected tissue were examined in roots of 20-30 year-old western redcedar {Thujaplicata), western hemlock (Tsuga heterophylla) and Douglas-fir (Pseudotsuga menziesii) trees infected by Armillaria ostoyae. Microscopic investigation o f abiotically wounded roots, as well as roots naturally infected and inoculated with A. ostoyae revealed distinct differences in the types and frequency o f host responses between cedar and the other two conifers. Following invasion by A. ostoyae, a higher frequency o f successful resistance reactions was induced in western redcedar compared to Douglas-fir and western hemlock. Breaching of non-suberized impervious tissue (NIT) and N P was common in Douglas-fir and western hemlock trees. The barrier zone in cedar formed by the uninjured cambium was comprised of axial parenchyma with pigmented deposits and provided a permanent barrier to spread by the fungus. Unique resistance mechanisms in cedar involving induced rhytidome formation impart increased resistance to the spread o f A. ostoyae in host tissue. In three inoculation trials, penetration o f l iving bark on host roots by A. sinapina did not differ from A. ostoyae. However, the frequency of successful resistance reactions induced following invasion by A. sinapina in Douglas-fir and western hemlock was significantly higher than the same species infected with A. ostoyae. Inoculum potential and host-pathogen interactions were key determinants o f pathogenicity o f A. sinapina on all hosts. In a survey o f twenty juvenile mixed species plantations throughout the southern Interior of B . C . , cumulative mortality in Douglas-fir trees was significantly higher than in western redcedar trees (p < 0.001). The incidence o f mortality decreased with increasing tree size for both species, however the rate o f decrease was markedly greater among cedar compared to Douglas-fir trees. The proportion of trees that showed compartmentalization and callusing at the root collar increased with increasing tree size, but the increase was markedly greater for cedar than Douglas-fir and occurred much earlier even when the trees were relatively small. Results indicate that the higher degree of resistance against A. ostoyae in western redcedar may help alleviate long-term impacts of root disease when regenerated on sites infested with Armil lar ia root disease. TABLE OF CONTENTS Abstract 1 1 Table of Contents i i i List o f Tables v i i i List o f Figures x List o f Abbreviations xxv i i i Acknowledgements xxix C H A P T E R I Introduction and General Literature Review 1 1.1 General Introduction 1 1.2 ArmiHaria Taxonomy and Species Identification 3 1.3 Armil lar ia Species in British Columbia 5 1.4 Inoculum and Rhizomorphs 6 1.5 Infection Biology ; 8 1.6 Host Response to Infection 9 1.6.1 Exudate Production 9 1.6.2 Meristematic Activi ty 10 1.6.3 Biochemical Defense 14 1.6.4 Compartmentalization of Decay in Trees (CODIT) 16 1.7 Host Susceptibility 19 1.8 Armil lar ia root disease in the Southern Interior o f British Columbia 22 1.9 Conclusions 26 C H A P T E R II: Macro- and Microscopic Host Response to Abiot ic Wounding, Inoculation with Armillaria ostoyae and Natural Infections 28 2.1 Introduction 28 2.1.1 Periderm Anatomy 32 2.1.2 N P Formation in Response to Wounding 34 2.1.3 NIT: A Tissue Essential for Regeneration of N P 36 2.1.4 Factors Affecting N I T Development and N P Formation 39 2.1.5 N I T Development and N P Formation: Their Role in Resistance Against Armil lar ia root disease 40 2.2 Materials and Methods 42 2.2.1 Study Sites 42 2.2.2 Inoculum Block Preparation 42 2.2.3 Inoculation and Sampling Technique 44 2.2.3.1 Winter Inoculations 47 2.2.3.2 Examination of Roots Naturally Infected with A. ostoyae 48 2.2.4 Sample Treatment 49 2.2.5 Re-Isolation o f A. ostoyae from lesions 52 2.2.6 Statistical Analysis : 53 2.3 Results and Discussion 54 2.3.1 Inoculation Trial: Frequency of Infection 54 2.3.2 Characterization o f Healthy Root Bark Tissues 59 2.3.3 Characterization of Abiotically Wounded Root Bark Tissues 62 2.3.4 Control Blocks 75 2.3.5 Host Responses to Inoculation with A. ostoyae in Roots 77 2.3.5.1 Characterization of the Different Stages o f Host Response to Infection 80 2.3.5.2 Model for Host-Pathogen Interactions 81 2.3.6 Host Variables 126 2.3.7 Winter Inoculation Trial 133 2.3.8 Lesions Analysis of Naturally Infected Individuals 138 ; 2.4 Conclusions 148 C H A P T E R III: Host response to infection by Armillaria sinapina in the roots of Douglas-fir, western hemlock and western redcedar 152 3.1 Introduction 152 3.2 Materials and methods 153 3.2.1 Study Sites 153 3.2.2 Inoculum Block Preparation, Inoculation and Sampling Technique 154 3.2.3 Re-Isolation o f A. sinapina from Lesions 155 3.2.4. Statistical Analysis 155 3.3 Results 155 3.3.1 Inoculation Trial: Frequency of Infection 155 3.3.2 Host Response to Inoculations with A. sinapina in Roots 159 3.3.3 Re-Isolation of A. sinapina from Lesions 180 3.4 Discussion 181 C H A P T E R I V : Above-ground Symptoms Development and Mortality Rates by Species in Juvenile M i x e d Conifer Stands 189 4.1 Introduction . 189 4.2 Materials and Methods 191 4.2.1 Site selection : 191 4.2.2 Plot selection and Measurement of trees 192 4.2.3 Statistical Analysis 193 4.3 Results : 194 4.4 Discussion 211 C H A P T E R V : Summary and Conclusions 219 Bibliography 224 Appendix I Map highlighting the Okanagan-Shuswap and Arrow-Boundary Forest Districts in the southern interior o f British Columbia where inoculation trials were conducted 239 Appendix II Chi-square tests of the frequency of infection by A. ostoyae among species for all field trials and winter inoculation trials... 240 Appendix III Chi-square tests o f the frequency o f host responses among species for all field trials 251 Appendix IV Chi-square tests o f the frequency of successful resistance reactions as a percentage of successful penetrations by A. 253 ostoyae among species on the different harvest dates and field trials Appendix V Chi-square tests o f the frequency o f N I T and N P formed, N I T and N P breached, the number o f roots that showed cambial invasion and compartmentalization among species 258 v i Appendix V I Relationship between root age and the frequency o f successful resistance reactions in Douglas-fir, western hemlock and western redcedar following penetration by A. ostoyae 264 Appendix VII Relationship between inner bark thickness and the frequency of successful resistance reactions in Douglas-fir, western hemlock and western redcedar following penetration by A. ostoyae 265 Appendix VIII Relationship between tree age and the frequency of successful resistance reactions in Douglas-fir, western hemlock and western redcedar following penetration by A. ostoyae 266 Appendix I X Relationship between tree size and the frequency o f successful resistance reactions in Douglas-fir, western hemlock and western redcedar following penetration by A. ostoyae 267 Appendix X Chi-square tests o f the frequency o f infection by A. sinapina among species for all field trials 268 Appendix X I Chi-square tests of the frequency o f successful resistance reactions following inoculation with A. sinapina among species 272 Appendix XI I Chi-square tests o f the frequency of N I T and N P initiated, N I T and N P breached, the number of roots that showed cambial invasion and compartmentalization among species in roots 274 infected with A. sinapina and A. ostoyae , Appendix XIII Disease incidence on hardwood species from all 20 sites combined 280 Appendix X I V The number o f trees tallied by species and disease status for all twenty sites 281 Appendix X V Logistic regression model analysis between Douglas-fir and western redcedar 303 Appendix X V I Tree size distribution o f western hemlock at 10 sites 313 Appendix XVII Chi-square tests o f the frequency o f mortality caused by A. ostoyae among species for 10 sites combined 314 v i i Appendix XVIII Chi-square tests of the frequency o f mortality (dead trees) and progressive lesions at the root collar (dying trees) among species 315 Appendix X I X Chi-square tests o f the frequency o f mortality, progressive lesions at the root collar, and callused lesions at the root collar as a proportion of the total number of trees showing aboveground symptoms of disease among species 316 Appendix X X Stand Establishment Decision A i d for Armil lar ia root disease for the Southern Interior Region of British Columbia 319 v i i i L I S T O F T A B L E S 2.1. Location and characteristics of three sites used in the four inoculation trials at which the host response to infection by Armillaria ostoyae in the roots of western hemlock, western redcedar and Douglas-fir trees was investigated 43 2.2 Number of root inoculations on Douglas-fir, western hemlock and western redcedar and the number and proportion producing infection by A. ostoyae at various harvest times in four field trials 55 2.3. The total number of trees sampled and the total number of roots examined following abiotic wounding of Douglas-fir, western hemlock and western redcedar roots at the different harvest dates for each field trial. Roots were initially wounded in early May of each year 63 2.4. The total number of control blocks in both the Kingfisher and Nakusp field inoculation trials by species, the number of control blocks colonized by on-site inoculum, and the number of roots that resulted in a lesion in the bark at surface contact with the control block; 75 2.5. Frequency of successful resistance reactions following infection by A. ostoyae in the roots of Douglas-fir, western hemlock, and western redcedar observed from four separate field trials 2002-2004 118 2.6 Individual species comparisons of the frequency of successful resistance reactions following inoculation with A. ostoyae for different harvest dates and inoculation trials 119 2.7 The number of root inoculations that resulted in successful penetration by A. ostoyae, the frequency of roots showing no visible or ineffective host response, initiation and breaching of non-suberized impervious tissue (NIT) and necrophylactic periderm (NP), killed cambium and compartmentalization, and the number and percentage of roots showing successful resistance reactions in Douglas-fir, western hemlock, and western redcedar trees for all field trials combined 121 2.8 Individual species comparisons of the frequency of roots showing NIT initiated, NIT breached, NP formed, NP breached and compartmentalization following inoculation with A. ostoyae in Douglas-fir, western hemlock, and western redcedar trees for all harvest dates in four inoculation trials 122 2.9 The number of Douglas-fir, western hemlock and western redcedar inoculated with A. ostoyae in the fall and the frequency of the resulting infection the next spring 136 2.10 The average age, average tree size, total number of trees sampled and number of lesions examined from the roots of Douglas -fir, western hemlock, and western redcedar trees naturally infected with A. ostoyae for all species at both sites near Hidden Lake 138 ix 3.1 Number of root inoculations on Douglas-fir, western hemlock and western redcedar trees and the number and proportion of roots that showed infection by A. sinapina at various harvest times in three field trials 157 3.2. Frequency of successful resistance reactions following infection by A. sinapina in the roots of Douglas-fir, western hemlock, and western redcedar observed from three separate field trials 2002-2004 163 3.3. Frequency data from Douglas-fir, western hemlock and western redcedar roots inoculated with A. sinapina and A. ostoyae harvested at 4-5 months and 1 -year from all field inoculation trials 164 3.4. Individual species comparisons of the frequency of the type of host response induced following infection by A. sinapina and A. ostoyae in Douglas-fir, western hemlock and western redcedar trees 165 4.1. Characteristics of the twenty sites surveyed for incidence of infection and mortality by A. ostoyae in juvenile mixed conifer stands in the Okanagan-Shuswap and Arrow-Boundary Forest Districts in the southern interior of B.C 195 4.2. Total number of conifer trees tallied by species in different disease status categories for all twenty sites combined 198 4.3. The proportion of the total number of trees by disease status category for Douglas-fir and western redcedar 199 4.4. Incidence of progressive infections, callused infections and mortality in Douglas-fir and western redcedar as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae 201 4.5. The proportion of the total number of trees by disease status category for Douglas-fir, western hemlock and western redcedar from ten sites 206 4.6. Incidence of progressive infections, callused infections and mortality in Douglas-fir, western hemlock and western redcedar as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae in 10 sites 207 X LIST OF FIGURES P A G E 1.1 Diagrammatic views of the anatomical model for non-specific defense mechanisms that follow (a) penetration of the bark surface, (b) penetration of the vascular cambium, and (c) penetration of the sapwood (From Mullick 1977) 11 2.1. A transverse section of stem of Sambucus nigra L . showing an early stage in the development of periderm with phellogen and its derivative tissues 32 2.2. Phellem are formed by periclinal divisions of the phellogen. The division of the phellogen cell gives either a phellem or phelloderm cell (cell 1). Over time, additional phellem and phelloderm derivatives are formed either externally or internally abutting the phellogen, respectively. The first derivatives of the phellogen are pushed outward so that they appear farthest from the phellogen. Phellem production is generally greater than phelloderm (from Mullick 1977) 32 2.3. Inoculation technique shown in situ. A. ostoyae inoculum block placed alongside a healthy western hemlock root 45 2.4. A n inoculum block as it was placed against a western redcedar root prior to harvesting from the ground. Clusters of A. ostoyae rhizomorphs are shown emerging from the inoculum block and adhering to the outer surface of the root. 45 2.5 Excavation of roots systems between a recently killed Douglas-fir and healthy western redcedar. Root contacts are flagged for examination of Armillaria-caused lesions at root contact 49 2.6. A sample of root bark from a healthy 19-year-old western redcedar tree 61 2.7. A sample of a root bark from a healthy 19-year-old western hemlock tree 61 2.8. A sample of root bark from a healthy 31-year-old Douglas-fir tree 61 2.9. A cryofixed section of healthy western redcedar root bark, BF 61 2.10. Same section shown in Fig. 2.9., B L 61 2.11. Same section shown in Fig. 2.9, U V 61 2.12. A cryofixed section of healthy western hemlock root bark, BF 61 2.13. Same section shown in Fig. 2.12., B L . Note clusters of sclereids visible in Fig. 2.12 and 2.13 61 2.14 Same section shown in Fig. 2.12, U V 61 2.15. A cryofixed section of healthy Douglas-fir root bark, BF 61 2.16. Same section shown in Fig. 2.15, B L . Note thick-walled stone phellem in Figures 2.15 and 2.16 61 x i 2.17. Same section shown in Fig. 2.15, U V 61 2.18. A phloroglucinol-HCl-treated section of a healthy western redcedar root. Both the thin-walled phellem (arrow) and phloem fibres stain positively for lignin, viewed under BF 62 2.19. The same section viewed under U V which shows the same phellem cells fluorescence brightly for suberin 62 2.20. A Sudan III stained section of a healthy Douglas-fir root bark showing alternating layers of thin-wal led (suberized) phellem shown here as bright orange, and thick-walled (stone) phellem, B L , x 45 62 2.21. A sample of Douglas-fir root bark 5 weeks after wounding, x 12 65 2.22. A cryofixed section of abiotically wounded Douglas-fir root bark. Freeze-killed tissue and outer boundary of NIT zone fluoresce bright-yellow green with B L , x 45 65 2.23. A sample of Douglas-fir root bark, 9 weeks after wounding showing distinct zone of clear-white tissue underlying browned tissue, x 12. Note: the original periderm sloughed off during the process of cryofixation 65 2.24. A phloroglucinol-HCl treated section of abiotically wounded Douglas-fir root bark, sampled 5 months after wounding showing dark staining of sieve cells comprising the NIT zone and positive staining for lignin in the thick-walled stone phellem overlying radially compressed thin-walled phellem 65 2.25. A Sudan III treated section of abiotically wounded Douglas-fir root bark, sampled 5 months after wounding showing suberized thin-walled phellem internally abutting two layers of stone phellem, x 45 65 2.26 A sample of western hemlock root bark, 5 weeks after wounding, x 12. Note clusters of sclereids 69 2.27. A phloroglucinol-HCl treated section of abiotically wounded western hemlock root bark, sampled 5 months following wounding, BF. NIT zone is distinct as the zone of hypertrophied tissue that stains very strongly for lignin 69 2.28. A Sudan II treated section of abiotically wounded western hemlock root bark, sampled 5 weeks after wounding, B L 69 2.29. A sample of a western hemlock root bark, 5 months following abiotic wounding showing a colorless zone of modified tissue internally abutting the necrotic zone indicative of a newly restored zone of phellogen, x 12. Note clusters of sclereids (arrows) 69 x i i 2.30. A sample of a western hemlock root, 7 cm in diameter, sampled 5 months following abiotic injury. A single band of phellem (stone phellem) can be seen as the clear zone of tissue internally abutting the necrotic zone and externally abutting a zone of thin-walled phellem (dark purple pigmented line). This clear zone comprised a band of thick-walled phellem. In this image the phellogen was separated from the necrotic tissue since OCT (shown in white) occupied the space between the necrotic tissue and underlying phloem, x 25 69 2.31. A cryofixed section of abiotically wounded western hemlock showing incomplete NP formation around clusters of sclereids, B L 69 2.32. A cryofixed section of wounded root bark sampled 5 months following injury showing a well organized NP with up to 9 thin-walled phellem in neat radial files, U V 69 2.33. A cryofixed section of wounded root bark sampled 1 year following injury, showing at least 4 rows of thin-walled phellem internally abutting the radially compressed phellem produced the year prior, BF, x 45 69 2.34 A sample of western redcedar root bark showing freeze-killed tissue to approximately half the bark thickness and a NP bordering injured tissue. Note pigmented phellem wrapped around phloem fibers, x l2 74 2.35. A cryofixed section of abiotically wounded western redcedar root showing bright yellow-green fluoresced necrotic tissue and a NP with 2 layers of thin-walled phellem 74 2.36. A Sudan III treated section of abiotically wounded western redcedar root bark, sampled 9 weeks after wounding, B L . Significantly more phellem production around individual fibers than adjacent areas is seen 74 2.37. A phloroglucinol-HCl treated section of wounded root bark, sampled 5 months following wounding, BF. NIT zone is inconspicuous and masked by lignification in the walls of the thin-walled phellem 74 2.38. The same section as in 2.38 but viewed under U V . Lignified, thin-walled phellem now fluoresce bright blue-violet indicating suberization of cell walls 74 2.39. A sample of western redcedar root bark showing NP formation around necrotic tissue and successive periderm formation (induced rhytidome formation) in adjacent phloem tissue. The new NP will extend proximally and distally away from the primary lesion to become continuous with the original periderm 74 2.40. A cryofixed section of abiotically wounded western redcedar showing a distinct zone of hypertrophy and hyperplasia associated with the induced rhytidome response. Internal to this zone, a meristematic layer of cells produced a single layer of thin-walled phellem, x 45, U V 74 2.41. A Sudan III treated section of tissue showing the suberized phellem formed internal to the induced rhytidome merging with the original periderm at the distal end of the lesion, x 45, B L 74 x i i i 2.42. A cryofixed section of Fig. 2.41 showing a second periderm resulting from induced rhytidome formation deeper in the bark tissue. Eventually all cells external to the last formed periderm become moribund and fluoresce bright yellow-green, B L 74 2.43. Monopodially branched rhizomorphs typical of A. sinapina produced in culture from an isolate from the control block 76 2.44. Incompatibility reaction in dual culture between the isolate obtained from the control block and isolate 87-01 used in the field trials. A dark pigmented zone developed in the medium between the two opposing mycelia (white arrow) and radial growth became flattened in the zone of inhibition 76 2.45. A Sudan III treated section of a western hemlock root showing slight browning of phloem tissue at surface contact with a control block after 9 weeks. Phellogen restoration was complete and at least 3-4 layers of thin-walled phellem were formed. B L , x 45 76 2.46. A photomacrograph of an infected Douglas-fir root showing resin exudation (arrows) on the surface of the root 1 year following inoculation with A. ostoyae 79 2.47. A photomacrograph of an infected western hemlock root showing resinosus (arrows) on the surface of the root 5 months following inoculation with A. ostoyae. Note darker discoloration of the bark showing the extent of necrosis in the underlying phloem 79 2.48. A photomacrograph of a western redcedar root showing uneven irregularities on the surface of the roots. Note rhizomorph adhering to the outer surface 79 2.49. A cryofixed section of a Douglas-fir root showing a lateral branch of a rhizomorph penetrating the outer cork layer, BF 79 2.50. A cryofixed section of a Douglas-fir root showing a rhizomorph adhering to the outer bark surface. A narrow zone of phellogen activity (typically 1 cell layer wide) was seen in the area immediately underlying the rhizomorph (arrow) compared to a more active phellogen zone (2-4 cells wide) on either side of the rhizomorph, B L 79 2.51. A western hemlock root showing a distinct zone of necrotic tissue resulting from inoculation with A. ostoyae. Fungal mycelia are absent and necrosis likely resulted from secretions of fungal enzymes/toxins from the cambial surface of the inoculum block 79 2.52. A photomacrograph of an Armillaria-lesion on a Douglas-fir root showing mycelial fans and the extent of cambial necrosis in the root 79 2.53. A mycelial fan invading the phloem of a western hemlock root. Distinct browning of tissue occurs ahead of mycelial colonization 79 xiv 2.54. A Douglas-fir root showing progressive browning of host tissue ahead of the penetrating mycelial fan 79 2.55 Schematic diagram showing the anatomical model for non-specific defense mechanisms induced following (Fig. 2.55.1) shallow injury to the bark where the injury is limited to the vicinity of the living phellogen and underlying phloem, (Fig. 2.55.2) deeper injury to the bark where necrosis extends to tissues in close vicinity, but not directly affecting the vascular cambium, and (Fig. 2.55.3.) injury to living phloem, vascular cambium and functional sapwood 82 2.56. A cryofixed section of a Douglas-fir root inoculated with A. ostoyae showing an abrupt demarcation between necrotic and adjacent healthy tissue at the infection front. Necrotic tissue has cell walls that fluoresce bright yellow-green under B L . 88 2.57. Another cryofixed section from an infected Douglas-fir root showing a diffuse boundary between necrotic and adjacent healthy tissue at the infection front, B L . 88 2.58. A photomacrograph of a Douglas-fir root showing a large wedge of mycelium invading the bark and cambial zone 88 2.59. A cryofixed section of Douglas-fir root showing a progressive lesion at the infection front. Cells in the adjacent phloem appear moribund as cells show sporadic hypertrophy and intercellular spaces are stained yellow-brown, B L 88 2.60. A cryofixed section of a Douglas-fir root showing the early stages of redifferentiation as a zone of non-fluorescence internal to a zone of hypertrophied tissue (NIT), B L 88 2.61. A phloroglucinol-HCl treated section of a Douglas-fir root inoculated with A. ostoyae showing a distinct zone of NIT comprised of several cell layers, BF 91 2.62. Phloroglucinol-HCl treated section of abiotically wounded Douglas-fir root showing the lignified NIT zone comprising only 1-2 cell layers followed by three layers of stone phellem, BF 91 2.63. A Sudan III treated section of Douglas-fir root 11 weeks following inoculation with A. ostoyae showing a newly restored periderm with up to 4 layers of thin-walled, suberized phellem. Note resin blister in the adjacent phloem with epithelial cells also staining positively for suberin, B L 91 2.64. 1-year following inoculation with A. ostoyae, a newly restored phellogen produced up to 7 layers of thin-walled phellem. Note the stone phellem externally abutting the thin-walled phellem, B L 91 2.65. A photomacrograph of a Douglas-fir root with its bark removed showing a smaller secondary root colonized by the fungus distally and the infection checked (compartmentalized) at the junction of the larger diameter root 91 XV 2.66. A paraffin embedded section of a Douglas-fir root showing a barrier zone formed by the uninjured cambium comprised of a tangential series of traumatic resin ducts following invasion by A ostoyae. The surrounding axial and ray parenchyma appear occluded with polyphenolic bodies. The vascular cambium (not shown) lies above the tracheids in this micrograph 91 2.67. Immediately adjacent to the killed vascular cambium, resin duct formation appears disorganized and is comprised of polyphenolic-rich axial parenchyma; deposits also accumulate in the ray cells. The vascular cambium (not shown) lies above the tracheids and ray parenchyma in this micrograph 91 2.68. A western hemlock root showing resin exudation and rhizomorphs on the root surface following inoculation with A. ostoyae 92 2.69. A photomacrograph of western hemlock root showing rhizomorph penetration and necrosis of the inner bark 92 2.70. A cryofixed section of Fig. 2.69 showing the large mycelial fans degrading and digesting the phloem tissue following invasion by the fungus, B L 92 2.71. A photomacrograph of a western hemlock root following inoculation with A. ostoyae showing a lack of hypertrophy on the proximal infection front and significant hypertrophy at the distal infection front. Note distinct zone of necrosis extended to the depth of the vascular cambium. Mycelial fans and large wedges of resin can be seen throughout 94 2.72. A cryofixed section of the proximal infection front shown in Fig. 2.70 showing a lack of hypertrophy at the infection front (no host response) in advance of a penetrating mycelium, U V 94 2.73. A cryofixed section of the distal infection front shown in Fig. 2.70 showing significant hypertrophy in the adjacent phloem and small zones of re-differentiated tissue as clusters of phellem embedded within the phloem, U V 94 2.74. A photomacrograph of a western hemlock root showing rhizomorph penetration of the inner phloem. The boundary between the necrotic tissue and adjacent healthy tissue is quite abrupt 94 2.75. A cryofixed section of Fig. 2.74 showing no visible host response at the infection front (lack of cell hypertrophy in the adjacent phloem), B L . 94 2.76. The same cryofixed section shown in Fig. 2.75 but view under U V 94 2.77. A cryofixed section of a western hemlock root following inoculation with A. ostoyae showing a distinct zone of dedifferentiated tissue (NIT) internally abutting a zone of necrosis, BF 94 2.78. The same section shown in Fig. 2.77 but viewed under B L . Note cell wall hypertrophy in the hypertrophied phloem and polyphenolic deposits occurring in phloem parenchyma cells. The zone of redifferentiation is not obvious here 94 xvi 2.79. Another cryofixed section of a western hemlock root showing a distinct zone of redifferentiation internally abutting a zone of NIT. The newly restored phellogen zone appears non-fluorescent in B L 94 2.80. A phloroglucinol-HCl treated section of a western hemlock root inoculated with A. ostoyae showing lignification of phloem parenchyma underlying necrotic tissue, BF 98 2.81. A phloroglucinol-HCl treated section of an infected hemlock root showing the lignified NIT comprising several cell layers and extending for some distance along the length of the sample close to the vascular cambium, BF 98 2.82. A cryofixed section of a western hemlock root inoculated with A. ostoyae showing incomplete differentiation of NIT around clusters of sclereids in the bark, BF 98 2.83. The same section shown in Fig. 2.82 viewed in B L . Note erratic hypertrophy and cell wall fluorescence in moribund tissue and adjacent phloem in the areas of incomplete dedifferentiation of NIT 98 2.84. A cryofixed section of a western hemlock root 11 weeks following inoculation with A. ostoyae showing a typical resistant reaction involving the complete formation of a NP around infected, necrotic tissue, B L '. 98 2.85. A Sudan III treated section of a hemlock root 11 weeks following inoculation with A. ostoyae showing the suberized phellem of the new NP becoming continuous with the original periderm, B L 98 2.86. A cryofixed section of a western hemlock root sampled approximately 1 year following inoculation with A. ostoyae. The resulting lesion was bound by a NP in which the thin-walled phellem appear to be radially compressed. At the time of sampling in the late spring, the current year's phellogen activity had already produced up to 5 layers of phellem 98 2.87. A cryofixed section of a larger diameter hemlock root showing a newly formed NP comprised of thick and thin-walled phellem, U V 98 2.88. A photomacrograph of a hemlock root showing initial NP formation in the bark and breaching of the newly formed periderm and necrosis extending down to the vascular cambium 98 2.89. A cryofixed section of a hemlock root showing incomplete differentiation of NP around clusters of sclereids, BF 98 2.90. A photomacrograph of a western hemlock root showing lateral ingrowth of callus following cambial invasion by A. ostoyae. Note newly differentiated NP in the outer periphery of the callus ] 00 XVII 2.91. A cryofixed section of western hemlock sample shown in Fig. 2.90 but viewed in B L . A NP became established in the outer periphery of the callus and a new vascular cambium was restored within the callus. Derivatives of the new cambial initials are disoriented in radial section 100 2.92. Callus formation originates via progressive hypertrophy and hyperplasia of cambial initials from the uninjured cambium as well as proliferation of the living xylem ray cells; section stained in phloroglucinol-HCl 100 2.93. The same section viewed in Fig. 2.92 but under 45X magnification showing more clearly the proliferation of xylem rays 100 2.94. A phloroglucinol-HCl treated section of a western hemlock root that showed temporary disruption of the normal cambial activity resulting in callus formation. The tissue derived from callus eventually develops secondary walls, become lignified, and remain embedded between tracheids in the annual growth ring 100 2.95. A western hemlock root sampled approximately 1-year following inoculation with A. ostoyae showed a series of traumatic resin canals following injury to the vascular cambium. Cells surrounding the resin ducts appear occluded 100 2.96. A cryofixed section of a western hemlock root showing callus formation at the edge of the killed cambium. Early differentiation of resin ducts show an oblique orientation but their structure becomes increasingly normal with increased distance from the area of killed cambium, B L 100 2.97. A western redcedar root sampled 1-year following inoculation with A. ostoyae. Symptoms of diseased roots are quite inconspicuous other than irregularities in the bark surface 104 2.98. A cryofixed section of a cedar root showing rhizomorph penetration of the inner bark 104 2.99. A cryofixed section of a western redcedar root showing rhizomorph penetration via enzymatic degradation and mechanical ingress of the fungus resulting in the collapse of the underlying phloem. 104 2.100. A photomacrograph of a cedar root showing expansion of necrosis in the bark following infection by A. ostoyae 104 2.101. A phloroglucinol-HCl treated section of a western redcedar root showing lignification of the fibers and thin-walled phellem. The fiber internally abutting the newly formed phellem stained more strongly for lignin, BF 104 2.102. The same section shown in Fig. 2.101 but viewed under U V . The same lignified phellem here appear suberized 104 2.103. A phloroglucinol-HCl treated section of a cedar root showing NP formation in the bark in response to infection by A. ostoyae. The NP is comprised of 1-2 layers of thin-walled lignified phellem, BF 104 xv i i i 2.104. The same section shown in Fig. 2.103 but viewed under U V showing the same lignified phellem fluorescing bright blue-violet indicating suberization. Note prolific formation of phellem around fibers in the vicinity of the newly differentiated periderm 104 2.105. A photomacrograph showing a typical resistance response involving NP formation in the bark following invasion by A. ostoyae 104 2.106. A cryofixed section of cedar root bark showing necrotic tissue bound by a NP. The phellem accumulate red pigments, BF 104 2.107. The same section shown in Fig. 2.106 but viewed under B L 104 2.108. A photomacrograph of a western redcedar root showing a HR in the bark. The excessive bark hypertrophy results in an increase in bark thickness up to 3X the original thickness of the bark 107 2.109. A cryofixed section of a western redcedar root showing changes in florescence characteristics in the zone of excessive hypertrophy indicative of the early stages of dedifferentiation 107 2.110. A photomacrograph showing induced rhytidome formation in cedar extending for some distance beyond the primary wounded tissue (the original NP) 107 2.111. A cryofixed section of a western redcedar root showing excessive hypertrophy and dedifferentiation in the phloem proximal and distal to the primary wounded tissue. A new meristematic phellogen differentiates immediately abutting the zone of hypertrophy and produces radial files of thin-walled lignified phellem.... 107 2.112. A western redcedar root showing massive swelling and irregularities in the otherwise smooth bark surface following inoculation with A. ostoyae. Note rhizomorphs adhering to the outer surface of the bark 107 2.113. A cryofixed section of a western redcedar root showing successive periderm formation deeper in the bark under a zone of bark hypertrophy, BF 105 2.114. The same section shown in Fig. 2.113 viewed under B L . The newly formed periderm is comprised of 1-2 layers of thin-walled phellem 107 2.115. A photomacrograph of a western redcedar root showing a NP formation in the bark and a series of small circular or elliptical zones of tissue which differentiated into traumatic resin canals in the phloem 110 2.116. A cryofixed section of cedar root bark shown in Fig. 2.115 showing fully differentiated resin ducts in the phloem surrounding the margin of the lesion 110 2.117. A cryofixed section of cedar root bark showing the initiation of traumatic resin duct formation in the phloem via excessive hyperplasia of phloem parenchyma either in mid-phloem or closer to the vascular cambium, B L 110 xix 2.118. Zones of actively dividing cells are non-fluorescent when viewed under B L 110 2.119. A cryofixed section of cedar showing the cavity of a fully differentiated resin duct resulting from schizogeny and lysigeny of the expanding hyperplasic tissue, BF 110 2.120. The same section shown in Fig. 2.119 but viewed under B L 110 2.121. A paraffin embedded section of a cedar root showing epithelial cells lining the lumen of the resin duct, BF 110 2.122. A paraffin embedded section showing traumatic resin ducts separated by axial and ray phloem parenchyma, BF 110 2.123. A cryofixed section showing phloem fibers fixed in the resin ducts. Note phellem-like cells wrapped around fibers, B L 110 2.124. A tangential section of a cedar root showing traumatic resin duct formation in the phloem following inoculation with A. ostoyae. The length of resin ducts vary but are generally between 2-5 mm, while the width range between 100-300 um, BF.. . 111 2.125. A cryofixed section of a cedar root showing phloem fibers fixed in the resin duct with pigmented phellem-like cells wrapped around the fiber, U V ; I l l 2.126. A photomacrograph of a cedar root showing NP formation and traumatic resin duct formation deep in the phloem tissue. The tissue surrounding the differentiating resin ducts appear as small circular or elliptical zones of necrotic tissue bound by phellem-like cells that accumulate the normal cell wall constituents, namely lignin and suberin, as well as the pigmented phlobaphene cell contents. With increasing distance from the area of primary wounding where a NP had differentiated, traumatic resin ducts appear normal I l l 2.127. A paraffin-embedded section of cedar root bark showing traumatic resin canal formation in the phloem adjacent to a newly differentiated NP. The parenchyma cells surrounding the lumen of the duct acculumated phellem-like pigments, BF.. . 111 2.128. A transverse section of cedar root bark collected from the root collar area of a healthy tree showing tangential series of traumatic resin canals in the phloem, BF H I 2.129. A tangential section of cedar root bark adjacent to a newly differentiating NP showing an anastomosing network of resin ducts of varying lengths and irregular accumulation of phellem-like pigments in the epithelial and/or parenchyma cells lining the resin ducts, BF I l l 2.130. A photomacrograph of a cedar root showing effective compartmentalization and callusing following cambial invasion by A. ostoyae. Note rhizomorph embedded within phloem tissue. A NP formed around necrotic tissue in the bark and lateral ingrowth of callus is evident at the margin of the killed cambium 117 XX 2.131. A cryofixed section of leading callus edge of the lesion shown in Fig. 2.130 viewed under polarized light. A NP developed in the outer periphery of the callus tissue and a new vascular cambium regenerated within the callus (shown as a darker zone of tissue) and amalgamated with the original vascular cambium. 117 2.132. A typical barrier zone in cedar is comprised of a higher than average number of axial parenchyma that accumulate polyphenols deposits. The zone of polyphenolic parenchyma tissue is generally confined to a small zone of tissue immediately adjacent to an area of killed cambium and does not extend tangentially around the circumference of the root 117 2.133. A cryofixed showing misalignment of cells at the leading edge of the callus. Cells become oriented in the appropriate plane with increasing distance from the callus edge and/or following fusion of opposing vascular cambia 117 2.134. A cedar root showing a barrier zone formed by the uninjured cambium comprised on axial parenchyma with pigmented phenolics. Lateral ingrowth of callus and woundwood extend over the area of killed cambium 117 2.135. Successful resistance reactions as a proportion of the total number of roots showing successful penetration by the A. ostoyae in Douglas-fir, western hemlock and western red cedar trees. Note: successful penetration includes cases where the original periderm is not breached but there was evidence of browning of the living phloem 125 2.136. Successful resistance reactions as a proportion of roots penetrated by A. ostoyae in relation to root diameter 127 2.137. Successful resistance reactions as a proportion of roots penetrated A. ostoyae in relation to the distance from the root collar 127 2.138. Successful resistance reactions as a proportion of roots penetrated by A. ostoyae in relation to tree vigour 132 2.139. A Douglas-fir root infected with A. ostoyae and showing traumatic resin canals formed in the next annual growth ring soon after the onset of growth in the next annual ring 134 2.140. An isolate obtained from a naturally infected western redcedar tree (tree #7905) paired with A. ostoyae isolate 87-01. After 8 weeks, incomplete intermingling of mycelia between the two opposing colonies was evident but lacked pigmentation in the media 139 2.141 Isolate obtained from the same tree identified in Fig. 2.140 but paired against A. sinapina isolate (Merritt). A dark demarcation line (arrow) developed in the pigment between the two opposing colonies indicating interspecific crosses (isolates of two different species of Armillaria) 139 xx i 2.142. A photomacrograph of a progressive lesion on a Douglas-fir root naturally infected with A. ostoyae. Large wedges of mycelium are seen advancing in the inner bark and cambium. Browning tissue in advance of mycelial colonization is seen 142 2.143. Another photomacrograph showing typical necrosis in the bark in advance of a penetrating mycelium 142 2.144. A cryofixed section of a Douglas-fir root showing sharp demarcation between infected and adjacent phloem tissue, B L 142 2.145. A cryofixed section of a Douglas-fir root showing a diffuse boundary between necrotic and apparently healthy tissue. Adjacent phloem appears hypertrophied but lacked any further differentiation of tissue leading to NIT development, BL. . . 