Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Assessment of neonectria neomacrospora as a potential biological control agent for hemlock dwarf mistletoe Rietman, Lea Mae 2004

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

Item Metadata

Download

Media
831-ubc_2004-0292.pdf [ 8.67MB ]
Metadata
JSON: 831-1.0074989.json
JSON-LD: 831-1.0074989-ld.json
RDF/XML (Pretty): 831-1.0074989-rdf.xml
RDF/JSON: 831-1.0074989-rdf.json
Turtle: 831-1.0074989-turtle.txt
N-Triples: 831-1.0074989-rdf-ntriples.txt
Original Record: 831-1.0074989-source.json
Full Text
831-1.0074989-fulltext.txt
Citation
831-1.0074989.ris

Full Text

Assessment of Neonectria neomacrospora as a Potential Biological Control Agent for Hemlock Dwar f Mistletoe By Lea Mae Rietman B.Sc., The University of British Columbia, 2001 A THESIS 'SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF M A S T E R OF SCIENCE In THE F A C U L T Y OF G R A D U A T E STUDIES Faculty of Forestry, Department of Forest Sciences We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 2004 © Lea Mae Rietman, 2004 L i b r a r y A u t h o r i z a t i o n In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Lea Rietman 08/04/2004 Name of Author (please print) Date (dd/mm/yyyy) Title of Thesis: Assessment of Neonectria neomacrospora as a potential biological control agent for hemlock dwarf mistletoe Degree: Master of Science Department of Forest Sciences The University of British Columbia Vancouver, BC Canada Year: 2004 Abstract The purpose of this study was to assess the potential of Neonectria neomacrospora as an inundative biological control agent for hemlock dwarf mistletoe (Arceuthobium tsugense). Based on the growth characteristics of six isolates of N. neomacrospora, isolate PFC 2559 was selected as the inoculum source. A field trial was conducted on Vancouver Island to determine the requirement of wounding for N. neomacrospora infection and to measure the impact of TV. neomacrospora infection on A. tsugense vigour. Conidia were mass produced and suspended in "Stabileze" formulation, which was sprayed onto entire hemlock dwarf mistletoe infections (swellings and shoots). Treatments involved unwounded A. tsugense swellings, cut A. tsugense aerial shoots, and wounded A. tsugense swellings. The number of healthy and diseased^, tsugense aerial shoots was recorded at 2 weeks and 1, 2, 3.5, 5, 6, and 9 months after treatment. At 10 months, a sub-portion of the treated A. tsugense swellings was destructively sampled and an attempt to re-isolate N. neomacrospora was done. The greatest amount of re-isolation of N. neomacrospora occurred for the wounded, inoculated treatment (73%), followed by the cut shoots, inoculated treatment (60%), and the unwounded, inoculated treatment (55%); (35%, 20%, and 21% for controls, respectively). Based on the extent of bark necrosis, the occurrence of sporodochia, and re-isolation of TV", neomacrospora, the wounded, inoculated treatment had the greatest impact on N. neomacrospora infection when compared to the unwounded, inoculated treatment and the cut shoots, inoculated treatment. There was no evidence that N. neomacrospora could infect wounded A. tsugense shoots. Once infection by TV. neomacrospora was established, a 38% decline in the number of healthy A. tsugense aerial shoots was observed at 9 months. Applying inoculum to unwounded A. tsugense swellings is predicted to reduce the number of A. tsugense shoots by 14% in the field at 9 months. This is a conservative estimate. Due to the extensive bark necrosis and girdling of the A. tsugense swelling, infection by N. neomacrospora is expected to have a greater long-term impact on A. tsugense vigour than that observed in the present field trial. Table of Contents Abstract ii Table of contents iii List of tables vii List of figures ix Acknowledgments • xii Chapter 1 Literature Review 1.10 Overview of the role of forest pathogens 1 1.2.0 Dwarf mistletoes : 1 1.2.1 Background on dwarf mistletoes.. 1 1.2.2 Symptoms and signs of dwarf mistletoes ..2 1.2.3 Life cycle of dwarf mistletoes 3 1.2.4 Biology of dwarf mistletoes. 4 1.2.5 Distribution of dwarf mistletoes 7 1.2.6 Hosts of dwarf mistletoes 7 1.2.7 Impact of dwarf mistletoe in forestry 8 1.2.8 Spread and intensification of dwarf mistletoes 9 1.3.0 Management strategies for dwarf mistletoes 10 1.3.1 Traditional control of dwarf mistletoes 12 1.3.2 Chemical control of dwarf mistletoes 14 1.3.3 Genetic control of dwarf mistletoes 15 1.3.4 Biological control of dwarf mistletoes 17 1.3.4.1 Hyperparasites of dwarf mistletoe shoots 19 1.3.4.2 Hyperparasites of dwarf mistletoe swellings 22 1.4.0 Introduction to Arceuthobium tsugense 23 1.4.1 Symptoms and signs of A. tsugense.. 23 1.4.2 Biology of A tsugense 23 1.4.3 Distribution of A tsugense 25 1.4.4 Taxonomy and hosts of A. tsugense 26 1.4.5 Spread and intensification of A tsugense 27 1.4.6 Impact of A tsugense in forestry 28 1.4.7 Control of A tsugense 30 1.5.0 Introduction to N. neomacrospora 30 1.5.1 Taxonomy and hosts of N. neomacrospora 31 1.5.2 Symptoms and signs of N. neomacrospora 33 1.5.3 Biology of TV. neomacrospora ..36 1.5.4 A potential biological control agent of Arcuethobium tsugense. 37 1.6.0 Research objectives 38 Chapter 2 Selection of an isolate of Neonectria neomacrospora 1 2.1.0 Introduction 39 2.2.0 Materials and methods 39 2.2.1 Collection of Neonectria neomacrospora isolates 40 2.2.2 Selection and screening of the N. neomacrospora isolates ..41 2.2.3 Linear colony growth at varying temperatures 41 2.2.4 Percentage and rate of spore germination at varying temperatures. 42 2.2.5: Mass production of inoculum on different solid -based media .44 2.2.6 Inoculation of A. tsugense in a small field trial 44 2.3.0 Results 48 2.3.1 Selection and screening of the TV", neomacrospora isolates 48 2.3.1.1 Linear colony growth at varying temperatures 48 r 2.3.1.2 Percentage and rate of spore germination at varying temperatures 50 2.3.1.3 Mass production of inoculum on different solid-based media 52 2.3.2 Inoculation of A. tsugense in the small field trial 54 2.4.0 Discussion 56 2.4.1 Growth characteristics of N. neomacrospora 56 2.4.2 Small field trial 57 Chapter 3 Inoculation of Arceuthobium tsugense in a large field trial 3.1.0 Introduction 60 3.2.0 Materials and methods 61 3.2.1 Stabileze formulation .....61 3.2.2 Site description 62 3.2.3 Experimental design 63 3.2.4 Pre-treatment assessment 65 3.2.5 Treatment application dates 65 3.2.6 Treatment assessment 65 3.2.7 Sampling at 10 months 65 3.2.8 Statistical analysis 67 3.2.8.1 Statistical analysis of the impact of wounding on infection 67 3.2.8.2 Statistical analysis of infection on A. tsugense vigour 68 v- 3.3.0 Results .....69 3.3.1 Stabileze formulation 69 3.3.2 Symptoms and signs of N. neomacrospora in the field. 70 3.3.3 Lost experimental units 10 months following inoculation 72 3.3.4 Importance of wounding for establishment of N. neomacrospora 73 3.3.4.1 Extent of bark necrosis 74 3.3.4.2 Proportion of A. tsugense with sporodochia 75 j 3.3.4.3 Re-isolation of N. neomacrospora from A. tsugense swellings 78 3.3.5 Impact of TV. neomacrospora on A. tsugense vigour 81 3.3.5.1 Number of healthy A. tsugense shoots for unconfirmed and confirmed groups 79 3.3.5.2 Proportion of A tsugense infections with zero shoots for unconfirmed and confirmed groups 83 3.3.5.3 Extent of bark necrosis and girdling of A tsugense swellings for unconfirmed and confirmed groups 83 3.4.0 Discussion 85 3.4.1 Stabileze formulation 85 3.4.2 Importance of wounding for establishment of N. neomacrospora 86 3.4.3 Effect of N. neomacrospora infection on A. tsugense vigour 89 3.4.4 Measuring.the impact of treatment A on A. tsugense shoot vigour 90 3.5.0 Closing Remarks 92 3.6.0 Future Research 93 Literature Cited 94 Appendix I-HOBO data for small field trial- temperature 105 Appendix II-HOBO data for small field trial- relative humidity 106 Appendix III-HOBO data for large field trial-temperature 107 Appendix IV-HOBO data for large field trial-relative humidity 108 vii List of Tables Table 1.1: Morphology and host characteristics of Nectria fuckeliana var fuckeliana and Neonectria neomacrospora (modified from Funk 1981; Booth 1979a, 1979b) 32 Table 2.1. The origin of the six isolates of N. neomacrospora selected for screening 41 Table 2.2. Treatment description and number of replicates for each treatment used in the small field trial 46 Table 2.3. Sporodochia and successful re-isolation of N. neomacrospora (N. n.) for each treatment from the small field trial. The proportion of A. tsugense swellings colonized by N. neomacrospora was based on the occurrence of sporodochia only 55 Table 3.1. Treatment description and number of replicates used in the field trial designed to assess N. neomacrospora as a biological control agent for A tsugense 64 Table 3.2. The number of experimental units collected from the trial 10. months after treatment. Note that the total number of experimental units ("Total") takes into account A. tsugense infections that were lost during the trial (Table 3.4) 66 Table 3.3. Break down of the proportion of A. tsugense swellings with confirmed and unconfirmed N. neomacrospora infection for each treatment 69 Table 3.4. The number of A. tsugense infections prior to treatment and the number of A. tsugense infections remaining at the termination of the trial 72 Table 3.5. Number of destructively sampled A. tsugense swellings at 10 months with and without sporodochia and successful re-isolation of N. neomacrospora 79 List of Figures Figure 1.1 Typical signs and symptoms of Neonectria neomacrospora. A . Perithecia from the sexual stage of N. neomacrospora, B . Sporodochia from the asexual stage of N. neomacrospora, C. Macroconidia from the asexual stage of N. neomacrospora, D. Mycelium of the asexual stage of N. neomacrospora grown on potato dextrose agar for 21 days, E. Resinosis and cankering of the A. tsugense swelling, F. Destructively sampled A. tsugense swelling with top layer of bark removed to show bark necrosis, G. Cross section of an A. tsugense swelling showing bark necrosis, H . Diseased and dead A. tsugenseshoots. Pictures A and E courtesy Dr. Simon Shamoun, Pacific Forestry Centre, Victoria, B C (2003) 33-34 Figure 2.1. A sample graph of the percentage of germinated spores plotted against time for PFC 2546 at 20°C. In this case, r 2 = 0.95 43 Figure 2.2. Linear colony growth (mm/year) of six isolates of N. neomacrospora at temperatures ranging from 4 to 30°C. Bars measure Standardn error of the mean 48 Figure 2.3. Colony radius (mm) across time for six isolates of N. Neomacrospora at 20°C. Bars measure standard error of the means 49 Figure 2.4. Percent conidial germination for six isolates of N. neomacrospora 24 hours after inoculation at temperatures ranging from 4°C to 35°C 50 Figure 2.5. Number of hours for 50% spore germination for six isolates of N. neomacrospora at temperatures ranging from 4°C to 30°C .'• 51 Figure 2.6. Mean number of spores produced by six isolates of N. neomacrospora averaged across four different media types (brown rice, millet, wheat bran and slow oats) at room temperature (22°C). Bars measure standard error of the mean 52 Figure 2.7. Mean number of spores produced on four media types averaged across six isolates of N. neomacrospora at room temperature (22°C). Bars measure standard error of the mean 53 Figure 2.8. Mean number of healthy A. tsugense shoots observed at 1 and 12 months after treatment. Bars measure standard error of the mean 54 Figure 2.9. Symptoms of A. tsugense infected with a mycelial plug of TV. neomacrospora into a wound at 1 month (A) and 9 months (B). The -arrow in (A) points to the site of inoculation. Note the diseased, dead A. tsugense aerial shoots and girdling of the swelling at 9 months (B) 58 Figure 3.1. A treated A. tsugense swelling at 2 weeks (A) and 9 months (B) after treatment. Treatment involved wounding the A. tsugense swelling and applying inoculum (treatment C). Infection by N. neomacrospora was confirmed 71 Figure 3.2. A treated A. tsugense swelling at 2 weeks (A) and 9 months (B) after treatment. Treatment involved applying inoculum to unwounded A. tsugense swellings (treatment A). Infection by N. neomacrospora was not confirmed 71 Figure 3.3. Mean extent of bark necrosis for A. tsugense swellings collected 10 months after treatment. Bars measure standard error of the mean 74 Figure 3.4. Proportion of A. tsugense swellings collected 10 months after treatment bearing confirmed sporodochia 75 Figure 3.5. Cumulative proportion of A. tsugense swellings bearing sporodochia 3.5,6 and 9 months after treatment 76 Figure 3.6. The non-cumulative proportion of A. tsugense swellings bearing sporodochia. A l l treatments are grouped together for this data 77 Figure 3.7. Proportion of A. tsugense swellings from which N. neomacrospora was successfully re-isolated 10 months after treatment 78 Figure 3.8. Linear relationship of the extent of bark necrosis and the proportion of A. tsugense swellings from which N. neomacrospora was successfully re-isolated. Regression analysis of Y=Yo + aX found r 2 - 0.93 80 Figure 3.9. Mean number of healthy A. tsugense shoots 9 months after treatment for unconfirmed and confirmed groups. Cut shoot treatments were removed from analysis. Bars measure standard error of the mean 81 Figure 3.10. Mean number of healthy A tsugense shoots observed for unconfirmed and confirmed N. neomacrospora infected groups. Bars measure standard error of the mean 82 Figure 3.11. Mean proportion of A. tsugense infections with no shoots 10 months after treatment for unconfirmed and confirmed groups 83 Figure 2.20. Proportion of girdled A. tsugense swellings 10 months after treatment for unconfirmed and confirmed groups ...84 xi Acknowledgements Funding for this project was provided by Forest Research British Columbia (FRBC) and Forestry Innovation Investment (FII). Many thanks are owed to my supervisors, Dr. Bart van der Kamp and Dr. Simon Shamoun, and my committee member, Dr. Bob Copeman for their commitment to this project. I would also like to thank Sue Askew, Dr. Hugh Barclay, Cherly Konoff, John Leaky, Dr. Les Safranyik, Holly Williams, and my coworkers for their contributions to this work. M y family and friends, thank you for supporting me through this experience. xn Chapter 1: Literature Review 1.1.0 Overview of the role of forest pathogens "Disease" can be defined as the deleterious effects resulting from injurious biotic agents other than insect damage or animal wounding (Callan 2001). A biotic agent that incites disease in a host is termed a "pathogen" (Castello, Leopold, and Smallidgel995). Both pathogens and abiotic disturbances recycle essential elements and influence forest development and landscape, often selecting for less vigourous or genetically susceptible individuals. Pathogens are unique in that they require the presence of a susceptible host and favourable environmental conditions (Callan 2001). Pathogens are major agents of forest diversity (van der Kamp 1991). The primary route in which pathogens impact forest succession is through tree mortality. Pathogens may impact (and be impacted by) plant species distributions, forest structure and composition, succession, biodiversity, and landscape patterns (Castello, Leopold, and Smallidgel995). Because forest pathogens can alter forest species composition and reduce host vigour, most forest pathogens are considered undesirable in the forest industry. Dwarf mistletoes, a genus of plant parasites that attack conifers, are a good example of problematic forest pathogens. Dwarf mistletoes cause extensive timber losses through direct and indirect mortality and growth reduction. They are especially damaging in parts of the Northern Hemisphere, particularly in western North America. In the USD A Forest Service survey from 1979-1983, the area of US commercial forest affected by dwarf mistletoe reached 22.2 millions of acres, surpassing root pathogens, fusiform rust, southern pine beetle, western spruce budworm and gypsy moth (USDA Forest Service 1988). The impact of dwarf mistletoes in forestry is significant and currently acceptable control methods are limited. 1.2.0 Dwarf mistletoe 1.2.1 Background on dwarf mistletoe Dwarf mistletoes (Arceuthobium) belong to the family Viscaceae and are obligate, perennial plant parasites of Pinaceae and Cupressaceae in the Old and New Worlds. 1 Dwarf mistletoes rely on their host for support, mineral nutrients and water, some carbon compounds, and possibly other growth factors. The group consists of 46 recognized taxa, which comprise of 42 species, 4 with 2 subspecies each, and 1 with 2 formae speciales (Hawksworth and Wiens 1996). Note: The term "infection" can be defined as the process of a pathogen establishing itself on a host and obtaining food (Agrios 1997) or as the pathological state resulting from having been infected (R. J. Copeman, Univeristy of British Columbia, Vancouver, B C , pers. comm.). In this document, both definitions were used. For example, infection may refer to the process of dwarf mistletoe parasitizing its host, or it may be used to refer to the entire dwarf mistletoe plant and parasitized host region. 1.2.2 Symptoms and signs of dwarf mistletoes In most cases, the first symptom associated with infection by dwarf mistletoe is swelling of the host tissue. From the swelling arise dwarf mistletoe buds that develop into shoots. Dwarf mistletoe shoots are perennial and the fruit they bear is the sole means of reproduction. After a shoot is broken or falls off, a unique structure called a basal cup remains attached to the host (Hawksworth and Wiens 1996). Older dwarf mistletoe infections often cause a phenomenon known as witches' broom, which consists of dense, profusely branched masses of host branches at the site of infection (Baranyay, Hawksworth, and Smith 1971). Witches' brooms can either be systemic (isophasic) or non-systemic (anisophasic). Typically, it is the parasite species rather than the host that determines the type of witches' broom formed. A few species of dwarf mistletoe, such as Arceuthobium aureum subsp. aureum and A. globosum subsp. globosum, do not induce witches' broom (Hawksworth and Wiens 1996). Host vigour and wood quality may also be reduced by dwarf mistletoe infection. Heavily dwarf mistletoe infected trees have been shown to have shorter needles and needle bearing stem length, reduced leaf surface and number of needles, and chlorosis of the needles (Korstian and Long 1922). Wood quality of infected dwarf mistletoe trees is reduced by the production of large knots, development of abnormal grain, and reduction in strength (Anonymous 1985). 2 1.2.3 Life cycle of dwarf mistletoes Dwarf mistletoes are dioecious seed plants. Depending on the species, the life cycle of mistletoe usually takes 5 or more years to complete (Hawksworth and Wiens 1996). The fruit contains a single seed and embryo. The seed is forcibly ejected and dispersed in late summer and early fall using a hydrostatically controlled explosive mechanism (Hinds, Hawksworth, and McGinnies 1963). Seeds can be discharged at velocities averaging 24 m per second (Hinds and Hawksworth 1965). Animal vectors may also play a role in dwarf mistletoe seed dispersal by disseminating seeds stuck to their feathers or fur. Some birds, such as ruffled grouse, blue grouse, and black-capped chickadee, have been reported to eat dwarf mistletoe seeds (Hawksworth 1975). Squirrels have also been known to feed on dwarf mistletoe infected bark (Baranyay, 1968, Hawksworth 1975). Although dwarf mistletoe seeds cannot survive through the digestive tract, feeding provides an opportunity for seeds to stick to the feathers or fur of the vector, which are removed later by grooming (Hudler, Oshima, and Hawksworth 1979; as cited in Hawksworth and Wiens 1996). Dwarf mistletoe seeds adhere to their landing surface by means of a sticky, viscous coating (viscin) that surrounds the seed (Roth 1959). Roth (1959) observed that seeds were more prone to stick to needles rather than branch parts that were bare of needles, as the seeds tended to bounce off the more solid branch parts. Hawksworth (1965) noted that of those Arceuthobium americanum and A. vaginatum seeds that landed on a host, 90% and 94% of the seeds landed on the host needles, respectively. Once wet with rain, the hygroscopic viscin becomes well lubricated and allows the seed to slide to the base of an upright needle. As the viscin dries, it becomes solid and cements the seed to the shoot surface. Only a small fraction of the seeds that land on susceptible hosts are transferred to the twigs; most seeds fall off the host. In the cases of A. americanum and A. vaginatum, an average of 20% of the seeds that landed on a host needle was transferred to twigs (Hawksworth 1965). Dwarf mistletoe seeds must remain attached to the host over winter. In the spring, the seed germinates and develops a radicular apex. Once this structure comes in contact with an obstruction, such as a needle base, a holdfast is formed. The centre of the holdfast develops a penetration wedge that forces into the host cortex (Scharpf and Parmeter 1967). The penetration structure appears to crush or push host cells out of the way as it enters the host (Hunt, Owens and Smith! 996). Once the host periderm has been penetrated, the host branch is invaded. A n incubation period of 2-5 years elapses before the first shoots appear. One to two years after shoots develop, the dwarf mistletoe infection may begin to flower. Flowering typically lasts 4-6 weeks and, in most cases, occurs during annual flowering periods. There are a few examples of tropical species of dwarf mistletoe, such as Arceuthobium aureum sp. aureum, which appear to flower continuously throughout the year. After pollination, fruits mature on the female infection the following year and the cycle is repeated (Hawksworth and Wiens 1996). Dwarf mistletoes live on the living tissue of the host tree and can survive as long as the host branch is alive (Anonymous 1985). As the host grows, dwarf mistletoe infections on the lower branches may be shaded out, loose all their shoots and become dormant. Gi l l and Hawksworth (1961) have suggested that dwarf mistletoe may live indefinitely in the dormant state. These dormant dwarf mistletoe infections can be reactivated i f conditions improve. Increasing light levels through selective logging, for example, may reactivate dormant dwarf mistletoe infections (van der Kamp 1986). 1.2.4 Biology of dwarf mistletoe The dwarf mistletoe infection can be separated into two parts: the shoots and the root-like system that forms in the host tissue. The shoots are the stem-like portions of the dwarf mistletoe infection outside the host that function primarily in reproduction (Baranyay, Hawksworth, and Smith 1971). Although dwarf mistletoe shoots are capable of some photosynthesis, the parasite derives a large proportion of its photosynthate and all of its water and minerals from the host (Leonard and Hull 1965). Dwarf mistletoes are therefore considered hemiparasitic, or "capable of some autotrophic carbon fixation, but dependent on the host for an unknown quantity of reduced carbon compounds." (Miller and Tocher 1975). The root-like portion of the dwarf mistletoe infection that forms in the host tissue has been termed the endophytic system (Thoday and Johnson 1930). Depending on the dwarf mistletoe species, the host, and possibly the site of original infection, the endophytic system can be localized or systemic (Kuijt 1960; as cited in Hawksworth and Wiens 1996). It has also been suggested that the site of inoculation, whether in primary or secondary tissue, has a significant impact on the form of endophytic system that develops for those dwarf mistletoe species capable of systemic infection (Alosi and Calvin 1984). In localized infections, the dwarf mistletoe endophytic system can only keep up with cambial growth and is contained within a limited section of the host branch. The endophytic system often causes swelling of the infected host branch, and the shoots remain localized near the original site of infection. In systemic infections, the endophytic system keeps pace with apical and cambial growth of the infected branch. The dwarf mistletoe endophytic system ramifies extensively through the host tissue of broom branches arising from infected bark, and swelling of the host branch is not so pronounced. The shoots of systemic infections are scattered along the host branch or concentrated at branch nodes. Few dwarf mistletoes are capable of systemic infection, and it has been argued that systemic infection is more evolutionary advanced (Hawksworth and Wiens 1972). Most dwarf mistletoes are only able to infect secondary tissue and have limited ability to move distally along the host branch (Alosi and Calvin 1984). \ The endophytic system of dwarf mistletoes differs from typical plant roots in that it lacks a root cap, pericycle, endodermis, and continuous phloem and xylem (Cohen 1954; Alosi and Calvin 1985; Calvin, Hawksworth, and Knutsonl984). Beginning with a primary haustorium at the infection site, a localized infection develops haustorial (cortical) strands, which extend longitudinally and sometimes circumferentially through the host cortex and outer host phloem (Alosi and Calvin 1984). Other structures termed "sinkers" have radially oriented growth and extend from the haustorial strands into the host secondary vasculature. Although given a different name, sinkers are not considered to be distinct organs of the haustorial system; rather, the name is used in reference to radially oriented portions of the endophytic system or tissue types (Alosi and Calvin, 1984). Sinker tissue is almost always associated with host rays (Thoday and Johnson 1930; Cohen 1954; Srivastava and Esau 1961). If young sinkers develop in a region 5 where ray cells do not occur, new ray cells quickly develop in relationship with the sinker cells. The resulting structure, composed of radially-orientated sinker cells surrounded by host ray cells, has been termed an "infection ray" (Srivastava and Esau 1961). The infection ray may consist of host ray parenchyma, host albuminous cells, sinker ground parenchyma, sinker xylem, and sinker sheath cells. The structure of the systemic endophytic system differs significantly from the localized endophytic system in the primary growth region of the host. In the case of Arceuthobium douglasii, no sinkers or multiseriate haustorial strands develop in the primary endophytic form. The parasite cells do not specialize in this region, and only the inner cortex, phloem, and procambium are invaded. Once the host branch matures and starts to develop secondary tissue, the endophytic system of A. douglasii changes. The primarily uniseriate strands of the mistletoe develop into more complex, multiseriate, longitudinal strands that cease elongation and expand radially. Apical strands of the parasite grow towards the cambium, and sinker cells develop, resulting in an endophytic structure that is more similar to that of localized infections (Alosi and Calvin 1984). Possibly as a result of continual host-parasite association, systemic infections tend to have less anomalous wood when compared to localized swellings. For example, infection of Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) by A douglasii did not alter the size of the growth rings of the host, or induce traumatic resin ducts, which are common in localized dwarf mistletoe swellings (Alosi and Calvin 1984). However, abnormal longitudinal shrinkage and twisting of the wood on drying may still occur as a result of increased micorfibril angle in tracheids in infected wood (B. J. van der Kamp, Univeristy of British Columbia, Vancouver, B C , pers. comm.). Due to their location within the host, sinkers are believed to be associated with parasite nutrient uptake (Alosi and Calvin 1984). The transfer of water and minerals, host-originating photosynthate, and possibly organic compounds from the host to the dwarf mistletoe occurs in the common apoplast (Alosi and Calvin 1985). Tainter (1971) suggested that conversion of carbohydrates into lipid would create a sufficient concentration gradient to maintain flow of carbohydrates from the host to the dwarf mistletoe infection. Once absorbed, these are then carried through the symplast of the parasite to the shoots. 6 The exact transport method of nutrients from the dwarf mistletoe endophytic system to the shoots is not understood. Depending on the parasite species, phloem and xylem are reduced or absent in the dwarf mistletoe endophytic system and shoots (Calvin, Hawksworth, and Knutson 1984; Alosi and Calvin 1984). A possible route for nutrient mobilization from the endophytic system to the shoots may involve coupling diffusion potentials to water pressure (Alosi and Calvin 1985). ; 1.2.5 Distribution of dwarf mistletoe Of the 42 species of dwarf mistletoe, 34 occur in the New World. Only one species (Arceuthobium juniper-procerae) occurs in the southern hemisphere. In the new World, dwarf mistletoes are found from southeastern Alaska, northern Canada, and Newfoundland through most of the western United States and Mexico to Central America (Honduras), as well as in an outlying population on the island of Hispaniola. Greatest species diversity occurs in northwestern Mexico and the western United States. Of all the species in the New World, Arceuthobium douglasii has the widest distribution, ranging from southern British Columbia (BC), Canada, southward throughout most of the western United States to southern Durango, Mexico (Hawksworth and Wiens 1996). Nineteen species of dwarf mistletoe occur in the United States, and 5 of these occur in Canada. These include Arceuthobium americanum, A. douglasii, A. laricis, A: pusillum, and A. tsugense. A l l but A. pusillum are found in B C (Hawksworth and Weins 1996). 1.2.6 Hosts of dwarf mistletoes The hosts of dwarf mistletoe are commonly ranked according to susceptibility to infection using the categories "principal", "secondary", "occasional", and "rare" (Hawksworth and Wiens 1972). Conifers that are not parasitized by a particular dwarf mistletoe species are categorized as "immune". Hawksworth and Weins (1972) designed these classifications based on an infection index, which is defined by percentage classes of infected trees for each host species within 6 m of a heavily parasitized host. Principle hosts have infection levels that are greater than 90%, secondary hosts between 50-90 %, 7 occasional hosts between 5-50%, and rare host between 0^5%. hrimune species are expected to have 0 % infection levels. The principal hosts for dwarf mistletoes worldwide are typically restricted to a single genus or for Pinus to.a single subgenus. Only Arceuthobium microcarpum, A. laricis, and A. tsugense have principle hosts belonging to different genera. These are Picea and Pinus; Larix and Tsuga; and Abies, Pinus and Tsuga, respectively (Hawksworth and Wiens 1996). In the New World, the principal hosts of dwarf mistletoe include Abies, Larix, Picea, Pinus, Pseudotsuga, and Tsuga. In the Old World, principle hosts are Abies, Keteleeria, Juniperus, Picea, and Pinus. Host specificity of dwarf mistletoe species is broad, ranging from a single primary host, such as in the case of Arceuthobium apachecum on southwestern white pine (Pinus strobiformis Engelm.), to several primary hosts. Occasionally, artificial inoculation may result in infection of tree species that are not considered to be natural hosts. In Canada, the trees most seriously affected by dwarf mistletoe are lodgepole pine (Pinus contorta Dougl. ex Loud.), Jack pine (P. banksiana Lamb.), western hemlock (Tsuga heterophylla (Raf.) Sarg.), western larch (Larix occidentalis Nutt.,), and black spruce (Picea mariana (Mill.) BSP); (Hawksworth and Wiens 1996). In B C , dwarf mistletoes attack hemlock on the coast, lodgepole pine in the interior and Douglas fir and larch in the southeastern interior (Anonymous 1985). In the western United States, the most seriously damaged trees are ponderosa pine (Pinusponderosa P. Laws Ex C. Laws), lodgepole pine, Jeffrey pine (P. jeffreyi Grev. & Balfj, Douglas fir, California red fir (Abies magnifica Dougl. ex D. Don), white fir (A. concolor (Gord. & Glehd.) Lindl. ex Hildebr.), western larch, and western hemlock (Hawksworth and Wiens 1996). Cedar, juniper, cypress, redwood, and giant sequoia appear to.be immune to dwarf mistletoe infection in North America (Johnson and Hawksworth 1985). 1.2.7 Impact of dwarf mistletoe in forestry Dwarf mistletoes are the single most destructive pathogen of commercially valuable coniferous timbers in several regions of Mexico, western Canada, western United States, and parts of Asia (Hawksworth and Wiens 1996). It has been estimated 8 that dwarf mistletoes cause economic losses amounting to 11.3 million cubic meters of wood annually in the western United States, and 3.8 million cubic meters in western Canada (Drummond 1982; Sterner and Davidson 1982; as cited in Hawksworth and Wiens 1996). In cases of high infection, both the height and diameter growth rate of infected trees can be reduced. In one study, lodgepole pine heavily infested with dwarf mistletoe were on average 6-8 feet shorter in total height when compared to non-infected trees (Dobie and Britneff 1975). In another study, Hawksworth and Hinds (1964) observed a 5% cumulative reduction in diameter growth per decade for lodgepole pine (as cited in Hawksworth and Johnson 1989). Although the mechanism by which dwarf mistletoe reduces tree growth is not yet fully understood, it is believed that dwarf mistletoe infections alter the host metabolic balance, making the site of infection a nutrient sink (Leonard and Hull 1965). In this way, dwarf mistletoe infections reduce the amount of host nutrients available for host growth and development, thus reducing host vigour. Infection of a conifer by dwarf mistletoe may also reduce cone and seed production, and predispose the host to infection by other decay fungi, insects, and environmental factors (Johnson and Hawksworth 1985; Hawksworth and Johnson 1989). Under severe conditions of infection, dwarf mistletoe can kill the host. The rate at which an infected tree dies depends on its age when it was first infected and the number of infections. Older trees are slower to show measurable effects of dwarf mistletoe infection (Johnson and Hawksworth 1985). 1.2.8 Spread and intensification of dwarf mistletoe The distribution and intensity of dwarf mistletoe infections in young, regenerating stands depends on the distribution of residual trees, the disease within the residual tree, and the distance between the residual tree and the susceptible host (Smith 1977). In some cases, removal of the original residual tree after infection has established in a regenerating stand does not change the pattern of infection, indicating that internal re-infection of the regenerating stand greatly influences the pattern of dwarf mistletoe spread (Smith 1977). 9 The growth rate of dwarf mistletoe is hypothesized to be faster on vigorous hosts when compared to suppressed hosts. Arceuthobium americanum on lodgepole pine and A. vaginatum f. sp. cryptopodum on ponderosa pine both formed new shoots nearly twice as rapidly on vigorous hosts when compared to suppressed hosts (Hawksworth 1960). In dense stands, dwarf mistletoe infections in the middle and upper crown are most damaging. The occurrence of mistletoe in the lower crown region is of less concern because the relative spread is small and the infections may be shaded out (van der Kamp 1986). In addition, older dwarf mistletoe infections are more susceptible to attack by secondary pathogens and mechanical damage. Simulation models of dwarf mistletoe infection and spread have been developed for lodgepole pine, ponderosa pine, and western hemlock stands (Bloomberg, Smith, and van der Wereld 1980). More recent models include the Dwarf Mistletoe Spread and Intensification model (DMSI) developed by ESSA in cooperation with the U S D A Forest Service and the Dwarf Mistletoe Impact Modeling System (DMLM), which was designed to run in conjunction with the Forest Vegetation Simulator (FVS, also known as PROGNOSIS); (Hawksworth et al. 1995). These models are important for estimating the long-term impact of different control strategies and are used to determine which control strategy will work best under certain conditions. 1.3.0 Management strategies for dwarf mistletoes Dwarf mistletoe has been associated with extensive timber loss in many parts of the world, particularly in North America. Because of this, effort has been made to develop efficient control methods for this parasite. Management of dwarf mistletoe within a stand requires detection, evaluation, prevention, and suppression of the pathogen (Johnson and Hawksworth 1985). Because of the life cycle and small size of dwarf mistletoe, it is not always easy to detect early stages of infection in the field. This is especially true when the shoots have not yet developed. In addition, various stem deformities or accumulation of organic debris may be misidentified as infections, and high up infections may be hidden by dense foliage and difficult to detect (Shaw, Freeman, and Mathiasen 2000). In a study of the accuracy of ground-based hemlock dwarf mistletoe rating, one observer rated 98% of the 10 trees having infection, when in fact a crane survey estimated only about 53% of the trees having infection (Shaw, Freeman, and Mathiasen 2000). Other observers underestimated the amount of infection in the field, indicating that there is a range in the accuracy of ground-based dwarf mistletoe rating and often strong observer bias. In general, the easiest indication of dwarf mistletoe infection is the presence of shoots or basal cups. A polymerase chain reaction (PCR) technique designed to detect early dwarf mistletoe infection within the host branch has recently been developed (Marler et al. 1999) and may prove useful in detecting dwarf mistletoe infections before visible signs of infection become established. Despite some difficulties in detecting dwarf mistletoe in the field, numerous dwarf mistletoe-rating systems have been developed. Hawksworth and Lusher (1956) produced a 6-class rating system for dwarf mistletoe infection (as cited in Hawksworth 1977). The 6-class rating system is used to quantify the degree of infection so that stand management priorities can be established, and to quantify the mistletoe-infection hazard of overstory trees or stands to understory stands. In addition, the rating system helps estimate growth loss and mortality and define which trees are suitable for seed trees (Hawksworth 1977). The first step for determining the disease rating of an infected tree involves dividing the live crown into thirds. Each third is then rated as: 0, no mistletoe, 1, light mistletoe (less than half of the branches infected) or 2, heavy mistletoe (more than half the branches infected). As a last step, the ratings of each third are added together to give a final disease rating for the tree. Although it was observed that field personnel have a bias to either over or under estimate the true dwarf mistletoe rating, the 6-class rating system is still useful for predicting management implications and understanding the biology of dwarf mistletoe (Shaw, Freeman, and Mathiasen 2000). Prevention is much a more effective means of control than removing dwarf mistletoes after they have infected a particular stand. With proper management of tree stands, it is possible to reduce the spread of dwarf mistletoe to nearby cut blocks. For example, designing treatment blocks that take advantage of natural or man-made barriers (i.e. rivers, roads) will reduce the spread of dwarf mistletoe from nearby infected stands (Johnson and Hawksworth 1985). It is also advisable to use clear cuts 20 acres (8 ha) or 11 larger in infected stands, and to make sure regenerating stands are disease free. The use of resistant species may also reduce dwarf mistletoe spread and intensification, and maintaining rapid tree growth can sometimes allow hosts to outgrow dwarf mistletoe infection (Richardson and van der Kamp 1972). Ideal strategies for management of forest pests should be cost effective, efficacious, environmentally sound, and sustainable. Prior to logging, forest fires were the main way in which dwarf mistletoes were eradicated from diseased stands (Alexander and Hawksworth 1975). Fire sanitized dwarf mistletoe infected stands by killing the host (and the parasite) and changing forest composition (Johnson and Hawksworth 1985). Once logging became established, silvicultural control of dwarf mistletoe was most commonly used. Today, alternative methods of control for dwarf mistletoes, such as chemical, genetic, and biological control are presently being sought (Shamoun and DeWald 2002). 1.3.1 Traditional control of dwarf mistletoes Dwarf mistletoes are ideal candidates for silvicultural management because they require a living host, have high host specificity, have a long life cycle, spread relatively slowly, and are easy to identify in the field when signs of disease are present (Johnson and Hawksworth 1985). Silvicultural control can be considered a natural means of management that uses preventative and suppressive mechanisms. Because of the relatively long life cycle of dwarf mistletoes, it can take many years and several generations of dwarf mistletoe infection before levels reach damaging proportions (Scharpf and Parameter 1967). If silvicultural control is successful and dwarf mistletoe is removed from a stand during the first rotation, then dwarf mistletoe should not be a serious problem for the subsequent rotations (Baranyay and Smith 1977). Prevention and removal of infected trees can therefore be sufficient in many cases for control. Occasionally, selective burning of infected trees can also be employed as a control strategy for dwarf mistletoe (Johnson and Hawksworth 1985). Due to the high variability of dwarf mistletoe disease incidence, a suitable control method must be determined on a case by case situation and take into account stand composition, stand age, number of years to harvest, disease incidence and pattern and 12 length of time the stand has been infected (Baranyay and Smith 1977).. In addition, acceptable forest practices for the region of interest must be considered. The most effective way to control dwarf mistletoe is to remove infected trees at harvest, particularly taller infected trees that may spread infection more rapidly (Anonymous 1985). This silvicultural approach uses either clearcutting or selective removal of diseased trees to remove dwarf mistletoe from a stand. Clearcutting is arguably one of the best ways to control dwarf mistletoe in that it often removes all susceptible trees from a stand, thereby eliminating dwarf mistletoe infection. In some instances, small advance vegetation that was not removed during clearcutting may have residual dwarf mistletoe infection and require further treatment. Larger cut block sizes in infected stands are superior to smaller cut sizes because they reduce the ratio of edge residual trees to regenerating stands (Johnson and Hawksworth 1985; Buckland and Marples 1952). In many cases, clearcutting can lead to total and long term control of dwarf mistletoe. A n alternative to clearcutting is selective removal of dwarf mistletoe infected trees, which can be integrated into the logging operations and typically involves mechanical removal or in some cases burning of the infected trees. However, there are several disadvantages to using selective removal as a means to control dwarf mistletoe. Because of the latency period in most dwarf mistletoe life cycles, detection and complete eradication of this disease using sanitation cutting and pruning is a virtual impossibility (Scharpf and Parameter 1967). In addition, intermediate cutting of dense dwarf mistletoe infected stands may reactivate dormant mistletoe infections by increasing light within the stand (Richardson and van der Kamp 1972; van der Kamp 1986). Thinning a stand may also reduce obstructions to dwarf mistletoe seed dispersal. These factors combined may temporarily increase the rate of upward advance and intensification of dwarf mistletoe within the stand (Richardson and van der Kamp 1972). Current forest practices in BC (online at http://www.for.gov.bc.ca/tasb/legsregs/ fpc/) make silvicultural control of dwarf mistletoe more difficult than in the past by reducing the size of clearcut blocks and increasing riparian reserves and the use of partial cutting. Research has shown that the infection level of second growth trees was proportional to the number of residual diseased trees, indicating that current forest 13 practices may favour the spread of dwarf mistletoe (Bloomberg and Smith 1982). Furthermore, silvicultural control of dwarf mistletoe is sometimes not possible. For example, it is not always preferable to mechanically remove dwarf mistletoe infected trees from parks, watersheds, or riparian zones (Hawksworth and Wiens 1996). Because of these difficulties, alternative methods for controlling dwarf mistletoes are presently being investigated. These include chemical, genetic, and biological control. 1.3.2 Chemical control of dwarf mistletoes Chemical control is advantageous in that it does not require mechanical removal of the host and is less labour intensive than silvicultural control. Chemical control of dwarf mistletoes in ornamental landscapes is especially alluring because the only alternative method for control, pruning, is expensive, laborious and reduces the aesthetics of the host (Adams, Frankel, and Lichter 1993). However, there has been little success in developing a chemical herbicide for dwarf mistletoe. Nearly 60 different chemicals were tested against dwarf mistletoe in the 1960's and 1970's, but none proved effective in killing the dwarf mistletoe infection without harming the host (Johnson, Hilderbrand, and Hawksworth 1989). In some cases, application of test herbicides damaged the host tree and had little impact on the dwarf mistletoe infection (Quick 1964). Today, many of these chemicals are considered unsafe and have been removed from the market. Ethephon [(2-chloroethyl) phosphoric acid] is the most promising chemical herbicide for dwarf mistletoe (Hawksworth and Johnson 1989). Ethephon is registered in the US under the trade name Florel® but is not registered in Canada. Its mechanism of action involves the release of ethylene during absorption by plant tissue, which speeds up the natural ripening process and results in abscission of dwarf mistletoe shoots. Shoot abscission rates for Arceuthobium vaginatum treated with ethephon ranged from 73-98 % (Johnson, Hilderbrand, and Hawksworth 1989). Ethephon does not ki l l the dwarf mistletoe endophytic system, however, and new dwarf mistletoe shoots may develop from the dwarf mistletoe swelling. Retreatment may therefore be required. Although ethephon does not ki l l the dwarf mistletoe infection, it may still be effective in reducing dwarf mistletoe in the field. Applying ethephon three to five years 14 r after thinning a dwarf mistletoe infected stand may lessen the spread and establishment of dwarf mistletoe long enough for the stand canopy to close in, thereby shading out any remaining dwarf mistletoe infections (van der Kamp 1986). Unfortunately, there are as of yet some unresolved problems with using ethephon in the field. Toxicity of ethephon has been observed in some host trees (Adams, Frankel, and Lichter 1993). Needle chlorosis and branch dieback occurred in black spruce and lodgepole pine after treatment with ethephon (Livingston and Brenner 1983; Frankel and Adams 1989). In addition, there is growing public resistance to the application of chemical herbicides in the field (Jobidon 1991). 1.3.3 Genetic control of dwarf mistletoes Breeding for resistance is applicable to situations where exotic pathogens are causing disease in trees that have not naturally evolved resistance or to native pathogens that have been intensified through silviculture. In addition, breeding for resistance is appropriate for situations where a high level of natural resistance to a native pathogen has failed to develop (Roth 1974). Dwarf mistletoes apply to the last of these conditions. There are several criteria to consider when searching for a resistant host. Smith, Wass, and Meagher (1993) suggest that both the frequency and the vigour of dwarf mistletoe infection should be considered when searching for resistant western hemlock stock. In addition, inoculation procedures testing for resistance should be consistent in the age of branch segment, branch order, and positioning of the seeds along each segment. Because of the slow rate of spread and uneven distribution of dwarf mistletoe in the field, development of dwarf mistletoe host resistance is not as advanced as that of other hosts with air-borne, more evenly distributed pathogens such as foliage diseases and stem rusts (van der Kamp 1991). The unequal exposure to seed source and infection variables makes it difficult to detect resistant varieties in the field (Roth 1974). As a result of the difficulties in detecting resistant hosts in the field, inoculation trials have been used to screen for host resistance to dwarf mistletoe. Besides having fewer or no dwarf mistletoe infections, resistant hosts may have slower dwarf mistletoe endophytic development or fewer dwarf mistletoe shoots (Smith 1974). For example, infection of larch and Douglas fir by Arceuthobium tsugense resulted in slow endophytic 15 development (Smith 1974). Only 1 of 38 western larches infected by A tsugense produced shoots, and shoot development was slower than that observed for susceptible hosts. Studies screening for variation in resistance against dwarf mistletoe infection have been much more rigorous between host species than within-host species (Shamoun and DeWald 2002). Research supports the existence of natural within-host species resistance to dwarf mistletoe infection. One inoculation trial found western hemlock and ponderosa pine varied in their susceptibility to Arceuthobium tsugense and A. campylopodum, respectively (Smith, Wass, and Meagher 1993). Within-host species resistance has also been observed in Arceuthobium vaginatum subsp. cryptopodum on ponderosa pine, A. campylopodum on ponderosa pine and Jeffery pine (Pinus jeffreyi Greville & Balfour), and A chinense on Evelyn keteleeria (Keteleeria evelyniana Mast.); (Bates 1927; Roth 1953; Scharpf 1984; Tong and Ren 1980; as cited in Hawksworth and Wiens 1996). Resistance mechanisms may occur at any stage of the dwarf mistletoe life cycle and range from a reduction in seed germination to apparent host immunity (Smith 1974). The ability of dwarf mistletoe seeds to remain attached to a host and penetrate the host surface is the first point at which some resistance may occur. Conifers with drooping or open foliage, such as western larch and pines, have been shown to have low dwarf mistletoe seed retention (Smith 1974). In addition, selection for shorter needle bases may reduce the penetration leverage of germinated dwarf mistletoe seeds and reduce the rate of successful infection (Hunt, Owens, and Smith 1996). Host response to infection may also impart resistance. In the case of red firs, penetration by Arceuthobium campylopodum f. sp. abietinum seeds was resisted by the development of an abscission zone at the base of the needle (Scharpf and Parmeter 1967). The abscission zone was followed by the development of wound cork, which inhibited penetration by the parasite. In another study, Tainter and French (1971) observed the formation of wound periderm in response to infection of eastern larch (Larix laricina (DuRoi) K . Kosh.) by Arceuthobium pusillum. As more research breeding and screening for genetic resistance to dwarf mistletoe infection takes place, a deeper understanding of the mechanisms involved in resistance wil l emerge. Recently, Nowicki, DeWald, and Moser (1999) found three neutral markers 16 with alleles using isozyme analysis that appear to be specifically associated with Arceuthobium douglasii infected Douglas fir trees. The purpose of these alleles is not fully understood, and more research into this field is required. Resistance screening programs for Douglas fir and western hemlock are currently underway at Northern Arizona University and the Canadian Forestry Service, Pacific Forestry Centre (Cartwright et al. 2003; Shamoun and DeWald, 2002). 