142 2.146. A basal disk collected from the base of an infected Douglas-fir tree revealed repeated infections over several years. The fungus killed the cambium, most likely during a period of dormancy and then the host formed a barrier zone in the spring to contain the infection. Subsequent formation and breaching of barriers in the bark and wood continued for several years. Note mycelial fans in the bark and cambial zone 142 2.147. Another disk collected from a Douglas-fir naturally infected with A. ostoyae revealed a resin barrier and callus formation at the onset of a new annual growth ring (yellow arrow) as well as single and double resin barriers (black arrows) within growth rings from previous years 142 2.148. A photomacrograph of a Douglas-fir root naturally infected with A. ostoyae where the host formed a NP in advance of the penetrating mycelia. Note thick rhytidome tissue external to the last formed periderm 142 2.149. A hemlock root naturally infected with A. ostoyae showing mycelial fans and necrosis in the inner bark and cambial zone 145 2.150. A phloroglucinol-HCl treated section of a western hemlock root naturally infected with A. ostoyae showing a large zone of hypertrophied and heavily lignified NIT internal to a zone of necrotic tissue, BF 145 2.151. A Sudan III treated section of western hemlock root showing NP formation in the bark. Note phellem wrapped around large cluster of sclereids, B L 145 2.152. A macrophotograph of a western hemlock root naturally infected with A. ostoyae. The fungus was initially walled off within a NP but then the fungus breached the barrier and continued to progressively advance in the inner bark 145 2.153. A photomacrograph of a larger diameter western hemlock root showing a resistant reaction involving NP formation with multiple bands of thick and thin-walled phellem 145 xxii 2.154. A cryofixed section of a western hemlock root shown in Fig. 2.153 showing up to 4-5 layers of thick-walled phellem between bands of radially compressed thin-walled phellem. No breaching of these types of barriers was observed, B L 145 2.155. A phloroglucinol-HCl treated section of an infected hemlock root showing a short series of traumatic resin canals formed in the wood. Normal tracheid production occurred immediately following, suggesting a temporary disruption of cambial activity as the fungus advanced in the bark 145 2.156. A phloroglucinol-HCl treated section of an infected hemlock root showing callus tissue (irregularly hypertrophied cells) embedded between normal tracheids in an annual growth ring, also suggesting a temporary disruption of normal cambial activity. The tissue derived from callus eventually developed secondary walls and become lignified 145 2.157. A cryofixed section of an infected western hemlock root showing pigmented phenolic deposits in the phloem parenchyma and traumatic resin duct formation as the fungal mycelium advances in close proximity to the vascular cambium. Such defense reactions were incapable of halting further spread of the fungus 145 2.158. Adventitious rooting in western hemlock in response to infection by A. ostoyae... 145 2.159. Infection of a smaller diameter root is confined at the junction of a larger diameter root. Note callus surrounding the central column of decay 145 2.160. Compartmentalization in western hemlock and subsequent expansion of callus over the face of the lesion 145 2.161. A photomacrograph of a western redcedar root naturally infected with A. ostoyae. Note large mycelial fans in the inner bark and browning of phloem in advance of mycelial colonization 147 2.162. A photomacrograph showing a typical resistant reaction in western redcedar as NP formation in the bark and effective compartmentalization with the callus edge growing over the face of the wound 147 2.163. A cryofixed section of the western redcedar root in Fig. 2.161 showing traumatic resin duct formation in the phloem bordering a zone of necrosis 147 2.164. A photomacrograph of a cedar root collected from the root collar area of a healthy tree showing normal resin duct formation in the phloem associated with older tissue at or near the root collar 147 2.165. Barrier zone formation in western redcedar comprising a distinct zone of polyphenolic-rich parenchyma adjacent to an area of killed cambium. Callus is formed by the uninjured cambial initials and a new meristematic vascular cambium regenerates within the callus tissue to produce normal xylem and phloem derivatives 147 xxiii 2.166. An older lesion on a western redcedar root shows evidence of the barrier zone formed in response to infection as pigmented deposits within axial parenchyma... 147 2.167. A basal disk obtained from a western redcedar tree naturally infected with A. ostoyae. Compartmentalization is shown as a barrier zone immediately adjacent to the area of killed cambium (barely visible at this magnification as a distinct pink/red hues in the wood. Note black zone lines formed by the fungus in the underlying wood. Annual growth does not appear to be impeded by the fungus. The fungus was never shown to breach the barrier zone in subsequent years unlike the other conifer species in this study : 147 3.1. Extensive rhizomorph production of A. sinapina from inoculum segments 1-year following inoculation of roots 156 3.2. A Douglas-fir root showing^, sinapina rhizomorphs adhered to the outer bark (arrows) and exudation of pitch on the surface of the root 161 3.3. A western hemlock root showing pitch and rhizomorphs on the root surface four months following inoculation with A. sinapina 161 3.4. A western redcedar root showing bark swelling (at bottom) associated with the hypersensitive response involving rhytidome formation. Several A. sinapina rhizomorphs are showing on the surface of the root 161 3.5. A western hemlock root showing epiphytic growth of an A. sinapina rhizomorph on the root. Lateral branches of the rhizomorph formed two points of attachment to the outer periderm (arrows). Penetration of the inner bark was lacking 161 3.6. A cryofixed section of a Douglas-fir root showing an A. sinapina rhizomorph penetrating the outer phellem as it formed a point of attachment. The underlying phellogen zone, shown as the narrow zone of non-fluorescence, appeared narrow suggesting the suppression of phellogen activity by the fungus. Note the contiguous phellogen zone is at least 2-3 times wider than that occurring immediately underneath the penetrating rhizomorph, B L 161 3.7. A Douglas-fir root showing two separate rhizomorph-initiated lesions following inoculation with A. sinapina. Note rhizomorphs on the surface of the root. Wedges of mycelium can be seen within the necrotic tissue. The fungus killed host tissue down to the vascular cambium 161 3.8. A photomacrograph of a Douglas-fir root showing mycelium of A. sinapina immobilized within a larger zone of necrotic tissue. Hypertrophy of cells within the necrotic zone is evident 161 3.9. A photomacrograph of a Douglas-fir root inoculated with A. ostoyae showing the mycelial fan located at the leading edge of an advancing infection front 161 XXIV 3.10. The frequency of successful resistance reactions formed in the roots of Douglas-fir, western hemlock and western redcedar trees following invasion by A. sinapina and A. ostoyae as a proportion of the total number of trees that showed penetration at 4-5 months and 1 year for all field trials combined. Different letters denote significant differences within a species 166 3.11. A photomacrograph of a Douglas-fir root showing invasion of the inner bark tissue by A. sinapina 170 3.12. A cryofixed section of the same root shown in Fig. 3.11 viewed in B L . Necrosis can be seen in advance of the penetrating mycelium and no visible host response was observed in the adjacent phloem at the infection front 170 3.13. A photomacrograph of a Douglas-fir root that showed no host response following invasion by A. sinapina. Large zone of phloem appeared hypertrophied, clearly visible at the macroscopic level (arrows), but a distinct zone of NIT was not identified 170 3.14. A photomacrograph of a Douglas-fir root infected with A. sinapina showing a sharp demarcation between the boundary of infected and adjacent, healthy tissue. 170 3.15. A cryofixed section of a Douglas-fir root showing NP formation in the bark following invasion by A. sinapina, BF 170 3.16. A photomacrograph of a Douglas-fir root which identified the zone of clear-white tissue immediately underlying necrotic phloem as the newly restored N P . . . 170 3.17. Another photomacrograph of a Douglas-fir root showing two wedges of mycelium in the zone of necrotic tissue. A new phellogen zone was restored around killed tissue 170 3.18. A Sudan III stained section of a Douglas-fir root with up to 22 layers of thin-walled, suberized phellem comprised the newly formed NP, B L 170 3.19. A photomacrograph of a western hemlock root showing necrosis in the bark and cambial zone following invasion by A. sinapina 174 3.20. A cryofixed section of a hemlock root showing no visible host response and lack of cell hypertrophy in the adjacent phloem at the infection front. The boundary between the necrotic zone and the adjacent living phloem was abrupt, U V 174 3.21. A cryofixed section of a western hemlock root showing no host response. Irregular hypertrophy occurred in the adjacent phloem at the infection front and the boundary between necrotic and living tissue was rather diffuse, B L 174 3.22. A cryofixed section of a western hemlock root showing incomplete differentiation of NIT and NP around large clusters of sclereids. These areas were commonly breached by the fungus, B L 174 XXV 3.23. A photomacrograph of a hemlock root sampled 1 -year following inoculation with A. sinapina. The rhizomorph did not penetrate to great depth in the living bark before the host contained the fungus within a new NP 174 3.24. A cryofixed section of a hemlock root showing a recently formed NP around killed tissue. The new phellogen appears as the zone of non-fluorescence internally abutting the two layers of thin-walled phellem. The NIT zone was relatively inconspicuous as it only comprised 1-2 layers of hypertrophied and lignified phloem externally abutting the phellem 174 3.25. A cryofixed section of a hemlock root showing browning of tissue in the adjacent phloem internal to the newly formed NP. Breaching of NP's were common through clusters of sclereids, BF 174 3.26. A photomacrograph of a hemlock root inoculated with A. sinapina showing breaching of the NP barrier at the junction with the vascular cambium 174 3.27. A cryofixed section of a hemlock root showing barrier zone formation in the wood comprised of traumatic resin ducts (arrows), B L 174 3.28. A western redcedar root showing abundant rhizomorphs growing on the surface of the root following inoculation with A. sinapina 177 3.29. A photomacrograph of a western redcedar root showing NP formation around killed tissue following invasion by A. sinapina in the bark 177 3.30. A phloroglucinol-HCl treated section of a cedar root showing NP formation in the bark. The NIT zone is difficult to discern as thin-walled phellem also stain positive for lignin, BF 177 3.31. The same section shown in Fig. 3.30 viewed under U V fluorescence showing the same lignified thin-walled phellem also to be suberized 177 3.32. A cryofixed section of a western redcedar root showing NP formation in the bark. Killed tissue fluoresces bright yellow-green and the phellem comprising the newly formed barrier are typically 2 cells wide, B L 177 3.33. A western redcedar root showing large zones of hypertrophied phloem following inoculation with A. sinapina 177 3.34. A photomacrograph of a western redcedar root showing the degree of hypertrophy in the inner bark induced by the H R associated with rhytidome formation. This image shows the rhytidome at the proximal end of the lesion where it becomes continuous with the original periderm 177 3.35. A cryofixed section of a cedar root showing the large zone of excessive hypertrophy involved in rhytidome formation in the bark. This section shows the hypertrophied cells as non-fluorescent indicating the onset of meristematic activity, B L 177 xxvi 3.36. A cryofixed section of a cedar root showing a later stage of rhytidome formation with a newly restored periderm internally abutting the zone of hypertrophy. Cells are now moribund and fluoresce bright yellow. The structure of cells within this zone deteriorates which facilitates in the en masse sloughing of necrotic tissue from the surface of the root, B L . 177 3.37. A phloroglucinol-HCl treated section of a western redcedar root showing lignification of the thin-walled phellem comprising the newly formed NP barrier in the bark, BF 179 3.38. The same section in Fig. 3.37 viewed under U V . This image shows the degree of difficulty a newly restored phellogen has forming around killed tissue when phloem fibers occur within the zone of redifferentiation. Newly formed phellem wrap around the phloem fibers, sometimes enveloping the entire length of the cell protruding from the necrotic zone until the phellogen becomes continuous.... 179 3.39. Another photomacrograph of a cedar root showing induced rhytidome formation in the bark and extent of bark hypertrophy 179 3.40. A western redcedar root showing browned tissue in the inner bark and traumatic resin ducts in the phloem at the margin of the necrotic tissue. Slight pigmentation (shown in red) can be seen in this image. Parenchyma cells differentiating into epithelial cells surrounding the lumen of the resin duct may accumulate phellem-like pigments, particularly where resin ducts are differentiating in close proximity and in conjunction with a newly formed periderm barrier 179 3.41. Barrier zone formation in cedar following invasion by A. sinapina comprised fewer polyphenol ic-rich axial parenchyma than that induced following invasion by A. ostoyae 179 3.42. An isolate obtained from an infected western redcedar tree (tree #12993) paired with the known A. sinapina isolate used in the inoculation trial (on the right). After several weeks, complete intermingling of mycelia and rhizomorphs within the culture was evident 180 3.43. An example of an incompatible reaction between two different isolates of A. sinapina. The isolate on the left is the known isolate used in the inoculation trial paired against A. sinapina isolate #29-2-8C on the right. Incomplete intermingling of mycelia between the two opposing colonies was evident after several weeks. Growth at the leading edge of the mycelial colony was inhibited. 180 4.1. Location of study sites numbered 1 to 20 used for surveying species mortality caused by Armillaria root disease in mixed conifer stands in the southern interior of British Columbia 194 4.2. Tree size distribution as number of trees per 10 mm D B H class for western redcedar and Douglas-fir trees for all 20 sites combined 197 xxvi i 4.3. Compartmentalization of Armillaria-caused lesions in the root collar area on a western redcedar tree. Two basal lesions resulted from an infection advancing on a single lateral root. The infection was checked once it reached the root collar. Barrier zone formation can be seen as a pinkish hue in the injured sapwood. Trunk fluting appears as a result of lateral ingrowth of callus over the face of the basal lesion. Note zone line formation in the infected sapwood 202 4.4. Compartmentalization of A. ostoyae infection in the root collar area on Douglas-fir (also shown as Fig. 2.146 in Section 2.3.8 of Chapter 2). Continuous formation and breaching of barriers in subsequent years was evident. Traumatic resin canals extended tangentially around the circumference of the stem the year that the lesion was compartmentalized. Subsequent years also show shorter tangential series of traumatic resin canals 202 4.5. A basal stem disk of Douglas-fir shows resin canal formation in the year prior to cambial invasion at the root collar. Note mycelial fans and resin embedded within infected bark. Callus formation and vascular regeneration enabled the tree to grow over the infected tissue, yet the formation of traumatic resin canals was continuously stimulated in subsequent years. The fungus breached the temporary barriers in subsequent years to colonized additional cambial tissue 202 4.6. Mortality in Douglas-fir (Fd) and western redcedar (Cw) as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae by size class 204 4.7. Callused lesions at the root collar in Douglas-fir (Fd) and western redcedar (Cw) as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae by size class 204 4.8. Progressive lesions at the root collar on Douglas-fir (Fd) and western redcedar (Cw) trees caused by A. ostoyae expressed as a proportion of the total number of infected trees by size class 205 4.9. Percent dead and dying Douglas-fir (Fd), western hemlock (Hw) and western redcedar (Cw) as a proportion of the total number of trees tallied 206 4.10. Mortality in Douglas-fir (Fd), western hemlock (Hw) and western redcedar (Cw) as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae by size class 209 4.11. Callused lesions at the root collar on Douglas-fir (Fd), western hemlock (Hw) and western redcedar (Cw) as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae by size class 209 4.12. Progressive lesions at the root collar in Douglas-fir (Fd), western hemlock (Hw) and western redcedar (Cw) as a proportion of the total number of trees with above-ground signs or symptoms of A. ostoyae by size class 210 L I S T O F A B B R E V I A T I O N S Acronyms and labels for macro- and micrographs: APh, adjacent phloem; BPh, browned phloem; BF, viewed in bright field; BL, viewed under blue light fluorescence; BZ, barrier zone CT, callus tissue HPh, healthy phloem; IT infected tissue; My, mycelium; NIT, non-suberized impervious tissue; NP, necrophylactic periderm; OCT, optimum cutting temperature compound; P, original periderm; Pg, phellogen zone; Pe, thin-walled phellem; Rh, rhizomorph; Rhyt, rhytidome; TD, traumatic resin ducts TPRD, traumatic phloem resin ducts; RZ, redifferentiated zone; Sci, sclerieds; SP, stone phellem; VC, vascular cambium xxix A C K N O W L E D G E M E N T S I would like to take this opportunity to thank the many people who have supported my journey and made this thesis possible. M y sincerest gratitude goes out to my Ph.D. research supervisor, Dr. Bart van der Kamp. His enthusiasm, inspiration, sound advice, attention to detail, and sense o f practicality helped me to develop my skills as a researcher. He has spent many hours with me explaining concepts and I am thankful he was so generous with his expertise. Throughout my Ph.D. degree, he provided encouragement, financial support, critical thought and comments on the thesis drafts, and ample opportunities for professional development that would aid in my quest for a career as a Forest Pathologist. I am indebted to Dr. Duncan Morrison for taking me under his wing and 'unearthing' the exhilarating world of root disease. He is an incredible resource and inspiration. He was instrumental into introducing me to forestry and forest pathology during my employment with the Canadian Forest Service prior to starting graduate studies at U B C . His constructive critique o f writing style and willingness to discuss ideas has helped shape this thesis. His encouragement, support and kindness were greatly appreciated. I truly need to express my heartfelt gratitude to both Bart and Duncan for their confidence in my ability to undertake this study. They were both outstanding supervisors. Their vigorous support and encouragement, knowledgeable guidance in experimental design, stimulating discussions, positivism, writing and signing countless recommendation letters and grant applications, and invaluable advice during the entire study period were greatly appreciated. They have both been inspirational mentors. M u c h thanks to Garry Jensen who sacrificed many o f his 'retired' hours to helping me and teaching me the craft o f plant microtechnique and microscopy, discussing the concepts of periderm formation and interpreting host-pathogen interactions - his expertise in this field is truly a gift and his experience enriched this research. Garry's editing suggestions, stimulating discussion, constructive comments, and precise sense o f language contributed to the final copy of this thesis. I also wish to thank Dr. Joerg Bohlmann for many insightful discussions while taking his directed study course and for reading the manuscript. I thank the Canadian Forest Service, Pacific Forestry Centre for providing a second 'home' for which I was able to work and live. Without the use of the laboratory facilities this thesis would not have been possible. I would also like to thank U B C Forest Sciences office staff for managing administrative aspects o f this research. Many thanks to the following funding agencies for providing financial support to this research: National Science and Engineering Council o f Canada, Forest Investment and Innovation Ltd. , Forest Investment Account, Canadian Forest Service - Pacific Forestry Centre, U B C Forest Sciences, I M A J O Cedar Management funds, and in-kind support from the Ministry o f Forests and Range, Tolko Industries and Pope and Talbot Ltd . X X X I would like to thank many of my colleagues who have assisted me with this work, extended support in the development o f this thesis, and provided a stimulating environment in which to learn and grow over the past four years: Chef Terry Holmes, Shawn Cleary, Jason Holland, Sue Askew, Rona Sturrock, Fred Peet, Lesley Manning, Hadrian Merler, Liyousa Shafei, Bob Johnson, Tom Johnston, Trevor Ryan, and David Jackson. I also wish to thank E d Setliff at Lakehead University for exposing me to the field of forest pathology and the fascinating fungus which has become the focus o f my study. I wish to thank my long-time friends Tania, Sarah and Laura who have remained extremely supportive despite my abandoning the circle in London, and particularly in the midst of the many other important events which unfolded this past year. Thanks for all the emotional support, camaraderie, and caring. I wish to thank my family and extended family for providing a loving environment for me. M y sister and brother-in-law were particularly supportive and understanding of the pressures of academia and I 'm extremely grateful for all the 'perspective' pep-talks. Lastly, and most importantly, I wish to thank my parents, Ruth and A l f Cleary whose foresight and values paved the way for an advanced education. They have supported and loved me and have never doubted my ability to succeed. A n d to my husband Jason, who witnessed first hand the hardships and successes that this academic journey has brought to both our lives. I thank him for his patience, particularly during the several months o f separation each year while carrying out this research, his assistance with several aspects o f the field work, his continuous encouragement, understanding, support, and love. 1 CHAPTER ONE: I N T R O D U C T I O N A N D G E N E R A L L I T E R A T U R E R E V I E W 1.1 G E N E R A L I N T R O D U C T I O N Species o f Armillaria causing root diseases o f trees, shrubs and some non-woody perennials occur in forested areas worldwide. Armillaria ostoyae (Romangnesi) Herink, is the primary pathogen causing root disease o f conifers in the northern hemisphere (Morrison 1981, Guillaumin et al. 1993, Klein-Gebbinck et al. 1993, Berube and Dessureault 1998, McLaughl in 2001). In British Columbia (B.C.) , A. ostoyae is a natural component o f undisturbed forests and is most abundant and frequent in the mid-elevation forests in the Interior-Cedar-Hemlock (ICH) biogeoclimatic zone (Lloyd et al. 1990) in the southern Interior o f B . C . Within this zone, A. ostoyae causes considerable losses in immature stands by ki l l ing natural and planted coniferous trees, causing growth loss on trees sustaining non-lethal infection, and predisposing trees to attack by other pathogens and insects. It can also act as a secondary parasite and the immediate cause of death o f trees infected by root diseases such as Phellinus sulphurascens Pilat (syn. P. weirii) and Inonotus tomentosus (Fr:Fr) S. Teng. or weakened by other pathogens or stress factors. Annual losses in timber are estimated to be more than half a mi l l ion cubic metres as a result o f mortality, growth repression and regeneration delay (Taylor 1986). Armil lar ia root disease is particularly problematic in new plantations where the fungus is carried over to the next rotation in colonized stump and root systems. Cumulative mortality in juvenile stands can be as much as 20% by age 20 years (Morrison & Pellow 1994) and numerous small disease centres may coalesce to form unstocked or understocked openings in the stand. Few options are available to mitigate potential losses due to ArmiHaria root disease. Mechanical removal of stumps is a very effective means of reducing the amount of woody inoculum that would otherwise be available to the fungus and minimizing the extent of Armillaria-caused mortality in the regenerating stand. Another less intrusive 2 option is to plant coniferous species that have a low susceptibility to ki l l ing by A. ostoyae. The use of resistant species in Armillaria-infested stands is economically practicable so long as the resistant hosts are a high-value merchantable species and well adapted to the particular site conditions. There are no woody hosts that show complete immunity to Armillaria. It is generally accepted that all conifers less than 15-years o f age are highly susceptible to ki l l ing by Armil lar ia root disease (Morrison et al. 1992). However, with the exception of Douglas-fir {Pseudotsuga menziesii (Mirb.) Franco) and western larch (Larix occidentalis Nutt.) (Robinson and Morrison 2001), very little is known about the relative susceptibility o f conifers in B . C . Western redcedar (Thuja plicata Donn ex D . Don) and western hemlock (Tsuga heterophylla (Raf.) Sarg.) are two conifers that regenerate naturally in the I C H zone, often comprising a significant component as serai stands develop and becoming the climax species in the mature overstory (Lloyd et al. 1990), However, the mechanisms by which both species are able to tolerate Armillaria and survive in the presence o f inoculum are not well understood. Traditionally, western hemlock and western redcedar are not planted, with the possible exception o f the latter in wetter microsites within a cutblock. Both species hold high value in terms of usage in the forest industry. Western redcedar accounts for approximately 7% of the total lumber production in B . C . , while hemlocklAbies spp. account for more than 12% (Council o f B . C . Forest Industries 2000). Western redcedar heartwood in particular contains several potent, antimicrobial toxins and is well known for its natural durability and decay resistance for exterior uses o f timber which make it a desirable species for high-yield forest management. Other species also contain various toxins in the heartwood. However, such chemical barriers in the heartwood would not likely influence infection and spread by Armillaria in the bark, vascular cambium and outer sapwood tissue. There is a paucity o f information concerning the ability o f certain conifers, including western redcedar and western hemlock, to activate defense mechanisms in those tissues that would enable them to respond to and halt infection by A. ostoyae. Knowledge of precisely which mechanisms are effective against the fungus and other host factors that might influence the expression 3 o f resistance would provide the necessary information to support the use o f either species for regeneration on sites where Armillaria may be o f concern. The aim o f this study was to compare the host response to infection by A. ostoyae in the roots o f western redcedar and western hemlock, at the macroscopic and microscopic level, specifically targeting the anatomical changes in cells and tissues that lead to necrophylactic periderm formation in the inner bark and barrier zone formation associated with the compartmentalization of infected woody tissue. Douglas-fir was included in this study as a susceptible control, providing a benchmark to which to compare host reactions of the other species. It was also the intent of this study to determine whether the pattern of basal lesion development and mortality in mixed stands of the three species was consistent with the efficiency o f natural resistance mechanisms operating in the roots o f these species. 1.2 ARMILLARIA T A X O M O N Y A N D SPECIES I D E N T I F I C A T I O N In early literature, there was much confusion surrounding the nature o f host-Armillaria interactions because the pathogen used to be considered a single variable or polymorphic species, Armillaria mellea (Vahl ex Fr.) Kummer (Watling 1982). Hintikka (1973) demonstrated that Armil lar ia has a bifactorial, heterothallic mating system in which two haploid monokaryons with compatible mating-type alleles would anastomose to form a secondary diploid mycelium. His work enabled subsequent studies to be undertaken that conclusively differentiated "biological species" o f Armillaria in both Europe and North America (Korhonen 1978, Anderson and Ul l r i ch 1979). Anderson et al. (1980) subsequently observed that certain species from Europe were partially compatible with those from North America. Since the division of Armillaria mellea sensu lato into a number of distinct biological species, a number o f studies of pathogenicity and virulence on species of Armil lar ia have been carried out which have helped explain much o f the variation in host-pathogen interactions and damage described for different host species 4 worldwide. Watling et al. (1991) suggests the genus Armillaria contains about 40 species. Somatic incompatibility mating tests are commonly performed using haploid tester strains (monospore isolates). Single spore cultures from Armillaria mushrooms typically have white or light brown aerial mycelium whereas cultures from basidiocarps, rhizomorphs and mycelial fans generally appear to be dark, flat, crustose and lacking aerial mycelium (Hintikka 1973). The pairing reaction o f single spore isolates follow a tetrapolar pattern which suggests nuclei o f crustose mycelia are diploid rather than dikaryotic (Hintikka 1973). Two diploid genets of the same species paired in a Petri dish w i l l form a nearly hyphae-free gap between opposing mycelia, whereas the same genotypes w i l l show complete intermingling or fusion o f opposing mycelia to form a single homogenous colony (compatible reaction). Failure o f mycelia to grow together indicates dissimilar genotypes (incompatible reaction) and a dark demarcation line w i l l develop in the growth medium between the two opposing mycelial colonies (Korhonen 1978, Anderson and Ul l r i ch 1979, Mallet et al. 1989). Vegetative mycelium of Armillaria species can sometimes discolour the media dark brown which makes judging reactions of opposing mycelia difficult. Furthermore, Armillaria isolated from lesions is frequently cultured on malt extract agar amended with Benomyl, in order to inhibit common contaminants (Worrall 1991). However, this can sometimes induce somatic haploidization of diploid mycelia, resulting in fluffy segregants of mycelial cultures and this can further increase difficulties with species identification (Guillaumin et al. 1991). More recently, molecular techniques using polymerase chain reaction (PCR) focussed on restriction site differences within the first and second ribosomal intergenic spacer region have proven to be a fast, accurate, and relatively inexpensive technique for confirming species identification (Harrington and Wingfield 1995; White et al. 1998). Random amplified polymorphic D N A ( R A P D ) analysis have been used to describe the population structure of Armillaria species in the Interior forests o f British Columbia (Dettman and van der Kamp 2001a; Dettman and van der Kamp 2001b). 5 1.3 ARMILLARIA SPECIES I N B R I T I S H C O L U M B I A Using interfertility tests of vegetative mycelia, Morrison et al. (1985) identified six intersterile groups of Armillaria in British Columbia originally designated A - F and later converted to the Roman numeral system for classifying intersterility groups as North American Biological Species ( N A B S ) according to the system used by Anderson and Ul l r i ch (1979). Armillaria ostoyae ( N A B S I), and Armillaria sinapina Berube & Dessureault ( N A B S V ) have widespread distribution throughout the coastal and interior forest regions o f B . C . The range o f A. ostoyae is from 49° to 53°N latitude and it is found primarily attacking conifer species but can also occur on hardwoods (Morrison et al. 1985; Morrison et al. 1992). The range of A. sinapina is similar to that o f A. ostoyae but extends farther north up to about 57°N latitude, and is primarily found on both dead conifer and hardwood trees and stumps (Morrison et al. 1985, Dettman and van der Kamp 2001a). Pathogenicity tests have shown that A. ostoyae is highly pathogenic (Rishbeth 1982, (Morrison & Pellow 2002). A. ostoyae rhizomorphs possess a dichotomous branching habit whereas those of A. sinapina exhibit monopodial-type branching. Morrison (2004) suggested that Armillaria species possessing dichotomous branching tend to be more pathogenic than those producing monopodial branching. Armillaria gallica Marxmuller & Romagn. ( N A B S VII) and Armillaria nabsnona V o l k & Burdsall ( N A B S IX) have been found on l iving and dead hardwoods in the Coastal-Douglas-Fir (CDF) biogeoclimatic zone in southwestern B . C . Armillaria cepistipes Velenovsky ( N A B S XI ) has been collected only twice from broadleaved hosts from two geographically very distant sites, Hope and Stewart, B . C . Another species, as yet unnamed and referred to as N A B S X , was collected from a relatively small area in southeastern B . C . on coniferous and hardwood hosts (Morrison et al. 1985). 6 1.4 I N O C U L U M A N D R H I Z O M O R P H S Armillaria spp. have a very wide host range, attacking many conifer and broad-leaved trees, shrubs, and herb species (Raabe 1962). Armil lar ia may survive for decades in colonized stump and root systems, depending on the size of the substrate. Although tree stumps occasionally become infected by basidiospores produced by the fruiting bodies of some Armillaria species (Rishbeth 1985, Legrand et al. 1996, Hood et. al 2002), spore infection by A. ostoyae is thought to be a very rare event because genets tend to occupy large areas (Anderson et al. 1979, Ferguson et al. 2003). For this reason, the natural infection pathway(s) o f A. ostoyae have not been identified. Vegetative growth is the primary mode o f spread at root contacts or through the soil v ia rhizomorphs. Rhizomorphs are highly differentiated vegetative organs that enable the fungus to grow out o f a suitable nutrition base and through the soil in search of new substrates to colonize, thereby increasing the infective potential o f the pathogen (Garraway et al. 1991). Inoculum potential is defined as "the energy o f growth of a parasite available for infection at the surface o f the host organ to be infected" (Garrett 1970), and is influenced by the size o f the inoculum source and species, time since colonization, the distance between the inoculum and the host and any environmental factors affecting growth o f the fungus (Redfern and Fi l ip 1991). Inoculum potential is maximized where healthy roots and inoculum (stumps) are in contact. Where infection occurs primarily at root contact, the distance from the inoculum source becomes less important in estimating inoculum potential than the size o f the inoculum itself. Conversely, when infection occurs primarily by rhizomorphs, the inoculum potential o f the fungus decreases with increasing distance between the inoculum source and the host (Redfern and Fi l ip 1991). Hence, when large gaps are bridged by rhizomorphs, such great distances may affect the infective capacity o f the fungus and consequently the host response to infection at the point o f penetration on the root. Although rhizomorphs are largely dependent on a supply of nutrients translocated from the food base to their growing tips, they are also able to absorb ions and oxygen through 7 the unpigmented, apical region of the growing tip. Morrison (1975) found that ammonium ion uptake was not translocated to the food source suggesting that nutrient supplies available to rhizomorphs from a food base may also be supplemented by uptake from the soil. The histology of rhizomorphs was studied in detail by Schmid and Liese (1970) and Motta (1969). Their structure consists of well-defined cellular regions that include a melanized and densely packed outer cortex that surrounds a subcortical layer of fungal cells and the medulla which consists of a loose mesh of wide-diameter hyphae that eventually form a central canal (Garraway et al. 1991). In older rhizomorphs, hyphae with thicker cell walls f i l l in the central canal o f the medulla. Schmid and Liese (1970) suggest that the central canal of the medulla is not hollow, but rather comprised of fungal hyphae that are loosely arranged. The primary meristem is located in the apical region o f the rhizomorph which is well-endowed with dense cytoplasm, protein and nucleic acids (Motta 1969). Secondary meristems associated with lateral growth are located in the subcortical layer distal to the apical center (Motta 1969). The medulla region is responsible for the transport of water and nutrients (Jennings 1984) and oxygen (Smith and Griffin 1971). A gelatinous sheet and a mucilaginous layer surrounds the rhizomorph at its apex and its melanized rind helps protect the fungus against dessication or damaging external agents in the soil, or both (Garraway et al. 1991). Early mycologists in the mid-nineteenth century like Schmitz (1848), as cited in Garraway et al. (1991), did not associate rhizomorphs found beneath the bark and in the soil with Armil lar ia root disease causing mortality on coniferous and broadleaved hosts. First described as Rhizomorphafragilis Roth., the fungus was later divided into two subforms, R. subterranean to describe the melanized strands o f hyphae growing through the soil and R. subcorticalis to describe the form that produces white mycelial fans in the bark and cambial tissue (Garraway et al. 1991). Hartig (1874) later resolved this issue by confirming the association o f rhizomorphs in the soil with that o f the Honey Fungus Agaricus melleus L . , now known as Armillaria. 8 1.5 INFECTION BIOLOGY The external tissues of woody plants, comprised of periderm and rhytidome, provide an important protection barrier to underlying tissues from microbial invasion. However, it offers little resistance to penetration by Armillaria species. The infection process by rhizomorphs has been described in some detail for both conifers and hardwoods (Day 1927, Thomas 1934, Woeste 1956, Rykowski 1975, 1980). Rhizomorphs may become firmly attached to the outer bark at several points along the surface of the root by hardening of the mucilaginous substance surrounding the growing tip (Day 1927, Thomas 1934). Next, a lateral branch or mycelial wedge, still connected to the rhizomorph, penetrates the rhytidome as a single unit rather than individual hyphae (Thomas 1934). The ability of rhizomorphs to penetrate the rhytidome and inner bark depends on the enzymatic capacity of the fungus (Rykowski 1975). The fungus is capable of producing suberinase, a suberin-degrading enzyme, to facilitate the destruction and dissolution of suberized walls in the outer phellem layers (Swift 1965, Zimmermann and Seemuller 1984). Penetration of the outer bark is similar among susceptible and resistant hosts and involves both mechanical and enzymatic processes (Rykowski 1975, 1980). However, the ability of some Armillaria species to penetrate host tissue, including those that are typically 'weak parasites', may also depend on the inoculum potential of the fungus. Once the fungus has penetrated the periderm, it causes necrosis of the living bark and cambial tissue. Enzymatic degradation of cells is caused by extracellular, phenol-oxidizing enzymes produced by the fungus such as peroxidase, tyrosinase and laccase (Garraway et al. 1991). Browning of phloem parenchyma cells occurs ahead of mycelial colonization. Such tissue typically contains high levels of phenol oxidases compared with adjacent healthy tissue (Wargo 1984). Cell hypertrophy was observed in phloem tissue located immediately underneath a penetrating rhizomorph of A. mellea sensu lato (Day 1927). In almost all vigorous hosts, one or more periderms were formed in an attempt to contain the infection in the bark. However, at times no periderms were formed and the fungus appeared to be advancing freely in the host tissue without any host reaction (Day 1927). Thomas (1934) reported 9 that in black walnut (Juglans nigra L.), which he considered to be moderately resistant to attack by Armillaria, a high proportion of infections was apparently halted and walled off by periderm tissue. Day (1927) and Thomas (1934) noted that in susceptible hosts, the fungus spreads rapidly in the bark and cambial invasion often resulted in girdling of the root and the advance of the fungus towards the root collar. In resistant roots, cankers often formed around sites of penetration and the spread of the fungus was checked by the formation of callus at the edge of the lesion (Thomas 1934, Rykowski 1975). 1.6 HOST RESPONSE TO INFECTION Rishbeth (1985) stated that the most important host responses occur after the bark has been infected by Armillaria mellea sensu stricto. On plum rootstocks, penetration of A. mellea (sensu stricto) was similar for both susceptible and resistant rootstocks and resistance was mainly due to post-infection reactions (Guillaumin et al. 1989). Day (1927) observed A. mellea sensu lato to be contained by a new cork layer that was formed deeper in the cortex following invasion in the roots of a Corsican pine (Pinus nigra ssp. laricio Maire) tree. Mullick (1977) reported that host reactions in the living bark induced in response to initial penetration by a pathogen will determine the outcome of a particular reaction as being either susceptible or resistant. Host responses in tissues affected by Armillaria root disease fall into four categories: exudate production, meristematic activity, biochemical interaction (Morrison et al. 1991), and compartmentalization (Shigo andTippett 1981). 1.6.1 E X U D A T E PRODUCTION Conifer hosts will exude resin in response to mechanical injury, pathogenic invasion, or both (Buckland 1953, Rishbeth 1972, Redfern 1978). Increased resin synthesis in the roots of Corsican pine was considered to be an important factor in resistance of pines against further ingress of Heterobasidion annosum (Fr.) Bref (Prior 1975, 1976). Rishbeth (1972) reported that the volatile turpentine component of resin produced by Scots pine (Pinus sylvestris L.) partially inhibited growth of A. mellea sensu lato in 10 culture. Rykowski (1975) also observed that resin produced by conifers in wood and callus limited further spread of A. mellea (= A. ostoyae). Vigorous resin flow in roots may become toxic to the fungus and/or play some role in terms of forcing infected tissue away from the root wood (Rishbeth 1982, 1985). Prior (1975) suggested that the main effect of oleoresin on the fungus is by physical blockage of the tracheids, rather than direct toxicity. Robinson (1997) observed that copious resin production formed in the roots of Douglas-fir and western larch help facilitate sloughing of infected tissue away from the root surface. Broadleaved hosts infected with Armillaria do not produce resin but instead, exude a sticky plant exudate called gum which is deposited in xylem vessels (Thomas 1934, Rishbeth 1985). Eucalyptus trees showing resistance against Armillaria luteobubalina Watling & K i l e often exhibited kino exudation and callusing around the base of infected trees (Podger et al. 1978, K i l e 1980, Shearer and Tippett 1988). 1.6.2. M E R I S T E M A T I C A C T I V I T Y Meristematic activity involves the renewal of tissues affected (killed) by the fungus. A t least three categories can be distinguished, namely phellogen renewal and the formation of a secondary periderm in the bark (Day 1927, Thomas 1934, Mul l i ck and Jensen 1973), the formation of callus tissue and new vascular cambium (Sharpies and Gunnery 1933, Shigo and Tippett 1981, Dujesiefken et al. 2001), and adventitious rooting (Rishbeth 1972, 1985). The regeneration of these tissues following invasion by the fungus serves to renew and replace lateral meristems, prevent desiccation and maintain gas exchange, and restrict the ingress o f plant pathogens. Early investigations by Thomas (1934) give complete and detailed descriptions o f resistance reactions in the bark following penetration by A. mellea sensu lato. In resistant hosts, a layer of cork was often formed which restricted further spread of the fungus in the root while in susceptible hosts, the fungus often breached new cork layers (Thomas 1934). 11 Mullick (1977) conducted a series of developmental studies on wound healing in bark and was the first to term this new periderm barrier as 'necrophylactic periderm' (Mullick and Jensen 1973). He considered the process of necrophylactic periderm formation to be non-specific, induced in response to either biotic or abiotic injury and/or simply age, and triggered whenever the phellogen is rendered non-functional (Mullick 1977). Mullick's (1977) model for non-specific wound healing in bark is presented in Figure 1.1. NIT NIT ep NIT NP-VCl iK VCfn), mmmmfi NON CONDUCTING SAPWOOD' NP - NECROPHYLACTIC PERIOERM ep - EXOPHYLACTIC PERIDERM VC(n) - POSITION OF VASCULAR CAMBIUM AT TIME OF PHELLOGEN RESTORATION VC(i)- POSITION OF VASCULAR CAMBIUM AT TIME OF INJURY ggggj - CONDUCTING SAPWOOD FORMED AFTER INJURY - CONDUCTING SAPWOOD EXTANT AT TIME OF INJURY - TRANSFORMED CAMBIAL-PHLOIC ZONE D%Sg| - ZONE OF NEWLY RESTORED VC I'M"') - ZONE OF NON - CONDUCTING SAPWOOD Fig. 1.1. Diagrammatic views of the anatomical model for non-specific defense mechanisms that follow (a) penetration of the bark surface, (b) penetration of the vascular cambium, and (c) penetration of the sapwood (From Mullick 1977). The first process shown in Figure 1.1 (a) is that of phellogen restoration which is triggered whenever the phellogen becomes non-functional (Mullick 1977). The necrophylactic periderm (NP) that establishes is preceded by a layer of non-suberized impervious tissue (NIT). Periderm formation in response to injury or infection and various stages of development involved in phellogen restoration is discussed further in Chapter 2. Shallow injury to the preexisting periderm and the first few layers of living bark cells trigger only the process of phellogen restoration. 12 In Figure 1.1 (b), Mullick (1977) illustrates vascular cambium restoration when penetration or injury reaches a certain depth of the bark. Thus, deep injuries to the bark may cause a disruption in the normal cambial activity without any direct injury to the cambium itself. The vascular cambium directly underneath the injury stops producing normal xylem derivatives while xylary and phloic derivatives continue to be produced on either side of the injured tissue (Mullick 1977). Mullick (1977) describes the deranged vascular cambium as a modified cambial-phloic zone which is characterized by enlarged cells. Presumably, this modified cambial-phloic zone is callus (isodiametric, undifferentiated tissue). Subsequently, xylem-like cells appear sporadically at the external boundary of the transformed cambial-phloic zone and a new vascular cambium develops between the sporadic xylem-like cells and the enlarge phloic cells. The newly restored vascular cambium will continue to produce normal xylem and phloem derivatives (Mullick 1977). NIT and NP are essential to completion of the second process and a new vascular cambium is restored (Mullick 1977). Figure 1.1 (c) illustrates blockage of sapwood conduction following even deeper injuries to the bark which Mullick (1977) characterizes as non-conductive discoloured wood. Air is prevented from spreading through the xylem proximal and distal to the point of injury because of tracheid occlusion. This blockage or barrier that forms is analgous with Wall 1 of the CODIT model (Shigo and Marx 1977). In addition to NP formation in the bark, the living vascular cambium at the periphery of the wound forms callus. A new phellogen differentiates at the outer part of the callus tissue and meristematic tissue forms as a replacement of the vascular cambium in the inner part of the callus (Mullick 1977). Several studies have since shown necrophylactic periderm to be involved in resistant reactions in woody plants in response to mechanical injury or pathogenic invasion (Blanchette and Biggs 1992, Wahlstrdm and Johansson 1992, Robinson and Morrison 2001)! The formation of new roots, including adventitious roots, is important in the process of wound healing largely because the survival of a tree depends, to some extent, upon the 13 balance between killing and regeneration of roots (Rishbeth 1985). Adventitious rooting occurs primarily in situations where a root has been girdled to restore the critical functions of absorption through the initiation of new root tips. The production of callus tissue and the regeneration of a new vascular cambium within the callus have been described for several hosts following direct injury to the cambium or the xylem (Sharpies and Gunnery 1933, Warren-Wilson and Warren-Wilson 1960, Oven and Torelli 1999, Dujesiefken et al. 2001). Many of the tissues involved in callus formation are also involved in barrier zone formation during the process of compartmentalization and will be discussed in more detail in the following section. The contribution of different tissues to the formation of callus differs between gymnosperms and angiosperms. In angiosperms, callus originates mainly by the living vascular ray cells in addition to the parenchyma of the phloem and xylem (Sharpies and Gunnery 1933). Novitskaya (1998) showed that in European white birch (Betula pendula Roth) callus originated from proliferation of xylem mother cells on the exposed surface of the wood. Dujesiefken et al. (2001) also demonstrated that in wounded hardwoods callus can form from tissue within the wound area, but only when the xylem remains undamaged. Oven and Torelli (1999) described a similar sequence of histological changes in the cambial zone of several different conifer hosts following wounding. However, the authors noted callus formation to occur as a result of progressive hypertrophy and hyperplasia of cambial cells as well as phloem parenchyma cells. Increased frequency of transverse divisions was observed tangentially in the fusiform cells of the cambial zone (Oven and Torelli 1999). The rate of callus production following wounding varies by tree species (Gallagher and Sydnor 1983, Oven and Torelli 1999), but may also vary depending on the size of the wound, the season of wounding, the location of the wound on the tree (Wensley 1966), and tree vigour (Buckland 1953). van der Kamp and Hood (2002) showed that Armillaria species can invade and girdle trees near the root collar, killing the vascular cambium and phloem in a single dormant season when the tree is unable to compartmentalize the infection. Similarly, Robinson et al. (2004a) showed that roots 14 were incapable o f initiating such defense responses immediately following wounding during the winter. Buckland (1953) observed Douglas-fir trees exhibiting good vigor frequently checked the advance o f A. mellea (= A. ostoyae) by laying down a resin barrier and callus around the infected tissue, forming a latent or dormant canker. Johnson et al. (1972) suggested that the ability of Douglas-fir infected by A. mellea (=A. ostoyae) to form callus at lesions increases between the age o f 5 and 20 years. The ability o f trees to form callus w i l l prevent mortality of the host tree (Johnson et al. 1972). Vigorous trees typically show more resinosus and callus formation around lesions than less vigorous trees. K i l e (1981) reported that inverted V-shaped lesions with callus ingrowth at the margin on several Eucalypts hosts prevented the girdling of stems by A. luteobubalina. 1.6.3. B I O C H E M I C A L D E F E N S E In as much as the physical, structural barriers may slow the infection process on roots, the underlying mechanisms must also involve chemical changes in host tissue in the form o f mobilized constituents in response to invasion by the fungus (Garraway et al. 1991). In his early investigations o f the infection biology of A. mellea, Thomas (1934) suggested that newly formed periderms confine the fungus to a lesion in the bark only after other means of resistance, presumably chemical, halted the fungus. Many trees produce a variety of inhibitory and toxic compounds in the bark and in the wood which endow them with various degrees of resistance against an invading pathogen (Asiegbu et al. 1998, Biggs et al. 1984, Lindberg 1991, Wargo 1984, Woodward and Pearce 1988). Phenolic compounds play an important role in both constitutive defenses, those that occur prior to fungal invasion, and inducible defenses following initial invasion by the fungus. Some host enzymes can detoxify pathogen toxins enabling the host to ward off potential infections. Moreover, some compounds which individually do not offer much disease resistance may act synergistically to provide resistance against an invading pathogen (Kozlowski 1969, Lindberg et al. 1992). Vance and Garraway (1973) found that phenols that accumulate in Armillaria thalli can inhibit mycelial growth and rhizomorph development in vitro. Phenols may also inhibit enzymes which could be 15 important in mycelial growth and rhizomorph morphogenesis (Vance and Garraway 1973). Phlobaphenes are pigmented phenolics that typically accumulate in the phellem of periderm tissue and consist of mixtures o f several distinct compounds (Mul l ick 1977). Using thin-layer chromatography, Mul l i ck (1969a) concluded that the pigments belonged to a new class o f non-anthocyanic reddish-purple compounds, although a minor portion of the pigments consisted of anthocyanidin occurring as such in nature (Mul l ick 1969b, 1969c, 1969d). Thuja plicata was reported to have at least eight distinct pigments and Tsuga heterophylla at least nine; the nature of pigments varied between species (Mul l ick 1969a), although cyanidin and pelargonidin were common to both (Mul l ick 1969b). Guillaumin et al. (1989) reported growth of A. mellea in the bark and sapwood of Prunus species was considerably less in resistant rootstocks. This slow growth was associated with pink or purple discoloration o f the tissue surrounding the lesion. The authors attributed this resistance to de novo synthesis of anthocyanic compounds (Guillaumin et al. 1989). This type o f active defense mechanism may inhibit the spread o f Armillaria long enough so that necrophylactic periderm formation can be completed. Phellem cells of Picea abies (L.) Karst. root bark infected with Heterobasidion annosum, contained granular material (most likely phenols) that varied from sparse to dense masses filling the whole cell (Heneen et al. 1994b). Hyphae penetrating phellem cells were sometimes seen embedded in dense granular phenolic compounds that entrapped and immobilized penetrating hyphae suggesting a possible toxic effect on the fungus (Heneen et al. 1994a, 1994b). Similar observations were recorded for Scots pine infected with A. ostoyae, where termination and dissolution o f single hyphae caused by fungitoxic compounds or lytic enzymes occurred as a post-infection host response (Wahlstrom and Johansson 1992). The inoculation technique described by Wahlstrom and Johansson (1992) involved wounding of the phellogen followed by inoculation with a small strip of infected bark directly into the wound which would certainly have facilitated penetration in which single hyphae may been observed in tissue samples. However, observations of natural infection by A. ostoyae on conifers in British Columbia suggests the phenomenon 16 of penetration by single hyphae does not occur. Instead, the fungus penetrates as a larger unit (i.e. rhizomorph) similar to that described by Day (1927), Thomas (1934), and Rykowski (1980). Western redcedar bark contains many complex substances including tannins, phlobaphenes, vanill in, catechin and fatty acids (Barton and MacDonald 1971). Chemical analysis o f the bark showed that it contains 31% lignin, not all that different from the lignin content o f the wood, and the extracts are quite acidic (Gonzalez 1997). Lignins are generally resistant to microbial attack. Increased lignification o f cells may increase resistance to compressive forces because cell walls are more resistant to mechanical penetration (Asigbu et al. 1998). Lignification may also restrict diffusion of enzymes and toxins from pathogen to host and can equally restrict availability o f water and nutrients from host to fungus (Asigbu et al. 1998). Defense reactions involving increased synthesis o f proteins and enzymes may disrupt fungal growth by causing cell lysis o f hyphal tips. Wargo (1975) found chitinase and (3-1,3-glucanase in the phloem of Acer saccharum Marsh, and several Quercus spp. and suggested that these enzymes could lyse hyphal walls of A. mellea. Robinson et al. (2000) also found higher concentrations of an endochitinase-like protein in the phloem of Douglas-fir trees infected with A. ostoyae and P. weirii (= P. sulphurascens). The presence of such enzymes, induced following pathogenic invasion, may act to inhibit fungal growth and may be part o f a multi-stage system of defense. Pathogens, after triggering the process o f phellogen renewal, interact not only with chemicals present in the normal bark with also with new chemicals produced during the course of cellular dedifferentiation (Mull ick 1977). 1.6.4. C O M P A R T M E N T A L I Z A T I O N O F D E C A Y I N T R E E S (CODIT) The C O D I T model describes the defense response of trees following injury to cambium or sapwood, or both, that results in the formation of boundaries which resist the spread of decay within the xylem. Hence, compartmentalization, which occurs in both stems and roots, w i l l conserve the mechanical support system of the tree (Shigo and Marx 1977). 17 The C O D I T model has two parts. Part I consists of 3 walls that surround and compartmentalize infected tissue in all planes. The formation o f each boundary (wall) involves chemical changes in the l iving parenchyma cells of the sapwood. Wal l 1 forms the upper and lower transverse boundaries of the compartment and serves to protect the water-conducting tracheids above and below the injured area from invading air (Shigo 1984). This is achieved by tracheid occlusion with deposits o f polyphenols and border pit aspiration. Wa l l 2 lies in a tangential plane and resists inward (radial) spread by utilizing the last cells of each annual growth ring. These cells are usually thick-walled and heavily lignified. Wal l 3 resists tangential spread and utilizes the xylem ray cells. Walls 1 to 3 represent the path o f least resistance, with Wal l 1 being the weakest and Wal l 3 being moderately strong (Shigo and Marx 1977, Shigo 1984). The strength of Wal l 2 depends on the resistance o f the latewood tracheids, particularly the thickness of cell walls and the degree o f lignification. The relative holding strength of each wall to fungal colonization is variable and may depend to a large extent on the host species and the overall vigour of the tree. Part II of the C O D I T model consists of a fourth wal l , also known as the "barrier zone" (Tippett and Shigo 1980, 1981). A barrier zone is comprised o f a unique layer o f cells formed by the uninjured cambium that restricts the fungus to xylem formed prior to injury (Shigo and Tippett 1981). The barrier zone in some conifers may be comprised largely o f distorted tracheids, increased number o f axial parenchyma containing polyphenols or traumatic resin ducts or both (Tippett and Shigo 1980,1981). Traumatic resin ducts may be formed around a wound in response to injury in all conifers, including those that do not form resin ducts in normal xylem. Such tissues comprising the barrier zone were effective in restricting A. mellea to wood formed prior to infection (Tippett and Shigo 1981, Shigo and Tippett 1981). The barrier zone is the strongest of all four walls and results in the most effective compartmentalization. Barrier zones may be quite large, encircling the entire stem or extending for only a limited distance tangentially and longitudinally from the wounded area. The ability of barrier zone to resist spread of a pathogen w i l l differ among and 18 within a host species and depends to some extent on the physiological activity of parenchyma cells and the availability and mobilization of stored material (Liese and Dujesiefken 1996). Hence, vigorous trees may react and compartmentalize quickly and more efficiently than trees with low vigour. The position of the barrier zone in the wood w i l l vary depending on the time o f year in which the vascular cambium was injured. Since phellogen and vascular cambium activity ceases in the winter as a result of tree dormancy, host reactions involving barrier zone formation would not be expected to take place until the tree resumes growth in the spring. Bannan (1936) showed that following injuries to stems of Tsuga canadensis (L.) Carr. during tree dormancy, the vascular cambium showed no response until the beginning of the next growing season and the newly formed cells produced at the beginning o f the annual ring were traumatic resin canals. Shigo and Tippett (1981) also suggested that the position o f the barrier zone in the earliest portion o f the growth ring indicates that spread o f A. mellea occurred during a period o f dormancy or soon after the onset of growth. Once formed, the barrier zone comprising wall 4 of the C O D I T model remains stationary whereas walls 1, 2, and 3 may adjust to an expanding column of decay. Liese and Dujesiefken (1996) suggest that a barrier zone should not be restricted to the small layer of cells formed immediately after wounding. In conifer roots, the natural alignment of the wood promotes more spread o f fungi in the axial direction through the cell lumina of tracheids. A s a result, wal l boundaries may extend as decay spreads either distal or proximal from its original point o f infection. Tippett et al. (1982) suggested that barrier zones may also actively extend as pathogens spread from a site o f initial infection on a root up towards the root collar. This was evidenced by the presence o f concentric tangential bands o f traumatic resin ducts produced in successive growth rings (Tippett et al. 1982). 19 1.7. H O S T S U S C E P T I B I L I T Y The extent of damage on host species depends not only on the host (relative susceptibility, the size, age, and vigour o f the tree), but also on the pathogen (pathogenicity and virulence o f the species/isolate and inoculum potential), and the influence of the environment (soil moisture, stress factors such as drought, and the season of injury). The relative resistance and susceptibility o f several conifers to ki l l ing by A. ostoyae in British Columbia has been compared (Morrison 1981, Morrison et al. 1992). Inoculation experiments on juvenile conifers in the northwestern United States (Entry et al. 1992) and observations on naturally infected roots o f saplings, juvenile and mature trees from B . C . (Robinson and Morrison 2001, Morrison 2000, Robinson 1997, Morrison et al. 1992) show Douglas-fir to be highly susceptible to ki l l ing by A. ostoyae. Reports on host susceptibility for other conifers including western redcedar and western hemlock are few (Morrison et al. 2000, Koenigs 1969), although Morrison et al. (1992) considered western redcedar and western hemlock to have similar susceptibility to ki l l ing by A. ostoyae. Indirect evidence from previous studies (Morrison et al. 2000, DeLong 1997) suggests distinct differences in host susceptibility, particularly where western redcedar is concerned. Clearly, more information is needed to elucidate the relative susceptibility o f these conifers to ki l l ing by A. ostoyae. In British Columbia, information pertaining to host resistance in juvenile conifers to Armil lar ia root disease is generally restricted to Douglas-fir and western larch (Robinson 1997, Robinson and Morrison 2001). Host response, and the type of damage resulting from infection by A. ostoyae varies by host species and also with tree age (Robinson and Morrison 2001). Younger trees are generally kil led more quickly than older trees (Morrison et al. 1992). Tree mortality in the new plantations usually begins about 5-7 years after stand establishment and mortality can continue throughout a rotation with mortality rates o f approximately 1-2% per annum (Morrison and Pellow 1994.) A decline in the initial flush o f mortality w i l l be observed usually starting around the age of 15 years which is usually attributed to increasing host resistance and a decline in 20 inoculum potential in stumps and roots left over from the previous stand. Early expression o f defense responses and the ability o f host species to successfully contain the infection is a characteristic feature of trees that show greater resistance against the fungus. Rykowski (1980) characterized susceptible roots as those in which the host cells become brown and filled with a mass o f grainy structures as a result o f pathological changes in the l iving protoplasm while under the influence of a penetrating mycelium. Susceptible roots did not frequently initiate defense reactions and necrosis often advanced to the cambium and xylem tissue (Rykowski 1980). On the other hand, resistant roots generally underwent cytological changes in the affected tissue to form a demarcation zone in the form o f a secondary cork layer and successfully isolated mycelium from adjacent healthy tissue (Rykowski 1980). The author also noted sloughing of infected root tissue from the surface o f the root following successful periderm formation. Following invasion o f the cambium, resistant roots frequently formed a resin barrier and callus around the margin of the lesion (Rykowski 1980). The location o f an infection on the root system is an important factor in host resistance and disease development (Morrison et al. 1991). Morrison (1972) found that on Corsican pine, inoculations with isolates o f A. mellea sensu lato at the root collar resulted in larger lesions than inoculations on lateral roots. Two root collar inoculations resulted in larger lesions than a single inoculation (Morrison 1972). Shaw (1980) noticed that infections at the root collar and/or taproot usually killed the host more rapidly than infections at other sites. This phenomenon was also noted on young Douglas-fir (Buckland 1953). However, the type of host response induced in terms of bark reactions may also vary depending on the location o f the infection on the root. Robinson and Morrison (2001) found that the most effective necrophylactic periderm barriers were those that had multiple bands o f thick and thin-walled phellem and these were most frequently formed on older, larger diameter roots. 21 Western larch shows considerable resistance to the fungus, but only after it reaches about 20-25 years o f age (Robinson and Morrison 2001). Robinson (1997) noted that western larch was several times more successful at containing infections to a lesion on the root than Douglas-fir and this resistance was due in large part to the multiple bands of thick-and thin-walled phellem of periderms formed around infected bark tissue. However, in young regenerating stands in the southern Interior, mortality in western larch can be extremely high and may exceed that observed in Douglas-fir (Morrison et al. 1988, H . Merler, unpublished data). Fast growing, shade intolerant hosts, such as western larch may become infected earlier due to their more extensive root system and greater probability of contacting inoculum. In contrast, shade tolerant hosts such as western redcedar and western hemlock grow much more slowly in height and diameter as they are quickly overtopped by the more competitive, shade-intolerant serai species. Indirect evidence from several studies conducted in the southern Interior o f B . C . suggests that western redcedar may be more tolerant of A. ostoyae than other common conifers. Woods (1994) noted that in a juvenile mixed conifer stand of Douglas-fir, lodgepole pine {Pinus cortorta Dougl. ex Loud.), western hemlock and western redcedar, cedar was among the conifers least affected by A. ostoyae. In a conifer species trial near Hidden Lake, B . C . , H . Merler (unpublished data) also noted that other conifers like Douglas-fir and western larch were more severely affected than western redcedar. In a 22-year-old Douglas-fir plantation infested with A. ostoyae, B . J. van der Kamp (unpublished data) noted western redcedar mortality to be approximately one tenth that o f Douglas-fir. Morrison et al. (2000) reported a lower frequency of infection on cedar roots exposed to A. ostoyae compared to other conifers like western hemlock and Douglas-fir. Results from a 30+ year-old stumping trial showed mortality in western redcedar to be significantly lower than other planted conifers such as Douglas-fir and western larch (D.J. Morrison, unpublished data). Relationships have been found between host vigour and the relative susceptibility o f trees or stands to pathogen attack (Rosso and Hansen 1998). Host resistance is either directly or indirectly affected by the ability o f host trees to synthesize biochemical defenses and initiate structural defense responses at the point o f infection. Singh (1983) reported that 22 the roots of vigorous seedlings generally showed more resinosus and callus formation. The incidence and severity o f disease in conifer seedlings was greater in non-vigorous seedlings than in vigorous seedlings, with the former becoming infected much earlier (Singh 1983). Buckland (1953) suggested that i f host vigor was maintained, A. mellea (= A. ostoyae) would not be fatal to the host. Growth Efficiency (GE) has served as a measure o f host vigour by defining thresholds above which resistance to insects and disease is conferred (Mainwaring and Maguire 2004). Lieutier and Ferrell (1988) used G E to quantify an index of the last annual ring or the last five annual rings. G E has been known to have a strong correlation with the parameters associated with tree health and with the trees' ability to synthesize resources towards defense in the event o f an attack by some biotic factor (Lieutier and Ferrell 1988). Bos and Parlevliet (1995) defined susceptible hosts as the relative inability to impede attack by a parasite and resistant hosts have the ability to hinder growth and development (i.e. rates o f spread) within host tissue. Resistance in any particular host can be estimated by exposing a host species to a pathogen under uniform conditions and measuring disease severity. Day (1927) noted that cankers formed on roots infected with A. mellea sensu lato, a characteristic typical of resistant hosts, was commonly found on T. plicata and T. heterophylla and that many of these trees showed no signs of dying. In the study reported here, resistance was described by examining the frequency of successful resistance reactions induced following inoculation with pathogenic A. ostoyae and by comparing symptom development and species mortality rates on naturally infected individuals across several stands throughout the I C H zone in the southern Interior of B . C . 1.8 A R M I L L A R I A R O O T D I S E A S E I N T H E S O U T H E R N I N T E R I O R O F B R I T I S H C O L U M B I A . Armillaria is one of the most damaging disease agents affecting coniferous and deciduous trees and shrubs in both natural and artificial forested settings worldwide. Armillaria species may act as either primary pathogens parasitizing healthy, uninjured trees; as secondary pathogens invading trees that are stressed or weakened by other abiotic/biotic 23 means; or as saprophytes colonizing moribund woody substrates and utilizing them as a nutrient source for growth. Clearly, the ability of Armillaria species to assume multiple roles as they encounter hosts o f different species, age and vigour, and their ability to survive in stumps and roots during stand renewal events has contributed to their widespread presence and longevity in forests. A. ostoyae and A. sinapina are endemic throughout the southern interior o f British Columbia. The species co-exist in the same stand, generally occupy the same ecological niche and can commonly be found on the same host tree. A. ostoyae may assume the role of both a primary and/or secondary pathogen as well as a saprophyte, although its competitive ability as a saprophyte appears to be somewhat limited since it must colonize host substrates prior to tree death in order to survive as long-lived inoculum (Garrett 1970). In contrast, A. sinapina may be considered a weak pathogen competitor that at times assumes the role o f a secondary parasite attacking stressed trees, but typically colonizes hosts trees only after they have been kil led by A. ostoyae or by some other agent. A. sinapina is distinguishable from A. ostoyae in the field based on the morphology of their respective rhizomorphs. A. ostoyae typically has dichotomous branching while A. sinapina has monopodial branching (Morrison 2004). According to Morrison (2004), A. sinapina was better able to colonized moribund tissue than A. ostoyae. Both A. sinapina and A. ostoyae have a reduced capacity to colonize substrates after they have been invaded by other micro-organisms (Morrison 2004). Although it is generally considered a weak pathogen, A. sinapina is highly competitive and has evolved strategies for survival such as the ability to produce an extensive rhizomorph system that enables it to take advantage o f position over A. ostoyae during the exploitation o f new substrates. Redfern and F i l ip (1991) suggested that persistence o f the more pathogenic A. ostoyae in forest stands was largely due to the longevity o f inoculum in stumps, whereas the persistence o f weakly pathogenic species, such as A. sinapina, may be aided by the behaviour o f its extensive rhizomorph system and more frequent spore infection because as Dettman and van der Kamp (2001a) found, A. sinapina genets tend to be small and numerous. However, successful rhizomorph 24 penetration and colonization of woody substrates by weak pathogens such as A. sinapina may depend to a large extent on the inoculum potential of the fungus. A n extensive rhizomorph system suggests lower inoculum potential at distant tips. Based on field observations, Morrison et al. (1992) suggested that A sinapina was generally not capable of attacking healthy conifers. However, Dettman and van der Kamp (2001a) showed that A. sinapina was responsible for some conifer mortality in the central interior o f B . C . There is a paucity of information concerning the behaviour of A. sinapina, the mechanisms involved in the infection biology o f this species, and whether host reactions induced by less pathogenic species of Armillaria, such as A. sinapina differ from those induced following invasion by A. ostoyae. In the coastal forests of B . C . , A. ostoyae behaves quite differently compared to its range in interior forests. On the coast, A. ostoyae causes mortality in young plantations but mortality generally ceases after about 20-years o f age (Morrison 1981). Cruickshank et al. (1997) found that the bole volume and frequency o f callus formation at A. ostoyae root lesions on juvenile Douglas-fir was significantly greater for coastal than interior trees. In the interior, A. ostoyae may continue to k i l l trees throughout a rotation (Morrison 1981), however, mortality among interior Douglas-fir trees appears to be somewhat varied between the Interior Douglas-fir (IDF) biogeoclimatic zone and the Interior Cedar-Hemlock (ICH) zone depending on the extent o f disturbance and/or stress factors influencing the development o f disease on those sites. The I C H biogeoclimatic zone is the most productive zone in the interior forests of B . C . and supports the greatest diversity o f tree species. The I C H has warm, dry summers and cold, wet winters with mean annual precipitation levels between 500-1200 mm, with 25-50% of that precipitation occurring as snowmelt (Meidinger and Pojar 1991). The structure and composition o f forests located i n the I C H zone have been influenced by a number of disturbance agents including harvesting, fire, insect outbreaks and pathogens such as root disease. Hence, Armil lar ia root disease likely plays a very important role in these forests influencing stand structure and affecting species composition. In fact, the 25 impact of Armillaria within this zone is higher than in any other zone in B . C . , although the IDF, Montane Spruce (MS) , and Engelmann Spruce-Subalpine Fir (ESSF) zones may also experience significant loss. Most plantations located in the I C H w i l l have some level of A ostoyae damage. Morrison et al. (2001) found that up to 80% of trees in mature, undisturbed stands had Armillaria-caused lesions on their root systems and lacked aboveground symptoms of disease. The absence of disease symptoms in mature stands suggests that the host and fungus have reached an equilibrium whereby infections exist primarily as discrete lesions on their roots. Occasionally, the balance between this equilibrium is disrupted because of changing environmental or biological conditions which causes a fluctuation in disease levels in favour o f the pathogen. The precise triggers causing the change from quiescent to active disease centres are not well understood, but may be related to the gradual increase in the number of lesions on a root system over time. In addition, stress caused by periods o f drought or insect attack may play some role by reducing the overall vigor o f the tree and its ability to marshal enough energy resources to the point o f attack, quickly and efficiently. Following harvest, stumps and root systems remain alive for a year or two, surviving on stored carbohydrates during which time they w i l l become colonized by Armillaria emerging from latent or dormant root lesions. The ability o f the fungus to quickly colonize host substrates is an important characteristic o f the pathogen enabling it to survive many decades in large colonized trees and ultimately increases the inoculum potential of the site, thereby increasing the risk of infection in residual and regeneration trees. Forestry practices that create stumps wi l l increase the amount of inoculum on a site. Brushing of hardwoods increases the proportion of stumps that the fungus can use as a food base. Woods (1994) found that mortality incidence in an unbrushed plantation was less than in a brushed plantation. Pre-commercial or commercial tWnning (partial cutting) also increases inoculum in stumps. The probability o f infection by A. ostoyae increases with tree size (Morrison 2000). However, this increase is substantially more pronounced in stands that have been selectively cut than in undisturbed stands and is 26 greatest in the dry climatic region (Morrison et al. 2001). Mortality in the understory residuals begins about 10 years following selective cutting and in the overstory and regeneration after 15 years (Morrison et al. 2001). Multiple stand entries maintain high fungal energy by continuously supplying the fungus with suitable substrates to utilize as a food base. The predominance o f the fungus in such stands in all probability reduces site potential and productivity. 