1.3.4 Biological control of dwarf mistletoe Although the concept of using plant pathogens to control forest pests is not new, development of biopesticide technology is still in its infancy (Evans 1995). The biological control approach involves enhancing natural plant pathogens such that the pathogen reduces the vigour of unwanted pests. The aim of an efficient biological control strategy is not complete eradication of the pest, which may upset the ecosystem, but rather to effectively manage the pest by reducing inoculum (Shamoun 1997). There are two types of biological control: classical control and inundative control. Classical biological control is based on the phenomenon that introduced organisms, in this case a pathogen or insect, may achieve dominance in a new environment because they will have few or no natural enemies or competitors (Evans 1995). Because of growing concern about introducing new organisms, the use of classical control is losing attractiveness. However, it is still promising as a means to control introduced organisms. Inundative biological control, on the other hand, uses indigenous pathogens to control target forest pests. This approach involves mass production and application of inoculum on the pest such that the pathogen interferes with the pest life cycle (Evans 1995). Only a small number of inundative biological control agents have been registered for use against forest pests. Chondrostereum purpureum (Pers.:Fr) is a wood-inhabiting fungus that invades the cambium and sapwood of mostly hardwood trees and shrubs. It is currently marketed as a wood decay promoter and as a biological control agent for woody weeds in Europe under the trade name BioChon™ (Ravensberg 1998; as cited in Evans, Frohlich, and Shamoun 2001). Chondrostereum purpureum is also registered in eastern Canada as Myco-Tech ™. A collaborative research project between MycoLogic 17 Innovative Biologicals Inc and the Canadian Forest Service is in the process of registering a different isolate of C. purpureum in western Canada under the tradename Chontrol™ (S. F. Shamoun, Pacific Forestry Centre, Victoria, BC, pers. comm.). Stumpout™ is another example of an inundative biological control agent used to control forest pests. Stumpout™ was registered in South Africa in 1997 and is currently used as a mycoherbicide to prevent re-growth of cut wattle stumps and to control introduced tree species (Morris, Wood, and Den Breeyen 1998; Lennox, Morris, and Wood 1999; as cited in Evans, Frohlich, and Shamoun 2001). It consists of the white-rot fungus Cylindrobasidium laeve, which is known to occur in South Africa and North America (Ginns and Lefebvre 1993; as cited in Evans, Frohlich and Shamoun 2001). Inundative biological control can also be used to control forest pathogens. For example, the wood rotting fungus Phlebiopsis gigantea (Fr.) Jul. has been used to reduce the spread of the causal agent of annosus root and butt rot, Heterobasidion annosum (Fr.) Bref. (Rishbelth 1963). Inoculation of fresh cut pine stumps with formulated spores of P. gigantea reduced colonization of the stump by H. annosum. Phlebiopsis gigantea been registered in the United Kingdom as P G Suspension™ and in Finland as RotStop™ (Rishbelth 1963; Roy et al. 2003; Pratt 1999). Biological control of the root pathogen Armillaria ostoyae (Romagn.) Herink using Hypholoma fasciculare (Huds. Ex. Fr.) has also been investigated in Canada (Chapman and Xiao 2000). Although there is potential for biological control of forest pests and pathogens, the majority of inundative biological control research has focused on weeds in agricultural systems. Because of this, more biological control agents have been registered for agricultural weeds than for forest pests and pathogens. Devine®, a formulation of Phytophthora palmivora (Butl.) Butl., was registered to control Morrenia oderata Lindl., strangler vine or milkweed vine, in citrus grooves (Ridings 1986). BioMal™ was registered in 1992 for control of round leaf mallow in Canada using Colletotrichum gloeosporioides Penz. Sacc. f. sp. malvae (Makowski and Mortensen 1992). However, due to the high cost of inoculum production, BioMal™ was never marketed (Watson et al. 2000). Potential biological control agents for management of dwarf mistletoe must meet certain criteria. These attributes include: 1) a distribution which coincides with that of the 18 target pathogen, 2) ecological amplitude sufficient to assure persistence within its host range, 3) production of abundant inoculum (spores) for establishment of epiphytotics, 4) high infectivity, 5) high virulence, and 6) an efficient mode of action for curtailing development of target disease (Wicker and Shaw 1968). In addition, the ideal biological-control agent should be able to cause infection without the necessity of wounding, and be easily formulated for field application and long-term storage (Shamoun 1997). Some insects endemic to Pakistan have been identified as potential biological control agents for dwarf mistletoes (Mushtaque and Baloch 1979). Mushtaque and Baloch (1979) observed 27 insects and mites known to damage mistletoes. Six of these occurred in reasonable numbers, appeared to be restricted eaters, and caused appreciable u damage to dwarf mistletoe. Although tests on the biology and host specificity of the 6 insects showed at least 3 of them are promising potential biological control agents for dwarf mistletoe, further research into their use in North America has not been done. Numerous fungi have been isolated from dwarf mistletoe infections (Baranyay 1966; Hawksworth, Wicker, and Scharpf 1977; Kuijt 1963). Some of these fungi appear to parasitize the dwarf mistletoe shoots, while others have been isolated from diseased dwarf mistletoe swellings. 1.3.4.1 Hyperparasites of dwarf mistletoe shoots Pathogens that attack dwarf mistletoe shoots are advantageous in their rapid disease development and direct impact on flower and berry development. These pathogens are useful in that they can reduce the ability of dwarf mistletoe to spread and intensify in the field. However, shoot pathogens may not enter the dwarf mistletoe swelling, and therefore do not eradicate the dwarf mistletoe infection (Wicker and Shaw 1968). At least six hyperparasites are known to attack the shoots of dwarf mistletoes in North America. The most common fungi with the widest dwarf mistletoe host range are Caliciopsis arceuthobii (Peck.) Barr., Cylindrocarpon gillii (Ellis) J. A . Muir, and Colletotrichum gloeosporioides (Pen.z) Penz. & Sacc. The other fungi (Pestalotia heterocornis Guba, Cylindrocarpon sp., Pestalotia maculiformans Guba & Zeller, 19 i Metaspheria wheeleri Lider, and Alternia alternata (Fr.) Keissler) are rare or only occur locally and will not be discussed (Hawksworth, Wicker, and Scharpf 1977). Caliciopsis arceuthobii (formerly Wallrothiella arceuthobii (Pk.) Sacc.) parasitizes the pistillate flowers of Arceuthobium pusillum and A douglasii in the United States and Canada (Peck, 1875; as cited in Wicker and Shaw 1968) as well as A. americanum and, rarely, A. vaginatum subsp. cryptopodum (Hawksworth, Wicker, and Scharpf 1977). The fungus is restricted to the floral organs of the pistillate shoots of dwarf mistletoes and prevents normal seed development. Caliciopsis arceuthobii occurs in two geographical locations: the northern location (BC, Oregon, Washington, Idaho, Montana) and the southern location (Colorado, Arizona, New Mexico, and Mexico); (Hawksworth, Wicker, and Scharpf 1977; Ramsfield 2002). The fungus has not been found in the region between these two locations. It has been suggested that C. arceuthobii may only be able to infect spring flowering dwarf mistletoe species (Kuijt 1963), but this remains speculation. The extent of damage caused by C. arceuthobii is variable, more than 90% of one year's crop of berries may be destroyed one year, and yet the next year the fungus may be absent from the same location (Hawksworth Wicker, and Scharpf 1977). A study conducted by Wicker and Shaw (1968) suggests that the amount of disease within a stand may vary from high to absent within as little as 2 years. In addition, although C. arceuthobii has been grown in culture, spores have not been produced in vitro (Knutson and Hutchins 1979; Parker 1970). These combined limit the use of this fungus as a biological control agent for dwarf mistletoe. Cylindrocarpon gillii (formerly Septogloeum gillii Ellis) attacks all parts of the dwarf mistletoe shoots (Hawksworth Wicker, and Scharpf 1977; Muir 1973). Known hosts of C. gillii are Arceuthobium abietinum f. sp. concoloris, A. abietinum f. sp. magnificae, A. americanum, A. apachecum, A. blumeri, A. californicum, A. campylopodum, A. cyanocarpum, A. divaricatum, A. douglasii, A. laricis, A. microcarpum, and A tsugense (Hawksworth Wicker, and Scharpf 1977). This fungus occurs more prevalently on the east slopes of the Cascade Mountains rather than in eastern Washington. It is also widely distributed in western parts of Canada, including B C and Alberta (Kuijt 1963). Early symptoms of infection are small, yellowish-white lesions on the shoots, which gradually enlarge, coalesce, and erupt through the epidermis, 20 releasing conspicuous, masses of white spores (Ellis 1946). The hyphae of the fungus invade all the tissues of the shoot except for the vascular elements (Ellis 1939). The shoot is often girdled, killing the distal portion of the shoot. Cylindrocarpon gillii is similar to Caliciopsis arceuthobii in its ecological amplitude, and can vary in the amount of disease from high to absent within a couple of years (Wicker and Shaw 1968). A field trial conducted by Mielke (1959) that involved inoculating A. .americanum with C. gillii found the parasite died out on four experimental plots only 3 years after inoculation. The lack of rain, which is required for spore dissemination and infection, was used to explain the reduction of C. gillii observed. Previous studies by Ellis (1946) showed that C. gillii is not a wound parasite and is able to penetrate unwounded host tissue. Abundant inoculum production in culture is also possible with this fungus. Colletotrichum gloeosporioides parasitizes the shoots of many dwarf mistletoe species, including Arceuthobium abietinum f. sp. concoloris, A. abietinum f. sp. magnificae, A. americanum, A. apachecum, A. californicum, A. campylopodum, A. divaricatum, A. douglasii, A. laricis, A. microcarpum, A. occidentale, and A. tsugense (Hawksworth Wicker, and Scharpf 1977). The first symptoms of infection by C. gloeosporioides involve the development of small, necrotic lesions on the shoots. As the disease progresses, infection causes shoot dieback or blight, often resulting in girdling of the shoots. Signs of disease can be seen on the shoots, where acervuli and setae may occur. The mycelium of C. gloeosporioides is entirely intercellular and occurs primarily in the exterior portion of the shoot cortex. Colletotrichum gloeosporioides is more persistent than Caliciopsis arceuthobii and Cylindrocarpon gillii in terms of disease abundance over time. Abundant inoculum production in culture is also possible with this fungus (Wicker and Shaw 1968). Recently, a study assessing the use of C. gloeosporioides as a biological control agent for lodgepole pine dwarf mistletoe (Arceuthobium americanum Nutt. ex Engelm.) was conducted in the BC interior (Ramsfield 2002). Results from this trial were not able to show a significant reduction in mistletoe seed production as a result of inoculation with C. gloeosporioides. However, the fungus was not dismissed as a potential biological control agent for A. americanum and merits further investigation 21 Other, more location and species-specific fungi that parasitize dwarf mistletoe shoots have also been observed. For example, the brown felt fungus Herpotrichia juniperi (Duby) Petr. is known to colonize and kil l shoots of Arceuthobium abietinum f. sp. magnificae growing on California red fir in the higher elevations of the southern cascades in California (Scharpf 1986). In this case, H. juniperi is not a feasible biological control agent because of its limited host range and its requirement of persistent snow pack to develop. , 1.3.4.2 Hyperparasites of dwarf mistletoe swellings Fungi that attack dwarf mistletoe swellings are advantageous in that disease symptoms are pronounced and the pathogen may possibly ki l l the dwarf mistletoe infection. However, as the host tree may be damaged as well, hyperparasites of dwarf mistletoe swellings must undergo additional laboratory study (Shamoun and DeWald 2002). Cytospora abietis Sacc, is associated with dwarf mistletoe swellings and is common on California red fir and white fir parasitized by Arceuthobium abietinum (Scharpf 1969; Scharpf and Bynum 1975). This canker-causing fungus can invade the dwarf mistletoe swelling and kil l the host branch. On occasion, C. abietis parasitizes non-mistletoe infected branches. Further research into the interactions between C. abietis, A. abietinum, and its host is required before evaluation of this fungus as a potential biological control agent for dwarf mistletoe can begin (Shamoun and DeWald 2002). Resin disease has been observed on Arceuthobium americanum swellings on lodgepole pine. The symptoms of resin disease include excessive fesin exudation from the dwarf mistletoe swelling, necrotic and discoloured bark tissue, repeated formation of necrophylactic periderms, retention of dead pine needles, and dead, resin infiltrated dwarf mistletoe shoots (Mark, Hawksworth, and Oshima 1976). Several weakly pathogenic fungi have been shown to cause resin-disease symptoms; with Alternaria alternata (Fries : Fries) von Keissler and Aureobasidium pullulans (de Bary) Arnaud most commonly isolated from diseased A. americanum swellings (Mark, Hawkwworth, and Oshima 1976). Neonectria neomacrospora (Booth & Samuels) Mantiri & Samuels (formerly Nectria neomacrospora Booth & Samuels) is also known to parasitize dwarf mistletoe swellings. This fungus has been observed as a virulent pathogen of A. occidentale swellings on pine;> A. abietinum on white fir, and A. tsugense swellings on hemlock (Byler and Cobb 1972a; Funk and Baranyay 1973; Funk, Smith, and Baranyay 1973). Neonectria neomacrospora is discussed in greater detail in section 1.5. 1.4.0 Introduction to Arceuthobium tsugense 1.4.1 Symptoms and signs of A. tsugense Hemlock dwarf mistletoe (Arceuthobium tsugense (Rosendahl) G. N . Jones) causes non-systemic infection of the host branch, often resulting in swelling formation. Shoots range in colour from greenish to reddish and the average shoot is 5-7 cm tall. Mature fruit occur on some female infections in late summer and early fall. The occurrence of basal cups on the dwarf mistletoe swelling is also a good sign of disease (Hawksworth and Wiens 1996). The formation of witches' brooms is fairly common for A. tsugense infections on hemlock and can develop as soon as 14 years after exposure to inoculum (Smith 1977). Because of the erratic growth pattern of hemlock, it is sometimes difficult to detect low levels of A. tsugense infection based on symptoms alone. Although the PCR technique described by Marler et al. (1999) used primers that were designed to detect Arceuthobium douglasii and A. laricis in host tissue, it is expected that these same primers will work for A. tsugense and other dwarf mistletoe species. 1.4.2 Biology of A. tsugense Studies monitoring A. tsugense seed dispersal and natural development of A. tsugense infection in the field have been done in Alaska, B C , and Oregon (Shaw and Loopstra 1991; Carpenter, Nelson, and Stewart 1979; Smith 1971, respectively). Field observations show that about 50% of seeds that land on foliage moved to twigs; germination on the host needles was rare (Carpenter, Nelson, and Stewart 1979). Of the A. tsugense seeds that landed on a susceptible host and remained attached overwinter, approximately 60-65% germinated and 6-33% established infection. Although the highest 23 percentage of successful infection was observed for the trial in Alaska, A. tsugense shoots were often broken and damaged and no seeds were produced over the 11-year trial period (Shaw and Loopstra 1991). Differences in seed germination and infection were attributed to pathogen attack, environmental conditions and self-pruning of the host branch. Based on close observations of the wedging of the holdfast in the acute angle of the needle base, the levering off of the needle, and the crushing of host cells, Hunt, Owens, and Smith (1996) support the importance of mechanical force for germinated A tsugense seeds to penetrate the host branch. No study of enzymatic activity has been performed; it is still not clear i f enzymatic activity works in combination with mechanical forces. Arceuthobium tsugense develops a localized infection on susceptible hosts. Swellings can develop as early as the first year following A. tsugense infection, but in most cases do not develop until the second year or later. The endophytic structure of A. tsugense consists of the haustorial (cortical) strands and sinkers similar to those observed in other dwarf mistletoe species. However, A. tsugense differs from some other species in that both the inner and outer haustorial strands bear sinkers, and the outer haustorial strands bear two distinct sheaths (Hunt, Owens, and Smith 1996). Cohen (1954) did not observe sinker cells on the outer haustorial strands of A. campy lopodum. Tainter (1971) described only one sheath layer in A. pusillum Peck.. In addition, the A. tsugense endophytic system lacks sieve elements (Calvin, Hawksworth, and Knutson 1984). The maximal longitudinal growth rate of the endophytic system of A. tsugense was approximately 1.5cm/year on hemlock and 6.6 cm/year on shore pine (Smith, 1971). Initial A. tsugense shoots develop in the second and third years after infection. The life span of A. tsugense shoots varies from less than 1 year up to 7 years, averaging 2-3 years. The average number of shoots per infection for 3-7 years after infection was approximately 10 shoots (Smith 1971). The growth rate of healthy, undisturbed male and female shoots is the same for the first two years following emergence of the shoots (approximately 20 mm by year 1 and 45-55 mm by year 2); (Smith 1971). The growth rate decreases for both male and female shoots after 2 years, and by the third year the female shoots are typically taller than the male shoots. 24 Flowering typically occurs 3-4 years after infection between July-August, and seeds are dispersed in late September and early October the following year. Both male and female shoots can produce at least 3 crops of flowers, although the average is much less as many shoots die before flowering or after producing only one crop of flowers (Smith 1977). Once the seeds are discharged, the life cycle is repeated. Dwarf mistletoes are perennial and may produce more than one crop of flowers and, i f female, berries. The location of the A. tsugense infection within the host may influence A. tsugense shoot production. For light, moderate, and severe host infections, the number of shoots observed in the top portion of the crown was equal to or greater than the middle and lower portions of the crown (Smith 1969). For all classes, the proportion of infections with shoots or fresh cups was greatest for the top portion (96%) when compared to the middle portion (84%) and the lower portion (50%). Recent work has developed a means to germinate A. tsugense seeds in vitro using Harvey's medium with varying concentrations of auxin (2,4-dichlorophenoxyacetic acid) arid cytokinin (6-benzylaminopurine); (Deeks, Shamoun, and Punja 1998; Deeks Shamoun and Punja 2001). Development of a technique to grow dwarf mistletoe in vitro is important because it could provide detailed information on dwarf mistletoe physiology and aid in developing chemical or biological control (Gill and Hawksworth 1961). 1.4.3 Distribution of A. tsugense Arceuthobium tsugense is one of four dwarf mistletoes found in BC. The species occurs along the Pacific coast from near Haines, Alaska to central California, and is particularly prominent on the Queen Charlotte Islands, northern Vancouver Island, Texada Island, and on the mainland near Prince Rupert and Vancouver (Baranyay and Smith 1977). In regions where A. tsugense occurs, the parasite can often reach high levels of infection. A detailed survey of southeastern Vancouver Island found an average of 85% of pine trees over 1.4 m tall were infected with A. tsugense (Wass 1976). For reasons unknown, A. tsugense is not found in the hemlock stands in BC's Interior Wet Belt. Inoculation trials of western hemlock in the Interior Western Hemlock Zone using A. tsugense inoculum from the coast found that A. tsugense was able to establish and develop normally in the Interior (Smith and Wass 1979). The interior population of 25 hemlock in B C and the United States is geographically separated from the coastal population. 1.4.4 Taxonomy and hosts of A. tsugense Because of its unusually broad host range, it has been recommended that A. tsugense be separated into three subspecies (mertensianae, tsugense and contortae); (Smith 1974; Wass and Mathiasen 2003). Two of the subspecies differ somewhat morphologically. The shoots of A. tsugense subsp. tsugense are 30% taller than A. tsugense subsp. mertensianae. In addition, subsp. tsugense flowers 1-2 weeks prior to subsp. mertensianae (Hawksworth and Wiens 1996). It is unclear i f morphological differences are due to genetic or environmental factors. The primary host for A. tsugense subsp. mertensianae is mountain hemlock (Tsuga mertensiana (Bong.) Carr.). Pacific silver fir (Abies amabilis Dougl. Forbes), subalpine fir (A. lasiocarpa (Hook.) Nutt. var lasiocarpa), and noble fir (A. procera Rehder) have also been reported as principal hosts for A. tsugense subsp. merensianae (Hawksworth and Wiens 1996). Whitebark pine (Pinus albicaulis Engelm.) is a secondary host, and western white pine (P. monticola Dougl.) is an occasional host (Hawksworth and Wiens 1996). Arceuthobium tsugense subsp. tsugense has western hemlock as a primary host, and shore pine (Pinus contorta Dougl. ex Loud var. contorta) and Rocky Mountain lodgepole pine (Pinus contorta Dougl. ex Loud var. latifolia Engelm.) as occasional hosts. In addition, Pacific silver fir, subalpine fir, and noble fir have also been tentatively reported as principal hosts for A. tsugense subsp. tsugense (Mathiasen 1994). Further division of A. tsugense subsp. tsugense into two races (shore pine and western hemlock) was under debate for many years. Reports did not agree on the significance of differences in morphology and geographical distribution between the shore pine and western hemlock races (Wass 1976; Wass and Mathiasen 2003). However, it was apparent that the two races differed in host affinities (Smith, 1974; Smith and Wass 1976,1979; Wass 1976; Wass and Mathiasen 2003). Field observations by Smith and Wass (1976) found that western hemlock appeared more susceptible to infection originating from A. tsugense parasitizing western hemlock than to A. tsugense 26 parasitizing shore pine. Shore pine also appeared more susceptible to infection originating from A. tsugense parasitizing shore pine than from A. tsugense parasitizing western hemlock. Inoculation trials concurred with these observations (Smith and Wass 1979). Low levels of infection occurred when shore pine was inoculated with seeds from the western hemlock race. In contrast, inoculation of western hemlock with seeds from the shore pine race resulted in moderate levels of infection, but only a sub-portion of successful infections produced shoots. Based on these observations, Wass and Mathiasen (2003) elevated the shore pine race to the species level as Arceuthobium tsugense subsp. contortae. The hosts for A. tsugense subsp. contortae are shore pine, occasionally hemlock, and rarely western white pine. 1.4.5 Spread and intensification of A. tsugense Arceuthobium tsugense is arguably one of the fastest spreading dwarf mistletoes in BC. The life cycle of A. tsugense may be completed in as little as 4 years; other dwarf mistletoe species typically take 5-6 years to complete their life cycle (Baranyay and Smith 1977). Infection rates of A. tsugense are highly variable and depend on environmental conditions and site productivity (van der Kamp 1986). Richardson and van der Kamp (1972) found the number of new A. tsugense infections per host tree can double every four years. In another study, the number of infected hemlock surrounding infected residuals increased logarithmically from 5-30 years after logging (Stewart 1976). Knowledge of the relatively short life cycle and rate of spread of A tsugense is important in planning silvicultural and other control methods (Smith 1971). The productivity of a site may influence the impact of A. tsugense infection on hemlock height growth. That is, there may be a critical rate of hemlock height growth above which A. tsugense infection no longer causes incremental loss (Richardson and van der Kamp 1972; van der Kamp 1986). In a good site, trees are expected to have a fast growth rate, and because of this may be able to outgrow A. tsugense infection and spread. The critical rate of height growth is expected to vary accordingly with site productivity and tree age (Richardson and van der Kamp 1972). Stand density may influence A. tsugense spread and intensification. In one estimate, the vertical spread of A. tsugense in coastal western hemlock in southern BC 27 was approximated to be 30 cm/year in a dense stand and 60 cm/year in an open stand (Richardson and van der Kamp 1972). Over a 5 year period, an average of 26 new infections per tree and a vertical rate of spread of 0.58 m/year was observed for the open stand, and 43 new infections per tree and a vertical rate of spread of 0.27 m/year was observed for the dense stand. These data suggest that the rate of vertical spread and the rate of intensification may act independently of one another. For A tsugense infection of hemlock, Smith (1969) modified the 6-class rating system so that only the middle third of the hemlock live crown is classified (Hawksworth 1977). This approach was taken because the bottom portion of hemlock branches are often dead, broken, or missing, and the top portion of the tree is often partially hidden. In addition to containing the greatest number of living infections, the middle-third assessment was more effective and easier to use than the 6-class rating system for the whole-tree. A simulation model of A. tsugense infection on hemlock has also been established (Bloomberg, Smith and van der Wereld 1980). The computer model D W A R F was created to predict spread and intensification of A tsugense under variable conditions. D W A R F takes into consideration stocking, size, and spatial distribution of residual source trees and of regeneration; severity of infection of residual source trees; topographic slope; tree growth rate and environmental effects (e.g. wind or early frost) on mistletoe seed production, dispersal, and infection. When comparing predicted model data to actual data, the predicted average number of infections per tree was generally slightly lower than recorded number, but was not significantly different (Bloomberg, Smith and van der Wereld 1980). Applying the model to a broader range of conditions found the number of predicted infections averaged 107% of the number of recorded infections in residual trees and 128% of the infections in second-growth trees (Bloomberg and Smith 1982). Bloomberg and Smith (1982) concluded that the model predictions were generally compatible with recorded observations and provided an estimate of the effects of factors and their interactions on dwarf mistletoe incidence. 