1.9 C O N C L U S I O N S Armillaria root disease is a serious forest health concern in the southern interior o f British Columbia. A. ostoyae poses a long-term threat to forest productivity and sustainable forest management because current silviculture practices increase the amount and potential of Armillaria inoculum and put regeneration or residual trees at risk o f becoming infected by the fungus. The problem is exacerbated when susceptible host species like Douglas-fir are used for regenerating infested sites. Trees planted near stumps are kil led fairly quickly. Expanding disease centres w i l l occur when root contact occurs between adjacent and recently kil led trees that have adequate inoculum potential to cause disease. This threat can be reduced by modifying silvicultural practices to minimize exposure o f susceptible host species to Armillaria inoculum in managed second-growth stands. There is a need for more research on host-pathogen interactions and natural resistance mechanisms operating in trees that show effective resistance against Armil lar ia root disease (Wiensczyk 2001). Macroscopic observations o f host response have been recorded for a variety o f hosts, however, less is known of these interactions at the cellular level and the types o f defense responses that are effective against further ingress of the fungus. The first objective of this project was to describe host responses to invasion of the roots o f western redcedar, western hemlock and Douglas-fir by A. ostoyae at the macro- and microscopic level, specifically targeting the anatomical changes in the bark tissue leading to necrophylactic periderm formation, as well as barrier zone formation associated with compartmentalization o f infected woody tissue. This was achieved by examining host 27 reactions and lesion development on inoculated and naturally infected roots and the relationship between host response and root size, distance from the root collar, tree vigor and the timing of infection. A second objective was to develop a better understanding o f the behaviour and infection process of A. sinapina which frequently co-exists in the same stand with A. ostoyae (Morrison etal. 1985, Morrison et al. 1991). A. sinapina is generally considered to be a weak pathogen on deciduous hosts and behaves for the most part as a saprotroph on coniferous hosts. In the central Interior of B . C . , Dettman and van der Kamp (2001a), however, showed that A. sinapina can be more pathogenic to coniferous hosts than previously reported. Clearly, more information is needed to elucidate the nature o f pathogenicity o f A. sinapina and whether there is any variation in susceptibility or host response to infection by A. sinapina compared to that of A. ostoyae. The final objective was to demonstrate that resistance mechanisms operating in the roots o f western redcedar, western hemlock and Douglas-fir and that the frequency o f resistant reactions expressed in roots following invasion by the fungus help explain the pattern of symptom development and mortality rates that are observed in juvenile stands in the ICH. Results w i l l be used to revise the table of susceptibility ratings for conifers (Anon. 1995) support the development of a decision aid and suggest recommendations for Armil lar ia root disease management in the southern Interior o f B . C . 28 C H A P T E R T W O : M A C R O - A N D M I C R O S C O P I C H O S T R E S P O N S E T O A B I O T I C W O U N D I N G , I N O C U L A T I O N W I T H ARMILLARIA OSTOYAE A N D N A T U R A L I N F E C T I O N S 2.1 I N T R O D U C T I O N The external tissues o f the bark, comprised o f both periderm and rhytidome, provide an important protective barrier to underlying tissues from microbial invasion. Like other plant pathogenic fungi, direct penetration of Armillaria w i l l trigger a series of predictable and coordinated events to replace injured tissue and meristems, maintain gas exchange and prevent desiccation, and restrict the ingress o f the pathogen. These non-specific host responses include periderm formation (Day 1927, Thomas 1934, Rykowski 1975, Mul l i ck 1977), barrier zone formation and callus tissue (Johnson et al. 1972, Shigo and Tippett 1981) and regeneration of new vascular cambium (Sharpies and Gunnery 1933, Warren-Wilson and Warren-Wilson 1960, Mul l i ck 1977, Oven and Torelli 1999, Dujesiefken et al. 2001). Once pathogens induce the process o f phellogen restoration, they interact not only with chemicals present in the normal bark but also with a whole suite of new chemicals produced during the cellular dedifferentiation (Puritch and Jensen 1980). Although chemical resistance plays a key role in defense responses o f plants, the formation o f structural barriers in tissues challenged by Armillaria is also an important mechanism by which further spread of the fungus may be halted (Garraway et al. 1991). This chapter describes the host response of western redcedar, western hemlock, and Douglas-fir to infection on roots at the macroscopic and microscopic levels, particularly targeting the structural barriers formed in the bark (necrophylactic periderm) and in the wood (barrier zone formation associated with compartmentalization o f infected tissue). Resistance responses o f woody plants to plant pathogenic fungi are classified as either 'passive' or 'active' (Blanchette and Biggs 1992). Passive resistance involves preformed morphological and/or chemical barriers to penetration or colonization by microorgansims. Examples o f preformed barriers may include the suberized cork cells of the periderm or various toxic extractives contained in the outer layers o f the bark. Active resistance involves induced morphological and/or chemical barriers to microbial colonization that develop when the host plant responds to injury. Two examples are the 29 formation of necrophylactic periderm (NP) in the bark and barrier zone formation in the wood. Both types o f responses occur in stems and roots and serve to protect tissues from further desiccation or necrosis and limit the spread o f the invading pathogen in healthy host tissue. The underlying mechanisms o f active defense must also involve chemical changes in the host tissues so that they resist enzymatic attack or become inhibitory or toxic to pathogens. Guillaumin et al. (1989) reported that growth of A. mellea in the bark and sapwood was considerably less in resistant plum rootstocks. This slow growth was associated with a pink or purple discoloration o f the tissue surrounding the lesion and was attributed to de novo synthesis of anthocyanic compounds involved in the resistance. This defense mechanism may inhibit the spread o f Armillaria enough so that the host is able to complete N P formation and confine the fungus to a lesion in the bark. However, the ability of trees to develop wound periderms in response to pathogen invasion quickly and efficiently to successfully contain the fungus varies between species. Robinson et al. (2004b) reported that western larch was capable o f successful N P development in advance of infection by A. ostoyae at a greater frequency and at a younger age than Douglas-fir. Resistant and susceptible hosts can be differentiated by comparing host-pathogen interactions on different species and assessing the degree o f interference in the process o f phellogen restoration (Struckmeyer and Riker 1951, M u l l i c k 1977). Tree roots respond to mechanical wounds or any factor that kil ls or removes bark by forming barriers in bark and wood in ways that are parallel to those observed in response to Armillaria invasion. However, the ability o f a host to form such barriers under the influence o f an aggressive pathogen may be interfered with, resulting in delays in N P formation in the bark or inhibited altogether. Host reactions induced following invasion by pathogenic fungi may exhibit a broad range o f temporal variability depending on the inherent resistance of the host species, the size and quality o f the inoculum, the season o f wounding, and the influence o f the environment on host-pathogen interactions. Moreover, observations have shown that host response within all species depends on age, size, and vigour o f the tree and the 30 location of the infection on the root (Morrison 1972, Morrison et al. 1991, Pearce et al. 1986, Redfern 1978, Robinson 1997, Rosso and Hansen 1998). Morrison et al. (1991) reported that the location o f an infection on the root system is an important factor in disease development. Robinson (1997) showed that infected Douglas-fir and western larch roots located close to the root collar produced N P ' s with multiple bands of thick- and thin-walled phellem, a structural characteristic that increased the trees' resistance to spread of the fungus. The frequency at which either species produced this type o f N P decreased with increasing distance from the root collar (Robinson 1997). However, Morrison (1972) found that on Corsican pine, inoculations with isolates of A. mellea at the root collar resulted in larger lesions than inoculations on the lateral roots. Furthermore, two root collar inoculations resulted in larger lesions than a single inoculation (Morrison 1972). Shaw (1980) noticed that infections initiated at the root collar and/or taproot usually killed the host more frequently than infections initiated on distal lateral roots. This phenomenon was also noted on young Douglas-fir (Buckland 1953). Pearce et al. (1986) reported that more vigorous, dominant trees were ki l led less frequently than the less vigorous, suppressed trees. To build upon Morrison's (1972) and Robinson's (1997) observations, several host variables were measured on trees inoculated with A. ostoyae to investigate whether size, age, tree vigour, or distance o f the infection from the root collar have any effect on the host response to abiotic wounding and/or A. ostoyae infection in roots o f Douglas-fir, western hemlock and western redcedar. The time of year in which trees become infected by Armillaria, together with the inherent growth characteristics o f those trees w i l l determine the nature and extent o f injury (Kozlowski 1969). Field observations (van der Kamp, pers. comm.) indicate that in many cases when larger trees succumb to death within one or two years, much of the damage may be caused by the fungus advancing and ki l l ing tissue while the tree is dormant and unable to initiate defense responses to stop the spread o f the fungus. Examination of basal lesions from trees infected with A. ostoyae indicates that the ki l led cambial surface is almost always coincident with the end o f an annual ring. However, invasion and partial or total girdling of the root collar often arises from an infection that was 31 established perhaps several years previously. Preliminary examinations of basal stem disks under this study showed traumatic resin canals as the first cell types formed at the start of an annual ring which suggests that the fungus progressively killed host tissue while the tree was dormant. Shigo and Tippett (1981) showed that trees infected with Armillaria formed a barrier zone in the early portion of the growth ring and the position of that barrier zone indicated that the spread of A. mellea occurred during the dormancy period or soon after the onset of growth. Fahn et al. (1979) showed that wounding during months of no cambial activity usually resulted in traumatic ducts being produced at the onset of the new growing season. To build on these observations, inoculation trials were conducted to investigate whether A. ostoyae can indeed successfully penetrate and cause infection in host species during the winter months. To compare the infection process and host reactions with those that were inoculated at the beginning of the growing season, trees of Douglas-fir, western hemlock and western redcedar were used. Woody substrates provide the only effective base from which Armillaria can spread and cause infection (Redfern and Filip 1991). Under experimental conditions, excised rhizomorphs of sufficient length were capable of forming new growing tips with adequate inoculum potential to infect healthy seedlings (Redfern 1973). Many reports support the view that hardwoods are generally better substrates for the production of Armillaria rhizomorphs than conifers (Morrison 1972, Rishbeth 1972, 1978). Garry oak (Quercus garryana Dougl.) branchwood is a high quality substrate for Armillaria inoculum to be used in field trials (Robinson 1997, Morrison, pers. comm.) and therefore was used as the substrate in the field inoculation trials in this study. Understanding the role of wound responses and the resistance of woody plants to Armillaria should lead to innovative control measures for managing root disease in new plantations. The A. ostoyae inoculation trials were in essence time-course studies that enabled complete characterization of wound related phenomena and lesion development over time on the three conifer species studied by systematically sampling roots several weeks and months following inoculation. 32 2.1.1 P E R I D E R M A N A T O M Y The periderm is a protective tissue o f secondary origin that replaces the epidermis in stems and roots that have continual secondary growth (Fahn 1960, Esau 1965). Periderm formation is also an important phase in the development o f protective layers at sites of leaf abscission, and injured tissues resulting from mechanical wounding, insect feeding and fungal infection (Mull ick 1975, Mul l i ck and Jensen 1973, Struckmeyer and Riker 1951). In roots, periderm arises from the pericycle after rupture of the cortex and endodermis during secondary growth (Fink 1999). In stems, the first periderm originates from anaplasia of the epidermis, of outer cortical parenchyma cells, or of phloem cells (Fink 1999). The periderm consists of three distinct tissues: the meristematic phellogen (cork cambium), the phellem (cork), and the phelloderm (Figure 2.1). 2.2 PHELLEM FORMATION 0 F ig . 2.1. A transverse section of stem of Sambucus nigra L. showing an early stage in the development of periderm with phellogen and its derivative tissues: phellem and phelloderm. Abbreviations: epi, epidermis; phe, phellem; pg, phellogen; pd, phelloderm. (courtesy of P.G. Mahlberg, cited from Dickison 2000). F ig . 2.2. Phellem are formed by periclinal divisions of the phellogen. The division of the phellogen cell gives either a phellem or phelloderm cell (cell 1). Over time, additional phellem and phelloderm derivatives are formed either externally or internally abutting the phellogen, respectively. The first derivatives of the phellogen are pushed outward so that they appear farthest from the phellogen (far right). Phellem production is generally greater than phelloderm (from Mullick 1977). Phellogen cells have thin walls which are primary in nature. They are generally oblong or rectangular in transverse and radial sections, and polygonal or irregularly shaped in tangential section. The phellogen usually consists o f just a single layer o f cells but may also appear as a zone of meristematic tissue depending on the season of wounding or its stage of development during the process of phellogen restoration. Divisions of the 33 phellogen are mostly periclinal while periodic anticlinal divisions keep pace with the increase in circumference of the axis (Esau 1965). Derivatives o f the phellogen usually form in a radial file that extends outward through the cork and inward through the phelloderm (Esau 1965). Each phellogen cell w i l l divide to form two cells, the outer daughter cell matures into a phellem cell while the inner cell remains the phellogen (Figure 2.2). Occasionally, the inner cell matures into a phelloderm and the outer cell remains a phellogen cell. Typically, the phellogen produces more phellem than phelloderm (Figure 2.2). In the coastal region of B . C . , phellogen renewal in the stems of conifers is usually initiated in M a y or June (Mull ick 1977), slows down in the autumn and virtually ceases in the winter. Robinson (1997) reported similar observations on the roots of western larch in the southern interior o f B . C . Phelloderm cells are non-suberized, l iving parenchyma cells that resemble cortical parenchyma cells in shape and content. Their radial arrangement internally abutting the phellogen usually distinguishes them from normal cortical phloem cells. However, in some plants, phelloderm may not be produced at al l or may exist as only a single layer o f cells internally abutting the phellogen following initial periclinal division (Srivastava 1964, Esau 1965). Phellem cells comprise the outer protective tissues of the bark that are formed external to the phellogen. Phellem are typically elongated with the long axis running parallel to the axis o f the stem (Lier 1952). There are two main types o f phellem cells: tihin-walled phellem and thick-walled phellem. Thin-walled phellem are suberized which offer chemical impermeability (Esau 1965, Grozdits et al. 1982). Suberin is a complex and variable polymer, composed of both l ipid and phenolic components. The presence o f lipids, including long chain fatty acids and fatty alcohols, w-hydroxy fatty acids and dicarboxylic acids, render the polymer hydrophobic, a property that contributes to the relatively inert nature o f suberized plant tissues (Kolattukudy 1984, Fink 1999). Thick walled phellem are heavily lignified (Esau 1965) and provide physical and mechanical protection (Grozdits et al. 1982). The occurrence and arrangement o f these two different cell types in the periderm varies by species. Phellem cells are non-living at maturity 34 (Esau 1965). They usually appear compacted and lack intercellular spaces except at lenticels. The walls of the cork cells may be colored brown or yellow, and the lumina contain colored resinous or tanniniferous materials (Esau 1965), non-anthocyanic and anthocyanic compounds (Mull ick 1977), and/or hydrolysable tannins (Fengel and Wegener 1984) which vary among species. Rhytidome consists of all tissues external to the phellem of the most recent phellogen. In conifers, the secondary phloem which comprises the l iving inner bark is composed of a variety of different cell types including ray and axial parenchyma, functional and non-function sieve cells, sclerenchyma cells (fibers and sclereids) and secretory cells (i.e. resin canals). The secondary phloem of Douglas-fir is made up of primarily axial parenchyma arranged in longitudinal rows. Sieve cells are invariably interspersed throughout. Ray parenchyma cells extend radially through the phloem. Sclereids and resin cysts are sometimes found in low abundance in the mid- and outer phloem. The same cell types occur in the bark of western hemlock, although the sclereids are both larger and more abundant than in Douglas-fir. In western redcedar, the phloem tissue generally contains fibers, sieve cells, and axial and ray parenchyma cells. The fibers typically alternate with parenchyma cells as single-layered tangential bands. 2.1.2 N P F O R M A T I O N I N R E S P O N S E T O W O U N D I N G The ability of trees to halt the ingress o f Armillaria infections in their roots depends on intricate sequences of physiological, chemical and anatomical events in the challenged cells of the l iving bark. The non-specific nature of defense in the bark o f conifer and hardwood species has been well described through a series of developmental studies and investigations of host-pathogen interactions (Mul l ick 1971; Mul l i ck and Jensen 1973; Mul l i ck 1975, 1977; Biggs 1984, 1985; Biggs etal. 1984). Mul l i ck (1971) observed distinct differences in the pigment composition in the periderms o f different coniferous trees and utilized cryofixation techniques to examine and characterize these tissues. Mul l i ck and Jensen (1973) classified all periderms in conifers into three types belonging to two basic categories: exophylactic and necrophylactic. 35 For many tree species, the first exophylactic periderm (FEP) replaces the epidermis within its first year and serves to protect underlying tissues from the external environment. F E P phellem are renewed every year so long as the F E P phellogen remains functional (Mul l ick 1977). A t a certain age, which varies among conifers, F E P phellogen becomes non-functional and is replaced by the second category of periderm: necrophylactic periderm (Mul l ick 1971, Mul l i ck and Jensen 1973). Hence, a necrophylactic periderm arises as a non-specific response whenever phellogen is rendered non-functional because o f age or in response to physical or pathogenic injury to the l iving bark tissue. The process o f necrophylactic periderm (NP) formation, also synonymous to the term wound periderm, enables the reestablishment of an intact periderm forming in successively deeper layers of bark. Dead tissues that accumulate external to N P on the surface of stems and/or roots comprise the outer dead bark (rhytidome), and give the surface of the bark its characteristic rough appearance (Mull ick 1977). Successive N P formation in the bark of some species like Douglas-fir, combined with retention of all the dead tissues, can give rise to very thick rhytidome tissue, especially on older stems, branches and roots. In western redcedar, thin bark may be attributed to thin phellem layers, N P ' s that cut off relatively thin layers of phloem, lower rates of phloem production and exfoliation of older bark to give a relatively smooth surface. A sequent exophylactic periderm (SEP) may develop internally abutting N P . M u l l i c k (1977) observed the separation o f necrotic tissue following SEP formation resulting in sloughing or scaling of the outer rhytidome tissue, and leaving SEP intact at the surface of the bark (Mull ick and Jensen 1973). Pigments of the phellem o f N P and that of F E P and SEP in Pacific silver fir (Abies amabilis (Dougl.) Forbes), grand fir (A. grandis (Dougl.) Lindl.) , western redcedar and western hemlock were chemically distinct (Mull ick 1971, Mul l i ck and Jensen 1973). The terminology used for bark anatomy in stems first described by Mul l i ck and Jensen (1973) may also apply to secondary tissues on roots. However, the age at which F E P 36 changes to a N P in roots certainly lags behind that o f stems. In some conifers like western redcedar which normally does not form very thick rhytidome layers, F E P may function for a longer period of time. In other conifers, F E P may be comprised of multiple layers o f phellem whereby a new layer o f phellem is formed each year. However, the distinction between F E P , N P , and SEP in coniferous roots requires further scrutiny. The occurrence of N P and SEP likely varies by root age and by species. Evidence for N P and SEP are more visible on conifer roots that typically produce thick rhytidome like Douglas-fir. In contrast, the number of layers o f phellem comprising the periderm on western redcedar roots is markedly lower than Douglas-fir or western hemlock. Field observations in this study suggest that sloughing o f outer phellem layers on cedar roots is common. However, in this study accurate classification o f periderm tissues prior to injury into either the exophylactic or necrophylactic categories was not possible without biochemical analysis of pigmentation o f phellem layers. Mul l i ck (1971) reported distinctness o f pigmentation in the first (i.e. FEP) and sequent (i.e. N P ) periderms in each o f the 40 species representing 13 genera o f conifers. Due to the variation in root age it was possible that, for all species, a N P that formed in response to abiotic injury or infection by A. ostoyae deeper in the bark tissue could have formed underneath the F E P or a N P that formed previously. 2.1.3 N O N - S U B E R I Z E D I M P E R V I O U S T I S S U E (NIT): A T I S S U E E S S E N T I A L F O R R E G E N E R A T I O N O F N P There has been some controversy surrounding the non-suberized nature o f the impervious zone of tissue formed during the initial stages o f phellogen restoration. Mul l i ck (1975) demonstrated the development of a zone of impervious tissue bordering injured bark which invariably precedes formation o f a new periderm. Histochemical tests revealed that imperviousness was not due to suberization of cells. Thus Mul l i ck coined the term 'non-suberized impervious tissue' (NIT). These observations were also confirmed by Soo (1977) when a sequential application o f a ferric chloride and then a potassium ferricyanide solution (F-F test), applied to prepared samples through either the wound or cambial surface, failed to penetrate a zone o f hypertrophied tissue. Histochemical tests confirmed that impervious tissue stained positively for lignin, but not suberin. 37 Lignin is predominantly composed of different polymers of phenyl-propanes including coniferyl alcohol (Fink 1999). Precisely how lignin inhibits attack by potential pathogens is uncertain, but several possible mechanisms have been postulated including the physical strength o f the polymer which makes cell walls more resistant to mechanical or enzymatic breakdown, restriction of diffusion of toxins from infected into non-infected cells and the transport of nutrients in the opposite direction, the physical shielding of host cell wall polysaccharides from attack by fungal enzymes, and the effects of the phenolic precursors of lignin on the pathogen themselves (Fink 1999). Subsequent developmental studies revealed that N I T does not develop from meristematic activity, but instead v ia hypertrophic dedifferentiation o f extant phloem cells around the periphery o f the injured tissue. The cell hypertrophy reduces flow or diffusion of pathogens and their detrimental products by elimination of the apoplastic pathway (through cell walls and intercellular spaces). Furthermore, N I T always preceded formation of N P , and N P developed specifically from tissues internally abutting N I T (Mul l ick 1977). Hence, Mul l i ck hypothesized that the development o f NIT around wounded bark provides the underlying tissue the time and conditions necessary for phellogen restoration to occur. However, subsequent studies of periderm formation in response to physical wounding or pathogenic injury failed to confirm that this zone of impervious tissue was indeed non-suberized. Biggs (1984,1985) reported the presence of intracellular suberin lamellae associated with the layer of lignified, impervious cells. Using a more sensitive technique to detect suberin, Biggs (1984,1985) demonstrated that suberized cells were always found internally abutting the non-suberized lignified cells. The author proposed that impermeability o f this tissue was attributed to cell wall suberization and that the term N I T should no longer be used (Biggs 1985). Woodward and Pearce (1988) also demonstrated suberization in the impervious layer formed in Sitka spruce (Picea sitchensis (Bong.) Carr.) trees challenged with Phaeolus schweinitzii (Fr.) Pat following treatment of the impervious tissue with chlorine dioxide to remove the lignin in cell walls. The authors suggest that masking o f suberin staining by phenolic compounds likely caused earlier confusion by Mul l i ck . Removal of the wal l -38 bound phenolics revealed the early onset of suberization and the F-F test confirmed that imperviousness was delineated by this layer o f suberized tissue (Woodward and Pearce 1988). Very few researchers have adopted the terminology proposed by Mul l i ck (1975). Subsequent studies typically recognize this zone o f impervious tissue as a ligno-suberized zone (LSZ) (Lindberg 1991, Wahlstrom and Johannson 1992, Oven and Torelli 1994, Oven et al. 1999, Simard et al. 2001). However, Mul l i ck (1977) recognized that while tissue underlying the N I T zone may be in the process o f redifferentiation, sporadic immature phellem-like cells may occur around the periphery o f NIT . A newly restored phellogen may form phellem as early as two days after impermeability sets in and at this time, suberin may be detected in both the phellem cell walls and the cell walls o f some N I T cells immediately adjacent to the newly formed phellem (Mull ick 1977). Often, there may be cells caught between the N I T and the zone of redifferentiation where the new phellogen is differentiating which develop suberized linings and take on the appearance o f phellem cells. However, these cells are neither N I T nor phellem cells. Robinson (1997) reported that it was the cells caught between the N I T and the developing phellogen that become suberized, not the impervious tissue itself. This study did not use other histochemical staining techniques for lignin other than Phloroglucinol + H C l and as such, no confirmation o f suberization in the impervious zone was conducted aside from viewing a Phloroglucinol-HCl stained section under both B F (to distinguish a lignified zone o f NIT) and U V fluorescence (to distinguish between suberized and non-suberized tissue). Sudan III was also used on separate sections to confirm presence o f suberin in cells. Because Mul l i ck (1977) acknowledged lignification and hypertrophic dedifferentiation o f tissues leading to the formation o f impervious tissue and that the suberization o f cells may appear immediately following impermeability, the provisional view of this author is to continue to use the terminology proposed by Mul l i ck (1977). 39 2.1.4 F A C T O R S A F F E C T I N G N I T D E V E L O P M E N T A N D N P F O R M A T I O N N I T and N P may be triggered by a number of factors including insect (Mull ick 1975) and fungal attack (Struckmeyer and Riker 1951; Biggs 1984, 1985; Wahlstrom and Johansson 1992; Robinson 1997; Simard et al. 2001; Solla et al. 2002), the time o f year in which the tree is wounded (Mull ick and Jensen 1976), and other environmental stress factors including drought (Puritch and Mul l i ck 1975), all o f which may slow the rate of N I T or N P formation in the bark. Mul l i ck postulated that N I T and N P formation surrounding bark occupied by the balsam woolly adelgid (Adelges piceae Ratz.) may be inhibited or delayed by chemicals secreted by the adelgid's stylets (Mull ick 1975, 1977). Wahlstrom and Johansson (1992) found that A. ostoyae infection in the roots of Pinus sylvesteris delayed the formation o f N P but increased the number o f lignified cell layers i n the bark relative to bark with non-challenged wounding. A subsequent study showed similar results whereby P. sitchensis roots inoculated with A. ostoyae or A. mellea delayed formation of impervious tissue and wound periderm compared to superficial wounds where the rhytidome tissues were removed or deeper wounding to the vascular cambium (Solla et al. 2002). Robinson et al. (2004b) reported that NPs formed in response to invasion by A. ostoyae halted further fungal advance on 68% and 45% of the roots examined from 10- and 27-year old western larch trees, respectively. The majority o f Douglas-fir roots in the same relative age classes showed frequent breaching o f NPs resulting in progressive lesions in the bark and cambial zone (Robinson et al. 2004b). The rate o f N I T formation varies at different times o f the year and similar to other physiological processes, N I T formation displays a cyclical pattern (Mul l ick and Jensen 1976). Following mechanical wounding to conifer stems, the rate of NIT formation was fastest in June, slowed down in August/September, virtually ceased in the winter and then resumed slowly again in the spring. Trees that were wounded in November during the dormant period did not complete N I T formation until the following spring in M a y (Mul l ick and Jensen 1976). Robinson (1997) reported similar results in western larch roots following injuries made to the bark at various times throughout the year. Following 40 mechanical injuries to the stems of western hemlock and western redcedar in July, Mul l i ck and Jensen (1976) reported that N I T was complete after about 30 days for both species and the authors speculated that impermeability may occur after just 2 weeks. However, these experiments were conducted in the coastal region of B . C . where the climatic conditions are generally more favourable for tree growth and vigour. Puritch and Mul l i ck (1975) showed that water stress could significantly delay the development of N I T tissues in A. grandis seedlings. Adverse effects o f stress factors such as drought on the susceptibility o f host species to attack by pathogens is a familiar concept, however the indirect effects of drought on the pathogen are less well known. Pathogenesis depends on specific interactions between the pathogen and the initiation of defense responses in the host involving complex biochemical and structural cellular alterations such as N I T and N P formation. The outcome of this interaction results in either susceptible or resistant host reactions. 2.1.5 N I T D E V E L O P M E N T A N D N P F O R M A T I O N : T H E I R R O L E I N R E S I S T A N C E A G A I N S T A R M I L L A R I A R O O T D I S E A S E . Non-specific responses including N I T development and N P formation are induced following direct injury to the phellogen which renders the meristem non-functional. Mul l i ck (1977) suggested that while N P per se may not be involved in host-pathogen interactions, the process o f its restoration, including N I T formation, is, and that the development of NIT in response to wounding may be the physiological basis of host response to disease in conifers (Mul l ick 1975). Robinson (1997) reported that N P ' s with multiple bands o f phellem were more frequently formed in the roots o f western larch and more effective at halting further spread of A. ostoyae in adjacent host tissue than in Douglas-fir. Less is known about the resistance responses of other conifers commonly found in the southern interior region o f B . C . Host susceptibility occurs when N I T formation is not triggered and/or when a fungus is able to circumvent the process of phellogen renewal (Mull ick 1977). In the case of Armillaria, penetration of the outer bark is similar in both susceptible and tolerant hosts, however host responses initiated following penetration o f the fungus w i l l determine 41 whether a particular reaction is susceptible or resistant. In this study, resistance or susceptibility was determined by how successfully the host was able to complete the process of phellogen restoration and N P formation while under the influence of the fungus. In situations where the fungus advanced to and killed the vascular cambium, resistance reactions were judged as those which appeared to have successfully compartmentalized the infection and showed lateral ingrowth o f callus tissue over the wounded tissue. To understand the effects Armillaria had on the initiation and development of N I T and N P formation, an understanding of the host response in the absence o f Armillaria was required. Thus, characterization of the process o f wound healing in abiotically wounded root bark was used for comparison. Cryoflxation, fluorescence microscopy and standard histological techniques were developed by Mul l i ck (1971) to characterize fresh samples of bark tissue and the anatomical changes in cells in response to wounding. This method preserves the chemical integrity of fresh tissue samples and permits physiological interpretation o f infected host tissue by observing changes in fluorescence characteristics when compared to healthy host tissue. Similar methods were used by Robinson (1997) to describe the host response to infection by A. ostoyae in the roots of western larch and Douglas-fir. To build on Robinson's work, similar techniques were employed to describe the host response to abiotic wounding, inoculation with A. ostoyae, and roots naturally infected with A. ostoyae in Douglas-fir, western hemlock, and western redcedar trees. 42 2.2 MATERIALS AND METHODS 2.2.1 S T U D Y SITES Four inoculation trials were conducted at three different sites in the Okanagan-Shuswap and Arrow-Boundary Forest Districts of the Southern Interior Forest Region (Appendix I). The three Douglas-fir sites were 20-30-years old and contained naturally regenerated western redcedar and western hemlock more than 15-years old. Table 2.1 gives details o f site locations and characteristics. 2.2.2 I N O C U L U M B L O C K P R E P A R A T I O N The method used to prepare inoculum blocks was based on that described in Redfern (1970, 1975), Morrison (1972) and used by Robinson (1997) with some modification. Inoculum blocks were prepared using known isolates of A. ostoyae (provided by D . Morrison, CFS) grown on segments o f Garry oak branch wood. Isolate No ' s . Y - 4 (collected from 70-Mile House) and 87-01 (collected from Victoria) were used to prepare inocula in the Hidden Lake (2002) and Kingfisher (2003) inoculation trials, whereas only isolate 87-01 was used in the Kingfisher (2004) and Nakusp (2004) trials. Different methods for making "starter discs" were used from year-to-year, all o f which resulted in successful inoculation of garry oak segments. In the first method, A. ostoyae was grown in Petri dishes containing 3% malt extract-1.5% agar medium. Autoclaved beech (Fagus grandijlora Ehrh.) and garry oak discs approximately 1 cm in diameter and 5 mm thick were added to mycelial colonies o f A. ostoyae. Once discs were fully colonized by the fungus they were used as "starter discs" to inoculate fresh segments o f Garry oak. In the second method, fully colonized inoculum blocks (approximately 2 x 8 cm) were cut into discs (approximately 5 mm thick) using a band saw and then cut into smaller pie-shaped "starter discs" to be used to inoculate fresh segments o f Garry oak. Garry oak branch wood (approximately 2-3 cm in diameter) was collected from living trees and taken back to the lab. A l l surface organisms such as lichen and mosses were scraped from the surface of the bark. Using a band saw, branches were then cut into 8 cm segments, surface sterilized in a 1% solution o f bleach for 15 minutes and rinsed in water. Table 2.1. Location and characteristics of three sites used in the four inoculation trials at which the host response to infection by Armillaria ostoyae in the roots of western hemlock, western redcedar and Douglas-fir trees was investigated. Site Year(s) BEC Trial was subzone/ Initiated site series Location Elevation Species Composition2 (m) Planted Natural Soil Type3 Plant Associations Hidden 2002 ICHmw2 50°31'N, 600 Fd, Lw Cw, Hw, Brunisol Paxistima myrsinites Lake 03 118°50'W Pw, Ep, Clintonia uniflora At, Se Frageria virginiana Linnaea borealis Rubus parviflorus Pleurozium schreberi Viola glabella Gymocarpium dryopteris King-fisher 2003 & 2004 ICHmw2 03 50°42'N, 118°44'W 600 Fd, Pl Cw, Hw Brunisol Paxistima myrsinites Pw, Ep, Clintonia uniflora Pl, Act, Athyrium filix-femina Tw, Se, Pleurozium schreberi W Peltigera aphthosa Linnaea borealis Rubus parviflorus Viola glabella Frageria virginiana Cladina spp. Gymocarpium dryopteris Naksup 2004 ICHmw2 05 50°17'N, 117°44'W 700 Fd Cw, Hw Luvisol Paxistima myrsinites Ep, Tw Clintonia uniflora Athyrium filix-femina Pleurozium schreberi Peltigera aphthosa Gymocarpium dryopteris Mahonia aquifolium Acer glabrum Rubus parviflorus Viola glabella Frageria virginiana Cladina spp. Streptopus roseus Galium triflorum Linnaea borealis See Meidinger and Pojar (1991). Actual soil hygrotopes and tropotopes were read from tables in Lloyd et al. (1990) and Braumandl and Curran (1992). 2 At, trembling aspen; Act, black cottonwood; Cw, western redcedar; Fd, Douglas-fir; Ep, paper birch; Hw, western hemlock; Lw, western larch; Pl, lodgepole pine; Pw, western white pine; Se, Engelmann spruce; Tw, pacific yew; W, willow 3 Soil orders were determined by examining soil profiles and determining soil texture from two soil pits dug at each site and using the Canadian System of Soil Classification, 3rd Ed. Ottawa. Agriculture and Agri-Food Canada and http://www.enviromnent.ualberra.ca/SoilsERM/class.html 44 A colonized disc was pinned over the cambium at the end of a fresh garry oak segment using small brass pins. Intimate contact between the starter disc and the fresh cambial surface of the garry oak segment is essential for colonization to take place (Morrison 1972). Inoculum blocks were then placed in buckets o f pre-washed, non-sterile coarse sand. After 8-12 weeks, the fungus fully colonized bark and cambial tissue and actively growing rhizomorphs emerged from the cambial edge at the opposite end from the starter disc and from cracks in the bark along the length o f the segment. Inoculum blocks showing abundant rhizomorph production and/or mycelium in the cambial region were used to inoculate healthy trees in the field. Blocks partially colonized by other fungi such as Xylaria sp. were still used for inoculum so long as the Armillaria mycelium was still present as emerging rhizomorphs. 2.2.3 I N O C U L A T I O N A N D S A M P L I N G T E C H N I Q U E In the field, root systems of western redcedar, western hemlock and Douglas-fir trees which did not exhibit any aboveground signs or symptoms o f root disease were carefully excavated up to a meter away from the root collar. A t least ten trees o f each species were inoculated in each trial. Existing rhizomorphs were removed from inoculum blocks in order to stimulate new rhizomorph growth and blocks were then placed either directly below or alongside the root without injuring the root itself. When possible, all inoculations on a single tree (up to four) occurred on separate primary roots. Flagging tape was tied around the root distal to the point o f inoculation. The root diameter and the distance from the root collar to the point o f inoculation were then accurately measured and then the root and the block were covered with mineral soil. In the first inoculation trial at Hidden Lake in 2002, the selection criterion for root diameter and distance from the root collar was usually in the range o f 2-3 cm and between 30-60 cm, respectively. Secondary roots were utilized when primary roots were unable to fulfill the selection criteria. In subsequent trials, a range o f root diameters and distances were obtained to determine what, i f any, difference these host variables may have on the type and success o f the host response. Trees were tagged and measured for diameter at 1.3 m (DBH) . 45 A s s imi la r techniques were used for producing i n o c u l u m i n the lab and the same technique was used for inocula t ing roots i n the f ie ld , experimental condi t ions for Armillaria i n o c u l u m w i t h respect to the v i ab i l i t y and i n o c u l u m potential were more or less homogenous f rom year to year. F igure 2.3 shows the inocula t ion technique in situ w i t h a garry oak segment co lon ized by A. ostoyae p laced alongside a healthy root without in jur ing the root itself. F igure 2.4 shows profuse rh izomorphs emerging f rom the i n o c u l u m b l o c k and adhering to the outer surface o f the root. Fig 2.3. Inoculation technique shown in situ. A. ostoyae inoculum block placed alongside a healthy western hemlock root. Fig. 2.4. An inoculum block as it was placed against a western redcedar root prior to harvesting from the ground. Clusters of A. ostoyae rhizomorphs (arrows) are shown emerging from the inoculum block and adhering to the outer surface of the root. In a l l but the first inocula t ion t r ia l , abiot ic w o u n d i n g to the roots was used as an addi t ional treatment to a l l o w compar i son between it and host reactions to Armillaria invas ion . A b i o t i c w o u n d i n g was performed o n healthy roots separate from those roots used for inoculat ions. W o u n d i n g was carr ied out wi thout phys ica l ly disrupt ing the outer bark by soak ing a smal l c i rcu lar polyurethane sponge (approximate 8-10 m m i n diameter) i n l iquef ied SITUR/Freeze™ cryogen spray (Tr iangle B i o m e d i c a l Sciences, D u r h a m , N C , U . S . A . ) w h i c h has temperatures o f -55 to - 6 0 ° C . U s i n g tweezers, the soaked sponge was appl ied to the exterior surface o f the root for a g iven length o f t ime depending o n the thickness o f the bark (usually between 10-15 seconds). T h i s method achieved a freezing injury i n the ph loem tissue to at least h a l f the bark thickness and thereby permitted characterization o f phel logen restoration and host response i n the absence o f Armillaria. 46 Flagging tape was tied around the root approximately 3 cm distal to the point of injury so that the wound could be easily located at some later sampling date. Both treatments (abiotic wounding and inoculations) covered the same range of root diameters and distances from the root collar. In the last two field inoculation trials, a further control treatment was added, namely placement o f uninfected fresh segments of garry oak next to uninfected roots. This control treatment was applied to the same trees but on separate roots from those used for abiotic wounding and inoculations. Trees were inoculated on different dates depending on the field trial. In the Hidden Lake field trial, inoculations occurred on two separate dates: July 8, 2002 and August 22, 2002. Roots were,harvested at 11 weeks and approximately 1 year following each inoculation date. In all subsequent field trials, trees were inoculated in early M a y and roots were harvested at two or three of the following intervals: 5 weeks, 8-9 weeks, 11 weeks, 5 months, and 1 year after inoculation depending on the specific trial. A t each harvesting date, the number of samples collected was relatively even among all species. Roots were carefully excavated and examined in situ to ensure that the placement of the inoculum block against the root had not shifted. In situations where there was clearly no contact between the cambial surface of the inoculum block and the root, the root was omitted from any further analysis. Also , the viability of inoculum at the time of harvesting was assessed by removing the bark from the inoculum unit. If mycelia were absent from the inoculum block, it was considered a non-viable unit. Roots associated with inoculum blocks with non-viable inoculum (lacking rhizomorphs) were excluded from the analysis. Roots that escaped disease because the fungus was not present at the root surface were also excluded from the analysis. Roots were marked with paint to define the point of contact with the inoculum block after the root had been harvested from the ground. Where evidence o f infection was present on the root (i.e. resinosus, or clumps of rhizomorphs adhering to the surface), sections of 47 the root up to 60 cm in length were removed from the ground. Root sections were washed with water to remove soil from the root surface and dissected for examination of host response to infection. Macroscopic observations of host response (i.e. the type of host reaction in the bark, bark hypertrophy, resin exudation, compartmentalization and callusing) were noted. In addition to root diameter and the distance from the root collar, the following measurements were also recorded: root age, inner bark thickness, and lesion length. Tree vigour was determined by taking cores at four cardinal directions at the root collar area from each tree showing successful inoculations. Tree vigour index was calculated as the most recent average annual increment divided by the 5-year average. The previous years' annual average annual increment was used for samples harvested during the growing season while the current average annual increment was used for samples harvested at the end of the growing season. Tree age was also determined from the core sample. Abiotic wounds were sampled concurrently with inoculations. Bark samples from roots with control blocks were collected in the same manner as that for abiotic wounding and inoculations, described below. Healthy bark samples were collected from treated roots at least 30 cm proximal or distal to the point of injury. 2.2.3.1 W I N T E R I N O C U L A T I O N S Three additional inoculation trials were implemented to investigate whether A. ostoyae can successfully penetrate and cause infection in Douglas-fir, western hemlock, and western redcedar in the winter months during tree dormancy, and what, i f any host response is initiated during that time. The inoculation trials were conducted at the Hidden Lake and Kingfisher sites in 2002-2004 (refer to Table 2.1 for site locations and characteristics). Inoculation and sampling techniques followed the same methodology outlined in Section 2.2.2, Inoculum block 48 Preparation, and Section 2.2.3, Inoculation and sampling technique. Neither abiotic wounding nor controls were used in any o f the winter inoculation trials. Trees were inoculated on different dates in each field trial. Inoculations occurred in early November in 2002, in mid-October in 2003 and in mid-September in 2004. Roots were harvested the following spring at the onset o f growth (late April/early May). In 2003, additional roots o f each species were left in the ground and subsequently harvested at different intervals over the course o f the growing season to observe how infection developed on inoculated hosts over time. The viability of the inoculum block was observed at the time of collection by removing the bark from the inoculum unit and assessing the presence o f mycelial fans. Sample treatment was carried out only on roots showing visible evidence of infection (browning o f the phloem at inoculum contact or rhizomorph penetration) according to the methods described in Section 2.2.4. 2.2.3.2 E X A M I N A T I O N O F R O O T S N A T U R A L L Y I N F E C T E D W I T H A OSTOYAE In the summer of 2002, roots naturally infected with A. ostoyae were examined to determine the host reactions involved in lesion development in Douglas-fir, western hemlock and western redcedar trees, enabling a comparison with artificial inoculations o f those species. Two Douglas-fir plantations that had naturally regenerated hemlock and cedar and that had Armillaria-caused mortality were selected for this component of the study. The sites were located near Hidden Lake, 30 km east o f Enderby in the I C H mw2 (Lloyd et al. 1990) in the Okanagan Shuswap Forest District. In the field, root systems of trees showing no above-ground evidence o f Armil lar ia root disease and growing in close proximity to a recently kil led tree were carefully excavated to reveal root contacts (Figure 2.5). Roots were examined for symptoms of infection caused by A. ostoyae (i.e. resinosus on the surface o f the root or irregularities on the surface of the bark) at root contact as well as proximal and distal regions for rhizomorph-initiated lesions. Root samples were collected and examined in the same manner as that outlined in Section 2.2.4, Sample Treatment. 49 Tit-Fig. 2.5. Excavation of roots systems between a recently killed Douglas-fir and healthy western redcedar. Root contacts are flagged for examination of Armillaria-caused lesions at root contact. 2.2.4 SAMPLE T R E A T M E N T In the field, tissue sections were collected by taking a 2 x 1 cm sample of bark either at the point where the inoculum block contacted the root or from the proximal and distal infection fronts at the margin of A. ostoyae-caused lesions. Infection fronts were recognized as either a necrophylactic periderm or necrotic host tissue separating apparently healthy (uncolonized) tissue from infected (colonized) tissue either proximal or distal to the point of invasion. When infection had spread to the vascular cambium, samples were collected from the infection front in the inner bark immediately below the original outer periderm as well as from the infection front at the vascular cambium. The radial face of the sample contained all the details of interest. Samples were trimmed to have the infection front centred on the radial face. Abiotically wounded samples were also collected with the margin of the lesion centred on the radial face. In order to minimize reactions to cutting and trimming during tissue harvest, samples were immediately coated in OCT® (Optimum Cutting Temperature compound; Thermo Shandon, Pittsburgh, PA, USA) and frozen in liquid nitrogen. 5 0 For histological studies of woody tissue, samples from root lesions showing evidence o f cambial invasion and/or compartmentalization and callusing were trimmed to approximately 2 x 1 cm and fixed in F A A (70% ethanol, formalin, glacial acetic acid, glycerin). In the lab, the sample blocks were dehydrated through a gradual series o f alcohols, embedded in paraffin wax, and the radial face was mounted and sectioned on a rotary microtome. Sections 16 urn thick were transferred to a glass microslide (Fisher Finest Premium Superffost, Fisher Scientific, Pittsburgh, P A , U S A ) , with approximately 6 sections per slide. In preparation for staining, the paraffin wax was removed from the slides using Hemo-De (Fisher Scientific, Pittsburgh, P A , U S A ) clearing solution and then hydrated through a gradual series of alcohols. Slides were stained with 0.2% Fast Green in 95% ethanol and 1% Safranin in H2O (Jensen 1962), a common stain used to delineate between cellulosic cell walls of plant tissue, and then mounted with Permount™ (Fisher Scientific, Pittsburgh, P A , U S A ) . Samples were examined under the light microscope to describe the host response following cambial invasion and barrier zone anatomy associated with compartmentalization for each species. In the lab, healthy, abiotically wounded, and infected bark samples were transferred from liquid nitrogen to a cryostat chamber at -20°C housing a microtome. Samples were mounted and dissected to reveal the radial (longitudinal) face of the sample. Host reactions including necrophylactic periderm formation, compartmentalization and callusing were examined macroscopically while still mounted in the cryostat at x6 to x25 magnification using a W I L D M 5 stereo zoom microscope ( W I L D Heersbrugg, Switzerland) and photographed using Kodak Elitechrome 100 color slide f i lm (Eastman Kodak Co. , Chicago, IL, U S A ) using a 35-mm camera mounted to a phototube. Sections 14 um thick were picked up off the microtome knife using pre-chilled slides, mounted in pre-chilled cryostat o i l (Cryo-cut Microtome Lubricant, American Optical Corporation) and covered with a pre-chilled 22 x 22 mm N o . l . glass coverslip (Fisher Finest™ Premium cover glass, Fisher Scientific, U S A ) . Usually two sections were mounted per slide. 51 The frozen sections mounted on slides were examined microscopically on a Carl Zeiss Photomicroscope II (Carl Zeiss, Oberkochen, Germany) equipped with a freezing stage maintained at approximately - 3 5 ° C . Slides were examined under tungsten-illuminated bright field (BF) and incident mercury-illuminated fluorescence. Fluorescence was achieved using an O S R A M HBO® 50 W mercury lamp and fluorescence filter combinations for blue light (BL) excitation in the 390-440 nm range and ultra-violet ( U V ) excitation at 365/366 nm. The microscope was equipped with a Carl Zeiss epi-fluorescence condenser and with filter sets 48 77 05 (comprising a B P 390-440 exciter filter, F T 460 chromatic beam splitter and an L P 475 barrier filter for viewing in B L ) and 48 77 01 (comprising a B P 365/10 exciter filter, F T 390 Chromatic beam splitter and an L P 530 barrier filter) for viewing in U V . Carl Zeiss Neofluar objectives were also fitted to the microscope. Additional cryostat sections were air-dried and stained with a saturated aqueous solution of phloroglucinol in 20% H C l (Ruzin 1999) to detect lignin in cell walls. Phloroglucinol-H C l w i l l stain lignin red. One to two drops o f phloroglucinol-HCl solution was added to air-dried sections, covered with a glass coverslip and examined immediately under B F . Separate individual sections were also stained with 0.1% toludine blue in H2O (Jensen 1962) to visualize fungal hyphae in host tissue. Sudan III is a highly specific stain for cutinized and suberized cell walls (Jensen 1962). A few drops o f Sudan III in ethylene glycol were added to air-dried sections and placed in a desiccator. After 12-24 hours, sections were generously rinsed with 85% ethylene glycol, covered with a glass coverslip and examined on the microscope. In B F , suberized cell walls appear orange-red. Suberized cell walls fluorescence bright orange in B L . Suberin was also detected in sections following staining with phloroglucinol-HCl. Phloroglucinol-HCl stains the aldehyde groups o f both lignin and suberin but autofluoresence of lignin is quenched whereas suberin is unaffected and suberized cells fluoresce bright blue-violet under U V . Although phloroglucinol-HCl is highly specific for detecting lignin (Jensen 1962, Ruzin 1999), and suberin (Biggs 1984) in plant cells, Perez-de-Luque et al. (2006) suggest that its specificity might also be expanded to include polyphenols. Safrannin-Fast green was used for staining serial cryostat sections o f cedar root bark. Photomicrographs were taken 52 of both frozen and stained cryostat sections, as well as paraffin-embedded sections with Kodak Ekrachrome 160 Tungsten fi lm (Eastman Kodak Co. , Chicago, IL , U S A ) . 2.2.5 R E - I S O L A T I O N O F A. OSTOYAE F R O M L E S I O N S Re-isolation of the fungus was attempted from a sample of roots that showed distinct Armillaria-caused lesions. Mycel ia l fans or rhizomorphs surface-sterilized with 95% ethanol were transferred to Petri-dishes or tubes containing 3% M E A amended with B D S solution (Worrall 1991), sealed with Parafilm® M (Pechiney Plastic Packaging, Menasha,WI, U S A ) , and stored in the dark at room temperature. For inoculations, the isolate collected from the inoculated root was challenged in a diploid-diploid pairing with the corresponding isolate of A. ostoyae from the inoculum block used to inoculate the root to confirm that the resulting infection was not caused by on-site inoculum. To demonstrate incompatibility reactions in culture, the same isolate was paired against different known isolates of A. ostoyae collected elsewhere in British Columbia (BC-9 collected from Campbell River, provided by D . Morrison, CFS) and Britain ( A g l and Sp6 collected from Elveden and Lynford, respectively, provided by D . Morrison, CFS) , and known British Columbia isolates o f A. sinapina: one from Merritt (provided by D . Morrison) and one from M i c a Creek (Isolate #29-2-8C provided by M . Cruickshank, CFS) . For naturally infected roots, the isolate collected from Armillaria-cmsed lesions was paired with known isolates o f A ostoyae (Canadian Centre for the Culture o f Microorganisms ( C C C M ) N o . 63) and A. sinapina ( C C C M N o . 64) to examine intra- and inter-specific reactions and confirm which Armillaria species caused infection on the trees sampled. A l l pairings were done in triplicate on 3% M E A and incubated at room temperature in the dark. Pairings were carefully monitored and scored as either compatible or incompatible after 6-8 weeks depending on the distance between the two opposing mycelial colonies and the rapidity of mycelial growth. Compatible-type matings display complete intermingling o f opposing mycelia. Incompatible mating as a result o f intraspecific crosses (two isolates o f the same species) display incomplete intermingling of mycelia at the opposing edges, sometimes appearing as a hyphal-free zone between the two opposing mycelia. Incompatible matings as a result of interspecific crosses (two isolates of different species) develop a dark pigmented demarcation line in 53 the agar at the margin where the two opposing colonies meet (Korhonen 1978). Photographs were taken to demonstrate interspecific and intraspecific reactions between isolates. 2.2.6 S T A T I S T I C A L A N A L Y S I S Chi-square tests were used to analyze whether there were any differences in (i) the total number of root inoculations that resulted in successful penetration of the outer bark between species at each harvest date within a given field trial, (ii) the frequency o f infection between the different harvest times for each species within a given field trial, and (iii) the frequency o f infection between field trials at a given harvest date (i.e. 1 year) within a given species. The number of sites or years or both, of field data represent a good random sample o f all sites in the moist-warm subzone of the I C H zone over time whereby host reactions induced following invasion by the fungus are visible at all harvest dates. Data were pooled from all field trials across sites, years and time since inoculation and Chi-square tests were used to determine whether (iv) there was any difference in the frequency of successful resistance reactions induced following invasion by A. ostoyae between species. Frequency data for all field trials showing the different stages o f host response induced in each species following invasion by A. ostoyae were analyzed using Chi-square tests to determine whether there was any significant variation between species in the frequency of (v) N I T formation, (vi) N I T breached, (vii) N P formation, (viii) N P breached, and (ix) compartmentalization and callusing. Data for host variables (i.e. root diameter, distance from the root collar, and tree vigor) collected from all infected trees in each field inoculation trial were pooled. Successful resistance reactions as a proportion o f the total number o f roots penetrated by A. ostoyae were plotted against each o f the host variables to determine whether there was any relationship between host response and a particular host variable. 5 4 2.3 RESULTS AND DISCUSSION 2.3.1 I N O C U L A T I O N T R I A L : F R E Q U E N C Y O F I N F E C T I O N The number of roots sampled at each harvest date was generally divided evenly for all species. However, uneven sample sizes among species resulted from the exclusion of individual host roots that were not considered to be challenged by the fungus as described below. Table 2.2 shows the frequency of infection caused by A. ostoyae in the roots o f Douglas-fir, western hemlock and western redcedar at each harvest date for each trial. A l l inoculum blocks produced rhizomorphs. However, the amount o f rhizomorphs produced was quite variable. Some inoculum blocks produced abundant rhizomorphs at the opposite end from where the starter disc was pinned to the cambial edge of the Garry oak segment. It was projected that the fungus would grow out o f a starter disc into the bark and cambial region of a fresh Garry oak segment, and its progressive colonization strategy would result in the fungus emerging at the opposite end of the block where it would produce tufts o f mycelium or rhizomorphs or both (D. Morrison, pers. commun.). This end was then placed directly alongside the root so that fresh, newly emerging mycelium or rhizomorphs would be in direct contact with the host. However, there was a number o f situations where the fungus produced significantly more rhizomorphs around the starter disc and very little from the cambial edge o f the block situated against the root. Furthermore, there were some situations in which considerably more rhizomorphs emerged from the cracks in the bark along the length o f the block and had not yet grown to attach themselves onto the host root at the time o f sampling. In addition, because o f the diversity of and interactions among microbial communities in the soil, inoculum blocks were often colonized by common cord-forming fungi like Xylaria sp. Blocks that had surface colonization of Xylaria sp. were nevertheless still considered to be viable inoculum because other fungi were strictly superficial on the bark yet the inner bark and cambium were fully colonized by A. ostoyae and blocks produced rhizomorphs. Indeed, many of these blocks resulted in successful penetration by the fungus. However, cases where competing fungi occupied the cambial zone o f the block and inhibited A. ostoyae 55 Table 2.2. Number of root inoculations on Douglas-fir, western hemlock and western redcedar and the number and proportion producing infection by A. ostoyae at various harvest times in four field trials. Harvest No. of root No. of Frequency Species Field Trial Time inoculations infections of infection Douglas-fir Hidden Lake (2002) 11 weeks 11 9 0.82 1 year 10 10 1.00 Kingfisher (2003) 5 weeks 2 0 0.00 9 weeks 4 0 0.00 1 year 30 10 0.33 Kingfisher (2004) 9 weeks 5 0 0.00 5 months 14 9 0.64 1 year 13 12 0.92 Nakusp (2004) 8 weeks 2 0 0.00 5 months 2 2 1.00 1 year 2 2 1.00 Western hemlock Hidden Lake (2002) 11 weeks 19 14 0.74 1 year 6 5 0.83 Kingfisher (2003) 5 weeks 7 1 0.14 9 weeks 8 3 0.38 1 year 30 23 . 0.77 Kingfisher (2004) 9 weeks 7 1 0.14 5 months 14 11 0.79 1 year 17 12 0.71 Nakusp (2004) 8 weeks 3 2 0.67 5 months 7 6 0.86 1 year 3 2 0.67 Western redcedar Hidden Lake (2002) 11 weeks 19 6 0.32 1 year 10 4 0.40 Kingfisher (2003) 5 weeks 10 0 0.00 9 weeks 5 2 0.40 1 year 24 8 0.33 Kingfisher (2004) 9 weeks 4 1 0.25 5 months 20 16 0.80 1 year 12 10 0.83 Nakusp (2004) 8 weeks 2 1 0.50 5 months 7 7 1.00 1 year 2 2 1.00 56 rhizomorphs from emerging from the end of the block, were omitted from further analysis. A l so , where the inoculum blocks failed to produce rhizomorphs at surface contact with the root, the host was not considered to be 'challenged' by the fungus and those samples were excluded from the data set. A s fungal infection is necessary for subsequent analysis o f defense mechanisms induced in the roots o f the different conifer species, the success of fungal inoculation was a critical and a fundamental aspect of this study. Assessment of host response to infection could only be done on roots showing successful penetration of the outer root bark. C h i -square tests among and between species for all field trials is given in Appendix II. There was no difference in the frequency o f infection between the two isolates of A. ostoyae at either the 2002 Hidden Lake trial or the 2003 Kingfisher trial. Both isolates colonized the Garry oak branch segments and rhizomorphs emerged from slits in the bark and from the cambial zone o f the block. The fungus penetrated intact (non-wounded) root bark on each species. However, there was variation in the frequency o f infection among species on the different harvest dates for the first two field trials. In the Hidden Lake trial, Douglas-fir and western hemlock had a higher proportion o f infected roots than western redcedar at 11 weeks (x2, p < 0.01) and 1-year (x2, p < 0. 1) following inoculation, whereas in al l subsequent trials infection on cedar was as frequent as the other conifers. The low frequency o f infection in cedar in the first inoculation trial may have been due to microsite differences or variations in microbial root surface communities among species, or both. Cedar trees selected for inoculations were limited to those within a 20-30-m-wide strip bordering a separate study site and many individual trees within that area had a tendency to be clumped particularly in wetter microsites such as depressions. Excavation of roots on cedar trees growing in these areas revealed soils rich in organic matter and it was not uncommon for roots to be growing through large pieces o f decaying wood. Although roots selected for inoculation were buried with mineral soil obtained from adjacent areas, the difference in soil type and microbial root surface communities may 57 have inhibited infection o f Armillaria at surface contact with the root, particularly with respect to cedar. Morrison (1972) found that rhizomorph production (measured as the cumulative dry weight) was positively correlated with the amount of organic matter in the soil. In fact, in this study it was not uncommon to unearth a profuse network o f A. sinapina rhizomorphs inhabiting the organic matter and the large pieces of decaying wood. General observations suggest that rhizomorphs production from blocks buried in organic soils was as good as from those in mineral soil. In the 2003 trial, infection rates were lower on Douglas-fir and western redcedar trees than on western hemlock trees. Very little infection occurred during the 5 weeks following inoculation, except for 1 hemlock root. In subsequent weeks, more infection occurred, particularly on western hemlock and western redcedar, however, there was no statistically significant difference in the frequency o f infection among species at either 5 weeks or 9 weeks. Douglas-fir roots examined at 9 weeks showed an abundance o f rhizomorphs at root contact and rhizomorphs adhering to and miming underneath outer bark scales. However because Douglas-fir has relatively thick rhytidome tissue, more time was required for the fungus to degrade the outer rhytidome layers and penetrate the inner bark. The summer of 2003 was particularly dry which contributed to the high incidence and severity of forest fires throughout the province. In the study area, the last moderate precipitation event occurred in mid-June, approximately 2 weeks before the 9 week sampling date. Subsequently, soil conditions became especially arid and remained as such until the next decent rainfall in October. Little infection occurred on roots sampled at 9 weeks. In-situ examinations of a few inoculated trees in the weeks following revealed very little evidence o f infection (i.e. lack o f resinosus on the surface o f the root at inoculum contact) and further sampling of roots was postponed until the following year. A t 1-year post inoculation, hemlock showed significantly more infection than Douglas-fir (X 2, p < 0.001) or western redcedar (x 2, p < 0.01). The difference in infection rates between hemlock and the other conifers is not well understood, however the high 58 frequency of infection observed in hemlock after 1 year in the 2003 trial parallels infection rates for that species in the other field trials. The above results are interpreted to mean that the arid soil conditions in 2003 delayed or inhibited infection on host roots. Adverse effects o f seasonal drying on rhizomorph growth in upper soil layers has been reported elsewhere. Morrison (1976) suggests the spread of rhizomorphs in the soil depends in large part on soil moisture and that periods of dry weather may affect growth of A. mellea rhizomorphs in the upper 5 cm of the soil. Pearce and Malajczuk (1990) found that no rhizomorphs grew in dry soil and that periods of dry weather may account for the paucity of rhizomorphs in forest soils of western Australia. Cmickshank et al. (1997) also confirmed this by reporting higher incidence o f Armillaria rhizomorphs on wetter sites in the drier hygrotopes. In this study, although A. ostoyae was viable in inoculum blocks after 1 year, seasonal drying the year prior may have reduced its infective capability and overall incidence of infection. Smith and Griffin (1971) suggested the growing tips o f rhizomorphs tend to melanize in the absence of a fi lm of water. Similarly, the low soil moisture content of the soil in mid-late summer in 2003 probably inhibited rhizomorph growth and development on host roots where they remained latent until more favourable conditions became available for renewed growth. In 2004, two trials located at two geographically separate sites showed similar results with respect to infection frequency on the three conifer species. N o difference in the rate of infection occurred among species at 8-9 weeks, 5 months or 1 year following inoculation at both sites. These results suggest that infection does not appear to occur sooner on some species than others. On all species, infection frequency was related to the amount of time since inoculation. Inoculum blocks were primed in the field before they were buried by disturbing the already melanized rhizomorphs to stimulate new rhizomorph growth. It generally takes several weeks for rhizomorphs to grow and adhere to the root surface before they can penetrate the l iving root. A l l species showed more 59 infection at 5 months than at 9 weeks. N o more infection was observed at 1 year following inoculation than what was observed at 5 months. In future inoculation studies, harvesting roots 4-5 months following inoculation in the spring would be more desirable since adequate time would have passed to allow the fungus to grow from the inoculum block to the root and for the host to initiate defense mechanisms in an attempt to contain the fungus. However, during a period o f summer drought the frequency o f infection may be significantly less. 2.3.2 C H A R A C T E R I Z A T I O N O F H E A L T H Y R O O T B A R K T I S S U E S The macro- and micrograph sample features in Figures 2.6-2.166 are radially sectioned and oriented so that the rhytidome is at or above the top of the figure and the vascular cambium is at or below the bottom o f the figure, unless otherwise stated. The magnification of all macrographs is x 15, and the magnification o f cryofixed sections viewed with bright field (BF), blue light excitation (BL) , and ultra violet excitation ( U V ) is x 35 unless otherwise stated. Other features of interest which are made reference to in the text are indicated with an arrow. Acronyms describing the specific tissues o f interest are defined on page xxv i i i . Macroscopically, fresh samples of the l iving phloem tissue of western redcedar roots were white in colour, the phellem of the outer periderm was a deep red, and all dead tissue external to the last formed periderm was brown (Figure 2.6). In western hemlock, the l iving phloem appeared pink and the phellem of the outer periderm was a dark red-purple colour (Figure 2.7). In Douglas-fir roots, the phloem tissue was pale yellow/orange and the phellem o f the outer periderm was dark brown (Figure 2.8). Typical frozen sections of healthy root bark were viewed with B F , B L excitation, and U V excitation for western redcedar (Figures 2.9-2.11), western hemlock (Figures 2.12-2.14), and Douglas-fir (Figures 2.15-2.17), respectively. The walls of healthy phellogen cells 60 did not fluoresce. In the late spring or early summer when the phellogen was actively dividing, a zone o f non-fluorescent meristematic cells was evident and it was sometimes difficult to distinguish between phellogen and immature phellem because early stages o f differentiation of phellem are nonsuberized and do not yet have pigmented contents. A typical periderm in healthy bark tissue of western redcedar consisted of 1-2 rows o f thin-walled phellem, a single row of phellogen and 1-2 rows of phelloderm cells. The bark surface was invariably smooth. In most coniferous trees, thin-walled phellem cells are typically suberized whereas thick-walled (stone) phellem cells are heavily lignified. However in western redcedar the phellem cells are both lignified and suberized (Figures 2.18 and 2.19). The periderm of western hemlock roots usually had 2-4 rows o f thin-walled phellem, a single row of phellogen and 1-2 rows of phelloderm cells. 1-3 rows o f stone phellem were observed on older larger diameter roots external to the thin-walled phellem. Mature thin-walled phellem were suberized and had dark brown or reddish-purple pigmented contents. Irregularities in the bark surface existed as small bumps on the surface o f the root (Figure 2.7). 61 Fig. 2.6. A sample of root bark from a healthy 19-year-old western redcedar tree. Fig. 2.7. A sample of a root bark from a healthy 19-year-old western hemlock tree. Fig. 2.8. A sample of root bark from a healthy 31-year-old Douglas-fir tree. Fig. 2.9. A cryofixed section of healthy western redcedar root bark, BF. Fig. 2.10. Same section shown in Fig. 2.9., BL. Fig. 2.11. Same section shown in Fig. 2.9, UV. Fig. 2.12 A cryofixed section of healthy western hemlock root bark, BF. Fig. 2.13. Same section shown in Fig. 2.12., BL. Note clusters of sclereids visible in Fig. 2.12 and 2.13. Fig. 2.14. Same section shown in Fig. 2.12, UV. Fig. 2.15. A cryofixed section of healthy Douglas-fir root bark, BF. Fig. 2.16 Same section shown in Fig. 2.15, BL. Note thick-walled stone phellem in Figures 2.15 and 2.16. Fig. 2.17. Same section shown in Fig. 2.15, UV. Arrows in Figures 2.9-2.17 point to the phellogen. Bar on photomacrographs and photomicrographs are 1 mm and 25 nm, respectively, and are applicable to successive photomacrographs and photomicrographs in this Chapter unless stated otherwise. 62 Periderms in Douglas-fir usually had 2-4 rows o f thin-walled phellem, 1-4 rows of stone phellem, a single row o f phellogen and 1-3 rows o f phelloderm cells. Mature thin-walled phellem were suberized and had orange-pigmented contents. Sometimes, Douglas-fir periderms had a band of stone phellem on the outside and several rows o f thin-walled phellem externally abutting the phellogen. The rhytidome was comprised o f dead phellem layers as well as any dead phloem tissue isolated by the last formed periderm (Figure 2.20). Fig. 2.18. A phloroglucinol-HCl-treated section of a healthy western redcedar root. Both the thin-walled phellem (arrow) and phloem fibres stain positively for lignin, viewed under BF. Fig. 2.19. The same section viewed under UV which shows the same phellem cells fluorescence brightly for suberin. Fig. 2.20. A Sudan III stained section of a healthy Douglas-fir root bark showing alternating layers of thin-walled (suberized) phellem shown here as bright orange, and thick-walled (stone) phellem, BL, x 45. 2.3.3 C H A R A C T E R I Z A T I O N O F A B I O T I C A L L Y W O U N D E D R O O T B A R K T I S S U E S Tissue samples were collected at various dates to determine the host response to abiotic wounding in the bark and the length o f time required for wound healing. Wounding was conducted in three of the four inoculation trials: Kingfisher 2003, 2004, and Nakusp 2004. Table 2.3 gives the total number o f roots examined following abiotic wounding in Douglas-fir, western hemlock and western redcedar trees in the different field trials. 