28 1.4.6 Impact of A. tsugense in forestry • • 3 In B C , the annual volume loss caused by dwarf mistletoe is about 4.2 million m of western hemlock and lodgepole pine (Baranyay and Smith 1977). Only decay fungi surpass this huge loss. Infection of hemlock by A. tsugense reduces the rate of host growth and wood density (Wellwood 1956). A reduction in lumber yield is expected to occur as a result of swellings, which produce abnormal tissue, develop large knots in the wood, and introduce secondary pathogens and decay (Wellwood 1956). Lightly A. tsugense infected hemlock trees may escape significant growth impact (Smith 1977). However, heavily infected A. tsugense hemlock trees can have significantly reduced host vigour. The growth of lightly infected trees was 84% and 41 % greater in height and volume, respectively, when compared to severely infected trees over a 7-year period (Smith 1969). At 15 inches diameter, the average lightly infected tree was 14 feet taller and 14 cubic feet greater in volume when compared to severely infected trees. These observations apply to a mature hemlock stand with individuals averaging 110 years old. One factor that contributes to reduced host vigour is the redirection and removal of nutrients and carbon compounds from the host to the A. tsugense infection. The ratio of CO2 fixed by photosynthesis to CO2 produced from respiration varied from 25-30% for A. tsugense shoots (Miller and Tocher 1975). The remaining carbon requirements are most likely derived from the host. Seasonal variation was also observed during the fall, winter and spring conditions, with A. tsugense having the highest rate of photosynthesis in the spring. Another factor that may influence host vigour is the introduction of secondary fungi via the dwarf mistletoe infection. The dead bark and cankered regions often associated with the A. tsugense swelling provide an ideal environment for invading fungi, as evidenced by the amount of stain and decay at or near most cankered areas (Baranyay 1966). Studies have shown that infection of hemlock by A tsugense can alter the bark moisture of invaded regions (Baranyay 1964). The extent and direction of impact on bark moisture varied depending on the age of the dwarf mistletoe infection, the time of year, and the growing site. 29 In one study, Bier (1959) noted that the relative turgidity of 5-10 year old bark of repressed hemlock trees was less (62-85%) than that of open grown trees of the same age (82-93%o). In addition, lower branches with thinner foliage were found to have lower relative turgidity when compared to more vigorous, higher up branches. Bier (1959) observed that the canker disease caused by Cephalosporium spp. only developed when the relative turgidity of the bark was below the threshold value of 80%, and suggested that measurements of the relative turgidity of the bark may provide a useful index of tree susceptibility to disease. 1.4.7 Control of A. tsugense Although silvicultural management is an effective means to control A. tsugense, current forest practices in BC are making this form of control difficult. Studies of host resistance to A. tsugense infection are underway, but as of yet no conclusive resistance mechanism has been found (Smith, Wass, and Meagher 1993, Cartwright et al. 2003). Chemical control of A. tsugense in BC is not feasible at present, as Ethephon is not registered in Canada and the use of chemical herbicides is loosing public favour. While developing a biological control strategy for A. tsugense is still in its early stages, some promising potential biological control agents have been observed. Fungal pathogens of A. tsugense have been collected from numerous locations throughout B C (Funk and Baranyay 1973; Shamoun 1997; Kope and Shamoun 2000). Cylindrocarpon gillii and Colletotrichum gloeosporioides have been isolated from diseased A tsugense shoots (Kuijt 1963; Kope, Shamoun and Oleskevich 1997; Kope and Shamoun 2000). In addition, Neonectria neomacrospora has been isolated from diseased A. tsugense swellings (Funk, Smith and Baranyay 1973; Kope and Shamoun 2000). 1.5.0 Introduction to Neonectria neomacrospora Neonectria neomacrospora (Booth & Samuels) Mantiri & Samuels (anamorph Cylindrocarpon cylindroides Wollenw.) has been found in Canada (Quebec and BC), parts of the United States (Oregon), and Europe (Norway); (Funk, Smith and Baranyay 1973; Byler, Cobb and Parameter 1972a, 1972b). Several sources have suggested the use of TV. neomacrospora as a biological control agent for dwarf mistletoe and western gall 30 rust (Funk, Smith and Baranyay 1973; Byler and Cobb 1972; Byler, Cobb and Parameter 1972a). 1.5.1 Taxonomy and hosts of A7, neomacrospora This ascomycete fungus belongs to the class Pyrenomycetes, the order Hypocreales, and the family Nectricacea. The teleomorph stage of N. neomacrospora has at least 5 synonyms. Originally named Nectria cucurbitula Tode ex Fr. var macrospora Wollenw, this species was renamed Nectria fuckeliana Booth after realization that the fungus had been misidentified. Later, Booth (1966) divided Nectria fuckeliana into two varieties, which were primarily distinguished on the basis of the size of their microconidia (Byler, Cobb and Parameter 1972b). These were Nectria fuckeliana Booth var fuckeliana (anamorph Cylindrocarpon cylindroidesvar tenue), and Nectria fuckeliana Booth var macrospora (Wr.) Booth (anamorph Cylindrocarpon cylindroides Wollenw). Ouellette (1972) raised Nectria fuckeliana var macrospora to the species level and renamed it Nectria macrospora (Wr.) Ouellette. This name, however, had already been given to previous fungal species, and it was changed to Nectria neomacrospora C. Booth & Samuels (Booth 1981). Nectria neomacrospora was considered a part of the genus Nectria s. lat. until recently, when the group was broken down into several genera. Based on its Cylindrocarponanamorph stage and the morphology of its perithecia, N. neomacrospora was moved into the new genus Neonectria (Rossman et al. 1999). Molecular work using mitochondrial ribosomal D N A sequencing supported the Neonectria genus as monophyletic (Mantiri et al. 2001). The name of this fungus was changed accordingly to Neonectria neomacrospora (Booth & Samuels) Mantiri & Samuels (Mantiri et al. 2001). Nectria fuckeliana var fuckeliana is a wound parasite of Pinaceae, including Abies, Larix, and Picea and should not be confused with Neonectria neomacrospora (Booth 1979a). Some symptoms of infection by N. fuckeliana var fuckeliana are similar to those observed for N. neomacrospora and include die back from the tip of infected branches, production of resinous swellings, cankering of the host bark, and in some cases death of the host tree (Lang 1981). 31 In addition, the size of perithecia and ascospores of N. fuckeliana and N. neomacrospora overlap (Table 1.1). Nectria fuckeliana also occurs in the same geographical regions as Neonectria neomacrospora (Europe and North America). Cultures of Neonectria neomacrospora may be separated from Nectria fuckeliana based on the size of microconidia (Table 1.1); (Booth 1979b). Hosts also differ somewhat between the two species. Although both varieties parasitize Abies, only N. neomacrospora parasitizes Tsuga and Pinus (Funk 1981). Table 1.1 Morphology and host ranges of Nectria fuckeliana var fuckeliana and Neonectria neomacrospora (modified from Funk 1981; Booth 1979a, 1979b). Species Anamorph Perithecia (approx.) Ascospores Micro-conidia Hosts Nectria fuckeliana var fuckeliana Cylindrocarpon cylindroides var tenue 300-400 um 13-16X5-6 um 4-7 X 2.5-3 pirn Abies, Larix, Picea, other Neonectria neomacrospora Cylindrocarpon cylindroides 400-600 fan 16-22 X.5-7 um 8-12 X 4-6 um Tsuga, Pinus, Abies Neonectria neomacrospora has been reported as a canker pathogen on Pinaceae and balsam fir {Abies balsamea L . Mill . ) ; (Booth 1979b). In other studies, N. neomacrospora has been isolated from swellings caused by several species of dwarf mistletoe and rust galls. In California and Washington, N. neomacrospora has been observed on Arceuthobium occidentale swellings on Bishop pine (Pinus muricata Dougl. ex D. Don) and A. abietinum swellings on white fir (Byler and Cobb 1972). In B C , N. neomacrospora has been observed on A. tsugense swellings on western hemlock (Funk, Smith, and Baranyay 1973). Neonectria neomacrospora is also known to infect and kil l galls caused by Peridermium harknessii in the western United States (Byler and Cobb 1972). In the cases of dwarf mistletoe and rust gall infections, N. neomacrospora appeared to be specific to the diseased swelling region (Funk and Baranyay 1973; Byler, Cobb and Parameter 1972a, 1972b). Although some studies have categorized N. neomacrospora as a conifer, pathogen (Ouellette and Bard 1966; Ouellette 1972; Booth 1979b), others have shown N. 32 neomacrospora to be specific to diseased conifer swellings and only mildly pathogenic to healthy conifer host tissue (Byler and Cobb, 1972, Funk, Smith, and Baranyay 1973). One possible explanation for the differences in pathogenicity observed for N. neomacrospora is the occurrence of different races (Ouellette 1972). Ouellette (1972) suggested that N. neomacrospora strains from Quebec were much more virulent on Abies than other isolates from B C and Europe. Alternatively, due to the complex taxonomy of this species, misidentification of the fungus may have occurred. The N. neomacrospora isolates used in the phylogenic study conducted by Mantiri et al. (2001) were isolated from Douglas fir rather than dwarf mistletoe swellings. However, the co-author Dr. Gary Samuels later confirmed the identity of several fungal i isolates taken from A. tsugense swellings on western hemlock to be Neonectria neomacrospora (USDA-ARS Systemic Botany and Mycology Lab, Beltsville, USA; pers. comm.). 1.5.2 Symptoms and signs ofN. neomacrospora Typical symptoms and signs associated with N. neomacrospora are shown in / Figure 1.1. Infection of dwarf mistletoe by N. neomacrospora is often associated with resinosis, sunken cankers, and, in some cases, girdling and death of the infected branch (Funk, Smith and Baranyay 1973; Smith and Funk 1980). Similar symptoms were observed for N. neomacrospora infected rust galls caused by Peridermium harknessii (Byler, Cobb, and Parameter 1972a, 1972b). Other symptoms of this fungus include dead or significantly fewer and smaller dwarf mistletoe shoots, shorter maximal distance between dwarf mistletoe shoots, and a slower rate of swelling and elongation of the dwarf mistletoe endophytic system (Smith and Funk 1980). Figure 1.1. Typical signs and symptoms of Neonectria neomacrospora. A . Perithecia from the sexual stage of TV", neomacrospora, B. Sporodochia from the asexual stage of N. neomacrosp'ora, C. Macroconidia from the asexual stage of N. neomacrospora, D. Mycelium of the asexual stage of N. neomacrospora grown on potato dextrose agar for 21 days, E. Resinosis and cankering of the A. tsugense swelling, F. Destructively sampled A. tsugense swelling with top layer of bark removed to show bark necrosis, G. Cross 33 section of an A. tsugense swelling showing bark necrosis, H. Diseased and dead A. tsugense shoots. Pictures A and E courtesy Dr. Simon Shamoun, Pacific Forestry Centre, Victoria, BC (2003). 34 In one study conducted by Funk, Smith and Baranyay (1973), infection of A tsugense swellings by N. neomacrospora resulted in an average reduction of 30 % in the number of A tsugense shoots when compared to the control. Forty-four percent of the A tsugense swellings colonized by N. neomacrospora were void of shoots (compared to 16% for the uncolonized A tsugense swellings). The growth rate of N. neomacrospora along the A tsugense swelling on potted hemlock was 39 mm/year, approximately the same rate of growth as the A tsugense swelling itself (Funk, Smith and Baranyay 1973). Signs associated with infection of A tsugense swellings by TV. neomacrospora include perithecia (perfect stage) and sporodochia (imperfect stage). Perithecia are small, reddish-brown, oval to conical ascospore producing structures that develop on the surface of dwarf mistletoe swellings. These structures measure 400-600/wn and can be found throughout the year in the field. The ascospores of are 1-septate and measure 16-22 X 5-7 /mi. Perithecia of N. neomacrospora can be produced in the laboratory from single spore crosses (Byler, 1970). Sporodochia are conidia-producing structures that resemble a small mass of mycelium on the dwarf mistletoe swellings. These white, irregular structures are most commonly found in the field on N. neomacrospora infected A tsugense swellings in the fall (Funk, Smith and Baranyay 1973). The imperfect stage of TV. neomacrospora has characteristic macro- and microconidia that can be cultured in vitro. Macroc'onidia are predominately straight or slightly curved, 3-7 septate spore structures and measure 33-85 X 4-7 jum. Microconidia are aseptate, oval to cylindrical spores formed from loose-branched conidiophores and measure from 8-12 X 4-6 ^m (Booth 1979b). Both life stages of TV", neomacrospora have been observed in the field. A study conducted by Byler, Cobb and Parameter (1972b) that made crosses of single spore isolates of TV. neomacrospora on P D A suggested that N. neomacrospora may be heterothallic. Other fungi may move into the A tsugense swelling once N. neomacrospora is established and magnify disease symptoms. For example, Nitschikia molnarii has been observed as a secondary invader of cankers caused by N. neomacrospora (Funk 1979). Many other fungi have been isolated from A tsugense swellings and shoots (Baranyay 1966; Hawksworth, Wicker, and Scharpf 1977; Kuijt 1963). 35 1.5.3 Biology of N. neomacrospora In numerous studies, N. neomacrospora has been isolated from the bark and wood of swollen conifer branches infected by dwarf mistletoe (Funk, Smith and Baranyay 1973; Byler and Cobb 1972). Field data suggested that N. neomacrospora is selective for the dwarf mistletoe swelling regardless of the host species. Neonectria neomacrospora has been found to be a virulent pathogen on dwarf mistletoe infected shore pine, true fir and western hemlock (Byler and Cobb 1972; Filip, Hadfield, and Schmidt 1979; Funk, Smith, and Baranyay 1973). Research by Byler, Cobb, and Parameter (1972a, 1972b) showed that N. neomacrospora could also infect pine galls caused by the rust fungus Peridermium harknessii. Data suggested that N. neomacrospora could eliminate P. harknessii on several species of pine by killing the gall. Byler, Cobb and Parameter(1972b) found that 90 % of the pine gall rust galls treated with N. neomacrospora were dead within 70 weeks after inoculation. Non-galled tissues remained resistant to invasion by N. neomacrospora. The ability of ./V. neomacrospora to infect dwarf mistletoe swellings and pine galls suggests that the fungus colonizes fast growing, nutritious, diseased coniferous tissue (Byler, Cobb and Parameter 1972b). Healthy coniferous tissue does not appear to be susceptible to infection. In addition to providing a nutritious environment in the swelling, dwarf mistletoe infection modifies bark moisture, which may also select for N. neomacrospora infection (Baranyay 1964). It is not known which spore type of TV. neomacrospora is the primary source of natural infection in the field, or i f wounding of the host is required for TV. neomacrospora to incite infection. Funk, Smith and Baranyay (1973) were able to establish infection on unwounded A. tsugense swellings using mycelium inoculum. However, a field trial conducted by Smith and Funk (1980) was unable to establish significant infection by applying a conidia and mycelium mixture onto unwounded A. tsugense swellings. The location of N. neomacrospora infection in P. harknessii rust galls also suggests that wounding may facilitate N. neomacrospora infection (Byler, Cobb and Parameter 1972b). 36 The location of N. neomacrospora within the dwarf mistletoe swelling is also poorly understood. The fungus has been isolated both from the bark and the wood of the dwarf mistletoe swelling, but it is not clear i f N. neomacrospora inyades the dwarf mistletoe endophytic system. Histological study of N. neomacrospora inoculated callus and germinating A. tsugense seeds showed that N. neomacrospora is able to infect A. J tsugense tissue both inter- and intra-cellularly (Deeks, Shamoun and Punja 2002). Cell wall degradation occurred as early as 0.5 days post mycelial contact, and infection reduced the growth rate of A. tsugense germinated seeds. Histological examination of rust galls inoculated with TV", neomacrospora found both the rust and pine components of the gall killed (Byler, Cobb and Parameter 1972b). The fungal hyphae of N. neomacrospora tended to be localized in and around living parenchymatous cells of the gall, such as phloem, parenchyma, epithelial cells of resin ducts, and wood ray parenchyma. 1.5.4 A potential biological control agent of Arceuthobium tsugense. Neonectria neomacrospora is a promising biological control agent for A. tsugense because it meets many of the criteria of an ideal biological control agent (Wicker and Shaw 1968). The natural distribution of N neomacrospora corresponds to the distribution of the target parasite A. tsugense (Funk, Smith and Baranyay 1973; Byler, Cobb, and Parameter 1972a, 1972b), and the fungus has sufficient ecological amplitude to assure persistence within its host range. In addition, N. neomacrospora produces both asexual and sexual spores in abundance and has proven its ability to establish and spread naturally in the field (Byler, Cobb and Parameter 1972a). Neonectria neomacrospora is able to diminish dwarf mistletoe spread by reducing the number of healthy shoots and in some cases killing the dwarf mistletoe infection. The mechanism ofN. neomacrospora is unique in that it attacks the dwarf mistletoe swelling and acts perennially. -The unknowns of N. neomacrospora, particularly the requirement of wounding for infection to occur, must be determined before this fungus is pursued further as a biological control agent for A tsugense. 37 1.6.0 Research objectives This study is part of a larger program to develop an inundative biological control agent for Arceuthobium tsugense using Neonectria neomacrospora. The research objectives are: 1. To select a lead isolate of TV", neomacrospora based on growth characteristics. These include the mycelial growth over a range of temperatures, conidial germination over a range of temperatures, and conidial production on a number of solid media. 2. To determine the requirement of wounding for N. neomacrospora infection. 3. To measure the impact of N. neomacrospora infection on A. tsugense shoot and swelling health. Chapters 2 and 3 describe the experimental methods used to achieve these objectives and end with an assessment of the probable efficacy ofN. neomacrospora as an inundative biological control agent. 38 Chapter 2 Selection of an isolate of Neonectria neomacrospora 2.1.0 Introduction Prior to commencing a field trial, it is advisable to conduct inoculation studies under controlled conditions. It is of particular importance to screen for the most virulent strain of the species under examination when developing a potential biological control agent. This idealistic approach becomes difficult, however, when working with hyperparasitic fungi. Two attempts to develop a system that could support A. tsugense infections under controlled conditions were unsuccessful (data not shown). Potted hemlock seedlings inoculated with A. tsugense seeds under greenhouse conditions did not establish mature A. tsugense infections in time for this study. A n alternative approach, which involved placing freshly cut hemlock branches that were naturally infected by A. tsugense into nutrient saturated Styrofoam blocks, did not maintain healthy A. tsugense infections long enough for N. neomacrospora to develop symptoms and signs of disease. As a consequence of the difficulties in developing A. tsugense experimental units under controlled conditions, it was not possible to screen N. neomacrospora isolates for virulence. Instead, it was decided that selection of an isolate of N. neomacrospora for the large field trial would be based on growth characteristics. In addition, a small field trial was conducted to develop inoculation techniques and to monitor disease development. 2.2.0 Materials and methods Unless specified otherwise, identification of A7! neomacrospora was based on colony and spore morphology as described in Chapter 1, Section 1.5.2 and a mycolbgical resource monograph (Booth 1966b). Dr. Gary Samuels (pers. comm.) confirmed the identity of the N. neomacrospora isolate used in the small field trial. Statistical analysis was done using SigmaStat version 2.03 (SPSS Inc., Chicago, IL). A l l figures were drawn using SigmaPlot version 7.0 (SPSS Inc., Chicago, IL). ) 39 2.2.1 Collection of A7, neomacrospora isolates Diseased A. tsugense swellings having symptoms and signs associated with N. neomacrospora were collected in June and July of 2001 and January of 2002 at Home Lake, Holt Creek, Mesachi Lake, and Muir Creek on Vancouver Island, British Columbia (BC). Previously collected and identified isolates of N. neomacrospora from the Pacific Forestry Centre (PFC) Culture Collection were also re-activated. Needles and A. tsugense shoots were removed from the freshly collected samples in the laboratory. Sections of diseased (i.e. necrotic) bark were removed and surface sterilized for 2-minute washes in each of 95% ethanol, 10% bleach (sodium hypochlorite, 5%), and three washes of sterile, distilled water. Sterilized sections were placed on potato dextrose agar (PDA, Difco Laboratories, Detroit, MI) and incubated at room temperature (22°C) for 1-2 weeks. If fruiting bodies were present, perithecia were crushed in sterile, distilled water and streaked onto a P D A plate using a sterile loop. Sporodochia were scraped off the sample and placed onto P D A using sterile tweezers. Small sections of resulting fungal cultures were re-plated on P D A and incubated for another 2-3 weeks at room temperature. Fungal colonies were then microscopically examined and identified. Four freshly isolated N. neomacrospora colonies and two N. neomacrospora isolates from the PFC culture collection were selected for the growth trials. v 40 \ 2.2.2 Selection and screening of the TV. neomacrospora isolates Six isolates of TV. neomacrospora were selected for screening (Table 2.1). A l l isolates were collected on Vancouver Island, BC. The growth characteristics determined for each of the six isolates included the rate of mycelial growth at varying temperatures (linear growth), the percentage and rate of spore germination at varying temperatures, and the number of spores produced on different solid-based media (rice, millet, slow oats, and wheat bran). Table 2.1. The origin of six isolates of TV. neomacrospora selected for screening. Isolate PFC No. Site and Date of Collection (dd/mm/yy) GPS (site) Fresh Sample or Culture Collection? PFC 4564 Spider Lake, Vancouver Island, 12/01/02 49°21'00" lat. 124°38'00 long. Fresh Sample PFC 4562 Spider Lake, Vancouver Island, 12/01/02 49°21'00" lat. 124°38'00 long. Fresh Sample PFC 4563 Spider Lake, Vancouver Island, 12/01/02 49°21'00" lat. 124°38'00 long. Fresh Sample PFC 4561 Holt Creek, Vancouver Island, 19/07/01 49°21'00" lat. 124°38'00 long. Fresh Sample PFC 2559 Holt Creek, Vancouver Island, 27/04/98 51°20'00" lat. 117°00'00 long. Re-activated from Culture Collection PFC 2570 Bowser, Vancouver Island, 13/05/98 56°25'00" lat. 129°30'00 long. Re-activated from Culture Collection 2.2.3 Linear colony growth at varying temperature Linear colony growth measured colony growth rate and was defined as the mean increase in radius (mm) of the colony per day. The linear colony growth was determined for the six isolates of TV. neomacrospora at temperatures ranging from 4-30°C. The purpose of this growth trial was to establish which isolate and temperature facilitated the fastest colony growth. Mycelial P D A plugs of TV. neomacrospora measuring 8mm in diameter were placed in the centre of fresh PDA plates and incubating at 4, 10, 15, 20, 25 and 30°C such that there were three replicates (plates) for each isolate at each temperature. The radius of each colony was measured along two points at 1, 2 and 3-41 weeks after inoculation. The experiment was terminated when the hyphal margin of the fastest growing colony reached the edge of the P D A plate at 3 weeks. 2.2.4 Percentage and rate of spore germination at varying temperatures Neonectria neomacrospora produces spores both sexually (ascospores) and asexually (conidia). Since conidia can be mass-produced in culture while ascospores cannot be produced in culture, it was decided that conidia would be used for the inoculum source in the small and large field trials. Because of this, the percentage and rate of spore germination was only measured for conidia. The percentage and rate of conidia germination at temperatures ranging from 4-30°C was determined for the six isolates of N. neomacrospora. The purpose of this trial was to determine which isolate and temperature resulted in the highest percentage and the fastest rate of spore germination. Conidia were mass-produced by inoculating 20 g of sterile rice and 20 ml of sterile, distilled water with 5 PDA plugs of mycelium in 250 ml Erylmeyer flasks. These were grown at room temperature for 7-10 days with infrequent agitation. At the end of the incubation period, 100 ml of sterile, distilled water was added to each flask. The flasks were placed on a rotary shaker at 225 rpm for 30 minutes. The contents of each flask was strained through sterile cheesecloth into a sterile beaker and transferred to multiple 50 ml centrifuge tubes. The centrifuge tubes were then centrifuged at 2500 G for 10 minutes, after which the supernatant was decanted and conidia were resuspended in 5 ml of sterile distilled water. The contents of each centrifuge tube that originated from the same isolate were combined such that there was only one centrifuge tube per isolate. Conidia were stored at 4°C and screened for bacterial contamination by plating a sub-sample of the spore solution on PDA and growing it at 30°C. Spores were used within 2 days after isolation or discarded. Spore germination was monitored for all six isolates at 4, 10, 15, 20, 25 and 30°C Prior to commencing the trial, the percentage of germinated spores was estimated for each isolate. If the percentage of germinated spores exceeded 10%, the spore suspension was not used in the trial. For each isolate and each temperature, 500 pl of a spore suspension containing l x l 0 5 spores/ml was plated onto water agar plates. There was only 4 2 one replicate (plate) for each isolate at each temperature. At least 300 spores were counted at 2, 4, 6, 8,10, 12 and (when necessary) 24 hours after inoculation by placing the plate directly under the microscope. Counting stopped once the percentage of germinated spores exceeded 95% or the time limit of 24 hours was reached. Counting was done in two rounds, counting three isolates at a time with staggered temperatures so that timing could be fairly accurate. In situations where the germination rate needed to be slowed for counting, plates were stored at 4°C. Spores were considered germinated i f the length of the germ tube exceeded the diameter of the spore. The percentage of germinated spores was plotted against time for each isolate at each temperature using a 3 r d parameter sigmoid regression formula Y = A/[(l + Exp (X-Xo))/B], where A= maximum y-value, B=rate at which the maximum y-value was reached, Y= percentage germination at any given time (hours), X= time (hours), and Xo=time to reach half the maximum y-value. The r values for the 36 graphs varied from 0.7984 to 0.994 with a mean of 0.932 (Figure 2.1). For each isolate at each temperature, the number of hours required for 50% spore germination was estimated using the sigmoid regression formula. These data were then used as a measurement of the rate of spore germination. 