63 Table 2.3. The total number of trees sampled and the total number of roots examined following abiotic wounding of Douglas-fir, western hemlock and western redcedar roots at the different harvest dates for each field trial. Roots were initially wounded in early May of each year. Species/Field Trial Date of sample No. of trees No. of roots Lesion length Douglas-fir collection sampled examined (mm) Kingfisher (2003) 5 weeks 3 4 8-20 9 weeks 2 4 11-26 1 year 17 27 10-21 Kingfisher (2004) 9 weeks 2 4 12-22 5 months 8 13 12-28 1 year 3 3 14-26 Nakusp (2004) 8 weeks 2 3 15-20 5 months 2 5 14-25 1 year 1 2 20-21 Western hemlock Kingfisher (2003) 5 weeks 4 4 10-20 9 weeks 2 8 9-18 1 year 17 25 11-22 Kingfisher (2004) 9 weeks 2 4 11-23 5 months 4 8 14-22 1 year 2 5 16-25 Nakusp (2004) 8 weeks 2 2 16-25 5 months 2 3 20-22 1 year 1 1 23 Western redcedar Kingfisher (2003) 5 weeks 8 9 16-25 9 weeks 3 8 12-25 1 year 14 23 10-61 Kingfisher (2004) 9 weeks 2 3 11-20 5 months 9 10 15-25 1 year 4 6 13-24 Nakusp (2004) 8 weeks 1 2 13-34 5 months 3 3 20-25 1 year 1 1 20 64 Douglas-fir After 5 weeks, roots showed a distinct zone of necrotic tissue that extended to a depth of approximately half the bark thickness (Figure 2.21). The freeze-killed tissue appeared as a mass of necrotic tissue that fluoresced yellow-green when viewed under B L (Figure 2.22). Some of these cells were situated along the outer boundary o f a N I T zone. During the early stages of dedifferentiation, a distinct zone o f irregularly-shaped hypertrophied cells with reticulated contents developed internally abutting necrotic tissue. A t the time of sampling (5 weeks post-wounding), a well developed, but narrow zone of N I T was seen under the necrotic zone. Ce l l walls fluoresced bright yellow and all tissue external to NIT, including the freeze-killed phellogen and phelloderm of the original periderm and underlying phloem, also fluoresced bright yellow (Figure 2.22). Staining with phloroglucinol-HCl clearly distinguished the N I T zone from surrounding host tissue because cell walls comprising the N I T zone stain more intensely for lignin. Macroscopically, a clear-white zone of modified l iving tissue appeared internal to the zone of brown hypertrophied cells (Figure 2.23). This zone of modified tissue represented an active, meristematic zone o f phellogen 3-5 cells wide, but narrower at the junction o f the original N P . Cells had thin walls and were non-fluorescent when viewed under B L or U V , indicating a zone of active cell division. Newly formed derivatives were produced in radial files. Single rows o f phellem cells were continuous along the newly restored phellogen and eventually merged with the phellogen cells o f the original periderm. Ce l l walls internal to the N I T zone stained positively for suberin following treatment with Sudan III. After 9 weeks, a N P was fully differentiated and usually comprised o f 1-2 rows of stone phellem and 1-2 rows o f thin-walled phellem. A newly restored phellogen produced stone phellem cells first which were positioned in the central region underneath the N I T zone. These cells typically had thicker walls than the adjacent N I T cells. They were rectangular or square shaped in radial section, were aligned in radial files with thin-walled phellem, and stained very strongly with phloroglucinol-HCl (Figure 2.24). The appearance of the freezing lesion was no different at 5 months than what was observed at 65 9 weeks with the exception of the additional layers of thin-walled phellem cell and narrowing o f the phellogen zone which usually appeared as a single layer o f cells. A t 5 months, coincident with the end o f the growing season, the N P had about 3-4 rows of stone phellem and 4-6 of thin-walled phellem (Figure 2.25). There were no more rows of phellem observed on wounded roots at 1 year than were present at 5 months. Apparently phellogen activity typically ceased during tree dormancy. Fig. 2.21. A sample of Douglas-fir root bark 5 weeks after wounding, x 12. Fig. 2.22. A cryofixed section of abiotically wounded Douglas-fir root bark. Freeze-killed tissue and outer boundary of NIT zone fluoresce bright-yellow green with BL, x 45. Fig. 2.23. A sample of Douglas-fir root bark, 9 weeks after wounding showing distinct zone of clear-white tissue underlying browned tissue, x 12. Note: the original periderm sloughed off during the process of cryofixation. Fig. 2.24. A phloroglucinol-HCl treated section of abiotically wounded Douglas-fir root bark, sampled 5 months after wounding showing dark staining of sieve cells comprising the NIT zone and positive staining for lignin in the thick-walled stone phellem overlying radially compressed thin-walled phellem. Fig. 2.25. A Sudan III treated section of abiotically wounded Douglas-fir root bark, sampled 5 months after wounding showing suberized thin-walled phellem internally abutting two layers of stone phellem, x 45. 66 In larger diameter roots of Douglas-fir, freezing injuries did not extend past the rhytidome layer and therefore did not permit characterization o f the host response to abiotic wounding. On roots where injury did occur, N P formation was complete at 5 weeks in all 4 (100%) of the root samples examined. Only 1 of the 8 (12%) Douglas-fir samples that were examined at 9 weeks showed necrosis extending beyond the zone of incomplete redifferentiation. The sample had initially formed a N I T zone. Phellogen restoration occurred in mid-phloem in the central area of the lesion but a zone of redifferentiation was still developing at the junction o f the original periderm on either side of the lesion. Incomplete phellogen restoration in these areas permitted extension of necrosis into the adjacent phloem. Beyond this zone of necrosis a new zone of dedifferentiating tissue had not yet developed at the time the root was sampled. However, it is conceivable that N I T and N P formation would be complete in approximately 3-4 weeks because according to Mul l i ck (1977) the rate o f phellogen restoration is typically greatest during the summer months. A l l samples collected at 5 months (n=18) and 1 year (n=32) showed NPs to be complete and at the latter harvesting date, many roots displayed en masse sloughing o f the freeze-killed tissue external to the new periderm. Only 1 of the 18 (5%) roots sampled at 5 months showed necrosis extending down to the vascular cambium. Compartmentaliz-ation and callusing was evident around the margin of the wounded cambium. Western hemlock Similar to Douglas-fir roots, western hemlock roots that were abiotically wounded showed distinct zones of necrotic tissue to at least half the thickness of the bark when sampled after 5 weeks (Figure 2.26). Necrotic tissue fluoresced yellow-green with B L . Tissue sections treated with phloroglucinol-HCl showed a well developed N I T zone comprised of hypertrophied and heavily lignified cells internally abutting necrotic phloem (Figure 2.27). A l l roots sampled at 5 weeks (n = 4) had formed a N P with approximately 1-3 layers of thin-walled phellem cells (Figure 2.28). However, N P formation was not always continuous along the periphery of the necrotic zone. N I T and N P typically form last at the junction with the original periderm and discontinuities in 67 periderm development occurred around clusters of sclereids. Sclereids are abundant in the phloem and were readily visible at the macroscopic level (Figure 2.26 and 2.29). A t 9 weeks, a clear zone of modified tissue was visible underlying the necrotic zone (Figure 2.29). Microscopically, this modified tissue showed a well defined meristematic phellogen zone, approximately 2-3 cells wide which did not fluoresce under B L or U V . Newly formed phellem cells accumulated dark red-purple contents and were organized in radial files. After 9 weeks, fully differentiated NPs consisted of up to 5 rows of thin-walled phellem. One to two rows of stone phellem plus 2-4 rows of thin-walled phellem were noticed only on two roots, both of which were more than 7 cm diameter (Figure 2.30). The newly restored phellogen producing single rows of phellem cells was discontinuous around clusters o f sclereids. Often a phellogen was fully restored adjacent to a cluster o f sclereids while a zone o f redifferentiation was still developing directly underneath a cluster of sclereids. Phellem production on either side of the cluster o f sclereids was far more advanced than immediately underneath (Figure 2.31). Eventually, a new phellogen became established around the cluster o f sclereids, however the timing of phellogen renewal around this cell type lagged behind that which occurs in phloem parenchyma. Sclereids essentially became part of the N I T zone as indicated by more intense staining for lignin following treatment with phloroglucinol-HCl. A zone o f dedifferentiated tissue that becomes hypertrophied, impermeable, and heavily lignified serves to seal off the affected region and provide the underlying cells the time and conditions necessary for phellogen restoration to occur. However, because sclereids lack protoplasm (Esau 1965) they also lack the ability to hypertrophy. The cell walls of sclereids are moderately thick and typically contain numerous pits. These anatomical characteristics suggest that in hemlock roots which normally contain an abundance of sclereids in the phloem, periderm formation may be delayed around the periphery of the wound, particularly in the region internally abutting clusters of sclereids. 68 A t 5 months, the phellogen produced organized rows of phellem in neat radial files. U p to 9 rows o f thin-walled phellem were observed at this sampling date (Figure 2.32). Sclereids no longer presented an obstacle to N P formation because tissue had sufficient time to redifferentiate underneath a cluster o f sclereids. However, the delay in phellogen restoration caused by the presence o f sclereids is evidenced by the higher number of phellem cells that occurred on either side of the cluster of sclereids than those which occurred directly underneath. The phellogen zone typically became much narrower at the end of the growing season, usually appearing as a single layer of cells. A number o f lesions on abiotically wounded roots collected at 5 months and 1 year showed en masse sloughing of the ff eeze-killed tissue with an intact periderm established underneath. When the sampling of roots occurred at the 1 year harvest date in late M a y or early June, the phellem produced around the necrotic tissue in the year o f injury appeared radially compressed and the current year's phellogen produced up to 3-5 rows o f thin-walled phellem (Figure 2.33). A higher proportion of roots with kil led cambium following rapid freezing was observed in western hemlock roots than in Douglas-fir roots. Freeze-killed tissue extended down to the vascular cambium on 2 o f 4 (50%) o f the roots examined at 9 weeks and 6 of 11 (55%) o f the roots examined 5 months following wounding in the Kingfisher trial. In all but one sample, compartmentalization and callusing were evident. The one root that lacked evidence o f compartmentalization and callusing was sampled at 5 months. N P formation was initiated but incomplete around clusters of sclereids and necrosis breached through the sclereids and extended down to the vascular cambium. The lack o f host response of the cambial tissue after 5 months suggests that perhaps the extension o f necrosis occurred later in the growing season, coinciding with the onset o f tree dormancy at the time the root was sampled. In this situation, a host response involving compartmentalization would not be expected to take place until growth resumes the following spring. 69 BPh Pe BPh 2.28 Fig. 2.26 A sample of western hemlock root bark, 5 weeks after wounding, x 12. Note clusters of sclereids. Fig. 2.27. A phloroglucinol-HCl treated section of abiotically wounded western hemlock root bark, sampled 5 months following wounding, BF. NIT zone is distinct as the zone of hypertrophied tissue that stains very strongly for lignin. Fig. 2.28. A Sudan III treated section of abiotically wounded western hemlock root bark, sampled 5 weeks after wounding, BL. Fig. 2.29. A sample of a western hemlock root bark, 5 months following abiotic wounding showing a colorless zone of modified tissue internally abutting the necrotic zone indicative of a newly restored zone of phellogen, x 12. Note clusters of sclereids (arrows). Fig. 2.30. A sample of a western hemlock root, 7 cm in diameter, sampled 5 months following abiotic injury. A single band of phellem (stone phellem) can be seen as the clear zone of tissue internally abutting the necrotic zone and externally abutting a zone of thin-walled phellem (dark purple pigmented line). This clear zone comprised a band of thick-walled phellem. In this image the phellogen was separated from the necrotic tissue since OCT (shown in white) occupied the space between the necrotic tissue and underlying phloem, x 25. Fig. 2.31. A cryofixed section of abiotically wounded western hemlock showing incomplete NP formation around clusters of sclereids, BL. Fig. 2.32. A cryofixed section of wounded root bark sampled 5 months following injury showing a well organized NP with up to 9 thin-walled phellem in neat radial files, UV. Fig. 233 A cryofixed section of wounded root bark sampled 1 year following injury, showing at least 4 rows of thin-walled phellem internally abutting the radially compressed phellem produced the year prior, BF, x 45. 70 Western redcedar Abiotically wounded roots of western redcedar sampled at 5 weeks showed necrosis extending to approximately half the thickness o f the bark (Figure 2.34). The freeze-killed tissue fluoresced yellow-green when viewed microscopically under B L (Figure 2.35). One to two rows of thin-walled phellem were formed immediately abutting the necrotic zone (Fig. 2.36) In contrast to Douglas-fir and western hemlock roots, a zone of white modified tissue was not visible in cedar roots on any o f the harvest dates. This was primarily due to the fact that the zone of white-clear modified tissue underlying necrotic tissue usually represents a wide zone o f re-differentiating tissue (a new meristematic phellogen zone) that could expand up to 4 cell layers wide during the late spring or early summer. The phellogen in cedar roots usually appeared as a single layer o f cells that was non-fluorescent when viewed under B L or U V . In cedar, the phellem of the original periderm was typically 1-2 cells wide and unlike the other species studied, the periderm did not accumulate additional layers o f phellem throughout the growing season. N P formation was complete in all samples examined at 5 weeks, as well as at subsequent harvest dates. K i l l e d cambium was found only on 2 of 13 (15%) and 2 of 30 (7%) of the roots sampled at 5 months and 1 year, respectively. In all cases, compartmentalization had occurred, callus tissue was evident around the margin o f the lesion and N P formation was complete in the bark. The chemical nature of cell wal l constituents associated with thin-walled phellem in cedar roots differed from Douglas-fir and western hemlock roots. Whereas thin-walled phellem in hemlock and Douglas-fir appear to be suberized, the same cell type in cedar was both suberized and lignified. Following treatment of cedar sections with phloroglucinol-HCl, the thin-walled phellem o f a newly formed N P stained very strongly for l ignin (Figure 2.37). These phellem cells were not considered to be stone phellem. Stone phellem typically have thick cell walls and are rectangular- or square-shaped in radial section. Examination o f the same sections treated with phloroglucinol-HCl under 71 U V revealed that the same lignified, thin-walled phellem were also suberized (Figure 2.38). Suberization in walls of phellem cells was also confirmed following staining of the sections from the same samples with Sudan III. The N I T zone in wounded cedar bark tissue was very difficult to discern in most o f the sections examined following staining with phloroglucinol-HCl. There is no evidence suggesting N I T formation does not precede N P formation as many studies documenting the early stages o f development in N P formation were carried out on western redcedar trees (Mul l ick and Jensen 1976). However in this study, N I T zones were not well defined. It is possible that N I T formation comprised just a single layer of phloem parenchyma cells and/or fibres, the latter cell types being heavily lignified already. In addition, cell hypertrophy in a single layer o f cells that is radially oriented in alternating rows of phloem parenchyma and fibers may be more difficult to distinguish from cases where a larger number o f cell layers become hypertrophied throughout the inner phloem. In cedar, new immature phellem cells may be produced around the periphery o f a single layer o f impervious tissue and lignin may be detected in both the phellem cell walls and the cell walls of the parenchyma comprising the single layer o f N I T externally abutting the newly formed phellem. Further developmental studies are warranted to confirm that western redcedar does indeed follow the classic model o f non-specific defense in the bark of conifers as defined by Mul l i ck (1977), particularly with respect to NIT formation, but also because many unique and never-before described defense mechanisms are involved in the process of wound healing and resistance against the ingress of pathogens in cedar (see below). Unlike Douglas-fir and western hemlock, cedar did not produce additional phellem layers throughout the growing season after a N P was formed around injured tissue. N o more phellem was observed around root lesions at 1 year than was detected at 9 weeks and 5 months, whereas phellem production on western hemlock and Douglas-fir increased with time since injury. One obstruction to phellogen restoration in the phloem o f cedar bark may be heavily lignified cells, such as fibers. During the process o f phellogen restoration, the cell hypertrophy and dedifferentiation of tissue caused the phloem fibers, 72 which are normally longitudinally aligned in radial sections, to become displaced and this served as an obstacle to continuous phellogen renewal around these displaced cells at the periphery of the wound. The result was a plethora o f phellem wrapped around individual fibers (Figure 2.38). In contrast to hemlock roots in which clusters of sclereids presented a temporary barrier to phellogen restoration, fibers in the phloem of cedar did not affect the rate of phellogen renewal since N P formation was always complete at each sampling date. In addition to N P formation in the bark, a unique phenomenon involving a type of localized response was induced in some roots. This response involved the activation of a cell death program, similar to the process o f apoptosis defined by Fink (1999), in tissue adjacent to freeze-killed phloem. This host reaction was not an immediate response to injury and was predictably induced at or following completion of N P formation in the bark. The induced response was indicative of rhytidome formation resulting in the formation o f a successive periderm, one that is structurally and biochemically the same as a N P ; however, the cellular alterations used to derive this new periderm were quite different. The resulting lesion showed a N P bordering injured tissue and additional periderms initiated in the mid-phloem area that extended proximally and distally for some distance (up to 2 cm) from the primary freezing injury. Eventually, the new periderm merged with the original periderm, forming one continuous periderm layer that separated necrotic from adjacent l iving tissue (Figure 2.39). The author is unaware o f any literature documenting this type o f host reaction and in this study the term "induced rhytidome" w i l l be used to describe the host reaction which involves induced apoptosis and successive periderm formation. Various stages of induced rhytidome formation were observed in 4 of 21 (19%), 2 o f 13 (15%), and 10 o f 30 (30%) o f the cedar roots examined at 8-9 weeks, 5 months and 1 year following wounding, respectively. N o evidence of induced rhytidome formation was observed on roots sampled at 5 weeks. 73 The earliest stage of induced rhytidome formation appeared as an amorphous mass of tissue that underwent hypertrophy and hyperplasia. The hypertrophy, in particular, was excessive and significantly more distended than that normally observed during NIT development at an abiotic wound (Figure 2.40). However, this zone of tissue was non-lignified. During the early stages of development, the zone of hypertrophy and hyperplasia appeared non-fluorescent when viewed under B L or U V indicating a zone of dedifferentiation or an active zone of meristematic activity. Many of these cells were elongated and had reticulated contents, some forming cross-walls. Over time, cell wall fluorescence developed and a narrow zone of re-differentiating tissue occurred internally abutting the excessively hypertrophied phloem. The presence of a single radial row of cells (phellem) indicated that a new phellogen had developed (Figure 2.40). Lignification and suberization of phellem cells preceded the accumulation of pigmented contents. The single row of phellem was continuous along the restored phellogen, extending for some distance, usually up to 2 cm, beyond the margin of the NP and eventually merged with the phellem of the original periderm and the new NP (Figure 2.41 and 2.42). Another unique response in cedar involving traumatic resin duct formation in the phloem was observed in only 2 of 30 (7%) of the roots sampled at 1 year. This response was not detected in any roots on any other date. Traumatic phloem resin duct (TPRD) formation in cedar was detected in a higher number of roots following inoculation with A. ostoyae than in roots with abiotic wounding. In many cases, induced rhytidome and TPRD's were induced simultaneously immediately following NP formation. As the intensity of the responses involving induced rhytidome and TPRD's was significantly higher in roots inoculated with A. ostoyae than in roots with abiotic wounding, a thorough description of these host responses including the early developmental stages and lesion development are presented in the Section 2.3.5 documenting host response to inoculation with A. ostoyae. 74 Fig. 2.34. A sample of western redcedar root bark showing freeze-killed tissue to approximately half the bark thickness and a NP bordering injured tissue. Note pigmented phellem wrapped around phloem fibers, xl2. Fig. 235. A cryofixed section of abiotically wounded western redcedar root showing bright yellow-green fluoresced necrotic tissue and a NP with 2 layers of thin-walled phellem. Fig. 236. A Sudan III treated section of abiotically wounded western redcedar root bark, sampled 9 weeks after wounding, BL. Significantly more phellem production around individual fibers than adjacent areas is seen. Fig. 237. A phloro-glucinol-HCl treated section of wounded root bark, sampled 5 months following wounding, BF. NIT zone is inconspicuous and masked by lignification in the walls of the thin-walled phellem. Fig. 238. The same section as in 2.37 but viewed under UV. Lignified, thin-walled phellem now fluoresce bright blue-violet indicating suberization of cell walls. Fig. 239. A sample of western redcedar root bark showing NP formation around necrotic tissue and successive periderm formation (induced rhytidome formation) in adjacent phloem tissue. The new NP will extend proximally and distally away from the primary lesion to become continuous with the original periderm. Fig. 2.40. A cryofixed section of abiotically wounded western redcedar showing a distinct zone of hypertrophy and hyperplasia associated with the induced rhytidome response. Internal to this zone, a meristematic layer of cells produced a single layer of thin-walled phellem,x45, UV. Fig.2.41. A Sudan III treated section of tissue showing the suberized phellem formed internal to the induced rhytidome merging with the original periderm at the distal end of the lesion, x 45, BL. Fig. 2.42. A cryofixed section of Fig. 2.40 showing a second periderm resulting from induced rhytidome formation deeper in the bark tissue. Eventually all cells external to the last formed periderm become moribund and fluoresce bright yellow-green, BL. 75 2.3.4. C O N T R O L B L O C K S A total of 46 control blocks (uncolonized segments o f Garry oak branchwood) were used in the Kingfisher trial and a total of 24 control blocks were used in the Nakusp trial. Control blocks were placed in early M a y and sampled at 8-9 weeks, 5 months and 1-year later. Table 2.4 shows the number o f control blocks used in both inoculation trials for each species, the number of control blocks that became colonized by on-site inoculum, the percentage of roots that showed lesions at root surface in contact with the control block, their corresponding lesion length and host response to injury Table 2.4. The total number of control blocks in both the Kingfisher and Nakusp field inoculation trials by species, the number of control blocks colonized by on-site inoculum, and the number of roots that resulted in a lesion in the bark at surface contact with the control block. Host Species N o . Control blocks N o . control blocks colonized by Armillaria N o . of roots with lesion at root contact Lesion length (mm) Host response Douglas-fir 27 1 1 10 N P Western hemlock 18 0 3 5-18 N P Western redcedar 25 0 1 5 N P Only one control block was found to be fully colonized by an Armillaria species. A t the time o f sampling (1 year) this block produced fresh rhizomorphs at the cambial edges and through cracks in the bark along the length of the unit. The fungus isolated in culture showed prolific monopodial branched rhizomorphs (Figure 2.43). This isolate was somatically incompatible with isolate 87-01 (A. ostoyae) and produced a typical demarcation line in the agar when tested in dual culture (Figure 2.44). The control block 76 was apparently colonized by on-site A. sinapina inoculum. N o infection was observed in the host root associated with the control block colonized by A. sinapina. In field inoculation trials involving western larch and Douglas-fir, Robinson (1997) reported similar results whereby control blocks became colonized by naturally occurring A. sinapina which subsequently failed to infect the host root. Rhizomorphs of the fungus were mostly superficial on the surface o f the host roots (Robinson 1997). Lesions in the bark at contact with a control block were found only on only 1 o f 27 (4%) and 1 of 25 (4%) o f the Douglas-fir and western redcedar roots respectively. In both cases, the lesion length was usually less than 10 mm. Three of the 18 (16.7%) western hemlock roots treated with the control had shallow lesions, approximately one-third the thickness o f the bark and lesion length varied between 5-18 mm. In all cases, the outer periderm did not appear to be ruptured and any evidence of physical injury to the root was lacking. However, slight browning o f the inner phloem tissue was evident and when examined microscopically, phloem tissue appeared to be slightly hypertrophied and fluoresced bright yellow-green under B L . Further examination o f a western hemlock root sampled at 9 weeks that showed slight browning o f the inner phloem tissue at surface contact with the control block revealed a distinct zone of NIT and 3-4 layers of thin-walled phellem (Figure 2.45). A l l roots that showed abnormal discoloration in the phloem (i.e. browning) formed a N P in response to injury. Fig. 2.43. Monopodially branched rhizomorphs typical of A. sinapina produced in culture from an isolate from the control block. Fig. 2.44. Incompatibility reaction in dual culture between the isolate obtained from the control block and isolate 87-01 used in the field trials. A dark pigmented zone developed in the medium between the two opposing mycelia (white arrow) and radial growth became flattened in the zone of inhibition. Fig. 2.45. A Sudan III treated section of a western hemlock root showing slight browning of phloem tissue at surface contact with a control block after 9 weeks. Phellogen restoration was complete and at least 3-4 layers of thin-walled phellem were formed. BL, x 45. 77 Only 5 of 70 (7%) of the control blocks caused lesions in the phloem underlying the block. Since the large majority of roots treated with controls showed no evidence of lesions in the underlying phloem and no blocks showed evidence of physical disruption of the outer bark tissues, the probability is low that inoculations with A. ostoyae were facilitated by initial injury to the root. Fungal infection usually takes longer than 2 months to occur given the time required for rhizomorphs to grow and adhere to the outer surface of the root. If a root was injured at the time of inoculation, prior to infection by the fungus, it is more likely that a N P would be formed within the first 5 weeks following block placement and the N P would be evident upon examination. It is possible that the mere pressure of a block placed against a root may cause slight injury to a few layer o f cells underlying the phellogen, particularly during times of elevated cambial activity (i.e. at the onset of the annual period of tree growth) when tissues are loosely arranged and conceivably more susceptible to damage. 2.3.5. H O S T R E S P O N S E T O I N O C U L A T I O N W I T H A. OSTOYAE I N R O O T S During this study, infected lesions on several roots of each of the species studied were harvested at various intervals following inoculation with A. ostoyae in order to permit characterization over time o f the host responses to infection and lesion development on roots on the three conifer species (Table 2.2). Infected roots of Douglas-fir and western hemlock commonly exuded resin on the surface of the root (Figure 2.46 and 2.47). Discoloration on the external surface o f the bark was sometimes noticeable, delineating the zone o f necrotic tissue underlying the periderm (Figure 2.47). Western redcedar roots did not produce surface resin in response to infection and discolouration o f the outer bark was difficult to detect. However, infected cedar roots sometimes exhibited irregular swellings usually associated with the induced rhytidome response following invasion by the fungus (Figure 2.48). The mode of penetration by A. ostoyae rhizomorphs in all hosts appeared to follow the generally accepted pattern described by Thomas (1934). Rhizomorphs adhered to the 78 outer surface of the root and formed a lateral branch that penetrated the phellem as a larger unit (Figure 2.49). Typically, the phellogen remained unharmed (i.e. still functional) until the penetrating mycelia degraded the most recent layers of phellem and exposed the l iving meristem. However, at times there was evidence o f suppression of phellogen activity in the area underlying a penetrating rhizomorph and/or stimulation o f phellogen activity immediately adjacent to the penetrating rhizomorph (Fig. 2.50). This reaction was observed more often on Douglas-fir than in other conifers. The phloem tissue underlying a penetrating rhizomorph appeared deformed indicative o f mechanical pressure exerted on those cells during the initial penetration phase. Infection also resulted without direct penetration by a rhizomorph. These reactions were typically seen as small necrotic lesions in the bark and always occurred directly underneath the cambial end of the inoculum block (Fig. 2.51). Infections o f this type were also associated with resin. Both mechanisms of entry (direct and indirect) resulted in the collapse of underlying host cells. Approximately 36% (n=143) o f the total number of roots sampled in the 2003 and 2004 field trials showed indirect penetration of the fungus whereas the majority of lesions on roots were rhizomorph initiated. However, considerably more western hemlock and western redcedar roots showed evidence o f indirect penetration by A. ostoyae than did Douglas-fir. Whitney et al. (1989) also reported non-rhizomorph infections in conifers following inoculation with A. ostoyae. The above results are interpreted to indicate that the healthy bark o f roots w i l l react to toxic substances produced by the fungus (i.e. extracellular cell-wall degrading enzymes or toxic metabolites, or both) exuding from the cambial end o f the inoculum block as it situated against a root. Similarly, Wargo (1983) suggested that extracellular secretions o f laccase and peroxidase causes the browning o f tissue in advance of a penetrating mycelia. Once the fungus penetrated the periderm, the phellogen and underlying phloem were killed. Lateral and tangential proliferation o f mycelial fans or wedges of mycelium were seen colonizing inner bark tissue and often advanced to the vascular cambium (Figure 2.52). The fungus always kil led tissue in advance o f the mycelium and lateral spread o f mycelia in the bark always preceded cambial colonization (Figure 2.53 and 2.54). 7 9 Fig. 2.46. A photomacrograph of an infected Douglas-fir root showing resin exudation (arrows) on the surface of the root 1 year following inoculation with A. ostoyae. Fig. 2.47. A photomacrograph of an infected western hemlock root showing resinosus (arrows) on the surface of the root 5 months following inoculation with A. ostoyae. Note darker discoloration of the bark showing the extent of necrosis in the underlying phloem. Fig. 2.48. A photomacrograph of a western redcedar root showing uneven irregularities on the surface of the roots. Note rhizomorph adhering to the outer surface. Fig. 2.49. A cryofixed section of a Douglas-fir root showing a lateral branch of a rhizomorph penetrating the outer cork layer, BF. Fig. 2.50. A cryofixed section of a Douglas-fir root showing a rhizomorph adhering to the outer bark surface. A narrow zone of phellogen activity (typically 1 cell layer wide) was seen in the area immediately underlying the rhizomorph (arrow) compared to a more active phellogen zone (2-4 cells wide) on either side of the rhizomorph, BL. Fig 2.51. A western hemlock root showing a distinct zone of necrotic tissue resulting from inoculation with A. ostoyae. Fungal mycelia are absent and necrosis likely resulted from secretions of fungal enzymes/toxins from the cambial surface of the inoculum block. Fig. 2.52. A photomacrograph of an Armillaria-lesioa on a rjtouglas-fir root showing mycelial fans and the extent of cambial necrosis in the root. Fig. 2.53. A mycelial fan invading the phloem of a western hemlock root. Distinct browning of tissue occurs ahead of mycelial colonization. Fig. 2.54. A Douglas-fir root showing progressive browning of host tissue ahead of the penetrating mycelial fan. 80 2.3.5.1. CHARACTERIZATION OF THE DIFFERENT STAGES OF HOST RESPONSE TO INFECTION Following invasion of A. ostoyae in host roots, trees attempt to contain the fungus by initiating defense response involving NP formation in the bark and/or barrier zone formation in the wood. The ability of trees to successfully contain the fungus will depend on the speed of formation and efficacy of these barriers and the degree of interference by the fungus. This process, to some degree, depends on the relative susceptibility of the host, host genetics, the inoculum potential of the fungus, and environmental factors such as low temperature and drought stress. Susceptibility occurs when the pathogen is able to overcome non-specific defense responses associated with wound healing. A key event in the development of a new periderm barrier in the bark is the early formation of an impervious zone in the vicinity of infected tissue. Deposition of lignins and cell wall thickening occur concomitantly in a zone of dedifferentiating tissue, called NIT in this thesis. If the host fails to trigger the development of NIT, the fungus is commonly seen to be advancing in the inner bark tissue with no visible host response. The classification of reaction as 'no or ineffective host response' in this thesis was done on the basis of tissues either lacking cell hypertrophy altogether or appearing as irregularly hypertrophied but ineffective lacking distinctive microscopic characteristics indicative of a zone of NIT including strong lignification in cells walls following staining with phloroglucinol + HCI and cell wall fluorescence in the zone of hypertrophy. A zone of NIT can be formed initially, but then the fungus can breach the barrier either directly through mature NIT or by circumventing the zone of NIT where dedifferentiation of tissue is incomplete (i.e. through clusters of sclereids, at the junction of the original periderm, or at the junction with the vascular cambium). The same can be said for NP formation. In most cases, NIT and NP do not form synchronously along the whole necrotic boundary zone. The cells and subsequent tissues form first in the mid-phloem between the wound edge and the vascular cambium and last in the region of the original phellogen. Therefore, within the same histological section, one can often view wound tissues ranging from no visible host reaction to those exhibiting NIT regeneration and NP 81 formation. In addition, any breaching o f barriers detected in samples harvested on different dates may not be the final outcome in terms o f lesion development, particularly for those samples harvested within the first 5-11 weeks following inoculation which presumably would be the early stages o f infection on the host roots. It is possible that continual formation and breaching o f barriers may occur before the outcome of the infection event is conclusively determined. Nevertheless, roots harvested at 5 months, coincident with the end o f the growing season, and 1-year following inoculation should give a good indication of the probable progression o f the fungus in host tissue in the early stages o f infection and its ability to contain the fungus to a lesion on the root. Cambial invasion occurs in one of three ways: (1) when the fungus progresses through the inner bark and precludes the development of NIT or N P , eventually reaching the cambium; (2) when N I T is initiated, then breached before a N P develops and the fungus advances to the vascular cambium; (3) when N I T and N P are formed, then breached either directly through the newly formed barrier or where N I T and N P become discontinuous around, clusters of sclereids or at the junction o f the vascular cambium. Cambial invasion takes time. In this study, A. ostoyae ki l led the vascular cambium on 11 of 40 (27%) of infected roots harvested at 5-11-weeks and on 70 of 151 (46%) of roots harvested at 5-months or later. 2.3.5.2. M O D E L O F H O S T - P A T H O G E N I N T E R A C T I O N S Figure 2.55 illustrates a schematic representation of the non-specific responses associated with wounding or pathogenic invasion of the bark and cambial tissue and the variation in the type o f host response induced with the depth o f injury on the species examined in this study. This model o f non-specific host response is a modified and expanded version of that presented by M u l l i c k (1977) in Figure 1.1. Fig . 2.55.1 Fig . 2.55.2 Fig. 2.55. Schematic diagram showing the anatomical model for non-specific defense mechanisms induced following (Fig. 2.55.1) shallow injury to the living bark where the injury is limited to the vicinity of the living phellogen and underlying phloem, (Fig. 2.55.2) deeper injury to the living bark where necrosis extends to tissues in close vicinity, but not directly affecting the vascular cambium, showing modified vascular cambium within the zone of callus and (Fig. 2.55.3.) injury to living phloem, vascular cambium and functional sapwood. Successive letters denote a sequence of time consisting of several weeks. Abbreviations: B, bark tissue; C T , callus tissue; K T , killed tissue; NP, necrophylactic periderm; P, original periderm; V C , original vascular cambium; V C 1 , new position of vascular cambium; X , original xylem; X 1 , xylem formed since injury D. decayed wood; E , empty space; T C , xylem tissue derived from callus; RD, traumatic resin ducts. Note: NIT formation was indiscernable in western redcedar. 84 The first processes shown in Figure 2.55.1 (a) shows a shallow injury to the l iving bark causing a disruption of the l iving phellogen and is similar to that reported by Mul l i ck (1977) in Figure 1.1 (a). The normal and simplest response following initial penetration o f the first few cell layers o f the l iving bark involves the development of N I T and then a N P around the periphery of the wound which has been described in detail above. This new barrier is usually formed first in the central region o f the wound (2.55.1 (b)) and eventually becomes continuous with the original periderm and its normal function is resumed (Figure 2.55.1 (c)). The second process shown in Figure 2.55.2 (a) arises from deeper injuries to the bark, closer to the vascular cambium. Similar to that described by Mul l i ck (1977) in Figure 1.1 (b), deeper injuries in the bark may cause a disruption in the normal cambial activity without any direct injury to the cambium itself. In the central region, the original vascular cambium forms callus tissue instead of normal tracheids and ray parenchyma while xylem and phloem derivatives continue to be produced on either side o f the injured tissue (Figure 2.55.2 (b)). Callus (undifferentiated, isodiametric parenchyma) arises from increased divisions o f the cambial intials as well as recent phloem derivatives. In contrast to what Mul l i ck (1977) describes, sporadic zones of NIT and N P may be initiated in the bark at the margin of kil led tissue prior to or concurrent with vascular cambium restoration. The N P develops either uniformly or non-uniformly along the periphery of the wound. In the latter case, a N P is usually formed last in the area contiguous with the original periderm. Eventually, the altered vascular cambium resumes normal activity to produce normal xylem and phloem cells and may do so even while a new N P is still differentiating (Figure 2.55.2. (c)). Tissue derived from callus remains embedded within the xylem and over time develops secondary walls and becomes lignified. Sometimes, semi-differentiated or fully differentiated traumatic resin canals may be formed in place of callus in the modified cambial zone in species like Douglas-fir and western hemlock (not shown). These resin canals may extend further tangentially from the modified cambial zone. Hence, when the cambial and xylem surface remains intact underneath a zone of necrotic bark, callus or resin ducts w i l l develop across the periphery of the affected tissue. 85 The third process shown in Figure 2.55.3 (a) involves deep injuries to the sapwood in which vascular cambium and the outer xylem (ray parenchyma) below the dead vascular cambium are also kil led. The interpretation o f the sequence o f events involved in wound healing (e.g. callus formation and regeneration o f vascular cambium) is an expanded version of that described by Mul l i ck (1977) in Figure 1.1. In addition to N P formation in the bark, the l iving vascular cambium at the periphery of the wound w i l l form callus by progressive hyperplasia o f parenchyma cells in the uninjured cambial zone (Figure 2.55.3 (b) ). Callus originates mainly from the uninjured vascular cambium immediately adjacent to kil led cambial tissue, but sometimes phloem parenchyma cells in close proximity to the vascular cambium can also participate in the formation of callus. Dead tissue w i l l remain between the two callus curls and sometimes as the wounded bark tissue is lifted as a result o f tissue expansion over the face o f the wound, an empty space w i l l occur. Callus curls w i l l eventually merge over the wounded cambium which can be seen as a zone of undifferentiated parenchyma external to the kil led cambium (Figure 2.55.3 (c) ). N I T formation followed by phellogen restoration occurs in the outer part of the callus tissue and meristematic tissue forms as a replacement o f the vascular cambium in the inner part o f the callus, usually starting in the callus tissue contiguous with the detached ends o f the original vascular cambium and then eventually meeting in the middle (Figure 2.55.3 (d)). Soon after the formation of these meristems, derivative tissues are produced as phellem and phelloderm to the outside and inside of the phellogen, and phloem and xylem elements to the outside and inside o f the vascular cambium. In some species like Douglas-fir and western hemlock, traumatic resin ducts (TD) may be produced instead of normal tracheids in xylem (Figure 2.55.3 (b-d)). In others like western red cedar, a zone o f axial parenchyma is formed in the xylem immediately adjacent to the kil led cambium (not shown). The tissue derived from callus in the xylem area develops secondary, lignified cell walls. A t this point, this tissue no longer participates in active division, except perhaps at the outermost edge o f the lesion or the callus curl where cells may still be non-lignified and actively dividing. In most lesions, the callus tissue bridges the surface o f the wound and the newly regenerated vascular cambium w i l l form a uniform meristematic layer. Some callus tissue located in the modified cambial zone w i l l become part of the phloem tissue itself. The tissue 86 derived from callus which ends up in the phloem is more difficult to detect than callus tissue which remains embedded in the xylem. N o specific tests were conducted to determine the extent o f non-conductive sapwood as reported in Mul l i ck (1977) although it's assumed that following direct injury to the vascular cambium, trees w i l l undergo the process of compartmentalization according to the C O D I T model proposed by Shigo and Marx (1977). Douglas-fir Roots o f Douglas-fir typically showed an initial host reaction involving resinosus in the bark which commonly exuded to the outer surface o f the rhytidome. Large wedges o f mycelium were often found interspersed with clumps o f resin in the necrotic tissue. However, the presence o f resin did little to deter the fungus from advancing to adjacent host tissue. The fungus kil led host tissue ahead o f mycelial colonization and on several occasions advanced rapidly enough to preclude the initiation of any host defenses leading to N P formation. A t any one sampling date, a percentage o f Douglas-fir roots infected with A. ostoyae showed no visible or ineffective host response as the fungus advanced in the bark. In the Hidden Lake trial, host reactions were not initiated on 2 of 9 (22%) and on 3 o f 10 (30%) of infected Douglas-fir roots sampled at 11 weeks and 1 year, respectively (Appendix III). In subsequent trials, between 50-100% (n=36) of Douglas-fir roots that showed initial penetration by the fungus showed no host response to infection. The adjacent phloem appeared brown and either lacked significant hypertrophy or appeared irregularly hypertrophied. Host reactions which failed to develop N I T typically showed a distinct zone of necrotic tissue. The boundary between the necrotic zone and the apparently healthy tissue at the infection front was either abrupt (Figure 2.56) or diffuse (Figure 2.57). In more advanced stages o f infection, the fungus decayed tissue in the area underlying the initial site of penetration and necrosis continued to expand in the bark and cambial zone (Figure 2.58). 87 When viewed under fluorescence, infected tissue appeared as an amorphous mass of cells with bright yellow-green cell walls (Figure 2.56). Progressive lesions commonly showed more advanced stages of tissue degradation. This stage was characterized by a zone of cells devoid of contents with bright yellow cell walls bordering a zone of moribund cells with sporadic hypertrophy and intercellular spaces stained yellow-brown (Figure 2.59). Despite the hypertrophy in the zone of necrosis or at the infection front, sections did not stain positively for lignin following phloroglucinol-HCl treatment which would otherwise indicate NIT development. The occurrence of NIT in infected Douglas-fir roots was quite variable. Early stages of NIT development were recognized under BL by cell wall fluorescence along the outer perimeter of the zone of dedifferentiation. However, this was rarely observed, probably due to the timing of sampling. Distinct zones of NIT were more recognizable following staining with phloroglucinol-HCl or under fluorescence as the abrupt zone of cell hypertrophy immediately external to a zone of redifferentiation (Figure 2.60). Even after 1 year, NIT developed in only 15 of 34 (44%) of the roots sampled. Since NIT invariably precedes NP formation, more than half the infected Douglas-fir roots had not contained their lesions in the bark. The low frequency of NIT formation in roots infected with A. ostoyae permitted the fungus to advance to the cambium. Of the 34 roots sampled at 1 year that showed successful penetrations, bark tissues in 13 (38.2%) were killed to the vascular cambium and none of these showed any evidence of compartmentalization and callusing (Appendix III). In Douglas-fir, sporadic lignification of adjacent phloem tissue was sometimes seen in advance of a penetrating mycelium. However, distinct zones of NIT were frequently lacking or the fungus penetrated beyond the developing NIT or both. Breaching of NIT was frequently observed in Douglas-fir roots. NIT generally developed first in the mid-phloem area underlying or adjacent to a zone of necrotic tissue and formed last at the junction of the original periderm. Breaching of NIT usually occurred in areas where zones of dedifferentiation were incomplete (e.g. at the original periderm) or deeper in the bark tissue at the junction with the vascular cambium. 88 Fig. 2.56. A cryofixed section of a Douglas-fir root inoculated with A. ostoyae showing an abrupt demarcation between necrotic and adjacent healthy tissue at the infection front. Necrotic tissue has cell walls that fluoresce bright yellow-green under BL. Fig. 2.57. Another cryofixed section from an infected Douglas-fir root showing a diffuse boundary between necrotic and adjacent healthy tissue at the infection front, BL. Fig. 2.58. A photomacrograph of a Douglas-fir root showing a large wedge of mycelium invading the bark and cambial zone. Fig. 2.59. A cryofixed section of Douglas-fir root showing a progressive lesion at the infection front. Cells in the adjacent phloem appear moribund as cells show sporadic hypertrophy and intercellular spaces are stained yellow-brown, BL. Fig. 2.60. A cryofixed section of a Douglas-fir root showing the early stages of redifferentiation as a zone of non-fluorescence internal to a zone of hypertrophied tissue (NIT), BL; x45. In tissue sections showing breaching o f NIT, NIT typically appeared as an incomplete zone of hypertrophied, heavily lignified parenchyma and farther browning of cells was observed deeper into the phloem. Interestingly, the number o f cell layers involved in N I T development was noticeably higher in tissue infected with A. ostoyae (Figure 2.61) compared to that which develops following abiotic wounding (Figure 2.62). The above results are interpreted to indicate that the increased frequency o f lignified cell layers involved in N I T development may be a delayed response to N P formation under the influence o f an advancing fungus. Wahlstrom and Johansson (1992) reported similar results in Pinus sylvesteris seedlings mechanically wounded and inoculated with A. 89 ostoyae. More cell layers were lignified in roots challenged by A. ostoyae, than non-challenged roots (Wahlstrom and Johansson 1992). The breaching o f N I T may depend on its speed o f formation relative to the speed at which the fungus or the toxins produced by the fungus moves through host tissue. In this study, all species showed breaching o f N I T although breaching was consistently more frequent in Douglas-fir and western hemlock than western redcedar (Appendix III). N P formation in Douglas-fir roots occurred in only 20 of 54 (37%) of roots that had successful penetrations. A newly restored phellogen produced layers o f thin-walled phellem that ranged between 2-4 cells wide at 11 weeks (Figure 2.63), 3-6 cells wide after 5 months, and 4-7 cells wide after 1 year (Figure 2.64). Stone phellem was not always associated with N P ' s , however when it was present, periderms had between 2-6 layers of thick-walled phellem with thin-walled phellem internally abutting it. More than half o f the NPs that initially formed on roots were breached by the fungus. The resulting lesion exhibited either an abrupt or diffuse boundary o f moribund and healthy phloem. Cells internal to breached barriers displayed irregular hypertrophy and fluorescent yellow-green cell walls when viewed under B L . Following the breaching o f N P ' s , second attempts at N I T formation were seen in only a few roots as a faint staining for lignin in the adjacent hypertrophied phloem. However, in all cases the host was unsuccessful at containing the infection within a new periderm barrier within the time frame from which it was sampled (i.e. at 1 year). The host might have attempted N P formation in the subsequent growing season, although i f the fungus was rapidly advancing in the inner bark and cambial tissue, changes in inoculum potential would most certainly occur as the fungus utilizes more and more host tissue to the point where the fungus might overwhelm any type o f host response to contain the infection. The fungus may colonize the root distally and infection may or may not be checked by the host when the fungus reaches the junction o f a larger diameter root or the root collar (Figure 2.65). 90 Overall, A. ostoyae advanced to and kil led the vascular cambium in 20 of 54 (37%) of the Douglas-fir roots that showed successful penetration by the fungus. In al l situations, cambial invasion was observed only on those roots harvested at 5 months and 1 year following inoculation (Appendix III). Therefore, cambial invasion in Douglas-fir depended on time since infection. O f the number of roots with kil led cambium only 2 o f 20 (10%) showed evidence of compartmentalization and callusing. Cross-sectional views o f a barrier zone formed in a Douglas-fir root 5 months following inoculation revealed that the cambium was injured mid-season as evidenced by the location o f the traumatic resin canals in the middle of that year's annual growth ring (Figure 2.66). Traumatic resin canals typically occurred in tangential bands. Epithelial cells surrounded the large lumen of individual resin ducts and the cells surrounding the traumatic resin ducts had pigmented deposits. Ray parenchyma adjacent to resin ducts in the vicinity of the injured cambium also had pigmented contents. Resin duct formation was disorganized in the area adjacent to where a new vascular cambium differentiated within the callus (Figure 2.67). The tracheids formed after the tangential series of resin canals showed similar disorganization. Several rows o f tracheids in the earlywood zone were slightly disoriented and some cells appeared to be occluded. The tracheids formed by the newly differentiated cambium became more organized with increasing distance from the callus edge. The xylem underlying the area o f kil led cambium showed slight discoloration. A. ostoyae interfered with the initiation o f active defense mechanisms involving N P formation and compartmentalization in more than half the infected Douglas-fir roots. The remaining roots responded by forming N I T and N P adjacent to the advancing fungus in the bark. However, breaching of these barriers and absence o f compartmentalization in roots following cambial invasion enabled the fungus to progressively develop in the roots. 91 Fig. 2.61. A phloroglucinol-HCl treated section of a Douglas-fir root inoculated with A. ostoyae showing a distinct zone of NIT comprised of several cell layers, BF. Fig. 2.62. Phloroglucinol-HCl treated section of abiotically wounded Douglas-fir root showing the lignified NIT zone comprising only 1-2 cell layers followed by three layers of stone phellem, BF. Fig. 2.63. A Sudan III treated section of Douglas-fir root 11 weeks following inoculation with A. ostoyae showing a newly restored periderm with up to 4 layers of thin-walled, suberized phellem. Note resin blister in the adjacent phloem with epithelial cells also staining positively for suberin, BL. Fig. 2.64. 1-year following inoculation with A. ostoyae, a newly restored phellogen produced up to 7 layers of thin-walled phellem. Note the stone phellem externally abutting the thin-walled phellem, BL. Fig. 2.65. A photomacrograph of a Douglas-fir root with its bark removed showing a smaller secondary root colonized by the fungus distally and the infection checked (compartmentalized) at the junction of the larger diameter root. Fig. 2.66. A paraffin embedded section of a Douglas-fir root showing a barrier zone formed by the uninjured cambium comprised of a tangential series of traumatic resin ducts following invasion by A. ostoyae. The surrounding axial and ray parenchyma appear occluded with polyphenolic bodies. The vascular cambium (not shown) lies above the tracheids in this micrograph. Fig. 2.67. Immediately adjacent to the killed vascular cambium, resin duct formation appears disorganized and is comprised of polyphenolic-rich axial parenchyma; deposits also accumulate in the ray cells. The vascular cambium (not shown) lies above the tracheid and ray parenchyma in this micrograph. 92 Western hemlock Resinosus on the surface o f the bark was also a common symptom of an A. ostoyae infection in roots of western hemlock trees (Figure 2.68). Similar to Douglas-fir, resin soaked phloem apparently did little to deter advance o f the fungus in the host. On several root samples, the fungus appeared to have advanced fairly rapidly as evidenced by the extent of lateral spread in bark tissue, thin mycelial fans, and a low frequency o f roots that initiated defense responses leading to N P formation in the bark. Phloem necrosis always occurred ahead of mycelial colonization (Figure 2.69 and 2.70). Fig. 2.68. A western hemlock root showing resin exudation and rhizomorphs on the root surface following inoculation with A. ostoyae. Fig. 2.69. A photomacrograph of western hemlock root showing rhizomorph penetration and necrosis of the inner bark. Fig. 2.70. A cryofixed section of Fig. 2.69 showing the large mycelial fans degrading and digesting the phloem tissue following invasion by the fungus, BL. The proportion o f hemlock roots showing no host response at the infection front varied between sampling dates and field trials (Appendix III). Overall, 29 o f 80 (36%) o f western hemlock roots showed no visible host response following penetration by A. ostoyae (refer to earlier definition o f 'no host response' on page 80). Cel l hypertrophy in roots lacking any visible host reaction was variable. The host either showed no cell hypertrophy at the infection front, irregular hypertrophy, or significant hypertrophy in advance o f a penetrating mycelium (Figures 2.71-2.73). Roots showing massive zones of bark hypertrophy did not stain positively for lignin following treatment with 93 phloroglucinol-HCl, nor was there any cell wal l fluorescence detected in the outer boundary o f that zone which otherwise might indicate the early stages o f N I T formation. The boundary between the necrotic zone and the adjacent, healthy or moribund phloem was always quite abrupt (Figure 2.74). When viewed under B L or U V , the large zone of ki l led tissue fluoresced bright yellow-green or blue-violet, respectively (Figures 2.75 and 2.76). In most instances where the host showed no visible host response, the fungus had advanced to and killed the vascular cambium. Fifty-one o f 80 (63%) o f western hemlock roots examined throughout this study developed N I T in response to invasion by A. ostoyae. Figures 2.77 and 2.78 show a typical zone o f N I T in hemlock as an amorphous mass o f hypertrophied cells underlying necrotic (browned) tissue. After a zone of dedifferentiation became established, cells internally abutting N I T redifferentiated to form a meristematic phellogen which appeared non-fluorescent when viewed under B L (Figure 2.79). L ike Douglas-fir, the N I T comprised several layers o f phloem cells that stained positively for lignin following treatment with phloroglucinol-HCl (Figure 2.80), considerably more layers than were detected in roots with abiotic wounds. Breaching o f NIT was common and was observed on several sampling dates. It occurred at higher frequency than in Douglas-fir (Appendix III). A number of infected hemlock roots showed large zones o f induced lignification in phloem tissue and at times, lignification extended for some distance along the tissue sample (Figure 2.81). These observations suggest that the host triggered NIT development and as mycelium advanced, phloem tissue increasingly underwent dedifferentiation in an attempt to form an impervious zone of tissue. 94 NIT RZ APh 2.79 Fig. 2.71. A photomacrograph of a western hemlock root following inoculation with A. ostoyae showing a lack of hypertrophy on the proximal infection front and significant hypertrophy at the distal infection front. Note distinct zone of necrosis extended to the depth of the vascular cambium. Mycelial fans and large wedges of resin can be seen throughout. Fig. 2.72. A cryofixed section of the proximal infection front shown in Fig. 2.70 showing a lack of hypertrophy at the infection front (no host response) in advance of a penetrating mycelium, U V. Fig. 2.73. A cryofixed section of the distal infection front shown in Fig. 2.70 showing significant hypertrophy in the adjacent phloem and small zones of redifferentiated tissue as clusters of phellem embedded within the phloem, UV. Fig. 2.74. A photomacrograph of a western hemlock root showing rhizomorph penetration of the inner phloem. The boundary between the necrotic tissue and adjacent healthy tissue is quite abrupt. Fig. 2.75. A cryofixed section of Fig. 2.74 showing no visible host response at the infection front (lack of cell hypertrophy in the adjacent phloem), BL. Fig. 2.76. The same cryofixed section shown in Fig. 2.75 but view under UV. Fig. 2.77. A cryofixed section of a western hemlock root following inoculation with A. ostoyae showing a distinct zone of dedifferentiated tissue (NIT) internally abutting a zone of necrosis, BF. Fig. 2.78. The same section shown in Fig. 2.77 but viewed under BL. Note cell wall hypertrophy in the hypertrophied phloem and polyphenols deposits occurring in phloem parenchyma cells. The zone of redifferentiation is not obvious here. Fig. 2.79. Another cryofixed section of a western hemlock root showing a distinct zone of redifferentiation internally abutting a zone of NIT. The newly restored phellogen zone appears non-fluorescent in BL. 95 The above results are consistent with those reported for Scots pine seedlings infected with A. ostoyae (Wahlstrom and Johansson 1992). In this study, the lack of adjoining N I T in the mid- and outer-phloem region allowed the fungus to advance in the bark. Sporadic lignification in hypertrophied phloem tissue was seen underlying infected phloem tissue since this is the area in which N I T is usually initiated. However, distinct zones of NIT were absent at the junction of the original periderm, which enabled the fungus to circumvent the developing N I T and continue to advance in the inner bark. In addition, N I T was regularly incomplete in areas occupied by large clusters o f sclereids (Figure 2.82) and necrosis commonly extended deeper into the phloem immediately underlying clusters of sclereids (Figure 2.83). O f the total number o f hemlock roots sampled, more than half formed at least a partial N P . After 11 weeks, the newly restored phellogen had produced 1-3 layers of thin-walled phellem (Figure 2.84). Mature phellem commonly developed pigmented contents and cell walls stained positive for suberin after treatment with Sudan III (Figure 2.85). The number o f phellem cells produced by the phellogen ranged between 2-6 and 3-9 after 5 months and 1 year (Figure 2.86), respectively. The wide range of phellem cells produced at the latter harvesting dates was likely a reflection of the timing of infection and the ability of the host to trigger defense reactions leading to N I T and N P formation while under the influence of the fungus. Stone phellem was observed in only 3 roots that formed a N P . In all three cases, these roots were harvested 1-year following inoculations and root diameters ranged between 3.5-5.5 cm. N P ' s showed a single band o f stone phellem (1-2 cells wide) followed by 1-3 rows of thin-walled phellem (Figure 2.87). Robinson (1997) suggested that N P ' s with multiple bands o f thick- and thin-walled phellem were a structural characteristic that helped impart increased resistance to the spread of A ostoyae in roots of western larch. In this study, lesion sizes on hemlock were noticeably smaller on roots with a band o f thick- and thin-walled phellem compared to roots with N P ' s consisting o f only thin-walled phellem. 96 Field observations showed small necrotic lesions underlying the cambial edge o f the inoculum block where it was situated against the root. It is assumed that there is no penetration by single hyphae and that penetration occurs primarily by mechanical and enzymatic degradation by rhizomorphs. However, in several cases, a penetrating rhizomorph was absent. The necrosis likely resulted from fungal toxins or exudates affecting tissue in the area immediately underlying the inoculum block. Lesions on such roots were commonly contained within a new N P barrier and lesion lengths were considerably smaller compared to lesions where the fungus was present as a large wedge of mycelium or advancing mycelial fan in the bark or cambial zone. The above results suggest that the lack o f the physical presence o f the fungus in the host may reduce the inoculum potential of the fungus because fungal toxins may be acting upon the host's defense mechanisms to a lesser degree, enabling the host to have a higher probability of forming a N P than it would under the influence o f an advancing mycelium. Rykowski (1975) reported that further stages o f penetration of the phloem parenchyma depend not only on host reactions induced in response to infection but also on the morphological form of the infecting mycelium. For example, i f after penetration, the mycelium does not keep the rhizomorphous structure, and instead invades the phloem in the form of single hyphae, the fungus may encounter a lysigenous zone that inhibits further ingress o f the pathogen and allows host defenses to wall out the pathogen (Kusano 1911 as cited in Rykowski 1980). In this study, visible mycelium occurred as larger units (i.e. mycelial fans/wedges or rhizomorphs) and no further investigation was conducted to determine whether individual hyphae were observed in lesions resulting from indirect penetration. Fifty-one percent (n=43) o f hemlock NPs that initially formed in roots were breached by the fungus (Figure 2.88 and 2.89). Internal to breached periderms a rather diffuse boundary between the necrotic tissue and the adjacent phloem was evident. Adjacent to an expanding zone o f necrosis, cells appeared hypertrophied and walls fluoresced bright yellow-green under B L . Again, two possible areas o f weakness where the fungus was 97 able to circumvent the developing or newly formed periderm were identified: at the junction with the vascular cambium and around clusters of sclereids (Figure 2.83). Oven et al. (1999) suggested that neither phloem rays nor groups o f sclereids in the bark of European beech lessened the ability of the trees to form a zone of impervious tissue and N P following injury to the bark. Histological evidence in this study revealed that sclereids become part of the N I T zone, as indicated by more intense staining for lignin compared to sclereids found in adjacent healthy tissue. However, staining of secondary phloem parenchyma cells internal to groups of sclereids was not always evident. Impermeability testing of these sclereids comprising the N I T zone was not conducted in this study. Despite the enhanced lignification in sclereids within the zone o f NIT , it appeared that fungal toxins were able to pass presumably through the apoplastic pathway or numerous pits found in the cell walls causing necrosis deeper in the bark. These results suggest that the presence o f sclereids in the phloem o f hemlock delays or results in discontinuities in NIT development and N P formation, allowing the fungus to grow through the bark before these boundaries are complete. A second attempt at N P formation was not observed on any roots examined on the different harvest dates. A. ostoyae advanced to and kil led the vascular cambium in 37 of 80 (46%) of western hemlock roots showing initial penetration by the fungus. O f those roots with ki l led cambium, only 3 of 37 (8%) showed successful compartmentalization and callusing from the margin o f the lesion. The formation o f callus tissue is a normal reaction to repair injured cambia. The cataplasmic structure of callus (i.e. undifferentiated, isodiametric parenchyma) originates as a primary reaction following wounding o f the vascular cambium. In the uninjured cambium, elongated fusiform initials are usually the first cells that segment transversely and then segregate into isodiametric parenchyma cells (Fink 1999). However, since most l iving cells are totipotent (i.e. have the ability to reset their developmental program and differentiate into new cell types), a variety o f cell types in the xylem and phloem may actively contribute to the formation o f callus. 98 Fig. 2.80. A phloroglucinol-HCl treated section of a western hemlock root inoculated with A. ostoyae showing lignification of phloem parenchyma underlying necrotic tissue, BF. Fig. 2.81. A phloroglucinol-HCl treated section of an infected hemlock root showing the lignified NIT comprising several cell layers and extending for some distance along the length of the sample close to the vascular cambium, BF. Fig. 2.82. A cryofixed section of a western hemlock root inoculated with A. ostoyae showing incomplete differentiation of NIT around clusters of sclereids in the bark, BF. Fig. 2.83. The same section shown in Fig. 2.82 viewed in BL. Note erratic hypertrophy and cell wall fluorescence in moribund tissue and adjacent phloem in the areas of incomplete dedifferentiation of NIT. Fig. 2.84. A cryofixed section of a western hemlock root 11 weeks following inoculation with A. ostoyae showing a typical resistant reaction involving the complete formation of a NP around infected, necrotic tissue, BL. Fig. 2.85. A Sudan III treated section of a hemlock root 11 weeks following inoculation with A. ostoyae showing the suberized phellem of the new NP becoming continuous with the original periderm, BL. Fig. 2.86. A cryofixed section of a western hemlock root sampled approximately 1 year following inoculation with A. ostoyae. The resulting lesion was bound by a NP in which the thin-walled phellem appear to be radially compressed. At the time of sampling in the late spring, the current year's phellogen activity had already produced up to 5 layers of phellem. Fig. 2.87. A cryofixed section of a larger diameter hemlock root showing a newly formed NP comprised of thick and thin-walled phellem, UV. Fig. 2.88. A photomacrograph of a hemlock root showing initial NP formation in the bark and breaching of the newly formed periderm and necrosis extending down to the vascular cambium. Fig. 2.89. A cryofixed section of a hemlock root showing incomplete differentiation of NP around clusters of sclereids, BF. 99 On injvjred hemlock roots, both the l iving xylem ray cells and the cambial initials of the uninjured cambium contributed to the formation of callus tissue (Figures 2.90-2.93). Progressive hyperplasia of cells created a large zone o f callus tissue that expanded over the surface o f the wound. Callus that formed in areas suffering from temporary disruption in normal cambial activity eventually developed secondary cell walls and became lignified and the resulting lesion appeared as a zone of hypertrophied cells in the middle o f an annual ring (Figure 2.94). The barrier zone in western hemlock was similar to that in Douglas-fir. The uninjured cambium formed a series of traumatic resin ducts that extended for some distance tangentially away from the primary infected tissue (Figure 2.95). Some of the first resin ducts that differentiated adjacent to the area o f kil led cambium sometimes showed fusion of individual canals to form larger cavities. The longitudinal extent of resin canal formation in such lesions were not measured. However, Cruickshank et al. (2006) reported that traumatic resin canals could be traced up to 110 cm proximally along an infected Douglas-fir root from a more distal lesion. In red pine (Pinus resinosa Ait . ) roots showing compartmentalization, polyphenolic parenchyma cells formed first immediately adjacent to the ki l led cambium and resin ducts were seen extending tangentially away from the kil led cambium (Tippett and Shigo 1980). Thus, physiologically active parenchyma that accumulate secondary metabolites that are associated with traumatic resin ducts in the barrier zone o f some conifers may be more important in resisting spread o f the fungus along the vascular cambium than traumatic resin ducts by themselves. In this study, parenchyma and tracheids surrounding the resin ducts appeared occluded and somewhat distorted in shape. Ray parenchyma formed between individual resin ducts also showed some occlusions (Figure 2.96), but became less occluded with increasing distance from the kil led cambium. Tracheids laid down immediately following vascular Fig. 2.90. A photomacrograph of a western hemlock root showing lateral ingrowth of callus following cambial invasion by A. ostoyae. Note newly differentiated NP in the outer periphery of the callus. Fig. 2.91. A cryofixed section of western hemlock sample shown in Fig. 2.90 but viewed in BL. A NP became established in the outer periphery of the callus and a new vascular cambium was restored within the callus. Derivatives of the cambial initials are disoriented in radial section. Fig. 2.92. Callus formation originates via progressive hypertrophy and hyperplasia of cambial initials from the uninjured cambium as well as proliferation of the living xylem ray cells; section stained in phloroglucinol-HCl. Fig. 2.93. The same section viewed in Fig. 2.92 but under 45X magnification showing more clearly the proliferation of xylem rays. Fig. 2.94. A phloroglucinol-HCl treated section of a western hemlock root that showed temporary disruption of the normal cambial activity resulting in callus formation. The tissue derived from callus (CT) eventually develops secondary walls, become lignified, and remain embedded between tracheids in the annual growth ring. Fig. 2.95. A western hemlock root sampled approximately 1-year following inoculation with A. ostoyae showed a series of traumatic resin canals following injury to the vascular cambium. Cells surrounding the resin ducts appear occluded. Fig. 2.96. A cryofixed section of a western hemlock root showing callus formation at the edge of the killed cambium. Early differentiation of resin ducts show an oblique orientation but their structure becomes increasingly normal with increased distance from the area of killed cambium, BL. 101 cambium regeneration, particularly in the area where the callus curl appeared to be growing over the lesion, were often obliquely or horizontally oriented but as growth continued, normal alignment of cells was restored. A new periderm was formed in the outer periphery of the callus. Traumatic resin ducts were also induced in the xylem in the absence of cambial invasion, primarily where the fungus was superficial in the bark. Even when the cambium was not directly affected (i.e. killed), stimuli triggered host defense mechanisms which caused a disruption to the normal cambial activity and formation of traumatic resin ducts. Incomplete closure of the wound is common in trees that have sustained injuries to the vascular cambium as a result of either mechanical wounding or pathogenic invasion. I f the vascular cambium turns inward during the process of wound healing, continued growth over a long period of time w i l l enable the host to form thick rhytidome tissue which may prevent physically co-mingling o f vascular cambium and a "bark inclusion" may be visible as a fine crack between converged callus ribs (Fink 1999). Such reactions on roots infected with A. ostoyae may more likely result in species that have a slower rate of healing or individual trees with low tree vigor. A t the time of sampling, no injuries to the cambium were completely healed over by callus. A. ostoyae interfered with the initiation o f host defense mechanisms. Although a higher percentage o f hemlock roots triggered the formation of NIT and N P than Douglas-fir, weaknesses in the development o f these barriers allowed the fungus to breach or circumvent them in the bark. A low frequency of compartmentalization and callusing was observed on roots following cambial invasion. This frequency was not all that different from Douglas-fir. Thus, the low frequency o f resistance reactions observed in roots following inoculation with A. ostoyae indicates that western hemlock and Douglas-fir are equally susceptible to infection by the fungus. 102 Western redcedar Following infection by A. ostoyae, western redcedar roots showed different and less frequent external symptoms o f infections on roots. Unlike Douglas-fir and western hemlock, cedar roots did not show resinosus at lesions. Rhizomorphs were observed on the surface of the root and at times irregularity in the otherwise smooth bark surface was indicative of A. ostoyae lesions (Figure 2.97). A s cedar has relatively thin, fibrous bark, the outer periderm was readily breached. Several roots showed a lateral branch o f a rhizomorph penetrating the outer phellem via mechanical and enzymatic degradation (Figure 2.98). A s a result, the underlying phloem cells collapsed giving the appearance of compressed tissue (Figure 2.99). Penetration o f the l iving bark rendered the phellogen non-functional and necrosis expanded into the underlying phloem (Figure 2.100). Only 1 of 57 (2%) cedar roots examined in this study showed no visible or ineffective host response. This percentage is in strong contrast to the 36% (n=80) and 54% (n=54) of roots showing no host response in hemlock and Douglas-fir, respectively (Appendix III). On this single root harvested at 5 months, sporadic hypertrophy was observed in phloem cells, however neither NIT nor N P was observed following staining of sections with either phloroglucinol-HCl or Sudan III. NIT was presumably initiated in nearly all samples examined on each harvesting date. However distinct zones of N I T were difficult to discern. Whereas the N I T in Douglas-fir and hemlock comprised several layers of lignified phloem underlying necrotic tissue or at an advancing infection front, the N I T formed in cedar never comprised more than 1 cell layer; however in most cases, a distinct zone o f N I T was lacking. It is customary during N I T development for secondary lignification to occur in phloem tissue. This was most easily noticed when phloem fibers which normally stain positive for lignin following treatment with phloroglucinol-HCl showed more intense staining for lignin within the vicinity where NIT would be expected to lie (i.e. externally abutting newly differentiated phellem) (Figure 2.101). However, lignification in a zone o f N I T may have also be 103 masked by the fact that thin-walled phellem cells were not only suberized, but also lignified (Figure 2.102). The above results are interpreted to indicate that the more intense staining of phloem fibers may in fact be caused by the thin-walled phellem which typically wrap around phloem fibers (Figure 2.103 and 2.104) rather than the fibers themselves assuming a greater role in the formation o f NIT . Fibers, like sclereids, have thick secondary walls, but are considerably more elongated and usually occur in tangential rows in the secondary phloem (Esau 1965). It may be that lignification in the cell walls of a single row o f parenchyma occurs only on the side o f the wall that borders a newly differentiated phellogen. When phellem cells are mature they develop deep reddish-purple contents, and their walls become suberized and lignified. Given the fact that cedar produces only 1-2 layers o f phellem compared to several layers of thin-walled and possibly thick-walled phellem like that found in Douglas-fir and western hemlock, it is not unreasonable to suspect that the N I T zone itself would also be minimal. N P formation occurred in 56 of 57 (98%) o f cedar roots inoculated with A. ostoyae (Figure 2.105). A newly restored phellogen produced a layer of phellem that was typically 1-2 cells wide after 8-9 weeks (Figure 2.106). Surprisingly, the number of phellem layers observed in a N P was essentially the same at 5 months or 1 year following inoculation than was observed at 8-9 weeks (Figure 2.107). This differed considerably from Douglas-fir and hemlock, where the number o f phellem in a N P depended on time since phellogen renewal. 