1201 : , co CD _ O C L CO "O CD -4—» CD C CD C D 100 80 60 40 20 —i— 10 15 20 Time (hours) 25 30 Figure 2.1. A sample graph of the percentage of germinated spores plotted against time for isolate PFC 2546 at 20°C. In this case, r 2 = 0.95. 43 2.2.5 Mass production of inoculum on different solid-based media The spore production trial consisted of six isolates, four media types, and two replications. The purpose of this trial was to measure the number of spores produced by each isolate on different solid-based media. Although different ratios of media and water was expected to influence the number of spores produced by N. neomacrospora, the amount of water and media used in this trial was altered to best suit the media type. Twenty ml of sterile distilled water was added to 20 g each of sterile rice and millet, 30 ml of sterile distilled water to 20 g of sterile slow oats, and 15 ml of sterile distilled water to 10 g of wheat bran. The media and water were mixed in 500 ml Erlenmeyer flasks and autoclaved twice to reduce contamination. Five mycelial P D A plugs measuring 8 mm in diameter were added to each flask such that there were two replicates (flasks) of each isolate for each media type. Cultures were grown at room temperature for 12 days. Conidia were extracted using the wash technique described for the spore germination trial in section 2.2.4. Sterile, distilled water was added to the suspension such that the final total volume of the suspension was 100 ml. The number of spores in 10 ul of suspension was then estimated using a haemocytometer. 2.2.6 Inoculation of A. tsugense in a small field trial The purpose of the small field trial was to conduct a preliminary inoculation trial using three different inoculation techniques. The trial was designed to provide information on which inoculation techniques are easy to apply and permitted infection. The small field trial also served as an early warning system. Since the small field trial was situated in a slightly warmer climate and treatments were applied 1 month before the large field trial, patterns observed in the small field trial were expected to occur sooner than those observed in the large field trial. The small field trial was located at the Pacific Forestry Centre (PFC), Victoria, BC. It consisted of eight planted hemlock trees that were inoculated with A. tsugense 44 seeds 4 years prior to treatment. At the start of the trial (July, 2002), there were a total of 21 healthy, mature A. tsugense infections. Prior to treatment, notes were taken on the health of the A. tsugense infections and the hemlock host. The number of healthy A. tsugense shoots was not counted prior to treatment. The growth trials were incomplete at this time, so it was not possible to select an isolate ofN. neomacrospora based on its growth characteristics. Instead, isolate PFC 4561 was selected for this trial because the A. tsugense swelling from which it was isolated was severely diseased Two types of inoculum and three treatments were used. For the first type of inoculum, conidia were mass produced on sterile rice media for 12 days and collected using the wash method described for the spore germination trial in section 2.2.4. The spore concentration of the resulting solution was determined using a haemocytometer, and the desired concentration of spores was adjusted with sterile distilled water to a concentration of 1.0 X 10 7 conidia per ml. The spores were then stabilized into a granulated formulation called "Stabileze" (Quimby et al. 1999). The Stabileze formulation was produced as follows: five grams of Waterlock ™ B-204 water absorbent starch (Grain Processing Corp., Muscatine, IA, USA) was mixed with 5 ml of corn oil (Spectrum Naturals, Inc., Petaluma, C A , USA) under sterile conditions. The mixture was then heated on "high" in the microwave oven for 1 minute and allowed to cool to room temperature. Once cooled, 20 ml of conidial suspension was slowly blended into the mixture, followed by 20.0 g of confectioners' sucrose (G & W Branch, Western Sugar Co., Denvor, CO, USA) and 7.0 g of Hi-Sil 233 ® (PPG Industries Inc., Pittsburgh, PA, USA) hydrated silica. The mixture was spread onto foil covered pans in the flow hood and allowed to air dry for 48 hours. The dried product was not passed through a series of sieves, as described in the protocol. Once dry, the product was stored in a sealed container at 4°C. For control treatments, a control Stabileze formulation was made using sterile distilled water in place of the conidial suspension. On the day of treatment application, a small amount of the dried Stabileze formulation was added to approximately 250 ml of sterile distilled water. The resulting solution was hand shaken until the formulation was mixed thoroughly. The final spore concentration used in the small field trial was unknown. For each treatment, the Stabileze 45 solution was hand sprayed onto the A. tsugense swelling and shoots until run off. In the case of control treatments, the control Stabileze was used. The second type of inoculum used in this trial consisted of mycelial P D A plugs. Cultures of isolate PFC 4561 were grown on P D A at room temperature for 2 weeks. Once the colonies were mature, 8 mm mycelial plugs were removed from the edge of the colonies and applied immediately to the mistletoe infected bark. Treatments for the small field trial involved: 1) spraying unwounded A. tsugense swellings with Stabileze (treatment A) or control Stabileze (treatment D), 2) wounding the A. tsugense swelling with a cork borer and inoculating the wound with a mycelial P D A plug ofN. neomacrospora (treatment B) or a sterile P D A plug (treatment E), then wrapping the wound with parafilm (Pichiney Plastic Packaging Inc., WI, USA), and 3) cutting off some A. tsugense shoots 0.5 mm from the shoot base and spraying the A. tsugense infection with Stabileze (treatment C) or control Stabileze (treatment F); (Table 2.2). Treatment B was applied on July 31 s t, 2002, and treatments A and C were applied on August 6 t h, 2002. A Hobo data logger (Onset, Bourne, M A ) was placed in the field trial site on July 31 s t, 2002 to monitor temperature and humidity at 15-minute intervals for 3 weeks after inoculation (Appendices I and II). Table 2.2. Treatment description and number of replicates for each treatment used in the small field trial. Treatment Treatment Description Number of Replicates A Isolate PFC 4561 formulated in "Stabileze" sprayed on A. tsugense swellings until run off 4 B Mycelial P D A plug of isolate PFC 4561 inserted into wound in A. tsugense swelling and wrapped with paraffin 4 C Isolate PFC 4561 formulated in "Stabileze" sprayed on cut A. tsugense shoots until run off 4 D Control "Stabileze" sprayed on A. tsugense swellings until run off 3 E Sterile P D A plug inserted into wound in A. tsugense swelling and wrapped with paraffin 3 F Control "Stabileze" sprayed on cut A. tsugense shoots until run off 3 46 Assessment occurred at 1 month after treatment and continued on a monthly basis for 1 year. Assessment measured the number of healthy and diseased A. tsugense shoots. Descriptive data was also collected on resinosis, cracking, and the occurrence of perithecia and sporodochia. In addition, pictures of each A tsugense infection were taken at each assessment using a digital camera (Nikon Coolpix995, Chiyoda-Ku, Tokyo, Japan) and stored for future reference. Occasionally, samples from diseased A. tsugense shoots and sporodochia were collected during the trial and plated onto PDA. These were grown at room temperature for 1-3 weeks and resulting fungal colonies were examined. The identity of N. neomacrospora was confirmed based on colony and spore morphology. At the end of the small field trial, a sub-set of the A. tsugense swellings were collected and destructively examined. Re-isolation ofN. neomacrospora from the necrotic bark and wood of the A. tsugense swellings was attempted. No re-isolation from A. tsugense shoots was done at this time. Re-isolation involved cutting small sections from the A. tsugense swelling and surface sterilizing them for 2 minutes in each of 10% bleach, 95% ethanol, and three washes of sterile distilled water. The sterile sections were then plated onto PDA, grown at 20°C and monitored on a weekly basis for 3 weeks. Fungal colonies were sub-cultured and N. neomacrospora was confirmed based on morphological and spore characteristics. 47 2.3.0 Results 2.3.1 Selection and screening of the TV. neomacrospora isolates 2.3.1.1 Linear colony growth at varying temperatures At week 3, isolate PFC 2559 grew marginally faster at 10, 15 and 20°C. A l l isolates grew best at temperatures ranging from 15°C to 20°C and peaked at 20°C (Figure 2.2). The colony radius appeared to be most severely inhibited at 30°C (mean radius at 21 days= 7.2 mm). The greatest variability in linear colony growth amongst isolates occurred at 25°C. 2.5 0.0 -I 1 1 1 1 1 1 1 0 5 10 15 20 25 30 35 Temperature (C) Figure 2.2. Linear colony growth (mm/day) for six isolates of TV. neomacrospora at temperatures ranging from 4 to 30°C. Bars measure standard error of the mean. 48 When the colony radius at 20°C was plotted over time, the relationship for isolates PFC 4564, PFC 2559, and PFC 2570 was slightly curved (Figure 2.3). The rate of colony growth during the second week was marginally faster than during the third week. The remaining isolates showed more linear relationships and consistent rates of colony growth. 45 40 35 E & 30 _ TD O C CO ^ ° c o o O 20 15 10 PFC 4564 P F C 2559 PFC 2570 PFC 4562 P F C 4563 P F C 4561 1 week 3 weeks 2 weeks Time after inoculation Figure 2.3. Colony radius (mm) across time for six isolates of N. neomacrospora at 20°C. Bars measure standard error of the means. 4 9 2.3.1.2 Percentage and rate of spore germination at varying temperatures At 24 hours after plating, at least 78% of the spores for each isolate had germinated across a broad temperature range (Figure 2.4). Lower and higher temperatures did not appear to inhibit spore germination for the six isolates of TV. neomacrospora. Isolate PFC 4562 (96-97%) had the smallest range and the second highest percentage of germinated spores across all temperatures when compared to the other isolates. At 20°C, isolate PFC 2559 had the lowest percentage of germinated spores (82.5%o) when compared to the other isolates. After completion of the germination trial, spore germination was monitored at 35°C, but no significant increase in the percentage of germinated spores was observed 24 hours after inoculation and counting was terminated. 100 80 A _ 60 40 20 10 15 20 25 Tern peratu re (C ) 30 35 Figure 2.4. Percent conidial germination for six isolates of TV. neomacrospora 24 hours after inoculation at temperatures ranging from 4°C to 35°C. 50 The estimated time required to reach 50% germination was highly variable among isolates (Figure 2.5). Based on calculations, conidia from isolate PFC 2570 took less time to reach 50% germination than the other isolates at all temperatures except for 30°C. At 20°C, isolate PFC 2570 reached 50% germination in approximately 1.4 hours. The rate of spore germination for all isolates was slowest at 4°C and 10°C. The rate of spore germination was fastest at 30°C (mean = 3.0 hours for 50% germination). It is important to note that, although spores were capable of germinating at relatively low and high temperatures, the health of germinated spores and early mycelial development was poorer at temperature extremes. This was especially true for spores that germinated at 30°C or higher. 12 10 A 8 H w i_ 3 O £ 6 H 0 E 4 H 2 H _ 10 15 20 Temperature (C) P F C 4564 P F C 2559 P F C 2570 P F C 4562 P F C 4563 P F C 4561 25 30 Figure 2,5. Number of hours for 50% spore germination for six isolates of Neonectria neomacrospora at temperatures ranging from 4°C to 30°C. 51 2.3.1.3 Mass production of inoculum on different solid-based media Both macroconidia and microconidia were produced in culture. Isolate PFC 2559 produced the greatest number of spores (74 spores per fi\) when compared to the other isolates (Figure 2.6). Isolate PFC 4654 (11 spores per ul) produced the fewest number of spores. The number of spores produced by isolates PFC 2570, PFC 4562, PFC 4563, and PFC 4561 ranged from 16-34 spores per u\. 0 CL CD o CL V> CD -O E C C co CD 120 100 H 1 1 1 1 1 ' i • ' i — " ' 1 r PFC 4564 PFC 2559 PFC 2579 PFC 4562 P F C 4663 PFC 4561 Isolate Figure 2.6. Mean number of spores produced by six isolates of N. neomacrospora averaged across four media types (brown rice, millet, wheat bran and slow oats) at room temperature (22°C). Bars measure standard error of the mean. 52 When comparing the four media types, rice supported the greatest production of spores (51 spores per al); (Figure 2.7). This was followed by millet (40 spores per ul) and wheat bran (30 spores per ul) produced an intermediate numbers of spores. Slow oats produced the least number of spores (5 spores per ul). There was also a range in spore size and consistency across the media types, where different media tended to produce larger, more uniform spores (data not shown). wheat bran rice oats millet Media Figure 2.7. Mean number of spores produced on four media types averaged across six isolates of N. neomacrospora at room temperature (22°C). Bars measure standard error of the mean. Two-way repeated measures analysis of variance on ranks detected a significant difference in spore production among the different media types (P=0.005). Tukey's test found rice, wheat bran, and millet each had significantly more spores than slow oats (P=0.005, P=0.009, P=0.010, respectively). Statistical analysis did not detect a significant 53 difference between isolate types or a significant interaction between media and isolate types. 2.3.2 Inoculation of A. tsugense in the small field trial The temperature and relative humidity for the small field trial is shown in Appendices I and II. Treatment B and its respective control were applied in the late afternoon on July 31 s t, 2002, and the highest temperature was 24°C at 4:48 pm. The lowest relative humidity was 38% at 5: 03 pm. On the day of inoculation for treatments A and C and their respective controls (August 6 t h, 2002), the highest temperature was 23°C at 12:48 pm. The lowest relative humidity was 50% at 4:18 pm. Trends in the number of healthy A tsugense shoots were apparent (Figure 2.8). Since the number of healthy A tsugense shoots was not recorded before treatment, data from the first assessment at 1 month is also included in Figure 2.8. CO -«—« o o JC CO CD CO C CD D l =5 co CO CD CD E Z3 80 60 40 20 H 1 month 12 mon ths T _ A B C D Trea tments Figure 2.8. Mean number of healthy A. tsugense shoots observed at 1 and 12 months after treatment. Bars measure standard error of the mean. 54 The mean number of shoots for the unwounded, Stabileze inoculated treatment (A) and its respective control (D) did not vary at 1 and 12 months after treatment. The cut shoots, Stabileze inoculated treatment (C) and its respective control (F) appeared to recover at 12 months from treatment. However, the wound, mycelial plug inoculated treatment (B) had a large reduction in the number of healthy dwarf mistletoe shoots at 12 months when compared to its respective control (E). There were not enough replicates used in the small field trial to detect statistically significant differences in the number of healthy A. tsugense shoots between treatments. Confirmation that TV. neomacrospora had colonized A. tsugense swellings in the small field trial was based on the occurrence of sporodochia (Table 2.3). A n attempt to re-isolate TV. neomacrospora from a few of the A. tsugense swellings was also made. Based on these data, treatment B appeared to have the highest number of colonized A. tsugense swellings (100%) when compared the other treatments. One year after treatment, sporodochia occurred as far as 40 mm away from the site of inoculation on an A. tsugense swelling from treatment B. Dr. Gary Samuels confirmed the identity of two isolates of TV. neomacrospora isolated from dead A. tsugense shoots and basal cups (pers. comm.). Table 2.3. Sporodochia and successful re-isolation of TV. neomacrospora (N. n.) for each treatment from the small field trial. The proportion of A. tsugense swellings colonized by TV. neomacrospora was based on the occurrence of sporodochia only. Treatment Number of Sporodochia Re-isolation Proportion of A. tsugense Replications Observed of N . n. swellings colonized by N . n. A 4 2/4 2/4 50% B 4 4/4 4/4 100% C 4 1/4 0/4 25% D 3 1/3 0/0 33% E 3 2/3 0/1 67% F 3 1/3 0/0 33% 55 2.4.0 Discussion 2.4.1 Growth characteristics ofN. neomacrospora It was assumed that the rate of colony growth of N. neomacrospora on P D A approximates N. neomacrospora vigour in the field under similar growth conditions. The linear colony growth of the six N neomacrospora isolates was fairly uniform for temperatures at and below 20°C (Figure 2.2). A l l six isolates grew best at temperatures ranging from 15°C to 20°C and had fastest linear colony growth at 20°C. These findings corroborate previous studies in which 6 of 7 isolates of N. neomacrospora (as Nectria macrospora) were shown to grow best at 21°C (Ouellette, 1973). Isolate PFC 2559 grew marginally faster than the other isolates at 10°C, 15°C, and 20°C, suggesting that it was the best candidate for the field trial in terms of colonization. The percentage and rate of spore germination are important measurements because they provide an estimate of infection potential, particularly the minimum duration of the infection window. The percentage of spores germinated at 24 hours was fairly uniform for the six isolates at all temperatures (Figure 2.4). Lower and higher temperatures did not appear to reduce the total percentage of germinated spores at 24 hours. Based on calculations, the conidia from isolate PFC 2570 reached 50% germination faster than any other isolate at 4, 10, 15, 20 and 25°C (Figure 2.5). The production of spores on four different media types was measured to determine which isolate and solid-based media resulted in the greatest mass production of inoculum. Of the six isolates, isolate PFC 2559 produced the greatest number of spores averaged across the four solid-based media, but this was not significantly greater than the other isolates. Rice produced significantly more spores than the other solid-based media types, and slow oats produced the fewest spores. Statistical analysis failed to detect a significant interaction between media and isolates. Based on these findings, it was decided that growing isolate PFC 2559 on rice media would result in the greatest amount of conidia production. When taking all three growth trials into account, isolate PFC 2559 was selected for the large field trial. Isolate PFC 2559 had the fastest colony growth and produced the greatest number of spores. These characteristics wil l make it easy to mass-produce inoculum for this isolate. The rate and percentage of spore germination for isolate PFC 56 2559 was marginally smaller than that observed for isolate PFC 2570, but was assumed to be sufficient for use in the field trial. 2.4.2 Small field trial. At 12 months, the wounded, plug inoculated treatment (B) resulted in a greater reduction in the number of healthy A tsugense shoots than the other inoculated treatments (A and C); (Figure 2.8). This suggests that applying mycelial inoculum to wounded A. tsugense swellings has a greater impact on A. tsugense shoot vigour than applying inoculum to cut A. tsugense shoots or unwounded A. tsugense swellings. The number of healthy A tsugense shoots observed for the unwounded, inoculated treatment (A) did not vary significantly during the field trial. In the case of the cut shoots, inoculated treatment (C), the A. tsugense infections had mostly recovered from treatment at 12 months. Sporodochia first appeared October 15 t h, 2002 (10 weeks after treatment) and were most prominent at 3.5, 4.5 and 7.5 months after treatment. Of the inoculated treatments, the wounded, plug inoculated treatment (B) had the highest occurrence of sporodochia (100% of A. tsugense swellings); (Table 2.3). The occurrence of sporodochia for the remaining inoculated treatments and uninoculated controls ranged from 33-50%. The occurrence of sporodochia for the uninoculated treatments (D, E, and F) was unexpected. There were no known nearby natural N. neomacrospora inoculum sources at the field trial site. It is unclear i f A. tsugense infections in the uninoculated treatments were contamination by the inoculated treatments, either during the inoculation process or by secondary infection. Secondary infection via sporodochia is less likely, as sporodochia first appeared on control treatments only 2 weeks after the first appearance of sporodochia on inoculated treatments. The treated A. tsugense swellings were within close proximity of one another (0.5- 5.0 m), allowing for easy spread of inoculum. The maximal distance between sporodochia and the inoculation site observed in the small field trial at 1 year was 40mm. Funk, Smith and Baranyay (1973) observed a similar maximal distance between sporodochia and the inoculation site (39 mm/year). These data suggest that N. neomacrospora is successfully moving through the A. tsugense swelling. In addition, the rate of movement observed in both trials approximates the rate 57 of A. tsugense spread in the hemlock host (Funk, Smith and Baranyay 1973). These data indicate that N. neomacrospora may be able to compete with the growth rate of A. tsugense and thereby have a long-term impact on A. tsugense vigour. Disease symptoms were greatest for the wounded, plug inoculated treatment (B); (Figure 2.9). A l l four A tsugense swellings from the wounded, plug inoculated treatment (B) were completely girdled, killing the lateral A. tsugense swelling and host branch. The earliest experimental unit to become girdled occurred at 8 months after treatment. A l l the A. tsugense shoots beyond the point of girdling died quickly after girdling occurred. In many cases, "mummified" dark brown and black shoots remained attached to the swelling. Several attempts to re-isolate N. neomacrospora from these diseased and dead shoots were made. In two cases, the fungus was successfully re-isolated from the dead A. tsugense shoots and cups, suggesting that N. neomacrospora may be able to move into diseased and dead A. tsugense shoots as a saprophyte. Figure 2.9. Symptoms of A. tsugense infected with a mycelial plug of N. neomacrospora into a wound at 1 month (A) and 9 months (B). The arrow in (A) points to the site of inoculation. Note the diseased, dead A tsugense shoots and girdling of the swelling at 9 months (B). One of the two girdled A tsugense swellings from the unwounded, inoculated treatment (A) was situated 0.5 m directly below an A tsugense infection from a wounded, plug inoculated treatment (B). Sporodochia were observed on the wounded, plug inoculated infection (B) as early as January 2002, and numerous diseased and dead A tsugense shoots fell from the higher wounded, plug inoculated infection (B) onto the 58 lower unwounded, inoculated infection (A) during the field trial. Because of this, it is not known i f infection of the unwounded, inoculated infection (A) resulted from the treatment itself or by natural infection. In either event, the A. tsugense swelling was not mechanically wounded. Results from the small field trial suggest that the wounded, plug inoculated treatment (B) had the greatest impact on A. tsugense shoot health and number, and also caused the greatest amount of girdling of the A. tsugense swelling. In addition, sporodochia were most commonly produced by the wounded, plug inoculated treatment (B). The unwounded, inoculated treatment (A) and cut shoots, inoculated treatment (C) had relatively small impacts on A. tsugense shoot and swelling vigour when compared to the wounded, plug inoculated treatment (B). Observations from the small field trial also confirmed that,<once established, N. neomacrospora is able to reduce A. tsugense shoot and swelling vigour (Figure 2.8, 2.9). 59 Chapter 3 Inoculation of Arceuthobium tsugense in a large field trial 3.1.0 Introduction The infection process of TV. neomacrospora is not well understood. This species has two separate life stages and produces spores both sexually (ascospores) and asexually (conidia). It is not known which of these spore types is the primary, natural source of inoculum in the field. Conidia can be easily mass-produced in the laboratory, but a technique to mass-produce ascospores in the laboratory has not yet been developed. Both conidial and perithecial stages occur in the field. Smith and Funk (1980) used a mycelium and conidia mixture suspended in water to inoculate unwounded A. tsugense swellings. Only 8 of the 20 the treated and 3 of the 20 control A. tsugense swellings became colonized by TV. neomacrospora. It was not known i f the form of inoculum or the inoculation process itself was responsible for the low infection, and it was suggested that an improved inoculum technique must be developed. No published trials have used ascospores as an inoculum source. With these considerations in mind, it was decided that conidia would be incorporated into a Stabileze formulation for the large field trial. The Stabileze formulation was selected because it is simple, inexpensive, and can improve survival of fungal pathogens (Quimby et al. 1999). Other studies with different species of fungi have found that formulation can extend the period of time in which the fungus remains viable, reduce the dew period requirements for infection, and reduce the spray volume by 100-fold (Boyette 1994). Preliminary work with TV. neomacrospora conidia formulated in Stabileze showed that the fungus was able to germinate from dry granules on P D A within 48 hours (data not shown). This is the first experiment to use TV. neomacrospora spores formulated in Stabileze. There is conflicting evidence about whether wounding of the A. tsugense swelling or shoots is required for infection by TV. neomacrospora to occur. The low infection observed for a trial in which mycelium and conidia mixed in water were applied to unwounded A. tsugense infections suggests that TV. neomacrospora can not penetrate 60 unwounded A. tsugense swellings (Smith and Funk 1980). In contrast, another trial that applied mycelial inoculum over basal cups and young shoots showed successful inoculation of unwounded A tsugense swellings (Funk, Smith and Baranyay 1973). Further research is required to clarify these findings. Measurement of symptom development, with particular emphasis on the rate of A tsugense shoot reduction and the impact on A tsugense swelling health also deserve more study. Only one previous inoculation trial has measured the impact of TV. neomacrospora infection on the number of A tsugense shoots (Smith and Funk 1980), but data from this study was only taken at the end of the trial and was not measured on a monthly basis. Another trial that monitored natural TV. neomacrospora infection on dwarf mistletoe attributed a 30% reduction in the number of A tsugense shoots to infection (Funk, Smith and Baranyay 1973). These data, however, applied to different aged TV. neomacrospora infections and also failed to measure the impact of infection on a monthly basis. The purpose of the large field trial was to determine i f wounding of the A tsugense swelling or shoots is required for TV. neomacrospora infection. In addition, the large field trial measured the impact of TV. neomacrospora infection on A tsugense shoot production and bark necrosis of the A tsugense swelling. 3.2.0 Materials and methods In similar fashion to the small field trial, the identification of TV. neomacrospora was based on colony and spore morphology as described in Chapter 1, Section 1.5.2 and a mycological resource monograph (Booth 1966b). The mycologist Dr. Gary Samuels confirmed the identity of the TV. neomacrospora isolate used in the large field trial (pers. comm.). 3.2.1 Stabileze formulation Mass inoculum of TV. neomacrospora isolate PFC 2559 was produced on sterile rice media and harvested at 12 days using the wash method previously described for the spore germination trial in section 2.2.4. The spores were then incorporated into Stabileze using the technique described for the small field trial in section 2.2.6. For control 61 treatments, a control Stabileze formulation was made using sterile distilled water in place of the conidial suspension. On the day of the treatment application, 5.0 g of the dried formulation was added to 500 ml of sterile, distilled water. The solution was alternatively hand shaken and stirred on a rotary shaker for 30 minutes, or until the formulation was mixed thoroughly. A total of three batches of 500 ml of Stabileze solution and three batches of 500 ml of the control Stabileze solution were made. The final concentration of spores was approximately 54000 conidia per ml. For each treatment and control, the Stabileze solution was hand sprayed onto the A. tsugense swelling and shoots until run off. Prior to commencing the field trial, germination experiments using formulated conidia were done in the laboratory to ensure that N. neomacrospora spores remained viable in the Stabileze formulation. These preliminary trials showed successful germination and colony development of N. neomacrospora on P D A for Stabileze in both dry and water suspended forms. The exact spore number and percentage germination could not be determined because spores were hidden in the Stabileze aggregates. Other laboratory work failed to detect self-inhibited spore germination at the spore concentration used in the large field trial. 3.2.2 Site description The field trial site was established in a mixed conifer stand heavily infested by A. tsugense near Parksville, Vancouver Island, BC. The field site was distributed across approximately 2 hectares in a Coastal Douglas-fir Zone following the biogeoclimatic ecosystem classification system employed by the British Columbia Ministry of Forests (1992). Hemlock was the primary tree species and ranged in age from 5 to 40 years. Other common tree species included Arbutus (Arbutus menziesii Pursh.), white pine, Douglas fir and occasionally cedar. The site had been thinned irregularly and had patches of open areas. Salal (Gaultheria shallon Pursh.), some Rubus species, sword fern (Polystichum munitum (Kaulf.) Presl) and bracken fern (Pteridium aquilinum (L.) Kuhn) were common. The site was heavily infested with Arceuthobium tsugense, with almost all hemlock having at least one and as many as 15 or more A. tsugense infections. 62 Naturally occurring inoculum of Neonectria neomacrospora and the A. tsugense shoot pathogen Colletotrichum gloeosporioides were present at the site and colonizing A. tsugense infections. A Hobo data logger (Onset, Bourne, M A ) was installed at the site on the day of inoculation to record temperature and relative humidity (%). The Hobo data logger collected data at 30-minute intervals for 3 months following treatment. 3.2.3 Experimental design Selection of A tsugense experimental units was based on the following criteria: 1) the A. tsugense shoots and swelling were disease free, with only a single A. tsugense infection on the host branch, 2) the host hemlock branch was vigourous, and 3) the infection was accessible and easy to monitor. Both male and female A. tsugense infections of various sizes were used in this trial. A total of 237 A. tsugense infections were selected, tagged, and randomly assigned one of seven treatments (Table 3.1). There were 35 replicates of treatments A , B, C, J, K , and L, and 27 replicates of treatment N . Treatments were assigned randomly, regardless of the sex, size, or location of the A. tsugense infection. In some cases, there was more than one A. tsugense infection on the same tree; it was assumed that all host trees would act similarly in the field and these A. tsugense were also randomly assigned treatments. The number of young hemlock trees used for each treatment was as follows: treatment A (26), B (34), C (29), J (30), K (28), L (29), and N (27). There was some overlap where different treatments occurred on the same tree. 63 Table 3.1. Treatment description and number of replicates used in the field trial designed to assess N. neomacrospora as a biological control agent for A tsugense. Treatment Treatment Description Number of Replicates / A Isolate PFC 2559 formulated in "Stabileze" sprayed on A. tsugense infections to run off 35 B Isolate PFC 2559 formulated in "Stabileze" sprayed on cut shoots of A. tsugense to run off 35 C Isolate PFC 2559 formulated in "Stabileze" sprayed on wounded swelling of A. tsugense infections to run off 35 J Control "Stabileze" sprayed on A. tsugense infections to run off 35 K Control "Stabileze" sprayed on cut shoots of A. tsugense to run off 35 L Control "Stabileze" sprayed on wounded swelling of A. tsugense infections to run off 35 N The A. tsugense infection was not treated . 27 The purpose of treatments A and J was to determine i f applying formulated conidia of N. neomacrospora onto unwounded A. tsugense is sufficient to permit infection. Treatment involved spraying unwounded'A tsugense with Stabileze (treatment A) or control Stabileze (treatment J) until run off. The purpose of treatments B and K was to determine i f applying formulated conidia of N. neomacrospora onto cut A tsugense shoots facilitates infection. Data from these treatments wil l ascertain i f TV. neomacrospora is able to enter the A tsugense swelling by colonizing wounded A tsugense shoots. For these treatments, all A tsugense shoots were cut 0.5 mm from the base using scissors sterilized in 95% ethanol. The entire infection was then sprayed with Stabileze (treatment B) or control Stabileze (treatment K) until run off. Treatment A served as the unwounded, inoculated control and treatment K as the cut shoots, uninoculated control (^respective control). The purpose of treatments C and L was to determine i f wounding A tsugense swellings facilitates infection. Treatment involved applying numerous incisions that penetrated beyond the cambium along the A tsugense swelling using a razor blade sterilized in 95% ethanol. The incisions were made perpendicular to the host branch and did not girdle the swelling. The A tsugense swelling and shoots were then sprayed with 64 Stabileze (treatment C) or control Stabileze (treatment L) until run off. Treatment A served as the unwounded, inoculated control, and treatment L as the wounded, uninoculated control (respective control). Treatment N received no treatment and was compared to treatment J to determine i f spraying A. tsugense swellings and shoots with control Stabileze increased natural infection via naturally occurring TV. neomacrospora inoculum. Treatment N was also a measurement of the proportion of experimental units that appeared healthy during the selection process but were naturally infected by TV. neomacrospora either before or during the trial. The Stabileze and control Stabileze formulations used in the field trial were grown on P D A directly following treatment application. This was done to confirm that the spores used in the trial were viable. The plates were incubated at 20°C for 1-2 weeks. 3.2.4 Pre-treatment assessment Assessment of each experimental unit was done prior to treatment application on August 21 s t and 22 n d , 2002. Pre-treatment assessment measured number of healthy A. tsugense shoots and the number of A. tsugense buds (shoots measuring less than 5 mm in height). 3.2.5 Treatment application dates A l l treatments were applied on August 29 t h, 2002. The weather during the day of treatment was sunny and warm with temperatures ranging from 19 to 24°C during the day down to 7°C at night. The highest temperature was 24°C and occurred at 1:35 pm. Relative humidity ranged from 52 to 80% during the day with a minimal humidity of 53% at 3:35 pm. The maximal relative humidity that night occurred at 11:35 pm and measured 91%. The following week had moderate temperatures ranging from 7 to 23°C, and relative humidity ranging from 40-100% (Appendices III and IV). The trial was intended to continue for 1 year. 65 3.2.6 Treatment assessment Assessment after treatment measured the number of healthy A tsugense shoots and the number of A. tsugense buds. The presence of sporodochia, perithecia, resinosis and cracking of the bark were also noted. A l l treatments were assessed at 2 weeks and 1, 2, 3.5, 5, 6, and 9 months after treatment application. Also, along with each assessment, a photographic record of a few representative A. tsugense infections from each treatment was taken using a digital camera (Nikon Coolpix995, Chiyoda-Ku, Tokyo, Japan) and saved for future reference. 3.2.7 Sampling at 10 months Because of the high occurrence of sporodochia observed in the field and concerns about secondary infection of treatments, the large field trial was terminated early at 9 months after treatment. At 10 months, a total of 20 experimental units were randomly collected from each of treatments A , B, C , K , and L, 21 samples from treatment J, and 11 samples from treatment N (Table 3.2). These units were studied more thoroughly in the laboratory, and measurements of bark necrosis and the occurrence of sporodochia were taken. An attempt to re-isolate N. neomacrospora from A. tsugense swellings was also made. Table 3.2. The number of experimental units collected from the trial 10 months after treatment. Note that the total number of experimental units ("Total") has taken into account A tsugense infections that were lost during the trial (Table 3.4). Treatment A B C J K L N Collected 20 20 20 21 20 20 11 Remaining 13 15 13 14 14 15 16 Total 33 35 33 35 33 35 27 The extent of bark necrosis for each A tsugense swelling was based on the circumference of bark necrosis at the widest point divided by the total circumference of the swelling at that point. The circumference was used to measure the extent of bark necrosis rather than a longitudinal measurement because it gives an estimate of how close 66 the swelling is to being girdled. Girdling of the swellings kills the distal part of the A. tsugense infection and therefore has a significant impact on A. tsugense shoot and seed production. The occurrence of sporodochia on destructively sampled A. tsugense swellings at the end of the field trial was treated as non-cumulative data. Microscopic observations of the spore-producing structures confirmed the identity of sporodochia. Data measuring the occurrence of sporodochia during the large field trial were kept separate and were not statistically analyzed. Field data on sporodochia were not analyzed because sporodochia may have been missed or falsely identified during assessment. To re-isolate N. neomacrospora, 10 small sections of bark and wood were cut from the A. tsugense swelling and surface sterilized for 2 minutes in each of 10% bleach, 95% ethanol, and three washes of sterile distilled water. When cutting out sections from the A. tsugense swelling, preference was given to the margins of necrotic bark and nearby wood regions. In cases where necrotic bark did not occur, sections were taken at random from the bark and wood of the A. tsugense swelling. The sterile sections were plated onto PDA, incubated at 20°C and monitored on a weekly basis for 3 weeks. Fungal colonies were sub-cultured and N. neomacrospora was confirmed based on culture and spore characteristics. 3.2.8 Statistical analysis A l l statistical analyses were performed using SigmaStat version~2.03 (SPSS Inc., Chicago, IL) and graphs were drawn using SigmaPlot version 7.0 (SPSS Inc., Chicago, IL). In situations where multiple comparisons between treatments were made, the alpha level of the test was not altered and remained a=0.05. This is because comparisons of interest were determined prior to experimentation; these include treatments A vs. C, A vs. J, A vs. B , B vs C, B vs. K , and C vs. L. 3.2.8.1 Statistical analysis of the impact of wounding on infection The extent of bark necrosis, the occurrence of sporodochia and successful re-isolation of N. neomacrospora for each treatment were used to address the first objective. The extent of bark necrosis, re-isolation and the non-cumulative occurrence of 67 sporodochia at 10 months were based only on the destructively sampled A. tsugense swellings. Data measuring the extent of bark necrosis did not meet the assumptions of analysis of variance and transformation using arcsin-square-root (x) did not remedy this problem. Analysis of was therefore done using a two-way analysis of variance on ranks for treatments A , C, J, and L and for treatments A , B, J, and K . Treatments were grouped in this manner for practical purposes; there was more interest in determining where, i f any, interactions between treatment types occurred in addition to the significance of differences. By separating treatments, it was possible to narrow the location of interaction. The Tukey test was used for pairwise comparisons. Chi-Square analysis was used to test i f significant differences occurred between treatments for the non-cumulative occurrence of sporodochia and re-isolation of N. neomacrospora. The Fisher Exact test was used to compare differences between two treatments. Data collected during the field trial on the occurrence of sporodochia were not statistically analyzed. Comparison of successful re-isolation of N. neomacrospora versus the occurrence of sporodochia on the destructively sampled A. tsugense swellings was done using Chi-Square analysis. The relationship between different classes of bark necrosis and the percentage of A. tsugense swellings with confirmed N. neomacrospora infection was analyzed using linear regression. 3.2.8.2 Statistical analysis of the impact of infection on A. tsugense vigour The impact of N. neomacrospora infection on A. tsugense vigour took into account the number of healthy A. tsugense shoots, proportion of A tsugense infections with zero shoots, and the extent of bark necrosis and girdling of the A. tsugense swelling. A l l data measuring the impact of infection on A. tsugense vigour were based on destructively examined samples only. The A. tsugense infections from all treatments were re-grouped into two groups: those that were confirmed to be colonized by N. neomacrospora (=confirmed group), and those that were not confirmed to be colonized by A7, neomacrospora (=unconfirmed group); (Table 3.3). Confirmation that infection by 68 N. neomacrospora had occurred was based on successful re-isolation of the fungus and/or confirmation of sporodochia at the end of the large field trial. Table 3.3. Break down of the number of A. tsugense swellings with confirmed and unconfirmed TV. neomacrospora infection for each treatment. Treatments A B C J K L N No. confirmed 11 14 18 6 4 7 2 No. unconfirmed 9 6 2 15 16 - 13 9 The assumptions of a t-test were not met, and so a Mann-Whitney Rank Sum test was used to statistically analyze the number of healthy A tsugense shoots, and the extent of bark necrosis for the confirmed and unconfirmed groups. The proportion of A tsugense infections with zero shoots and the proportion of girdled A. tsugense swellings were statistically analyzed using Chi-square analysis followed by the Fisher exact test. In the case of the number of healthy A tsugense shoots and the proportion of A tsugense swelling with zero shoots, treatments B and K were removed from data analysis because the treatment process itself (i.e., cutting shoots) may have influenced the variable of interest. 3.3.0 Results 3.3.1 Stabileze formulation The field trial used the formulation technique described by Quimby et al. (1999). The Stabileze formulation technique was simple to perform and the product easy to store. In the dry form, the formulation consisted of an aggregate of small clumps, measuring a maximum of 5 mm in diameter. When the dry formulation was placed on PDA, TV. neomacrospora mycelium emerged from these clumps within 48 hours. The dry formulation could be suspended fairly easily in water at 1% solution and did not clog the hand held atomizer. The larger clumps took longer to suspend in water; these could have been removed i f the sieve method described by Quimby et al. (1999) had been used. 69 The P D A plates sprayed with the Stabileze solution used in the field trial confirmed that the conidia were viable. Within one week, numerous colonies ofN. neomacrospora had established for the Stabileze used on treatments A , B, and C. The control Stabileze (no spores) used for treatments J, K and L did not give rise to N. neomacrospora colonies. 3.3.2 Symptoms and signs ofN. neomacrospora in the field r Since N. neomacrospora infects the A tsugense swelling, it is difficult to monitor early stages of disease development. For the first 2 months following treatment application, there were no evident symptoms or signs that infection of A tsugense by N. neomacrospora had occurred. Sporodochia-like structures were observed on many experimental units later on in j the field trial. At no time throughout the trial were perithecia observed on any of the experimental units. A l l A. tsugense infections showed a decline in the total number of healthy A. tsugense shoots throughout the trial period. The percentage of dead and diseased A tsugense shoots remained below 25 % for all treatments until 6 months after inoculation. At this time, some treatments began to show an increase in broken, diseased and dead shoots. Diseased A tsugense shoots did not show localized disease symptoms, but rather the entire A tsugense shoot became "mummified" and turned dark brown as the tissues died. In some cases, the dead A tsugense shoots remained attached to the swellings (Figure 3.1, 3.2). It is speculated that bark necrosis and girdling of the A tsugense swelling contributed to reduced A tsugense shoot vigour. 70 'A" ." - • Figure 3.1. A treated A. tsugense swelling at 2 weeks (A) and 9 months (B) after treatment. Treatment involved wounding the A. tsugense swelling and applying inoculum (treatment C). Infection by N. neomacrospora was confirmed. -H H H H H f _ • B - ? Figure 3.2. A treated A. tsugense swelling at 2 weeks (A) and 9 months (B) after treatment. Treatment involved applying inoculum to unwounded A. tsugense swellings (treatment A). Infection by N. neomacrospora was not confirmed. 71 3.3.3 Lost experimental units 10 months following inoculation Across the large field trial period, 6 experimental units were lost (Table 3.4). Table 3.4. The number of A. tsugense (A. t.) infections prior to treatment and remaining at termination of the trial. Treatment Number of A . t. infections prior to treatment Number of lost A . t. infections Number of A . t. infections at termination of the trial A' 35 2 33 B 35 0 35 C 35 2 33 J 35 0 35 K 35 ' 2 . 33 L 35 0 35 N 27 0 27 Death of the A. tsugense swelling as a result of self-pruning of the host branch was rare. Selection of experimental units prior to treatment application discriminated against A. tsugense infections on lower hemlock branches that were prone to self-pruning. As a result, few (<1 %) of the A. tsugense swellings were lost to self-pruning. Instances where the host branch died beyond the A. tsugense swelling were much more common and were caused by girdling of the swelling. Since N. neomacrospora attacks the A. tsugense swelling, and because its mechanism to reduce A. tsugense vigour is poorly understood, A. tsugense infections killed by girdling or host branch death were not removed from data analysis. Dead or girdled A. tsugense swellings were screened the same way as non-girdled A. tsugense branches. The occurrence of girdling was not used as a measure of successful infection by N. neomacrospora. Squirrel damage was also minimal in the field trial (<1%) and therefore swellings affected by such damage also remained in data analysis. A windblown tree fell onto one 72 hemlock host containing several experimental units, and some of these units were lost. Tags from some other experimental units were also lost. 3.3.4 Importance of wounding for establishment ofN. neomacrospora During destructive sampling, there appeared to be a strong association between staining of the wood in the A. tsugense swelling and bark necrosis. Staining has also been observed in diseased Nectria galligena apple wood in apple orchards (McCracken et al., 2003). Bark necrosis and cankering of the A. tsugense swelling associated with N. neomacrospora infection may provide a habitat for stain and decay fungi (Baranyay, 1966) It is not known i f TV. neomacrospora or another fungal species is responsible for the staining observed in this trial. In addition to staining, numerous insects were found within the bark of destructively sampled A. tsugense swellings. Many of the A. tsugense swellings had insect cavities, and 9 of the 132 swellings still had larval insects in the bark. These insects caused extensive bark necrosis within the A. tsugense swelling. Byler, Cobb and Parameter (1972b) also noted the occurrence of insects in N. neomacrospora colonized rust galls and suggested that insects may serve as vectors for some fungi. No attempt was made to identify the species of insects found on and within the A. tsugense swellings. 73 3.3.4.1 Extent of bark necrosis The extent of bark necrosis for each destructively examined treatment is shown in Figure 3.3. Statistical analysis of treatments A , C, J, and L using a two-way analysis of variance on ranks found a significant interaction between wound treatments and inoculum treatments was also observed (P=0.015). Further analysis detected significant differences between treatments C and L and A and C (P<0.001, PO.001, respectively). Statistical analysis of treatments A , B, J and K failed to detect significant differences between wound treatments or a significant interaction between wound and inoculum treatments. There was, however, a significant difference between inoculum treatments (P=0.013). C J K Treatments Figure 3.3. Mean extent of bark necrosis for A tsugense swellings collected 10 months after treatment. Bars measure1 standard error of the mean. Eighty-five percent of the A. tsugense swellings from treatment C were completely girdled or had girdled side branches (data not shown). This is much higher than the proportion of girdled A. tsugense swellings observed for treatments A , B, and the respective control (55%, 55%, and 45%, respectively). 74 3.3.4.2 Proportion of A. tsugense with sporodochia The proportion of A. tsugense swellings bearing sporodochia (presence or absence) 10 months after treatment is shown in Figure 3.4. Lumped chi-square analysis of treatments A , B, C, J, K and L found significant differences between treatments (P = <0.001). Significantly more A. tsugense swellings from treatment C had sporodochia when compared to treatments A and L (P = 0.014, P<0.001 respectively). Treatments A and B did not significantly increase the occurrence of sporodochia when compared to their respective controls (treatments J and K) . The occurrence of sporodochia for treatment N were not significantly greater than that measured for treatment J. No sporodochia were observed for the control treatments J, K or L at 10 months after treatment. C J K T r e a t m e n t s Figure 3.4. Proportion of A. tsugense swellings collected 10 months after treatment bearing confirmed sporodochia. The occurrence of sporodochia observed during the large field trial period was looked at both as cumulative and non-cumulative data. Since the occurrence of sporodochia observed in the field was based on observational data and not confirmed, statistical analysis of the data was not done. 75 For the cumulative data, the occurrence of sporodochia was treated as additive from one assessment to the next, such that all units that bore sporodochia at any time were considered to have sporodochia. Treatments B and C had more sporodochia than their respective controls at 3.5, 6 and 9 months (Figure 3.5). Treatment A did not appear to increase the occurrence of sporodochia when compared to its respective control. CD CO C & J 1 1 <1 o g. C CO .2 sz p o 0.8 0.6 0.4 0.2 0.0 I 3.5 Months Post I 6 Months Post • 9 Months Post C J K Treatments N Figure 3.5. Cumulative proportion of A tsugense swellings bearing sporodochia 3.5, 6 and 9 months after treatment. 76 The non-cumulative approach looked at the occurrence of sporodochia at each assessment and did not consider data from previous assessments. No apparent trends were observed between treatments for non-cumulative occurrence of sporodochia (data not shown). However, the non-cumulative occurrence of sporodochia for all treatments combined increased over time (Figure 3.6). The greatest increase in the proportion of A. tsugense swellings with sporodochia occurred 2-4 months after treatment (between October and December). The non-cumulative occurrence of sporodochia was highest in May at the end of the large field trial period. 