104 Fig. 2.97. A western redcedar root sampled 1-year following inoculation with A. ostoyae. Symptoms of diseased roots are quite inconspicuous other than irregularities in the bark surface. Fig. 2.98. A cryofixed section of a cedar root showing rhizomorph penetration of the inner bark. Fig. 2.99. A cryofixed section of a western redcedar root showing rhizomorph penetration via enzymatic degradation and mechanical ingress of the fungus resulting in the collapse of the underlying phloem. Fig. 2.100. A photomacrograph of a cedar root showing expansion of necrosis in the bark following infection by A. ostoyae. Fig. 2.101. A phloroglucinol-HCl treated section of a western redcedar root showing lignification of the fibers and thin-walled phellem. The fiber internally abutting the newly formed phellem stained more strongly for lignin, BF. Fig. 2.102. The same section shown in Fig. 2.101 but viewed under UV. The same lignified phellem here appear suberized. Fig. 2.103. A phloroglucinol-HCl treated section of a cedar root showing NP formation in the bark in response to infection by A. ostoyae. The NP is comprised of 1-2 layers of thin-walled lignified phellem, BF. Fig. 2.104. The same section shown in Fig. 2.103 but viewed under UV showing the same lignified phellem fluorescing bright blue-violet indicating suberization. Note prolific formation of phellem around fibers in the vicinity of the newly differentiated periderm. Fig. 2.105. A photomacrograph showing a typical resistance response involving NP formation in the bark following invasion by A. ostoyae. Fig. 2.106. A cryofixed section of cedar root bark showing necrotic tissue bound by a NP. The phellem accumulate red pigments, BF. Fig. 2.107. The same section shown in Fig. 2.106 but viewed under BL. 105 A unique localized host reaction was induced in cedar following inoculation with A. ostoyae. Adjacent to an area o f infected tissue where the host initially formed a N P , large zones o f phloem parenchyma became markedly hypertrophied, increasing bark thickness by 2-3 times normal (Figure 2.108). Large zones appeared non-fluorescent suggesting hyperplasia and meristematic activity. The walls o f hypertrophied cells appeared thin and some cells developed cross walls. Cells within the zone o f excessive hypertrophy underwent changes in fluorescence characteristics which were indicative of the early stages o f tissue dedifferentiation (Figure 2.109). A new periderm was formed on either side of the lesion, and extended for some distance beyond the primary wounded tissue bound by a N P (Figure 2.110). This particular host reaction resembled rhytidome formation and was observed in 27 o f 57 (47%) o f cedar roots that showed penetration by the fungus. However, this particular host reaction differed from normal rhytidome formation with respect to the intensity of the hypertrophy and hyperplasia occurring in tissue and the extent to which the successive periderm extended along the length o f the root to the point at which it became continuous with the original periderm. The new periderm that establishes underneath the zone o f hypertrophy and hyperplasia (Figure 2.111) either joined with the initially formed N P at or near its deepest point, or else it laid wholly below (and to either side) o f the initially formed N P . In both cases it joined the pre-existing periderm some distance distal and proximal to the infection point. The intensity o f the localized response that was expressed depended on the extent o f colonization o f the bark tissue by the fungus. For example, the lesion size or the length at which induced rhytidome formation occurred beyond the obvious infection front depended on the extent o f necrosis in the bark. Thus for shallow bark injuries to one-third or one-half the bark thickness, the host usually responded by forming a N P around the necrotic tissue. Slightly deeper injuries also induced N P formation; however, an expanding zone of necrosis stimulated the induced rhytidome response. In roots showing extensive colonization of the l iving bark upon initial penetration by the fungus, induced rhytidome formation extended for up to 20 cm, either proximal or distal from the lesion. Roots showing this type of response were sometimes recognized prior to sampling by massive swelling o f the outer bark around inoculum contact (Figure 2.112). This particular host reaction is coined 'induced rhytidome' in this thesis. 106 Apoptosis, otherwise known as a type of cell death program induced in plants, is sometimes described in association with a hypersensitive response that may be triggered by toxic microbial products or elicitors of host origin (Kacprzak et al. 2001). Several studies have shown that mycotoxins secreted by fungal species are capable of inducing apoptosis in plant and animal tissues (Gilchrist 1998, Heath 1988, Khurana et al. 2005). In this study, cedar responded to invasion o f the l iving tissues by initiating a cell death program in the specific form o f induced rhytidome formation to restrict further penetration of the pathogen. In no case did A. ostoyae penetrate beyond the induced rhytidome layer. Hence, when cedar bark was injured, the host formed a necrophylactic periderm to confine the infection and then induced rhytidome formation on either side of the lesion so that one continuous rhytidome layer would be formed and eventually be sloughed from the surface of the root. Roots sampled at 5 months and 1 year following inoculation already displayed en masse sloughing o f necrotic tissue, particularly in the zone of excessive hypertrophy (= induced rhytidome). Thus it appears that induced rhytidome formation not only served to confine the fungus within a new periderm barrier, but it also facilitated sloughing o f infected tissue from the surface of the root thereby reducing the risk of subsequent infections from inactive lesions. Induced rhytidome formation was only observed in two roots from the first inoculation trial. A t the time of examination, these roots did not display the pronounced effect o f bark hypertrophy seen in many o f the other cedar root samples in subsequent years. However, it did show early development of a new N P differentiating underneath necrotic phloem proximal to the area of bark necrosis (Figure 2.113 and 2.114). Re-examination of the micrographs of these samples revealed anatomical changes in the bark and changes in fluorescence characteristics indicative of induced rhytidome formation, similar to what was observed in subsequent years. 107 Fig. 2.108. A photomacrograph of a western redcedar root showing induced rhytidome formation in the bark. The excessive bark hypertrophy results in an increase in bark thickness up to 3X the original thickness of the bark. Fig. 2.109. A cryofixed section of a western redcedar root showing changes in florescence characteristics in the zone of excessive hypertrophy indicative of the early stages of dedifferentiation. Fig. 2.110. A photomacrograph showing induced rhytidome formation in cedar extending for some distance beyond the primary wounded tissue (the original NP). Fig. 2.111. A cryofixed section of a western redcedar root showing excessive hypertrophy and dedifferentiation in the phloem proximal and distal to the primary wounded tissue. A new meristematic phellogen differentiates immediately abutting the zone of hypertrophy and produces radial files of thin-walled lignified phellem. Fig. 2.112. A western redcedar root showing massive swelling and irregularities in the otherwise smooth bark surface following inoculation with A. ostoyae. Note rhizomorphs adhering to the outer surface of the bark. Fig. 2.113. A cryofixed section of a western redcedar root showing successive periderm formation deeper in the bark under a zone of bark hypertrophy, BF. Fig. 2.114. The same section shown in Fig. 2.113 viewed under BL. The newly formed periderm is comprised of 1-2 layers of thin-walled phellem. 108 Another unique phenomenon induced in cedar following inoculation with A. ostoyae was the formation of traumatic resin ducts in the phloem (Figure 2.115 and 2.116). Western redcedar trees form neither normal nor traumatic resin canals in the xylem and resin canals are rarely observed in healthy phloem root tissue. The formation of traumatic resin ducts in western redcedar has not been described in the literature. However, similar defense responses have been reported in another conifer belonging to Cupressaceae, namely in Japanese cypress trees (Yamada et al. 2003, Yamanaka 1984, 1989). Various stages of development in the formation o f traumatic phloem resin ducts (TPRD) were observed in numerous root samples infected by A. ostoyae. Because resin ducts do not form synchronously, early stages o f development in the formation o f T P R D ' s , as well as fully differentiated resin ducts were observed. This permitted a complete characterization o f the various stages of development in T P R D ' s in western redcedar roots. Detailed examination of cryofixed bark samples and serial sections of paraffin embedded cedar bark revealed that T P R D ' s arise by reorganization of preexisting phloem parenchyma in close proximity to the vascular cambium. Assuming that the youngest mature parenchyma cells are physiologically, more sensitive to stimulation, expansion and division of cells leading to the formation o f T P R D ' s may occur preferentially in this zone. Unlike induced rhytidome, this zone o f cells did not appear to be excessively hypertrophied. However, it appeared to undergo hyperplasia as evidenced by the numerous small cell derivatives within a relatively narrow zone of tissue. Excessive proliferation o f cells, followed by schizogenous and lysigenous separation of cells, resulted in a series of longitudinal resin ducts arranged in tangential bands in the inner and mid-phloem region. Stages of development of traumatic phloem resin ducts and their corresponding photomicrographs are as follows: 109 1. A zone of pre-existing axial and ray phloem parenchyma cells, closest to the vascular cambium, undergoes hyperplasia and begins to divide and expand radially, forming small regularly spaced clusters of redifferentiating tissue (RZ) (Figure 2.117). 2. The zones of meristematic activity are seen as large areas of non-fluorescence when under B L (Figure 2.118). Recent derivatives repeatedly divide periclinally such that tissue in this meristematic zone is comprised of numerous small parenchyma cells which are capable of further division. 3. Cells situated in the centre of a mass of hyperplasic tissue wil l begin to separate schizogenously (separation of cells occur along their middle lamellae forming intercellular spaces). Schizogenous separation may also occur concomitantly with lysigeny (dissolution of the common middle lamellae), expanding to create a cavity in the central region (Figure 2.119 and 2.120). 4. The space in the canals becomes enlarged over time via additional periclinal and also anticlinal divisions of parenchyma cells surrounding the canals. Schizogeny and lysigeny subsequently follow. 5. Cells lining the newly formed canals become epithelial cells (Figure 2.121). These cells typically have thin walls and participate in the synthesis of resin. 6. Eventually, circular or elliptic resin canals are formed in tangential rows, separated by phloem ray parenchyma (Figure 2.122). Phloem fibers may appear in the zone of redifferentiating tissue (Figure 2.123). The size of a resin canal varies depending on its developmental stage, although, a fully differentiated resin canal is approximately 100-300 pm in diameter and between 2-5 mm in length (Figure 2.124). 7. Where TPRD's are formed in close proximity to a newly differentiated NP , epithelial cells surrounding the lumen of a resin duct may accumulate phlobaphenes and cell wall constituents, namely lignin and suberin that are normally found in phellem (Figure 2.124-2.127). This is particularly evident where displaced phloem fibers in the zone of phellogen restoration merge with parenchyma cells involved in the formation of TPRD's . Parenchyma cells surrounding the resin duct may be influenced by metabolic and restorative processes occurring within neighbouring cells. However, with increasing distance from a N P , resin ducts appear normal. 8. In transverse section, resin canals appear as neat tangential rows (Figure 2.128) and in tangential longitudinal section, sections may appear as an anastomosing network of resin ducts (Figure 2.129). 110 2.118 Fig. 2.115. A photomacrograph of a western redcedar root showing a NP formation in the bark and a series of small circular or elliptical zones of tissue (arrows) which differentiated into traumatic resin canals in the phloem. Fig. 2.116. A cryofixed section of cedar root bark shown in Fig. 2.115 showing fully differentiated resin ducts in the phloem surrounding the margin of the lesion. Fig. 2.117. A cryofixed section of cedar root bark showing the initiation of traumatic resin duct formation in the phloem via excessive hyperplasia of phloem parenchyma either in mid-phloem or closer to the vascular cambium, BL. Fig. 2.118. Zones of actively dividing cells are non-fluorescent when viewed under BL. Fig. 2.119. A cryofixed section of cedar showing the cavity of a fully differentiated resin duct resulting from schizogeny and lysigeny of the expanding hyperplasic tissue, BF. Fig. 2.120. The same section shown in Fig. 2.119 but viewed under BL. x45. Fig. 2.121. A paraffin embedded section of a cedar root showing epithelial cells lining the lumen of the resin duct, BF. Fig. 2.122. A paraffin embedded section showing traumatic resin ducts separated by axial and ray phloem parenchyma, BF. Fig. 2.123. A cryofixed section showing phloem fibers fixed in the resin ducts. Note phellem-like cells wrapped around fibers, BL. I l l Fig. 2.124. A tangential section of a cedar root showing traumatic resin duct formation in the phloem following inoculation with A. ostoyae. The length of resin ducts varies but is generally between 2-5 mm, while the width ranges between 100-300 um, BF. Fig. 2.125. A cryofixed section of a cedar root showing phloem fibers fixed in the resin duct with pigmented phellem-like cells wrapped around the fiber, UV. Fig. 2.126. A photomacrograph of a cedar root showing NP formation and traumatic resin duct formation deep in the phloem tissue. The tissue surrounding the differentiating resin ducts appear as small circular or elliptical zones of necrotic tissue bound by phellem-like (arrows) cells that accumulate the normal cell wall constituents, namely lignin and suberin, as well as the pigmented phlobaphene cell contents. With increasing distance from the area of primary wounding where a NP had differentiated, traumatic resin ducts appear normal. Fig. 2.127. A paraffin-embedded section of cedar root bark showing traumatic resin canal formation in the phloem adjacent to a newly differentiated NP. The parenchyma cells surrounding the lumen of the duct acculumated phellem-like pigments, BF. Fig. 2.128. A transverse section of cedar root bark collected from the root collar area of a healthy tree showing tangential series of traumatic resin canals in the phloem, BF. Fig. 2.129. A tangential section of cedar root bark adjacent to a newly differentiating NP showing an anastomosing network of resin ducts of varying lengths and irregular accumulation of phellem-like pigments in the epithelial and/or parenchyma cells lining the resin ducts, BF. 112 Yamada et al. (2002) found that traumatic resin canals were induced in the functional phloem of Chamaecyparis obtusa Endlicher (Hinoki cypress) trees following artificial wounding. The authors also reported the presence of traumatic resin canals in the phloem in diseased stems of C. obtusa infected with the Rooshi pitch canker (Cistella japonica Suto et Kobayashi). Early stages o f resin canal formation were evident after just one month (Yamanaka 1984). In this study, the formation o f traumatic resin ducts in the phloem was not an immediate host response to wounding. Its initiation usually coincided with the completion o f N P around the primary wounded tissue. Hence, T P R D ' s would not be evident until at least several weeks following initial penetration by the fungus. N o cedar roots inoculated with A. ostoyae sampled at 5, 8, 9 or 11 weeks showed evidence o f T P R D ' s in the bark. However, T P R D were observed in 43% (n=23) and 75% (n=12) o f the infected cedar roots sampled at 5 months and 1 year following inoculation, respectively. Only 3 of 30 (10%) abiotically wounded roots showed evidence of T P R D ' s in the bark after 1 year. The lack o f both induced rhytidome formation and T P R D ' s observed in the majority o f the freezing injuries may be due to the shallow depth of the bark injuries which are more often walled out by a single N P . Yamanaka (1989) reported that within 11 days o f wounding in the phloem o f C. obtusa, the conducting system for resin develops within the proliferating parenchyma cells. Nagy et al. (2000) also suggested that traumatic resin duct formation in the bark requires up to 2 weeks for formation of structures with secretory capacity. In this study, the margin of the lesions was clearly defined at 5 months and 1 year following inoculation with A. ostoyae, and T P R D ' s extended up to 6-cm proximal and distal from the point where the periderm associated with induced rhytidome became continuous with the original periderm. Presumably the function o f resin synthesis in infected roots had already transpired and the fungus would be interacting not only with chemicals involved in the process o f phellogen renewal associated with N P and induced rhytidome formation, but also with resin synthesis in response to the wounding. There has been some work suggesting that resin canals induced under pathological stimulation contain a resin with 113 quantitatively and qualitatively different composition compared to regular resin canals (e.g. the resin may have a higher concentration o f diterpene resin acids and monoterpene olefins) (Gref and Ericsson 1985, Fink 1999). A x i a l resin duct activation was induced by methyl jasmonate in "incipient" phloem resin ducts o f Cupressus macrocarpa Gord. (Hudgins et al. 2004). Dormant axial resin ducts in the secondary phloem of other conifers responded to similar chemical stimuli that mimics part of the wound response by activating epithelial cells lining the resin ducts to produce resin (Hudgins et al. 2004). It is interesting that T P R D ' s in cedar occurred proximally, distally and tangentially around the periphery of the Armillaria-cansed lesion. Lesions on cedar typically do not girdle the root. Since the development o f T P R D ' s occurs in conjunction with or immediately following the development of a N P to confine the fungus to a lesion in the bark, the presence of these T P R D ' s in the phloem may play some secondary role in terms of limiting the spread o f the fungus in root tissue. Yamanaka (1984) reported that the metabolic response o f resin synthesis and flow may be promoted secondarily by the same stimuli on a continuous basis and epithelial cells can remain active for years. Hence, i f A. ostoyae infection induced traumatic phloem resin canals, further stimulation by an expanding zone o f necrosis or an advancing mycelium would interact with mechanisms already in place for resin synthesis which may cause lysis of fungal material. It is well known that resin production in Douglas-fir and western hemlock does little to deter the advance and colonization o f host tissue by A. ostoyae. Precisely i f or how oleoresin in cedar differs from that found in Douglas-fir of western hemlock and the nature of the biochemical induction o f terpene synthesis in cedar with respect to its resistance against A. ostoyae warrants further investigation. In this study, field observations on a small sample of cedar trees revealed that resin canals in the phloem were usually associated with older tissue located at or near the root collar (Figure 2.128). This was not surprising since the minute amount o f resinosus observed on cedar trees usually occurs at the base of the stem. These observations suggest that on 22-28-year old cedar trees, resin canals are normally found at the root collar but at approximately 25-30 cm from the root collar no resin canals are usually found. 114 However, when roots are injured, T P R D ' s may be induced in large numbers in phloem tissue that normally lacks resin canals. Yamanaka (1984) reported similar results in the phloem of C. obtusa, in association with the 'shoe-string rot', Armillaria mellea where resin canals appeared to be concentrated in the lower stem. In this study, resin canals in the phloem seem to occur naturally in older tissue and they are frequently formed in roots infected with A. ostoyae, and to a lesser extent following abiotic injury. Thus, the response appears to be non-specific. Like the induced rhytidome, the intensity o f the response with respect to the induction and the distal and proximal expansion of T P R D seemed to depend on the extent of injury to the bark. Observations on infected roots of cedar suggested that the expansion of resin ducts occurs far beyond the periphery of the induced rhytidome (i.e. 20 + cm beyond the already expanded lesion). The mechanisms by which such responses are transmitted to distant tissues are unknown. However, Franceschi et al. (2000) suggests signals through polyphenolic parenchyma (PP) cells and interconnected radial rays may be one means by which this occurs. Conceivably, this could also be attributed to fungal elicitors or stimuli as A. ostoyae colonizes and advances in host tissue. Yamanaka (1984) reported that the formation o f phloem resin canals in the branches of C: obtusa was accelerated by the insertion of pins coated with paraquat or with cycloheximide. Cheniclet (1987) also showed higher levels o f terpene production in tissues following fungal invasion compared to abiotic wounding. Hence, the distance over which these resistance responses are induced in roots inoculated with A. ostoyae, over and above that which was observed in abiotically wounded roots, suggests that some chemical stimuli specific to the host-pathogen interaction in cedar has a fundamental and key role in the resistance of that host species to A. ostoyae. Ninety-two percent (n=24) of cedar roots that showed kil l ing to the cambium exhibited compartmentalization o f the infection and displayed active callusing around the margin of the lesion (Figure 2.130). Callus was formed at the margin of the lesion by progressive hyperplasia of both cambial initials and phloem parenchyma located in close proximity to the vascular cambium. In the outer periphery of the callus tissue a N P developed, while 115 in the inner- or mid-region of the callus a new vascular cambium regenerated, probably differentiating from the point where the original vascular cambium was still intact and amalgamated with the original vascular cambium (Figure 2.131). Eventually, the tissue derived from callus developed secondary cell walls and became lignified. Examination of the barrier zone induced in cedar following cambial invasion revealed some unique characteristics. The anatomy o f cedar wood parallels that o f Douglas-fir and western hemlock with respect to normal tracheid and ray parenchyma tissue. However, unlike the other conifers, cedar does not form traumatic resin ducts in the xylem following injury o f the cambium. Instead, a high density o f axial parenchyma was formed adjacent to the kil led cambium which appeared occluded with various deposits. The chemistry of the secondary metabolites that accumulate in this tissue and their biological significance in relation to wound defense, specifically against A. ostoyae, remains to be determined. However, several reports suggest that a complex range o f polyphenolic compounds are involved that are implicated as having an antimicrobial role against pests and diseases (Eyles et al. 2003, Woodward and Pearce 1988). Secondary metabolites appeared to accumulate in the parenchyma cells derived from callus as well as the parenchymatous zone formed by the uninjured cambium restricting the spread of infection to adjacent healthy cambial tissue (Figure 2.132). The intensity o f the response diminished with increasing tangential distance from the area of kil led cambium. Like in other conifers, tracheids derived from a newly differentiating vascular cambium in cedar may be oriented in a variety o f planes (radial, tangential, and transverse), however tissues became progressively more organized with increasing cell division and distance from the callus edge (Figure 2.133). Cedar appeared to have a superior ability to compartmentalize and develop protective boundaries than the other conifers. This notion was supported by the fact that cedar roots were rarely girdled by the fungus and that longitudinal necrosis in the cambium was considerably less in cedar than in hemlock or Douglas-fir. The wood underlying the area of kil led cambium showed slight discoloration, however deposits in the ray parenchyma 116 in the wood were less frequent than in those formed after injury. The high frequency of compartmentalization observed in cedar roots may be attributed to the barrier zone comprised of a higher density of axial parenchyma cells that accumulate deposits (presumably phenolic in nature) which act as a biological deterrent to the fungus (Figure 2.134). Comparison of Douglas-fir, western hemlock and western redcedar Susceptibility has been defined as the 'inability to defend itself against, or to overcome the effects of invasion by a pathogenic organism' (as cited in Bos and Parlevliet 1995). Thus susceptible hosts do not possess the ability to impede attack (growth or development) o f a parasite (Bos and Parlevliet 1995). In this study, resistance or susceptibility was determined by a trees' ability to complete N P formation or compartmentalization, or both, while under the influence o f the fungus within the frame in which the tree was sampled. Susceptible reactions included situations where there was no visible host response in the bark or cambial region, breaching o f either N I T or N P and lack o f subsequent barrier formation, or evidence of compartmentalization and callusing but while the fungus was still considered progressive in the bark. Resistance refers to the ability of a host to hinder growth and activity o f a parasite and encompasses a range o f mechanisms including resistance to the ingress, establishment and spread of the parasite itself (Bos and Parlevliet 1995). Resistant reactions were classified as such when the infection was contained within a N P or when sections o f woody tissue showed barrier zone formation in the wood and callusing at the margin of the wound. Table 2.5 shows the frequency of successful resistance reactions following invasion by A. ostoyae for Douglas-fir, western hemlock and western redcedar at the different harvest dates in each field trial. Chi-square analyses o f the frequency o f the different stages/events involved in host reactions and the frequency o f successful resistance reactions by al l three species was only attempted where a sufficient number o f instances (i.e. predicted frequency o f the phenomenon in question > 1) was available for analysis. A s stated earlier, cases where 117 Fig. 2.130. A photomacrograph of a cedar root showing effective compartmentalization and callusing following cambial invasion by A. ostoyae. Note rhizomorph embedded within phloem tissue. A NP formed around necrotic tissue in the bark and lateral ingrowth of callus is evident at the margin of the killed cambium. Fig. 2.131. A cryofixed section of leading callus edge of the lesion shown in Fig. 2.130 viewed under polarized light. A NP developed in the outer periphery of the callus tissue and a new vascular cambium regenerated within the callus (shown as a darker zone of tissue) and amalgamated with the original vascular cambium. Fig. 2.132. A typical barrier zone in cedar is comprised of a higher than average number of axial parenchyma that accumulate polyphenolic deposits. The zone of polyphenolic parenchyma tissue is generally confined to a small zone of tissue immediately adjacent to an area of killed cambium and does not extend tangentially around the circumference of the root. Fig. 2.133. A cryofixed showing misalignment of cells at the leading edge of the callus. Cells become oriented in the appropriate plane with increasing distance from the callus edge and/or following fusion of opposing vascular cambia. Fig. 2.134. A cedar root showing a barrier zone formed by the uninjured cambium comprised on axial parenchyma with pigmented phenolics. Lateral ingrowth of callus and woundwood extend over the area of killed cambium. Table 2.5. Frequency of successful resistance reactions following infection by A. ostoyae in the roots of Douglas-fir, western hemlock, and western redcedar observed from four separate field trials 2002-2004. Frequency of Successful Resistance Reactions as a proportion of successful penetrations by A. ostoyae Field Trial Year Harvest Time Douglas-fir Western hemlock Western redcedar Hidden Lake 2002 11 weeks 0.33 (n=9) 0.07 (n=14) 0.67 (n=6) 1 year 0.30 (n=10) 0.00 (n=5) 1.00 (n=4) Kingfisher 2003 5 weeks - (n=0) 0.00 (n=l) _ (n=0) 9 weeks - (n=0) 0.33 (n=3) 0.50 (n=2) 1 year 0.20 (n=10) 0.35 (n=23) 0.75 (n=8) Kingfisher 2004 9 weeks (n=0) 0.00 (n=l) 1.00 (n=l) 5 months 0.11 (n=9) 0.36 (n=l l ) 0.94 (n=16) 1 year 0.08 (n=12) 0.42 (n=12) 0.80 (n=10) Nakusp 2004 8 weeks - (h=0) 0.50 (n=2) 1.00 (n=l) 5 months 0.00 (n=2) 0.00 (n=6) 0.71 (n=7) 1 year 0.00 (n=2) 0.50 (n=2) 1.00 (n=2) n = the number of cases of successful penetration of roots by A. ostoyae. 119 the pathogen failed to penetrate the root (initial infection) were excluded from analysis of frequency of N I T and N P formation and breaching. Failure to penetrate was not treated as a form of resistance, but rather attributed, in most instances to failure of proper inoculum development or to soil and environmental factors which prevented normal rhizomorph development and adhesion to root surfaces. Thus, 5 week and 9 week harvest dates from the second, third and fourth inoculation trials were excluded from the analysis due to the small sample size and lack of infection in some species on those dates, predominantly in Douglas-fir. Table 2.6 shows which differences in the frequency of successful resistance reactions between species and at different times and places were statistically significant. C h i -square tests revealed there were significant differences in the frequency of successful resistance reactions between western redcedar and either or both o f Douglas-fir and western hemlock on the different harvest dates and field trials (Appendix IV) . Table 2.6. Individual species comparisons of the frequency of successful resistance reactions following inoculation with A, ostoyae in Douglas-fir (Fd), western hemlock (Hw) and western redcedar (Cw) for different harvest dates and inoculation trials. Significance 1 o f Species Comparisons for each Field Trial and Harvest Date Species Comparisons H L 11 wks (2002) 1 yr K F (2003 1 yr ) K F 5 mths (2004) 1 yr Nakusp (1 5 mths •004) 1 yr Fd /Hw ns (n=9/14) ns (n=10/5) ns (n=10/23) ns (n=9/ll) p<0.1 (n=12/12) nt (n=2/6) ns (n=2/2) Fd/Cw ns (n=9/6) p < 0.025 (n=10/4) p <0.025 (n=10/8) p < 0.001 (n=9/16) p < 0.001 (n=12/10) p<0.1 (11=2/7) p < 0.05 (n=2/2) H w / C w p < 0.01 (n=14/6) p < 0.01 (n=5/4) p < 0.05 (n=23/8) p < 0.01 (n=l 1/16) p<0.1 (n=12/10) p < 0.01 (n=6/7) ns (n=2/2) 1 The p-value denotes the significance level of which the null hypothesis (i.e. no difference between species in the frequency of successful resistance reactions) would be rejected, ns = not significant at a = 0.05; nt = not tested n = the number of cases of successful penetration of roots by A. ostoyae 120 Successful resistance reactions following infection by A. ostoyae ranged between 0-33% and 0-50% for Douglas-fir and western hemlock trees, respectively. L o w frequencies of resistant responses were consistent in Douglas-fir and hemlock roots on each sampling date in each of the four inoculation trials. Hemlock and Douglas-fir did not differ with respect to their ability to contain infections within a lesion in the bark or wood. Although western red cedar showed a significantly higher proportion of successful resistance reactions than Douglas-fir on all harvest dates in every field trial (Table 2.5), the difference was not significant in the 2002 Hidden Lake trial at 11 weeks ( j 2 , p < 1) or at 5 months in the 2004 Nakusp trial (x2, < 0.1). Similarly, cedar consistently showed a significantly higher percentage of successful resistance reactions compared to western hemlock, but the difference was less significant at 1 year in the 2004 Kingfisher trial (x2, p < 0.1) and at 1 year in the 2004 Nakusp trial (x2, p < 0.1). There was no significant difference in the frequency of successful resistance reactions between the different harvest dates for each species. Furthermore, no difference in the frequency of successful resistance reactions was detected between the different field trials/sites for each species at 1 year. The above results are interpreted to indicate that resistance did not depend on the length of time that the root remained in the ground in contact with the inoculum block nor did any site factors influence the frequency o f the resistance reactions formed in a given host species. For this reason, frequency data for all field trials showing the different stages/events involved host responses leading to N P formation in the bark and compartmentalization in the wood on different harvest dates and in different field trials were pooled for each species and analyzed as such. Table 2.7 shows the frequency of N I T and N P formed, N I T and N P breached, the number of roots that showed cambial invasion and compartmentalization, and the number of roots with successful resistance reactions as a percentage of successful penetrations by A ostoyae in Douglas-fir, western hemlock and western redcedar for all field trials combined. Frequencies of the different stages o f host response are not strictly sequential. For example, not all instances of N I T Table 2.7. The number of root inoculations that resulted in successful penetration by A. ostoyae, the frequency of roots showing no visible or ineffective host response, initiation and breaching 0 of non-suberized impervious tissue (NIT) and necrophylactic periderm (NP), kil led cambium and compartmentalization, and the number and percentage of roots showing successful resistance reactions in Douglas-fir, western hemlock, and western redcedar trees for all field trials combined. Successful No. o f resistance reactions Species N o . of No. roots CODIT successful as a percentage Successful N H R N I T N I T N P N P with killed & resistance of successful penetrations initiated breached formed breached cambium* callusing reactions penetrations Douglas-fir 54 29 25 . 1 0 20 11 20 2 10 1 8 . 5 2 % Western hemlock 80 29 51 18 43 22 37 3 21 2 6 . 2 5 % Western redcedar 57 1 5 6 c 3 53 11 24 22 47 8 2 . 4 6 % NHR, no visible or ineffective host response; NIT, non-suberized impervious tissue; NP, necrophylactic periderm; CODIT, compartmentalization of decay in trees " breaching refers to direct penetration through barriers as well as circumvention in places where the host did not complete the formation of the barrier. b killing of the vascular cambium was primarily a reflection of the time and the location of the colonizing fungus in the tissue at the time of sampling. The killing of cambium resulted from a combined number of some, but not all roots showing NHR and breaching of barriers in the bark. c although NIT was generally indistinguishable in cedar, it is the provisional view of this author that in all places where a NP formed, NIT had developed internally abutting that zone. to 122 breaching wi l l result in N P not being formed. N P formation occurs within a few days of N I T initiation and both N I T and N P w i l l be complete in some areas (e.g. central region of the wounded tissue) but incomplete in others (e.g. contiguous with the original periderm). Table 2.8 shows which differences in frequency o f host reactions shown in Table 2.7 were significant. The corresponding Chi-square tests can be found in Appendix V . Table 2.8. Individual species comparisons of the frequency of roots showing NIT initiated, NIT breached, NP formed, NP breached and compartmentalization following inoculation with A. ostoyae in Douglas-fir (Fd), western hemlock (Hw), and western redcedar (Cw) trees for all harvest dates in four inoculation trials. Significance 0 level of differences between pairs of species for different stages/events o f host reactions in the bark and wood Type of host response Fd /Hw Fd /Cw H w / C w N I T (p < 0.05) (p< 0.001) (p< 0.001) initiated (n"=54/80) (n=54/57) (n=80/57) N I T ns (p< 0.001) (p< 0.001) breached (n=25/51) (n=25/56) (n=51/56) N P ns (p< 0.001) (p< 0.001) formed (n=54/80) (n=54/57) (n=80/57) N P ns (p<0.01) (p<0.01) breached (n=20/43) (n=20/53) (n=43/53) C O D I T & ns (p< 0.001) (p< 0.001) callusing (n=20/37) (n=20/24) (n=37/24) The p-value denotes the significance level of which the null hypothesis (i.e. no difference between species in the frequency of a particular host reaction) would be rejected, ns = not significant at a = 0.05 ' the sample size (n) changes for different stages of host reactions whereby for NIT initiated and NP formed, n = the number of cases of successful penetration of roots by A. ostoyae; for NIT and NP breached, n = the number of cases that resulted in NIT initiation and NP formation, respectively; for CODIT & callusing, n = the number of cases in which the vascular cambium was killed as a combined result of some, but not all roots showing no visible or ineffective host response and breaching of barriers in the bark 123 Chi-square tests revealed significant differences among species in the frequency of N I T initiation (x2, p< 0.001). Western redcedar differed significantly from both Douglas-fir and western hemlock. Although N I T formation in Douglas-fir was slightly lower than western hemlock, the difference was still significant. A higher percentage of western hemlock and Douglas-fir roots showed no visible host response following invasion by A. ostoyae than cedar. The above results are interpreted to indicate that the capacity for trees to trigger the process of phellogen renewal is less frequent in Douglas-fir and western hemlock than in western redcedar. O f the total number of western redcedar roots sampled at various intervals over the course of 3 years, only 1 o f 57 (2%) roots did not develop N I T compared to 36% (n=80) in western hemlock and 54% (n=54) in Douglas-fir. The breaching of N I T depends on its speed of formation relative to the speed at which the fungus or the toxins produced by the fungus move through host tissue. N I T was breached in all species although breaching was consistently more frequent in Douglas-fir and western hemlock than western redcedar. N P formation varied at each sampling date and was primarily dependent on whether the host developed NIT . In most cases, both N I T and N P were observed in roots examined on each harvesting date. A N P can differentiate within several days following the establishment of NIT, even when a fully differentiated zone of N I T is not continuous with the original periderm. Hence, one may find both N I T and N P differentiated in the mid-phloem underlying necrotic tissue while NIT , and therefore N P ' s are absent or still in the developing stages near the junction o f the original periderm. Although N I T was difficult to discern histologically, it 's generally assumed that all species that formed N I T also formed a N P . Western redcedar always formed N P ' s with higher frequency than either Douglas-fir and western hemlock (Table 2.8). Similar to N I T formation, the frequency of N P formation in cedar was significantly greater than in Douglas-fir and western hemlock. Again, there was no difference between Douglas-fir and western hemlock. 124 The fact that N I T and N P were initiated in nearly all cedar roots penetrated by the fungus and those barriers showed significantly less breaching by the fungus suggests that Douglas-fir and western hemlock may be slower at completing host defense structures in response to infection by A. ostoyae than cedar. A slower rate at which host defenses are induced in Douglas-fir and western hemlock are also suggested by the wider zone of N I T in those species and incomplete differentiation of tissue leading to N P formation. Breaching of NPs was common among all species, particularly in Douglas-fir and western hemlock. Chi-square tests revealed significant differences between cedar and hemlock and between cedar and Douglas-fir. The frequencies at which NPs were breached in either Douglas-fir or western hemlock were similar. There were no differences among species in the frequency of kil led cambium following successful penetration by A. ostoyae. Overall, A. ostoyae advanced to and ki l led the cambium in 37% (n=54), 46% (n=80), and 42% (n=57) of the Douglas-fir, western hemlock and western redcedar roots, respectively. In most situations, cambial invasion was more frequent on roots harvested at 5 months and 1 year compared to the first 11 weeks following inoculation. However, strong differences were evident among species in the frequency of compartmentalization and callusing (x2, p< 0.001). Compartmentalization took place in only 2 o f 20 (10%) Douglas-fir roots and in only 3 o f 37 (8%) western hemlock roots. These low percentages are in large contrast to the 92% (n=24) of western redcedar roots that showed compartmentalization f