0.30 T 1 % 0.25 Months Figure 3.6. The non-cumulative proportion of A. tsugense swellings bearing sporodochia. A l l treatments are grouped together for this data.., 77 3.3.4.3 Re-isolation of A7, neomacrospora from A tsugense swellings Results for re-isolation of N. neomacrospora from A. tsugense swellings are shown in Figure 3.7. Chi-square analysis detected significant differences between treatments A, B, C, J, K and L (P = 0.003). Treatment A had much higher re-isolation of N. neomacrospora than expected based on the extent of bark necrosis and the occurrence of sporodochia, and was significantly greater than its respective control (P = 0.048). Treatments B and C also resulted in significantly higher re-isolation of N. neomacrospora than their respective controls (P=0.022, P = 0.025, respectively). Treatment C had the highest re-isolation of N. neomacrospora of all treatments. The difference between treatment J and treatment N was negligible. A B C J K Treatments Figure 3.7. Proportion of A tsugense swellings from which TV. neomacrospora was successfully re-isolated 10 months after treatment. 78 Comparison of the confirmed occurrence of sporodochia at the end of the field trial and successful re-isolation of N. neomacrospora from the destructively sampled A. tsugense swellings suggested that re-isolation may fail to detect as much as 30% of infected A tsugense swellings (Table 3.5). Sporodochia were absent from 78% of A. tsugense swellings from which N. neomacrospora was successfully re-isolated. Chi-square analysis detected significant differences between groups (P=0.020). Table 3.5 Number of destructively sampled A. tsugense swellings at 10 months with and without sporodochia and successful re-isolation of N. neomacrospora. Sporodochia Observed Sporodochia Not Observed Re-Isolation Successful 12 43 Re-Isolation Unsuccessful 5 72 \ 79 When the extent of bark necrosis was broken down into 4 classes of 25% each and plotted against the proportion of A. tsugense swellings with confirmed N. neomacrospora infection for each class, a strong linear relationship emerged (Figure 3.8). The best linear model fit was Y= -0.034 -0.189X, where Y= the proportion of the A. tsugense swellings with confirmed infection and X= the class of bark necrosis (r2=0.93). Confirmed N. neomacrospora infection was based on successful re-isolation of N. neomacrospora and/or confirmed sporodochia at the end of the field trial. Figure 3.8. Linear relationship of the extent of bark necrosis and the proportion of A. tsugense swellings from which N. neomacrospora was successfully re-isolated. Regression analysis of Y=Yo + aX found 0.93. 80 3.3.5 Impact ofN. neomacrospora on A. tsugense vigour 3.3.5.1 Number of healthy A. tsugense shoots for confirmed and unconfirmed groups Arceuthobium tsugense swellings were regrouped into confirmed and unconfirmed groups regardless of treatment. The confirmed group (N=60) consisted of the destructively sampled A. tsugense swellings from which N. neomacrospora was successfully re-isolated or which bore sporodochia at 10 months. The unconfirmed group (N=72) consisted of the remainder of destructively sampled A. tsugense swellings which failed to successfully re-isolate N. neomacrospora and did not bear sporodochia. In the case of shoot analysis, cut shoot treatments (B and K) were removed from analysis because the treatment process itself may have influenced the variable of interest (i.e. number of shoots). Analysis of the number of healthy A. tsugense shoots 9 months after treatment found significant differences between confirmed and unconfirmed groups (Figure 3.9); (P=0.025). The confirmed group had 1.63 fewer shoots than the unconfirmed group, showing a 38% decrease. unconfirmed confirmed Infection by N. neomacrospora Figure 3.9. The mean number of healthy A. tsugense shoots 9 months after treatment for unconfirmed and confirmed groups. Cut shoot treatments were removed from analysis. Bars measure standard error of the mean. 81 A general decline in the total number of A. tsugense shoots was observed for both confirmed and unconfirmed groups throughout the trial period (Figure 3.10). The largest decrease in the number of healthy A tsugense shoots occurred between pre-treatment assessment (0 months) and 0.5 months after treatment. 0 2 4 6 . 8 10 Months after treatment Figure 3.10. Mean number of healthy A. tsugense shoots per swelling for unconfirmed and confirmed N. neomacrospora infected groups. Bars measure standard error of the mean. 82 3.3.5.2 Proportion of A. tsugense infections with zero shoots for unconfirmed and confirmed groups The confirmed group had a significantly greater proportion of A. tsugense infections with no shoots than the unconfirmed group 10 months after treatment (Figure 3.11); (P=0.002). Approximately 57% of the A. tsugense infections from the confirmed group did not have shoots, and 33% of the A. tsugense infections from the unconfirmed group did not have shoots. unconfirmed confirmed Infection by N. neomacrospora Figure 3.11. Proportion of A. tsugense infections with no shoots 10 months after treatment for unconfirmed and confirmed groups. 3.3.5.3 Extent of bark necrosis and girdling of A. tsugense swellings for unconfirmed and confirmed groups At 10 months after treatment, the mean extent of bark necrosis for the confirmed group (73%) was nearly twice as great as that observed for the unconfirmed group (37%). Statistical analysis using a one-way analysis of variance on ranks detected significant differences between the two groups (PO.001). 83 Chi-square analysis of the proportion of girdled A tsugense infections detected a significant difference between the confirmed and unconfirmed groups (P<0.001); (Figure 3.12). The confirmed group had just over two times as many girdled A. tsugense swellings (68%) when compared to the unconfirmed group (33%). unconfirmed confirmed Infection by N. neomacrospora Figure 3.12. Proportion of girdled A. tsugense swellings 10 months after treatment for unconfirmed and confirmed groups. 84 3.4.0 Discussion Wound-related response to fungal attack is a complex process that involves necrosis (death) of host tissue at the site of pathogen attack (Woodward and Pearce 1988; Biggs, Merrill, and Doris 1984). Successful pathogens are able to initiate infection and overcome host defence and compartmentalization. Although there is much potential for using necrotizing fungi as a biological control strategy, only a few have been evaluated. Early work by Byler, Cobb and Parmeter (1972a; 1972b) found the necrotizing fungus Neonectria neomacrospora was able to infect and kil l galls caused by Peridermium harknessii on Bishop pine and Monterey pine. Further investigation found that this fungus was a virulent pathogen of dwarf mistletoe infected pine tissues, but only weakly pathogenic to healthy conifer tissue (Byler and Cobb 1972). Nectria ditissima Tul., a wound pathogen that attacks red alder (Alnus rubra Bong.), is another necrotizing fungus that has been investigated as a potential biological control agent (Dorworth 1995; Dorworth et al. 1996). Dorworth (1995) developed an inoculation strategy for this fungus (ALDERKILL™), and registered the formulated form of N. ditissima as PFC-MYCOCHARGE®. Inoculation trials showed that the inoculation strategy was able to produce 100% infection, and that once infection was established, cankers longer than 0.5 m developed in 30 months (Dorworth 1995). Although disease symptoms caused by N. ditissima infection on red alder are slow to develop, the PFC-M Y C O C H A R G E technique is still recommended for pre-commercial stand improvement. Establishment of an inoculation procedure for N. neomacrospora is necessary before it can be further assessed as a biological control agent for hemlock dwarf mistletoe. While there has been some previous inoculation work with N. neomacrospora on dwarf mistletoe (Byler and Cobb 1972; Funk, Smith and Baranyay 1973; Smith and Funk 1980), this is the first large field trial to use different wounding techniques. Results from this trial were broken down according to two main objectives: to determine whether wounding is required for infection by N. neomacrospora to occur and to measure the impact of TV. neomacrospora infection on A. tsugense shoot health and number and swelling health. 85 Because of concern for secondary infection of experimental units, the large field trial was terminated 9 months after treatment. Sporodochia were common at the field site and occurred on experimental units and other A. tsugense swellings that were not used in the trial. Perithecia were also observed on A. tsugense swellings that were not used in the trial. At 9 months, 50% or more of the A. tsugense swellings from treatments B, C, L and N had had sporodochia at one point in time (Figure 3.5). Neonectria neomacrospora was successfully re-isolated from 16-35% of the A. tsugense swellings from the uninoculated treatments (J, K and L), indicating that some background infection had occurred (Figure 3.7). Early termination of the field trial prevented the long-term study of infection by TV". neomacrospora on A. tsugense vigour. 3.4.1 Stabileze formulation Smith and Funk (1980) stressed the need for an improved inoculation technique to assess N. neomacrospora as a potential biological control agent for A. tsugense. Because of this, it was decided that the inoculum source for the field trial would consist of conidia from the anamorph stage of N. neomacrospora formulated in Stabileze. Formulation can improve spore viability once applied in the field and permits long-term storage of inoculum (Quimby et al. 1996; Boyette 1994). Results from the large field trial show that Stabileze formulated conidia of TV. neomacrospora permitted infection of the A. tsugense swelling. Application of formulated conidia onto A. tsugense swellings increased infection by 39-40 % based on re-isolation data (Figure 3.6). This level of infection is higher than the 25% increase in infection observed in the non-formulated inoculation trial conducted by Smith and Funk (1980). It is not clear, however, i f differences are due to formulation of the spores or other experimental factors. A trial that directly compares the efficacy of Stabileze formulated spores versus non-formulated spores is required before any definitive conclusions about Stabileze can be made. 3.4.2 Importance of wounding for establishment of TV. neomacrospora The extent of bark necrosis, occurrence of sporodochia, and successful re-isolation of TV. neomacrospora from destructively sampled A. tsugense swellings were 86 used to determine the requirement of wounding for TV. neomacrospora infection. Because bark necrosis is much more useful in determining TV. neomacrospora infection than the occurrence of sporodochia or re-isolation of TV. neomacrospora, it was given precedence over the other two measurements. Bark necrosis of the A. tsugense swelling is a symptom commonly associated with TV. neomacrospora infection that often leads to cankering and girdling of the swelling (Funk, Smith and Baranyay 1973; Smith and Funk 1980). It is advantageous to use bark necrosis to quantify TV. neomacrospora infection because it is easy to measure and is consistently correlated with TV. neomacrospora infection. Although there was some bark necrosis in almost all A. tsugense swellings regardless of treatment, a strong, linear relationship between different classes of bark necrosis and the proportion of A. tsugense swellings from which TV. neomacrospora was successfully re-isolated was observed (r2=0.93); (Figure 3.8). The occurrence of sporodochia and re-isolation of TV. neomacrospora from the A. tsugense swelling are useful in that they confirm TV. neomacrospora infection, but failure to detect sporodochia or to re-isolate TV. neomacrospora does not necessarily mean infection has not occurred. Comparison of those A. tsugense swellings from which sporodochia were observed to those from which TV. neomacrospora was successfully re-isolated verified that neither of these measurements were perfect indicators of infection (Table 3.5). Sporodochia were observed on only 21% of the A. tsugense swellings from which TV. neomacrospora was re-isolated, and TV. neomacrospora was re-isolated from only 70% of the A. tsugense swellings that had sporodochia. In future trials, it may be worthwhile to develop molecular markers for detection of TV. neomacrospora in the A. tsugense swelling. Recent work by Langrell (2002) developed a species-specific PCR technique for detection of Neonectria galligena from wood of young apple trees. A similar technique could be modified for detection of TV. neomacrospora in the bark and wood of A. tsugense swellings. For the wounded, inoculated treatment (C), the extent of bark necrosis was greater than the sum of wounding and inoculation alone (Figure 3.3). Statistical analysis detected a significant interaction between wounding and inoculation (P=0.015), suggesting that this treatment caused the greatest amount of infection. The extent of bark necrosis for the 87 unwounded, inoculated treatment (A) and the cut shoots, inoculated treatment (B) was only marginally higher than the controls. The non-cumulative occurrence of sporodochia at 10 months and re-isolation of N. neomacrospora concur with these findings (Figures 3.4, 3.7). The wounded, inoculated treatment (C) had significantly more sporodochia than the other treatments and had the highest success of re-isolation of N. neomacrospora. The occurrence of sporodochia and successful re-isolation of TV. neomacrospora for the cut shoots, inoculated treatment (B) were slightly higher but not significantly different from the; unwounded, inoculated treatment (A). Based on these findings, the wounded, inoculated treatment (C) had the greatest impact on N. neomacrospora infection. Although there was some indication that infection was occurring in the other inoculated treatments, the impact of these treatments on A tsugense swelling bark necrosis was marginal in comparison to the wounded, inoculated treatment (C). There was no strong evidence that N. neomacrospora is able to colonize the A. tsugense swelling via cut A. tsugense shoots. These findings do not concur with an inoculation study done by Funk, Smith and Baranyay (1973), where N. neomacrospora successfully established after applying mycelial inoculum to undisturbed A. tsugense basal cups or young shoots. However, inoculated A. tsugense swellings in the trial conducted by Funk, Smith and Baranyay (1973) were monitored for almost 2 years after treatment and some A. tsugense swellings did not show signs of infection until as late as 21 months. The disease development of N. neomacrospora is slow and can take a year or longer to establish symptoms (Funk and Baranyay 1973). The long-term impact of N. neomacrospora infection (i.e. 1 year or longer) was not measured in the large field trial because it was terminated early at 9 months. Although the extent of bark necrosis for the unwounded, inoculated treatment (A) was not significant, the treatment had 39% more N. neomacrospora re-isolation success than its respective control (P=0.048); (Figure 3.7). It is possible that development of symptoms and signs of disease (i.e. bark necrosis, sporodochia) in the unwounded, inoculated treatment (A) is slower than the wounded, inoculated treatment (C). The impact of the unwounded, inoculated treatment (A) on A. 88 f tsugense shoot health may therefore be larger in the long term when disease has had sufficient time to develop. The large field trial was terminated early because of the high occurrence of background N. neomacrospora inoculum. Ideally, the large field trial should have been run on a site free of natural N. neomacrospora background inoculum. Despite efforts, a site with high A. tsugense infection and little to no 7Y. neomacrospora infection could not be found. Several sites on southern Vancouver Island were assessed, and there appeared to be a strong correlation between high levels of A. tsugense infection and the presence of N. neomacrospora. It was necessary to choose a site with high A. tsugense infection in order to have a sufficient number of experimental unit replicates. 3.4.3 Effect of TV. neomacrospora infection on A. tsugense vigour For the second objective, experimental units were re-grouped into two groups (confirmed and unconfirmed) regardless of treatment. The confirmed group consisted of A. tsugense swellings from which N. neomacrospora had been successfully re-isolated, or which had sporodochia at the end of the trial. The unconfirmed group consisted of all other A. tsugense swellings and may have contained N. neomacrospora infected A. tsugense swellings that escaped detection. The impact of N. neomacrospora infection on A. tsugense vigour was determined by measuring the number of healthy A. tsugense shoots, the proportion of A tsugense swellings with zero shoots, the extent of bark necrosis, and the proportion of girdled A tsugense swellings for confirmed and unconfirmed groups. Measuring the number of healthy shoots, in particular, was important because it estimated how much N. neomacrospora infection reduced A tsugense reproduction. Since N. neomacrospora acts perennially, shoot reduction is expected to continue and negatively impact the ability of A tsugense to produce inoculum the year(s) following treatment. The impact of N. neomacrospora infection on the number of A tsugense berries on female plants could not be measured directly because there was an insufficient number of healthy, female A tsugense plants at the field site. In addition, since berries are produced annually, at least 1 complete year of data would be required before the impact of N. neomacrospora infection on the number of healthy A tsugense berries could 89 be determined. Arceuthobium tsugense shoots are produced throughout most of the year and remain attached to the A. tsugense infection. It is not known i f A. tsugense plants are pollen limited. The impact of N. neomacrospora infection on the extent of bark necrosis and the occurrence of girdling were measured because these factors are expected to impact the number and health of A. tsugense shoots. Although all A. tsugense swellings lost shoots over the trial period (Figure 2.19), the confirmed group lost significantly more A. tsugense shoots than the unconfirmed group (P=0.025); (Figure 3.10). In addition, more A. tsugense swellings from the confirmed group did not have A. tsugense shoots (Figure 2.20). Overall, infection of A tsugense by N. neomacrospora reduced the number of healthy shoots by 1.6 shoots, or 38%. These data were similar to those observed by Funk, Smith and Baranyay (1973) who attributed a 30% reduction in the number of A tsugense shoots to 7Y. neomacrospora infection. The confirmed group also had significantly more bark necrosis and girdling than the unconfirmed group (Figure 2.21). Although several A tsugense swellings were completely girdled at the end of the trial, bark necrosis in other A tsugense swellings was i localized to one region of the swelling and had minimal impact on A tsugense shoots in healthy parts of the swelling. Although parts of the A tsugense swelling were girdled or had a large extent of bark necrosis, the A tsugense swelling could still bear numerous healthy A tsugense shoots in unaffected areas. As bark necrosis develops in infected A tsugense swellings, the impact of N. neomacrospora on the number of healthy A tsugense shoots is expected to increase. 3.4.4 Measuring the impact of treatment A on A. tsugense shoot vigour At present, the unwounded, inoculated treatment (A) is the only feasible treatment that could be used in the field to control A tsugense. The necessity of wounding for the other two treatments (B and C) makes them impractical for large-scale control. Wounding vectors, frost damage, and emerging A tsugense buds and basal cups-may provide some natural wound entry points for N. neomacrospora infection (Byler, Cobb and Parmeter 1972b). Insect galleries and larvae were observed in some of the 90 destructively sampled A. tsugense swellings, but the identity of the insects wasMiot determined. Further research into manipulation of wound vectors to increase N. neomacrospora infection is required. The impact of applying formulated conidia to unwounded A. tsugense swellings on the number of healthy A. tsugense shoot can be estimated via two routes. The first was simply to look at the average number of healthy A. tsugense shoots for the unwounded, inoculated treatment (A; 3.5 shoots) and compare that with its respective control (J; 5.9 shoots). In this case, applying inoculum to unwounded A. tsugense swellings reduced the number of A. tsugense shoots by an average of 2.4 shoots, or 41%. This is significantly different (P=0.008). However, the number of A. tsugense shoots observed for the respective control (J) was unusually high when compared to other uninoculated treatments (K, L and N). If the unwounded, inoculated treatment (A) were compared to the untreated treatment (N), the difference in number of shoots would not be significant (in this case, a 6% increase). The second way to calculate the number of healthy shoots required two steps. The first step determined the proportion of A. tsugense swellings from the unwounded, inoculated treatment (A) and its respective control (J) that were infected by N. neomacrospora. In this case, N. neomacrospora was confirmed for 55% of the unwounded, inoculated treatment and 29% of the respective control (Table 3.3). Next, the impact of infection by N. neomacrospora on the number of healthy A. tsugense shoots was determined. A l l treatments were combined and re-grouped; the confirmed group had an average of 2.9 A. tsugense shoots and the unconfirmed group had an average of 4.6 A. tsugense shoots. By calculating the difference between the number of A. tsugense shoots expected for the unwounded, inoculated treatment (A) and the number of A. tsugense shoots expected for its respective control (J), the impact of the unwounded, inoculated treatment (A) on A. tsugense shoot vigour was estimated. Although this method can be considered "the long way around", it is more conservative in that infection by N. neomacrospora was confirmed. Also, this method does not depend on the unusually high number of shoots observed for the unwounded, uninoculated treatment (J). Accordingly, using the first method, the unwounded, inoculated treatment (A) reduced the number of healthy A tsugense shoots in the field by about 41%. This 91 measurement is much larger than that calculated by the second method, which suggested that the unwounded, inoculated treatment (A) reduced the number of healthy A tsugense shoots about 14%. The true impact of the unwounded, inoculated treatment (A) on the number of A. tsugense shoots at 9 months is most likely in between these two values. The long-term impact of N. neomacrospora infection on A. tsugense shoot and swelling vigour is expected to be larger than that observed in the large field trial. 3.5.0 Closing Remarks Neonectria neomacrospora is a unique biological control agent for A. tsugense because of its host specificity and its infection mechanism. This fungus attacks the swelling of A tsugense, and has the potential of exerting a long-term negative impact on A tsugense seed production. In some cases, N. neomacrospora may girdle and kil l the A tsugense swelling (Funk, Smith and Baranyay 1973). To study the feasibility of using N. neomacrospora as a biological control agent of A tsugense, a large field trial that involved inoculating A. tsugense swellings with Stabileze formulated conidia was conducted. Three treatments were assessed: inoculation of unwounded A tsugense swellings, cut A tsugense shoots, and wounded A tsugense swellings. The data from this study suggest that wounding the A tsugense swelling facilitates N. neomacrospora infection, resulting in the greatest amount of bark necrosis, sporodochia production, and successful re-isolation of N neomacrospora. Although the unwounded, inoculated and cut shoots, inoculated treatments had greater re-isolation of N. neomacrospora than their respective controls, the impact of these treatments on N. neomacrospora infection was marginal when compared to the wounded, inoculated treatment, The unwounded, inoculated and cut shoots, inoculated treatments did not significantly increase bark necrosis or the occurrence of sporodochia when compared to their respective controls. Unless a wounding mechanism is discovered, applying N. neomacrospora inoculum to unwounded A tsugense swellings is the only economically feasible approach to control A tsugense. Although the initial impact of the unwounded, inoculated treatment on N. neomacrospora infection appeared minor, it cannot be dismissed as a 92 potential biological control method. It is possible that disease development (i.e bark necrosis, sporodochia) is slower for the unwounded, inoculated treatment, and that the long-term impact on A. tsugense vigour will be greater than that observed in the large field trial. Results from the trial also suggest that, once established, N. neomacrospora can significantly reduce A. tsugense shoot number by 38% at 9 months. Based on the extent of bark necrosis and girdling of the A. tsugense swellings observed during destructive sampling, the impact of N. neomacrospora infection on A. tsugense shoot vigour was lower than anticipated. Again, the extent of bark necrosis and girdling of the A. tsugense swelling are expected to have a greater, long-term impact on A. tsugense shoot production and the A. tsugense life cycle. 3.6.0 Future Research In the future, it would be interesting to measure the long-term impact of N. neomacrospora infection on A. tsugense as well as other dwarf mistletoe species. The rate of bark necrosis spread in the A. tsugense swelling is of particular interest, as it wil l determine i f A. tsugense is able to outgrow N. neomacrospora infection. It would also be worthwhile to see i f N. neomacrospora isolates from different locations vary in levels of virulence and to confirm whether or not this fungus is able to infect other hosts such as Abies. Combining N. neomacrospora with another A. tsugense pathogen, such as Colletotrichum gloeosporoioid.es, is also worth investigating. The study of natural A. tsugense wound vectors, which may allow for increased infection of TV*, neomacrospora in the field, is also of interest. In addition, the utilization of molecular markers for early detection of A 7 neomacrospora infection is an important contribution to study the host-pathogen interaction. 93 Literature Cited Adams, D. H. , S. J. Frankel and J. M . Lichter. 1993. Considerations when using ethephon for suppressing dwarf and leafy mistletoe infestations in ornamental landscapes. Journal of Arboriculture. 19(6): 351-357. Agrios, G. N . 1977. Plant Pathology. 4th Ed. Academic Press, New York. Alexander, M . E. and F. G. Hawksworth. 1975. Wildland fires and dwarf mistletoes: A literature review of ecology and prescribing burning. General Technical Report R M -14. Fort Collins, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station. 12p. Alosi, C. M . and C. L. Calvin. 1984. The anatomy and morphology of the endophytic system of Arceuthobium spp. In: Hawksworth, F. G. and R. F. Scharpf, tech. coords. Biology of dwarf mistletoes: Proceedings of the symposium; 1984 August 8; Fort Collins, CO: General Technical Report RM-111. Fort Collins, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station: 40-52. Alosi, C. M . and C. L . Calvin. 1985. The ultrastructure of dwarf mistletoe (Arceuthobium spp.) sinker cells in the region of the host secondary vasculature. Canadian Journal of Botany. 63: 889-898. Anonymous, c. 1985. Forest Diseases and Management. Victoria: Information Services Branch, Ministry of Forests. Province of British Columbia. Canadian Cataloguing in Publications Data. 19p. Baranyay, J. A . 1964. Bark Moisture relationships in western hemlock branches infected by dwarf mistletoe. Canadian Journal of Botany. 42(10): 1313-1319. Baranyay, J. A . 1966. Fungi from dwarf mistletoe infections in western hemlock. Canadian Journal of Botany. 44: 597-604. Baranyay, J. A . 1968. Squirrel feeding on dwarf mistletoe infections. Canadian Department of Forestry Bi-Monthly Research Note. 24(5): 41-42. Baranyay, J. A . , F. G. Hawksworth, and R. B. Smith. 1971. Glossary of dwarf mistletoe terms. B C P-2-71. Victoria, B C : Canadian Forestry Service, Department of the Environment. Pacific Forest Research Centre. 42 p. Baranyay, J. A . and R. B. Smith. 1977. Dwarf mistletoes in British Columbia and recommendations for their control. Canada PFRC Information Report BC-X-72. Victoria, B C : Canadian Forestry Service, Pacific Forestry Research Centre. 18 p. 94 Bier, J. E. 1959. The relation of bark moisture to the development of canker disease caused by native, facultive parasites; III. Cephalosporium canker on western hemlock. Canadian Journal of Botany. 37: 1140-1142. Bloomberg, W. J. and R. B. Smith. 1982. Measurement and simulation of dwarf mistletoe infection of second-growth western hemlock on southern Vancouver Island. Canadian Journal of Forest Research. 12: 280-291. Bloomberg, W. J., R. B. Smith, and A., van der Wereld. 1980. A model of spread and intensification of dwarf mistletoe infection in young western hemlock stands. Canadian Journal of Forest Research. 10: 42-52. Booth, C. 1966. The genus Cylindrocarpon. Mycological Papers 104: 1-56. Booth, C. 1979a. Nectria fuckeliana. In CMI Description of Pathogenic Fungi and Bacteria (No. 624). Commonweath Mycological Institute. Great Britain, The Cambrian News (Abersytwyth) Ltd. Booth, C. 1979b. Nectria neomacrospora. In CMI Description of Pathogenic Fungi and Bacteria (No. 623). Commonweath Mycological Institute. Great Britain, The Cambrian News (Abersytwyth) Ltd. Booth C. and Samuels G. J. 1981. Nectria neomacrospora, nom. nov., a new name for Nectria macrospora (Wollenw.) Ouellette. Transactions of the British Mycological Society. 77 (3). 1981. 645. Buckland, D. C. and E. G. Marples.1952. Management of western hemlock infested with dwarf mistletoes. British Columbia Lumberman. 36(5): 50-51, 136,140. Byler, J. W. 1970. Non-rust fungi associated with galls caused by Peridermium harknessii. Berkeley, C A : University of California, Graduate Division. 78p. Ph.D. Thesis. Byler J. W. and F. W. Cobb. 1972. The occurrence and phytopathogenicity of Nectria fuckeliana on dwarf mistletoe in California. Canadian Journal of Botany. 50: 1162. Byler, J. W., F. W. Cobb, Jr., and J. R. Parameter, Jr. 1972a. Effects of secondary fungi on the epidemiology of western gall rust. Canadian Journal of Botany. 50: 1061-1066. Byler, J. W., F. W. Cobb, Jr., and J. R. Parameter, Jr. 1972b. Occurrence and significance of fungi inhabiting galls caused by Peridermium harknessii. Canadian Journal of Botany. 50: 1275-1282. Callan, B. 2001. Introduction to forest diseases. Forest Pest Leaflet 54. Natural Resources Canada, Victoria, B C : Canadian Forestry Service, Pacific Forestry Research Centre. 15 p. 95 Calvin, C. L. , F. G. Hawksworth, and D. M . Knutson. 1984. Phloem in Arceuthobium globosum (Viscaceae). Botanical Gazette. 145(4): 461-464. i Carpenter, L.R., E. E. Nelson, and J. L. Stewart. 1979. Development of dwarf mistletoe infections on western hemlock in Coastal Oregon. Forest Science. 25(2): 237-243. Cartwright, C , S. Shamoun, S. Askew, and L. Rietman. 2003. Provenance variation in susceptiblility of western hemlock (Tsuga heterophylla) to hemlock dwarf mistletoe (Arceuthobium tsugense). [Abstract]. Meeting of the Western Forest Genetics Association. 2003, July 28-31. Whistler, B C . 1 p. Castello, J. D., D. J. Leopold, and P. J: Smallidge. 1995. Pathogens, patterns, and processes in Forest Ecosystems. Bioscience. 45(1): 16-24. Chapman, B . and Xiao, G. 2000. Inoculation of stumps with Hypholoma fasciculare as a possible means to control armillaria root disease. Canadian Journal of Plant Pathology 10:297-310. Cohen, L . I. 1954. The anatomy of the endophytic system of the dwarf mistletoe, Arceuthobium campylopodum. American Journal of Botany. 41: 840-847. Deeks, S. J., S. F. Shamoun, and Z. K . Punja. 1998. In vitro culture of western hemlock dwarf mistletoe. In Sturrock, R., comp., Proceedings of the 45 t h western international forest disease work conference. 1997, September 15-19th. Prince George, B C . Victoria, B C : Natural Resources Canada, Canadian Forestry Service, Pacific Forestry Research Centre, p. 74. Deeks, S. J., F. S. Shamoun, and Z. K . Punja. 2001. In vitro germination and development of western hemlock dwarf mistletoe (Arceuthobium tsugense subsp. tsugense). Plant Cell, Tissue, and Organ Culture. 66: 97-105. Deeks, S. J., F. S. Shamoun, and Z. K . Punja. 2002. Histopathology of callus and germinating seeds of Arceuthobium tsugense subsepecies tsugenese infected by Cylindrocarpon cylindroides and Colletotrichum gloeosporioides. Journal of Plant Sciences 163(5) 765-773. Dobie, J. and A . A . Britneff. 1975. Lumber grades and volumes from lodgepole pine infected with dwarf mistletoe. Wood and Fiber. 7(2): 104-109. Dorworth, C. E. 1995. Biological control of red alder (Alnus rubra) with the fungus Nectria ditissima. Weed Technology 9:243-248. Dorworth, C. E., D. E. Macey. T. N . Sieber, and T. A . D. Woods. 1996. Biological control of red alder (Alnus rubra) with indigenous pathogenic Ascomycotina. Canadian Journal of Plant Pathology 18:315-492. 9 6 Ellis, D. E. 1939. A fungal disease of Arceuthobium. Phytopathology 29: 995-996. Ellis, D.E. 1946. Anthracnose of dwarf mistletoe caused by a new species of Septogloeum. Journal of the Elisha Mitchell Scientific Society. 62: 25-50. Evans, H . C. 1995. Fungi as biocontrol agents of weeds: a tropical perspective. Canadian Journal of Botany. 73 (Supp 1): S58-S64. Evans, H . C , J. Frohlich, and S. F. Shamoun. 2001. Biological Control of Weeds. In Pointing, S. B. and K . D. Hyde, eds. Bio-Exploitation of Filamentous Fungi. Fungal Diversity Press, Hong Kong. P. 349- 401. Filip, G. M . , J. S. Hadfield, and C. Schmidt. 1979. Branch mortality of the true fir in west central Oregon associated with dwarf mistletoe and canker fungi. Plant Disease Reporter. 63(3): 189-193. Frankel, S. and D. Adams 1989. Reduction of dwarf mistletoe with the plant growth regulator ethephon. Forest Pest Management Report 89-1. San Francisco, C A : U.S. Department of Agriculture, Forest Service, Pacific Southwest Region. 5 p. Funk, A . 1979. A new Nitschkia on western conifers and its probable microconidial state. Canadian Journal of Botany. 57: 2113-2115. Funk, A . 1981. Parasitic microfungi of western trees. Report BC-X-222. Victoria, B C : Canadian Forestry Service, Pacific Forest Research Centre. 190 p. Funk, A . and J. A . Baranyay, 1973. Additional fungi and a gall disease of dwarf mistletoe swellings in western hemlock. Canada Plant Disease Survey. 53(4): 182. Funk, A. , R. B . Smith and J. A . Baranyay. 1973. Canker of dwarf mistletoe swellings on western hemlock caused by Nectria fuckeliana var macrospora. Canadian Journal of Forest Research. 3: 71-74. Gi l l , L . S. and F. G. Hawksworth 1961. The mistletoes: a literature review. Technical Bulletin No. 1242. Washington, DC: United States Department of Agriculture. Forest Service, Rocky Mountain Forest and Range Experiment Station. 87 p. Hawksworth, F. G. 1960. Growth rate of dwarf mistletoe infections in relation to the crown class of the host. Research Notes RM-41. Fort Collins, CO: United States Department of Agriculture. Forest Service, Rocky Mountain Forest and Range . Experiment Station. 4 p. Hawksworth, F. G. 1965. Life tables for two species of dwarf mistletoe. I. Seed dispersal, interception, and movement. Forest Science 11(2): 142-151. 97 Hawksworth, F. G. 1975. Dwarf mistletoe and its role in lodgepole pine ecosystems. In Baumgartner, D .M. , ed.. Management of lodgepole pine ecosystems. Proceedings of a symposium; October 9-11, 1973; Pullman, W A : Washington State University, Cooperative Extension Service. 342-358. Hawksworth, F. G. 1977. The 6-class dwarf mistletoe rating system. General Technical Report RM-48. Fort Collins, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station. 7 p. Hawksworth, F. G. 1987. Taxonomy of hemlock dwarf mistletoe. In: Cooley, S.J., comp. Proceedings, 34 t h Annual Western International Forest Disease Work Conference; 1986 September 8-12; Juneau, A K . Portland, OR: United States Department of Agriculture, Forest Service, Forest Pest Management. 45-46. Hawksworth, F. G. and D. W. Johnson. 1989. Biology and management of dwarf mistletoe in lodgepole pine in the Rocky Mountains. General Technical Report R M -169. Fort Collins, CO: United States Department of Agriculture. Forest Service, Rocky Mountain Forest and Range Experiment Station. 38 p. Hawksworth, F. G., E. F Wicker, and R. F. Scharpf. 1977. Fungal parasites of dwarf mistletoes. GeneraFTechnical Report RM-36. Fort Collins, CO: United States Department of Agriculture. Forest Service, Rocky Mountain Forest and Range Experiment Station. 14p. Hawksworth F. G and D. Wiens 1972. Biology and classification of dwarf mistletoes (Arceuthobium). Agriculture Handbook No. 401. Washington, DC: United States Department of Agriculture, Forest Service, pgs 38-39. Hawksworth, F. G. and D. Wiens 1996. Dwarf mistletoes: Biology, pathology and systematics. Agricutlure Handbook No. 709. Washington, DC: United States Department of Agriculture, Forest Service. 410 p. Hawksworth, F.G., J.C. Williams-Cipriani, B.B. Eav, B .G. Geils, R.R. Johnson, M . A . Marsden, J.S. Beatty, G.D. Shubert and D.C.E. Robinson. 1995. Dwarf mistletoe impact modeling system: user's guide and reference manual. Report MAG-95-2. Fort Collins, CO. United States Department of Agriculture, Forest Service, Forest Pest Management, Methods Applications Group. 120 pp. Online at: w w w . f s . f e d . u s / foresthealth /technology/d_modVdm_ug.pdf Hinds, T. E. and F. G. Hawksworth. 1965. Seed dispersal velocity in four dwarf mistletoes. Science. 148: 517-519. Hinds, T. E., F. G. Hawksworth, and W. J. McGinnies. 1963. Seed discharge in Arceuthobium: a photographic study. Science. 140: 1236-1238. 9 8 \ Hunt, R. S., J. N . Owens, and R. B. Smith. 1996. Penetration of western hemlock, Tsuga heterophylla, by the dwarf mistletoe Arceuthobium tsugense, and development of the parasite cortical system. Canadian Journal of Plant Pathology, 18: 342-346. Jobidon, R. 1991. Some future directions of biologically based vegetation control in forestry research. Forest Chronicles. 67(5): 514-519. Johnson, D. W. and F. G. Hawksworth. 1985. Dwarf mistletoes: candidates for control through cultural managements In: Insect and disease conditions in the United States, 1979-1983. General Technical Report WO-46U.S. Washington, DC: United States Department of Agriculture, Forest Service. 48-55. Johnson, D. W., D. M . Hildebrand and F. G. Hawksworth. 1989. Effects of application and timing of ethephon treatments on abscission of ponderosa pine dwarf mistletoe. Forest Service Technical Report R2-44. Lakewood, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Region, Timber, Forest Pest and Cooperative Forestry Management. 9 p. Kuijt, J. 1960. Morphological aspects of parasitism in the dwarf mistletoe (Arceuthobium). Publ. Botany 30. Berkley, C A : University of California. 337-436. Kuijt, J. 1963. Distribution of dwarf mistletoes and their fungus hyperparasites in western Canada. National Museum of Canada Bulletin. 186: 139-148. Knutson, D. M . and A . S. Hutchins. 1979. Wallrothiella arceuthobii infection Arceuthobium douglasii: culture and field inoculation. Mycologia 71: 821-828. Kope, H . H. , S. F. Shamoun, and C. Oleskevich. 1997. First report of Colletotrichum gloeosporioides on Arceuthobium tsugense subs, tsugense in Canada. Plant Disease. 81(8): 1095. Kope, H . H. , and S. F. Shamoun. 2000. Mycroflora associates of western hemlock dwarf mistletoe plants and host swellings collected from southern Vancouver Island, British Columbia. Canadian Journal of Plant Disease. 80: 144-147. Korstian, C. F. and W. H . Long. 1922. The Yellow Pine Mistletoe: Effect on growth and suggestions for control. Bulletin 1112. Washington, DC: United States Department of Agriculture. 35p. Krokene, P., H . Solheim, and E. Christiansen. 2001. Induction of disease resistance in Norway spruce (Picea abies) by necrotizing fungi). Plant Pathology 50:230-233. Lang, V . K . 1981. Cylindrocarpon cylindroides var. tenue als pathogen an Abies concolor und einigen Tannenbastarden. European Journal of Forest Pathology. 11: 191-192. 9 9 Langrell, S. R. H . 2002. Molecular detection of Neonectria galligena (syn. Nectria galligena). Mycological Research 106(3): 280-292. Leonard, O. A . and R. J. Hull. 1965. Translocation relationships in and between mistletoes and their hosts. Hilgardia. 37(4): 115-153. Livingston, W. H . and M . L . Brenner 1983. Ethephon stimulates abscission of eastern dwarf mistletoe shoots on black spruce. Plant Disease. 67(8): 909-910. Makowski, R. M . D. and K . Mortensen. 1992. The first mycoherbicide in Canada: Colletortichum gloeosporioid.es f. sp. malvae for round-leaved mallow. In: Delfossse, E. S. and Scott, R. R., eds. Proceedings of the 1 s t international control congress. Melbourne, Australia. 298-300. . Mantiri, F. R., G. J. Samuels, J. E. Rahe, and B. M . Honda. 2001. Phylogenetic relationships in Neonectria species having Cylindrocarpon anamorphs inferred from mitochondiral ribosomal D N A sequences. Canadian Journal of Botany. 79: 334-340. Mark, W. R., F. G. Hawksworth, and N . Oshima. 1976. Resin disease: a new disease of lodgepole pine dwarf mistletoe. Canadian Journal of Forest Research. 6: 415-424. Marler, M , D. Pedersen, T. Mitchell-Olds, and R. M . Callaway. 1999. A polymerase chain reaction method for detecting dwarf mistletoe infection in Douglas-fir and western larch. Canadian Journal of Forest Research. 29: 1317-1321. Mathiasen, R. L. 1994. Natural infection of new hosts by hemlock dwarf mistletoe. Research Note RM-RN-530. Fort Collins, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station. 6 P-McCracken, A . R., A . Berne, D. J. Barbara, T. Locke, L . R. Cooke, K . Phelps, T. R. Swinburne, A. E. Brown, B. Ellerker, and S. R. H . Langrell. 2003. Relative significance of nursery infections and orchard inoculum in the development and spread of apple canker (Nectria galligena) in young orchards. Plant Pathology 52: 553-566. Mielke, J. L. 1959. Infection Experiments with Septogloeum gillii, a fungus parasitic on dwarf mistletoe. Journal of Forestry. 57: 925-926. Miller, J. R. and R. D. Tocher. 1975. Photosynthesis and respiration of Arceuthobium tsugense (Loranthaceae). American Journal of Botany. 62(7): 765-769. Muir, J. A . 1973. Cylindrocarpon gillii, a new combination for Septogloeum gillii on dwarf mistletoe. Canadian Journal of Botany. 51: 1997-19998. 100 Mushtaque, M . and G. M . Baloch. 1979. Possibilities of biological control of mistletoes Loranthus spp., using oligophagous insects from Pakistan. Entomophaga. 24(1): 73-81. Nowicki B. , L . E. DeWald, and L. P. Moser. 1999. Evidence for genetic resistance to dwarf-mistletoe [Abstract]. Westen Forest Genetics Association. 1999 Annual Meeting; 1999, July 27-29; Flagstaff, A Z : Northern Arizona Univeristy, Forestry. 1 p. Ouellette, G. B. 1972. Nectria macrospora (Wr.) Ouellette sp. nov. (=N. fuckeliana var macrospora): Strains, physiology and pathogenicity, and comparison with N. fuckeliana var fuckeliana. European Journal of Forest Pathology. 2: 172-181. Ouellette, G. B., and G. Bard. 1966. A perennial canker of balsm fir on Anticosta Island. Plant Disease Reporter. 50(10): 722-724. Parker, A . K . 1970. Growth of Wallrothiella arceuthobii on artificial media. Canadian Journal of Botany 48(5): 837-839. Pratt, J. 1999. PG Suspension for the control of formes root rot of pine. Forestry Commission Information Note 18. Edinburgh, U K : 4 p. Online at: ~ http://www.forestry.gov.uk/ website/PDF.nsf/pdf/fcinl 8.pdf/$FILE/fcinl 8.pdf. (accessed December 12th, 2003). Quick, C. R. 1964. Experimental Herbicidal Control of Dwarf mistletoe on some California conifers. Research Note PSW-47. Berkeley, C A : United States Department of Agriculture, Forest Service, Pacific Southwest Forest and Range Experiment Station. 9 p. Quimby, P.C. Jr, N . K . Zidack, C. D. Boyette, and W. E. Grey. 1999. A simple method for stabilizing and granulating fungi. Biocontrol Science and Technology 9: 5-8. Ramsfield, T. D. 2002. Investigations into a biological control strategy for lodgepole pine dwarf mistletoe. Vancouver, B C : University of British Columbia, Graduate Studies, Department of Forest Sciences. 185p. Ph.D. Thesis. Richardson, K . S. and B. J. van der Kamp. 1972. The rate advance and intensification of dwarf mistletoe on immature western hemlock. Canadian Journal of Forest Research. 2: 313-316. Ridings, W. H . 1986. Biological control of strangler vine in citrus, a researcher's view. Weed Science. 34(Supp 1): 31-32. Rishbeth, J. 1963. Stump protection against Fomes annosus. III. Inoculation with Peniophora gigantea. Annals of Applied Biology. 52: 63-77. 101 Rossman, A . Y . , G. J. Samrnuels, C. T. Rogerson and R. Lowen. 1999. Genera of Bionecticaceae, Hypocreaceae, and Nectricaceae (Hypocreales Ascomycetees). Studies in Mycology. 42. 231 p. Roth, L. F. 1959. Natural emplacement of dwarf mistletoe seed on ponderosa pine. Forest Science. 5: 365-369. Roth, L. F. 1974. Resistance of Ponderosa pine to dwarf mistletoe. Silvia Genetica. 23: 116-120. Roy, G., G. Laflamme, G. Bussieres and M . Dessureault. 2003. Field tests on biological control of Heterobasidion annosum by Phaeotheca dimorphospora in comparison with Phlebiopsis gigantean. Forest Pathology. 33(2): 127-140. Scharpf, R. F. 1969. Cytospora abietis associated with dwarf mistletoe on true firs in California. Phytopathology 59(11): 1657-1658. Scharpf, R. F. 1986. Dwarf mistletoe as a host for brown felt blight in California. Plant Disease. 70: 798-799. Scharpf, R. F. and H . H . Bynum. 1975. Cytospora cankers of true firs. Forest Pest Leaflet 146. Washington, DC: United States Department of Agriculture, Forest Service. 5p. Scharpf, R. F. and J. R. Parmeter. 1967. The biology and pathology of dwarf mistletoe Arceuthobium campylopodum f. abietinum parasitizing true firs (Abies spp.) in California. Technical Bulletin 1362, Washington, DC: United States Department of Agriculture, Forest Service. 42p. Shamoun, S. F. 1997. Managing dwarf mistletoe- a biological and genetic control approach. In: Information Forestry, December 1997. Natural Resources Canada, Victoria, B C : Canadian Forestry Service, Pacific Forestry Research Centre, p. 2. Shamoun, S. F. and L. DeWald. 2002. Management of Dwarf Mistletoes by Biological, Chemical, and Genetic Methods. In: Geils, B. W., J. C. Tovar, and B. Moody, tech. coords. Mistletoes of North American Conifers. General Technical Report RMS-GTR-98. Fort Collins, CO: United States Department of Agriculture, Forest Service, Rocky Mountain Research Station. 75-82. Shaw, D., E. A . Freeman, and R. L. Mathiasen. 2000. Evaluating the accuracy of ground-based hemlock dwarf mistletoe rating: a case study using the wind river canopy crane. Western Journal of Applied Forestry. 15(1): 8-14. Shaw C. G . , III, and E. M . Loopstra. 1991. Development of dwarf mistletoe infections on inoculated western hemlock trees in southeast Alaska. Northwest Science. 65: 48-102 Smith, R. B . 1969. Assessing dwarf mistletoe on western hemlock. Forest Science. 15(3): 277-285. Smith, R. B . 1971. Development of dwarf mistletoe (Arceuthobium) infections on western hemlock, shore pine, and western larch. Canadian Journal of Forest Research. 1:35-42. Smith, R. B . 1974. Infection and Development of dwarf mistletoes on plantation-grown trees in British Columbia. Report BC-X-97. Natural Resources Canada, Victoria, B C : Canadian Forestry Service, Pacific Forestry Research Centre, Department of Environment. 21 p-Smith, R. B. 1977. Overstory spread and intensification of hemlock dwarf mistletoe. Canandian Journal of Forest Research. 7: 632-640. Smith, R. B . and Funk, A . 1980. Assessing Nectria neomacrospora as a biological control agent for hemlock dwarf mistletoe. Canadian Department of Forestry. B i -monthly Progress Report. 36 (3): 11-12. Smith, R. B. and E. F. Wass. 1976. Field evaluation of ecological differentiation of dwarf mistletoe on shore pine and western hemlock. Canadian Journal of Forest Research. 6: 225-228. Smith, R. B. and E. F. Wass. 1979 Infection trials with three dwarf mistletoe species within and beyond their known ranges in British Columbia. Canadian Journal of Plant Pathology. 1: 47-57. Smith, R. B. , E. F. Wass, and M . D. Meagher. 1993. Evidence of resistance to hemlock dwarf mistletoe (Arceuthobium tsugense) in western hemlock (Tsuga heterophylla) clones. European Journal of Forest Pathology. 23: 163-170. Srivastava, L . M . and K . Esau. 1961. Relation of dwarf mistletoe (Arceuthobium) to the xylem tissue of conifers. I. Anatomy of parasite sinkers and their connections with host xylem. American Journal of Botany. 48: 159-167. Stewart, J. L . 1976. Dwarf mistletoe infection from residual western hemlock on cutover stands. Research Note PNW-278. Corvallis, OR: United States Department of Agriculture, Forest Service, Pacific Northwest Forest Range Experiment Station. 3 p. Tainter, F. H . 1971. The ultrastructure of Arceuthobium pusillum. Canadian Journal of Botany. 49: 1615-1622. Tainter, F. H . and D. W. French. 1971. The role of wound periderm in the resistance of larch and jack pine to dwarf mistletoe. Canadian Journal of Botany. 49: 501-504. 103 i Thoday, D. and E. T. Johnson. 1930. On Arceuthobium pussillum. Peck. I. The endophytic system. Annals of Botany. 44: 393-413. United States Department of Agriculture (USDA), Forest Service 1988. Forest Health through silviculture and integrated pest management-strategic plan. United States Government Printing Office, 1988-576-488. Washington, DC. United States Department of Agriculture. 26 p. ; Van der Kamp, B. J. 1986. Dynamics of dwarf mistletoe infected, immature western hemlock stands. In: Cooley, S. J., comp. Proceedings, 34 t h Annual Western International Forest Disease Work Conference; 1986, September 8-12; Juneau, A K . Portland, OR: United States Department of Agriculture, Forest Service, Forest Pest Management. 48-50. Van der Kamp, B. J. 1991. Pathogens as agents of diversity in forested landscapes. The Forest Chronicle 67(4): 353-354. Wass, E. F. 1976. Ecological study of shorepine stands infected with dwarf mistletoe on southeastern Vancouver Island. Report BC-X-142. Victoria, B C : Department of Environment, Pacific Forestry Research Centre. 33 p. Watson, A . K and R. E. Wall. 1995. Mycoherbicides: Their role in vegetation management in Canadian forests. In P. J. Charest and L. C. Duchesne, eds. Recent Progress in forest biotechnology in Canada. Information Report PI-X-120. Victoria, B C : Canandian Forerst Service. 74-82 Watson, A . K. , Gressel, J., Sharon, A . and Dinoor, A . 2000. Colletotrichum strains for weed control. In D. Prusky, S. Freeman, M . B. Dickman, eds., Colletotrichum host specificity, pathology and host-pathogen interaction. APS Press, St. Paul, Minnesota. 245-265. Wellwood, R. W. 1956. Some effects of dwarf mistletoe on western hemlock. Forestry Chronicles. 32: 282-296. Wicker, E. F. and C. G. Shaw. 1968. Fungal Parasites of dwarf mistletoes. Mycologia. 60: 372-383. 104 Appendix I -Temperature on inoculation day (July 31 s t and August 6 t h, 2002) and immediately following for the small field trial 07/31 2002 Date 08/14 08/16 S/N 353954 Figure appendix 1-1. Temperature on inoculation day and immediately following. Victoria, BC. Maximum temperature on inoculation day was 24°C on July 31 s t and 22.9°C on August 6 t h, 2002. 07/31 2002 Date Figure appendix 1-2. Temperature at inoculation site from July 31 s t to August 30th. Victoria, BC. 105 Appendix II- Relative humidity (RH) on inoculation day (July 31 s t and August 6 t h, 2002) and immediately following for the small field trial Figure appendix 4-1. Relative humidity on inoculation day and following. Victoria, British Columbia. Minimum relative humidity was 38.4% on July 31 s t and 50.2% on August 6 t h, 2002. 110-1 Date Figure appendix 4-2. Relative humidity at inoculum site from August 31 s t to October 27 t h, 2002. Near Parksville, British Columbia. 106 Appendix III- Temperature on inoculation day (August 31st) and immediately following for the large field trial 25 H 08/28 2002 09/03 Spider L a k e 09/09 S/N 353954 Date Figure appendix 3-1. Temperature on inoculation day and following. Near Parksville, British Columbia. Maximum temperature on inoculation day was 23.6°C at 1:35 PM. 08/28 2002 09/27 Spider L a k e Date 10/27 S/N 353954 Figure appendix 3-2. Temperature at inoculation site from August 31 s t to October 27 t h, 2002. Near Parksville, British Columbia. 107 Appendix IV- Relative humidity (RH) on inoculation day (August 31s t) and immediately following for the large field trial 08/28 2002 08/03 Spidef Lake 08/09 S/N 353954 Date Figure appendix 4-1. Relative humidity on inoculation day and following. Near Parksville, British Columbia. Minimum relative humidity on inoculation day was 52.8% at 3:35 P M . 08/28 2002 09/27 Spider Lake 10/17 10/27 S/N 353954 Date Figure appendix 4-2. Relative humidity at inoculum site from August 31 s t to October 27 t h, 2002. Near Parksville, British Columbia. 108 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0074989/manifest

Comment

Related Items